vol9 organic ligands

Transcript

1 C HERMODYNAMICS HEMICAL T OMPOUNDS AND C OMPLEXES OF OF C U, Np, Pu, Am, Tc, Se, Ni and Zr O ELECTED WITH RGANIC L IGANDS S Wolfgang Hummel (Chairman) Laboratory for Waste Management Paul Scherrer Institut Villigen (Switzerland) Giorgio Anderegg Linfeng Rao Glenn T. Seaborg Center Swiss Federal Institute of Technology Chemical Sciences Division (ETH) Lawrence Berkeley National Laboratory Zürich (Switzerland) Berkeley (California, USA) Osamu Tochiyama Ignasi Puigdomènech Institute of Multidisciplinary Research for Swedish Nuclear Fuel and Waste Management Co. (SKB) Advanced Materials (IMRAM) Stockholm (Sweden) Tohoku University Sendai (Japan) Edited by Federico J. M OMPEAN (Series Editor and Project Co-ordinator) LLEMASSÈNE (Volume Editor) and Jane P ERRONE Myriam I OECD Nuclear Energy Agency, Data Bank Issy-les-Moulineaux, France i

2

3 Preface This book is the ninth of the series “C hemical Thermodynamics” edited by the OECD Nuclear Energy Agency (NEA), and it is the fi rst review in this series focused on or- ganic ligands. This Organics Review was initiated by the Management Board of the NEA Thermochemical Database Project Phase II (NEA TDB II). Originally Wolfgang Hummel (PSI, Switzerland), Giorgio Andere gg (ETH Zürich, Switzerland), Ignasi Puigdomènech (SKB, Sweden), Osamu Tochiyama (Tohoku University, Sendai, Japan), ř emek Lubal (Masaryk University , Brno, Czech Republic) partici- Josef Havel and P pated in the Organics Review. In August 2003 time constraints and the pressure of other commitments forced Josef Havel and P ř emek Lubal to resign from the Review Team. In December 2003 Linfeng Rao (L BNL, Berkeley, USA) jo ined the Review Team. The first meeting of the Organics Review Group was held in August 1998 at the NEA Headquarters at Issy-les-Moulineaux (France) and six plenary meetings fol- lowed in October 1999 (Brno, Czech Republic), April 2000 (Sendai, Japan), November 2000 (PSI, Switzerland), September 2001 (Bregenz, Austria), October 2002 (PSI, Swit- zerland) and January 2004 (OECD, Paris, France). The Executive Group of the Man- agement Board provided scientific assistan ce and Hans Wanner, as the designated member of the Executive Group, participated in some of the Review Group meetings and maintained close contact with the chairman of th e Organics Review Team. At the NEA Data Bank the responsibility for the overall co-ordination of the Project was placed with Eric Östhols (from its initiation in 1998 to February 2000), with Stina Lundberg (from March 2000 to September 2000) and with Federico Mompean (since September 2000). Federico Mompean wa s in charge of the preparation of the successive drafts, updating the NEA thermodynamic database and editing the book to its present final form, with assistance from Myriam Illemassène and Jane Perrone. All the members of the Review Team contributed fully to the main text and the discussions, and the excellent internal communication of the team has to be noted here. However, the workload was distributed ac cording to the expertise of each member. Wolfgang Hummel was the principal author of the sections about oxalic acid, protonation and Na and K interactions of oxalate, while Josef Havel and P ř emek Lubal were initially involved in re viewing the compounds and complexes of oxalate. Ignasi Puigdomènech reviewed citric acid and et hylenediaminetetraacetic acid, as well as protonation and Na and K interactions of citrate and edta. Osamu Tochiyama was the metal citrate complexes, whereas Giorgio principal author of all sections concerning Anderegg carried out the review of edta complexes with the editorial help of Wolfgang v

4 PREFACE vi Hummel. After the resignation of Josef Havel and P ř emek Lubal their remaining tasks were reassigned: Linfeng Rao reviewed uranium oxalate compounds and complexes, amu Tochiyama Zr and Am oxalates and Wolfgang Hummel Ca and Mg oxalates, Os Ignasi Puigdomènech Ni, Tc, Se, Np and Pu oxalates. The chapter on iso-saccharinic acid and compounds and complexes of isa is a joint review of Linfeng Rao and Wolfgang Hummel. Villigen, Switzerland, March 2005 Wolfgang Hummel, Chairman

5 Acknowledgements d financial support from the For the preparation of this book, the authors have receive NEA TDB Phase II Project. Th e following organisations take part in the Project: ANSTO, Australia ONDRAF/NIRAS, Belgium RAWRA, Czech Republic POSIVA, Finland ANDRA, France IPSN (now IRSN), France FZK, Germany JNC, Japan ENRESA, Spain SKB, Sweden SKI, Sweden HSK, Switzerland NAGRA, Switzerland PSI, Switzerland BNFL, UK NIREX, UK DoE, USA The authors would like to express their gratitude to Richard M. Kettler (De- partment of Geology, University of Nebraska, Lincoln, Nebraska, USA), Ranko P. Bontchev (Sandia National Laboratories, Al buquerque, New Mexico , USA) and Marian Borkowski (Los Alamos National Laboratory, Los Alamos, NM, USA) for providing unpublished data, to Martin A. Glaus and Jan Tits (both from the Laboratory for Waste Management, Paul Scherrer Institut, Villigen, Switzerland) for providing preprints of Samoylov from the Nikolaev Institute of their most recent publications and to Pavel Inorganic Chemistry (RAS) for providing access to a monograph by N. M. Nikolaeva. vii

6 ACKNOWLEDGEMENTS viii The authors would like to acknowledge the contributions of Prof. Josef Havel and Dr. P ř emek Lubal (both from the Department of Analytical Chemistry, Masaryk University, Brno, Czech Republic) to the section of oxalate compounds and complexes, mainly for nickel and neptunium, during the early stages of the review process. At the NEA Data Bank, Pierre Nagel and Eric Lacroix have provided excellent software and advice, which have eased the ed itorial and database work. Cynthia Picot, Solange Quarmeau and Amanda Costa from NEA Publications have provided consider- able help in editing the present series. Their contributions and the support of many NEA staff members are highly appreciated. Jan Rosdahl (Inorganic Chemistry, Royal Institute of Technology, Stockholm, Sweden) helped through literature searches in preparing several figures showing mo- lecular structures of complexes. His contribution is grat efully recognized. Ingmar Grenthe (Inorganic Chemistry, Royal Institute of Technology, Stock- holm, Sweden) reviewed early versions of Chapters VII and VIII. His contributions are gratefully acknowledged. undergone a peer re The entire manuscript of this book has view by an inde- pendent international group of reviewers, according to the proced ures in the TDB-6 eer reviewers have seen and approved the Guideline, available from the NEA. The p modifications made by the authors in response to their comments. The peer review comment records may be obtained on reques ear Energy Agency. t from the OECD Nucl The peer reviewers were: Prof. Gregory R. Choppin, Florida State University, Tallahassee, Florida, USA Prof. Pier G. Daniele, Univer sity of Turin, Turin, Italy Prof. Alfredo Mederos, University of La Laguna, Tenerife, Spain Their contributions are gratefully acknowledged.

7 Note from the Chairman of the NEA TDB Project Phase II The need to make available a comprehensive, internationally recognised and quality- assured chemical thermodynamic database that meets the modelling requirements for the safety assessment of radioactive waste disposal systems prompted the Radioactive Waste Management Committee (RWMC) of th e OECD Nuclear Energy Agency (NEA) to launch in 1984 the Thermochemical Databa se Project (NEA TDB) and to foster its continuation as a semi-autonomous project known as NEA TDB Phase II in 1998. The RWMC assigned a high priority to the critical review of relevant chemical thermodynamic data of inorganic species and compounds of the actinides uranium, neptunium, plutonium and americium, as well as the fission product technetium. The first four books in this series on the chemic al thermodynamics of uranium, americium, neptunium and plutonium, and technetium originated from this initiative. B Project reflects the interest in many The organisation of Phase II of the TD OECD/NEA member countries for a timely comp ilation of the thermochemical data that would meet the specific requirements of their developing national waste disposal pro- grammes. The NEA TDB Phase II Review Teams, comprising internationally recognised experts in the field of chemical thermodynami cs, exercise their scie ntific judgement in an independent way during the preparation of the review reports. The work of these Review Teams has also been subjected to further independent peer review. Phase II of the TDB Project consisted of: (i) updating the existing, CODATA- compatible database for inorganic species and compounds of uranium, neptunium, plu- tonium, americium and technetium, (ii) extending it to include selected data on inor- ganic species and compounds of nickel, selenium and zirconium, (iii) and further adding data on organic complexes of citrate, oxalate, edta and iso-saccharinic acid (isa) with uranium, neptunium, plutonium, americium, technetium, nickel, selenium, zirconium and some other competing cations. The NEA TDB Phase II objectives were formulated by the 17 participating or- ganisations coming from the fields of radioactive waste management and nuclear regu- lation. The TDB Management Board is assisted for technical matters by an Executive In this second phase of the Project, the Group of experts in chemical thermodynamics. ix

8 P N TDB P ROJECT HASE II OTE FROM THE CHAIRMAN OF THE NEA x NEA acts as coordinator, ensuring the applica tion of the Project Guidelines and liaising with the Review Teams. The present volume is the fifth one published within the scope of NEA TDB Phase II. It contains a database for organic complexes of citrate, oxalate, edta and iso- saccharinic acid (isa) with uranium, ne ptunium, plutonium, americium, technetium, nickel, selenium, zirconium and some other competing cations. We trust that the efforts of the reviewers, the peer reviewers and the NEA Data Bank staff merit the same high recognition from the broader sc ientific community as receive d for previous volumes of this series. Mehdi Askarieh United Kingdom Nirex limited Chairman of TDB Project Phase II Management Board On behalf of the NEA TDB Project Ph ase II Participating Organisations: ANSTO, Australia ONDRAF/NIRAS, Belgium RAWRA, Czech Republic POSIVA, Finland ANDRA, France IPSN (now IRSN), France FZK, Germany JNC, Japan ENRESA, Spain SKB, Sweden SKI, Sweden HSK, Switzerland NAGRA, Switzerland PSI, Switzerland BNFL, UK Nirex, UK DoE, USA

9 Editor’s note This is the ninth volume of a series of expert reviews of the chemical thermodynamics of key chemical elements in nuclear technology and waste management. This volume is devoted to organic compounds and complexes of citrate, oxalate, edta and iso- saccharinic acid (isa) with uranium, ne ptunium, plutonium, americium, technetium, selenium, nickel, zirconium and some other competing cations. The tables contained in cted thermodynamic values within the NEA Chapters III and IV list the currently sele TDB Project. The database system developed at the NEA Data Bank, see Section II.6, assures consistency among all the selected and auxiliary data sets. The recommended thermodynamic data ar e the result of a critical assessment of published information. The values in the a uxiliary data set, see tables IV-1 and IV-2, have been adopted from CODATA key values or have been critically reviewed in this or earlier volumes of the series. xi

10 xii How to contact the NEA TDB Project Information on the NEA and the TDB Project, on-line access to sel ected data and com- puter programs, as well as many documents in electronic format are available at www.nea.fr. To contact the TDB project coordinator an d the authors of the review reports, send comments on the TDB reviews, or to request further information, please send e- mail to [email protected] If this is not possible, write to: TDB project coordinator OECD Nuclear Energy Agency, Data Bank Le Seine-St. Germain 12, boulevard des Îles F-92130 Issy-les-Moulineaux FRANCE The NEA Data Bank provides a number of services that may be useful to the reader of this book. • The recommended data can be obtained via internet directly from the NEA Data Bank. • The NEA Data Bank maintains a library of computer programs in various areas. This includes geochemical codes such as PHREEQE, EQ3/6, MINEQL, MINTEQ and PHRQPITZ, in which chemical thermodynamic data like those presente d in this book are required as the basic input data. These computer codes can be obtained on request from the NEA Data Bank.

11 Contents Preface v Acknowledgement vii Note from the chairman of the NEA TDB Project Phase II ix Editor’s note xi Part I Introductory material 1 I Introduction 3 I.1 Backgr ound ...3 I.2 Focus of the review...5 I.3 Review proce dure and results ...6 II Standards, conventions, and contents of the tables 9 II.1 Symbols, terminol ogy and nomenclature ...9 II.1.1 Abbreviations ...9 II.1.2 Symbols and terminology...11 II.1.3 Chemical formulae and nomenclature ...13 II.1.4 Phase designators...13 II.1.5 Processes ...15 II.1.6 Equilibrium constants ...15 II.1.6.1 Protonatio n of a lig and...16 II.1.6.2 Formation of metal ion complexes...17 II.1.6.3 Solubility constants ...18 II.1.6.4 Equilibria invol ving the addition of a gaseous ligand ...19 II.1.6.5 Redox equilibria...19 II.1.7 pH ...22 II.1.8 Order of formulae ...24 II.1.9 Reference codes...25 ersion factors ...25 II.2 Units and conv II.3 Standard and ref erence conditions ...28 II.3.1 Standard state ...28 II.3.2 Standard state pressure ...29 II.3.3 Reference temperature...32 II.4 Fundamental physical constants ...32 xiii

12 CONTENTS xiv estimates...33 II.5 Uncertainty The NEA-TDB system...33 II.6 II.7 Presentation of the selected data...35 Tables of selected data Part II 39 III Selected data for oxalate, citrate, edta and isa ... 41 IV Selected auxiliary data... 57 Part III Discussion of data selection 79 V Criteria for data evaluation and pa rticular problems encountered in the review procedure ... 81 V.1 ta evaluation...81 Criteria for da V.2 Particular problems of commonly used experimental methods ...83 V.2.1 Potentiometry ...84 V.2.2 Two-phase distribution (solvent extraction and ion exchange) ...87 V.2.3 Spectrophotometry ...91 V.3 Ionic medium effects on protonation constants for organic ligands ...92 V.3.1 Introduction ...92 V.3.2 Evaluation of protonation data excluding complexes with medium cations92 V.3.3 Complex formation with medium cations ...94 Complex formation with medium V.3.4 cations versus the SIT formalism ...98 V.3.5 Accesso ry data...99 on reaction enthalpies ...99 V.3.6 Ionic medium effects versus strong specific ion interaction...100 V.4 Weak complexes r oxalate compounds and complexes . 105 VI Discussion of data selection fo VI.1 Introduction...105 VI.2 Oxalic acid ...106 VI.2.1 H ox(cr) ...106 2 VI.2.2 H O(cr) ...110 ox·2H 2 2 ox(aq) ...112 VI.2.3 H 2 VI.3 Protonation constants of oxalate ...113 VI.3.1 Introduction ...113 VI.3.2 Analysis of K ...135 1 ...140 VI.3.3 Analysis of K 2 VI.3.4 Temperature effects ...143

13 CONTENTS xv constants for VI.3.5 Selected protonation oxalate ...149 VI.4 Alkali metal oxalate co mpounds and complexes ...152 VI.4.1 Sodium and potassi um oxalate compounds ...152 + + and K ...156 VI.4.2 Complexes with Na VI.5 Magnesium and calcium oxalat e compounds and complexes...159 VI.5.1 Magnesium and calci um oxalate compounds ...159 VI.5.2 Magnesium and calci um oxalate complexes ...177 VI.5.3 Selected values for Mg and Ca oxalate compounds and complexes...189 VI.6 Selenium oxalate comp ounds and complexes ...189 VI.6.1 Solid sele nium oxalates ...189 VI.6.2 Aqueous selenium oxalate complexes ...190 VI.7 Nickel oxalate comp ounds and complexes ...190 VI.7.1 Solid nick el oxalate s...190 VI.7.2 Aqueous nickel oxalate complexes...194 VI.7.2.1 Temperature effects ...199 VI.7.2.2 Ternary complexes...200 VI.8 mpounds and complexes...200 Technetium oxalate co Zirconium oxalate comp VI.9 ounds and complexes ...202 VI.9.1 Zirconium oxa late compounds ...202 VI.9.2 Zirconium oxalate complexes...204 VI.10 Uranium oxalate comp ounds and complexes...206 VI.10.1 Solid uranium oxalates ...206 VI.10.1.1 Solid uranium(IV) oxalates ...206 VI.10.1.2 Solid uranium(VI) oxalates ...212 VI.10.1.2.1 General comments...212 VI.10.1.2.2 UO O(cr)...214 ox·3H 2 2 VI.10.2 Aqueous uranium oxalate complexes ...225 VI.10.2.1 Aqueous uranium(III) oxalate complexes ...225 VI.10.2.2 Aqueous uranium(IV) oxalate complexes ...225 VI.10.2.3 Aqueous uranium(V) oxalate complexes ...228 VI.10.2.4 Aqueous uranium(VI) oxalate complexes ...229 VI.10.2.4.1 Binary U(VI) oxalate complexes...229 VI.10.2.4.2 Polynuclear U(VI) oxalate complexes ...241 VI.10.2.4.3 Ternary U(VI) oxalate complexes...242 VI.11 Neptunium oxalate comp ounds and complexes ...245 VI.11.1 Solid neptunium oxalates...245 VI.11.1.1 Solid neptunium(III) oxalates...245 VI.11.1.2 Solid neptunium(IV) oxalates...246 VI.11.1.3 Solid neptunium(V) oxalates ...248 VI.11.1.4 Solid neptunium(VI) oxalates...249

14 CONTENTS xvi VI.11.1.5 Solid neptunium(VII) oxalates ...249 VI.11.2 Aqueous neptunium oxalate complexes...249 VI.11.2.1 Neptunium(III) oxalate complexes ...249 VI.11.2.2 Neptunium(IV) oxa late complexes...250 VI.11.2.3 Neptunium(V) oxalate complexes ...254 VI.11.2.4 Neptunium(VI) oxalate complexes...261 VI.11.2.5 Neptunium(VII) oxalate complexes ...261 ounds and complexes ...261 VI.12 Plutonium oxalate comp VI.12.1 Solid pluton ium oxalates ...261 VI.12.1.1 Solid plutonium(III) oxalates ...262 VI.12.1.2 Solid plutonium(IV) oxalates ...268 VI.12.1.3 Solid plutonium(V) oxalates...273 VI.12.1.4 Solid plutonium(VI) oxalates ...274 VI.12.2 Aqueous plutonium oxalate complexes ...276 VI.12.2.1 Plutonium(III) oxalate complexes ...276 VI.12.2.2 Plutonium(IV) oxalate complexes ...277 VI.12.2.3 Plutonium(V) oxalate complexes ...282 VI.12.2.4 Plutonium(VI) oxalate complexes ...283 VI.13 Americium oxalate comp ounds and complexes ...284 VI.13.1 Solid americium oxalates...284 VI.13.1.1 Am(III) compounds ...284 VI.13.1.2 Am(V) compounds ...285 VI.13.2 Aqueous americium oxalate complexes...285 VI.13.2.1 Am(III) oxalate complexes...285 VI.13.2.2 Am(V) oxalate complexes ...294 VI.13.2.3 Am(VI) oxalate complexes...295 citrate compounds and complexes .. 297 VII Discussion of data selection for VII.1 Introduction...297 VII.1.1 Metal i on citrates ...300 VII.2 Citric acid ...303 VII.2.1 H cit·H O(cr) ...303 3 2 cit(cr) ...304 VII.2.2 H 3 VII.2.3 H cit(aq) ...304 3 Protonation constants for citrate ...307 VII.3 VII.3.1 Introduction ...307 VII.3.2 Analysis of K ...315 3 VII.3.3 Analysis of K ...317 2 VII.3.4 Analysis of K ...319 1 ...321 VII.3.5 Analysis of K 0

15 CONTENTS xvii constants for citrate ...322 VII.3.6 Selected protonation VII.3.7 Temperature effects ...326 VII.4 Alkali metal citrate co mpounds and complexes ...331 + + ...331 and K VII.4.1 Complexes with Na VII.5 e compounds and complexes...335 Magnesium and calcium citrat citrate complexes ...335 VII.5.1 Magnesium and calcium citrate compounds...350 VII.5.2 Magnesium and calcium VII.6 Selenium citrate comp ounds and complexes...353 VII.7 Nickel citrate comp ounds and complexes ...353 VII.8 Technetium citrate co mpounds and complexes ...361 Zirconium citrate comp ounds and complexes ...361 VII.9 VII.10 Uranium citrate comp ounds and complexes...363 VII.10.1 U(IV) citr ate complexes ...363 VII.10.2 U(VI) citr ate complexes ...363 VII.11 Neptunium citrate comp ounds and complexes ...373 VII.12 Plutonium citrate comp ounds and complexes ...378 VII.13 Americium citrate comp ounds and complexes ...380 VIII Discussion of data selection for ethylenediaminetetraacetate (edta) compounds and complexes... 391 VIII.1 Introduction...391 VIII.2 H edta(cr)...394 4 VIII.2.1 Thermodynamic properties...394 VIII.2.2 Solubility of H edta(cr) ...394 4 ta in aqueous solutions ...396 VIII.3 Acid-base equilibria of ed VIII.3.1 Introduction ...396 VIII.3.2 Analysis of K K ...405 and 6 5 ...407 K VIII.3.3 Analysis of 4 VIII.3.4 Analysis of K ...409 3 K VIII.3.5 Analysis of ...411 2 ...413 VIII.3.6 Analysis of K 1 4– VIII.3.7 Selected proton ation constants for edta ...416 VIII.3.8 Temperature effects ...427 VIII.4 Alkali metal edta comp ounds and complexes ...437 + + VIII.4.1 Complexes with Na and K ...437 VIII.5 Magnesium and calcium edta compounds and complexes...443 VIII.5.1 Magnesium and calci um edta compounds ...443 ta complexes ...444 VIII.5.2 Stability of ma gnesium and calcium ed

16 CONTENTS xviii VIII.5.3 Enthalpy of complex formation ...455 VIII.6 Selenium edta comp ounds and complexes ...459 VIII.7 Nickel edta compou nds and complexes ...459 VIII.7.1 Nickel edta compounds ...459 edta complexes ...461 VIII.7.2 Stability of nickel VIII.7.3 Enthalpy of nickel edta complex formation...468 VIII.8 Technetium edta comp ounds and complexes...470 VIII.8.1 Technetium ed ta compounds ...470 VIII.8.2 Technetium ed ta complexes ...471 VIII.9 Zirconium edta compounds and complexes ...472 VIII.9.1 Zirconium edta compounds ...472 VIII.9.2 Stability of zirconi um edta complexes ...472 VIII.9.3 Enthalpy of zirconium edta complex formation ...475 VIII.10 Uranium edta compou nds and complexes...475 VIII.10.1 Uranium edta compounds ...475 VIII.10.1.1 U(IV) edta compounds ...475 VIII.10.1.2 U(VI) edta compounds ...476 VIII.10.2 Uranium edta complexes...477 VIII.10.2.1 U(V) edta complexes...477 VIII.10.2.2 U(III) edta complexes...477 VIII.10.2.3 U(IV) edta complexes ...478 VIII.10.2.4 U(VI) edta complexes ...484 VIII.10.3 Enthalpy of uranium edta complex formation...489 VIII.11 Neptunium edta comp ounds and complexes...490 VIII.11.1 Neptunium edta compounds...490 VIII.11.2 Neptunium edta complexes ...490 VIII.11.2.1 Np(III) edta complexes...490 VIII.11.2.2 Np(IV) edta complexes ...491 VIII.11.2.3 Np(V) edta complexes...492 VIII.11.2.4 Np(VI) ed xes ...498 ta comple VIII.11.3 Enthalpy of neptunium edta complex formation...498 VIII.12 Plutonium edta comp ounds and complexes ...498 VIII.12.1 Plutonium edta compounds ...498 VIII.12.2 Plutonium edta complexes ...498 VIII.12.2.1 Pu(III) edta complexes ...498 VIII.12.2.2 Pu(IV) edta complexes ...501 VIII.12.2.3 Pu(V) and Pu(VI) edta complexes...502 VIII.12.3 Enthalpy of plutonium edta complex formation ...503

17 CONTENTS xix ounds and complexes...503 VIII.13 Americium edta comp VIII.13.1 Americium edta compounds ...503 VIII.13.2 Americium edta complexes...503 VIII.13.2.1 Am(III) edta complexes...504 VIII.13.2.2 Am(V) edta complexes...507 VIII.13.3 Enthalpy of americium edta complex formation...508 IX Discussion of data selection for isosaccharinic acid (isa) compounds and complexes ... 511 IX.1 Introduction...511 IX.2 Isosaccharin ic acid ...512 IX.3 Protonation of isosaccharinate...513 IX.3.1 Lactonisation of c acid ...514 isosaccharini IX.3.2 Protonation of isosaccharin ate...515 IX.3.3 Enthalpy and entropy of pr otonation of isos accharinate...519 IX.4 Alkali metal isa compounds and complexes...519 IX.5 Alkaline earth metal isa compounds and complexes...520 Selenium isa compou IX.6 nds and complexes ...524 IX.7 nds and complexes...525 Nickel isa compou IX.8 Technetium isa compounds and complexes ...526 IX.9 Zirconium isa compou nds and complexes...526 nds and complexes ...526 IX.10 Uranium isa compou IX.11 Neptunium isa compounds and complexes ...530 IX.12 Plutonium isa compou nds and complexes...531 IX.13 Americium isa compou nds and complexes ...532 IX.14 Other compounds and complexes of isa ...533 IX.14.1 Trivalent lanthanide is a compounds and complexes ...534 IX.14.1.1 La(III) isa complex ...534 IX.14.1.2 Eu(III) isa complexes ...535 IX.14.1.3 Tb(III) isa complexes ...537 IX.14.2 Iron(III) isa compounds and complexes ...537 IX.14.3 Th(IV) isa compounds and complexes ...538 Part IV Appendices...541 A Discussion of selected references ... 543 B Ionic strength corrections... 819 raction equations ...821 B.1 The specific ion inte

18 CONTENTS xx B.1.1 Background ...821 B.1.2 mperatures other than 298.15 K ...826 Ionic strength corrections at te Estimation of ion inte raction coeffi B.1.3 cients...828 B.1.3.1 Estimation from mean activity coeffici ent data ...828 B.1.3.2 Estimations based on experimental values of equilibrium constants at different ionic strengths ...829 B.1.4 interaction coef ficients...832 On the magnitude of ion B.2 Ion interaction coeffici ents versus equilibrium constants for ion pairs..833 B.3 Tables of ion intera ction coefficients ...833 C Assigned uncertainties ... 849 C.1 The general problem. ...849 Uncertainty estimates in the se C.2 lected thermodynamic data. ...851 C.3 One source datum ...852 Two or more indepe data ...853 C.4 ndent source C.4.1 Discrepancies...855 C.5 erent ionic st rengths...857 Several data at diff C.5.1 Discrepancies or insufficie nt number of data points...859 C.6 Procedures for data handling...861 C.6.1 ionic strength ...861 Correction to zero C.6.2 Propagation of errors ...863 C.6.3 Rounding ...864 C.6.4 Significan t digits ...865 Bibliography ... 867 List of authors ... 1035

19 List of Figures Figure II-1: Standard order of arrangement of the elements and compounds based on the periodic classificati on of the elements ...25 ity coefficients adjusted wi th a Debye-Hückel term for Figure V-1: Mean ionic activ some 1:1 electrolytes (sodium acetate and chloride, ammonium chloride and tetramethylammonium chloride)...94 + 4– 3– Figure V-2: Expected deviations in SIT plots for reaction: H U HL , caused + L + 4– 3– + by the simultaneous reaction: M + L ML U , where M is the alkali metal cation of the bac lyte...96 kground electro Figure V-3: The effect of alkali meta l complex formation on SIT plots...97 rmula of oxa lic acid ...105 Figure VI-1: Structural fo Multi-linear least-squares SI T regression plots for the reaction: Figure VI-2: 2– – + ox U Hox + H . ...137 Figure VI-3: Multi-linear least-squares SIT regression plots for the reaction: ... – + Hox H + H ox(aq). ...141 U 2 and multi-linear least-squares SIT regression Figure VI-4: Enthalpy changes at 25°C 2– + – plots for the reaction: ox Hox + H ...148 U Figure VI-5: Enthalpy changes at 25°C and least-squares SIT regression plot for the – + reaction: Hox U H + H ox(aq). ...148 2 Figure VI-6: Solubility of Na ox(cr) and NaHox·H O(cr) in water as a function of 2 2 temperature...154 Figure VI-7: Solubility of K ox·H O(cr), KHox(cr) and KH (ox) ·2H O(cr) in water 2 2 2 2 3 as a function of temperature. ...156 Figure VI-8: Solubility product of Mg(ox)·2H O(cr) according to Reaction (VI.10), 2 calculated from analytical solubility data and extrapolated to zero ionic strength using parameters di scussed in the text. ...163 Figure VI-9: Solubility products of calcium oxalate hydrates at 25ºC, calculated from analytical solubility data and extrapolated to zero ionic strength using parameters discussed in the text. ...172 Figure VI-10: Solubility products of calcium oxalate hydrates at 37ºC, calculated from analytical solubility data and extrapolated to zero ionic strength using in the text. ...173 parameters discussed

20 LIST OF FIGURES xxii Figure VI-11: Temperature dependence of solubility products of calcium oxalate hy- ο ο and log K H ∆ values se- drates. The lines are calculated using 10 ,0 m s sol eir assigned uncertainties. ...175 lected in this review with th Figure VI-12: Simulated titration curves in 0.3 molal NaCl (left) and 3 molal NaCl –4 –4 molal molal H ox(aq) (upper curve) and 2x10 (right) of 2x10 2 –4 H ox(aq) and 2x10 molal MgCl (lower curve). ...180 2 2 Figure VI-13: Distribution of complex species in the simulated titration of the Mg ox- alate system (Figure VI-1 2) in 0.3 m NaCl (left) and 3 m NaCl (right). .. ...181 Figure VI-14: Weighted least squares SIT- regression plot of equilibrium data from for the formation of Mg (ox)(aq) according to Reac- [2001CHO/BON] at 25°C. ...184 tion (VI.17) in NaCl Figure VI-15: Weighted least squares SIT- regression plot of equilibrium data from [82DAN/MAR] for the formation of Mg(o x)(aq) according to Reaction (VI.17) in Et NI at 37°C...185 4 Figure VI-16: Weighted least squares SIT- regression plot of equilibrium data from − 2 [2001CHO/BON] for the formation of Mg(ox) according to Reaction 2 (VI.18) in NaCl at 25°C. ...186 regression plot of equilibrium data from Figure VI-17: Weighted least squares SIT- [82DAN/MAR] for the formation of Ca(ox)(aq) according to Reaction (VI.19) in Et NI at 37°C...187 4 Figure VI-18: Comparison of experimental and calculated values for the solubility of ssium oxalate; and (b) sulphuric acid nickel oxalate at 25°C in: (a) pota solutions. ...193 Figure VI-19: Weighted least squares SIT-regression plot of equilibrium data for the n = 1. ...198 formation of Ni(ox)(aq) according to Reaction (VI.21) with Figure VI-20: Weighted least squares SIT-regression plot of equilibrium data for the 2 − formation of N i(ox) according to Reaction (VI.21) with n = 2. ...198 2 Figure VI-21: Structure of the μ –hydroxy bridged tetranuclear cation, 8+ ·8H (OH) , present in solid ZrOCl (OH ) O...202 ] [Zr 16 4 2 2 8 2 Figure VI-22: Structures of the vari ously aged “zirconium hydroxides” ...203 Figure VI-23: Solubility of U(ox) ·6H O(s) in HCl.. ...208 2 2 Figure VI-24: Solubility of U(ox) ·6H O(s) in HClO and Na ox at different tempera- 4 2 2 2 . ...209 tures reported in [79NIK] Figure VI-25: Experimental solubility constants (a) and en thalpy (b) for the Reaction (VI.22), original data from [79NIK] ...212

21 LIST OF FIGURES xxiii Figure VI-26: Solubility of UO ox·3H O(s) in water at different temperatures reported 2 2 ture. ...216 in the litera Figure VI-27: Solubility of UO ox·3H O(s) in nitric acid at different temperatures re- 2 2 ported in the literature. ...217 Figure VI-28: log ...221 K T (VI.25) as a funtion of 1/ 10 s Figure VI-29: SIT plot for Reaction (VI.26). ...224 Figure VI-30: Stability cons tants (a) and enthalpy (b) for Reactions (VI.32) and (VI.33), original data from [79NIK] ...228 Figure VI-31: Postulated binding modes of oxalate in U(VI) complexes. (a) “side-on” bidentate; (b) “end -on” bidentate; (c ) unidentate. ...230 of Reaction (VI.37) ...239 Figure VI-32: SIT plot Figure VI-33: SIT plot of Reaction (VI.38) ...239 Figure VI-34: Speciation of U(VI) at 25°C, using the stability constants of ternary U(VI)-hydroxyl-oxalato comp lexes in Table VI-44. ...244 Figure VI-35: Comparison of some of the experimental solubility data for Np(IV) ox- alate in HNO at 23 and 22°C, solutions reported in [64POR] , [83LUE] 3 respectively...246 Figure VI-36: Comparison of equilibrium constants for the complex formation be- 25ºC...253 ≈ and oxalic acid at tween Np(IV) – + 2– Figure VI-37: Extrapolation of data for reaction NpO ox NpO listed in U + ox 2 2 Table VI-50 according to the SIT model described in Appendix B. ..260 2– 3 + − U N pO + 2 ox N pO(ox) listed Figure VI-38: Extrapolation of data for reaction 2 2 in Table VI-50 according to the SIT model described in Appendix B.260 Figure VI-39: Comparison of experimental Pu(III) concentrations in solutions in con- tact with Pu(III)-oxalate and vary ing concentrations of the oxalate ligand...266 Figure VI-40: Comparison of the experimental solubility data for Pu(III) oxalate: a) in solutions containing ammonium oxalate and nitric acid; and b) in solu- tions containing oxalic and nitric acids. ...267 Figure VI-41: Experimental Pu(IV) concentrations measured in solutions in contact with Pu(IV)-oxalate. ...271 Figure VI-42: Comparison of some of the experimental solubility data , [58GEL/DRA2] , [58GEL/DRA3] for Pu(VI) oxalate [58DRA/MOS2] ng nitric acid...276 in solutions containi

22 LIST OF FIGURES xxiv Figure VI-43: Overview of the equilibrium constants at ≈ 25ºC listed in Table VI-59 42 − n Pu(ox) with n = 1 to 3. ...281 for the Pu(IV) complexes n Figure VI-44: Fitting of the values of log K (VI.60) at each ionic strength (Table VI- 10 63) to the SIT equation. ...289 . ...290 K (VI.60) against I Figure VI-45: Plot of log 10 m Figure VI-46: Fitting of the values of log (VI.61) at each ionic strength (Table VI- K 10 63) to the SIT equation. ...291 . ...292 K (VI.61) against I Figure VI-47: Plot of log 10 m es of citric acid in H Figure VII-1: The structur O(cr) [72ROE/KAN] and of cit- cit·H 3 2 . ...297 [86VIO/ROD] cit·5.5H O(cr) rate in Na 2 3 2 − Figure VII-2: The structures of: (a) [Ni(cit)(H O) ] [83BAK/BAK] , and (b) 222 5 − Fe(cit) [98MAT/RAP] ...299 2 Figure VII-3: The structure of SbNa(cit) (H O) [91HAR/SMI] ...300 2 2 2 Figure VII-4: The total concentration of citrate at equilibrium with Ca cit ·4H O in 2 2 3 calcium chloride solutions as a function of pH at 25°C. ...303 Figure VII-5: Multi-linear least-squares SIT regression plots for the reaction: – + cit(aq)...315 H cit + H U H 2 3 SIT regression plots for the reaction: Figure VII-6: Multi-linear least-squares + – 2– Hcit H cit U ...317 + H 2 Figure VII-7: Equilibrium constants for Reaction (VII.7). Data from Table VII-2 were extrapolated to 25°C and converted to molal units and plotted according to the SIT methodology. ...320 Figure VII-8: Calculated distribution of citrate species as a function of pH in 1 M NaCl at 25°C. ...323 3– + 2– at 25°C for the reaction: cit Figure VII-9: Enthalpy changes U plotted + H Hcit methodology. ...329 according to the SIT 2– + – at 25°C for the reaction: Hcit Figure VII-10: Enthalpy changes plotted U H cit + H 2 methodology. ...329 according to the SIT – + U cit + H cit(aq) H s at 25°C for the reaction: H Figure VII-11: Enthalpy change 2 3 plotted according to th e SIT methodology. ...330 –3 –3 Figure VII-12. Simulated titration curves of (1) 2x10 M citric M citric acid, (2) 2x10 2+ –3 = 0.1 M. ...341 M Ca at I acid and 2x10 –3 Figure VII-13: Distribution of complex species in the simulated titration of 2x10 M –3 2+ = 0.1 M. ...341 M Ca I at citric acid and 2x10

23 LIST OF FIGURES xxv − − 2+ 3 Figure VII-14: Fitting of the values of U Mg log of β + cit at each Mg(cit) 10 ionic strength (given in Table VII-13) to the SIT equation. ...345 β log Figure VII-15: Fitting of the values of (VII.10) at each ioni c strength (given in 1 10 Table VII-13) to th e SIT equation. ...346 2+ Figure VII-16: Fitting of the values of log β of Mg -citrate at each ionic strength 1 10 ...347 2+ Figure VII-17: Fitting of the values of log β of Ca -citrate at each ionic strength 1 10 ...348 Figure VII-18: The experimental solubility product data of Ca (cit) ·4H O(cr) and the 3 2 2 fitting of the selected values using equation (VII.19), including and ne- glecting the ∆ε and terms...351 na log H O 10 2 –3 –3 Figure VII-19: Simulated titration curves of (1) 5x10 M citric acid, (2) 5x10 M citric 2+ –3 acid and 5x10 M Ni at I = 0.1 M...357 –3 Figure VII-20: Distribution of complex species in the simulated titration of 5x10 M –3 2+ citric acid and 5x10 M Ni I = 0.1 M...358 at 2+ Figure VII-21: Fitting of the values of log for the Ni -citrate system at each ionic β 10 strength (given in Table VII-19) to equations (VII.43) and (VII.44). 360 Figure VII-22: Simulated titration curves of 0. 01 M citric acid (upper dashed line), 0.01 2+ M citric acid and 0.01 M UO (solid line), and 0.01 M strong acid 2 (lower dashed line) at I = 0.1 M. ...367 of the relevant species in the titration of Figure VII-23: Changes in the concentrations 2+ 0.01 M citric acid and 0.01 M UO shown in Figure VII-22. ...368 2 Figure VII-24: Fitting of the values of log β (VII.45) at each ionic strength (given in 10 Table VII-24) to th e SIT equation. ...370 Figure VII-25: A plot of . ...371 I log β (VII.45) against m 1 10 Figure VII-26: Fitting of the values of log (VII.46) at each ionic strength ( cf . Table β 10 VII-24) to the SIT equation (VII.52). ...372 Figure VII-27: Values of log β (VII.46) versus ionic strength. ...372 10 Figure VII-28: Fitting of the values of log strength (Table VII- β (VII.53) at each ionic 10 27) to the SIT equation. ...376 Figure VII-29: Plot of ionic strength. ...377 versus log β (VII.53) 1 10 Figure VII-30: Fitting of the values of log β (VII.58) at each ioni c strength (Table 10 1 VII-33) to the SIT equation. ...385

24 LIST OF FIGURES xxvi β against I for the reactions, (VII.58) and Figure VII-31: A plot of log m 10 33 +− − 3 + Am 2 cit Am(cit) U (VII.63) ...386 2 Figure VII-32: A Plot of log β against . Solid lines are drawn by using the results of I 10 m fitting equation (V II.64). ...387 of Am(III)-citrate complex species in solutions containing Figure VII-33: Distribution + –6 –3 log [H ] − 10 M citric acid as a function of M Am(III) and 10 I = at 10 0.1 M. ...389 Figure VIII-1: The structure of H edta as a double zwitterion...392 4 2– Figure VIII-2: The stru cture of: (a) Ni(edta) [84NES/POR2] , and (b) – Ni(Hedta)(H O) [86POL/FIL] ...392 2 2– Figure VIII-3: Structures of two mixed-ligand complexes: ZrCO [95MIS/SER] edta 3 3– and ThF [85MIK/LOB] ...393 edta 3 polynuclear edta complexes: Tc Figure VIII-4: Structures of O (H edta) 2 2 2 2 4– [81BUE/AND] ...393 F [85SHC/ORL] ) edta , and (UO 2 2 2 Figure VIII-5: Equilibrium constants for reaction: H edta(aq). ...396 edta(cr) U H 4 4 ( r –4) Figure VIII-6: Schematic structures of H edta ...397 r Figure VIII-7: Linear least squares SIT-regression plot for the Reaction (VIII.2). ...405 gression plot for R Figure VIII-8: Linear least squares SIT-re eaction (VIII.3). ...406 Figure VIII-9: Multi-linear least squares SI for the reaction: T-regression plots + – H edta U H + H edta(aq). ...408 4 3 linear least squares SIT-regressi on plots for the reaction: Figure VIII-10: Multi- 2– + – * H U H to be independent of ionic edta ε , assuming + H ∆ edta 2 3 3 * strength for sodium and potassium electrolytes, and I = f ( ε ∆ ) for m 3 tetraalkylammonium media. ...410 linear least squares SIT-regressi on plots for the reaction: Figure VIII-11: Multi- 3– + 2– * Hedta ∆ + H U , assuming H to be independent of ionic ε edta 2 2 * ε ∆ strength for sodium and potassium electrolytes, and f ( I ) for = m 2 tetraalkylammonium media. ...412 Figure VIII-12: Weighted least squares SIT-regression plots for the reaction: 4– + 3– 3– edta Hedta , assuming: a) the formation of Na(edta) U and + H 3– * Kedta ; b) that the values of ε ∆ are independent of ionic strength for 1 * sodium and potassium electrolytes, and c) that ) for ε ∆ = f ( I m 1 tetraalkylammonium media. ...415 Figure VIII-13: Calculated distribution of diss olved ethylenediamin etetraacetate species 1 M NaCl at 25°C. ...417 as a function of pH in

25 LIST OF FIGURES xxvii –4) r ( + ( r –5) + H H edta edta plotted U Figure VIII-14: Enthalpy changes for reactions: H r –1) r ( according to the SIT methodology. ...435 3– Figure VIII-15: Enthalpy data for the formation of Na(edta) plotted according to the SIT model. ...440 3– 3– Figure VIII-16: Equilibrium constants for the formation of Na(edta) and Kedta plot- ted according to the SI T methodology. ...442 –3 Figure VIII-17: Simulated titration curves of 1x10 molal H edta(aq) in 1 molal NaCl 4 –3 –3 (upper curve) and 1x10 edta(aq) and 1x10 molal MgCl m H in 1 4 2 molal NaCl (lower curve). ...448 Figure VIII-18: Distribution of complex species in the simulated titration of the Mg edta system (Figure VIII-17) in 1 m NaCl. ...448 Figure VIII-19: Mu ession plot for the reaction lti-linear least-squares SIT regr 2+ 4– 2– Mg + edta U Mg(edta) . ...452 Figure VIII-20: Extrapolation to infinite dilution of log for the formation of K 10 1 2– Ca(edta) in tetramethylammonium chloride solution at 25°C [83ARE/MUS] . ...453 Figure VIII-21: Extrapolation to infinite dilution of log K for the reaction 10 2– – + Mg(edta) + H MgHedta in NaCl at 25°C [2001CHO/BON] ...454 U Figure VIII-22: Weighted least squares SIT-regression plot of enthalpy data from 2– [63AND] , [76VAS/BEL2] for the formation of Mg(edta) ...458 Figure VIII-23: Weighted least squares SIT-regression plot of enthalpy data from [56CAR/STA] , [63AND] , [76VAS/BEL2] for the formation of 2– Ca(edta) . ...459 Figure VIII-24: Weighted least squares SIT-regression plot of enthalpy data from 2– [63AND] [76VAS/BEL3] for the formation of Ni(edta) , . ...469 Figure VIII-25: Weighted least squares SIT-regression plot of equilibrium data from [62KRO/ERM] for the formation of Uedta(aq) according to Reaction (VIII.23). ...483 Figure VIII-26: Weighted least squares SIT-regression plot of equilibrium data from 2– [98POK/BRO] edta according to Reaction for the formation of UO 2 (VIII.25). ...487 Figure VIII-27: Weighted least squares SIT-regression plot of equilibrium data from – [98POK/BRO] for the formation of UO (Hedta) according to Reaction 2 (VIII.26). ...488

26 LIST OF FIGURES xxviii Figure VIII-28: Weighted least squares SIT-regression plot of equilibrium data from – (H edta) according to Reac- for the formation of NpO [98POK/BRO] 2 2 tion (VIII.29 ). ...495 Figure VIII-29: Weighted least squares SIT-regression plot of equilibrium data from 2– [98POK/BRO] (Hedta) for the formation of NpO according to reaction 2 (VIII.31). ...496 Figure VIII-30: Weighted least squares SIT-regression plot of equilibrium data from 3– [98POK/BRO] for the formation of NpO edta according to Reaction 2 (VIII.33). ...497 Figure VIII-31: Weighted least squares SIT-regression plot of equilibrium data from – [99CHE/CHO] according to Reaction for the formation of Am(edta) (VIII.36). ...506 Figure IX-1: (a) -isosaccharinic acid and (b) β -isosaccharinic acid...512 α Figure IX-2: Dehydration of α -isosaccharinic acid to form α -isosaccharinate-1,4- lactone. ...512 Figure IX-3: Two representations of the calcium α -isosaccharinate st ructure in the solid Ca(isa) (cr) [68NOR/WER] ...520 2 Figure IX-4: Correlation between the stability constants of U(VI) carboxylate com- plexes and the p K t = 20 – 25°C, I = 1.0 M). ...529 of carboxylic acids ( a Solubility of UO Figure A-1: (ox)·3H O in solutions of mineral acids. ...546 2 2 Heat capacity of calcium oxalate monohydrate as a function of Figure A-2: temperature...550 Figure A-3: of calcium oxalate monohydrate as a function of Heat capacity temperature...550 Figure A-4: Deviations of observed heat capacities of calcium oxalate monohydrate [33LAT/SCH] from the values of the polynomial function fitted by this review. ...551 Figure A-5: First protonation constant for oxalate determined and extrapolated to I = 0 by [39HAR/FAL] model with the parameters C ∆ and the constant r,m p determined in this review. ...554 Figure A-6: First protonation constant for oxalate determined and extrapolated to I = 0 by [48PIN/BAT] and the constant C ∆ model with the parameters p r,m determined in this review. ...559 Figure A-7: Solubility of UO (ox)·3H O in nitric acid at 20°C reported in the 2 2 literature. ...564

27 LIST OF FIGURES xxix (ox)·3H O in aqueous solutions of oxalic acid at vari- Solubility of UO Figure A-8: 2 2 able temperatures...571 Figure A-9: Solubility of UO ox·3H at 25 and 50°C O in HNO .578 [58BOL/KOR] 3 2 2 [58FUG] Figure A-10: Experimental data of Fuger evaluated using the edta protona- [51CAB] tion constants of Cabell (grey symbols), and using protonation view (black symbols). ...583 constants selected in this re Figure A-11: Calculated solubilities of Pu(IV)-ox compared with the experimental data at 20°C reported in [58GEL/MOS2] , [58MOS/GEL3] , [58MOS/GEL] ...588 [60STA3] Figure A-12: Replot of Figure 4 from ...608 Figure A-13: Second protonation constant for oxalate determined and extrapolated to H ∆ I model with the pa- = 0 by [61MCA/NAN] and the constant rm rameters determined in this review...613 Figure A-14: Experimental solubility of H = 1, 2 edta(cr) in (H,Na)NO I solutions of 3 4 and 3 M at 25°C [62KRO/ERM] compared with calcu lated values using the equilibrium constant s in Table A-6. ...618 Figure A-15: Structural unit of the th ree types of zirconium hydroxides ...645 Figure A-16: Experimental electromigration data for the system Am – edta of Lebe- dev [67LEB/MAK] . ...650 et al Figure A-17: U(VI) speciation in th e U(VI)-oxalate-sulfite system ( I = 2.5 M) using [67ZAK/ORL2] the values from . ...657 Figure A-18: Second protonation constant for oxalate determined and extrapolated to = 0 by [69KUR/FAR] and the constant I H ∆ model with the parame- rm ters determined in this review...664 Figure A-19: Calculated species distribution in aqueous solutions of H SO equili- 2 4 brated with β –Ni(ox)·2H O at 25°C. ...699 2 Figure A-20: Experimental data from solutions equilibrated with –Ni(ox)·2H O at β 2 25°C...700 Figure A-21: U(VI) speciation ( C = 0.004M) ...706 U Figure A-22: Second protonation constant for oxalate determined in 1 M NaClO by 4 [76KAL] and the constant H ∆ model with the parameters determined rm in this review. ...708 Figure A-23: Experimental solvent extraction data from [76MCD/KEL] (open circles) re-fitted in this review assuming two complexes, Ca(ox)(aq) and 2 − ...709 Ca(ox) 2

28 LIST OF FIGURES xxx Figure A-24: Enthalpy of the first protonation constant for oxalate determined in the C ∆ model with and the constant indicated media by [76VAS/SHE] r,m p the parameters determined in this review. ...710 Figure A-25: Enthalpy of the second proton ation constant for oxalate determined in 0.15 M NaNO [76VAS/SHE] and the constant by C ∆ model with 3 r,m p the parameters determined in this review. ...711 + 4– 2+ ges for the reaction: edta Figure A-26: Enthalpy chan U plotted edta + 6 H H 6 according to the SIT methodology. ...718 Figure A-27: Enthalpy ch anges for the reaction: H U H edta(aq) plotted edta(cr) 4 4 methodology. ...727 according to the SIT 2+ + anges for the reaction: H Figure A-28: Enthalpy ch H U edta edta(aq) + 2 H plot- 6 4 ted according to the SI T methodology. ...728 Figure A-29: First protonation constant for oxalate determined by [90ROB/STE] in Et NI medium, corrected to the molal scale in this review, and the con- 4 stant model with the parameters determined in this review...767 C ∆ r,m p [92ROB/STE] Figure A-30: First protonation constant for oxalate determined by in NaCl medium and the constant C (A.102) model with the parame- ∆ r,m p ters determined in this review...774 [92ROB/STE] Figure A-31: First protonation constant for oxalate determined by in KCl medium and the constant C ∆ (A.102) model with the parame- p r,m ters determined in this review...774 Figure A-32: First protonation constant for oxalate determined by [92ROB/STE] in Et (A.102) model with the parame- NI medium and the constant ∆ C 4 p r,m ters determined in this review...775 Figure A-33: Protonation constants for citrate in ≈ 0.1 molal NaCl determined by [97BEN/PAL] and the constant model with the parameters de- ∆ C p r,m termined in this review. ...790 Figure A-34: First protonation constant for oxalate determined by [98KET/WES] in NaCF SO medium and the constant model with the parameters ∆ C 3 3 p r,m determined in this review. ...793 Figure A-35: First protonation constant for oxalate determined by [98KET/WES] in NaCl medium and the constant C ∆ model with the parameters de- r,m p termined in this review. ...794 Figure A-36: Second protonation constant for oxalate determined by [98KET/WES] in NaCl medium and the constant C ∆ model with the parameters r,m p determined in this review. ...794

29 LIST OF FIGURES xxxi + ox – Na ∆ ] versus x in the system H : calculated values [H SeO Figure A-37: Plot of 2 3 2 by this review ignoring complex formation between oxalate and selenite compared with experimental data from [98OBR/MIT] . ...797 Figure A-38: Experimental activities of water for mixtures of H cit and NaCl at 25°C 3 [2004SCH/MAU] compared with values calculated using ε (H cit,NaCl) 3 –1 = (0.03 ± 0.03) kg·mol . ...817 Figure A-39: Experimental activities of water and corresponding osmotic coefficients for Na cit solutions at 25°C compared with [2004SCH/MAU] 3 3– + values calculated using ,cit (Na ε ± 0.01) + (0.145 ± 0.020) ) = – (0.12 log ...817 m Na+ 10 Figure B-1: Plot of for Reaction (B.12), at 25°C and 1 bar I log 4 D + β versus m 1 10 ...830

30 List of Tables Table II-1: Abbreviations fo r experimental methods ...9 Table II-2: and terminology...11 Symbols Table II-3: Abbreviations used as subscripts of ∆ to denote the type of chemical process. ...15 Table II-4: Unit conv ersion factors ...26 Table II-5: Factors Š for the conversion of molarity, c , to molality, m , of a sub- B B stance B, in various media at 298.15 K ...27 Table II-6: Reference states for some el ements at the reference temperature of 298.15 K and standard pressure of 0.1 MPa...29 Table II-7: Fundamental physical co nstants ...33 Table III-1: Selected thermodynamic data for oxalate compounds and complexes.43 Table III-2: Selected thermodynamic data for reaction involving oxalate compounds and complexes ...45 Table III-3: Selected thermodynamic data for citrate compounds and complexes ..47 Table III-4: Selected thermodynamic data for reaction involving citrate compounds and complexes ...49 Selected thermodynamic data Table III-5: for edta compounds and complexes...52 Table III-6: Selected thermodynamic data for reaction involving edta compounds and complexes ...53 Table III-7: Selected thermodynamic data for reaction involving isa compounds and complexes...56 Table IV-1: Selected thermodynamic data for auxiliary compounds and complexes59 Table IV-2: Selected thermodynamic data for reaction involving auxiliary com- pounds and complexes...73 Table V-1: The ratio between molarity, c , and molality, m , for selected electrolytes. ...99 ο ∆ H Table VI-1: Literature values for . ...109 rm

31 LIST OF TABLES xxxiii Literature data on the protonation constants for oxalate considered in Table VI-2: this review. ...116 Table VI-3: Accepted data on the protonati on constants for oxal ate. The protonation its and extrapolated to 25°C where constants were corrected to molal un necessary. ...129 otonation constants for oxalate. ...133 Table VI-4: Accepted data on the pr ο * at 25°C. ...136 log K (VI.5) and ε ∆ Selected values of Table VI-5: 1 1 10 * ο Table VI-6: Selected values of ε log (VI.7) and at 25°C. ...140 ∆ K 2 2 10 Table VI-7: Literature data on the enthalpy and heat capacity changes for oxalate protonation, with the uncertainties assigned in this review. ...145 Table VI-8: Calculated equilibrium constants in Molar units for some + 2– + + ox /Na /K /H systems at 25°C...150 –1 Table VI-9: Selected sp ecific ion interaction coefficients (kg·mol ) for oxalate and + + + its protonated forms in Li and tetraethylammonium electro- , Na , K lytes. ...151 and potassium oxalate compounds...153 Table VI-10: Sodium Table VI-11: Experime ntal equilibrium data for Na and K oxalate systems. ...157 and calcium oxalate compounds in pure Table VI-12: Solubility data of magnesium water as reported in the literature. ...160 Table VI-13: Enthalpies of dehydration of calcium oxalate monohydrate reported in the literature...165 Table VI-14: Solubility products of magnesium and calcium oxalate compounds re- ported in the literature and re-e valuated in this review. ...167 Table VI-15: Enthalpies of dissolution of calcium oxalate hydrates reported in the literature and re-evaluated in this review. ...174 Table VI-16: Selected fo rmation data for Ca(ox)·H O(cr) and O(cr), Ca(ox)·2H 2 2 Ca(ox)·3H O(cr). ...177 2 Table VI-17: Experimental equilibrium data for the Mg and Ca oxalate systems ...178 Table VI-18: Accepted equilibrium data fo r the Mg and Ca oxalate systems...182 Table VI-19: Selected values for Mg and Ca oxalate compounds and complexes. .189 Table VI-20: Reported values for the solu bility of nickel oxal ate in water. ...191 Table VI-21: Literature values for the solubility products and thermodynamic data for 2+ 2– the reaction: Ni(ox)·2H O U Ni O(l). ...192 + ox + 2H 2 2

32 LIST OF TABLES xxxiv Table VI-22: The stability constants of Ni(II) oxalate complexes reported in literature ...194 Table VI-23: Accepted stabilit y constants of nickel(II) oxalate complexes at 25°C ed in this review. ...197 with the uncertainties assign alpy changes associated with the formation Table VI-24: Literature values for the enth of nickel(II) oxalate co mplexes at 25°C. ...199 Table VI-25: Selected literature data for the enthalpy changes associated with the formation of nickel (II) oxalate comple xes at 25°C. ...200 Table VI-26: Technetium ox alate compounds reported in the literature...201 Table VI-27: Literature data on the solubility product of the solid compounds of zir- conium with oxalate. ...204 Table VI-28: Literature data on the formation constants of oxalate complexes of zir- conium...205 Table VI-29: Solid U(IV) oxalates of which the solubilities are reported in the litera- ture...207 Table VI-30: Solubility constants of solid uranium(IV) oxalates reported in the litera- ture...210 Table VI-31: Experimental enthalpy of Reaction (VI.22) reported in [79NIK] ...211 Table VI-32: Solid U(VI) oxalates of which the structures by X-ray crystallography or the solubilities are reported in the literature. ...213 Table VI-33: Studies of the solubility of UO O(s) considered by this review. ox·3H 2 2 ...215 Table VI-34: Solubility constants of UO (ox)·3H O(s) reported in the literature...219 2 2 Table VI-35: Solubility constants of UO O obtained by re-evaluation of the (ox)·3H 2 2 data in Tabl e VI-34...223 Table VI-36: Stability constants of aqueous U(IV) oxalates reported in the literature. ...226 Table VI-37: Experimental enthalpy of Reactions (VI.32) and (VI.33) from [79NIK] . ...227 Table VI-38: Stability constants of aqueous U(VI) oxalates reported in the literature. ...231 Table VI-39: Stability consta nts of aqueous uranium(VI) oxalates at 25°C accepted by this review. ...236

33 LIST OF TABLES xxxv ο ο log β and log β by weighted and unweighted lin- Table VI-40: Calculation of 2 10 1 10 ear regression...238 ο β log Table VI-41: Calculation of at 298.25 K. ...240 10 3 Table VI-42: Selected stability constants of aqueous U(VI) oxalates at 298.15 K. .241 Table VI-43: Stability constants of dinuclear U(VI) oxalates reported in the literature. ...242 Table VI-44: Stability constants of ternary U(VI) oxalate complexes reported in the literature. ...243 Table VI-45: Solubility constants of solid Np(IV) oxalate reported in the literature. ...247 + 4+ Table VI-46: Solubility constants for reaction: Np(ox) O(cr) + 4 H U Np ·6H + 2 2 2 H ox(aq) + 6 H O(l), obtained from the re-evaluations described in 2 2 Appendix A of experimental literature solubilities of Np(IV) oxalate at 25°C. ...248 ≈ Table VI-47: Literature stability constants for neptunium(IV) oxalate complexes..250 Table VI-48: Results from the re-evaluatio ns described in Appendix A of experimen- tal literature data on the complex formation between Np(IV) and oxalic acid at 25ºC. ...252 ≈ Table VI-49: The stability constants of neptunium(V) oxalate complexes reported in the literature...254 Table VI-50: Accepted stab ility constants for neptunium(V) oxalate complexation, with the uncertainties assign ed by this review. ...257 Table VI-51: Pu-oxalate compounds reported in the literature. ...263 Table VI-52: Literature studies reportin g the solubility of Pu(III)-oxalate. ...264 Table VI-53: Literature studies reportin g the solubility of Pu(IV)-oxalate...268 Table VI-54: Literature solubility consta nts involving solid Pu(IV) oxalate. ...272 Table VI-55: Results from re-analysis performed by this review of literature data on Pu(IV) oxalate so lubility. ...273 Table VI-56: Literature solubility consta nts involving solid Pu(VI) oxalate. ...275 Table VI-57: Literature stability consta nts for Pu(III) oxal ate complexes...277 te complexes...278 Table VI-58: Literature stability consta nts for Pu(IV) oxala

34 LIST OF TABLES xxxvi for Pu(IV) oxalate co mplexes converted to Table VI-59: Stepwise stability constants 3+ , and when neces- molal units and corrected for the formation of PuOH + 3 PuNO sary for . ...280 3 tants for Pu(V) oxalate complexes. ...282 Table VI-60: Literature stability cons Table VI-61: Literature stability consta nts for Pu(VI) oxalate complexes...284 Table VI-62: Literature data on the solubility product of the solid compounds of Am(III) with oxalate...285 Table VI-63: Literature data on the formation constants of oxalate complexes of Am(III). ...286 Table VI-64: Accepted forma tion constants for oxalate co mplexes of Am(III) at 25°C lected values...288 used to derive the se 3+ Table VI-65: Selected formation cons tants for the oxalate complexes of Am at 25°C...294 Table VI-66: Literature data on the formation constants of Am(V) oxalate complexes ...294 ο Table VII-1: Literature values for ∆ H (VII.3). ...305 rm Table VII-2: Literature data on the protonation constants for citrate considered in this review. ...310 * ο Table VII-3: Selected values of (VII.5) and log ∆ ε at 25°C. ...316 K 3 10 3 ο * Selected values of Table VII-4: ∆ log at 25°C...318 K (VII.6) and ε 2 2 10 Table VII-5: Results from regressions of experimental data obtained with each of the background cations at 25°C...319 * ο Selected values of Table VII-6: at 25°C. ...320 K log (VII.7) and ∆ ε 1 10 1 4– Table VII-7: Literature data on the protonation constant for H cit included in the –1 review process. ...322 Table VII-8: Calculated equilibrium constants in Molar units for some + + 3– + cit /H /K systems at 25°C...324 /Na –1 Selected specific ion interaction coefficients (kg·mol Table VII-9: ) for citrate and + + + its protonated forms in Li , Na , K and tetramethylammonium electro- lytes. ...326 Table VII-10: Literature data on the enthalpy changes for citrate protonation, with as- signed uncertainties. ...326 Table VII-11: Literature data on the formation of citrate complexes with alkali cations + + + (M = either Na or K ). ...332

35 LIST OF TABLES xxxvii 2+ Table VII-12: Literature data on the formation constants for citrate complexes of Mg 2+ and Ca . ...335 2+ 2+ Table VII-13: Accepted fo rmation constants for citrate complexes of Mg and Ca lected values...342 used to derive the se 2+ 2+ Table VII-14: Selected fo rmation constants for the citrate complexes of Mg and Ca at 25°C...349 2+ energy of formation of the citrate complexes of Mg Table VII-15: Selected Gibbs – 2+ and Ca and selected molar entropy of Ca(cit) . ...350 2+ Table VII-16: Literature data on the solubility product of the solid compounds of Mg 2+ and Ca with citr ate. ...350 Table VII-17: Accepted literature values of t of the solid com- the solubility produc 2+ pounds of Ca with citrate to estimate the selected values at 25°C...352 2+ Table VII-18: Literature data on the formation constants for citrate complexes of Ni . ...354 2+ Table VII-19: Accepted fo rmation constants for c itrate complexes of Ni at 25°C used to derive the selected values. ...359 2+ tants for the citrate complexes of Ni Table VII-20: Selected formation cons at 25°C ...361 4+ Table VII-21: Literature data on the formation constants for citrate complexes of Zr ...362 4+ Table VII-22: Literature data on the formation constants for citrate complexes of U ...363 Table VII-23: Literature data on the formation constants for the citrate complexes of 2+ UO ...364 2 2+ Table VII-24: Accepted fo rmation constants for citrate complexes of UO used to 2 derive the selected va lues at 25°C. ...369 2+ UO Table VII-25: Selected fo rmation constants for the citrate complexes of at 25°C 2 ...373 ation constants of citrate complexes of Table VII-26: Literature data on the form + NpO . ...374 2 + rmation constants for citrate complexes of Table VII-27: Accepted fo pO used to N 2 derive the selected va lues at 25°C. ...375 3+ Table VII-28: Literature data for the formation constants of citrate complexes of Pu ...378

36 LIST OF TABLES xxxviii Table VII-29: Literature data for the formation constants of citrate complexes of 2+ PuO ...378 2 4+ Table VII-30: Experimental equilibrium data for Pu citrate system at 25°C...379 4+ Table VII-31: Literature data on the formation constants for citrate complexes of Pu 4+ or Th expressed as log β at 25°C. ...380 10 Table VII-32: Literature data on the formation constants for citrate complexes of 3+ Am . ...381 3+ Table VII-33: Accepted fo rmation constants for citrate complexes of Am at 25°C used to derive the se lected values...382 3+ Table VII-34: Selected formation cons tants for the citrate complexes of Am at 25°C ...388 Table VIII-1: Literature studies considered by this review where the solubility of + + H edta(cr) has been reported as a function of [H ] ) at con- ] (or [H 4 TOT stant ionic strength...395 Table VIII-2: Literature data considered by this review on the protonation constants, and K log K , log , log K , K log log K (Table VIII-2–a), 10 10 3 10 2 4 5 10 1 10 4– . ...399 (Table VIII-2–b), for edta log K 6 10 * ο log K ε ∆ Table VIII-3: Selected values of (VIII.4) and (VIII.4)...407 4 4 10 * ο Table VIII-4: Selected values of (VIII.5) and K log ∆ ε (VIII.5). ...409 3 3 10 * ο Table VIII-5: Selected values of log (VIII.6)...411 (VIII.6) and ε K ∆ 2 10 2 Table VIII-6: Results of non-linear weighed least-squares fitting of literature data for 4– the first protonation of edta at (25 ± 5)°C. ...414 –1 4– Table VIII-7: Specific ion inte raction coefficients (kg·mol ) for edta and its proto- nated species...416 4– Table VIII-8-a: Calculated values for the protonation constants of edta and for alkali- + + metal complex formation in Molar units in some Na /K electrolytes at 25 and 20°C...418 4– lues for the apparent first protonation constants of edta Table VIII-8-b: Calculated va in + + molar units in some Na /K background electrolytes at 25 and 20°C.425 Table VIII-9: Literature data on the enthal py changes for ethyle nediaminetetraacetate step-wise protonation reactions in sodium, potassium or tetraalkyl- ammonium media. ...428 Table VIII-10: Literature data considered by this review on the enthalpy changes for ethylenediaminetetraacetat e protonation, excluding step-wise reactions. ...431

37 LIST OF TABLES xxxix anges at 25°C for the stepwise protonation reactions of Table VIII-11: Heat capacity ch 4– ...433 edta + H ∆ enthalpy changes, Table VIII-12. Selected (VIII.9) and corresponding ∆ε (M ) L rm values at 25°C...434 + 4– + Table VIII-13: Literature data on the formation of edta complexes with Na and K considered in this review. ...438 Table VIII-14: Literature data on the enthalpy changes for the complex formation be- 4– 4– + + tween edta . The data correspond to reaction: Na + edta U and Na 3– Na(edta) . ...439 Table VIII-15: Magnesium and calcium edta compounds. References reporting solubil- ity data are marked with (sol.). ...443 Table VIII-16: Experimental equilibrium data for the Mg and Ca edta systems. The uncertainties are given as reported in the references. ...445 Table VIII-17: Accepted forma tion constants for the Mg and Ca edta systems to derive the selected values. ...450 Table VIII-18: Experi Mg and Ca edta systems. ...455 mental enthalpy data for the Table VIII-19: Accepted enthal py data for the Mg and Ca edta systems to derive the selected values...458 Table VIII-20: Ni porting solubility data are marked ckel edta compounds. References re with (sol.). References reporting X- ray single crystal structures are marked with (str.), (am.) refer to LAXS (Large-Angle X-ray Scattering) analyses of amorphous solids; (en) stands for ethylenediamine (C H N ) 8 2 2 and (big) represents biguanide (C )...460 H N 7 2 5 Table VIII-21: Experimental equilibrium data for the Ni edta system. The uncertainties are given as reported in the references. ...461 Table VIII-22: Experimental equilibrium data for the Ni edta X system where the ligand X forms a ternary complex Ni(edta)X. ...463 Table VIII-23: Accepted formation constants for the Ni ed ta systems to derive the se- lected values. ...465 rimental enthalpy data for the Ni edta system. ...468 Table VIII-24: Expe Table VIII-25: Expe rimental equilibrium data fo r the Zr edta system. ...473 Table VIII-26: Expe rimental enthalpy data for the Zr edta system. ...475 Table VIII-27: U( IV) edta compounds. A reference reporting solubility data is marked with (sol.). ...476

38 LIST OF TABLES xl Table VIII-28: U( VI) edta compounds. A reference reporting solubility data is marked with (sol.), and a reference reporting an X-ray single crystal structure is marked with (str.). ...477 Table VIII-29: Equilibri um data for the U(III) edta system. ...477 Table VIII-30: Experi the U(IV) edta system. ...478 mental equilibrium data for Table VIII-31: Experimental equilibrium data for the U(IV) edta X system where the ligand X forms a ternary complex UedtaX...480 Table VIII-32: Experi mental equilibrium data for the U(VI) edta system. ...484 Table VIII-33: Equilibri um data for the Np(II I) edta system. ...491 Table VIII-34: Experi e Np(IV) edta system. ...491 mental equilibrium data for th Table VIII-35: Experi mental equilibrium data for the Np(V) edta system. ...492 Table VIII-36: Experi mental equilibrium data for th e Pu(III) edta system. ...499 Table VIII-37: Experi mental equilibrium data for th e Pu(IV) edta system. ...501 the Pu(V) edta system. ...502 Table VIII-38: Experi mental equilibrium data for Table VIII-39: Experi mental equilibrium data for th e Pu(VI) edta system. ...502 mental equilibrium data for Table VIII-40: Experi the Am(III) edta system...504 Table VIII-41: Accepted forma tion constants for the Am edta system to derive the se- lected values. ...505 mental equilibrium data for Table VIII-42: Experi the Am(V) edta system...508 Table VIII-43: Expe rimental enthalpy data for the Am edta system. ...509 Table IX-1: Lactonisation constant of is osaccharinic acid in the literature. ...514 Table IX-2: Protonation constants of isos accharinic acid in the literature. ...515 Table IX-3: isosaccharinate compounds in the litera- Solubility constants of calcium ture...521 Table IX-4: Stability constants of calcium isosaccharinate complexes in the litera- ture...522 Table IX-5: Stability constants of isa complexes with Ni(II) and Cu(II) in the litera- ture...525 Table IX-6: Stability constants of U(IV) - isa complexes in the literature [2004WAR/EVA] ...527 Table IX-7: Stability constants of U(VI) - isa complexes in the literature [2004RAO/GAR] . ...528

39 LIST OF TABLES xli Table IX-8: Enthalpy and entropy of U(VI) isa complexation in the literature . ...529 [2004RAO/GAR] isosaccharinate complexe s in the literature. Table IX-9: Stability constants of Np(IV) ...530 s of Pu(IV) isosaccharinate complexes in the literature. Table IX-10: Stability constant ...532 lexes with Ln(III) in the literature...534 Table IX-11: Stability constants of isa comp (III) isa complexes in the literature Table IX-12: Stability constants of Fe [2004RAO/GAR] . ...538 Table IX-13: Enthalpy of Fe(III) isa complexation in the literature [2004RAO/GAR] . ...538 Table IX-14: Stability constants of Th(IV) isosaccharinate complexe s in the literature. ...539 Table A-1 Solubility of UO (ox)·3H O in water at variable temperatures. ...570 2 2 Calculated solubility products Table A-2 fined by reactions for U(IV)/oxalate de (A.13), (A.14) and (A.15)...579 Table A-3 Solubility constants for H ionic strengths. ...591 edta(cr) at different 4 log K Corrections of Table A-4 [59MOS/ZAK] for the complexa- (A.27) from s 10 tion of U(VI) with nitrate. ...596 Table A-5 Solubility constants for Np(IV) oxalate in hydrochloric acid medium [60KON/GEL] ...602 Table A-6 Solubility constants of H ionic strengths...618 edta(cr) at different 4 Table A-7 Equilibrium constants reported in [64BAN/SHA] for reactions: 42 − n 4+ 2– 4+ Np(ox) Np n ox O(cr) + + , and Np(ox) ·6H U U Np 2 2 n 2– 2 ox O(l). ...626 + 6H 2 Equilibrium constants for reactions (A.51) obtained in this review by Table A-8 fitting the solubility data reported in [64BAN/SHA] . ...628 Table A-9 Citrate protonation constants and U(VI) citrate formation constants as reported in [65RAJ/MAR] ...637 Table A-10 Accepted U(VI) c itrate formation constants w ith uncertainties estimated by this review. ...637 Table A-11 Equilibrium constants at zero ionic strength and 25°C from [70DEY/PEN] . ...698 , [75DEY/CAN]

40 LIST OF TABLES xlii Table A-12 fined by reactions Calculated solubility products for U(IV)/oxalate de (A.73), (A.74) and (A.75)...726 at Stability constants for the Np(V)–oxalate system in 1.0 M NaClO Table A-13 4 [92TOC/INO] 25°C ...776 Protonation constants obtained by [94ERT/MOH] Table A-14 [96CHO/CHE] and ...780 for the protonation of c itrate in 5 m NaCl. ...784 Literature data Table A-15 2– 4– Table A-16 Literature data for the protonation of ox and edta . ...784 Corrections of log Table A-17 for from [96BOR/LIS] and [2001BOR/MOO] β exp. 10 the complexation of U(VI) with chloride. ...786 Citrate equilibrium constants and SIT interaction parameters as reported Table A-18 in [96CHO/ERT] . ...787 Table A-19 Values of H ∆ and obtained in this review by fitting the origi- ∆ C rm p r,m [97BEN/PAL] nal equilibrium constants of at various temperatures as- suming that was constant...789 C ∆ r,m p quantities for the first and second protonation reaction Table A-20 Thermodynamic of oxalic acid ...793 Calculated hydrogen ion activities in the system H Table A-21 ox – Na SeO and 2 2 3 + experimental [H [98OBR/MIT] ...796 ∆ ] values from 3– + Equilibrium constants and correction factors for Na(edta) Table A-22 Cl and , UO 2 UO ted in this review...798 (aq) complexation as evalua Cl 2 2 Correction of the original data of [98POK/BRO] Table A-23 for the equilibrium 2– 4– 2+ + edta UO U UO edta ...798 2 2 Table A-24 Correction of the original data of [98POK/BRO] for the equilibrium 3– 2+ – . ...799 UO + Hedta U UO (Hedta) 2 2 Table A-25 Correction of the original data of [98POK/BRO] for the equilibrium + 3– 4– NpO + edta U NpO ...799 edta 2 2 Correction of the original data of [98POK/BRO] Table A-26 for the equilibrium + 2– 4– + + edta + H U NpO (Hedta) NpO . ...800 2 2 [98POK/BRO] Table A-27 Correction of the original data of for the equilibrium – + 4– + + edta N ...800 U NpO pO + 2H (H edta) 2 2 2 3– 2+ Table A-28 Equilibrium constants and correction factors for Na(edta) , AmCl and + AmCl complexation as evaluated in this review. ...802 2

41 LIST OF TABLES xliii for the equilibrium Correction of the original data of Table A-29 [99CHE/CHO] – 3+ 4– Am + edta ...803 U Am(edta) Correct list of experimental parameters for the Mg edta data in Table A-30 [2001CHO/BON] . ...810 Stability constant Table A-31 with M = Mg, Ca and Ni. s for the Reaction (A.116) ...812 Water activities Table B-1: for the most common ionic media at various con- a HO 2 raction approach and the interaction centrations applying Pitzer’s ion inte [91PIT] parameters given in . ...824 Table B-2: Debye–Hückel constants as a function of temperature at a pressure of 1 bar below 100°C and at the steam saturated pressure for t ≥ 100°C. 828 Table B-3: The preparation of the experimental equilibrium constants for the ex- trapolation to I = 0 with the specific ion interaction method at 25°C and Reaction (B.1 2)...829 1 bar, according to –1 – Ion interaction coefficients ( , ) jk Table B-4: , ) for cations j with k = Cl ε (kg·mol − − and N O [80CIA] , taken from Ciavatta ClO , [88CIA] unless indi- 4 3 cated otherwise. ...835 –1 + Table B-5: Ion interaction coefficients ( , ) jk (kg·mol k ) for anions j with , = Li ε + + Na , taken from Ciavatta , [88CIA] unless indicated oth- and K [80CIA] erwise. ...842 Ion interaction coefficients (1, , ) jk Table B-6: and (2, , ) jk ε for cations j with k = ε + – + − − Cl and N O ClO (first part), and for anions j with k = Li , , Na and 4 3 + K ε = ε . The + ε I (second part), according to the relationship log 10 2 m 1 data are taken from Ciavatta [80CIA] , unless indicated other- [88CIA] wise ...846 –1 SIT interaction coefficient ( j,k ) kg · mol Table B-7: ε for neutral species, j, with k, electroneutral combina tion of ions. ...847 Table C-1: Details of the calculation of equilibrium constant corrected to I = 0, us- ing (C.19) ...861

42

43 Part I Introductory material 1

44

45 Chapter I I Introduction I.1 Background The modelling of the behaviour of hazardous materials under environmental conditions is among the most important applications of natural and technical sciences for the pro- for example, the safety of a waste deposit, tection of the environment. In order to assess, it is essential to be able to predict the ev entual dispersion of its hazardous components r hazardous materials st ored in the ground in the environment (geosphere, biosphere). Fo or in geological formations, the most probable transport medium is the aqueous phase. An important factor is therefore the quant itative prediction of the reactions that are likely to occur between hazardous waste dissolved or suspended in ground water, and the surrounding rock material, in order to estimate the quantities of waste that can be transported in the aqueous phase. It is thus essential to know the relative stabilities of the compounds and complexes that may form under the relevant conditions. This infor- mation is often provided by speciation calculations using chemical thermodynamic data. The local conditions, such as ground water and rock composition or temperature, may not be constant along the migration paths of hazardous material s, and fundamental thermodynamic data are the indispensable basis for dynamic modelling of the chemical behaviour of hazardous waste components. In the field of radioactive waste mana gement, the hazardous material consists to a large extent of actinides and fission products from nucl ear reactors, in addition to lesser amounts from other sources such as waste from medicine, industry and research facilities. The scientific literature on thermodynamic data, mainly on equilibrium con- stants and redox potentials in aqueous solution, has been contradictory in a number of cases, especially in actinide mprehensive review of the chemistry. A critical and co available literature is necessary in order to establish a reliable thermochemical database that fulfils the requirements for rigorous modelling of the behaviour of the actinide and fission products in the environment. The International Atomic Energy Agency (IAEA) in Vienna published special issues with compilations of physicochemical properties of compounds and alloys of elements important in reactor technology: Pu, Nb, Ta, Be, Th, Zr, Mo, Hf and Ti be- started the publication of the series “The tween 1966 and 1983. In 1976, IAEA also Chemical Thermodynamics of Actinide Elements and Compounds”, oriented towards nuclear engineers and scientists. This international effort has resulted in the publication 3

46 I Introduction 4 ermodynamic properties of a given type of of several volumes, each concerning the th compounds for the entire actinide series. Th ese reviews cover the literature approxi- mately up to 1984. The latest volume in this series appeared in 1992, under Part 12: The [92FUG/KHO] Actinide Aqueous Inorganic Complexes . Unfortunately, data of impor- tance for radioactive waste management (for example, Part 10: The Actinide Oxides) is lacking in the IAEA series. Committee (RWMC) of the OECD Nu- The Radioactive Waste Management ed the need for an internationally acknowledged, high- clear Energy Agency recognis quality thermochemical database for applicati on in the safety assessment of radioactive waste disposal, and undertook the development of the NEA Thermochemical Data Base (TDB) project [85MUL] [88WAN] , [91WAN] , . The RWMC assigned a high priority to the critical review of relevant chemical thermodynamic data of compounds and com- es uranium, neptunium, plutonium and ameri- plexes for this area containing the actinid cium, as well as the fission product technetium. The first four books in this series on the chemical thermodynamics of uranium [92GRE/FUG] , americium [95SIL/BID] , techne- [99RAR/RAN] tium and neptunium and plutonium originated from [2001LEM/FUG] this initiative. Simultaneously with the NE A TDB project, other reviews on the physical and chemical properties of actinides ap peared, including the book by Cordfunke et al. [90COR/KON2] , the series edited by Freeman et al. [84FRE/LAN] , [85FRE/LAN] , [85FRE/KEL] , , [87FRE/LAN] , [91FRE/KEL] , the two volumes edited [86FRE/KEL] et al. [86KAT/SEA] by Katz et al. [92FUG/KHO] , and Part 12 by Fuger within the IAEA review series mentioned above. In 1998, Phase II of the TDB Project (TDB-II) was started to provide for the further needs of the radioactive waste management programs by updating the existing database and applying the TDB review methodology to other elements (nickel, sele- nium, zirconium) and to simple organic compounds and complexes. In TDB-II the overall objectives are set by a Management Boar d, integrated by the representatives of 17 organisations from the field of radioactive waste management. These participating organisations, together with the NEA, provide financial support for TDB-II. The TDB II - Management Board is assisted in technical matters by a group of experts in chemical thermodynamics (the Executive Group). Th e NEA acts in this phase as Project Co- ordinator ensuring the implementation of th e Project Guidelines and liaising with the Review Teams. After the update on the chem ical thermodynamics of uranium, neptu- nium, plutonium, americium and technetium [2003GUI/FAN] , and the books on chemi- and zirco- [2005OLI/NOL] , selenium [2005GAM/BUG] cal thermodynamics of nickel nium [2005BRO/CUR] , the present volume is the fifth to be published within this sec- ond phase of the TDB project.

47 I.2 Focus of the review 5 I.2 Focus of the review This NEA TDB organics review is within the scope and the spirit of the previous re- views aimed at helping to model the chemical behaviour of actinides and fission and activation products in the near and far field of a radioactive waste repository using con- sistent data. The present critical review deal s with compounds and complexes of U, Np, Pu, Am, Tc, Se, Ni, Zr as well as H, Na, K, Mg, Ca with oxalate, citrate, ethylenediami- netetraacetate (edta) and iso-saccharinate (isa). A prerequisite for a successful widely recognized critical evaluation of ther- modynamic data is the sensible decision which organic ligands the evaluation should mely the importance of the ligands in radio- comprise. This decision has two aspects, na active waste problems, and the availability of experimental data. The “Agreement of the OECD/NEA Thermochemical Data Base Project (Phase II)” states in Annex I that the ed to oxalate, citrate, edta and isa. evaluation of organic ligands should be limit From the viewpoint of importance for radioactive waste problems this set of ligands is very well posed. Oxalate is, with respect to its complexation strength, the ma- jor product of radiolytic degradation of bitumen, sometimes used for waste condition- ing, and ion exchange resins used in decontamination procedures. In addition, oxalate is one of the strong complexing natural organic ligands (besides humic substances). Cit- rate and edta are used in decontamination processes and thus, they become part of the radioactive waste inventory. In terms of comp lexation strength, oxalate, citrate and edta cover a wide range of complex stability, and they may be used in model calculations as representatives of dicarboxylic acids (oxala te), hydroxy-polycarboxylic acids (citrate) and polyamino-polycarboxylic acids (edta). Finally, from the viewpoint of complexa- tion strength, isa is the most important product of alkaline degradation of cellulose in cement pore waters. Thus, isa is of major concern in many performance assessments of planned radioactive waste repositories. Regarding the availability of experimental data, the situation is less clear. In the case of oxalate, citrate and edta a large body of experimental studies has been pub- lished and this review provides, based on the critical discussion of several hundreds of publications, a considerable set of selected thermodynamic values. However, in the case of isa the number of experimental studies is very limited and, despite of the importance of isa for performance assessments, only a few thermodynamic values could be selected. The critical review of experimental studies concerning isa mainly is a status report pointing out gaps in our present knowledge and further research needs. As the task of the present review on or ganic ligands is to complement the other reviews of the NEA TDB project, which are restricted to inorganic compounds and complexes of actinides and fission and activation products, a natural choice of elements comprises U, Np, Pu, Am, Tc, Se, Ni and Zr. However, this review cannot be restricted to these elements as it aims at a thermodynami c data set useful for practical application. In addition to the above mentioned actinides and fission and activation products this

48 I Introduction 6 of ground and surface waters which may review considers also the major constituents ., H, Na, K, Mg and Ca. Any geochemical interact with the selected organic ligands, i.e model including organic ligands should take into account these competing interactions ew provides a selected consistent set of these auxiliary and therefore, the present revi constants. + The protonation equilibria (complexation with H ) of the organic ligands be- long to the basic data of these systems and have been, in general, very well studied. They are considered in the present review as the main auxiliary data for the critical evaluation of metal – ligand complexation in the review itself, and as the most impor- tant auxiliary data for practical application. The situation is more complicated for Na and K interactions. Here, we are in the realm of ambiguity charact eak complexation or strong erised by the question: “W apters V, VI, VII and VIII the answer is specific ion interaction?”. As discussed in Ch different for different ligands. In the case of oxalate, Na and K interactions are treated solely as specific ion interactions. In the case of edta, Na and K complexation constants Na and K interactions with citrate are in had to be included in the speciation models. the review process both variants have been between these two extremes. In the course of evaluated and compared, and finally the review team decided to treat the effects of Na and K on citrate solely as rather strong specific ion interactions. The elements Ca and Mg are included in the present review because complexa- tion of the organic ligands with these competing cations is of importance in geochemi- cal models concerning the complexation of actinides and fission products in common ground and surface waters. Further auxiliary data are taken both from the publication of CODATA key values [89COX/WAG] and from the evaluation of additional auxiliary data in this series , , [99RAR/RAN] , [2003GUI/FAN] of NEA TDB reviews [92GRE/FUG] [2005GAM/BUG] and [2005OLI/NOL] and their use is recommended by this review. Care has been taken that all the selected thermodynamic data at standard state and conditions ( cf. Section II.3) and 298.15 K are internally consistent. For this purpose, special software within the NEA TDB database system has been used; cf. Section II.6. In order to maintain consistency in the application of the values selected by this review, it is essential to use the NEA TDB selected and auxiliary data when calculating equilib- rium constants involving actinide and fission product compounds and complexes. I.3 Review procedure and results The literature has been surveyed up to the end of 2001 for all ligands. For oxalate a few more recent references are included, and for is a the aim had been to consider all relevant literature up to the end of 2004. Experimental measurements published in the scientific

49 I.3 Review procedure and results 7 literature are the main source for the selection of recommended data. Previous reviews are not neglected. They have been primarily used as sources for original scientific litera- ture, but they also form a valuable source of critical information on the quality of pri- mary publications. In the realm of metal – organic complexes a plethora of experimental studies is found in the literature dealing with mixed complexes, i.e ., complexes containing a com- mon metal ion and two or more different ligands. In this review mixed complexes, in general, were considered if they contain co mbinations of oxalate, citrate, edta and isa with or without additional inorganic ligands. Mixed complexes comprising other or- ganic ligands are mentioned occasionally fo r qualitative comparis on only. From the viewpoint of application, by far the most important class of mixed complexed are ter- nary metal – hydroxide – organic ligand complexes. These hydrolysed organic com- plexes may predominate in alkaline ground and surface waters and in high pH cement pore waters and thus, they are important in assessing the influence of organic ligands on radionuclide complexation in cementitious repositories. The relevant literature about such complexes is discussed in the present re view, but only in a few cases reliable ther- modynamic constants could be selected. Also of importance in many ground and sur- face waters would be the class of metal – carbonate – organic ligand complexes. How- most complete lack of such data. ever, this review can only state the al The detailed discussion of organic compounds has been restricted in this re- view to the so-called “sparingly soluble” solids. These are mainly metal oxalates, which are considered in several sec tions of Chapter VI. The generally rather soluble citrate and edta compounds are discussed in qualitative terms only in Chapters VII and VIII. From the viewpoint of model application in performance assessments the most important solid is calcium oxalate, because the possible precip itation of this solid in many ground and dissolved oxalate to rather low levels. surface waters can limit the concentration of When necessary, experimental source data are re-evaluated by using chemical models that are either found to be more realistic than those used by the original author, or are consistent with side-reactions discussed in another section of the review (for ex- ample, data on metal complex formation might need to be re-interpreted to take into on reactions). Re-evaluation of literature account consistent values for ligand protonati r known systematic eff ects (for example, if values might also be necessary to correct fo metal – chloride complexation has been neglected in the original literature) or to make extrapolations to standard state conditions ( I = 0) by using the specific ion interaction (SIT) equations ( cf. Appendix B). For convenience, these SIT equations are referred to in some places in the text as “the SIT”. In order to ensure that consistent procedures are used for the evaluation of pri- mary data, a number of guidelines have been developed. They have been updated and improved since 1987, and their most r ecent versions are available at the NEA [2000OST/WAN] , . , [2000WAN/OST] , [99WAN] , [99WAN/OST] [2000GRE/WAN]

50 I Introduction 8 cf. Chapter II, Appendix B Some of these procedures are also outlined in this volume, and Appendix C. Particular problems encounte red in this review on organic ligands are discussed in detail in Chapter V. Once the critical review process in the NEA TDB project is completed, the resulting manuscript is reviewed independently by qualified experts nominated by the ocedures outlined in NEA. The independent peer review is perfor med according to the pr the TDB-6 guideline [99WAN] peer review is to receive . The purpose of the additional ssessments made by the primary reviewers, an independent view of the judgments and a to verify assumptions, results and conclusions, and to check whether the relevant litera- ture has been exhaustively considered. The independent peer review is performed by ect matter to be reviewed, to a degree at persons having technical expertise in the subj least equivalent to that needed for the original review. The thermodynamic data selected in th e present review (see Chapters III and cf. Sec- IV) refer to the reference temperature of 298.15 K and to standard conditions, tion II.3. For the modelling of real systems it is , in general, necessary to recalculate the standard thermodynamic data to non-standard conditions. For aqueous species a proce- dure for the calculation of the activity factors is thus required. This review uses the ap- proximate specific ion interaction method (SIT) for the extrapolation of experimental data to the standard state in the data evaluation process. For maximum consistency, this method, as described in Appendix B, should always be used in conjunction with the selected data presented in this review. The thermodynamic data selected in this review are provided with uncertainties representing the 95% confidence level. As disc ussed in Appendix C, there is no unique way to assign uncertainties, and the assignments made in this review are to a large ex- tent based on the subjective choice by the re viewers, supported by their scientific and technical experience in the corresponding area. The quality of thermodynamic models cannot be better than the quality of the data on which they are based. The quality aspect includes both the numerical values of and the “completeness” of the chemical the thermodynamic data used in the model model used, e.g. , the inclusion of all relevant dissolved chemical species and solid phases. For the user it is important to consider that the selected data set presented in this review (Chapters III and IV) is certainly not “complete” with respect to all the conceiv- able systems and conditions; there are gaps in the information. The gaps are pointed out in the various sections of Chapters VI, VII, VIII and IX, and this information may be used as a basis for the assignment of research priorities.

51 Chapter II II Standards, Conventions and Contents of the Tables Equation Section 2 This chapter outlines and lists the symbols, terminology and nomenclature, the units and conversion factors, the order of formulae, the standard conditions, and the fundamental physical constants used in this volume. They are derived from international standards ed for the TDB publications. and have been specially adjust II.1 Symbols, terminology and nomenclature II.1.1 Abbreviations Abbreviations are mainly used in tables wh ere space is limited. Abbreviations for meth- ods of measurement are listed in Table II-1. Table II-1: Abbreviations for experimental methods. aix Anion exchange Atomic Emission Spectroscopy AES cal Calorimetry chr Chromatography cix Cation exchange col Colorimetry con Conductivity cou Coulometry cry Cryoscopy dis Distribution between two phases DSC Differential Scanning Calorimetry DTA Differential Thermal Analysis EDS Energy Dispersive Spectroscopy em Electromigration emf Electromotive force, not specified (Continued on next page) 9

52 10 II Standards, Conventions and Contents of the Tables Table II-1: (continued) EPMA Electron Probe Micro Analysis Extended X-ray Absorption Fine Structure EXAFS FTIR Fourier Transform Infra Red Isotope Dilution Mass-Spectroscopy IDMS ir Infrared gl Glass electrode Ion selective electrode with ion X stated ise-x ix Ion exchange kin Rate of reaction Laser Induced Breakdown Detection LIBD MVD Molar Volume Determination Nuclear Magnetic Resonance NMR PAS Photo Acoustic Spectroscopy pol Polarography pot Potentiometry prx Proton relaxation qh Quinhydrone electrode red Emf with redox electrode Scanning Electron Microscopy SEM sp Spectrophotometry sol Solubility TC Transient Conductivity TGA Thermo Gravimetric Analysis TLS Thermal Lensing Spectrophotometry TRLFS Time Resolved Lase r Fluorescence Spectroscopy UV Ultraviolet vlt Voltammetry XANES X-ray Absorption Near Edge Structure XRD X-ray Diffraction ? Method unknown to the reviewers Other abbreviations may also be used in tables, such as SHE for the standard hydrogen electrode or SCE for the saturate d calomel electrode. The abbreviation NHE has been widely used for the “normal hydrogen electrode”, which is by definition iden- tical to the SHE. It should nevertheless be noted that NHE customarily refers to a stan- dard state pressure of 1 atm, whereas SHE alwa ys refers to a standard state pressure of 0.1 MPa (1 bar) in this review.

53 11 II.1 Symbols, terminology and nomenclature II.1.2 Symbols and terminology The symbols for physical and chemical quantities used in the TDB review follow the recommendations of the International Union of Pure and Applied Chemistry, IUPAC [79WHI] , [93MIL/CVI] . They are summarised in Table II-2. Table II-2: Symbols and terminology. Symbols and terminology l length height h radius r diameter d volume V mass m ρ density (mass divided by volume) time t frequency ν wavelength λ r e medium itself, disregarding boundary o internal transmittance (transmittance of th T container influence) ) /T A (1 , (decadic abso internal transmission density rbance): log 10 i A cl ε / molar (decadic) absorption coefficient: B τ relaxation time Avogadro constant N A (a) relative molecular mass of a substance M r thermodynamic temperature, absolute temperature T t Celsius temperature (molar) gas constant R Boltzmann constant k Faraday constant F S (molar) entropy m C (molar) heat capacity at constant pressure ,m p H (molar) enthalpy m G (molar) Gibbs energy m μ chemical potential of substance B B pressure p p p partial pressure of substance B: x B B fugacity of substance B f B (Continued next page)

54 12 II Standards, Conventions and Contents of the Tables Table II-2 (Continued) Symbols and terminology γ f / fugacity coefficient: p f,B B B (b) n amount of substance x mole fraction of substance B: B molarity or concentration of a solute substance B (amount of B divided by the , [B] c B (c) volume of the solution) (d) of B divided by the mass of the solvent) molality of a solute substance B (amount m B m factor for the conversion of mola rity to molality of a solution: / c Š B B νν ) νν ( + (e) ++ −− mean ionic molality m = , mmm ± ±+− activity of substance B a B γ / am activity coefficient, molality basis: B BB y / ac activity coefficient, concentration basis: B BB νν + νν ( ) (e) ++ −− , == aaaa a mean ionic activity ± ±+− B νν ) ( ν + ν (e) ++ − − , y γ=γγ mean ionic activity coefficient ± ±+− φ osmotic coefficient, molality basis 22 11 or I == cz I mz I ionic strength: ∑∑ ii ii c ii m 22 SIT ion interaction coefficient between substance B ) and substance B 1 2 (B , B ) ε 12 , positive for products (negative for reactants stoichiometric coefficient of substance B ν B ν B = general equation for a chemical reaction 0 ∑ B B (f) K equilibrium constant e for cations, negative for anions) charge number of an ion B (positiv z B charge number of a cell reaction n E electromotive force 1 − pH pH = log [ /(mol kg )] a −⋅ + 10 H κ electrolytic conductivity . (g) ° superscript for standard state 1 12 C. of nuclide (a) ratio of the average mass per formula unit of a substance to of the mass of an atom 12 . (b) . sections 1.2 and 3.6 of the IUPAC manual [79WHI] cf . A solu- [79WHI] nce concentration” in the IUPAC manual (c) This quantity is called “amount-of-substa 3 − mol dm ⋅ is called a 0.1 molar solution or a 0.1 M solution. tion with a concentration equal to 0 . 1 1 − mol kg ⋅ is called a 0.1 molal solution or a 0.1 m solution. A solution having a molality equal to 0.1 (d) ν ions, in an aqueous solution with =+ ) νν which dissociates into ( For an electrolyte (e) NX + ± − νν + − and = mm ) ν ( and activity coefficient are m molality , the individual cationic molality + + () / am = . A similar definition is used for the anioni c symbols. Electrical neutrality requires that γ + ++ . νν zz = −− ++ K Special notations for equilibrium constants ar e outlined in Section II.1.6. In some cases, is used to (f) c K a constant in molal units. indicate a concentration constant in molar units, and m (g) See Section II.3.1.

55 13 II.1 Symbols, terminology and nomenclature II.1.3 Chemical formulae and nomenclature [71JEN] , [77FER] , [90LEI] This review follows the recommendations made by IUPAC on the nomenclature of inorganic compounds and complexes, except for the following items: • The formulae of coordination compounds and complexes are not enclosed in square brackets [71JEN] (Rule 7.21). Exceptions are made in cases where square brackets are required to distinguish between coordinated and uncoordi- nated ligands. • e.g. , The prefixes “oxy–” and “hydroxy–” are retained if used in a general way, “gaseous uranium oxyfluorides”. For specific formula names, however, the [71JEN] IUPAC recommended citation e.g. , “uranium(IV) (Rule 6.42) is used, difluoride oxide” for UF O(cr). 2 An IUPAC rule that is often not followed by many authors [71JEN] (Rules 2.163 and 7.21) is recalled here: the order of arranging ligands in coordination com- pounds and complexes is the following: central atom first, followed by ionic ligands and then by the neutral ligands. If there is more than one ionic or neutral ligand, the alpha- betical order of the symbols of the ligating atoms determines the sequence of the − − (UO ) (OH) CO (UO ) CO (OH) ligands. For example, is non- is standard, 3 3 3 3 22 22 standard and is not used. Abbreviations of names for organic ligands appear sometimes in formulae. Following the recommendations by IUPAC, lower case letters are used, and if neces- sary, the ligand abbreviation is enclosed within parentheses. Hydrogen atoms that can be replaced by the metal atom are shown in the abbreviation with an upper case “H”, for − Hedta , example: Am(Hedta)(s) (where edta stands for ethylenediaminetetraacetate). 3 II.1.4 Phase designators Chemical formulae may refer to different chem ical species and are often required to be specified more clearly in order to avoid ambiguities. For example, UF occurs as a gas, 4 a solid, and an aqueous complex. The distinction between the different phases is made the chemical formula and appear in paren- by phase designators that immediately follow theses. The only formulae that ase designator are aqueous are not provided with a ph ions. They are the only charged species in this review since charged gases are not con- sidered. The use of the phase designators is described below. The designator (l) is used for pure liquid substances, e.g. , H • O(l). 2 • The designator (aq) is used for und issociated, uncharged aqueous species, e.g. , U(OH) (aq), CO (aq). Since ionic gases are not considered in this review, all 4 2 ions may be assumed to be aqueous and are not designed with (aq). If a chemi- cal reaction refers to a medium other than D O , 90% ethanol/10% , HO ( e.g. 2 2

56 14 II Standards, Conventions and Contents of the Tables O), then (aq) is replaced by a more explicit designator, , “(in H DO)” or e.g. 2 2 “(sln)”. In the case of (sln), the composition of the solution is described in the text. The designator (sln) is used for substances in solution without specifying the • actual equilibrium composition of the substance in the solution. Note the dif- ference in the designation of H O(l) in Reaction H O in Eqs.(II.2) and (II.3). 2 2 (II.2) indicates that i.e. , no solutes are present, H O is present as a pure liquid, 2 whereas Reaction (II.3) involves an HCl solution, in which the thermodynamic properties of H O(sln) may not be the same as those of the pure liquid H O(l). 2 2 In dilute solutions, however, this difference in the thermodynamic properties of H O can be neglected, and H O(sln) may be regarded as pure H O(l). 2 2 2 Example: UO Cl (cr) + 2 HBr(sln) UOBr (cr) + 2HCl(sln) U (II.1) 2 22 U ⋅⋅ UO Cl H O(cr) + 2 H O(l) UO Cl 3H O(cr) (II.2) 22 2 2 22 2 UO ( γ UO Cl (cr) + H O(sln) U (II.3) ) + 2 HCl(sln) 3222 • The designators (cr), (am), (vit), and (s) are used for solid substances. (cr) is used when it is known that the compound is crystalline, (am) when it is known that it is amorphous, and (vit) for glassy substances. Otherwise, (s) is used. • In some cases, more than one crystalline form of the same chemical composi- tion may exist. In such a case, the different forms are distinguished by separate designators that describe the forms more precisely. If the crystal has a mineral name, the designator (cr) is replaced by th e first four characters of the mineral name in parentheses, e.g. SiO (quar) for quartz and , SiO (chal) for chalced- 2 2 ony. If there is no mineral name, the desi gnator (cr) is replaced by a Greek let- ter preceding the formula and in dicating the structural phase, , , -UF α e.g. 5 -UF β . 5 Phase designators are also used in conjunction with thermodynamic symbols to define the state of aggregation of a compound to which a thermodynamic quantity re- fers. The notation is in this case the same as outlined above. In an extended notation ( cf . [82LAF] ) the reference temperature is usually given in addition to the state of aggrega- tion of the composition of a mixture. Example: + ο G (Na , 298.15 K) ∆ standard molar Gibbs energy of forma- fm + at 298.15 K tion of aqueous Na ο (UO (SO ) 2.5H O, cr, 298.15 K) S ⋅ standard molar entropy of 2 m24 UO (SO ) 2.5H O(cr) ⋅ at 298.15 K 24 2

57 II.1 Symbols, terminology and nomenclature 15 ο (UO , , 298.15 K) C α standard molar heat capacity of 3 ,m p UO at 298.15 K α − 3 (HF, sln, HF 7.8H O) H ∆⋅ enthalpy of formation of HF diluted fm 2 1:7.8 with water. II.1.5 Processes Chemical processes are denoted by the operator ∆ , written before the symbol for a [82LAF] property, as recommended by IUPAC . An exception to this rule is the equilib- rium constant, . Section II.1.6. The nature of the process is denoted by annotation of cf ∆ G ∆ H the ∆ e.g. , the enthalpy of sublimation, , , the Gibbs energy of formation, , sub m fm . The abbreviations of chemical pro cesses are summarised in Table II-3. etc Table II-3: Abbreviations used as subscripts of to denote the type of chemical process. ∆ ∆ Subscript of Chemical process separation of a substance into its at constituent gaseous atoms (atomisation) dehyd elimination of water of hydration (dehydration) dilution of a solution dil f formation of a compound from its constituent elements fus melting (fusion) of a solid hyd addition of water of hydration to an unhydrated compound mixing of fluids mix r chemical reaction (general) process of dissolution sol ation) of a solid sub sublimation (evapor tr transfer from one solution or liquid phase to another trs transition of one solid phase to another vap vaporisation (evapor ation) of a liquid ∆ ∆ H G the Gibbs and , The most frequently used symbols for processes are f f formation of a compound or complex from the elements in energy and the enthalpy of their reference states ( . Table II-6). cf II.1.6 Equilibrium constants The IUPAC has not explicitly defined the symbols and terminology for equilibrium constants of reactions in aqueous solution. The NEA has therefore adopted the conven- tions that have been used in the work Stability Constants of Metal Ion Complexes by Sillén and Martell [64SIL/MAR] , [71SIL/MAR] . An outline is given in the paragraphs below. Note that, for some simple reactions, there may be different correct ways to in- be preferable to indicate the number of dex an equilibrium constant. It may sometimes

58 16 II Standards, Conventions and Contents of the Tables ally in cases where several ligands are dis- the reaction to which the data refer, especi cussed that might be confused. For example, for the equilibrium: U (II.4) M L M + L mq mq (II.4) is accepted, too. Note that, and β (II.4) would be appropriate, and both β β qm , qm , epwise formation constant, and is used for the consecutive or st is used K β in general, for the cumulative or overall formation constant. In the following outline, charges are only given for actual chemical species, but ar e omitted for species containing general symbols (M, L). II.1.6.1 Protonation of a ligand ⎡⎤ HL r ⎣⎦ + = K HL U H + H L (II.5) 1 rr − 1, r + ⎡ ⎤⎡ ⎤ HHL 1 − r ⎦ ⎣ ⎦⎣ ⎡ ⎤ HL r ⎦ ⎣ + = H + L HL r U (II.6) β r 1, r r + ⎤⎡ ⎤ ⎡ HL ⎦⎣ ⎦ ⎣ , but it [64SIL/MAR] This notation has been proposed and used by Sillén and Martell has been simplified later by the same authors [71SIL/MAR] from to K . K r r 1, For the addition of a ligand, the notation shown in Eq.(II.7) is used. ⎡ ⎤ HL q ⎦ ⎣ (II.7). + L HL U HL K = 1 − qq q ⎡⎤ ⎡ ⎤ L HL − 1 & q ⎣ ⎦ ⎣⎦ . rmation constant of the species Eq.(II.8) refers to the overall fo HL rq ⎡ ⎤ HL rq ⎦ ⎣ + β H + L H L (II.8). rq U = qr rq , rq + ⎡ ⎤⎡ ⎤ HL ⎦⎣ ⎦ ⎣ r r can be omitted if In Eqs.(II.5), (II.6) and (II.8), the second subscript = 1, as shown in Eq.(II.7). Example: − 2 ⎡ ⎤ HPO 4 ⎦ ⎣ −− +3 2 = = H + PO HPO ββ U 1,1 1 44 +3 − ⎡ ⎤⎡ ⎤ HPO 4 ⎣ ⎦⎣ ⎦ − ⎤ ⎡ HPO 24 ⎣ ⎦ −− +3 = H PO U β 2 H + PO 424 1, 2 2 +3 − ⎡ ⎤⎡ ⎤ HPO 4 ⎣ ⎦⎣ ⎦

59 II.1 Symbols, terminology and nomenclature 17 Formation of metal complexes II.1.6.2 ⎡⎤ ML q ⎣⎦ ML + L ML U (II.9) = K qq − 1 q ⎤⎡ ⎤ ⎡ L ML − 1 q ⎦⎣ ⎦ ⎣ ⎤ ⎡ ML q ⎦ ⎣ = (II.10) β ML M + L U q q q q ⎤⎡ ⎤ ⎡ ML ⎣ ⎦⎣ ⎦ For the addition of a metal ion, i.e. , the formation of polynuclear complexes, the follow- ing notation is used, analogous to Eq.(II.5): ⎡⎤ ML m ⎣⎦ M + M L (II.11) M L U = K mm − 1 1, m ⎡ ⎤⎡ ⎤ MML 1 − m ⎣ ⎦⎣ ⎦ Eq.(II.12) refers to the overall formation constant of a complex ML. mq ⎤ ⎡ ML mq ⎦ ⎣ = mq (II.12) U β M + L M L qm , mq mq ⎡ ⎤⎡ ⎤ ML ⎣ ⎦⎣ ⎦ i.e. The second index can be omitted if it is equal to 1, , β β if becomes qm , q m 1. The formation constants of mixed ligand complexes are not indexed. In this = case, it is necessary to list the chemical reactions considered and to refer the constants to the corresponding reaction numbers. It has sometimes been customary to use negative values for the indices of the − protons to indicate complexation with hydroxide ions, . This practice is not OH − adopted in this review. If OH occurs as a reactant in the not ation of the equilibrium, it is treated like a normal ligand L, but in general formulae the index variable n is used . If H O occurs as a reactant to form hydroxide complexes, q H O is consid- instead of 2 2 e reaction is treated as described below in ered as a protonated ligand, HL, so that th as the index variable. For convenience, no general form is n Eqs.(II.13) to (II.15) using L . In many ex- H used for the stepwise constants for the formation of the complex M r q m periments, the formation constants of metal ion complexes are determined by adding a ligand in its protonated form to a metal ion solution. The complex formation reactions thus involve a deprotonation reaction of the ligand. If this is the case, the equilibrium constant is supplied with an asterisk, as shown in Eqs.(II.13) and (II.14) for mononu- r polynuclear complexes. clear and in Eq.(II.15) fo + ⎤ ⎡ ⎡⎤ ML H q ⎣⎦ ⎦ ⎣ + * ML + H + HL K U ML (II.13) = 1 q − qq ⎡ ⎤ ⎡⎤ ML HL q − 1 ⎣⎦ ⎦ ⎣ q + ⎤⎡ ⎤ ⎡ H ML q ⎦⎣ ⎦ ⎣ * + = qq U (II.14) M + HL ML + H β q q q ⎡ ⎤⎡ ⎤ MHL ⎣ ⎦⎣ ⎦

60 18 II Standards, Conventions and Contents of the Tables q + ⎤⎡ ⎤ ⎡ ML H mq ⎦⎣ ⎦ ⎣ + * = M + HL M L + H (II.15) mq q U β , mq qm mq ⎤⎡ ⎤ ⎡ MHL ⎣ ⎦⎣ ⎦ Example: ++ ⎡ ⎤⎡ ⎤ H UO F 2 ⎣ ⎦⎣ ⎦ ** + 2+ + = = U + H β K UO + HF(aq) UO F 11 22 2+ ⎡ ⎤⎡ ⎤ HF(aq) UO 2 ⎣ ⎦⎣ ⎦ 5 ++ ⎤⎡ ⎤ ⎡ (UO ) (OH) H 23 5 ⎦⎣ ⎦ ⎣ + * 2+ + = β 3 UO + 5 H O(l) (UO ) (OH) + 5 H U 22 235 5,3 3 2+ ⎡⎤ UO 2 ⎣⎦ Note that an asterisk is only assigned to the formation constant if the proto- nated ligand that is added is deprotonated during the reaction. If a protonated ligand is added and coordinated as such to the metal ion, the asterisk is to be omitted, as shown in Eq.(II.16). ⎤ ⎡ M(H L) rq ⎦ ⎣ = q β (II.16) U M(H L) M + H L rrq q q ⎡ ⎤⎡ ⎤ MHL r ⎣ ⎦⎣ ⎦ Example: − ⎡ ⎤ UO (H PO ) 22 43 ⎣ ⎦ 2+ −− = UO + 3 H PO β U UO (H PO ) 224 2243 3 3 2+ − ⎡ ⎤⎡ ⎤ H PO UO 224 ⎣ ⎦⎣ ⎦ Solubility constants II.1.6.3 nstants involving a solid compound are denoted as “sol- Conventionally, equilibrium co ubility constants” rather than as formation constants of the solid. An index “s” to the equilibrium constant indicates that the consta nt refers to a solubility process, as shown in Eqs.(II.17) to (II.19). ab (II.17). ab U M + L ML(s) = M L K [][] ab ,0 s is the conventional solubility product, and the subscript “0” indicates that K s ,0 the equilibrium reaction involv es only uncomplexed aqueou s species. If the solubility constant includes the formation of aqueous complexes, a notation analogous to that of Eq.(II.12) is used: mb q − () mmb ⎛⎞ a ⎡⎤ ⎡⎤ K (II.18). = M L L L M L (s) M L + q − U m q sqm ,, mq ab ⎜⎟ ⎣⎦ ⎣⎦ aa ⎝⎠ Example: − + + − ⎤⎡ ⎤ ⎡ . UO F (cr) U KK = = UO F UO F + F F 2 ,1 ,1,1 22 2 ss ⎣ ⎦⎣ ⎦

61 19 II.1 Symbols, terminology and nomenclature Similarly, an asterisk is added to the solubility constant if it simultaneously in- volves a protonation equilibrium: mb mmb ⎛⎞ ⎛⎞ + −− qq M L (s) + HL M L + H U mq ab ⎜⎟ ⎜⎟ aa a ⎝⎠ ⎝⎠ mb − () q a ⎡⎤ ⎡⎤ ML HL mq ⎣⎦ ⎣⎦ * (II.19) K = ,, sqm mb () − q + a ⎡⎤ H ⎣⎦ Example: − +2+ U(HPO ) 4H O(cr) + H UHPO + H PO + 4 H O(l) ⋅ U 2 4 4 2 42 2 2+ − ⎤ ⎡ ⎤⎡ UHPO H PO 424 ⎦ ⎦⎣ ⎣ ** . = = KK ss ,1,1 ,1 + ⎡⎤ H ⎣⎦ II.1.6.4 Equilibria involving the addition of a gaseous ligand A special notation is used for constants describing equilibria that involve the addition of a gaseous ligand, as outlined in Eq.(II.20). ⎡⎤ ML q ⎢⎥ ⎣⎦ (II.20) K = U ML + L(g) ML p, − 1 q qq ⎡⎤ p ML L 1 − q ⎢⎥ ⎣⎦ The subscript “p” can be combined with any other notations given above. Example: CO (aq) [ ] 2 CO (g) CO (aq) U K = p 22 p CO 2 − 2+ 6 + (UO ) (CO ) 12 H + U 3 UO + 6 CO (g) + 6 H O(l) 22 2 2336 12 6+ − ⎡⎤⎡⎤ H (UO ) (CO ) 23 36 ⎣⎦⎣⎦ * = β p,6,3 3 6 2+ ⎡⎤ UO p 2CO 2 ⎣⎦ − 2+ U + UO (CO ) 2 H UO CO (cr) + CO (g) + H O(l) 2 2 23 2 32 2 2+ − ⎡ ⎤⎡ ⎤ H UO (CO ) 232 ⎣ ⎦⎣ ⎦ * = K p, s, 2 p CO 2 complicated, it is recommended that In cases where the subscripts become or K β be used with or without subscripts, but always followed by the equation number of the equilibrium to which it refers. II.1.6.5 Redox equilibria Redox reactions are usually quantified in term s of their electrode (half cell) potential, E , which is identical to the electromotive force (emf) of a galvanic cell in which the elec-

62 20 II Standards, Conventions and Contents of the Tables 1 dard hydrogen electrode, SHE , in accordance with the “1953 trode on the left is the stan ntials are given as re- . Therefore, electrode pote Stockholm Convention” [93MIL/CVI] drogen electrode, whic h acts as an electron duction potentials relative to the standard hy donor. In the standard hydrogen electrode, is at unit fugacity (an ideal gas at unit H(g) 2 + is at unit activity. The sign of the electrode potential, , is H E pressure, 0.1 MPa), and that of the observed sign of its polarity when coupled with the standard hydrogen elec- ο trode. The standard electrode potential, E , the potential of a standard galvanic cell , i.e. cf. relative to the standard hydrogen electrode (all components in their standard state, Section II.3.1, and with no liquid junction pot ential) is related to the standard Gibbs ο ο G and the standard (or thermodynamic) equilibrium constant K ∆ . as energy change rm outlined in Eq.(II.21): 1R T οο ο −∆ = = ln EGK (II.21) rm FF nn ο and the potential, E E by: , is related to ο EE Tn a −ν (II.22) ln . = (R / F) ∑ ii For example, for the hypothetical galvanic cell: Fe(ClO ) a (aq, = 1) 4 2 2+ Fe (II.23) = 1, (g, p f = 1 1) bar) HCl(aq, = Pt H Pt a + 2 H H 2 Fe(ClO ) (aq, = 1) a 3 4 3+ Fe where denotes a liquid junction and a phase boundary, the reaction is: 1 3+ 2+ + Fe + H (g) Fe + H U (II.24) 2 2 ented by two half cell reactions, each Formally Reaction (II.24) can be repres − ”), as shown in the following involving an equal number of electrons, (designated “ e equations: 2+ 3+ − Fe + e U (II.25) Fe + − 1 (II.26) . U H + e H(g) 2 2 − The terminology is useful, although it must be emphasised “ e ” here does not represent the hydrated electron. − : a Equilibrium (II.26) and Nernst law can be used to introduce e T R ο E faa ln( /( )) (II.27) + E = (II.26) + − H H e 2 F 1 The definitions of SHE and NHE are given in Section II.1.1.

63 21 II.1 Symbols, terminology and nomenclature ο According to the SHE convention E (II.26) = 0, = 1, = 1, hence a f + H H 2 R T ln = Ea (II.28). − − e F This equation is used to calculate a numerical value of from emf meas- a − e vs. the SHE; hence, as for the value of E (V vs. the SHE), the numerical value urements depends on the SHE convention. Equilibrium constants may be written for these a of − e half cell reactions in the following way: a 2+ ο Fe (II.25) = K (II.29) aa ⋅ − 3+ Fe e ⋅ aa + − ο He (II.26) = = 1 (II.30) K (by definition) f H 2 ο ο ο In addition, (II.26) = 0, H ∆ (II.26) = 0 by definition, ∆ (II.26) = 0, ∆ S G rm rm rm ο ο ο G ∆ ∆ ∆ G G From . (II.24) (II.26) and (II.25) = at all temperatures, and therefore rm rm rm + (g) and H , the correspond- the values given at 298.15 K in selected auxiliary data for H 2 – can be calculated to be used in thermodynamic cycles involving half ing values for e cell reactions. The following equations describe the change in the redox potential of a (cf are equal to unity p and . Eq.(II.22)): Reaction (II.24), if + H H 2 ⎛⎞ a 2+ ο Fe E E (II.24) = (II.24) – Rln T (II.31) ⎜⎟ ⎜⎟ a 3+ Fe ⎝⎠ − = a 1 (by the convention expressed in For the standard hydrogen electrode e Eq.(II.30)), while rearrangement of Eq.(II.2 9) for the half cell containing the iron per- gives: chlorates in cell (II.23) ⎛⎞ a 2+ ο Fe log − = log K log (II.25)– a ⎜⎟ − 10 10 10 e ⎜⎟ a 3+ Fe ⎝⎠ and from Eq.(II.27): ⎛⎞ a 2+ ο Fe log − = (II.32) log K a (II.24)– log ⎜⎟ − 10 10 10 e ⎜⎟ a 3+ Fe ⎝⎠ F (II.24) (II.33) − log aE = and − 10 e R ln(10) T which is a specific case of the general equation (II.28). The splitting of redox reactions into two half cell reactions by introducing the − is related to the standard electrode potential, ”, which according to Eq.(II.27) symbol“ e – notation does not in any way refer to solvated electrons). is arbitrary, but useful (this e − e ”, and When calculating the equilibrium composition of a chemical system, both “ + H can be chosen as components and they can be treated numerically in a similar way: + H may be defined for both. However, while etc. equilibrium constants, mass balance,

64 22 II Standards, Conventions and Contents of the Tables represents the hydrated proton in aqueous solution, the above equations use only the − − activity of “ e ”. Concentration to activity con- ”, and never the concentration of “ e versions (or activity coefficients) are never needed for the electron ( Appendix B, cf. Example B.3). In the literature on geochemical modelling of natural waters, it is customary to ous solution with the symbol “pe” or represent the “electron activity” of an aque “p ”( ε ), and the redox potential of an ) by analogy with pH ( log =− a − = a log + − 10 10 H e aqueous solution relative to the standard hydrogen electrode is usually denoted by either “Eh” or ). E ” (see for example [81STU/MOR] , [82DRE] “ [84HOS] , [86NOR/MUN] , H ο ′ is used to denote the so called “formal poten- E In this review, the symbol [74PAR] tial” . The formal (or “conditional”) potential can be regarded as a standard potential for a particular medium in which the activity coefficients are independent (or ο ′ approximately so) of the reactant concentrations [85BAR/PAR] (the definition of E s” for equilibria). Therefore, from parallels that of “concentration quotient T R ο ′ −ν c EE ln = (II.34) ∑ i i n F ο ′ is the potential E for a cell when the ratio of the concentrations (not the activities) E on the right-hand side and the left-hand side of the cell reaction is equal to unity, and ∆ G R T οο rm ′ −ν EE = = ln Š γ − (II.35) ∑ i i F n n F m i 㠊 is are the molality activity coefficients and () , the ratio of molality where the c i i cf. Section II.2). The medium must be specified. to molarity ( II.1.7 pH Because of the importance that potentiometric methods have in the determination of equilibrium constants in aqueous solutions, a short discussion on the definition of “pH” and a simplified description of the experimental techniques used to measure pH will be given here. The acidity of aqueous solutions is often expressed in a logarithmic scale of the hydrogen ion activity. The definition of pH as: ) p H = log = log ( γ am −− +++ 10 10 HHH can only be strictly used in the limiting range of the Debye–Hückel equation (that is, in extremely dilute solutions). In practice the use of pH values requires extra assumptions as to the values for single ion activities. In this review values of pH are used to describe qualitatively the ranges of acidity of experimental studies, and the assumptions de- scribed in Appendix B are used to calculate single ion activity coefficients. The determination of pH is often performed by emf measurements of galvanic , [73BAT] . A common setup is a cell made up cells involving liquid junctions [69ROS]

65 II.1 Symbols, terminology and nomenclature 23 | AgCl(s) in a solution of constant chloride concentra- e.g ., Ag(s) of a reference half cell ( tion), a salt bridge, the test solution, and a glass electrode (which encloses a solution of constant acidity and an internal reference half cell): salt test KCl(aq) AgCl(s) Ag(s) Pt(s) Pt(s) Ag(s) AgCl(s) KCl(aq) solution bridge a b (II.36) where stands for a glass membrane (permeable to hydrogen ions). The emf of such a cell is given by: T R * − EE a E + ln = + j H n F * where is a constant, and E is the liquid junction potential. The purpose of the salt E j b bridge is to minimise the junction potential in junction “ ”, while keeping constant the junction potential for junction “ a ”. Two methods are most often used to reduce and con- concentration (the “salt bridge”) is a E . An electrolyte solution of high trol the value of j requirement of both methods. In the first method, the salt bridge is a saturated (or nearly saturated) solution of potassium chloride. A problem with a bridge of high potassium 1 inside the liquid junction concentration is that potassium perchlorate might precipitate when the test solution contains a high concentration of perchlorate ions. high concentration of the In the other method the salt bridge contains the same ). However, if the same inert electrolyte as the test solution (for example, 3 M NaClO 4 concentration of the background electrolyte in the salt bridge and test solutions is re- are dramatically increas ed. For example, if both the bridge and E duced, the values of j − [ClO ] = 0.1 M as background electrolyte, the dependence of the the test solution have 4 + 3 –1 at 25 ≈ − 440 × [ H [69ROS] ] mV·dm C ·mol ° ” on acidity is “b liquid junction at E j . 0.07 pH units = 2 of ≥ (p.110), which corresponds to an error at pH ng the liquid junction potentials and in Because of the problems in eliminati defining individual ionic activity coefficients, an “operational” definition of pH is given . This definition involves the measurement of pH differences by IUPAC [93MIL/CVI] between the test solution and standard solutions of known pH and similar ionic strength cancel each other when emf values are sub- γ and E (in this way similar values of + j H stracted). The measurement and use of pH in equilibrium analytical investigations cre- ates many problems that have not always been taken into account by the investigators, as discussed in many reviews in Appendix A. In order to deduce the stoichiometry and equilibrium constants of complex formation reactions and other equilibria, it is neces- sary to vary the concentrations of reactants and products over fairly large concentration ranges under conditions where the activity coeffi cients of the species are either known, 1 KClO C (cr) has a solubility of ≈ 0.15 M in pure water at 25 ° 4

66 24 II Standards, Conventions and Contents of the Tables or constant. Only in this manner is it possible to use the mass balance equations for the various components together with the measurement of one or more free concentrations [90BEC/NAG] , p. , , [97ALL/BAN] to obtain the information desired [61ROS/ROS] hydrogen ions, it is necessary to use concentration 326-327. For equilibria involving units, rather than hydrogen ion activity. For experiments in an ionic medium, where the concentration of an “inert” electrolyte is much larger than the concentration of reactants activity coeffi- and products we can ensure that, as a first approximation, their trace cients remain constant even for moderate variations of the corresponding total concen- trations. Under these conditions of fixed ionic strength the free proton concentration + [H ] (also often referred may be measured directly, thereby defining it in terms of – log 10 + ] and pH [H to as pcH) rather than on the activity scale as pH, and the value of – log 10 Equilibrium constants deduced from meas- will differ by a constant term, i.e ., log . γ + 10 H media are therefore constants, because they refer to conditional urements in such ionic the given medium, not to the standard state. In order to compare the magnitude of equi- librium constants obtained in different ionic media it is necessary to have a method for estimating activity coefficients of ionic species in mixed electrolyte systems to a com- mon standard state. Such procedures are discussed in Appendix B. + ] and pH is virtually [H Note that the precision of the measurement of – log 10 ± the same, in very good experiments, 0.001. However, the accuracy is generally con- siderably poorer, depending in the case of glass electrodes largely on the response of the etc. ), and to a lesser extent on the calibration method electrode (linearity, age, pH range, + [H ] calibration standards can be prepared employed, although the stoichiometric – log 10 far more accurately than the commercial pH standards. II.1.8 Order of formulae To be consistent with CODATA, the data tables are given in “ Standard Order of Ar- . This scheme is presented in Figure II-1 below, and shows [82WAG/EVA] rangement ” the sequence of the ranks of the elements in this convention. The order follows the ranks of the elements. For example, for uranium, this means that, after elemental uranium and its 4+ U ), the uranium compounds and complexes with oxygen , monoatomic ions ( e.g. en those with oxygen and hydrogen, and so would be listed, then those with hydrogen, th d combinations of the elements. Within a on, with decreasing rank of the element an rank elements go before increasing coeffi- class, increasing coefficients of the higher ple, in the U–O–F class of compounds and cients of the lower rank elements. For exam UOF (cr) , , , UOF (g) , UO F(aq) UOF (cr) complexes, a typical sequence would be 2 2 4 4 + − 2 − . UO F (aq) , etc UO F (cr) , , UO F (g) , UOF(cr) UO F , , UO F , UO F 22 22 22 2 4 2 236 23 . Formulae with identical stoichiometry are in alphabetical order of their [92GRE/FUG] designators.

67 25 II.1 Symbols, terminology and nomenclature Figure II-1: Standard order of arrangement of the elements and compounds based on the 82WAG/EVA ). periodic classification of the elements (from [ ]           #                                        $  "     !      " #        '( & % %   # & %   $  * #     )  % % % +  & II.1.9 Reference codes The references cited in the review are ordered chronologically and alphabetically by the first two authors within each year, as described by CODATA [87GAR/PAR] . A refer- ence code is made up of the final two digits of the year of appearance (if the publication is not from the 20th century, the year will be put in full). The year is followed by the first three letters of the surnames of the first two authors, separated by a slash. If there are multiple reference codes, a “2” will be added to the second one, a “3” to the third one, and so forth. Reference codes are always enclosed in square brackets. II.2 Units and conversion factors Thermodynamic data are given according to the Système International d'unités (SI units). The unit of energy is the joule. Some basic conversion factors, also for non- thermodynamic units, are given in Table II-4.

68 26 II Standards, Conventions and Contents of the Tables Table II-4: Unit conversion factors. To convert from multiply by to (non-SI unit symbol) (SI unit symbol) –10 × (exactly) 1 10 metre (m) ångström (Å) 5 standard atmosphere 10 (atm) pascal (Pa) 1 (exactly) . 01325 × 5 (exactly) bar (bar) pascal (Pa) 1 × 10 joule (J) thermochemical calorie (cal) 4.184 (exactly) −− 11 11 − − JK mol ⋅⋅ mol  e.u. cal K ⋅⋅ 4.184 (exactly) entropy unit ect deals with the thermodynamics of Since a large part of the NEA TDB proj aqueous solutions, the units describing the amount of dissolved substance are used very 3 − ⋅ ” mol dm frequently. For convenience, this review uses “M” as an abbreviation of “ − 1 mol kg ⋅ ” for , and, in Appendices B and C, “m” as an abbreviation of “ c for molarity, . It is often necessary to convert concen tration data from molarity to molality molality, m r the correction and extrapolation of equilib- and vice versa. This conversion is used fo rium data to zero ionic strength by the specific ion interaction theory, which works in molality units ( cf . Appendix B). This conversion is made in the following way. Molality m moles of substance B dissolved in 1 kilogram of pure water. Molarity is defined as B kilogram of pure water, moles of substance B dissolved in c ( ) cM − ρ is defined as B B –3 ρ is the density of the solution in kg·dm and M the molar weight of the solute in where –1 . From this it follows that: kg·mol c B = . m B ρ cM − B [76BAE/MES] Baes and Mesmer .439) give a table with conversion factors , ( p (from molarity to molality) for nine electrolytes and various ionic strengths. Conversion factors at 298.15 K for twenty one electrolytes, calculated using the density equations , are reported in Table II-5. [85SOH/NOV] reported by Söhnel and Novotný Example: 1.00 M NaClO 1.05 m NaClO  44 1.00 M NaCl 1.02 m NaCl  4.00 M NaClO 4.95 m NaClO  44 7.55 m NaNO 6.00 M NaNO  33 It should be noted that equilibrium constants need also to be converted if the concentration scale is changed from molarity to molality or vice versa. For a general equilibrium reaction, 0 = be expressed either in B , the equilibrium constants can ν ∑ B B molarity or molality units, K or K , respectively : m c log ν c K log = ∑ 10 B 10 c B B log log ν m K = ∑ 10 B m B 10 B

69 27 II.2 Units and conversion factors = mc mc = Š , or the relationship between , Š (log log ) log − (/) With 10 B B 10 10 BB and K becomes very simple, as shown in Eq.(II.37). K c m (II.37) Š log = log log + ν KK ∑ 10 mc B 10 10 B . Eq. ν is the sum of the stoichiometric coefficients of the reaction, cf ∑ B B (II.53) and the values of Š are the factors for the conversion of molarity to molality as tabulated in Table II-5 for several electroly te media at 298.15 K. The differences be- tween the values in Table I I-5 and the values listed in the uranium NEA TDB review [92GRE/FUG] ( p .23) are found at the highest concentrations, and are no larger than 3 –1 ± 003 dm . 0 , reflecting the accuracy expected in ·kg this type of conversion. The un- certainty introduced by the use of Eq.(II.37) in the values of log K will be no larger 10 m than ± 0.001 . ν ∑ B B Table II-5: Factors , to molality, Š for the conversion of molarity, , of a substance m c B B [85SOH/NOV] B, in various media at 298.15 K (calculated from densities in ). 3 c m / Š (dm = of solution per kg of H O) B 2 B ClO c HCl NaCl LiCl NaClO ) LiClO Ba(ClO NH (M) HClO 4 4 4 4 4 2 4 0.10 1.0077 1.0075 1.0074 1.0091 1.0108 1.0048 1.0046 1.0049 0.25 1.0147 1.0145 1.0141 1.0186 1.0231 1.0076 1.0072 1.0078 0.50 1.0266 1.0265 1.0256 1.0351 1.0450 1.0123 1.0118 1.0127 0.75 1.0386 1.0388 1.0374 1.0523 1.0685 1.0172 1.0165 1.0177 1.00 1.0508 1.0515 1.0496 1.0703 1.0936 1.0222 1.0215 1.0228 1.50 1.0759 1.0780 1.0750 1.1086 1.1491 1.0324 1.0319 1.0333 2.00 1.1019 1.1062 1.1019 1.2125 1.0430 1.0429 1.0441 3.00 1.1571 1.1678 1.1605 1.3689 1.0654 1.0668 1.0666 4.00 1.2171 1.2374 1.2264 1.0893 1.0930 1.0904 1.1147 1.1218 1.1156 5.00 1.2826 1.3167 1.1418 1.1423 6.00 1.3547 1.4077 MgCl NaNO CaCl Cl LiNO NaBr HNO (M) KCl NH c 3 2 3 4 3 2 0.10 1.0057 1.0066 1.0049 1.0044 1.0054 1.0056 1.0058 1.0059 0.25 1.0099 1.0123 1.0080 1.0069 1.0090 1.0097 1.0102 1.0103 0.50 1.0172 1.0219 1.0135 1.0119 1.0154 1.0169 1.0177 1.0178 0.75 1.0248 1.0318 1.0195 1.0176 1.0220 1.0242 1.0256 1.0256 1.00 1.0326 1.0420 1.0258 1.0239 1.0287 1.0319 1.0338 1.0335 1.50 1.0489 1.0632 1.0393 1.0382 1.0428 1.0478 1.0510 1.0497 2.00 1.0662 1.0855 1.0540 1.0546 1.0576 1.0647 1.0692 1.0667 3.00 1.1037 1.1339 1.0867 1.0934 1.0893 1.1012 1.1090 1.1028 4.00 1.1453 1.1877 1.1241 1.1406 1.1240 1.1417 1.1534 1.1420 5.00 1.2477 1.1974 1.1619 1.1865 1.2030 1.1846 6.00 1.2033 1.2361 1.2585 1.2309 (Continued on next page)

70 28 II. Standards, Conventions and Contents of the Tables Table II-5 (continued) 3 m Š / c O) = of solution per kg of H (dm 2 B B NaSCN NO CO Na H K SO c Na (M) NH SO PO (NH H ) SO CO 2 4 3 4 4 2 2 4 3 2 2 3 3 4 4 1.0077 1.0064 1.0044 1.0082 0.10 1.0074 1.0027 1.0042 1.0069 0.25 1.0151 1.0116 1.0071 1.0166 1.0143 1.0030 1.0068 1.0130 1.0276 1.0209 1.0127 1.0319 1.0261 1.0043 1.0121 1.0234 0.50 1.0405 1.0305 1.0194 1.0486 0.75 1.0383 1.0065 1.0185 1.0342 1.00 1.0539 1.0406 1.0268 1.0665 1.0509 1.0094 1.0259 1.0453 1.0818 1.0619 1.0441 1.1062 1.0773 1.0170 1.0430 1.0686 1.50 2.00 1.1116 1.0848 1.1514 1.1055 1.0268 1.0632 1.0934 3.00 1.1769 1.1355 1.2610 1.1675 1.1130 1.1474 4.00 1.2512 1.1935 1.4037 1.2383 1.1764 1.2083 1.3194 1.2560 1.2773 5.00 1.3365 1.2600 6.00 1.4351 1.3365 1.4131 1.3557 II.3 Standard and reference conditions II.3.1 Standard state A precise definition of the term “standard state” has been given by IUPAC [82LAF] . c parameters, but not their absolute values, The fact that only changes in thermodynami can be determined experimentally, makes it important to have a well-defined standard state that forms a base line to which the effe ct of variations can be referred. The IUPAC [82LAF] definition of the standard state has been adopted in the NEA–TDB project. ο The standard state pressure, p = . 1 MPa (1 bar), has therefore also been adopted, cf . 0 rd state principle to pure substances and Section II.3.2. The application of the standa noted that the standard state is always mixtures is summarised below. It should be cf linked to a reference temperature, . Section II.3.3. • The standard state for a gaseous substance, whether pure or in a gaseous mixture, is the pure substance at the standard state pressure and in a (hypothetical) state in which it exhibits ideal gas behaviour. • The standard state for a pure liquid substan ce is (ordinarily) the pure liquid at the standard state pressure. • ce is (ordinarily) the pure solid at the The standard state for a pure solid substan standard state pressure. • The standard state for a solu te B in a solution is a hypothetical liquid solution, at –1 ο = the standard state pressure, in which = 1 mol·kg , and in which the activ- mm B ity coefficient is unity. γ B ο ο ∆ H It should be emphasised that the use of superscript, implies e.g. , , in , fm that the compound in question is in the standa rd state and that the elements are in their reference states. The reference states of th e elements at the reference temperature ( cf .

71 29 II.3 Standard and reference conditions Section II.3.3) are listed in Table II-6. Table II-6: Reference states for some elements at the reference temperature of 298.15 K and standard pressure of 0.1 MPa [82WAG/EVA] , [89COX/WAG] [91DIN] , , [2005GAM/BUG] [2005OLI/NOL] . , O Zn crystalline, hexagonal gaseous 2 gaseous H Cd crystalline, hexagonal 2 He gaseous Hg liquid Ne gaseous Cu crystalline, cubic Ar gaseous Ag crystalline, cubic Kr gaseous Ni crystalline, fcc Xe gaseous Fe crystalline, cubic, bcc gaseous Tc crystalline, hexagonal F 2 gaseous V crystalline, cubic Cl 2 liquid Ti crystalline, hexagonal Br 2 crystalline, orthorhombic Am crystalline, dhcp I 2 S crystalline, orthorhombic Pu crystalline, monoclinic Se crystalline, trigonal Np crystalline, orthorhombic Te crystalline, hexagonal U crystalline, orthorhombic Th crystalline, cubic gaseous N 2 P Be crystalline, hexagonal crystalline, cubic (“white”) crystalline, rhombohedral (“grey”) As crystalline, hexagonal Mg Sb crystalline, rhombohedral Ca crystalline, cubic, fcc Bi crystalline, rhombohedral Sr crystalline, cubic, fcc C crystalline, hexagonal (graphite) Ba crystalline, cubic Si crystalline, cubic Li crystalline, cubic Ge crystalline, cubic Na crystalline, cubic crystalline, cubic Sn crystalline, tetragonal (“white”) K Pb crystalline, cubic crystalline, cubic Rb β , Cs crystalline, rhombohedral B crystalline, cubic Al crystalline, cubic II.3.2 Standard state pressure The standard state pressure chosen for all selected data is 0.1 MPa (1 bar) as recom- mended by the International Union of Pure and Applied Chemistry IUPAC [82LAF] . However, the majority of the thermodynamic data published in the scientific literature and used for the evaluations in this review, refer to the old standard state pres- sure of 1 “standard atmosphere” ( = 0.101325 MPa). The difference between the ther- modynamic data for the two standard state pr essures is not large and lies in most cases within the uncertainty limits. It is nevertheless essential to make the corrections for the change in the standard state pressure in order to avoid inconsistencies and propagation

72 30 II. Standards, Conventions and Contents of the Tables by the change between these two standard of errors. In practice the parameters affected state pressures are the Gibbs energy and entr opy changes of all processes that involve gaseous species. Consequently, changes occur also in the Gibbs energies of formation of species that consist of elements whose refe rence state is gaseous (H, O, F, Cl, N, and the noble gases). No other thermodynamic quantities are affected significantly. A large part of the following discussion has been taken from the NBS tables of chemical ther- modynamic properties [82WAG/EVA] , see also Freeman [84FRE] . of pressure on the properties of all The following expressions define the effect substances: ⎛⎞ ∂∂ HV ⎛⎞ V T VT = (1 ) =−α − (II.38) ⎜⎟ ⎜⎟ pT ∂∂ ⎝⎠ ⎝⎠ p T 2 C ∂ ⎛⎞ ⎛⎞ ∂ V p = − T (II.39) ⎜⎟ ⎜⎟ 2 p ∂ ∂ T ⎝⎠ ⎝⎠ p T ⎛⎞ SV ∂∂ ⎛⎞ = = V −α − (II.40) ⎜⎟ ⎜⎟ pT ∂∂ ⎝⎠ ⎝⎠ p T ⎛⎞ G ∂ = V (II.41) ⎜⎟ p ∂ ⎝⎠ T 1 V ∂ ⎛⎞ (II.42) α≡ where ⎜⎟ ∂ VT ⎝⎠ p V = R T / p and α = R / pV = 1 / T. The conversion equations listed below For ideal gases, (Eqs. (II.43) to (II.50)) apply to the small pressure change from 1 atm to 1 bar (0.1 MPa). The quantities that refer to the old sta ndard state pressure of 1 atm are assigned () atm and those that refer to the new sta ndard state pressure of 1 bar are , the superscript () bar assigned the superscript . For all substances the chan ges in the enthalpy of fo rmation and heat capacity are much smaller than the expe rimental accuracy and can be di sregarded. This is exactly true for ideal gases. (bar) (atm) () () = 0 HT H T ∆−∆ ff (II.43) (atm) (bar) (II.44) ( ) = 0 CT C T − ( ) pp For gaseous substances, the entropy difference is: (atm) ⎛⎞ p (atm ) ( bar ) ( ) ( ) = R ln ST S T = R ln 1.01325 − ⎜⎟ (bar) p ⎝⎠ –1 –1 ·mol (II.45) = 0.1094 J·K α = R / pV . This is exactly true for ideal gases, as follows from Eq.(II.40) with The entropy change of a reaction or process is thus dependent on the number of moles of gases involved:

73 II.3 Standard and reference conditions 31 (atm) ⎛⎞ p (atm) (bar) = δ SS R ln ∆−∆ ⋅ ⎜⎟ rr (bar) p ⎝⎠ –1 –1 = 0.1094 J·K × ·mol δ (II.46) δ is the net increase in moles of gas in the process. where Similarly, the change in the Gibbs ener gy of a process between the two stan- dard state pressures is: (atm) ⎛⎞ p (atm) (bar) = δ R ln GG T ∆−∆ −⋅ ⎜⎟ rr (bar) p ⎝⎠ –1 = – × δ at 298.15 K. (II.47) 0.03263 kJ·mol (bar) (atm) GG ∆−∆ , since the Gibbs energy of for- Eq.(II.47) applies also to ff describes the formation process of a co mpound or complex from the reference mation states of the el ements involved: (atm) (bar) –1 GG ∆−∆ = – δ × (II.48). 0.03263 kJ·mol at 298.15 K. ff The changes in the equilibrium constants and cell potentials with the change in the standard state pressure follows from the expression for Gibbs energy changes, Eq.(II.47), (bar) (atm) ∆−∆ GG (bar) (atm) rr KK = log log −− 10 10 T R ln 10 (atm) ⎛⎞ p ln ⎜⎟ (bar) (atm) p ⎛⎞ p ⎝⎠ δ δ = log ⋅ =⋅ ⎜⎟ 10 (bar) ln10 p ⎝⎠ (II.49) 0.005717 δ × = (bar) (atm) GG ∆−∆ (bar) (atm) rr = EE −− F n (atm) ⎛⎞ p R ln T ⎜⎟ (bar) p ⎝⎠ =⋅ δ F n 0.0003382 (II.50) δ V at 298.15 K =⋅ n It should be noted that the standard potential of the hydrogen electrode is equal to 0.00 V exactly, by definition. + −ο 1 def H 0.00V H (g) + U eE (II.51). 2 = 2

74 32 II. Standards, Conventions and Contents of the Tables H(g), is This definition will not be changed, although a gaseous substance, 2 involved in the process. The change in the potential with pressure for an electrode po- tential conventionally written as: − + Ag (cr) U Ag + e ed reaction that incl should thus be calculated from the balanc udes the hydrogen elec- trode, 1 ++ Ag + H (g) Ag (cr) + H U 2 2 . Hence, the contribution to from an electron in a half cell reaction is δ = − 0.5 Here δ the same as the contribution of a gas molecule with the stoichiometric coefficient of 0.5. as the combination with the hydrogen half cell. This leads to the same value of δ Example: (bar) (atm) +2+ Fe(cr) + 2 H Fe + H (g) = 1 U = 0.00017 V EE δ− 2 ) (bar) (bar 0.0057 = − KK CO (g) CO (aq) = 1 log − U log δ− 22 10 10 53 -1 (atm) (bar) δ∆−∆−⋅ U GG 0.008 kJ mol = NH (g) + O NO(g) + H O(g) = 0.25 2 r r 32 42 1 -1 −− (bar) (atm) ⋅ 0.098 kJ mol δ− U ∆ −∆ ClO = 3 Cl (g) + 2 O (g) + e GG = f f 22 4 2 II.3.3 Reference temperature The definitions of standard states given in Section II.3 make no reference to fixed tem- perature. Hence, it is theoretically possible to have an infinite number of standard states of a substance as the temperature varies. It is, however, convenient to complete the definition of the standard state in a particular context by choosing a reference tempera- , the reference temperature chosen in the ture. As recommended by IUPAC [82LAF] C. Where necessary for the discussion, ° = 298.15 K or t = 25.00 T NEA-TDB project is values of experimentally measured temperat ures are reported after conversion to the . The relation between th e absolute temperature T (K, kelvin) and the IPTS–68 [69COM] 15 K. . 273 = T = () where TT − t C) is defined by ° ( t Celsius temperature o o II.4 Fundamental physical constants To ensure the consistency with other NEA TDB Reviews, the fundamental physical . Those relevant to this constants are taken from a publication by CODATA [86COD] review are listed in Table II-7. Note that updated values of the fundamental constants can be obtained from CODATA , notably through its Internet site. In most cases, recal- culation of the NEA TDB database entries with the updated values of the fundamental constants will not introduce significant (with respect to their quoted uncertainties) ex- cursions from the current NEA TDB selections.

75 33 II.6 The NEA-TDB system Table II-7: Fundamental physical constants. These values have been taken from . The digits in parentheses are the one–standard–deviation uncer- [86COD] CODATA tainty in the last digits of the given value. Quantity Symbol Value Units –1 speed of light in vacuum 299 792 458 m·s c –7 –7 –2 10 permeability of vacuum = 12.566 370 614... 10 4 N·A μ π× 0 –1 2 2 –1 –12 μ ·m c = 8.854 187 817... ·J 10 1/ C permittivity of vacuum є 0 ο –34 h Planck constant 10 J·s 6.626 0755(40) –19 C e 10 elementary charge 1.602 177 33(49) –1 23 10 6.022 1367(36) mol N Avogadro constant A –1 Faraday constant F 96 485.309(29) C·mol –1 –1 R 8.314 510(70) J·K molar gas constant ·mol –23 –1 Boltzmann constant, R/N 10 J·K k 1.380 658(12) A Non-SI units used with SI: –19 electron volt, (e/C) J eV 1.602 177 33(49) 10 J –27 1.660 5402(10) atomic mass unit, u kg 10 12 1 ( ) mmC = 1u = u 12 II.5 Uncertainty estimates One of the principal objectives of the NEA TDB development effort is to provide an idea of the uncertainties associat ed with the data selected in the reviews. In general the uncertainties should define the range within which the corresponding data can be repro- , a full statistical treatment is limited or duced with a probability of 95%. In many cases impossible due to the availability of only one or a few data points. Appendix C de- scribes in detail the procedures used for the assignment and treatment of uncertainties, as well as the propagation of errors and the standard rules for rounding. II.6 The NEA-TDB system A database system has been developed at the NEA Data Bank that allows the storage of es as well as for reactions. The structure thermodynamic parameters for individual speci of the database system allows consistent derivation of thermodynamic data for individ- ual species from reaction data at standard conditions, as well as in ternal recalculations of data at standard conditions. If a selected value is changed, all the dependent values will be recalculated consistently. The mainte nance of consistency of all the selected data, including their uncertainties ( cf . Appendix C), is ensured by the software devel- oped for this purpose at the NEA Data Bank. The literature sources of the data are also stored in the database.

76 34 II. Standards, Conventions and Contents of the Tables lid at the reference temperature of The following thermodynamic parameters, va 298.15 K and at the standard pressure of 1 bar, are stored in the database: ο ∆ G the standard molar Gibbs energy of formation from the elements in fm –1 ) their reference state (kJ·mol ο H ∆ the standard molar enthalpy of formation from the elements in their fm –1 ) reference state (kJ·mol ο –1 –1 S ·mol ) the standard molar entropy (J·K m ο –1 –1 the standard molar heat capacity (J·K ·mol C ). p ,m ο ο ο S ∆ G ∆ H For aqueous neutral species and ions, the values of , and , m fm fm ο correspond to the standard partial molar quantities, and for individual aqueous C p ,m ions they are relative quantities, defined with respect to the aqueous hydrogen ion, ac- + ο ο + ∆ H S H H , T ) = 0 and that ) T that ( , ( [89COX/WAG] cording to the convention fm m = 0. Furthermore, for an ionised solute B containing any number of different cations and anions: ο οο H (cation, aq) + (anion, aq) HH (B , aq) = ∆ν∆ ν∆ ∑∑ + fm fm fm ±− − + οο ο (B , aq) = (cation, aq) + (anion, aq) SS νν S ∑∑ m m+m ±− − + As the thermodynamic parameters vary as a function of temperature, provision coefficients of empiri cal temperature func- is made for including the compilation of the tions for these data, as well as the temperat ure ranges over which they are valid. In or calculated at several temperatures many cases the thermodynamic data measured were published for a particular species, rath er than the deduced temperature functions. ed in this review to obtain the most sig- In these cases, a linear regression method is us nificant coefficients of the following empirical function for a thermodynamic parameter, : X −− 212 () = ln ln +⋅+⋅+⋅ +⋅ +⋅ + ⋅ TabTcTdTeTfTgTT X X XX X X X X (II.52) i 33 − X hT jTkT + ⋅⋅+⋅ + + . XXX T Most temperature variations can be described with three or four parameters. In , the thermal functions of the heat capacities of indi- i.e. , the present series, only CT () ,m p the database. They refer to the relation: vidual species are considered and stored in 212 − − ⋅+⋅ +⋅ +⋅ CT abTcTdT eT () = + ,m p (where the subindices for the coefficients ha ve been dropped) and are listed in the se- lected value tables. The pressure dependence of thermodyna mic data has not been the subject of critical analysis in the present compilation. Th e reader interested in higher temperatures and pressures, or the pressure dependency of thermodynamic functions for geochemical , applications, is referred to the specialised literature in this area, e.g. , [82HAM]

77 II.6 The NEA-TDB system 35 , [88SHO/HEL] [88TAN/HEL] , [89SHO/HEL] , [89SHO/HEL2] , , [84MAR/MES] , . [91AND/CAS] [90MON] Selected standard thermodynamic data re ferring to chemical reactions are also ”, involving reactants and products compiled in the database. A chemical reaction “ r ‘B”, can be abbreviated as: r 0 B =ν (II.53) ∑ B B r where the stoichiometric coefficients υ are positive for products, and negative for B ered in the NEA TDB system include: reactants. The reaction parameters consid ο log K the equilibrium constant of the reaction, logarithmic 10 r –1 ο G ∆ the molar Gibbs energy of reaction (kJ·mol ) rm –1 ο ∆ H the molar enthalpy of reaction (kJ·mol ) rm ο –1 –1 the molar entropy of reaction (J·K S ∆ ·mol ) rm –1 –1 ο C the molar heat capacity of reaction (J·K ·mol ). ∆ r,m p ailable, are stored acc ording to Eq.(II.52). The temperature functions of these data, if av ο ο ∆ K G according to the following , is related to The equilibrium constant, rm r relation: ο ∆ G ο rm − K = log 10 r T R ln(10) ο ∆ (B) G and can be calculated from the individual values of (for example, those given fm in selected values tables), according to: 1 οο r (B) log = −ν∆ (II.54) KG ∑ 10 f B m r R ln(10) T B II.7 Presentation of the selected data rwise indicated, they The selected data are presented in Chapters III and IV. Unless othe refer to standard conditions ( . Section II.3) and 298.15K (25 ° C) and are provided with cf an uncertainty which should correspond to the 95% confidence level (see Appendix C). Chapter III contains tables of selected thermodynamic data for individual com- pounds and complexes of oxalate, citrate and edta (Table III–1, Table III–3 and Table III–5, respectively), tables of selected reac tion data (Table III–2, Table III–4, Table III– 6 and Table III–7) for reactions concerning oxalate, citrate, edta and isa species, respec- tively. The selection of these data is discussed in Chapters VI to IX. Chapter IV contains, for auxiliary compounds and complexes, a table of the thermodynamic data for individual species (Table IV–1) and a table of reaction data

78 36 II. Standards, Conventions and Contents of the Tables [89COX/WAG] lues are the CODATA Key Values . The (Table IV–2). Most of these va [99RAR/RAN] , , [92GRE/FUG] selection of the remaining auxiliary data is discussed in and [2005OLI/NOL] . , [2005GAM/BUG] [2001LEM/FUG] All the selected data presented in Tables of Chapter III are internally consis- tent. This consistency is maintained by the internal consistency verification and recalcu- lation software developed at the NEA Data Bank in conjunction with the NEA-TDB data base system, cf . Section II.6. Therefore, when using the selected data for organic species, the auxiliary data of Chapter IV must be used together with the data in Chapter III to ensure internal cons istency of the data set. It is important to note that Table III–2, Table III–4, Table III–6, Table III–7 and Table IV–2 include only those species fo r which the primary selected data are reac- tion data. The formation data derived there from and listed in Table III–1, Table III–3, g auxiliary data, and their uncertainties Table III–5 and Table IV–1, are obtained usin are propagated accordingly. In order to mainta in the uncertainties originally assigned to the selected data in this review, the user is advised to make direct use of the reaction data presented in that Table III–2, Table III–4, Table III–6, Table III–7 and Table IV–2, rather than taking the derived values in Table III–1, Table III–3, Table III–5 and Table .(II.54). The later approach would imply a IV–1 to calculate the reaction data with Eq twofold propagation of the uncertainties and result in reaction data whose uncertainties would be considerably larger than those originally assigned. The thermodynamic data in the selected set refer to a temperature of 298.15 K ° C), but they can be recalculated to other temperatures if the corresponding data (25.00 . For example, the [97PUI/RAR] (enthalpies, entropies, heat capacities) are available temperature dependence of the standard reacti on Gibbs energy as a function of the stan- T = 298.15 K), and of the heat ca- e reference temperature ( dard reaction entropy at th 0 pacity function is: ο ⎛⎞ (T) C ∆ TT r,m p οο ο ο d , ( ) = ( ) + (T) d T T T ST ( ) GT H T C ∆∆ ∆−∆+ ⎜⎟ rm 0 r ,m r m 0 rm p ∫∫ ⎜⎟ TT 00 T ⎝⎠ and the temperature dependence of the standard equilibrium constant as a function of py and heat capacity is: the standard reaction enthal ο ⎛⎞ ∆ HT () 11 οο rm0 −− log ( ) = log KT KT ( ) ⎜⎟ 10 10 0 TT R ln(10) 0 ⎝⎠ ο TT ∆ CT () 11 r,m p ο d , ( ) d + −∆ T CTT r,m p ∫∫ R ln(10) R ln(10) TT TT 00 where R is the gas constant ( cf . Table II-7).

79 II.7 Presentation of the selected data 37 enthalpies of reaction are selected or In the case of aqueous species, for which can be calculated from the selected enthalpies of formation, but for which there are no selected heat capacities, it is in most cases possible to recalculate equilibrium constants ° to temperatures up to 100 to 150 C, with an additional uncertainty of perhaps about 1 to 2 logarithmic units, due to neglecting the heat capacity contributions to the temperature , species i.e. correction. However, it is important to observe that “new” aqueous species, ° C and therefore not detected, may be significant not present in significant amounts at 25 at higher temperatures, see for [87CIA/IUL] example the work by Ciavatta et al. . Addi- tional high-temperature experiments may therefo re be needed in order to ascertain that proper chemical models are used in the modelling of hydrothermal systems. For many species, experimental thermodynamic data are not available to allow a selection of pa- rameters describing the temperature dependence of equilibrium constants and Gibbs energies of formation. The user may find information on various procedures to estimate the temperature dependence of thes e thermodynamic parameters in [97PUI/RAR] . The thermodynamic data in the selected set refe r to infinite dilution for soluble species. Ex- ο , usually measured at high ionic strength, to trapolation of an equilibrium constant K K at I = 0 using activity coefficients , is explained in Appendix B. The corresponding γ Gibbs energy of dilution is: ο (II.55) ∆∆−∆ = GGG m m r m r dil (II.56) ∆ = R ln T γ − ± r ∆ S Similarly can be calculated from ln γ and its variations with T , while: ± m dil 2 ∂ ( ln ) ∆∆ HT (II.57) γ = R ± p dil m r ∂ T , which is neglected in this review, when no T γ with depends only on the variation of ’s are available. In this case the Gibbs energy of data on the temperature dependence of γ dilution fference. This entropy of reaction is G ∆ is entirely assigned to the entropy di m dil οο ο calculated using = GHTS . ∆∆−∆ , the above assumption H ∆ = 0, and G ∆ rm rm r m dil m dil m

80

81 Part II Tables of selected data 39

82

83 Chapter III III Selected data for organic compounds and complexes This chapter presents the chemical thermodynamic data set for compounds and com- plexes of U, Np, Pu, Am and Ni as well as H, Na, K, Mg and Ca w ith oxalate, citrate, ethylenediaminetetraacetate (edta) and iso-sacch arinate (isa) that has been selected in this review. Table III–1, Table III–3 and Table III–5 contain the recommended thermo- dynamic data of the compounds and species complexed with oxalate, citrate and edta, respectively. Table III–2, Table III–4, Table III–6 and Table III–7 contain the recom- mended thermodynamic data of chemical equilibrium reactions by which the oxalate, citrate, edta and isa compounds and complexes are formed, respectively. The species and reactions in the tables appear in standard order of arrange- ment. Table III–2, Table III–4, Table III–6 and Table III–7 contain information only on those reactions for which primary data selec tions are made in Chapter VI, VII, VIII and IX of this review, respectively. These selected reaction data are used, together with data for species and auxiliary data selected in the NEA TDB Project, to derive the corre- sponding formation data in Table III–1, Ta ble III–3 and Table III–5. The uncertainties associated with values for the species and the auxiliary data are in some cases substan- tial, leading to comparatively large uncertainties in the formation quantities derived in this manner. ο The values of for many reactions are known more accurately than would G ∆ rm ο be calculated directly from the uncertainties of the ∆ values in Table III–1, Table G fm I–5 and auxiliary data. The inclusion III–3, Table II of tables for reaction data (Table III–2, Table III–4, Table III–6 and Table III–7) in this report allows the use of equilib- rium constants with total un certainties that are based di rectly on the expe rimental accu- racies. This is the main reason for including both the table for reaction data (Table III–2, ο ο ο Table III–4, Table III–6 and Table III–7) and the tables of ∆ , G S values ∆ and H m fm fm (Table III–1, Table III–3 and Table III–5). 41

84 III. Selected data for organic compounds and complexes 42 A detailed discussion of the selection procedure is presented in Chapters VI to IX. It may be noted that these chapters contain data on more species or compounds than are present in the tables of Chapter III. The main reasons for this situation are the lack of information for a proper extrapolation of the primary data to standard conditions in some systems and lack of solid primary data in others. A warning: The addition of any aqueous species and their data to this internally consistent data base can result in a modified data set, which is no longer rigorous and can lead to erroneous results. The situation is similar when gases or solids are added.

85 III. Selected data for organic compounds and complexes 43 Table III–1: Selected thermodynamic data for compounds and complexes with oxalate. All ionic species listed in this table are aqueous species. Unless noted otherwise, all data i.e. refer to the reference temperature of , a pressure 298.15 K and to the standard state, = of 0.1 MPa and, for aqueous species, infinite dilution ( I 0). The uncertainties listed below each value represent total uncertainties and correspond in principle to the statisti- . cally defined 95% confidence interval. Values obtained from internal calculation, cf footnotes (a) and (b), are rounded at the third digit after the decimal point and may therefore not be exactly identical to those given in Part III. Systematically, all the values are presented with three digits after the decimal point, regardless of the significance of these digits. The data presented in this ta ble are available on computer media from the OECD Nuclear Energy Agency. ο ο ο ο ∆ S G H ∆ C p ,m m fm fm Compound –1 –1 –1 –1 –1 –1 (kJ · mol · mol ) (kJ · mol ) (J · K · mol ) ) (J · K (d) (b) (b) 2– – 680.134 – 830.660 47.597 ox 1.830 ± ± 1.592 ± 3.020 (d) (d) (b) – – 704.393 – 823.360 153.447 Hox 1.588 1.831 ± 3.045 ± ± (a) 105.900 115.600 – 698.549 – 828.800 -H ox α 2 1.502 ± 0.200 ± 1.500 ± 0.200 ± (d) (d) (b) – 820.060 – 712.384 ox(aq) 191.318 H 2 ± 1.508 3.523 ± ± 1.839 – 827.500 β -H ox 2 1.500 ± – 1427.000 . H ox 2H O(cr) 2 2 ± 1.500 (b) (b) (b) 15.159 – 885.672 Ni(ox)(aq) – 755.531 1.843 3.561 ± 1.998 ± ± (b) (b) (b) 2– – 1449.650 83.499 – 1724.132 Ni(ox) 2 3.317 6.423 ± ± 3.761 ± + (b) – 1315.991 Am(ox) 5.166 ± (b) – – 2020.099 Am(ox) 2 6.107 ± (b) 3– – 2713.304 Am(ox) 3 9.237 ± (b) – – 1610.160 NpO ox 2 5.945 ± (b) 3– – 2301.140 (ox) NpO 2 2 6.809 ± (Continued on next page)

86 III. Selected data for organic compounds and complexes 44 Table III–1 (continued) ο ο ο ο ∆ ∆ S G H C fm fm m ,m p Compound –1 –1 –1 –1 –1 –1 (kJ · mol · mol ) ) (J · K · mol (J · K ) ) (kJ · mol (a) (d) (b) UO 170.958 – 1673.383 – 1824.300 ox(aq) 2 62.039 ± 18.300 2.690 ± ± (b) (b) 347.517 – 2702.000 . – 2395.078 UO 3H O(cr) ox 2 2 18.000 ± 63.351 3.103 ± ± (b) 2– – 2379.317 (ox) UO 2 2 ± 4.145 (b) 4– – 3071.723 (ox) UO 2 3 10.320 ± (b) Mg(ox)(aq) – 1155.830 2.277 ± (b) 2– – 1845.154 Mg(ox) 2 ± 3.922 (b) (b) 153.340 – 1680.990 156.370 – 1519.912 . O(cr) Ca(ox) H 2 2.138 1.945 ± 2.000 ± ± 2.000 ± (b) (b) (b) – 1970.520 205.678 . – 1754.598 Ca(ox) 2H O(cr) 2 ± ± 5.006 2.139 2.180 ± (b) (b) (b) 258.429 – 2260.850 – 1991.110 . O(cr) 3H Ca(ox) 2 2.289 5.452 2.126 ± ± ± (b) – 1251.149 Ca(ox)(aq) 2.138 ± (b) 2– – 1936.020 Ca(ox) 2 3.016 ± οο ο ∆=∆− GHTS . (a) Value calculated internally using ∑ f m m, i fm i (b) Value calculated internally from reaction data (see Table III–2). (d) Value calculated internally from reaction data for a different key species.

87 III. Selected data for organic compounds and complexes 45 Table III–2: Selected thermodynamic data for reactions involving oxalate compounds and complexes. All ionic species listed in this table are aqueous species. Unless noted mperature of 298.15 K and to the standard otherwise, all data refer to the reference te I = 0). The i.e. , a pressure of 0.1 MPa and, for aqueous species, infinite dilution ( state, ent total uncertainties and correspond in uncertainties listed below each value repres principal to the statistically defined 95% confidence interval. Values obtained from in- ternal calculation, cf . footnote (a), are rounded at the third digit after the decimal point and may therefore not be exactly identical to those given in Part III. Systematically, all the values are presented with three digits after the decimal point, regardless of the sig- nificance of these digits. The data presented in this table are available on computer me- dia from the OECD Nuclear Energy Agency. Reaction Species ο ο ο ο ∆ H ∆ ∆ G S log K rm rm rm 10 –1 –1 –1 –1 ) · mol (J · K ) )(kJ · mol (kJ · mol 2– 2– – + ox Hox H + ox U (a) 24.259 – 7.300 – 105.850 – 4.250 ± ± 0.057 ± 0.386 0.010 0.100 ± – – + Hox H U H + Hox ox(aq) 2 (a) – 37.871 7.991 – 3.300 – 1.400 1.773 0.030 ± ± ± 0.171 0.500 ± . H 2H ox U ox(aq) O(l) + H 2H O(cr) H ox(aq) 2 2 2 2 2 35.280 0.130 ± 2– 2+ Ni + ox Ni(ox)(aq) U Ni(ox)(aq) (a) 99.362 – 29.625 0.000 5.190 0.228 0.040 ± 1.264 ± ± 0.300 ± 2– 2– 2+ 2– + 2 ox U Ni(ox) Ni 2 Ni(ox) 2 (a) 120.105 – 43.609 – 7.800 7.640 ± ± ± 1.676 0.400 0.300 0.070 ± 2– 3+ + + Am U Am(ox) + ox Am(ox) – 37.159 6.510 0.150 ± 0.856 ± 2– – 3+ – + 2 ox Am(ox) U Am 2 Am(ox) 2 – 61.133 10.710 0.200 1.142 ± ± 2– 3– 3+ 3– U + 3 ox Am(ox) Am 3 Am(ox) 3 – 74.205 13.000 ± 1.000 ± 5.708 – 2– + – NpO U ox + ox NpO 2 2 NpO ox 2 – 22.261 3.900 ± 0.571 ± 0.100 (Continued on next page)

88 III. Selected data for organic compounds and complexes 46 Table III–2 (continued) Reaction Species ο ο ο ο ∆ H G ∆ ∆ S log K rm rm 10 rm –1 –1 –1 –1 (J · K · mol ) ) )(kJ · mol (kJ · mol 2– 3– + 3– (ox) NpO + 2 ox U NpO 2 2 2 NpO (ox) 2 2 – 33.107 5.800 ± 0.200 1.142 ± 2– 2+ UO ox(aq) UO U + ox UO ox(aq) 2 2 2 – 40.698 7.130 ± 0.160 ± 0.913 . . UO O(cr) ox UO U 3H O(l) ox(aq) + 3H 2 2 2 2 UO ox 3H O(cr) 2 2 (a) 1.800 – 10.274 – 20.200 – 33.290 3.500 ± 12.827 1.541 ± 0.270 ± ± 2– 2+ 2– 2– + 2 ox U UO (ox) UO 2 2 2 (ox) UO 2 2 – 66.499 11.650 0.856 ± 0.150 ± 2– 4– 2+ 4– + 3 ox U (ox) UO UO 2 3 2 UO (ox) 3 2 – 78.771 13.800 8.562 ± 1.500 ± 2– 2+ Mg U + ox Mg(ox)(aq) Mg(ox)(aq) – 20.321 3.560 ± 0.040 ± 0.228 2– 2– 2+ 2– + 2 ox Mg(ox) U Mg 2 Mg(ox) 2 – 29.511 5.170 0.457 ± ± 0.080 2– 2+ . . Ca U + H O(cr) O(l) + ox H Ca(ox) 2 2 H Ca(ox) O(cr) 2 (a) 8.730 95.023 – 21.500 – 49.831 0.500 ± 0.342 ± 2.033 ± ± 0.060 2– 2+ . . Ca O(l) + ox + 2H 2H O(cr) U Ca(ox) 2 2 Ca(ox) 2H O(cr) 2 (a) 8.300 – 47.377 – 25.200 74.381 ± ± ± 0.342 ± 3.864 0.060 1.100 2– 2+ . . Ca 3H O(l) + ox O(cr) U Ca(ox) + 3H 2 2 Ca(ox) 3H O(cr) 2 (a) 8.190 57.182 – 46.749 – 29.700 0.040 ± 4.427 ± ± 1.300 0.228 ± 2– 2+ Ca + ox Ca(ox)(aq) U Ca(ox)(aq) – 18.209 3.190 ± 0.060 ± 0.342 2– 2– 2– Ca(ox) U Ca(ox)(aq) + ox 2 Ca(ox) 2 – 4.738 0.830 0.190 ± 1.085 ± οο ο ∆=∆−∆ (a) Value calculated internally using . GHTS rm r m rm

89 III. Selected data for organic compounds and complexes 47 Table III–3: Selected thermodynamic data for compounds and complexes with citrate. All ionic species listed in this table are aqueous species. Unless noted otherwise, all data i.e. refer to the reference temperature of , a pressure 298.15 K and to the standard state, of 0.1 MPa and, for aqueous species, infinite dilution ( I = 0). The uncertainties listed below each value represent total uncertainties and correspond in principle to the statisti- cally defined 95% confidence interval. Values obtained from internal calculation, cf . footnote (b), are rounded at the third digit after the decimal point and may therefore not l the values are pre- III. Systematically, al be exactly identical to those given in Part sented with three digits after the decimal point, regardless of the significance of these digits. The data presented in this table are available on computer media from the OECD Nuclear Energy Agency. ο ο ο ο S ∆ G H ∆ C fm m fm p ,m Compound –1 –1 –1 –1 –1 –1 (kJ · mol · mol (J · K ) ) · mol (J · K ) ) (kJ · mol (b) (b) (b) 3– – 1162.258 75.587 – 1519.920 cit 1.855 ± 2.014 2.070 ± ± (b) (b) (b) 2– – 1198.561 208.417 – 1516.620 Hcit ± 2.048 1.510 ± 2.011 ± (b) (b) (b) – – 1225.845 – 1519.020 291.879 H cit 2 1.110 ± ± 2.026 2.010 ± (b) 225.400 – 1541.730 252.100 cit(cr) – 1236.695 H 3 ± ± 2.019 2.010 0.200 ± 0.200 ± (b) (b) (b) H – 1523.520 cit(aq) 336.710 – 1243.712 3 ± 0.428 2.009 ± 2.003 ± 268.050 283.600 – 1838.460 – 1473.272 O(cr) cit ·H H 3 2 2.000 0.200 2.009 ± ± 0.100 ± ± (b) – – 1246.617 Ni(cit) 2.204 ± (b) 4– – 2418.807 Ni(cit) 2 4.694 ± (b) – 1268.079 Ni(Hcit)(aq) ± 2.228 + (b) – 1283.320 cit) Ni(H 2 2.583 ± (b) Am(cit)(aq) – 1809.759 5.288 ± (b) 3– – 3002.555 Am(cit) 2 8.451 ± + (b) – 1834.361 Am(Hcit) 7.696 ± (Continued on next page)

90 III. Selected data for organic compounds and complexes 48 Table III–3 (continued) ο ο ο ο ∆ ∆ S G H C fm fm m ,m p Compound –1 –1 –1 –1 –1 –1 (kJ · mol · mol ) ) (J · K · mol (J · K ) ) (kJ · mol (b) – – 3057.466 Am(Hcit) 2 ± 8.448 (b) 2– – 2091.029 cit NpO 2 ± 5.984 (b) – – 2165.953 UO cit 2 ± 2.838 (b) 2– – 4351.198 (cit) ) (UO 2 2 2 6.049 ± (b) (Hcit)(aq) – 2179.652 UO 2 6.299 ± (b) – – 1645.089 Mg(cit) 2.423 ± (b) – 1668.777 Mg(Hcit)(aq) ± 2.447 + (b) – 1688.698 cit) Mg(H 2 2.580 ± (b) (b) (b) – – 1742.463 – 2062.920 111.282 Ca(cit) ± 20.242 2.278 ± 6.425 ± (b) – 1768.035 Ca(Hcit)(aq) 2.304 ± + (b) – 1787.385 cit) Ca(H 2 2.445 ± (b) . – 5033.670 4H Ca O(cr) (cit) 3 2 2 ± 5.149 (b) Value calculated internally from reaction data (see Table III–4).

91 III. Selected data for organic compounds and complexes 49 Table III–4: Selected thermodynamic data for reactions involving citrate compounds and complexes. All ionic species listed in this table are aqueous species. Unless noted mperature of 298.15 K and to the standard otherwise, all data refer to the reference te I = 0). The state, i.e. , a pressure of 0.1 MPa and, for aqueous species, infinite dilution ( ent total uncertainties and correspond in uncertainties listed below each value repres principal to the statistically defined 95% confidence interval. Values obtained from in- . footnote (a), are rounded at the third digit after the decimal point ternal calculation, cf and may therefore not be exactly identical to those given in Part III. Systematically, all the values are presented with three digits after the decimal point, regardless of the sig- nificance of these digits. The data presented in this table are available on computer me- dia from the OECD Nuclear Energy Agency. Reaction Species ο ο ο ο G ∆ ∆ S H ∆ log K rm rm rm 10 –1 –1 –1 –1 ) · mol )(kJ · mol ) (J · K (kJ · mol 3– 2– + 3– Hcit H + cit U cit (a) 36.303 – 132.830 – 3.300 – 6.360 ± 0.300 1.077 ± 0.020 ± ± 0.114 2– – + 2– H H U cit + Hcit 2 Hcit (a) 27.284 2.400 – 83.463 – 4.780 0.300 ± 0.010 ± 1.024 0.057 ± ± – + – H cit(aq) + H cit H U 3 2 H cit 2 (a) 17.866 – 44.830 4.500 – 3.130 ± 1.024 0.057 0.010 ± ± 0.300 ± . H cit cit(cr) + H H O(l) O(cr) U H H cit(cr) 3 3 2 2 3 10.900 ± 0.200 . H cit(aq) + H cit H U O(l) H O(cr) H cit(aq) 2 3 2 3 3 (a) 1.328 29.110 123.060 – 7.580 ± ± 0.023 ± 0.377 0.004 0.110 ± – 3– 2+ – Ni + cit U Ni(cit) Ni(cit) – 38.586 6.760 ± 0.080 ± 0.457 3– 4– 2+ 4– Ni(cit) + 2 cit U Ni 2 Ni(cit) 2 – 48.518 8.500 ± 2.283 ± 0.400 2– 2+ Ni + Hcit U Ni(Hcit)(aq) Ni(Hcit)(aq) – 23.745 4.160 0.571 ± 0.100 ± (Continued on next page)

92 III. Selected data for organic compounds and complexes 50 Table III–4 (continued) Reaction Species ο ο ο ο ∆ ∆ G ∆ H S K log rm rm rm 10 –1 –1 –1 –1 )(kJ · mol · mol ) (J · K ) (kJ · mol – 2+ + + Ni + H cit U Ni(H cit) cit) Ni(H 2 2 2 – 11.701 2.050 ± 1.427 ± 0.250 3– 3+ Am U + cit Am(cit)(aq) Am(cit)(aq) – 48.804 8.550 1.142 ± ± 0.200 3– 3– 3+ 3– + 2 cit U Am(cit) Am 2 Am(cit) 2 – 79.342 13.900 5.708 ± ± 1.000 2– 3+ + + Am U Am(Hcit) + Hcit Am(Hcit) – 37.102 6.500 1.000 ± ± 5.708 2– – 3+ – + 2 Hcit Am(Hcit) U Am 2 Am(Hcit) 2 – 61.647 10.800 5.708 ± 1.000 ± 2– 3– + 2– NpO cit U + cit NpO 2 2 NpO cit 2 – 21.006 3.680 0.285 ± ± 0.050 – 3– 2+ – UO cit + cit U UO 2 2 UO cit 2 – 51.144 8.960 ± 0.970 ± 0.170 2+ 2– 3– 2– (cit) + 2 cit ) U (UO 2UO 2 2 2 2 (UO ) (cit) 2 2 2 – 121.581 21.300 0.500 ± ± 2.854 2– 2+ UO + Hcit UO (Hcit)(aq) U UO (Hcit)(aq) 2 2 2 – 28.540 5.000 5.708 ± 1.000 ± – 3– 2+ – Mg + cit U Mg(cit) Mg(cit) – 27.456 4.810 ± ± 0.030 0.171 2– 2+ Mg + Hcit Mg(Hcit)(aq) U Mg(Hcit)(aq) – 14.841 2.600 0.070 0.400 ± ± (Continued on next page)

93 III. Selected data for organic compounds and complexes 51 Table III–4 (continued) Species Reaction ο ο ο ο G H ∆ ∆ S ∆ log K rm rm rm 10 –1 –1 –1 –1 (J · K · mol )(kJ · mol ) ) (kJ · mol – + 2+ + Mg + H cit) cit U Mg(H cit) Mg(H 2 2 2 – 7.478 1.310 0.160 0.913 ± ± – 3– 2+ – Ca + cit U Ca(cit) Ca(cit) (a) – 27.399 0.000 91.895 4.800 0.171 ± ± 0.030 ± ± 6.000 20.132 2– 2+ Ca + Hcit U Ca(Hcit)(aq) Ca(Hcit)(aq) – 16.667 2.920 0.070 ± ± 0.400 – 2+ + + Ca cit) + H cit U Ca(H Ca(H cit) 2 2 2 – 8.733 1.530 ± 0.160 ± 0.913 3– 2+ . . 4H 3Ca O(l) + 2 cit O(cr) (cit) + 4H U Ca 2 2 3 2 Ca 4H (cit) O(cr) 3 2 2 17.900 – 102.174 0.571 0.100 ± ± οο ο . (a) Value calculated internally using GHTS ∆=∆−∆ rm rm r m

94 III. Selected data for organic compounds and complexes 52 Table III–5: Selected thermodynamic data for compounds and complexes with edta. All ionic species listed in this table are aqueous species. Unless noted otherwise, all data 298.15 K and to the standard state, , a pressure refer to the reference temperature of i.e. of 0.1 MPa and, for aqueous species, infinite dilution ( I 0). The uncertainties listed = and correspond in principle to the statisti- below each value represent total uncertainties cf . cally defined 95% confidence interval. Values obtained from internal calculation, footnote (b), are rounded at the third digit after the decimal point and may therefore not l the values are pre- be exactly identical to those given in Part III. Systematically, al sented with three digits after the decimal point, regardless of the significance of these available on computer media from the OECD digits. The data presented in this table are Nuclear Energy Agency. ο ο ο ο S G ∆ ∆ H C p m fm ,m fm Compound –1 –1 –1 –1 –1 –1 (kJ · mol (J · K · mol ) ) (J · K · mol ) (kJ · mol ) (b) 4– – 1704.800 edta 3.751 ± (b) 3– – 1724.600 Hedta 3.718 ± (b) 2– – 1739.800 edta H 2 3.696 ± (b) – – 1732.700 H edta 3 ± 3.674 – 1759.800 edta(cr) H 4 1.500 ± (b) – 1730.800 edta(aq) H 4 3.354 ± (b) 2– – 1785.912 Ni(edta) 3.873 ± (b) – – 2332.100 Am(edta) 4.508 ± (b) – – 2305.290 Pu(edta) 4.401 ± (b) 2– – 2152.000 Mg(edta) 3.820 ± (b) 2– – 2270.000 Ca(edta) 3.903 ± (b) 3– – 1949.140 Na(edta) ± 4.803 (b) Value calculated internally from reaction data (see Table III–6).

95 III. Selected data for organic compounds and complexes 53 Table III–6: Selected thermodynamic data for reactions involving edta compounds and complexes. All ionic species listed in this table are aqueous species. Unless noted oth- ure of 298.15 K and to the standard state, erwise, all data refer to the reference temperat = 0). The uncer- i.e. I aqueous species, infinite dilution ( , a pressure of 0.1 MPa and, for l uncertainties and correspond in principal tainties listed below each value represent tota to the statistically defined 95% confidence interval. Values obtained from internal cal- . footnote (a), are rounded at the third digit after the decimal point and may cf culation, therefore not be exactly identical to those given in Part III. Systematically, all the values are presented with three digits after the decimal point, regardless of the significance of these digits. The data presented in this ta ble are available on computer media from the OECD Nuclear Energy Agency. Reaction Species ο ο ο ο ∆ ∆ G ∆ S H K log rm rm rm 10 –1 –1 –1 –1 ) · mol )(kJ · mol ) (J · K (kJ · mol 4– 3– + 4– Hedta U H + edta edta (a) 64.158 – 148.779 19.800 – 11.240 0.500 ± ± 1.773 ± 0.030 ± 0.171 2– 3– + 3– H U H edta + Hedta 2 Hedta (a) – 79.204 38.815 15.200 – 6.800 ± ± ± 1.395 0.114 0.400 0.020 ± 2– – + 2– H edta edta U H + H 3 2 H edta 2 (a) – 84.120 – 7.100 17.980 – 3.150 0.114 ± ± ± 1.395 ± 0.400 0.020 – + – H edta(aq) U + H edta H 4 3 H edta 3 (a) – 49.066 12.729 – 1.900 – 2.230 0.050 ± 5.121 ± 0.285 1.500 ± ± H edta(aq) H edta(cr) U H edta(aq) 4 4 4 (a) 21.691 29.000 24.516 – 3.800 1.085 ± ± 10.699 0.190 ± 3.000 ± + + + H U + H edta H edta(aq) H edta 4 5 5 – 7.420 1.300 0.100 0.571 ± ± 2+ + + 2+ H + H edta U H edta H edta 6 5 6 2.854 – 0.500 0.200 ± ± 1.142 2– 4– 2+ 2– Ni + edta U Ni(edta) Ni(edta) (a) 305.696 – 117.243 – 26.100 20.540 2.827 0.400 ± 0.130 ± ± ± 0.742 (Continued on next page)

96 III. Selected data for organic compounds and complexes 54 Table III–6 (continued) Reaction Species ο ο ο ο ∆ ∆ H G ∆ S log K 10 rm rm rm –1 –1 –1 –1 ) · mol )(kJ · mol ) (J · K (kJ · mol – 2– + – Ni(edta) + H U Ni(Hedta) Ni(Hedta) – 20.891 3.660 0.913 0.160 ± ± – 4– 3+ – Am U Am(edta) + edta Am(edta) (a) 341.027 – 112.277 – 10.600 19.670 0.628 ± 7.031 ± 0.110 ± ± 2.000 – + Am(edta) U Am(Hedta)(aq) + H Am(Hedta)(aq) – 12.386 2.170 1.427 ± ± 0.250 4– – 3+ – Pu U Pu(edta) + edta Pu(edta) (a) – 115.188 357.163 – 8.700 20.180 ± ± 8.147 0.370 ± ± 1.200 2.112 – + Pu(edta) + H U Pu(Hedta)(aq) Pu(Hedta)(aq) – 10.503 1.840 1.484 0.260 ± ± 4– 4+ Np + edta Np(edta)(aq) U Np(edta)(aq) – 178.091 31.200 3.425 ± ± 0.600 3– 4– + 3– NpO edta + edta U NpO 2 2 NpO edta 2 – 52.685 9.230 0.742 ± ± 0.130 + 2– 3– 2– + Hedta NpO U (Hedta) NpO 2 2 NpO (Hedta) 2 – 33.221 5.820 0.110 ± ± 0.628 – 2– + – NpO U edta) + H (H edta NpO 2 2 2 2 NpO (H edta) 2 2 – 25.515 4.470 ± 0.140 ± 0.799 4– 4+ U + edta U Uedta(aq) Uedta(aq) – 168.387 29.500 1.142 ± 0.200 ± 2– 4– 2+ 2– UO UO edta + edta U 2 2 UO edta 2 – 78.200 13.700 1.142 ± ± 0.200 (Continued on next page)

97 III. Selected data for organic compounds and complexes 55 Table III–6 (continued) Species Reaction ο ο ο ο S ∆ H ∆ ∆ G log K rm rm rm 10 –1 –1 –1 –1 )(kJ · mol (J · K · mol ) ) (kJ · mol 4– 2+ 2UO edta(aq) + edta U (UO ) edta(aq) ) (UO 2 2 2 2 2 – 117.586 20.600 ± 0.400 ± 2.283 3– – 2+ – UO UO + Hedta (Hedta) U 2 2 UO (Hedta) 2 – 47.776 8.370 0.571 ± ± 0.100 2– 4– 2+ 2– Mg + edta U Mg(edta) Mg(edta) (a) – 62.218 19.800 275.089 10.900 ± ± 0.571 ± 2.338 0.100 0.400 ± – 2– + – Mg(edta) + H Mg(Hedta) U Mg(Hedta) – 25.686 4.500 ± 1.142 ± 0.200 4– 2– 2+ 2– Ca + edta Ca(edta) U Ca(edta) (a) 168.489 – 22.200 – 72.435 12.690 0.060 ± ± 1.766 0.342 ± 0.400 ± 2– – + – Ca(edta) U Ca(Hedta) + H Ca(Hedta) – 20.206 3.540 0.090 ± ± 0.514 3– 4– + 3– Na + edta U Na(edta) Na(edta) (a) 40.190 – 15.983 – 4.000 2.800 ± 0.200 ± 10.766 1.142 ± 3.000 ± 3– 4– + 3– U + edta Kedta K Kedta – 10.274 1.800 ± 0.300 ± 1.712 ο οο (a) Value calculated internally using GHTS ∆=∆−∆ . rm rm r m

98 III. Selected data for organic compounds and complexes 56 Table III–7: Selected thermodynamic data for reactions involving isa compounds and table are aqueous species. Unless noted oth- complexes. All ionic species listed in this ure of 298.15 K and to the standard state, erwise, all data refer to the reference temperat , a pressure of 0.1 MPa and, for I = 0). The uncer- aqueous species, infinite dilution ( i.e. tainties listed below each value represent tota l uncertainties and correspond in principal to the statistically defined 95% confidence interval. Systematically, all the values are point, regardless of the significance of these presented with three digits after the decimal available on computer media from the OECD digits. The data presented in this table are Nuclear Energy Agency Species Reaction ο ο ο ο ∆ H G S ∆ ∆ log K rm rm rm 10 –1 –1 –1 –1 (J · K ) )(kJ · mol ) · mol (kJ · mol – + (1) H U Hisa(aq)* + isa Hisa(aq)* – 22.832 4.000 2.854 ± 0.500 ± – + 2+ + Ca + isa U Ca(isa) Ca(isa) – 9.704 1.700 1.712 0.300 ± ± – 2+ + Ca U + isa Ca(isa )(aq) + H )(aq) Ca(isa –H –H 59.364 – 10.400 0.500 ± ± 2.854 – 2+ Ca + 2 isa U Ca(isa) (cr) (cr) Ca(isa) 2 2 – 36.531 6.400 1.142 0.200 ± ± (1) This reaction refers to a "composite" protonati on equilibrium where [Hisa(aq)*] = [Hisa(aq)] + (aq)], see discussion in Section IX.3.2. [isa L

99 Chapter IV IV Selected auxiliary data This chapter presents the chemical thermodynamic data for auxiliary compounds and complexes which are used within the NEA TDB Project. Most of these auxiliary species are used in the evaluation of the recommended data reported in Chapter III. Addition- ally, data for species of Ni, U, Np, Pu an d Am, selected in the corresponding NEA TDB Reviews, have also been included in Table IV–1 for the convenience of the reader. It is therefore essential to always use these au xiliary data in conjunction with the selected data for the organic ligands reviewed in this volume. The use of other auxiliary data can lead to inconsistencies and erroneous results. The values in the tables of this chapter for auxiliary compounds and complexes are either CODATA Key Values, taken from [89COX/WAG] , or were evaluated within the NEA TDB project, as described in the corresponding chapters of the uranium review [92GRE/FUG] , the technetium review [99RAR/RAN] , the neptunium and plutonium review [2001LEM/FUG] , the Update review [2003GUI/FAN] , the nickel review [2005GAM/BUG] . [2005OLI/NOL] and the selenium review Table IV–1 contains the selected thermodynamic data of the auxiliary species ta of chemical reactions involving auxil- and Table IV–2 the selected thermodynamic da iary species. The reason for listing both reaction data and entropies, enthalpies and Gibbs energies of formation is, as described in Chapter III, that uncertainties in reaction ο ο ο data are often smaller than the derived H ∆ , S G and ∆ , due to uncertainty accu- fm fm m mulation during the calculations. All data in Table IV–1 and Table IV–2 refer to a temperature of 298.15 K, the standard state pressure of 0.1 MPa and, for aqueous species and reactions, to the infinite dilution standard state ( I = 0). The uncertainties listed below each reac tion value in Table IV–2 are total un- atistically defined 95% confidence interval. certainties, and correspond mainly to the st The uncertainties listed below each value in Table IV–1 have the following signifi- cance: 57

100 IV. Selected auxiliary data 58 [89COX/WAG] , the ± terms have the meaning: “it • for CODATA values from is probable, but not at all certain, that the true values of the thermodynamic quantities differ from the recommended values given in this report by no ± terms attached to the recommended values”. more than twice the • for values from [92GRE/FUG] , [99RAR/RAN] , [2003GUI/FAN] , [2005GAM/BUG] and [2005OLI/NOL] , the ± terms are derived from total uncertainties in the corresponding e . Table cf quilibrium constant of reaction ( IV–2), and from the ± terms listed for the necessary CODATA key values. 2 − − HCO CO CODATA [89COX/WAG] (g), , values are available for CO , 2 3 3 ο ο ο − − 2 HPO and and, H HPO ∆ and ∆ S . From the values given for the values of G fm 24 m fm 4 consequently, all the relevant equilibrium constants and enthalpy changes can be calcu- lated. The propagation of errors during this procedure, however, leads to uncertainties in the resulting equilibrium constants that are significantly higher than those obtained from experimental determination of the constants. Therefore, reaction data for CO (g), 2 − − 2 HCO CO , which were absent form the corresponding Table IV–2 in , 3 3 [92GRE/FUG] , are included in this volume to provide the user of selected data for the organic ligands ( cf . Chapter III) with the data needed to obtain the lowest possible un- certainties on reaction properties. Note that the values in Table IV–1 and Table IV–2 may contain more digits [89COX/WAG] than those listed in either or in the chapters devoted to data selection in [92GRE/FUG] [95SIL/BID] , [2001LEM/FUG] , [2003GUI/FAN] , [2005GAM/BUG] , and [2005OLI/NOL] , because the data in the present chapter are retrieved directly from the computerised data base and rounded to three digits after the decimal point through- out.

101 IV Selected auxiliary data 59 Table IV–1: Selected thermodynamic data for auxiliary compounds and complexes [89COX/WAG] adopted in the NEA TDB project, including the CODATA Key Values and NEA TDB data for species of Ni, U, Np, Pu and Am needed for the calculation of the values in tables III-1, III-3 and III-5. A ll ionic species listed in this table are aqueous species. Unless noted otherwise, all data refer to 298.15 K and a pressure of 0.1 MPa and, for aqueous species, a standard state of infinite dilution ( I = 0). The uncertainties listed below each value represent total uncertainties and correspond in principle to the statistically defined 95% confidence interv al. Values in bold typeface are CODATA without further evaluation. Key Values and are taken directly from Ref. [89COX/WAG] . footnotes (a) and (b), are rounded at the Values obtained from internal calculation, cf third digit after the decimal point and may therefore not be exactly identical to those , [95SIL/BID] , given in the chapters devoted to data selection in [92GRE/FUG] , [2003GUI/FAN] , [2005GAM/BUG] and [2005OLI/NOL] . System- [2001LEM/FUG] atically, all the values are presented with th ree digits after the decimal point, regardless rence listed for each entry in this table indi- of the significance of these digits. The refe cates the NEA TDB Review where the corresponding data have been adopted as NEA TDB Auxiliary data. The data presented in th is table are available on computer media from the OECD Nuclear Energy Agency. ο ο ο ο G ∆ S H ∆ C Compound and p fm m fm ,m –1 –1 –1 –1 –1 –1 review where adopted · mol (J · K ) (J · K ) (kJ · mol ) · mol ) (kJ · mol (a) O(g) 21.912 161.059 249.180 231.743 [92GRE/FUG] 0.001 0.003 0.100 ± ± 0.100 ± ± O (g) 29.378 0.000 0.000 205.152 2 [92GRE/FUG] ± ± 0.003 0.005 (a) H(g) 20.786 114.717 217.998 203.276 [92GRE/FUG] ± 0.006 ± 0.002 ± ± 0.006 0.001 + H 0.000 0.000 0.000 0.000 [92GRE/FUG] (g) H 28.836 130.680 0.000 0.000 2 [92GRE/FUG] 0.003 0.002 ± ± (a) – – 10.900 – 230.015 – 157.220 OH 0.040 ± 0.072 ± 0.200 ± [92GRE/FUG] (a) H O(g) 33.609 188.835 – 241.826 – 228.582 2 [92GRE/FUG] 0.040 ± 0.010 0.030 ± ± 0.040 ± (a) H O(l) 75.351 69.950 – 285.830 – 237.140 2 [92GRE/FUG] ± 0.030 0.040 ± ± ± 0.041 0.080 H – 191.170 (aq) O 2 2 [92GRE/FUG] ± 0.100 He(g) 20.786 0.000 0.000 126.153 [92GRE/FUG] ± 0.002 ± 0.001 Ne(g) 20.786 0.000 146.328 0.000 [92GRE/FUG] ± 0.003 ± 0.001 (Continued on next page)

102 IV. Selected auxiliary data 60 Table IV–1: (continued) ο ο ο ο H ∆ S ∆ G C Compound and fm m fm p ,m –1 –1 –1 –1 –1 –1 review where adopted · mol ) (J · K ) · mol ) (kJ · mol ) (J · K (kJ · mol Ar(g) 20.786 0.000 0.000 154.846 [92GRE/FUG] ± 0.001 ± 0.003 Kr(g) 20.786 0.000 164.085 0.000 [92GRE/FUG] 0.001 0.003 ± ± Xe(g) 20.786 0.000 169.685 0.000 [92GRE/FUG] ± 0.001 ± 0.003 (a) F(g) 22.746 158.751 79.380 62.280 [92GRE/FUG] 0.300 ± 0.004 ± 0.300 ± 0.002 ± (a) – – 13.800 – 335.350 – 281.523 F ± 0.692 ± 0.800 ± 0.650 [92GRE/FUG] F (g) 31.304 202.791 0.000 0.000 2 [92GRE/FUG] 0.005 0.002 ± ± (a) HF(aq) 88.000 – 323.150 – 299.675 [92GRE/FUG] ± 3.362 0.702 0.716 ± ± (a) HF(g) 173.779 29.137 – 273.300 – 275.400 [92GRE/FUG] ± 0.003 ± 0.700 ± 0.002 ± 0.700 (a) – – 655.500 92.683 – 583.709 HF 2 2.221 ± 1.200 ± 8.469 ± [92GRE/FUG] (a) Cl(g) 165.190 21.838 121.301 105.305 [92GRE/FUG] 0.004 ± 0.001 ± ± 0.008 ± 0.008 (a) – 56.600 – 167.080 – 131.217 Cl ± ± 0.200 ± 0.117 0.100 [92GRE/FUG] Cl (g) 33.949 0.000 223.081 0.000 2 [92GRE/FUG] 0.002 0.010 ± ± – – 37.669 ClO 0.962 ± [92GRE/FUG] – 10.250 ClO 2 ± 4.044 [92GRE/FUG] (a) – 162.300 – 7.903 – 104.000 ClO 3 ± 1.000 ± ± 3.000 1.342 [92GRE/FUG] (a) – 184.000 – 128.100 – 7.890 ClO 4 ± 0.400 0.600 ± 1.500 ± [92GRE/FUG] (a) HCl(g) 186.902 29.136 – 92.310 – 95.298 [92GRE/FUG] ± 0.005 ± 0.100 ± 0.002 ± 0.100 HClO(aq) – 80.023 [92GRE/FUG] 0.613 ± – 0.938 HClO (aq) 2 [92GRE/FUG] ± 4.043 (Continued on next page)

103 IV Selected auxiliary data 61 Table IV–1: (continued) ο ο ο ο H S G ∆ ∆ C Compound and m fm fm p ,m –1 –1 –1 –1 –1 –1 review where adopted ) ) (J · K ) (kJ · mol · mol · mol ) (J · K (kJ · mol (a) Br(g) 175.018 20.786 111.870 82.379 [92GRE/FUG] ± 0.004 ± 0.120 ± 0.001 ± 0.128 (a) – 82.550 – 103.850 – 121.410 Br ± 0.167 ± 0.200 ± 0.150 [92GRE/FUG] Br 4.900 (aq) 2 [92GRE/FUG] 1.000 ± (a) Br (g) 245.468 36.057 30.910 3.105 2 [92GRE/FUG] 0.110 ± ± 0.142 ± 0.002 ± 0.005 Br (l) 0.000 0.000 152.210 2 [92GRE/FUG] 0.300 ± – – 32.095 BrO ± 1.537 [92GRE/FUG] (a) – 161.500 19.070 – 66.700 BrO 3 0.500 ± 0.634 ± 1.300 ± [92GRE/FUG] (a) HBr(g) 29.141 198.700 – 36.290 – 53.361 [92GRE/FUG] 0.166 ± 0.004 ± 0.160 ± 0.003 ± (b) HBrO(aq) – 81.356 [92GRE/FUG] 1.527 ± (a) I(g) 180.787 20.786 106.760 70.172 [92GRE/FUG] ± 0.004 ± ± 0.060 0.040 ± 0.001 (a) – 106.450 – 56.780 – 51.724 I ± 0.300 ± 0.050 ± 0.112 [92GRE/FUG] I (cr) 0.000 0.000 116.140 2 [92GRE/FUG] 0.300 ± (a) I (g) 36.888 260.687 62.420 19.323 2 [92GRE/FUG] 0.005 ± ± ± 0.002 0.120 0.080 ± (a) – 118.000 – 126.338 – 219.700 IO 3 ± 0.500 0.779 ± ± 2.000 [92GRE/FUG] (a) HI(g) 206.590 29.157 26.500 1.700 [92GRE/FUG] ± 0.004 ± ± 0.100 0.003 ± 0.110 HIO – 130.836 (aq) 3 [92GRE/FUG] 0.797 ± S(cr)(orthorhombic) 22.750 0.000 32.054 0.000 [92GRE/FUG] 0.050 ± 0.050 ± (a) S(g) 23.674 167.829 277.170 236.689 [92GRE/FUG] 0.151 ± 0.006 0.150 ± 0.001 ± ± 2– 120.695 S 11.610 ± [92GRE/FUG] (Continued on next page)

104 IV. Selected auxiliary data 62 Table IV–1: (continued) ο ο ο ο ∆ H ∆ G S C Compound and m fm fm p ,m –1 –1 –1 –1 –1 –1 review where adopted ) · mol (J · K (J · K · mol ) ) ) (kJ · mol (kJ · mol (a) S (g) 32.505 228.167 128.600 79.686 2 [92GRE/FUG] ± 0.010 ± 0.301 ± 0.010 ± 0.300 (a) (g) SO 248.223 39.842 – 296.810 – 300.095 2 [92GRE/FUG] ± 0.050 ± 0.200 ± 0.020 ± 0.201 2– – 487.472 SO 3 4.020 ± [92GRE/FUG] 2– – 519.291 S O 2 3 ± 11.345 [92GRE/FUG] (a) 2– 18.500 – 909.340 – 744.004 SO 4 0.400 0.418 ± 0.400 ± ± [92GRE/FUG] (a) – 67.000 12.243 – 16.300 HS ± 2.115 1.500 ± 5.000 ± [92GRE/FUG] (a) H S(aq) 126.000 – 38.600 – 27.648 2 [92GRE/FUG] 2.115 1.500 ± 5.000 ± ± (a) H S(g) 34.248 205.810 – 20.600 – 33.443 2 [92GRE/FUG] ± 0.050 ± 0.500 ± 0.010 ± 0.500 – – 528.684 HSO 3 4.046 ± [92GRE/FUG] – – 528.366 O HS 3 2 ± 11.377 [92GRE/FUG] – 539.187 SO (aq) H 3 2 [92GRE/FUG] ± 4.072 (a) – 131.700 – 886.900 – 755.315 HSO 4 1.000 ± 1.342 ± 3.000 ± [92GRE/FUG] Te(cr) 25.550 49.221 0.000 0.000 [92GRE/FUG] 0.050 ± 0.100 ± (a) 69.890 60.670 – 265.996 – 321.000 (cr) TeO 2 [2003GUI/FAN] ± 2.500 ± 0.150 ± 2.500 ± 0.150 (a) N(g) 153.301 20.786 472.680 455.537 [92GRE/FUG] ± 0.400 ± ± 0.001 0.400 ± 0.003 N (g) 29.124 0.000 0.000 191.609 2 [92GRE/FUG] 0.001 ± 0.004 ± (a) – 107.710 275.140 348.200 N 3 ± 1.000 2.000 ± ± 7.500 [92GRE/FUG] (a) – 146.700 – 206.850 – 110.794 NO 3 ± 0.417 ± 0.400 ± 0.400 [92GRE/FUG] (Continued on next page)

105 IV Selected auxiliary data 63 Table IV–1: (continued) ο ο ο ο S H ∆ ∆ G C Compound and fm m fm ,m p –1 –1 –1 –1 –1 –1 review where adopted ) ) (J · K (J · K · mol ) (kJ · mol ) · mol (kJ · mol HN 147.381 260.140 321.372 (aq) 3 [92GRE/FUG] ± 34.403 ± 10.050 ± 2.051 – 81.170 109.040 – 26.673 (aq) NH 3 [92GRE/FUG] 0.305 ± 0.913 ± 0.326 ± (a) NH (g) 35.630 192.770 – 45.940 – 16.407 3 [92GRE/FUG] 0.005 0.350 ± ± ± 0.350 ± 0.050 (a) + 111.170 – 133.260 – 79.398 NH 4 0.400 ± ± 0.278 0.250 ± [92GRE/FUG] P(am)(red) – 7.500 [92GRE/FUG] 2.000 ± P(cr)(white, cubic) 23.824 0.000 41.090 0.000 [92GRE/FUG] 0.250 ± 0.200 ± (a) P(g) 20.786 163.199 316.500 280.093 [92GRE/FUG] 1.003 ± 0.003 0.001 ± ± 1.000 ± (a) P (g) 32.032 218.123 144.000 103.469 2 [92GRE/FUG] 2.006 ± 0.004 2.000 ± ± 0.002 ± (a) P (g) 67.081 280.010 58.900 24.419 4 [92GRE/FUG] 0.500 ± 0.448 ± 1.500 ± 0.300 ± 3– – 1025.491 – 220.970 – 1284.400 PO 4 ± ± 1.576 ± 12.846 4.085 [92GRE/FUG] 4– – 1935.503 O P 7 2 ± 4.563 [92GRE/FUG] (a) 2– – 33.500 – 1299.000 – 1095.985 HPO 4 1.500 ± 1.567 ± 1.500 ± [92GRE/FUG] (a) – 92.500 – 1137.152 – 1302.600 H PO 2 4 ± 1.500 1.567 ± ± 1.500 [92GRE/FUG] 161.912 – 1294.120 H – 1149.367 PO (aq) 4 3 [92GRE/FUG] ± 2.575 1.576 1.616 ± ± 3– – 1989.158 O HP 7 2 ± 4.482 [92GRE/FUG] 2– – 2027.117 H O P 2 7 2 4.445 ± [92GRE/FUG] – – 2039.960 O P H 2 3 7 4.362 ± [92GRE/FUG] 274.919 – 2280.210 – 2045.668 H (aq) P O 7 4 2 [92GRE/FUG] ± 3.299 ± ± 6.954 3.383 24.640 0.000 0.000 35.100 As(cr) [92GRE/FUG] ± 0.600 ± 0.500 (Continued on next page)

106 IV. Selected auxiliary data 64 Table IV–1: (continued) ο ο ο ο ∆ ∆ G S H C Compound and m fm fm p ,m –1 –1 –1 –1 –1 –1 review where adopted · mol ) (J · K ) · mol ) (kJ · mol ) (J · K (kJ · mol (a) – 40.600 – 350.022 – 429.030 AsO 2 4.008 ± 0.600 4.000 ± ± [92GRE/FUG] (a) 3– – 162.800 – 648.360 – 888.140 AsO 4 ± 4.008 4.000 ± 0.600 ± [92GRE/FUG] (a) 116.520 105.400 – 782.449 As – 924.870 (cr) O 5 2 [92GRE/FUG] 8.016 ± 1.200 0.800 ± ± 8.000 ± (a) 214.200 191.290 As – 1152.445 O – 1313.940 (cubic) 4 6 [92GRE/FUG] ± 16.032 ± ± 0.800 16.000 ± 2.400 (a) 234.000 – 1154.009 – 1309.600 O (monoclinic) As 6 4 [92GRE/FUG] 16.000 ± 3.000 ± 16.041 ± (a) 408.600 – 1092.716 (g) O – 1196.250 As 6 4 [2005GAM/BUG] 16.000 ± 6.000 16.116 ± ± (a) 125.900 – 402.925 (aq) – 456.500 HAsO 2 [92GRE/FUG] ± ± 0.600 4.000 ± 4.008 (a) – 110.500 – 587.078 – 714.790 AsO H 3 2 ± ± 4.000 ± 4.008 0.600 [92GRE/FUG] (a) 195.000 – 639.681 AsO – 742.200 (aq) H 3 3 [92GRE/FUG] 1.000 ± ± ± 4.015 4.000 (a) 2– – 714.592 – 1.700 – 906.340 HAsO 4 ± 0.600 4.008 ± 4.000 ± [92GRE/FUG] (a) – 117.000 – 753.203 – 909.560 AsO H 4 2 4.015 ± ± 1.000 4.000 ± [92GRE/FUG] (a) 184.000 – 766.119 H – 902.500 AsO (aq) 3 4 [92GRE/FUG] ± 1.000 ± 4.000 ± 4.015 – 4248.400 . 5 H (As O O(cr) ) 3 5 2 2 ± 24.000 [92GRE/FUG] 25.260 0.000 0.000 45.520 Sb(cr) [92GRE/FUG] 0.200 0.210 ± ± 25.410 56.740 0.000 0.000 Bi(cr) [2001LEM/FUG] ± 0.200 0.420 ± C(cr) 8.517 5.740 0.000 0.000 [92GRE/FUG] 0.080 0.100 ± ± (a) C(g) 158.100 20.839 716.680 671.254 [92GRE/FUG] ± ± 0.003 ± 0.450 ± 0.001 0.451 (a) CO(g) 29.141 197.660 – 110.530 – 137.168 [92GRE/FUG] 0.002 ± ± ± ± 0.004 0.173 0.170 (a) CO (aq) 119.360 – 413.260 – 385.970 2 [92GRE/FUG] ± 0.200 ± 0.600 ± 0.270 (Continued on next page)

107 IV Selected auxiliary data 65 Table IV–1: (continued) ο ο ο ο G S ∆ ∆ H C Compound and fm fm m ,m p –1 –1 –1 –1 –1 –1 review where adopted · mol · mol (J · K (J · K ) (kJ · mol ) ) ) (kJ · mol (a) CO (g) 37.135 213.785 – 393.510 – 394.373 2 [92GRE/FUG] ± 0.010 0.130 ± ± 0.002 0.133 ± (a) 2– – 50.000 – 675.230 – 527.900 CO 3 0.250 1.000 0.390 ± ± ± [92GRE/FUG] (a) – 98.400 – 689.930 – 586.845 HCO 3 ± 0.200 ± 0.251 ± 0.500 [92GRE/FUG] (b) (b) (b) – 166.939 147.350 101.182 CN ± 8.475 2.519 ± 3.541 ± [2005OLI/NOL] (b) (b) (b) 114.368 103.750 131.271 HCN(aq) [2005OLI/NOL] ± 3.536 ± 8.440 ± 2.517 (a) 201.710 119.517 129.900 HCN(g) [2005OLI/NOL] 2.500 2.500 ± 0.100 ± ± (a) – 144.268 76.400 92.700 SCN ± 4.000 18.974 4.000 ± ± [92GRE/FUG] Si(cr) 19.789 0.000 0.000 18.810 [92GRE/FUG] ± ± 0.030 0.080 (a) Si(g) 22.251 167.981 450.000 405.525 [92GRE/FUG] 0.004 8.000 8.000 ± 0.001 ± ± ± (a) SiO α – quartz) ( 44.602 41.460 – 910.700 – 856.287 2 [92GRE/FUG] 1.002 ± 0.200 ± ± 1.000 0.300 ± 2– – 1.488 – 1381.960 – 1175.651 SiO (OH) 2 2 15.330 ± ± ± 51.592 1.265 [92GRE/FUG] – 88.024 – 1431.360 – 1251.740 SiO(OH) 3 ± 3.743 1.162 ± ± 13.144 [92GRE/FUG] (b) (b) (b) – 1307.735 189.973 – 1456.960 (aq) Si(OH) 4 [92GRE/FUG] ± 3.163 ± 1.156 ± 11.296 2– – 2269.878 (OH) O Si 2 4 3 ± 2.878 [92GRE/FUG] – – 2332.096 O (OH) Si 2 5 2 ± 2.878 [92GRE/FUG] 3– – 3048.536 Si (OH) O 3 6 3 3.870 ± [92GRE/FUG] 3– – 3291.955 (OH) O Si 5 5 3 3.869 ± [92GRE/FUG] 4– – 4075.179 (OH) O Si 4 8 4 5.437 ± [92GRE/FUG] 3– – 4136.826 (OH) O Si 7 4 5 4.934 ± [92GRE/FUG] (Continued on next page)

108 IV. Selected auxiliary data 66 Table IV–1: (continued) ο ο ο ο H G S ∆ ∆ C Compound and fm m fm ,m p –1 –1 –1 –1 –1 –1 review where adopted · mol ) (J · K ) (kJ · mol · mol (J · K ) ) (kJ · mol (a) (g) SiF 73.622 282.760 – 1615.000 – 1572.773 4 [92GRE/FUG] ± 0.500 ± 0.800 ± 0.500 ± 0.814 Ge(cr) 23.222 0.000 31.090 0.000 [92GRE/FUG] ± 0.100 0.150 ± (a) Ge(g) 30.733 167.904 372.000 331.209 [92GRE/FUG] 0.005 ± ± 3.000 0.001 3.000 ± ± (a) GeO (tetragonal) 39.710 50.166 – 580.000 – 521.404 2 [92GRE/FUG] 1.000 ± ± 1.002 ± 0.300 ± 0.150 (a) GeF (g) 301.900 81.602 – 1190.200 – 1150.018 4 [92GRE/FUG] ± 1.000 ± 0.584 ± 1.000 ± 0.500 Sn(cr) 27.112 0.000 0.000 51.180 [92GRE/FUG] ± 0.030 ± 0.080 (a) Sn(g) 21.259 168.492 301.200 266.223 [92GRE/FUG] 0.004 1.500 1.500 ± 0.001 ± ± ± 2+ (a) Sn – 16.700 – 8.900 – 27.624 [92GRE/FUG] 1.000 ± 1.557 ± 4.000 ± (a) SnO(tetragonal) 47.783 57.170 – 280.710 – 251.913 [92GRE/FUG] 0.300 0.220 0.200 ± 0.300 ± ± ± (a) SnO (cassiterite, – 515.826 53.219 49.040 2 – 577.630 tetragonal) 0.200 ± 0.100 0.204 ± ± 0.200 ± [92GRE/FUG] Pb(cr) 26.650 0.000 0.000 64.800 [92GRE/FUG] ± 0.300 0.100 ± (a) Pb(g) 20.786 175.375 195.200 162.232 [92GRE/FUG] 0.005 ± 0.800 ± 0.001 ± 0.805 ± 2+ (a) Pb 18.500 0.920 – 24.238 [92GRE/FUG] ± ± 1.000 ± 0.399 0.250 (a) PbSO (cr) 148.500 – 919.970 – 813.036 4 [92GRE/FUG] ± ± 0.600 0.400 0.447 ± B(cr) 11.087 0.000 0.000 5.900 [92GRE/FUG] ± ± 0.080 0.100 (a) B(g) 20.796 153.436 565.000 521.012 [92GRE/FUG] ± ± 0.015 5.000 5.000 ± 0.005 ± (a) B (cr) O 53.970 62.761 – 1273.500 – 1194.324 3 2 [92GRE/FUG] 1.400 ± ± 0.300 ± ± 0.300 1.404 (a) B(OH) (aq) 162.400 – 1072.800 – 969.268 3 [92GRE/FUG] ± ± 0.820 0.600 ± 0.800 (a) B(OH) (cr) 86.060 89.950 – 1094.800 – 969.667 3 [92GRE/FUG] 0.820 0.800 ± 0.600 ± ± ± 0.400 (Continued on next page)

109 IV Selected auxiliary data 67 Table IV–1: (continued) ο ο ο ο ∆ ∆ G S H C Compound and m fm fm ,m p –1 –1 –1 –1 –1 –1 review where adopted · mol ) (J · K ) · mol (J · K ) ) (kJ · mol (kJ · mol (a) BF (g) 50.463 254.420 – 1136.000 – 1119.403 3 [92GRE/FUG] ± ± ± 0.100 ± 0.803 0.200 0.800 Al(cr) 24.200 0.000 0.000 28.300 [92GRE/FUG] 0.100 ± 0.070 ± (a) Al(g) 164.554 21.391 330.000 289.376 [92GRE/FUG] ± 0.004 ± 4.000 ± 0.001 ± 4.000 3+ (a) Al – 325.000 – 538.400 – 491.507 [92GRE/FUG] 1.500 10.000 ± 3.338 ± ± (a) Al O (corundum) 79.033 50.920 – 1675.700 – 1582.257 2 3 [92GRE/FUG] 0.100 ± 1.300 ± 0.200 ± ± 1.302 (a) AlF (cr) 75.122 66.500 – 1510.400 – 1431.096 3 [92GRE/FUG] 1.309 ± 0.500 ± 1.300 ± 0.400 ± + Tl – 32.400 [99RAR/RAN] 0.300 ± Zn(cr) 25.390 0.000 41.630 0.000 [92GRE/FUG] 0.150 ± 0.040 ± (a) Zn(g) 20.786 160.990 130.400 94.813 [92GRE/FUG] 0.004 0.001 ± ± 0.402 ± ± 0.400 2+ (a) Zn – 109.800 – 153.390 – 147.203 [92GRE/FUG] 0.254 0.200 ± ± ± 0.500 (a) ZnO(cr) 43.650 – 350.460 – 320.479 [92GRE/FUG] ± 0.400 0.270 ± 0.299 ± Cd(cr) 26.020 51.800 0.000 0.000 [92GRE/FUG] ± ± 0.150 0.040 (a) Cd(g) 20.786 167.749 111.800 77.230 [92GRE/FUG] 0.004 0.205 0.200 ± 0.001 ± ± ± 2+ (a) Cd – 72.800 – 75.920 – 77.733 [92GRE/FUG] 1.500 ± 0.600 ± 0.750 ± (a) CdO(cr) 54.800 – 258.350 – 228.661 [92GRE/FUG] ± ± 1.500 ± 0.400 0.602 (a) . – 1464.959 229.650 – 1729.300 CdSO 2.667 H O(cr) 4 2 ± ± 0.400 0.800 ± 0.810 [92GRE/FUG] (a) Hg(g) 20.786 174.971 61.380 31.842 [92GRE/FUG] 0.054 ± 0.005 ± ± 0.040 0.001 ± Hg(l) 0.000 0.000 75.900 [92GRE/FUG] 0.120 ± 2+ (a) Hg – 36.190 170.210 164.667 [92GRE/FUG] 0.313 ± ± 0.800 0.200 ± (Continued on next page)

110 IV. Selected auxiliary data 68 Table IV–1: (continued) ο ο ο ο ∆ ∆ G S H C Compound and m fm fm p ,m –1 –1 –1 –1 –1 –1 review where adopted · mol (J · K ) · mol ) ) (J · K ) (kJ · mol (kJ · mol (a) 2+ 65.740 166.870 153.567 Hg 2 ± 0.559 ± ± 0.800 0.500 [92GRE/FUG] (a) HgO(montroydite, red) 70.250 – 90.790 – 58.523 [92GRE/FUG] 0.154 ± 0.300 0.120 ± ± (a) Hg Cl (cr) 191.600 – 265.370 – 210.725 2 2 [92GRE/FUG] 0.471 ± 0.800 0.400 ± ± (a) Hg SO (cr) 200.700 – 743.090 – 625.780 4 2 [92GRE/FUG] 0.400 ± 0.200 ± 0.411 ± Cu(cr) 24.440 0.000 0.000 33.150 [92GRE/FUG] ± 0.050 ± 0.080 (a) Cu(g) 20.786 166.398 337.400 297.672 [92GRE/FUG] 1.200 0.004 ± 1.200 ± 0.001 ± ± 2+ (a) Cu – 98.000 64.900 65.040 [92GRE/FUG] 1.557 ± ± 4.000 1.000 ± CuCl(g) 76.800 [2003GUI/FAN] 10.000 ± (a) (cr) CuSO 109.200 – 771.400 – 662.185 4 [92GRE/FUG] 1.206 ± 0.400 1.200 ± ± Ag(cr) 25.350 0.000 42.550 0.000 [92GRE/FUG] ± 0.100 0.200 ± (a) Ag(g) 20.786 172.997 284.900 246.007 [92GRE/FUG] 0.004 ± ± 0.800 0.001 0.802 ± ± + (a) Ag 73.450 105.790 77.096 [92GRE/FUG] 0.156 ± 0.400 0.080 ± ± (a) AgCl(cr) 96.250 – 127.010 – 109.765 [92GRE/FUG] ± ± 0.200 ± 0.098 0.050 2+ (a) – 46.100 – 55.012 Ni – 45.773 – 131.800 [2005GAM/BUG] 0.771 ± 1.400 ± 0.878 ± 7.500 ± Ti(cr) 25.060 0.000 0.000 30.720 [92GRE/FUG] ± 0.100 0.080 ± (a) Ti(g) 24.430 180.298 473.000 428.403 [92GRE/FUG] ± ± 0.010 ± 3.000 ± 0.030 3.000 (a) TiO (rutile) 55.080 50.620 – 944.000 – 888.767 2 [92GRE/FUG] 0.800 ± 0.300 0.806 ± ± 0.300 ± (a) TiCl (g) 95.408 353.200 – 763.200 – 726.324 4 [92GRE/FUG] ± ± ± 4.000 3.229 3.000 ± 1.000 (Continued on next page)

111 IV Selected auxiliary data 69 Table IV–1: (continued) ο ο ο ο ∆ H ∆ G S C Compound and m fm fm p ,m –1 –1 –1 –1 –1 –1 review where adopted ) · mol (J · K (J · K · mol ) ) ) (kJ · mol (kJ · mol 3+ (a) Am –201.000 –598.698 –616.700 [95SIL/BID] ± ± 15.000 1.500 4.755 ± 3+ (b) –184.510 –591.790 –578.984 Pu [2001LEM/FUG] ± 2.688 ± ± 6.154 1.964 4+ (a) (b) –426.390 –491.774 –556.022 Np [2001LEM/FUG] 12.386 4.185 ± 5.586 ± ± (a) + –907.765 –4.000 –45.904 –978.181 NpO 2 4.629 ± 10.706 ± 5.628 ± 25.000 ± [2001LEM/FUG] 27.660 50.200 0.000 0.000 U(cr) [92GRE/FUG] ± 0.050 ± 0.200 4+ (a) (b) –416.895 –529.860 –591.200 –220.000 U [2003GUI/FAN] 50.000 ± 3.300 ± ± 1.765 12.553 ± (a) 2+ –98.200 42.400 –1019.000 –952.551 UO 2 3.000 1.500 ± ± 3.000 1.747 ± ± [92GRE/FUG] Th(cr) 26.230 51.800 0.000 0.000 [92GRE/FUG] 0.500 ± 0.050 ± (a) Th(g) 20.789 190.170 602.000 560.745 [92GRE/FUG] ± ± 0.100 ± 6.000 0.050 6.002 ± (a) ThO (cr) 65.230 – 1226.400 – 1169.238 2 [92GRE/FUG] ± 0.200 ± 3.500 ± 3.504 Be(cr) 16.443 9.500 0.000 0.000 [92GRE/FUG] 0.060 0.080 ± ± (a) Be(g) 20.786 136.275 324.000 286.202 [92GRE/FUG] 0.003 5.000 ± ± 0.001 5.000 ± ± (a) BeO(bromellite) 13.770 25.565 – 609.400 – 580.090 [92GRE/FUG] ± 0.040 ± ± ± 0.100 2.500 2.500 Mg(cr) 24.869 0.000 0.000 32.670 [92GRE/FUG] ± ± 0.100 0.020 (a) Mg(g) 20.786 148.648 147.100 112.521 [92GRE/FUG] 0.800 ± 0.801 ± 0.001 ± ± 0.003 2+ (a) Mg – 137.000 – 467.000 – 455.375 [92GRE/FUG] ± ± 4.000 ± 1.335 0.600 (a) MgO(cr) 37.237 26.950 – 601.600 – 569.312 [92GRE/FUG] 0.200 ± ± ± ± 0.150 0.305 0.300 (a) MgF (cr) 57.200 61.512 – 1124.200 – 1071.051 2 [92GRE/FUG] ± ± 0.500 ± 1.200 ± 0.300 1.210 Ca(cr) 25.929 0.000 41.590 0.000 [92GRE/FUG] 0.400 ± 0.300 ± (Continued on next page)

112 IV. Selected auxiliary data 70 Table IV–1: (continued) ο ο ο ο ∆ ∆ G S H C Compound and m fm fm p ,m –1 –1 –1 –1 –1 –1 review where adopted · mol ) (J · K ) · mol ) (kJ · mol ) (J · K (kJ · mol (a) Ca(g) 20.786 154.887 177.800 144.021 [92GRE/FUG] 0.004 ± 0.800 ± 0.001 ± ± 0.809 2+ (a) Ca – 56.200 – 543.000 – 552.806 [92GRE/FUG] ± ± 1.000 ± 1.000 1.050 (a) CaO(cr) 42.049 38.100 – 634.920 – 603.296 [92GRE/FUG] 0.916 ± 0.400 ± 0.900 ± 0.400 ± CaF(g) 33.671 – 276.404 229.244 – 302.118 [2003GUI/FAN] ± 0.500 5.100 ± ± 0.500 ± 5.104 35.687 241.634 – 103.400 – 129.787 CaCl(g) [2003GUI/FAN] ± 0.300 5.000 ± ± 0.010 5.001 ± 55.700 0.000 0.000 Sr(cr) [92GRE/FUG] ± 0.210 2+ (a) – 31.500 – 563.864 – 550.900 Sr [92GRE/FUG] ± 0.781 ± 2.000 ± 0.500 (a) 55.440 – 559.939 – 590.600 SrO(cr) [92GRE/FUG] ± 0.500 0.900 0.914 ± ± (a) 114.850 – 784.974 – 833.850 (cr) SrCl 2 [92GRE/FUG] ± ± 0.420 ± 0.700 0.714 (a) 194.600 – 783.146 (cr) – 982.360 ) Sr(NO 2 3 [92GRE/FUG] ± 1.018 0.800 2.100 ± ± 0.000 0.000 62.420 Ba(cr) [92GRE/FUG] 0.840 ± 20.786 170.245 185.000 152.852 Ba(g) [2003GUI/FAN] ± 0.010 5.006 ± ± 0.001 ± 5.000 (a) 2+ 8.400 – 557.656 – 534.800 Ba [92GRE/FUG] 2.500 ± 2.582 ± 2.000 ± (a) 72.070 – 520.394 – 548.100 BaO(cr) [92GRE/FUG] 2.500 0.380 ± 2.515 ± ± 34.747 246.219 – 324.992 – 349.569 BaF(g) [2003GUI/FAN] ± 0.210 ± ± ± 0.300 6.700 6.705 (a) 123.680 BaCl – 806.953 (cr) – 855.200 2 [92GRE/FUG] 2.514 ± ± 0.250 ± 2.500 Li(cr) 24.860 29.120 0.000 0.000 [92GRE/FUG] 0.200 0.200 ± ± (a) Li(g) 138.782 20.786 159.300 126.604 [92GRE/FUG] ± ± 0.010 1.002 ± ± 0.001 1.000 (Continued on next page)

113 IV Selected auxiliary data 71 Table IV–1: (continued) ο ο ο ο S H G ∆ ∆ C Compound and fm m fm p ,m –1 –1 –1 –1 –1 –1 review where adopted · mol ) (J · K (J · K · mol ) ) ) (kJ · mol (kJ · mol + (a) Li 12.240 – 278.470 – 292.918 [92GRE/FUG] 0.080 0.150 ± 0.109 ± ± Na(cr) 28.230 51.300 0.000 0.000 [92GRE/FUG] 0.200 ± 0.200 ± (a) Na(g) 153.718 20.786 107.500 76.964 [92GRE/FUG] ± 0.003 0.700 ± ± 0.001 ± 0.703 + (a) Na 58.450 – 240.340 – 261.953 [92GRE/FUG] 0.060 ± 0.150 0.096 ± ± (a) NaF(cr) 51.160 – 546.327 – 576.600 [2001LEM/FUG] ± 0.704 ± ± 0.150 0.700 50.500 72.150 – 411.260 – 384.221 NaCl(cr) [2001LEM/FUG] 0.120 ± 0.200 0.147 ± ± – 467.580 NaNO (cr) 3 [2003GUI/FAN] ± 0.410 K(cr) 29.600 0.000 0.000 64.680 [92GRE/FUG] 0.200 ± ± 0.100 (a) K(g) 160.341 20.786 89.000 60.479 [92GRE/FUG] 0.003 ± ± ± 0.001 0.802 0.800 ± + (a) K 101.200 – 252.140 – 282.510 [92GRE/FUG] 0.116 ± ± 0.200 ± 0.080 KCl(cr) – 436.461 [2005GAM/BUG] 0.129 ± – 393.330 KBr(cr) [2005GAM/BUG] ± 0.188 – 329.150 KI(cr) [2005GAM/BUG] ± 0.137 Rb(cr) 31.060 0.000 0.000 76.780 [92GRE/FUG] ± 0.100 0.300 ± (a) Rb(g) 20.786 170.094 80.900 53.078 [92GRE/FUG] ± 0.003 ± 0.800 ± 0.001 0.805 ± + (a) Rb 121.750 – 251.120 – 284.009 [92GRE/FUG] 0.153 ± ± 0.250 0.100 ± Cs(cr) 32.210 0.000 0.000 85.230 [92GRE/FUG] ± ± 0.200 0.400 (a) Cs(g) 175.601 20.786 76.500 49.556 [92GRE/FUG] ± 1.000 ± 0.003 ± 1.007 ± 0.001 (Continued on next page)

114 IV. Selected auxiliary data 72 Table IV–1: (continued) ο ο ο ο H ∆ ∆ S G C Compound and m fm fm ,m p –1 –1 –1 –1 –1 –1 review where adopted · mol · mol ) (kJ · mol ) (J · K ) ) (J · K (kJ · mol + (a) Cs 132.100 – 258.000 – 291.456 [92GRE/FUG] 0.500 ± ± ± 0.500 0.535 (a) 101.170 52.470 – 413.807 CsCl(cr) – 442.310 [2001LEM/FUG] 0.160 ± 0.200 0.208 ± ± 112.940 – 405.600 52.930 – 391.171 CsBr(cr) [2001LEM/FUG] 0.305 ± ± 0.400 ± 0.250 ο οο . (a) Value calculated internally using GHTS ∆=∆− ∑ f m m, i fm i (b) Value calculated internally from reaction data (see Table IV–2).

115 IV Selected auxiliary data 73 for reactions involving auxiliary compounds Table IV–2: Selected thermodynamic data and complexes used in the evaluation of thermodynamic data for the NEA TDB Project data. All ionic species listed in this table are aqueous species. The selection of these [2005OLI/NOL] and . Unless noted data is described in Chapter VI of [92GRE/FUG] otherwise, all data refer to 298.15 K and a pressure of 0.1 MPa and, for aqueous species, = 0). The uncertainties listed below each value I a standard state of infinite dilution ( represent total uncertainties and correspond in principle to the statistically defined 95% confidence interval. Systematically, all the va lues are presented with three digits after the decimal point, regardless of the significan ce of these digits. The reference listed for each entry in this table indi cates the NEA TDB Review where the corresponding data have been adopted as NEA TDB Auxiliary data. The data presented in this table are the OECD Nuclear Energy Agency. available on computer media from Reaction Species and ο ο ο ο review where adopted ∆ ∆ H G ∆ S log K rm rm rm 10 –1 –1 –1 –1 ) · mol )(kJ · mol ) (J · K (kJ · mol – + U HF(aq) + H F HF(aq) (a) – 18.152 101.800 12.200 [92GRE/FUG] 3.180 0.300 0.114 1.077 ± 0.020 ± ± ± – – – + HF(aq) U HF F 2 HF 2 (a) 18.486 – 2.511 3.000 [92GRE/FUG] 0.440 2.000 ± 0.685 ± 7.090 0.120 ± ± – + – HClO(aq) U ClO + H ClO (a) 42.354 – 78.329 19.000 [92GRE/FUG] – 7.420 ± 9.000 ± 0.130 ± 30.289 0.742 ± – + – U ClO (aq) + H HClO 2 2 ClO 2 11.188 [92GRE/FUG] – 1.960 0.114 ± 0.020 ± – + Cl (g) + H Cl O(l) + H + HClO(aq) U HClO(aq) 2 2 25.900 [92GRE/FUG] – 4.537 ± 0.105 0.600 ± – + H O(l) + HClO(aq) U 2H + HClO (aq) + 2 e HClO (aq) 2 2 2 (b) [92GRE/FUG] – 55.400 316.230 3.996 ± 0.700 ± – + – U BrO HBrO(aq) + H BrO (a) 49.260 30.000 – 64.600 [92GRE/FUG] – 8.630 3.000 ± ± 0.171 10.078 0.030 ± ± – + Br + HBrO(aq) O(l) U Br (aq) + H + H HBrO(aq) 2 2 47.034 [92GRE/FUG] – 8.240 0.200 ± ± 1.142 (Continued on next page)

116 IV. Selected auxiliary data 74 Table IV–2 (continued) Reaction Species and ο ο ο ο review where adopted S H G ∆ ∆ ∆ log K rm 10 rm rm –1 –1 –1 –1 )(kJ · mol · mol ) ) (J · K (kJ · mol – + H (aq) HIO + IO U (aq) HIO 3 3 3 – 4.498 [92GRE/FUG] 0.788 0.166 ± ± 0.029 – 2– + 2– HS H + S U S 108.450 [92GRE/FUG] – 19.000 ± 2.000 11.416 ± – – 2– 2– 2– + SO O(l) + SO U 2OH + 2 e H 4 3 2 SO 3 (b) [92GRE/FUG] 179.230 – 31.400 0.700 ± 3.996 ± – – 2– 2– 2– O(l) + 2SO + S + 4 e O U 6OH 3H 3 3 2 2 S O 2 3 (b) [92GRE/FUG] – 39.200 223.760 7.991 ± ± 1.400 – + H + HS S(aq) U H H S(aq) 2 2 39.899 [92GRE/FUG] – 6.990 0.170 ± 0.970 ± – 2– + – HSO U + SO H 3 3 HSO 3 (a) 359.590 66.000 – 41.212 [92GRE/FUG] 7.220 ± ± 0.080 100.630 ± 30.000 ± 0.457 – 2– + – O + S HS O U H 3 2 2 3 O HS 3 2 – 9.076 [92GRE/FUG] 1.590 0.150 ± 0.856 ± – + H (aq) SO H + HSO U H (aq) SO 3 2 3 2 3 (a) – 10.503 88.891 16.000 [92GRE/FUG] 1.840 0.080 ± ± 16.840 ± 0.457 5.000 ± – 2– + – + SO U HSO H 4 4 HSO 4 – 11.302 [92GRE/FUG] 1.980 0.050 ± 0.285 ± – + H (aq) HN + N U HN (aq) 3 3 3 (a) 39.671 – 26.828 – 15.000 [92GRE/FUG] 4.700 ± ± 0.457 ± 33.575 0.080 10.000 ± + + NH (aq) U H + NH NH (aq) 3 4 3 (a) – 2.130 52.090 52.725 [92GRE/FUG] – 9.237 ± 0.210 ± 0.126 ± 0.821 ± 0.022 (Continued on next page)

117 IV Selected auxiliary data 75 Table IV–2 (continued) Reaction Species and ο ο ο ο review where adopted ∆ ∆ H G ∆ S log K rm rm 10 rm –1 –1 –1 –1 ) · mol )(kJ · mol ) (J · K (kJ · mol – + H + NO (aq) HNO U (aq) HNO 2 2 2 (a) 23.219 – 11.400 – 18.323 3.210 [92GRE/FUG] ± 0.913 ± 10.518 0.160 ± 3.000 ± 3– 2– + 3– + PO U H HPO 4 4 PO 4 (a) – 187.470 14.600 70.494 [92GRE/FUG] – 12.350 0.171 ± ± 12.758 ± 3.800 ± 0.030 3– 4– + 4– U O O H + P HP 2 7 7 2 O P 7 2 53.656 [92GRE/FUG] – 9.400 ± 0.150 ± 0.856 – 2– + – PO + HPO U H H 4 2 4 H PO 2 4 (a) – 41.166 – 3.600 126.000 [92GRE/FUG] 7.212 ± ± 0.074 ± 3.363 0.013 1.000 ± – + H + H PO (aq) U PO H H PO (aq) 4 2 3 4 3 4 (a) 69.412 8.480 – 12.215 [92GRE/FUG] 2.140 ± 2.093 0.171 ± 0.600 0.030 ± ± 3– 2– + 3– + HP O P H U O H 7 2 2 7 2 O HP 7 2 37.958 [92GRE/FUG] – 6.650 ± 0.100 ± 0.571 2– – + 2– O H P U O + H P H 7 2 3 7 2 2 H O P 7 2 2 12.843 [92GRE/FUG] – 2.250 0.150 ± 0.856 ± – + – P + H H O O (aq) U P H 2 7 3 4 7 2 H P O 3 2 7 5.708 [92GRE/FUG] – 1.000 0.500 ± ± 2.854 2H O P U (aq) O(l) + H PO H (aq) H O P (aq) 4 2 3 4 2 7 2 4 7 (a) 21.045 22.200 15.925 [92GRE/FUG] – 2.790 4.673 ± ± ± 1.000 0.970 ± 0.170 – + H + HCO O(l) (aq) + H U CO CO (aq) 3 2 2 2 – 36.269 [92GRE/FUG] 6.354 ± 0.114 ± 0.020 CO (aq) U CO (g) CO (g) 2 2 2 – 8.402 [92GRE/FUG] 1.472 0.020 ± ± 0.114 (Continued on next page)

118 IV. Selected auxiliary data 76 Table IV–2 (continued) Species and Reaction ο ο ο ο review where adopted ∆ ∆ H G ∆ S log K rm rm rm 10 –1 –1 –1 –1 ) · mol )(kJ · mol ) (J · K (kJ · mol – 2– + – U + H HCO CO 3 3 HCO 3 – 58.958 10.329 [92GRE/FUG] ± 0.114 0.020 ± – + – U CN HCN(aq) + H CN (a) – 30.089 52.571 43.600 [2005OLI/NOL] – 9.210 0.200 ± 0.114 ± 0.772 0.020 ± ± HCN(aq) HCN(g) U HCN(aq) (a) [2005OLI/NOL] – 70.439 – 5.149 – 26.150 0.902 ± 0.050 ± 8.440 ± 0.285 2.500 ± 2– + 2– + SiO (OH) 2H U (aq) Si(OH) 4 2 2 (OH) SiO 2 2 (a) – 191.460 75.000 132.080 [92GRE/FUG] – 23.140 ± 0.514 ± 50.340 ± 15.000 ± 0.090 – + – U (aq) + SiO(OH) H Si(OH) 4 3 SiO(OH) 3 (a) – 101.950 25.600 55.996 [92GRE/FUG] – 9.810 ± ± 6.719 ± 2.000 0.020 0.114 ± 2H U O(l) + SiO (aq) (quar) Si(OH) Si(OH) (aq) 4 2 2 4 (a) 8.613 22.832 25.400 [92GRE/FUG] – 4.000 0.571 0.100 10.243 3.000 ± ± ± ± 2– + 2– O O(l) + Si + H U 2H (aq) (OH) 2Si(OH) 4 2 2 3 4 O (OH) Si 3 2 4 108.450 [92GRE/FUG] – 19.000 1.712 ± 0.300 ± – + – O O(l) + Si + H (OH) H U (aq) 2Si(OH) 5 4 2 2 2 Si O (OH) 2 5 2 46.235 [92GRE/FUG] – 8.100 1.712 ± 0.300 ± 3– + 3– O O(l) + Si + 3H (OH) 3H U (aq) 3Si(OH) 4 2 3 6 3 Si (OH) O 3 3 6 163.250 [92GRE/FUG] – 28.600 0.300 ± 1.712 ± 3– + 3– O O(l) + Si (OH) U 3H + 2H (aq) 3Si(OH) 5 2 4 3 5 Si (OH) O 5 5 3 156.970 [92GRE/FUG] – 27.500 0.300 ± 1.712 ± (Continued on next page)

119 IV Selected auxiliary data 77 Table IV–2 (continued) Reaction Species and ο ο ο ο review where adopted G ∆ ∆ H S ∆ K log rm rm 10 rm –1 –1 –1 –1 (J · K )(kJ · mol · mol ) ) (kJ · mol 4– + 4– O O(l) + Si + 4H (aq) 4H U (OH) 4Si(OH) 4 2 4 4 8 Si (OH) O 8 4 4 207.200 – 36.300 [92GRE/FUG] ± ± 2.854 0.500 3– + 3– O O(l) + Si + 4H 3H (OH) U (aq) 4Si(OH) 4 2 4 5 7 Si (OH) O 7 4 5 145.560 [92GRE/FUG] – 25.500 0.300 ± 1.712 ± οο ο . GHTS ∆=∆−∆ (a) Value calculated internally using rm rm r m (b) Value calculated from a selected standard potential.

120

121 Part III Discussion of data selection

122

123 Chapter V Criteria for data evaluation and V particular problems encountered in the review procedure Equation Section 5 V.1. Criteria for data evaluation In this review the NEA guide and data selection were ap- lines for the review procedure plied. However, the review team felt it necessary to summarise in addition the criteria for data evaluation with special emphasis on organic ligands. As these criteria were ap- plied implicitly in all the review work, and as they are mentioned explicitly in the dis- cussion of publications in the following chapte rs and in Appendix A, they are presented and discussed here. The following check list for data evalua tion was proposed jointly on the occa- sion of the second plenary meeting: • Check the pH scale used. Is the calibration method for the pH-electrodes indicated in the publication? The pH-electrodes should have been calibrated preferably in the concentration scale, and not with standard buffers to the activity scale. That is, “pH” should + refer to – log ]. The latter quantity is sometimes reported as pcH or pH [H . In 10 c some cases it is reported that the glass electrodes were calibrated with standard + buffers, and [H ] calculated from pH, for example with the Davies equation. In general, this procedure is not accepted in this review. In the case of ligand pro- tonation constants references were discarded when they reported mixed equi- librium constants, i.e. , involving both, proton activities and ligand concentra- tion. 81

124 V Criteria for data evaluation and particular pr oblems encountered in the review procedure 82 • Check that a reliable referenc e electrode has been used. An unfortunate choice of an experimental set-up is using a calomel reference electrode and NaClO as background electrolyte. The calomel reference elec- 4 trode contains a saturated KCl or 1 M KCl solution. Under such conditions the measurements are very difficult becaus e of the potential precipitation of KClO 4 occurring at the contact between the two solutions (KCl and NaClO ). 4 • Check whether suitable metal salts have been used in the experiments. Sometimes, an experiment was mainly set-up from the viewpoint of conven- ience or analytical accuracy, and little atte ntion was paid to the ambiguities of e, sulphate metal salts have occasion- results caused by this choice. For exampl ally been used because these salts allow lutions with exact the preparation of so compositions by weighing. However, the drawback of this choice is the ambi- guity of data interpretation due to the additional effects caused by metal sul- phate complexation. • Check that the temperature at which the measurements have been done is given. This obviously important information for deriving reliable thermodynamic data is often missing, or given only in ambiguous terms. For example, the term “room temperature” varies with the year and the geographic situation of the in- vestigation between 15 and 30ºC. • Check whether the ionic strength was reasonably constant under the experimental conditions. A background electrolyte providing a constant ionic medium must be used. The nature of the background electrolyte, the ionic strength, and the total ≤ 0.1 M the condition I ligand concentrations must be given. For studies where of constant ionic medium is quite difficult to achieve. For example, pH values I < 0.1 M without disturbing substantially the below 2 cannot be reached at composition of the ionic medium. In the case of citrate, and even more so in ations of the reacting species with re- the case of edta, the use of large concentr the background electrolyte, e.g. , 0.1 M KNO spect to the concentration of , can 3 result in total ionic strengths much higher than 0.1 M. Sometimes this effect has been remedied in experimental studies by considering the contribution of the reacting species to the total ionic st rength and reducing the concentration of the added inert salt in order to reach = 0.1 M (KNO I ). In such cases the ionic 3 medium is shown in parenthesis in this review indicating that the concentration can be lower than 0.1 M. of KNO 3 • Check that the speciation model used for analysing the results is compatible with the one chosen by the TDB review. As an example, speciation models of metal – edta systems at low pH some- 2+ + times ignore the presence of the species H edta under these con- and H edta 6 5

125 V.2 Particular problems of comm only used experimental method 83 ditions. In such cases the data have been re-interpreted in this review if the ex- perimental data have been reported in a suitable form. Check that the protonation constants used are compatible with the values selected • in the TDB review. If necessary and possible, i.e. , if the experimental data have been reported in a suitable form, the data have been recal culated using protonation constants se- lected in this review. • Check the method used, if any, for extrapolation to zero ionic strength. If necessary and possible, , if the experimental data have been reported in a i.e. suitable form, the data have been recalcu lated in this review using SIT. The SIT approach has been found acceptable for ionic strengths up to about 5 mo- lal. • Check that all the complexes whose stab ility constants are extracted from the measurement data are significant under the experimental conditions. Sometimes chemical models are propos ed where the complexes purportedly the total concentration of the inves- being studied represent a minor fraction of tigated substance. If possible, , if the experimental data have been reported i.e. in a suitable form, in this review alternative chemical models have been used to analyse the data. In addition to this check list, special attention has been paid concerning the characterisation of solid compounds. A large number of metal – organic compounds have been reported in the literature, but thei r degree of characterisation varies consid- e, citrate, edta and isa com pounds are mentioned only if erably. In this review, oxalat t by elemental analysis, or it is indicated their stoichiometry has been confirmed at leas explicitly that the authors assumed a certain stoichiometry without further evidence. If solubility data or thermodynamic constants have been reported, an important criteria for the reliability of these values is the proper characterisation of the solid compound in equilibrium with the solution. In many cases the complete lack of any characterisation of the solid prevented the selection of thermodynamic data by this review. V.2. Particular problems of co mmonly used experimental methods Most commonly, the following methods are used to determine stability constants. (1) Potentiometry (2) Two-phase distribution (solvent extraction and ion exchange) Spectrophotometry (3)

126 V Criteria for data evaluation and particular pr oblems encountered in the review procedure 84 However, since, for the polyprotic acid H L, any of the dissociated species n r–n H ( r = 0 to L n – 1) may form complexes with metal ions, the following problems are r commonly encountered in the papers examined by this review: (a) Assignment of complex species. (b) Error propagation in the method of data processing (errors from the data and from the mathematical transformation). Error propagation from the dissociation constants used in the calculation. (c) (d) Side reactions (hydrolysis). Potentiometry V.2.1. − C In this method, titration curves, pH vs ), are obtained . (total concentration of OH OH for the solution containing known concentrations of H L ( C ) and the metal ion (C ). n M L The complex formation of the metal ion M with H L (in the following discussions, n several protons according to the reaction, charges will be omitted for brevity) liberates MHL M(HL) ( )H mq qnr ++− U (V.1) nmrq and the resulting proton concentration is meas ured potentiometrically. If the ratio of the number of protons to the number of ligands in the complex is more freely defined, the reaction may be considered as, MHL MHL( )H (V.2) mq qnr ++− U nmrq In this case, the complex is considered as a mixed ligand complex MHL mrq s L and OH. Note that it is not possible to discriminate ) or with H L ( r ≠ L and H with H s r r by potentiometry between these two kinds of mixed ligand complexes. ), which is to be determined, is de- The stability constant for the Reaction (V.3 fined as: MHL M(HL) + U (V.3) mq rmrq [M (H L) ] mrq β (V.4) = (H L), qm mq r [M] [H L] ⋅ r L and metal ions as a salt of a non- When organic acid is added as H n complexing anion, mass balance equations for ligand (L), metal ion (M) and proton (H) are: n [H L] [M (H L) ] =+ Cq , (V.5) ∑ ∑∑∑ L imrq 0 = imrq [M (H L) ] Cm =+ , (V.6) [M] ∑∑∑ M mrq mr q n qr [H] [OH] [H L] , (V.7) −−+ = + i C nC [M (H L) ] ∑ ∑∑∑ LOH imrq = 1 imrq where

127 V.2 Particular problems of comm only used experimental method 85 r rr H ′ , (V.8) == =α )[H] [L] β [H L] KC [H] [L] ( ∏ r i rr L i 1 = n ′ (V.9) [H L] C = ∑ i L i 0 = H r [H L] [H] β rr α= (V.10) = r nn i H β [H] 1 [H L] + ∑∑ ii iii 0 == . (V.11) = K and [OH] /[H] w q , r , m and If [H], [L], and [M] and all formation constants ( H , log β ) are known, equilibrium concentrations of all species can be β log 10 i ( H L ), qm 10 r and all formation C , C C , calculated straightforwardly. Alternatively, when OH M L H log β ) are known, these equations can be log β , q m r constants ( and , , i 10 qm ( H L ), 10 r solved to give [H], [L], and [M], which in turn enable to calculate equilibrium concen- q , r trations of all species. When some of m are larger than 1, Eqs. (V.4) - (V.11) and can be solved only implicitly. Potentiometric determination of stability constants is the estimation of the best C C , C log , and β and q , r , m from the various data sets of ( set of M OH L ( H L ), 10 qm r H β log ) with known C , , C C . In [H], where [H] is measured for various OH i L M 10 principle, one can carry out this by adopting the following procedure. . Prepare the data sets of [H] (observed) for various C C , , C (i) L M OH Make estimates of the set of q , . , m and β log r (ii) qm 10 ( H L ), r For each data set, solve Eqs. (V.4) - (V .11) to obtain [H], [L], and [M] for (iii) H C . C and β C using the above estimates and log , a given set of M OH L i 10 Compare [H](estimated) and [H](observed). (iv) Change estimates of the set of q , r , m and . (v) β log 10 ( H L ), qm r Repeat (iii)-(v) until [H](estimated) su fficiently agrees with [H](observed). (vi) Although this is numerically possible, it is somewhat difficult since the proce- dure is doubly implicit. To simplify Eqs. (V.5) - (V.7), among the terms, and appearing in , [M (H L) ] m qr q [M (H L) ] [M (H L) ] ∑∑∑ ∑∑∑ ∑∑∑ mrq mrq mrq Eqs. (V.5) - (V.7), at least two of them must be directly related to allow substitution of each other. Therefore, in many papers, cert ain assumptions on the complex species are made to simplify the problem. L) forms complexes containing one metal ion When only one ligand species (H r m = 1), the second term in the right hand side of Eq.(V.7) can be transformed as fol- ( lows: = r q qr . (V.12) [M(H L) ] [M(H L) ] ∑∑ ∑ rq rq rq q Substituting Eqs. (V.5), (V.8) and (V.11) into Eq.(V.7) gives:

128 V Criteria for data evaluation and particular pr oblems encountered in the review procedure 86 n ′ ′ . (V.13) K [H] [H] / ) nC C ( C i r C C −−+ = α+ − ∑ i LOH W L LL i = 1 ′ Only one unknown, in this equation can be explicitly estimated from C L H log n K β log (the average number , and , . Thus, C C measured [H] with OH L L i w 10 10 of ligand bound per metal ion) can be directly calculated as: [M(H L) ] q ∑ rq ′ − CC q LL . (V.14) == n L CC MM n Since [H can be expressed as a function L] can be calculated by Eq.(V.8), L r L]. of [H r q β [H L] qq [M(H L) ] ∑∑ qL r rq ) (H r qq (V.15) == n L q + C 1[HL] β ∑ M qL r ) (H r q ′ [H () CCC L] can be analysed (graphi- − vs . log Therefore, a plot of 10 LLM r cally or by nonlinear curve-fitting) to give values of . β log 10 qL ) ( H r In some papers, the above titration curve is compared with that of the metal-ion blank solution. In the absence of a metal ion, the proton balance is expressed by, n * . (V.16) −−+ = α [H] [OH] i C C nC ∑ i LOH L i 1 = By comparing Eq.(V.16) with Eq.(V.13) at the same pH, n * ′ . (V.17) =−=− α− ∆ CCC CC ir () ∑ i L OH OH L OH ( ) i 1 = Then is calculated as, n L ∆ C OH = n (V.18) L () α− Cir ∑ i M ′ n ), (without calculating C Although this seems to make it easier to obtain L L this is only apparently easier since the blan k titration can be calculated from a given set H of C . Also, this makes the error estimation difficult, , C β , log log and K L OH w 10 i 10 H β K log log since the errors in the blank titration and those in and will be in- 10 i w 10 troduced in a complicated manner. When only one complex species is fo rmed, Eqs. (V.5) - (V.7) can be ex- pressed as: n ′ , (V.19) [H L] [M (H L) ] [M (H L) ] Cq Cq =+ =+ ∑ mrq imrq LL i 0 = [M] [M (H L) ] Cm =+ (V.20) M mrq n . (V.21) [H] [OH] nC [M (H L) ] qr i C [H L] −−+ = + ∑ imrq LOH i 1 =

129 V.2 Particular problems of comm only used experimental method 87 Substituting Eq.(V.19) into Eq.(V.21) gives n ′ ′ [H] /[H] ( ) C r C i nC C K C , (V.22) −−+ = α+ − ∑ wi LOH LL L = i 1 ′ C . Then which enables us to calculate L ′ [M (H L) ] ( ) / =− CC q (V.23) mrq LL ′ [M] ) / CmCCq =− − (V.24) ( MLL ′ (V.25) C [H L] =α rr L are used to directly calculate log by Eq.(V.3). β ( H L ) 10 q r Except for the simplest cases discusse d above, titration curves cannot be re- solved to explicitly give β log without a complex nonlinear least-squares q ( H L ) 10 r method. Error propagation in the data processing (errors from the data and from the mathematical transformation) is quite complicated in this potentiometric determination . Thus, reproduction of the titration curve with a set of assumed species log β of 10 ( H L ) q r and their β and the comparison of those parameters with results obtained log ( H L ) q 10 r using different sets of assumed species and their are recommended to log β ( H L ) q 10 r judge the reliability of the work. Potentiometry is effective only when the concentrations of the different species change along the titration curve. Therefore the number of species that may be investi- gated using potentiometry depends strongly on the concentration ranges studied, both of ligand, metal and pH. Because of that a large and constant ionic medium is essential to avoid changes of activity coefficients, while a llowing a wide range of concentrations to be covered. When interpreting the data it is therefore useful to calculate the amounts of the different species along a titration. If the pe rcentage of the species is very small, per- haps < 5%, then it is very difficult to dete rmine its equilibrium constant. Also if there are only one or two titrations covering a small range of ligand concentrations, then the uncertainty is increased. Two-phase distribution (solvent extraction and ion V.2.2. exchange) D , of a metal ion between two phases in the presence of a com- The distribution ratio, plex-forming reagent is expressed by: C D M,o 0 , (V.26) == D q [M] [M(H L) ] + β 1[HL] + ∑∑ ∑∑ (H L) rq r q r rq rq where C is the concentration of M in the organic or ion exchanger phase and D is the M,o 0 distribution ratio in the absence of the ligand. Only monomeric complexes are consid-

130 V Criteria for data evaluation and particular pr oblems encountered in the review procedure 88 change studies, the term “distr ibution coefficient” and the ered for simplicity. In ion ex symbol are used in place of “distribution ratio” and K . D d When only one ligand species forms complexes, Eq.(V.26) can be simplified to: D 0 D = . (V.27) 2 ββ 1[HL] [HL] ++ + " rr 2( H L ) 1( H L ) rr In the past, when complicated mathematical treatment was unavailable, this equation was transformed into: 1 − DD 0 . (V.28) ββ " =+ + [H L] 2 ( H L ) 1( H L ) r rr [H L] r were obtained as an intercept and a slope of the plot of this equa- β and β 2(H L) 1( H L ) r r tion. When different ligand species (such as L, HL and H L ...) may form com- 2 plexes, Eq.(V.27) is transformed into, DD 00 , (V.29) == D H qrqq βββ 1[HL]1([H])[L] ++ ∑∑ ∑∑ (H L) (H L) qr qr rr rq rq and apparent constants H app rq (V.30) ββ β = ([H]) ∑ r qq (H L) r r app are obtained at a certain pH. From the dependence of log β on pH, β log q ( H L ) 10 q 10 r can be estimated. Check points that should be considered are: ο log D or : K log (a) 0 10 d 10 Dependence of on pH, concentrations of extractants and ionic strength (in log D 0 10 ο solvent extraction), or dependence of log K on pH, concentrations of competing 10 d ions and ionic strength must be carefully examined. (b) Side reactions: In connection with (a), the effect of possibl e side reactions must be checked. Hydrolysis and complex formation by the extractant are especially important. When some side re- should be expressed as: action is serious, Eq.(V.26)

131 V.2 Particular problems of comm only used experimental method 89 C M,0 D = , (V.31) ++ [M] [M(H L) ] [MX ] ∑∑∑ x rq xrq + << seems to indicate [M(H L) ] Although the condition [M] [MX ] ∑∑∑ rq x that the effect of complex formation by X is well masked by the main complex forma- tion with H [MX ] L, this does not guarantee that [M] [MX ]  . This means that ∑ ∑ r x x cannot be neglected in the analysis since Eq.(V.31) is transformed into: D 0 D = . (V.32) q β [H L] ∑∑ (H L) qr r rq 1 + x β 1[] X + ∑ x x (c) Distribution of metal-containing species. Complexes should not be carried into the organic or exchanger phase. Especially, in ion exchange, complex species may be also transferred into the exchanger phase. (d) Error propagation: Usually in solvent extraction, the concentra tions of the metal ion in both, organic and aqueous phases, are measured to obtain D values. If absolute errors in the determination ly equal, the relative errors in D of metal-ion concentration are near (/ ln) D ∆=∆ DD will be nearly constant. In this case, the best set of parameters will be obtained by minimising the residual sum of squares, 2 (V.33) =− SD D log log ∑ () calculated observed 10 10 L] [H is the value calculated by Eq.(V.26) with the variable log log D where r 10 calculated 10 β Since the variable(s) and parameters should be log and assumed parameters q 10 (H L) r selected in a way that their uncertainties will equally contribute to the residual sum, the variable and parameters should be taken as log [H L] (not [H L]), (not β log r 10 r q 10 (H L) r ). Thus, the best form that should be taken for a least- β ) and log D D (not 0 0 10 q (H L) r squares method is: log log [ H L ] β + q 10 10 qr ⎛⎞ (H L) r log log 10 (V.34) DD log 1 =−+ ∑∑ ⎜⎟ 10 10 0 rq ⎝⎠ [H D L] to this equation can be per- vs . log Curve fitting of the plots of log 10 r 10 formed since the right-hand side of Eq.(V.34) can be calculated straightforwardly using [H L]) and fitting parameters ( ). On the con- log D β and log the variable (log r 10 0 10 q 10 (H L) r trary, any transformation su ch as Eq.(V.28) will complic ate the error consideration. ) is / D /1 D . If D − / D or ( DD changes with the magnitude of D The absolute error in 0 0 0 taken as a fitting value for the non-weighted least-squares curve-fitting, the data with D / are larger with smaller D D small are heavily weighted since the absolute errors in 0 L] in the is divided by [H − DD /1 . This unjustified weighting is lessened since D r 0

132 V Criteria for data evaluation and particular pr oblems encountered in the review procedure 90 left hand side of Eq.(V.28). However, data plots with small [H L] will all be distributed r L]. The use- close to the y-axis because the right hand side of Eq.(V.28) is linear in [H r L] plots is where the predominant species is gradually ful parameter range of such [H r L) to M(H . Moreover, the error in D will give a systematic L) changing from M(H r 2 0 r lt to estimate the reliability of the values of error to the plots. As a result, it is difficu log β which are converted from obtained by this type of plot. Linearity β 10 q q (H L) (H L) r r → 0 should be carefully examined to check the L] of the plot in the region where [H r L] range. validity of Eq.(V.28) in the selected [H r On the other hand, usually in the ion exchange method, the concentration of metal ion in the aqueous phase C is measured and compared with the total concen- M,aq tration C M,t CC − M,aq M,t (V.35) D = C M,aq where is equivalent to K D . In this case, the absolute errors in C or the fraction re- M,aq d f erefore, the residual = C ) are nearly equal. Th / C mained in the aqueous phase ( M,t aq M,a sum of squares to be minimised is, 2 Sf =− f (V.36) ∑ () aq,calculated aq,observed where f can be calculated as: aq C 11 M,a (V.37) f == = aq D log 10 1 CD + + 1 10 M,t D is calculated by Eq.(V.34) straightforwardly using the variable where log 10 (log ). When the transforma- [H log L]) and fitting parameters ( β and log D 0 10 10 r 10 q (H L) r /1 DD − are similar since it is ex- tion given by Eq.(V.28) is used, relative errors in 0 pressed as: 0 − ff (1 / ) (1 / ) D aq aq 0 −= (V.38) 1 − ) 1 Df (1 / aq 0 0 f (1 / f and is the fraction re- (1 / ) ) are similar ( f where the absolute errors in aq aq aq mained in the aqueous phase in the absence of complexant). A similar discussion as for solvent extraction applies here, and it can be concluded that also in this case, it is diffi- cult to estimate the reliability of the values of . β log q 10 (H L) r Generally speaking, in the case of solvent extraction there is a new liquid phase, with uncertainties in activity coefficients in the organic phase, unknown parallel reactions, unknown species extracted, etc. This increases the uncert ainty of the method as compared with potentiometry and spectrophotometry where only one liquid phase is present.

133 V.2 Particular problems of comm only used experimental method 91 V.2.3. Spectrophotometry When the number of complexes is one or two, spectrophotometry with the analytical usually gives reliable method such as mole ratio method or cont inuous variation method information about the composition of the complex(es) and their molar absorptivity. The problem in this method is the error from the mass balance relationship. For example, in the simplest case, β is obtained as, q (H L) r [M(H L) ] [M(H L) ] rq rq == (V.39) β q q (H L) r −α − CCq [M(H L) ]) ( [M(H L) ])( [M][H L] ML r rq rq r where α is r H r β [H L] [H] rr (V.40) α= = r nn H i [H L] [H] 1 β + ∑∑ ii 0 == iii [M(H ) and the molar L) λ ] is obtained from the absorbance at a certain wavelength ( q r λ absorptivity ( ε . ) of the complex at λ M(H L) rq λ [M(H L) ] / =ε (V.41) A λ rq M(H L) rq λ ) is obtained from the absorbance of a ε Since molar absorptivity ( M(H L) rq 'known' concentration of the complex (usu ally under the condition where nearly all , it includes the same uncertainty as the de- metal ions are in the form of the complex) entration. Therefore termination of the conc and ([M(HL)]) C − M rq ([M(HL)]) − in Eq.(V.39) are reliable only when they are larger than the error Cq rq L limit of C C L] . That is, when the stability of the complex is very high, [M] or [H and L r M cannot be estimated reliably. Usually, such conditions are avoided since the experiment tration of the complex) is gradually chang- is done where the absorbance (or the concen ing. However, when the number of complexes is large, the situation may be more com- plex. When the number of complexes is large, the best set of ’s and molar ab- β q (H L) r λ sorptivities ( ε ’s) are estimated by minimising the following residual sum of M(H L) rq squares: 2 (V.42) =− SA A ∑∑ () λλ nn , ,calculated , ,observed λ n λ where the summation is carried out over several wavelengths ( n spectra. Total ) and all metal and total ligand concentrations give strict conditions for mass balances. In other words, they are considered not to contain any error. Since and ([M(HL)]) − C rq M may give serious C and C cannot become negative, the errors in − ([M(HL)]) Cq M L rq L systematic errors. If the same stock solution is used to prepare sample solutions, the problem is more serious. Although many wavelengths can be selected to improve the fitting, systematic errors may be introduced because each spectrum contains the error in the preparation of that solution. To minimise S by Eq.(V.42), spectrophotometry seems of complexes since larger number of pa- to have a tendency to assume larger number

134 V Criteria for data evaluation and particular pr oblems encountered in the review procedure 92 λ must be checked whether its β ’s and S . rameters ( ε S ’s) easily give smaller q M(H L) rq (H L) r size is reasonable as compared with the precisions of C and C . Also, the deviations of L M residuals must be examined whether they have some inclined tendency. Usually, these kind of examinations are not carried out in the papers, and the second best way may be to reproduce species distributions and absorption spectra from the reported parameters. In summary, in the case of spectrophotometry, additional parameters must be estimated (molar absorptivities). This is a disadvantage against potentiometry, where less fitting parameters need to be obtained. Also, in general, the precision of the meas- urements is lower for a spectrophotometer th an for a voltmeter. In case where species do not differ very much in absorption coefficients a distinction between species might be impossible (for example protonation of many ligands). And furthermore, spectropho- tometry (similar to solvent extraction) re quires in many cases that potentiometric + measurements (of lower precision) ar e made in any case to determine [H ]. So in general spectrophotometry requires double measurements: potentiometric and optical absorbance. However, clearly the advantage of this method is that there is positive en it has absorption in a characteristic confirmation that a complex is formed wh wavelength. V.3. Ionic medium effects on protonation constants for organic ligands V.3.1. Introduction The activity coefficients of aqueous ions and molecules at “low” ionic strengths, I < 0.1 M, are dominated by the Debye-Hückel term. Owing to this, the activity coeffi- cients are not highly dependent on the nature of the background electrolyte. However, at higher ionic strengths, > 0.1 M, the nature of the ionic components of the background I electrolyte will affect the value of the protonation constants. + For polycarboxylic acids, equilibrium constants for the formation of Na and + complexes have often been reported. These experimental studies are based on K elective glass electrodes. Such complexes potentiometric data, from both pH and ion-s would affect the protonation constants determined in sodium or potassium electrolyte media. However, it is in general not po ssible to differentiate between weak complex formation ( K < 10) and the variation of activity coefficients, based solely on potentiometric data. V.3.2. Evaluation of protonation da ta excluding complexes with medium cations y − The following nomenclature is used for the protonation of a ligand, L , cf . Section II.1.6.1:

135 V.3 Ionic medium effects on prot onation constants for organic ligands 93 ry − () ] [H L − y − 1) y ( r ) + ( r − r = K (V.43) L H U L H + H r ( − 1) r r ry +−− (1) L [H ][H ] r − (1) The standard SIT equations may be used to describe changes in activity coeffi- electrolytes of composition MX these equations are: cients. For 1:1 background o − KK γγ =++ log log (V.44) γ log log log ry ry () +−−− (1) 10 10 10 10 10 rr HHL HL − (1) rr The activity coefficients are according to the SIT methodology given by (Sec- tion B.1): 2 γ (V.45) =− log zD ik m + ( , ) ε ∑ 10 k ii k resulting in: 2o (V.46) log log −∆ε K zD K I −∆ = 10 10 rrm . Eqs. (B.6) to (B.8) in Appendix B): cf where ( AI m = (V.47) D + I 11.5 m 22 2 ()( 1)1222 zryry ry ∆=− −−− −= − − () ( 1) ry ry −−+ +− −+ , M ) (H L , M ) (H , X ) (H L − ε ∆ε = ε − ε rr (1) − and I is the ionic strength in molal units of the background electrolyte (for a 1:1 elec- m + = 0.509 at 25°C (Table B.2). is equal to [M ]), and A I trolyte with formula MX, m As discussed in Appendix B the ε -values are in general not dependent on ionic strength at moderate concentrations of background salt. This (lack of) dependence is s. There are however exceptions, such as illustrated in Figure V-1 for a few electrolyte the tetraalkylammonium halides, as indicated in the same figure. For these exceptions ) + ) = ε ( i , k , cf the following relationship has been proposed ( . Appendix B): ε ( i , k I 1 m ε ( i , k ) log -values are not dependent on the ionic strength, as is the general ( I ). If the ε m 10 2 case, then the experimental values for the protonation constants, corrected for the De- I bye-Hückel term as indicated in the left hand side of Eq.(V.46), when plotted against m ο log K are obtained. ∆ε and should produce a straight line from which the values of r 10 ∆ε For protonation constants it is possible to separate in two parts, so that all data for a given background cation may be treated simultaneously: 2o* +− ⋅ = log ε (H ,X ) log (V.48) ε K zD I K I −∆ ⋅ −∆ − 10 10 rmrm −−+ −+ *() (1) ry ry where depends only on the cationic com- ∆ε = ε (H L (H , M ) , M ) −ε L − rr (1) + + ). In this case the left hand side of , K , etc. position of the background electrolyte (Na log K , corrected both for the Debye- Eq.(V.48), with the experimental values of 10 r

136 V Criteria for data evaluation and particular pr oblems encountered in the review procedure 94 + should Hückel term and the specific interaction term for H , when plotted against I m * ο ∆ε log K and may be obtained. produce a straight line from which the values of r 10 Eq.(V.48) has been used in Sections VI.3.2 and VI.3.3 (Figures VI-2 and VI-3), Sec- tions VII.3.2 to VII.3.4 (Figures VII-5 to VII-7), and Sections VIII.3.2 to VIII.3.6 (Fig- ures VIII-7 to VIII-12). ity coefficients adjusted w Figure V-1: Mean ionic activ ith a Debye-Hückel term for some 1:1 electrolytes (sodium acetate and ch loride, ammonium chloride and tetrame- ) in Eq.(V.45) correspond to the slope of ε I thylammonium chloride). The values of k ( i , , m lines passing through the origin, and they are only slightly dependent on the ionic strength for most salts. The figure shows however that in the case of tetraalkylammo- nium halides the ionic interactions can be better described by assuming ion interaction strength rather than constants. Ciavatta coefficients that are functions of the ionic i ( ε ) = I , k , i ( ε I ) log k , i ( ε ) + k , i ( ε ) = ) + , k , i ( ε proposed the relationship I k , [80CIA] 1 m m m 10 1 2 i ε I ( , , and this function has been adopted by this review, see also Section B.3 k ) log 2 10 m and Table B-6. 0.20 Na(ac) a(ac ) N 0.15 N aCl NaCl H Cl N 4 NH4Cl ) Me NC l 4 Me4NCl 0.10 m I ( D + 0.05 ± γ 10 0.00 log −0.05 −0.10 0123 1 − m ) (mol·kg Complex formation with medium cations V.3.3. It is often claimed that polycarboxylic acids form complexes with alkali metal ions. 4 − 3 − , cit , and to a lesser extent species with lower electric charge, This could concern edta − 3 − 2 2 − , and ox , Hcit . In several studies the difference in protonation con- such as Hedta

137 V.3 Ionic medium effects on prot onation constants for organic ligands 95 stants obtained in tetraalkylammonium and al kali metal electrolytes has been used to + + . It is assumed in these studies that no and K derive formation constants, e.g. , for Na complexes are formed between the ligands and tetraalkylammonium ions, and that the observed medium effects can be ascribed to complex formation. However, as indicated in Figure V-1 the activity coefficients in tetraalkylammonium salts are quite different from those in alkali metal salts, although the differences do not necessarily implicate + and chloride. Therefore, different ., between Na alkali metal ion complex formation, e.g ionic medium effects on protonation constants when comparing Na/K electrolytes with tetraalkylammonium salts are not necessarily an indication of complex formation be- + + . or K tween the ligand and Na This section illustrates how the SIT model of activity coefficients may be ap- plied to estimate medium effects on protonation constants if complex formation indeed occurs between the ligand and the medium cation. y − L , is described as: The first protonation of a ligand, y − ) (1 ] [HL + y (1– y ) − H . (V.49) L HL U + = K H y + − [H ][L ] + , is consid- If the possibility of complex formation with the medium cation, M constant are expressed as: ered, the equilibrium and its y ) − (1 ] [ML (1– y − y ) + M ML . (V.50) + L = K U M y + − [M ][L ] Note that a special nomenclature is used in this section for the equilibrium con- ) (1– y is not negligible, the protonation constant that is stants. If the concentration of ML measured experimentally is in fact: y − ) (1 † [HL ] (V.51) = K H +− − (1 yy ) ] [H ] [L ] [ML + () − y ) (1– + y L From Eq.(V.50), [ML ][ [M ], and it follows that: K ] = M † + =−+ (V.52) log log log 1 [M ] KK K ) ( H 10 10 10 M H K may be calculated with the SIT model, The ionic media dependence of log 10 H cf . Eq.(V.46): o2 KzD I =+∆−∆ε (V.53) K ) ( log log m H 10 H 10 H H 2 2 2 yy −+ −+ +− ) (1 where 1. In- − , and ( ∆ z y ) = (1 − y ) − −ε ∆ε = ε (HL ,M) (L ,M) (H,X) −ε H H serting Eq.(V.53) in Eq.(V.52): + †2 o + (V.54) −∆ε ⋅ − K KzD K I [M ] 1 log −∆ ⋅ = log ) log ( ) ( m H M 10 10 H 10 H H

138 V Criteria for data evaluation and particular pr oblems encountered in the review procedure 96 A comparison of Eq.(V.54) with Eq.(V.46) reveals that the term ) y (1– + − log ]) accounts for the effect of ML K (1 + [M in SIT plots. Eq.(V.54) may be M 10 considered from two extreme situations: either unimportant or considerable complex + formation with M . For negligible complex formation with the background cation, + K ] [M  1, and Eq.(V.54) is essentially equal to Eq.(V.46). In that case the SIT plot M (1– ) + y K [M is significant, ] 4 1, and the term is unaffected. When the formation of ML M + log K [M plot, as indicated in Eq.(V.54). For (1 + ]) will have an effect on the SIT 10 M should also be evaluated with the SIT K consistency, the ionic strength dependence of M . Eq.(V.46)): methodology ( cf o2 (V.55) =+∆−∆ε KzD I K ) log log ( m M M 10 M 10 M 2 ) yy −+ −+ +− (1 where . It should be noted that ( ∆ z ) = −ε −ε (L , M ) (M ,X ) (ML , M ) ∆ε = ε H M 2 ) . z ∆ ( M + + Figure V-2 shows values of log (1 + K [M [M ], for some values of ]) versus M 10 ο 4– , when the charge of the ligand is − 4 ( e.g ., for edta ). In general, the value of ∆ε K M M in Eq.(V.55) is unknown, and this uncertainty is indicated in the figure. + 4 − 3 − Figure V-2: Expected deviations in SIT plots for reaction: H HL + L , caused by U + + − 4 − 3 the simultaneous reaction: M is the alkali metal cation of the + L , where M U ML background electrolyte. The con tinuous curves show the deviations caused by three o different values of d lines show the effect of K and ∆ε = 0 in Eq.(V.55). The dotte M M setting ∆ε = ± 0.2. M 2.0 − ∆ε = 1000 = ° 0.2 K M Μ ) 1.5 M K ] ∆ε = 0.2 + Μ 1.0 K ° = 100 M ( 1 + [M 10 0.5 log = 1 ° K M 0.0 1.5 1.0 3.0 0.0 0.5 2.5 2.0 + ] [M

139 V.3 Ionic medium effects on prot onation constants for organic ligands 97 The effect of complex formation with the background cation on an SIT plot is shown in Figure V-3. It may be shown that the parameters ∆ε and ∆ε have opposite H M effects on the slope of the plot, and they have a large degree of correlation, as illustrated in Figure V-3. Figure V-3: The effect of alkali metal complex formation on SIT plots. In this artificial + 3 − 2 − ο example the following reactions are used: H + L ( U = 6.5) and log K HL H 10 3 − 2 + + ο 2 2 − is ( z 6, and M log K = 1.4), for both reactions ( ∆ U ML ) − = ( ∆ z + L ) = M M H 10 M the alkali metal cation of the background electrolyte. The upper two lines show SIT plots in the absence of complex formation, calculated using Eq.(V.46) with ∆ε = − 0.4 H ∆ε (continuous line) and = − 0.1 (dotted line). The lower curves show the effects of H − 2 ∆ε = ∆ε = − 0.4 (con- formation, calculated with Eqs.(V.54) and (V.55) using ML H M tinuous curve) and ∆ε = − 0.1, ∆ε = − 0.01 (dotted curve), illustrating that similar H M ∆ε results are obtained with widely different values of and . ∆ε M H = 0.4 − H ∆ε 7.00 without ) = 0.1 − ∆ε H 2 m − I ML ( D 6.50 + 6 † H K 10 0.1 = − H ∆ε 6.00 with formation 0.4 = H ∆ε − log 2 = 0.01 ∆ε − M − of ML = 0.4 M ∆ε − 5.50 2.0 0.0 0.5 1.5 1.0 + ] [M Several effects may be inferred from Figure V-2 and Figure V-3 for non- negligible alkali cation complexation in SIT plots: ∆ε The slope of the SIT plot is affected, i.e ., the value of obtained is not neces- • ∆ε sarily equal to . H 0.2 molal, the SIT plot is not linear. At I 3 • m

140 V Criteria for data evaluation and particular pr oblems encountered in the review procedure 98 ο log K The value of obtained from data at • I 4 0.2 molal may have a sub- m H 10 + + ο stantial error, proportional to the value . Data obtained in Na K and K log M 10 ο log K more negative than those from data in media will result in values of H 10 tetraalkylammonium media. • can and be obtained from the ionic strength dependence of ∆ε not ∆ε M H † K log Figure V-3. Additional in- . This is due to their large covariance, cf. H 10 ε formation is needed in order to obtain the different -values. , When analysing data from several el ectrolytes, for example NaCl, NaClO 4 NaNO , etc. , it may be convenient to separate ∆ε in two parts, so that all data for a M 3 given background cation may be treated simultaneously, in a similar way as ∆ε (com- H that case Eq.( V.54) becomes, pare Eqs.(V.46) and (V.48)). In † 2 +− + o* ( (H ,X ) K zD I log −∆ ⋅ −ε ⋅ (V.56) ) K KI [M ] =−∆ε−+ 1 log log ( ) m H H 10 m 10 10 H H M −+ −+ *(1) yy where . Similarly Eq.(V.55) may be rewritten as: −ε (HL ∆ε = ε (L , M ) , M ) H o2 * +− K + ∆ ⋅ −∆ε ⋅ +ε = (V.57) I KzD I log log ( ) (M , X ) ⋅ mm 10 M 10 M M M 2 2 2 2 −+ −+ yy *(1) where − . Again, ( ∆ z ∆ ) y = ( 1. z − ) ) = (1 − y −ε ∆ε = ε (L , M ) , M ) (ML M H M +−+ ** (1) (1) − yy Furthermore, the difference is expected to (HL , M ) −ε (ML , M ) ∆ε − ∆ε = ε MH be a small value. V.3.4. Complex formation with medium cations versus the SIT formalism Figure V-3 shows that in the case where co mplex formation occurs between the ligand and the background cation a curved SIT plot is obtained. Similar curves may be ob- tained assuming an ionic strength dependence of the SIT parameters: ∆ε ( I ) = ∆ε + H m H,1 ). This corroborates the assertion that it is practically not possible to dis- I ( log ∆ε 10 m H,2 tinguish between weak complex formation and ionic activity effects. Each chemical system must be examined separately and a decision taken on the existence of alkali 4– metal ion complexes. During this review it has been seen that for edta the formation of 3– a complex such as Na(edta) may be justified, and that for oxalate there is no evidence case of citrate, both formalisms might be for the formation of sodium complexes. In the applied with equal success to explain the io nic strength dependence of protonation con- stants. However, as there is no additional evidence (such as spectroscopic) for the for- 2– mation of Na(cit) , it has been estimated in this review that the simplest approach is to use the SIT model for activity coefficients ta king into account, wh en needed, an ionic strength variation of the ∆ε parameter. The SIT approach has been found acceptable H for ionic strengths up to about 5 molal.

141 V.3 Ionic medium effects on prot onation constants for organic ligands 99 V.3.5. Accessory data The use of Eq.(V.48) has two requirements: conversion of equilibrium constants to + − ε (H , X ) . To convert equilibrium constants to molal units, molal units, and values of the density of the background electrolyte is needed as described in Section II.2. Molar to molal conversion factors for KNO , NaBr, and KBr, and for tetraalkylammonium 3 salts that are not tabulated in Table II.5 (Chapter II), are given in Table V-1. c , and molality, Table V-1: The ratio between molarity, , for selected electrolytes. m , [85SOH/NOV] and apparent molar Values calculated from densities [77ANS/SUR] [66CON/VER] volumes . Š = / c m c (M) NaBr KBr KNO NBr Me NI Pr NCl Me NBr Et NBr Et NCl Et 4 4 4 4 3 4 4 0.1 1.005 1.006 1.007 1.012 1.013 1.019 1.019 1.021 1.026 0.25 1.009 1.012 1.013 1.030 1.032 1.046 1.047 1.052 1.065 0.5 1.015 1.021 1.024 1.060 1.064 1.090 1.097 1.107 1.136 1 1.029 1.040 1.045 1.121 1.138 1.188 1.210 1.23 1.31 2 1.058 1.081 1.093 3 1.089 1.129 4 1.124 + + + Me , Et represent tetramethyl-, tetraethyl-, and N tetrapropylammonium ions respectively. , and Pr N N 4 4 4 + − − Specific ion interaction parameters for H were calculated from and I with Br : the activity coefficients tabulated in [59ROB/STO] –1 + − ± ⋅ mol ε 0.01) kg = (0.15 (H , Br ) + − –1 = (0.19 ± 0.01) kg ⋅ mol (H , I ) . ε Ionic medium effects on reaction enthalpies V.3.6. For small temperature intervals (between 0 and 50°C) changes of equilibrium constants : may be calculated with the constant heat-capacity method [97PUI/RAR] ⎛⎞ ∆ C ⎛⎞ ⎛⎞ ∆ HT T () T 11 p r 0 r0 KT =+ −+ −+ KT 1 ln log ( ) log ( ) . ⎜⎟ ⎜⎟ ⎜⎟ rr 10 10 0 ⎜⎟ T T TT R ln(10) R ln(10) 00 ⎝⎠ ⎝⎠ ⎝⎠ The ionic medium dependence of the reaction enthalpies may be estimated : [97GRE/PLY2] with the SIT model 22 ο =∆ + ∆ − ∆ (V.58) ∆ε HHzDTm R L L rm rm

142 V Criteria for data evaluation and particular pr oblems encountered in the review procedure 100 where: I A 3 Lm D , =× L 4 + I 11.5 m 3/2 − 1/2 A is the Debye-Hückel constant (at 25°C and 1 atm kg A ⋅ mol ⋅ = 1.986 kJ ), and L L (, ) , ∆ε = ν ε ij ∑ LL i i where the stoichiometric coefficients ν are positive for products and negative for reac- i teraction coeffi- py specific ion in i , j ) are the relative partial molar enthal ( ε tants, and L cients: ij (, ) ∂ε ⎛⎞ (, ) ε= ij L ⎜⎟ ∂ T ⎝⎠ P where i refers to a reactant and j to the counter-ion of th e background electrolyte. 2 ∆−∆ H zD against the mo- According to the SIT model, a diagram of rm L ο ∆ H , will produce a linear plot from which m lality of the background electrolyte, rm 2 T ⋅∆ε may be obtained. These equations apply to R H ∆ obtained either and L rm from calorimetric measurements or from the temperature variation of protonation con- stants in molal units. For moderate ionic strengths ( ≤ 1 M) and temperature intervals (0 to 50°C) if the equilibrium constants are instead given in molar concentration units, values will be found. ∆ε equivalent equations might used but different L Weak complexes versus strong specific ion V.4. interaction Common background electrolytes used in experiments to determine stability constants + + are NaClO with organic ligands has and NaCl. The interaction of Na , KNO and K 3 4 − − – ClO NO , and Cl been discussed in Section V.3. In this section, the interaction of 3 4 with metal ions, especially actinides is discussed. Within the framework of SIT, metal ion – perchlorate interactions are treated in all cases as specific ion interactions, usin g the corresponding SIT interaction coeffi- − x+ ClO ) as reported in Appendix B. , ε cients (M 4 – − NO the situation is less obvious, as there are two al- and Cl In the cases of 3 ternative methods to treat metal ion interactions with nitrate and chloride: (A) Simple treatment as specific ion inter action, as in the case of perchlorate, – x+ x+ − ) and , (M NO , Cl ). In this ε (M ε SIT coefficients with corresponding 3 procedure any possible effects of nitrate or chloride complexation are in- cluded in the respective ion interaction coefficients, and consequently nitrate or chloride complexes are not in cluded in the speciation model.

143 V.4 Weak complexes versus strong specific ion interaction 101 (B) Complex treatment using nitrate or ch loride complexation constants together with SIT interaction coefficients valid for perchlorate solutions, and conse- quently correcting experimental equilibrium constants measured in nitrate or chloride solutions for nitrate or chloride complexation effects. Note, that this treatment requires SIT interaction coeffi cients for all charged metal – nitrate or metal – chloride complexes included in the speciation model. In general, procedure (A) is used in NEA TDB reviews for metal ions and complexes of low charge, and procedure (B) is used in the case of highly charged metal ions. However, the procedures applied in different NEA TDB reviews are not consistent and hence somewhat confusing. This topic is not yet settled, and it should be addressed in future NEA TDB reviews more thoroughly. For the time being, in this review it was decided to follow procedure (B) if ni- trate or chloride complexation constants have been selected by previous NEA TDB re- selected values are available. As a conse- views, and to follow procedure (A) if no such quence, Np(V) and Pu(III) – ch loride interactions were treated according to procedure (A), whereas for U(VI) and Am(III) – chloride interactions procedure (B) was applied by explicitly considering chloride complexes. The procedures applied and decisions made in this review are illustrated by the following two examples. As a first example we discuss how Am – edta data measured in NaCl solution 2+ 3– , AmCl and for complexation effects of Na(edta) have been corrected in this review + AmCl . The following equilibria have been considered: 2 + 4– ο 3– Na Na(edta) + edta K U = (2.8 ± 0.2) ∆ε = – 0.27 (this review) log Na 10 3+ – 2+ ο Am U AmCl + Cl = – 0.13 (see text) ∆ε 0.03) ± = (0.24 K log Cl 10 1 + 3+ – ο + 2 Cl AmCl = – 0.38 (see text) ∆ε 0.5) U log K = – (0.74 ± Am 2 10 Cl 2 ο The log K values have been used to , valid at I = 0, together with their ∆ε 10 K at the appropriate ionic strength, given by the NaCl concentration, and sub- calculate log K values for complexation effects accord- sequently to correct the experimental 10 ing to the following formula: −− +2 log = log + log (1 + ⋅⋅⋅ [Cl ] + [Cl ] ) KK K KK [Na ]) + log (1 + corr 10 10 Na 10 Cl Cl 10 12 ο ο The stability constants K and log for the Am chloride com- K log Cl 10 Cl 10 1 2 . The ∆ε values were calculated from [2003GUI/FAN] plexes have been taken from − − – 3+ 2+ – 3+ 2+ ClO ClO , Cl , ) = ) ≈ ) = (0.49 ± 0.03), ε (AmCl , Cl ε ) ≈ ε (AmCl (Am , (Am ε 4 4 – + + – + − ) = (0.17 0.04), and AmCl , Cl , Cl ) ≈ ε ( ± AmF ) = , (Na ClO ε ( ε 0.04), ± (0.39 2 4 2 . ± [95SIL/BID] (0.03 0.01), all values taken from 3+ – Note that [2003GUI/FAN] (Am report a value , Cl ε ) = (0.23 ± 0.02) with the – 3+ , Cl ) (Nd ε remark (footnote G on page 729) that this value “is assumed to be equal to

144 V Criteria for data evaluation and particular pr oblems encountered in the review procedure 102 3+ which is calculated from trace activity coefficients of Nd ion in 0 – 4 m NaCl. These zer ion interaction parameters evaluated in trace activity coefficients are based on the Pit – NaCl and from osmotic coefficients in aqueous NdCl [97KON/FAN] 3 NdCl − .” Alternatively, inspection by this review of the osmotic coefficients for CaCl 2 3 3+ – also result in ε (Nd 0.02). The , Cl solutions given by ) = (0.23 ± [59ROB/STO] NdCl 3 latter value is derived ignoring Nd – chloride complexation. − 2+ – 2+ ± ) = (0.21 ε ( UO , Cl UO , ) = N O 0.02) and Very similar values ε ( 2 2 3 (0.24 ± 0.03) were obtained from osmotic coefficients reported by [59ROB/STO] for [92GRE/FUG] Cl solutions, respectively. For the latter case and UO (NO ) UO 3 2 2 2 2 lled that these coeffici remark (footnote (i) on page 695) “it is reca ents are not used in without taking [80CIA] were calculated by Ciavatta the present review because they − 2+ ClO UO ( ε chloride and nitrate complexation into account”. Instead, , ) = 4 2 (0.46 ± 0.03) should be used in chloride solutions together with the equilibrium + Cl Cl and UO . Hence, in analogy [92GRE/FUG] (aq) as evaluated in constants for UO 2 2 2 − 3+ ClO (Am ε to the U(VI) chloride system, this review used , ) = (0.49 ± 0.03) together 4 + 2+ AmCl given in . and [2003GUI/FAN] with the equilibrium constants for AmCl 2 As a second example, some subtleties are discussed in the following concern- ing U(VI) – chloride complexation effects. – + 2+ U UO Cl this review used the following SIT in- UO + Cl In the case of 2 2 ∆ε for the U(VI) – chlo- ) to calculate [92GRE/FUG] teraction coefficients (taken from ride complexation: 2+ − 2+ − 0.03) ) = ε ( ClO ) = (0.46 UO , ± UO , Cl ( ε 2 2 4 − + (Cl ε , Na ) = (0.03 ± 0.01) − + (UO ε , Cl Cl ) = (0.33 ± 0.03). 2 + 2+ – ∆ε = – 0.25, Cl UO + Cl , in contrast to U UO ∆ε This gives = – 0.16 for 2 2 , p. 195. Where does this discrepancy come from? [92GRE/FUG] as reported in The solution of this puzzle is found in [92GRE/FUG] , p.192, line 21: For de- + – + – , Cl , H ) ) = 0.33 from their fitted Cl = – 0.25 Grenthe et al. used ε (Cl ∆ε (UO ε riving 2 , but inspection of = 0.12. There is no further comment on this choice in [92GRE/FUG] for the SIT plot (especially Table V.22 and Figure V.10 re veals that the data they used ) media, and hence, the SIT the ones at high ionic streng th) were measured in H(Cl,ClO 4 + – – Cl interaction was used. interaction coefficient for H ∆ε But in the case of NaCl solutions one cannot just use the derived for – + ) media, and this review calculated an appropriate using ε (Cl ) = 0.03 , Na ∆ε H(Cl,ClO 4 – + ∆ε ) = 0.12. This results in a difference in of 0.09. , H (Cl ε instead of

145 V.4 Weak complexes versus strong specific ion interaction 103 2+ – UO + 2 Cl Cl U UO In the case of (aq) this review used the following 2 2 2 SIT interaction coefficients to calculate ∆ε for the second U(VI) – chloride complexa- tion: − 2+ 2+ – UO , Cl ε ( 0.03) UO , ± ClO ) = ) = (0.46 ε ( 2 2 4 – + ε (Cl 0.01) ) = (0.03 ± , Na ε (UO Cl (aq), NaCl) = 0. 2 2 – 2+ ∆ε This gives = – 0.52 for UO + 2 Cl (aq), in contrast to = U UO Cl ∆ε 2 2 2 0.62, as reported in − [92GRE/FUG] , p. 196. This puzzle is mo re complicated than the first one. ) in [92GRE/FUG] and Again, we have the difference in media: H(Cl, ClO 4 – + , H = ∆ε ) = 0.12 gives ∆ε = – 0.70, and not (Cl ε NaCl in this example. But applying − 0.62 as derived in Figure V.11 of [92GRE/FUG] . The solution to this puzzle is again , p. 192, last 4 lines: The authors realised this discrepancy but found in [92GRE/FUG] simply stated "which is compatible with the value obtained from linear regression". This is true considering the uncertainties assigned to the values. However, the real ques- ∆ε tion is not addressed by this statement: Following the SIT procedure, [92GRE/FUG] . ∆ε 0.17) from their fitted ± Cl )) = (0.08 (aq), H(Cl,ClO (UO ε should have calculated 4 2 2 Now, there is an ambiguity: We can either argue that this value is not significantly dif- ferent from zero (as assumed a priori for neutral species in the NEA TDB review), or we can decide that this value has to be put in the table of selected SIT interaction coeffi- cients in Appendix B. Obviously, without explicitly saying so, Grenthe et al. decided to follow the first line of argument; and this review decided to follow the (not explicitly . stated) decisions in [92GRE/FUG]

146 [*]

147 Chapter VI VI Discussion of data selection for oxalate compounds and complexes ( .1)Equation Section 6 VI.1 Introduction Oxalic acid (ethane-1,2-dioic acid) ha s the chemical formula HOOC-COOH (C ; H O 2 2 4 –1 molecular weight = 90.0355 g·mol ; CAS Registry Number: 144-62-7). The anhydrous acid is not found in natu re and must be prepared from the di- –1 hydrate (C O; molecular weight: 126.066 g·mol H ; CAS Registry Number: O ·2H 4 2 2 2 6153-56-6) even when produced industrially. Oxalic acid is widely distributed in the vegetal and animal kingdoms, nearly always in the form of its salts. The potassium salt is found in common sorrel ( oxalis acetosella ) and the name oxalic acid is derived from that plant. Oxalic acid appears to have been discovered by Scheele, but the first account of its properties has been published by Bergman in 1776 [1808THO] . In his seminal paper “On Oxalic Acid” Thomson [1808THO] reported detailed chemical analyses of oxalic acid, alkali and alkali earth oxalates concerning formation, water of crystallisa- tion and decomposition of these compounds. He interpreted his careful analytical work of elements and stated “from the knowledge in terms of the law of multiple proportions of this curious law, it is difficult to avoid co ncluding that each of these elements consist fixed proportions”, of atoms of determinate weight, which comb ine according to certain i.e ., this classic paper on oxalic acid contains the first systematic application of Dalton’s atomic hypothesis. Figure VI-1: Structural formula of oxalic acid O O C C — HO OH 105

148 VI Discussion of data selection for oxalate 106 d may dissociate in acidic aqueous so- The two carboxylic groups of oxalic aci lutions. In reactions and formulae in this review oxalic acid is denoted as H ox, and the 2 2– oxalate ligand in aqueous solutions is denoted as ox . In formulae where the abbrevia- tion is preceded by a lower case it is en closed in parenthesis, as in Ca(ox)·H O(cr) or 2 – – Na(ox) , but no parentheses are needed in formulae as Kox . Also the protonated form + in parenthesis, as in Ca(Hox) of the oxalate ligand is enclosed . Oxalate may coordinate metal ions with its two carboxylic groups. Because of its ability to form 5-membered chelate rings, oxalate is quite a strong complexing agent. Complexes with stoichiometry 1:x (x > 1) are common. However, the limiting complex actually formed in aqueous solution depends on the charge and size of the metal ion. this review are stability constants. In Most thermodynamic constants selected in order to convert these stability constants into values of standard Gibbs energy of forma- ο 2– tion the value of ∆ (ox , 298.15 K) is needed. The derivation of this constant is not G fm ο a straightforward procedure. The enthalpy of formation of oxalic acid ∆ (H ox, aq, H 2 fm 298.15 K), derived from heat of combustion and heat of dissolution of H ox·2H O(cr) 2 2 ο ( Section VI.2), and the standard enthalpy changes of oxalate protonation ( ∆ (1) cf. H rm ο ο 2– and cf. H Section VI.3) are used to calculate H ∆ ∆ (ox (2), , 298.15 K). Solubility fm rm ο 2– S and calorimetric data of Ca(ox)·H O(cr) are then used to derive , 298.15 K) ( cf . (ox 2 m ο 2– Section VI.5). The latter two quantities allow calculating , 298.15 K). There (ox ∆ G fm is no redundant information to conf irm the results of this procedure. VI.2 Oxalic acid The solid formed at equilibrium in the system oxalic acid – water is the dihydrate, H ox·2H O(cr). It forms colourless and odourless crystals and it is stable in air at room 2 2 temperature. Anhydrous oxalic acid, H ox(cr), is prepared by dehydration of 2 ox·2H C results in needle-shaped crystals of O(cr). Vacuum sublimation at 80 – 100 ° H 2 2 β -H ox, whereas slow evaporation of acetone or ether solutions under N atmosphere 2 2 deposits crystals of α -H ox [53BRA/COT] , [74DER/SMI] . 2 VI.2.1 ox(cr) H 2 crystalline forms, the orthorhombic Anhydrous oxalic acid exists in two -H ox (space α 2 group ) and the monoclinic β -H ox (space group P 2 Pcab / c ). The dimorphism of 2 1 H ox(cr) has been elucidated already in the ear ly days of crystallographic X-ray studies 2 and Hendricks who determined the [35HEN] [24HOF/MAR] by Hoffmann and Mark unit-cell dimensions, space groups, and the main structural features. The first detailed crystal structure analysis of α -H ox has been reported by Cox et al. [52COX/DOU] . 2 α -H Very precise structures of ox and β -H ox have been determined by Derissen and 2 2 Smit [74DER/SMI] . ox(cr) – H ox·2H O(cr) – The vapour pressure measurements in the system H 2 2 2 water vapour by Bradley and Cotson [53BRA/COT] showed that α -H ox is the stable 2

149 VI.2 Oxalic acid 107 modification at room temperature. ox·2H Their experiments indicate that H O(cr) gives 2 2 the -form on dehydration, and that a mixture of the α - and the β -form slowly reverts to β -from at temperatures below 100 ° C. The thermodynamic α → β transition the pure α C, as determined by differential scanning calo- temperature is in the range 120 – 123 ° [78DER/SMI] rimetry . ο ∆ H The heat of combustion, , of anhydrous oxalic acid, H ox(cr), has been 2 cm th studied extensively since the advent of combustion calorimetry in the 19 century , [1880REC] , [1882THO] , [1884STO/REC] , [1885STO] , [1886LOU] , [1875BER3] [1889JAH] . , [34BEC/ROT] , [64WIL/SHI] , [26VER/HAR] [1889STO/KLE] , calorimetry reveal rather consistent Although the early studies in combustion results (in Table VI-1, note that the discrepant results of [1884STO/REC] and [1885STO] have later been rejected by the same group on the basis of improved new measurements [1889STO/KLE] ), studies published before the 1930’s are in general rejected in this review. As discussed by Domalski [72DOM] , international agreement on a chemical standard to be used in calibrating bomb calorimeters came during the 1920’s, and substantial improvements in calorimetric procedures, measuring instru- ments, and calculative methods were achieved in the 1930’s. Hence, from the above list only the results of Becker and Roth [34BEC/ROT] and Wilhoit and Shiao [64WIL/SHI] are further considered by this review. ο In order to calculate the enthalpy of formation (H ∆ H ox, cr, 298.15 K) from 2 fm ο H ∆ the corresponding heat of combustion (H ox, cr, 298.15 K) according to the reac- 2 cm tion: C U O O(l) (VI.1) (cr) + ½O (g) + H (g) H 2CO 2 2 2 4 2 2 only the enthalpies of formation of CO (g) and H th constants are O(l) are needed. Bo 2 2 ο : 0.13) H , g, 298.15 K) = – (393.51 ∆ (CO ± taken from CODATA [89COX/WAG] 2 fm − 1 ο 1 − kJ·mol ± ∆ (H and O, l, 298.15 K) = – (285.83 . The enthalpy of 0.04) kJ·mol H 2 fm formation of O (g) is zero by definition. 2 In many papers the heats of combustion and the enthalpies of formation are ex- − 1 − 1 . In this review, values for the heats of combustion in kJ·mol pressed in kcal·mol − 1 were obtained by multiplying the values reported in kcal·mol by 4.184. Calculations of − 1 the enthalpies of formation in kJ·mol were obtained from the difference between the heats of combustion and the corresponding sum of the enthalpies of formation of the 1 − combustion products, both in kJ·mol . Following this path, rather than the multiplica- tion of the reported enthalpies of formation by 4.184, avoided rounding variations in the last place, which otherwise would occur. α -H The enthalpy of formation of ox has been calculated as a weighted mean 2 of the values reported by [34BEC/ROT] (who, according to their preparation proce- dures, most probably examined a rather pure α -phase) and [64WIL/SHI] (who explicitly

150 VI Discussion of data selection for oxalate 108 presence of the action to avoid the -phase), and β checked their samples by X-ray diffr the following value has been selected: ο − 1 H ∆ , 298.15 K) = − (H ± 1.5) kJ·mol (828.8 ox, . α 2 fm There are no heat of combustion data for β -H ox. However, the heat of the 2 ο α→β transition, H ∆ , has been derived from vapour pressure measurements trs m [53BRA/COT] [72DOM] , and it was directly determined by a and heats of solution data differential scanning calorimeter (DSC) study [78DER/SMI] . Note that the value –1 ∆ and quoted by H given in Table I of [88PET/TSY] (120ºC) = 1.3 kJ·mol m trs [96DOM/HEA] as new measurements actually were taken from [78DER/SMI] . As discussed in [78DER/SMI] , the value for ∆ , derived from heats of H m trs - and β [53BRA/COT] sublimation of α -oxalic acid by seems too large. This is corrobo- α rated by an unpublished re-determination of the vapour pressure of -H ox cited in 2 [78DER/SMI] and , and by an also unpublished study of heats of solution data of α [72DOM] -oxalic acids cited in β . Hence, this review prefers the direct measurement by –1 ∆ H DSC [78DER/SMI] ± 0.25) kJ·mol , and assumes that does (120-123ºC) = (1.30 trs m ining this value with the selected value not significantly vary with temperature. Comb ο for α ∆ (H , 298.15 K) the following value is selected: ox, H 2 fm ο 1 − H ∆ . ox, β , 298.15 K) = − (827.5 ± 1.5) kJ·mol (H 2 fm Heat capacity and entropy data of anhydrous oxalic acid has been reported in [39SAT/SOG] several papers . , , [64DAV] , [82LUF/REE] [85DAL/GUS] [39SAT/SOG] Sato and Sogabe used an ice calorimeter to measure the specific C. However, they only ° heat of anhydrous oxalic acid in the temperature range 0 – 100 ο 100 C 1 − 1 − report a mean value [] ox, cr) = 118.0 J·K , and according to their C ·mol (H 2 ,m 0 p ° C until its weight became i.e preparation procedure, ., “oxalic acid was dehydrated at 90 constant and was used in the measurement” , they might have measured a mixture of - α and -H β ox. Hence, the results of [39SAT/SOG] are not credited in this review. 2 David [64DAV] applied differential thermal analysis (DTA) and gives a mean 1 − 1 − value ox, cr) = 146 J·K of the anhydrous oxalic . Neither the preparation C ·mol (H 2 p ,m measurements are reported and thus, the acid nor the temperature range of the DTA [64DAV] results of are not considered in this review. Dalidovich [85DAL/GUS] used a dynamic calorimeter to measure the et al. heat capacities of H ox(cr) and H ox·2H O(cr) and give coeffi cients for a four- 2 2 2 parameter heat capacity equation in the range 150 – 360 K and 170 – 300 K for H experimental data are reported ox(cr) and H O(cr), respectively. However, no ox ⋅ 2H 2 2 2 in [85DAL/GUS] . Furthermore, they state that “an hydrous samples were prepared by drying the crystallohydrates in a thin layer at the dehydration temperature” but no at- tempts were reported to check the α - and β -phase problem. Hence, their value − 1 − 1 is not credited in this review. ox, cr, 298.15 K) = 104.8 J·K C (H ·mol 2 p ,m

151 VI.2 Oxalic acid 109 [82LUF/REE] report a careful low-te Luff and Reed mperature heat capacity study of -H ination of the material showed the α ox. They state that “petrographic exam 2 -form anhydrous oxalic acid” and report the α material to be essentially homogeneous erature range 10 – 320 K. The standard mo- experimental heat capacity data in the temp are selected lar heat capacity and entropies determined by Luff and Reed [82LUF/REE] in this review, with the following assigned uncertainties derived from deviations of [82LUF/REE] observed heat capacities from smoothed values, Fig. 1 in , − 1 ο − 1 ox, 0.2) J·K , 298.15 K) = (105.9 ± (H C ·mol α 2 ,m p ο − − 1 1 S α ± 0.2) J·K . ·mol (H , 298.15 K) = (115.6 ox, 2 m The standard Gibbs energy of formation of α -H ox is calculated from the val- 2 enthalpy of formation selected above: ues of the standard entropy and of the ο 1 − G ∆ α , 298.15 K) = ox, (698.5 ± 1.5) kJ·mol (H . − 2 fm The enthalpy of solution of H ox(cr) has been evaluated calorimetrically 2 − ο 1 [86APE] = (8.85 0.20) kJ·mol H ∆ by Apelblat ± . For a lower temperature and a sol m H ∆ ± = (8.49 0.24) obtained finite amount of water, Becker and Roth [34BEC/ROT] m sol − 1 kJ·mol in good agreemen t with the more accurate value of [86APE] . ο H ∆ Table VI-1: Literature values for . Uncertainties estimated in this review. rm − 1 Enthalpy changes (kJ Method mol ) Reference ⋅ H ox(cr) 2 cal H ∆ (17ºC) = – (252 ± 20) [1875BER3] cm cal ∆ (18ºC) = – (248 ± 21) [1880REC] H cm H ∆ = -250 [1882THO] cal cm H ∆ (18ºC) = – (214 ± 13) [1884STO/REC] cal cm H ∆ (18ºC) = – (215 ± 13) [1885STO] cal cm (17ºC) = – (265 ∆ ± 25) [1886LOU] H cal cm H ∆ (0-30ºC) = – (253 ± 5) [1889JAH] cal cm [1889STO/KLE] H ∆ (20ºC) = – (252 ± 5) cal cm ± H ∆ (18-50ºC)= – (25.8 2.2) [10JOR] vapour pressure hyd m H ∆ (21ºC) = – (251.5 ± 1.4) [26VER/HAR] cal cm [34BEC/ROT] H ∆ (20.5ºC, H ± ⋅ 2100H 0.24) O(l)) = (8.49 ox cal 2 2 m sol ο b, c ∆ (25ºC) = – (245.6 ± 2.3) [34BEC/ROT] H cal cm ο H [34BEC/ROT] (25ºC) = – (827.3 ± 2.3) ∆ cal fm (Continued on next page)

152 VI Discussion of data selection for oxalate 110 Table VI-1 (continued) − 1 Enthalpy changes (kJ ) Reference ⋅ Method mol ox(cr) H 2 ∆ = 4.9 [53BRA/COT] vapour pressure H m trs ο a (25ºC) = – (242.9 ± 2.0) H cal ∆ [64WIL/SHI] cm ο ∆ (25ºC) = – (830.0 ± H [64WIL/SHI] 2.0) cal fm –1 e ∆ [72DOM] H = 1.3 kJ·mol cal trs m H ∆ (120-123ºC) = (1.30 ± 0.25) [78DER/SMI] DSC trs m ο H → ∆ (25ºC, m 0) = (8.85 [86APE] 0.20) ± cal m sol ο 2.1) [86APE] (25ºC) = – (26.5 ± cal ∆ H hyd m H ox ⋅ 2H O(cr) 2 2 cal (?ºC) = – 224.7 [1895JOR/STA] H ∆ cm O(l)) = (36.06 ± cal ox ⋅ 2100H (19.7ºC, H ∆ 0.45) [34BEC/ROT] H 2 2 m sol ο c ∆ (25ºC) = – (217.5 ± 1.5) H [34BEC/ROT] cal cm ο H ∆ (25ºC) = – (1427.0 ± 1.5) [34BEC/ROT] cal fm ο d ∆ (25ºC, m → H 0) = (35.65 ± 0.38) [52SPE/MIL] cal sol m ο H ∆ (25ºC, m → 0) = (35.23 ± 0.14) [86APE] cal m sol ) = 26.5 / ∂ T [87APE/MAN] H ∆ ( m ∂ m sat sat sol m a: The conversion of combustion da [64WIL/SHI] for the process at con- ta presented in Table IV of ο ο ∆ , to that at constant pressure is made by means of the equation H ∆ = U stant volume, cm cm ο ∆ + ∆ n R T U ∆ n = 1.5 is the difference between the nu mber of moles of gaseous products , where cm 1 − at ., n R T = 3.72 kJ·mol i.e in combustion Reaction (VI.1), and gaseous reactants involved ∆ 298.15 K. ox·2H , of [34BEC/ROT] on H H ∆ O(cr) were combined with their data b: The combustion data, 2 2 cm O(cr) to obtain the value H ∆ , of H of ox(cr) and H ∆ ox ⋅ 2H H on the heats of solution, 2 2 2 cm m sol ox(cr). H 2 . [72DOM] H ∆ corrected to the refere nce temperature 298.15 K by c: cm . Recalculated and extrapolat ed to infinite dilution by d: [86APE] α - and Derived from the unpublished data of Taylor on the heats of solution of -oxalic acids as e: β . cited in [72DOM] H O(cr) ox VI.2.2 2H ⋅ 2 2 lline forms; both modifications are monoclinic Oxalic acid dihydrate exists in two crysta and are distinguished only in their space group settings: 2 O and / n P α -H ox·2H for 2 2 1 P 2 O(cr) was not realised in the early / a for β -H ox·2H ox·2H O. The dimorphism of H 2 2 2 1 2 . Decades later, Dunitz and Robertson [24HOF/MAR] days of X-ray crystallography [47DUN/ROB] and Ahmed and Cruickshank [53AHM/CRU] discuss and try to re- interpret older inconsistent data, actually describing the structure of α -H O, but ox ⋅ 2H 2 2 O as a possible explanation for β -H 2H ox ⋅ they still are not aware of the existence of 2 2 the unsatisfactory agreement of reported crysta l structures. Only the advent of neutron

153 VI.2 Oxalic acid 111 [69COP/SAB] , . Meanwhile, very diffraction resolved the puzzle [69COP/SAB2] accurate structure data are available and -H ⋅ 2H O(cr) is now used as a standard to α ox 2 2 [2000HER] . , calibrate X-ray diffractometers [94LEH/LUG] Concerning the differences between -H ox ⋅ 2H α O and β -H O, Coppens ox ⋅ 2H 2 2 2 2 [69COP/SAB] and Sabine found that “intramolecular dist ances and angles in the oxalic acid molecule are remarkably similar in the two polymorphs”. Moreover, a curious experience reported in [69COP/SAB] and β form α is important for this review: “The Craven at the Crystallographic Laboratory appear to grow at random. Initially Dr B.M. of the University of Pittsburgh attempted to grow the α form. The first crystal from the solution was α . The α crystal was lost and subsequently only the β and the second β form would grow at Pittsburgh. We repeated Craven’s experiments at Brookhaven and obtained only the α form. No attempt was made to examine this effect further.” No study of the thermodynamic properties of H ox ⋅ O(cr) could be identi- 2H 2 2 fied in this review which would distinguish between and β α form. However, consider- , the differences in thermo- ing the results and experiences quoted above [69COP/SAB] dynamic properties of the α β form are perhaps so small that they are already in- and cluded in the uncertainties assigned to the values selected in this review. The heat of combustion of oxalic acid dihydrate has been determined by Joris- [1895JOR/STA] sen and van de Stadt [34BEC/ROT] . Only the and by Becker and Roth latter study is considered in this review, and the selected enthalpy of formation value of [34BEC/ROT] as re-evaluated by Domalski [72DOM] is: 1 − ο 1.5) kJ·mol ∆ ⋅ 2H (H O, cr, 298.15 K) = − (1427.0 ± ox H . 2 2 fm As a consistency check, the enthalpy change for the hydration reaction: H ox( α ) + 2 H O(cr) (VI.2) O(l) U H 2H ox ⋅ 2 2 2 2 ο − 1 is calculated to be ∆ = − (26.5 ± , in agreement with the experimen- ⋅ mol H 2.1) kJ m hydr 1 ο − 1 − tal values = − (25.8 ± 2.2) kJ ⋅ mol and H [10JOR] − (26.34 ± 0.17) kJ ⋅ mol ∆ m hydr . [86APE] Dalidovich et al. report coefficients fo r a four-parameter heat [85DAL/GUS] capacity equation in the range 170 – 300 K for H ox ⋅ 2H O(cr). However, no experimen- 2 2 , and their value O, cr, 298.15 K) (H ox ⋅ 2H [85DAL/GUS] tal data are reported in C 2 2 p ,m 1 − − 1 = 185.1 J ⋅ K is not credited in this review. mol ⋅ No low-temperature heat capacity study could be identified by this review and O(cr) cannot be calculated ox ⋅ 2H hence, the standard Gibbs energy of formation of H 2 2 because of the missing value fo r the standard entropy of H ox ⋅ 2H O(cr). 2 2

154 VI Discussion of data selection for oxalate 112 VI.2.3 ox(aq) H 2 ox 2H O(cr) in pure water and in various acids has been studied The solubility of H ⋅ 2 2 [23HER/NEU] , [28FLO] , [31CHA/BEL] , [36TRA] , [49REA] , , extensively [12MAS] [51NOR] [82WIL/SUL] and , . In pure water the solu- , [72KUK/KOB] [87APE/MAN] bility of the dihydrate is quite high. The following values have been reported: 1.48 –1 –3 –3 mol·dm , 1.18 mol·dm , 1.21 mol·kg [12MAS] (25°C) (30°C) (25°C) [23HER/NEU] –1 –1 [28FLO] (30°C) (25°C) [87APE/MAN] . [51NOR] , and 1.31 mol·kg , 1.57 mol·kg [87APE/MAN] The value determined by is considered to be the most reliable one by this review. The solubility of oxalic acid dihydrate is of interest because it could be used to ο calculate the value of G (H ox, aq, 298.15 K) from the following reaction, ∆ 2 fm H ⋅ 2H U O(cr) ox H O(l) (VI.3) ox(aq) + 2 H 2 2 2 2 ο ο ο ο G (H 2H ox, aq) + 2 G ∆ ∆ (H ⋅ O, l) − ∆ (VI.3) = G G ∆ (H ox O, cr) 2 2 2 2 fm fm rm fm ο ο ∆ (VI.3) = − G log K (VI.3) / (R T ln (10)) = + aa 2 log log H O 10 10 10 rm H ox 22 Water activities, a , for binary oxalic acid – water mixtures have been HO 2 and 0.25 – 1.23 molal measured in the range 0.4 – 1.1 molal [96KIR/MAU] which, in combination with the reported oxalic acid solubility [2001MAF/MEI] ο ο G ∆ G ∆ , allow estimation of O, cr) ox (VI.3). However, (H ⋅ 2H [87APE/MAN] 2 2 fm rm low-temperature heat use of the lack of a could not be calculated in this review beca ⋅ O(cr). 2H ox capacity study for the determination of the standard entropy of H 2 2 The enthalpy change for Reaction (VI.3) may be determined either calorimetri- cally or from the temperature dependence of the solubility of the saturated solutions. The values found in the literature are given in Table VI-1. The calorimetric values in dilute solutions [34BEC/ROT] [52SPE/MIL] , , should be more accurate than the value derived from the temperature depend- [86APE] . Selecting the weighted [87APE/MAN] ence of concentration of saturated solutions average of the two values determined fro m solution calorimetry at 25°C and extrapo- , leads to: [86APE] [52SPE/MIL] lated to infinite dilution ο ο − 1 H ∆ (VI.3) = (35.28 ± 0.13) kJ ⋅ mol = H . ∆ sol m rm ο ∆ H Using the value of (VI.3) determined calorimetrically and rm ο H ∆ (H ox ⋅ 2H ected above leads to: O, cr, 298.15 K) sel 2 2 fm 1 − ο ∆ ox, aq, 298.15 K) = − (820.1 ± 1.5) kJ (H mol H . ⋅ 2 fm As a consistency check, the enthalpy change for the reaction: H ox(aq) (VI.4) U H ox(cr) 2 2 1 ο − is calculated to be , in agreement with the values ∆ (VI.4) = (8.7 ± 2.1) kJ ⋅ mol H m sol ο − 1 − 1 H ∆ H ∆ mol = (8.85 ± [86APE] , and mol ⋅ 0.20) kJ ⋅ (20.5ºC) = (8.49 ± 0.24) kJ m sol m sol

155 VI.2 Oxalic acid 113 [34BEC/ROT] obtained calorimetrically. ο The selected value of ∆ (H ox, aq, 298.15 K) and th e selected values of H 2 fm ο ∆ H for oxalate protonation ( cf . Section VI.3.5) were used to calculate: rm ο –1 2– , 298.15 K) = H (830.7 ± 1.6) kJ·mol (ox . − ∆ fm ο 2– From this value and the selected values S , 298.15 K) = (47.6 ± 3.0) (ox m ο − –1 2– 1 − 1 G ∆ mol , 298.15 K) = − (680.1 ± 1.8) kJ ⋅ and ·mol (ox ( cf . Section VI.5.1) J·K fm cf . Section VI.3.5) it is pos- together with the selected pr otonation constants of oxalate ( sible to calculate: ο − 1 ∆ G ox, aq, 298.15 K) = − (712.4 ± 1.8) kJ ⋅ mol (H 2 fm 2– which may be combined with the entropy for ox selected in Section VI.5.1 and the otonation of oxalate selected reaction entropy changes for the pr listed in Table III-2 yielding: ο –1 –1 ox, aq, 298.15 K) = (191.3 ± 3.5) J·K (H ·mol S . 2 m The partial molar heat capacity at infinite dilution for undissociated oxalic acid 1 − 1 − ο has been reported to be 3.0) J ⋅ K (H ± ⋅ mol ox, aq, 298.15 K) = (95.1 C 2 ,m p . [89SIJ/ROS] The partial molar volume of the undissociated acid in water has been reported ο 3 − 1 V in several papers: ± 1.0) cm (H ⋅ mol 0.6) ox, 298.15 K) = (49.1 [75HOI] , (49.4 ± 2 m 3 − 1 3 − 1 [89SIJ/ROS] ⋅ mol ⋅ mol . The average of this , and 46.33 cm [90APE/MAN] cm 3 − 1 ο ± 2) cm . For the ⋅ mol (H V ox, 298.15 K) = (48 somewhat discrepant data set is 2 m ο − V (Hox , 298.15 K) = determined [90APE/MAN] oxalate ions Apelblat and Manzurola m − 1 ο 2 − 3 3 − 1 V , 298.15 K) = (25 (ox ⋅ mol ± ± 2) cm 2) cm ⋅ mol and (uncertainties assigned (34 m by this review). VI.3 Protonation constants of oxalate VI.3.1 Introduction Oxalic acid is a diprotic acid. In this ox, and the review oxalic acid is denoted as H 2 r = 1 and 2: standard TDB nomenclature is used for the protonation of the ligand with (2) r − [H ox ] + r ( ( 3) − 2) − r r ox K + H ox = U H H ( r 1) − r r − + r (3) [H ] [H ] ox − r (1) − r (2) ] [H ox − + − 2) 2 r ( r = H U . H r β + ox ox r r −+ r 2 [ox ] [H ] + In one study [70LUK] ox the formation of H in strong acids has been postu- 3 lated based on electromigration experiments in concentrated perchloric acid solutions. + ox has not been confirmed by any other However, the existence of the species H 3 method, and this review does not consider this cationic species.

156 VI Discussion of data selection for oxalate 114 150) were found from a literature search for A large number of references ( ≈ the acid-base equilibria of oxalate. The majority of these references contain studies on metal complexation, where the authors needed values for the dissociation constants of oxalic acid under the same experimental conditions as the metal-complexation study. dvantageous to judge the Because of the large numb er of references, it was a quality of the experimental details quite rigorously. The following criteria were consid- ered when discarding references within the screening process: The dissociation constants of oxalic acid reported in the following references were • discarded from the Clear indication must be given that the acid-base constants were determined experimentally in the actual study, and not taken from another publication. The calibration method for the pH-electrode must be indicated. They must have been • calibrated in the concentration scale, and not with standard pH-buffers. That is, “pH” + − log [H ] . References where discarded when they reported mixed must refer to 10 , involving both proton activities and ligand concentrations. i.e. equilibrium constants, s electrodes were calibrated with standard In some cases it is reported that the glas + ] calculated from pH, for example with the Davies equation. This pro- buffers, and [H cedure is not accepted in this review. A background electrolyte providing a constant ionic strength must be used. • • The temperature, ionic strength, and the natu re of the background electrolyte must be given. The dissociation constants of oxalic acid reported in the following references were discarded from the review procedure because they did not fulfil one or more of the , , [28SIM] , [28SIM2] , [31GAN/ING] , [36BRI/JAR] [24LAR] criteria indicated above: , [40VOS/BEC] , [38CAN/KIB] , [57BAB/DUB] , , [54SCH/LAI] [37CLA/VOS] [58GEL/MAT] , [58MOS/GEL] , [60YAM/DAV] , [61GRI/AST] , , [57DUT/SUR] , [62BRU/KAI] , [62MCD/LON] , [63PAR/BAR] , [63STA] , [64SEK] , [61ZOL/MAR2] , , [66MAK/YUS] , [68HAR] [66LYL/NAQ] [69HAV] , [70COU/FAU] , , [65SEK4] [72MEN] , , [71GOR/MIK] , [71GEL] , [73CHU/SKO2] , [74MOR/SEK] , [70GRE/BRY2] , , [74SIN/TAN] , [75BOT] , [75FRI/PIZ] , [75SIN/TAN] , [74SIN/GHO] [74NAZ/FLY2] , , [76YAD/GHO] , [77DEL2] , [77DUC/BER] , [76BRI/ELD] [75VOT/BAR] , [77REB/BAR] , [78JAI/SHA] , [78SIN] , [79ELE/ELD] , [79FRI/SYC] , [77JAI/KUM] , [82INO/TOC] , [82SAC/CHA] [83BAR/SUS] , [83GUP/CHA] , , [80MIS/MIT] , , [84VEN/SWA] , [84ZHO/XU] , [86CHO/DAD] , [87GAM] [83OLI/WIK] , , [88GHA/MAN] , [88DAS/KAL] , [89PRA/RAO] , [89DAS/DAS] [88CAS/FAU] , [90ULL/BHA] , [92DAN/ROB] , [92TRI/TRI] , [96DAS/MOH] , [89RED/RAM] , [99CIA/IUL] , [2001MAF/MEI] . These references are not discussed in [98ALD/BIA] Appendix A. A few references found in the databases of Smith & Martell [2003SMI/MAR] obviously are typo errors. They do not contain dissocia- [2002PET/POW] and IUPAC , [50MEI2] , tion constants of oxalic acid or do not even refer to oxalate at all: [50MEI]

157 VI.3 Protonation constants of oxalate 115 , [60SCH/FIS] [70BRA/LEP] , [70SCH/HOW] , [79MAY/CHA] , [57GEL/MAT2] , . [89FUE/REB] Data from a few other references c ould not be accepted for other reasons. [39PAR/NIC] , , , [70DAV/WAT] , [49REA] [39PAR/GIB] These references [72NIK/ANT] , [81SIN/GHO] , [85RED/RAO] , [86RED/RAO] , , [70STE/PAZ] [94ERT/MOH] , [94RED/SHI] , [96ARA/ARC] , , , [96BOR/LIS] [90RED/SAT] are discussed in Appendix A. [98KHA/RAD] The data reported in the remaining references are listed in Table VI-2. If data were reported over a range of temperatures these values have been used T -variation of in this review for the determination of enthalpy changes from the log K [39HAR/FAL] , [48PIN/BAT] , , see Section VI.3.4. In these cases 10 [61MCA/NAN] , [70DAV/WAT] , [76KAL] [69KUR/FAR] [90ROB/STE] , , , [92ROB/STE] [98KET/WES] ( cf. Appendix A) only , log K data at 25°C and I > 0 10 ο were included in the weighted least-squares procedure for determining log . K r 10 If data were reported solely at temperatures not too different from 25°C the K data have been extrapolated to 25°C using the temperature parameters evalu- log 10 ated in this review. These are: [69GRE/GAR] , [72MAG/BIS] (20°C) and [76MAK/TOU] , , [82JAC/COS] , [83DAN/RIG] , [83DAN/RIG2] (37°C). [81DAN/RIG] thalpies selected in The corrections were obtained from reaction en Section VI.3.4. Ionic cf media corrections to the reaction en thalpies consist of two parts ( . Section V.3.6): a Debye-Hückel expression, and a specific ion interaction ( ∆ε ) term. The resulting reac- L tion enthalpies are all relatively small, and because of the limited temperature interval involved (– 5°C and + 12°C), the calculated corrections for log were always K r 10 0.05 log ≤ -units. 10 A single measurement in Bu NClO , the value reported at 40°C [69POS/VED] 4 4 in LiClO at 70°C [88KIT] , and the value reported in NaClO at 60°C [2000CIA/IUL] 4 4 were not included in the weig hted least-squares procedure. Uncertainties reported in the original publications have been multiplied by a i.e. , including factor 1.96 to obtain error limits closer to a 95% total uncertainty level, random and possible systematic deviations. In cases where no uncertainty limits were reported, a value of 0.1 log -units has been used in the weighted least-squares proce- ± 10 dure. Reported uncertainties below ± 0.01 in log K in the original papers were in- 10 1 creased to ± 0.02 ( ≈ ± 0.01 × 1.96), and reported uncertainties below ± 0.05 in K log 2 10 in the original papers were increased to ± 0.1 in the least-squares regression analysis. The data finally selected for multi-linear least-squares regressions for deter- ο K log mining are listed in Table VI-3 and Table VI-4. The protonation constants n 10 were corrected to molal un its and extrapolated to 25 ° C where necessary.

158 VI Discussion of data selection for oxalate 116 Table VI-2: Literature data on the protonation constants for oxalate considered in this review. Data in Uncertainties are included as italics were reported in molal units. reported in the references. I t ( ° C) (M) Method log Electrolyte Reference K log K 2 10 1 10 0 NaCl 0 4.228 pot → [39HAR/FAL] 5 4.235 10 4.244 15 4.255 20 4.268 25 4.286 30 4.308 35 4.331 40 4.356 45 4.388 50 4.417 25 4.276 1.27 [41DAR] con → 0 0 NaCl 0 pot [48PIN/BAT] → 4.201 5 4.207 10 4.218 15 4.231 20 4.247 25 4.266 30 4.287 35 4.312 40 4.338 45 4.369 50 4.399 ± 0.02) 25 (3.81 ± 0.02) [60MCA/NAN] (1.37 pot 0.1 NaClO 4 pot HCl 0 1.24 [61MCA/NAN] → 0 1.25 15 25 1.25 35 1.29 45 1.29 [65BAU/SMI] 5 (3.63 ± 0.01) (1.32 ± 0.01) pot 0.5 NaClO 4 15 (3.65 0.01) (1.23 ± 0.01) ± ± 0.01) (1.20 ± 0.01) 25 (3.67 [65BOT/CIA] 25 (3.57 ± 0.02) (1.07 ± 0.05) pot 1 NaClO 4 25 3.85 1.13 [65NAG/UMA] pot 0.1 NaClO 4 [66LHE/MAR] 25 (3.81 ± 0.01) pot 0.1 KNO 3 [66MOO/SUT] 25 (1.32 ± 0.02) pot 0.1 NaClO 4 (Continued on next page)

159 VI.3 Protonation constants of oxalate 117 Table VI-2: (continued) Method Electrolyte t ( ° C) I (M) log K Reference K log 2 10 10 1 ± 0.01) (1.08 pot 1 NaClO 0.01) [66MOO/SUT] 25 (3.55 ± 4 ± 0.01) (1.26 ± 0.03) 3 (3.80 pot 1 KNO 25 (3.62 0.01) 1.1 [67RAJ/MAR] ± 3 [68DEN/MEI] ± 0.03) (1.00 ± 0.05) 25 (3.50 pot 0.52 LiClO 4 0.05) 25 (3.82 0.01) (1.26 ± ± [69CON/MAR] pot 0.1 KNO 3 pot 1 NaClO 20 (3.55 ± 0.01) (1.01 ± [69GRE/GAR] 0.02) 4 → 25 (1.30 ± 0.01) [69KUR/FAR] 0 HCl pot 0.01) ± 30 (1.31 (1.32 35 0.01) ± ± 0.01) 40 (1.33 0.01) ± (1.34 45 50 (1.36 0.01) ± ± 0.01) 55 (1.36 40 NClO 1.27 4.27 [69POS/VED] pot 0.1 Bu 4 4 pot 0.1 NaClO 1.28 [69VOR/IVA] 20 3.90 4 0.02) [70ASC/BRI] ± pot 0.5 NaCl 25 (3.60 [70CIA/GRI] 25 (3.35 ± 0.01) (0.84 ± 0.05) pot 1 LiClO 4 pot 1 NaClO ± 0.01) ± 0.01) [72MAG/BIS] 20 (3.60 (1.13 4 25 3.81 1.31 [73ARM/DUN] pot 0.1 NaClO 4 40 (1.03 0.01) [76KAL] ± pot 1 NaClO 4 (1.04 45 0.02) ± ± 0.01) 50 (1.03 0.04) ± (1.06 55 KNO 35 (1.04 ± 0.02) 3 0.00) ± 40 (0.96 ± 0.01) 45 (1.05 50 (1.00 ± 0.01) ± 0.03) 55 (1.05 [76MAK/TOU] ± 0.01) (1.09 ± 0.03) 37 (3.68 pot 0.15 NaClO 4 pot 0.1 KNO ± 0.01) [77BRO/PET] 25 (3.88 3 0.01) ± 0.01) (1.01 ± [77HED/OLI] 25 (3.56 pot 1 NaClO 4 ± ± 0.02) (1.25 25 (3.85 0.05) [80GRA/MUS] pot 0.1 KNO 3 pot 0.1 Et NI 37 (3.98 ± 0.02) [81DAN/RIG] 4 ± 0.02) 0.3 (3.98 ± 0.03) (3.86 0.1 NaNO 3 0.3 (3.71 ± 0.01) ± 0.01) (3.89 0.1 KNO 3 ± 0.02) 0.3 (3.73 (Continued on next page)

160 VI Discussion of data selection for oxalate 118 Table VI-2: (continued) K (M) ( ° C) Method I log K Reference t log Electrolyte 2 10 1 10 (1.40 0.01) 0.01) [82JAC/COS] ± pot 0.15 NaCl 37 (3.76 ± 25 (3.82 ± 0.01) (1.27 ± pot 3 NaClO [83CRU/HEY] 0.01) 4 sp (3.84 (1.26 ± ± 0.01) 0.01) ± 37 (3.70 [83DAN/RIG2] 0.01) pot 0.15 LiNO 3 (3.82 ± 0.01) NaNO 3 KNO (3.84 ± 0.01) 3 NI 37 (3.99 ± 0.01) [83DAN/RIG] pot 0.10 Et 4 ± 0.01) (3.98 0.30 (4.08 0.56 0.01) ± 0.01) ± (4.18 0.96 25 (3.75 ± (0.99 ± 0.01) [84JOH/JON] 0.01) pot 0.15 KNO 3 pot 0.04 Et NI 25 (4.00 0.02) [85ROB/STE] ± 4 ± 0.02) (3.93 0.25 ± 0.02) 0.81 (4.09 pot 0.6 NaCl 25 (3.57 0.01) ± ± 0.01) [85SJO/OEH2] (0.97 ± 0.02) (0.89 ± 0.02) [86CRU/HEY] pot 1 NaCl 25 (3.52 70 (3.46 ± 0.03) (1.09 ± 0.03) [88KIT] pot 1 LiClO 4 pot 0.5 NaCl 25 (3.60 (1.16 0.010) [89FUE/REB2] ± ± 0.03) NI 5 (3.987 0.006) [90ROB/STE] ± pot 0.04 Et 4 ± 0.006) (3.903 0.16 (3.926 0.36 ± 0.006) ± 0.006) (4.014 0.64 ± 0.006) (4.135 1 15 (4.005 0.04 ± 0.006) 0.006) ± (3.918 0.16 ± 0.006) (3.940 0.36 (4.027 0.64 0.006) ± ± 0.006) 1 (4.147 ± 0.006) 0.04 25 (4.039 0.16 (3.949 ± 0.006) ± 0.006) 0.36 (3.970 ± 0.006) (4.056 0.64 (4.174 1 ± 0.006) ± 0.006) 0.04 35 (4.086 0.006) ± 0.16 (3.993 0.36 (4.012 0.006) ± ± 0.006) 0.64 (4.097 ± 0.006) 1 (4.215 (Continued on next page)

161 VI.3 Protonation constants of oxalate 119 Table VI-2: (continued) Method Electrolyte t ( ° C) I (M) log K Reference K log 2 1 10 10 ± [90ROB/STE] pot 0.04 Et 0.006) NI 45 (4.144 4 0.006) 0.16 (4.049 ± (4.065 0.36 ± 0.006) ± 0.006) (4.149 0.64 ± 0.006) 1 (4.266 55 (4.213 0.04 ± 0.006) ± 0.006) (4.114 0.16 ± 0.006) 55 (4.128 0.36 0.64 (4.210 0.006) ± ± 0.006) 1 (4.326 [92AZA/ELN] 25 (3.87 ± 0.04) pot 0.1 KNO 3 0.04 NaCl 10 ± 0.01) (1.14 ± 0.10) [92ROB/STE] pot (3.91 0.10) ± 0.01) (1.08 ± 0.17 (3.73 ± 0.10) 0.01) (1.05 ± 0.40 (3.62 0.76 (3.55 0.01) (1.04 ± 0.10) ± ± 0.01) (1.02 ± 0.10) 1.25 (3.50 ± 0.01) (1.17 ± 0.10) 0.04 (3.95 25 0.17 (3.76 0.01) (1.11 ± 0.10) ± ± 0.01) (1.08 ± 0.10) 0.39 (3.65 0.10) ± (1.06 ± 0.01) (3.58 0.74 1.23 (3.54 (1.04 0.10) ± ± 0.01) 0.10) ± 0.01) (1.19 ± (4.00 0.04 37 0.10) ± 0.01) (1.13 ± (3.80 0.17 0.40 (3.69 0.01) (1.09 ± 0.10) ± ± 0.01) (1.07 ± 0.10) 0.74 (3.61 ± 0.01) (1.05 ± 0.10) 1.22 (3.57 0.04 45 (4.04 0.01) (1.21 ± 0.10) ± ± 0.01) (1.14 ± 0.10) 0.16 (3.84 0.10) ± (1.11 ± 0.01) (3.72 0.40 0.75 (3.63 ± 0.01) (1.11 ± 0.10) 0.10) 0.01) ± (1.09 ± 1.21 (3.59 ± 0.01) (1.15 ± 0.10) (3.93 0.04 KCl 10 0.17 (3.78 0.01) (1.10 ± 0.10) ± ± 0.01) (1.08 ± 0.10) (3.71 0.40 ± 0.01) (1.07 ± 0.10) (3.67 0.77 1.32 (3.63 0.01) (1.07 ± 0.10) ± ± 0.01) (1.18 ± 0.10) 25 (3.96 0.04 0.01) ± (1.12 ± 0.10) (3.80 0.17 (Continued on next page)

162 VI Discussion of data selection for oxalate 120 Table VI-2: (continued) Method t ( ° C) I (M) log K Electrolyte Reference log K 2 1 10 10 [92ROB/STE] ± (1.10 0.41 0.10) 0.01) (3.71 pot KCl 25 ± (3.65 ± 0.010 (1.10 ± 0.77 0.10) (3.61 1.30 (1.09 0.10) 0.01 ± ± (1.19 ± 0.10) ± 0.01) 0.04 37 (4.01 0.01) (1.13 ± 0.10) ± 0.17 (3.83 (3.72 0.41 ± 0.10) 0.010 ± (1.11 (1.11 ± 0.10) 0.01) ± 0.77 (3.64 ± 0.01) (1.10 ± 0.10) 1.30 (3.59 45 (4.05 0.04 0.01) (1.21 0.10) ± ± ± 0.01) (1.15 ± 0.10) (3.85 0.17 0.10) ± 0.01) (1.13 ± (3.73 0.40 (3.64 0.76 0.01) (1.14 ± 0.10) ± ± 0.10) 0.01) (1.13 ± (3.59 45 1.29 (1.09 0.04 NI 10 (3.96 ± 0.01) ± 0.08) Et 4 0.17 (3.87 ± 0.01) ± 0.08) (1.02 ± 0.01) (1.05 ± 0.08) 0.39 (3.87 0.08) ± (1.19 ± 0.01) (3.92 0.73 1.22 (3.98 (1.26 ± 0.01) ± 0.08) ± (1.13 ± 0.08) 0.01) (4.00 25 0.04 ± 0.08) 0.01) (1.05 ± (3.90 0.17 (3.91 0.39 ± (1.16 ± 0.08) 0.01) 0.08) ± ± 0.01) (1.22 (3.96 0.73 ± 0.01) (1.29 ± 0.08) 1.23 (4.04 0.04 37 (4.06 0.010 (1.14 ± 0.08) ± ± 0.01) (1.07 ± 0.08) 0.17 (3.96 ± 0.01) (1.18 ± 0.08) (3.96 0.39 0.73 (4.03 ± 0.01) 0.08) (1.23 ± ± 0.01) (1.29 ± 0.08) 1.24 (4.12 ± 0.01) (1.13 ± 0.08) 0.04 (4.11 45 0.17 (4.01 0.01) ± 0.08) ± (1.18 0.08) ± ± 0.01) (1.19 (4.01 0.39 ± 0.01) (1.25 ± 0.08) (4.07 0.73 (4.16 1.24 ± 0.01) (1.32 ± 0.08) [93GLA/MAJ] pot 0.15 NaCl 25 3.76 37 3.95 0.01) (1.08 ± 0.09) [94KIS/SOV] ± pot 0.2 KCl 25 (3.73 0.02) [95LU/MOT] ± pot 0.1 KCl 25 (3.86 pot 0.1 NaCl 25 3.822 [95MIR/SAD] (Continued on next page)

163 VI.3 Protonation constants of oxalate 121 Table VI-2: (continued) Method Electrolyte t ( ° C) I (M) log K Reference K log 2 10 10 1 pot 0.3 NaCl 3.645 [95MIR/SAD] 3.572 0.5 1 3.504 2 3.509 3.621 3 4 3.756 3.809 0.1 NaClO 4 0.3 3.633 0.5 3.565 1 3.542 3.622 2 3.811 3 4.012 4 25 (3.73 0.01) [96CHO/CHE] ± pot 0.3 NaClO 4 0.5 (3.65 (1.28 ± ± 0.01) 0.03) ± 0.01) (1.33 ± 0.04) (3.60 1 0.01) ± (1.49 ± 0.02) 3 (3.84 (4.23 5 0.01) (1.42 ± 0.04) ± 0.01) ± 0.01) (1.75 ± (4.71 7 ± (2.13 ± 0.05) 0.04) (5.11 9 pot 0.1 Me NI 25 3.87 [96XUE/TRA] 1.15 4 25 (3.64 0.01) (1.0 ± 0.1) [97BAR/CEC] ± pot 0.5 NaClO 4 25 (3.83 ± 0.01) [98FER/MAN] pot 3 NaClO 4 0.185 NaCl 0 (3.655 pot 0.01) [98KET/WES] ± ± 0.016) (3.577 0.334 ± 0.014) 2.002 (3.506 3.149 (3.553 0.014) ± ± 0.014) 5.000 (3.834 0.010) ± (3.803 25 0.100 0.185 (3.706 ± 0.009) ± 0.015) 0.334 (3.613 ± 0.015) 0.334 (3.618 1.001 (3.541 ± 0.015) ± 0.015) (3.553 1.001 ± 0.015) (3.556 1.001 2.002 (3.467 0.014) ± ± 0.014) (3.496 2.002 ± 0.014) 3.149 (3.517 (Continued on next page)

164 VI Discussion of data selection for oxalate 122 Table VI-2: (continued) Method Electrolyte t ( ° C) I (M) log K Reference K log 2 10 10 1 (3.733 0.014) [98KET/WES] pot ± NaCl 25 5.000 (3.910 ± 0.009) 0.100 50 (3.801 0.185 0.008) ± 0.334 (3.706 ± 0.015) 0.100 (3.561 ± 0.014) 0.185 0.334 (3.457 0.014) ± 2.002 (3.535 ± 0.014) (3.526 3.149 0.014) ± 3.149 (3.529 ± 0.014) 5.001 (3.677 ± 0.014) (3.684 5.001 0.014) ± 75 0.100 (4.052 ± 0.008) 0.185 (3.940 ± 0.008) 0.334 (3.832 0.015) ± 1.001 (3.649 0.014) ± 2.002 (3.568 0.014) ± (3.562 3.150 ± 0.014) 3.150 (3.573 ± 0.014) 5.002 (3.665 ± 0.014 (4.219 0.100 100 0.008) ± (4.112 0.185 ± 0.008) 0.334 (3.980 0.014) ± 1.002 (3.767 0.014) ± (3.678 2.003 0.014) ± 2.003 (3.675 ± 0.014) 3.151 (3.617 ± 0.014) 5.004 (3.666 0.014) ± 0.100 125 (4.406 ± 0.008) 0.185 (4.291 0.007) ± (4.145 0.334 ± 0.014) (3.898 1.002 0.014) ± (3.777 2.004 0.014) ± 3.153 (3.683 ± 0.014) 5.006 (3.690 ± 0.014) 0.100 (4.605 150 0.007) ± 0.185 (4.494 ± 0.007) (4.322 0.334 ± 0.014) 2.006 (3.880 ± 0.014) (Continued on next page)

165 VI.3 Protonation constants of oxalate 123 Table VI-2: (continued) Method Electrolyte t ( ° C) I (M) log K Reference K log 2 10 10 1 (4.036 ± pot [98KET/WES] NaCl 150 1.003 0.014 (3.880 ± 0.014) 2.006 3.156 (3.762 ± 0.014) ± 0.014) (3.713 5.012 0.007) ± (4.836 175 0.100 (4.741 0.185 0.007) ± 0.014) ± (4.495 0.335 ± 0.014) 1.005 (4.216 2.009 (3.977 ± 0.014) ± 0.014) (3.850 3.161 0.014) ± (3.749 5.020 3.172 (1.001 0 ± 0.086) ± 0.012) 4.998 (1.351 0.296) ± (1.271 0.100 5 0.100 (1.183 0.369) ± ± 0.098) 0.316 (1.051 0.098) ± (1.050 0.316 0.601 (0.989 ± 0.086) ± 0.089) 0.602 (0.974 ± 0.066) (0.961 1.002 25 0.100 (1.282 ± 0.235) 0.294) ± (1.184 0.100 ± 0.0800) 0.316 (1.068 (1.064 0.316 0.081) ± ± 0.079) 0.601 (0.953 0.078) ± (0.966 0.602 1.002 (0.941 ± 0.059) ± 0.089) (0.957 2.001 ± 0.013) 3.002 (0.981 (1.104 3.172 0.065) ± ± 0.014) (1.033 4.002 ± 0.012) (1.203 4.999 0.100 50 (1.310 ± 0.183) ± 0.222) 0.100 (1.216 0.178) ± (1.174 0.101 0.316 (1.089 ± 0.066) ± 0.066) 0.316 (1.090 (Continued on next page)

166 VI Discussion of data selection for oxalate 124 Table VI-2: (continued) Method t ( ° C) I (M) log K Electrolyte Reference log K 2 10 10 1 (1.009 ± 0.013) [98KET/WES] 0.332 NaCl 50 pot (1.033 0.061) ± 0.601 (1.033 0.602 ± 0.061) 0.047) ± (1.000 1.002 ± 0.043) 1.030 (0.901 (0.826 1.982 ± 0.086) 0.092) ± (0.808 2.002 ± 0.013) (0.978 3.002 (1.010 3.172 0.060) ± ± 0.060) 3.172 (1.024 ± 0.012) (1.034 4.002 5.000 (1.135 ± 0.011) 0.142) ± (1.351 0.099 75 ± 0.162) 0.100 (1.284 (1.248 0.101 0.125) ± ± 0.051) (1.159 0.316 ± 0.051) (1.156 0.316 0.332 (1.080 ± 0.012) ± 0.047) 0.601 (1.122 ± 0.047) 0.601 (1.121 (1.059 1.002 ± 0.039) ± 0.039) 1.030 (0.953 0.072) ± (0.864 1.982 2.002 (0.851 ± 0.075) 0.011) ± (1.017 3.003 0.048) ± (1.121 3.172 3.173 (1.075 ± 0.050) 0.010) ± (1.051 4.003 ± 0.010) 5.001 (1.127 100 (1.385 0.099 0.112) ± ± 0.116) 0.100 (1.369 0.087) ± (1.339 0.100 0.315 (1.238 ± 0.039) ± 0.039) (1.240 0.315 ± 0.013) 0.331 (1.169 0.601 (1.178 ± 0.039) ± 0.039) (1.177 0.601 ± 0.033) (1.124 1.001 (Continued on next page)

167 VI.3 Protonation constants of oxalate 125 Table VI-2: (continued) Method t ( ° C) I (M) log K Electrolyte Reference log K 2 10 10 1 (1.036 ± 0.035) [98KET/WES] 1.030 NaCl 100 pot (0.925 0.060) ± 1.982 (0.909 2.002 ± 0.061) 0.009) ± (1.064 3.004 ± 0.042) 3.173 (1.123 (1.154 3.173 ± 0.041) 0.009) ± (1.100 4.004 ± 0.009) (1.143 5.003 125 (1.489 0.099 0.083) ± ± 0.063) 0.099 (1.438 ± 0.086) (1.472 0.100 0.315 (1.338 ± 0.031) 0.031) ± (1.340 0.315 ± 0.014) 0.331 (1.270 (1.272 0.601 0.033) ± ± 0.033) (1.273 0.601 ± 0.029) (1.214 1.002 1.030 (1.146 ± 0.033) 0.051) ± (1.028 1.982 0.050) ± (1.002 2.003 (1.114 3.006 ± 0.008) ± 0.037) (1.188 3.175 ± 0.037) (1.192 3.175 4.007 (1.185 0.008) ± ± (1.318 SO 5 0.142) 0.101 NaCF 3 3 ± 0.150) 0.101 (1.290 0.101 (1.289 ± 0.151) ± 0.065) (1.084 0.316 ± 0.065) (1.085 0.316 0.600 (1.082 ± 0.039) 0.046) ± (1.002 0.600 ± 0.061) (1.040 1.001 1.001 (1.040 ± 0.061) ± 0.150) (1.236 0.101 25 ± 0.147) (1.244 0.101 0.101 (1.231 ± 0.152) ± 0.160) (1.206 0.102 (Continued on next page)

168 VI Discussion of data selection for oxalate 126 Table VI-2: (continued) Method t ( ° C) I (M) log K Electrolyte Reference log K 2 10 1 10 pot SO 0.316 25 NaCF (1.060 ± 0.062) [98KET/WES] 3 3 (1.048 ± 0.063) 0.316 0.600 (1.044 0.039) ± ± 0.043) 0.600 (0.995 ± 0.058) 1.001 (0.995 (1.010 1.001 ± 0.058) ± 0.122) 0.101 (1.248 50 ± 0.123) (1.242 0.101 0.101 (1.238 ± 0.124) ± 0.134) (1.203 0.102 ± 0.051) (1.082 0.316 0.316 (1.084 0.051) ± ± 0.033) (1.060 0.600 ± 0.036) 0.600 (1.019 1.001 (1.066 0.049) ± ± 0.049) 1.001 (1.070 0.088) ± (1.311 75 0.101 0.101 (1.305 ± 0.089) ± 0.093) 0.101 (1.282 ± 0.087) (1.319 0.101 0.316 (1.150 ± 0.037) 0.026) ± (1.105 0.600 ± 0.027) 0.600 (1.092 (1.142 1.000 0.042) ± ± 0.043) 1.001 (1.127 0.060) ± (1.405 100 0.101 0.101 (1.390 ± 0.062) ± 0.061) (1.400 0.101 ± 0.026) 0.315 (1.253 (1.184 0.600 0.020) ± ± 0.038) (1.197 1.000 ± 0.038) (1.197 1.000 0.101 125 (1.503 ± 0.045) ± 0.045) 0.101 (1.498 0.015) ± (1.293 0.600 1.001 (1.313 ± 0.034) ± 0.034) 1.001 (1.303 (Continued on next page)

169 VI.3 Protonation constants of oxalate 127 Table VI-2: (continued) Method t ( ° C) I (M) log K Electrolyte Reference log K 2 10 10 1 SO 0.100 NaCF 5 (3.760 ± 0.011) pot [98KET/WES] 3 3 (3.769 ± 0.100 0.011) (3.764 0.100 0.011) ± ± 0.011) (3.769 0.100 ± 0.011) (3.633 0.273 0.316 (3.599 ± 0.011) ± 0.011) (3.539 0.600 ± 0.011) (3.540 0.600 1.000 (3.520 ± 0.011) 0.011) ± (3.821 25 0.100 ± 0.011) 0.100 (3.826 0.100 (3.819 0.011) ± ± 0.011) 0.100 (3.819 ± 0.011) 0.273 (3.675 0.316 (3.641 0.011) ± ± 0.011) (3.571 0.600 0.011) ± (3.573 0.600 1.000 (3.541 ± 0.011) ± 0.011) 0.100 50 (3.936 0.011) ± (3.937 0.100 (3.931 0.100 0.011) ± 0.011) ± (3.930 0.100 ± 0.011) 0.273 (3.772 0.316 (3.739 ± 0.011) 0.011) ± 0.600 (3.656 ± 0.011) 0.600 (3.658 1.000 (3.615 ± 0.011) ± 0.011) (4.086 0.100 75 ± 0.011) 0.100 (4.074 0.100 (4.074 ± 0.011) 0.011) ± (3.905 0.273 ± 0.011) 0.316 (3.871 0.600 (3.775 0.011) ± ± 0.011) (3.778 0.600 ± 0.011) (3.721 1.000 0.100 100 (4.256 0.011) ± ± 0.011) (4.257 0.100 ± 0.011) 0.100 (4.243 (Continued on next page)

170 VI Discussion of data selection for oxalate 128 Table VI-2: (continued) Method Electrolyte t ( ° C) I (M) log K Reference K log 2 10 10 1 [98KET/WES] (4.246 ± 0.011) 100 pot SO 0.100 NaCF 3 3 (4.060 ± 0.011) 0.273 (4.027 0.316 ± 0.011) ± 0.011) 0.600 (3.919 ± 0.011) 0.600 (3.918 (3.849 1.000 ± 0.011) ± 0.011) 0.100 125 (4.450 0.011) ± (4.438 0.100 0.100 (4.440 0.011) ± 0.011) ± (4.234 0.273 ± 0.011) 0.316 (4.200 0.601 (4.076 0.011) ± ± 0.011) (4.076 0.601 ± 0.011) 1.001 (3.992 0.3 pot NaCl 25 (3.67 0.02) 0.05) [99MIZ/BON] (1.06 ± ± ± 0.02) (0.9 ± 0.2) (3.52 1 ± 0.02) (0.9 ± 0.1) 2 (3.54 3 (3.55 ± 0.02) (0.86 ± 0.07) 0.05) ± ± 0.02) (1.04 (3.74 4 0.02) (1.2 ± 0.2) ± 5 (3.85 pot 0.2 KCl 25 (3.75 0.02) ± 0.2) [2000BUG/KIS] ± (1.3 60 (3.68 ± 0.01) (0.99 ± 0.05) [2000CIA/IUL] pot 1.0 NaClO 4 25 (3.57 ± 0.02) ± 0.03) [2000FER/IUL] (1.14 ise-ox 2.0 NaClO 4 3.0 (3.81 ± (1.14 ± 0.02) 0.02) ± 0.02) (1.00 ± 0.03) [2000VAS/CAR] 25 (3.57 ise-ox 1.0 NaClO 4 [2001CIA/TOM2] 25 (3.583 ± 0.003) (1.27 0.05) ± pot 1.0 NaClO 4 sp 3.0 NaClO 25 (3.83 ± (1.09 ± 0.01) [2002HAV/SOT] 0.01) 4 ± 0.005) (1.097 ± 0.025) [2004CRE/ROB] pot 0.10 NaCl 25 (3.841 0.011) (1.028 ± 0.028) ± 0.25 (3.705 0.50 (3.605 ± 0.010) (0.977 ± 0.027) ± 0.009) (0.938 ± 0.035) (3.530 1.00 ± 0.015) (0.942 ± 0.042) (3.532 2.00 3.00 (3.620 0.015) (0.998 ± 0.045) ± ± 0.012) (1.149 ± 0.050) 4.50 (3.854 ± ± 0.009) (1.131 NI (3.929 0.020) 0.10 Et 4 0.25 (3.908 ± 0.014) (1.112 ± 0.020) ± 0.009) (1.127 ± 0.030) 0.50 (3.973 0.020) ± 0.015) (1.176 ± (4.154 1.00

171 VI.3 Protonation constants of oxalate 129 ion constants for oxala te. The protonation Table VI-3: Accepted data on the protonat constants were corrected to molal units and extrapolated to 25 C where necessary. ° Uncertainties are estimated in this review. log log log ∆ K I ∆ log K K Reference K log Electrolyte ∆ K 1 1 10 2 2 10 10 10 10 (25°C) (25°C) (molal) (molal) 0.53 – 0.011 (3.49 ± 0.06) (0.99 ± 0.10) [68DEN/MEI] LiClO 4 ± (0.82 ± 0.10) [70CIA/GRI] 1.05 – 0.021 (3.33 0.02) LiNO 0.05) – 0.003 – 0.038 (3.66 ± [83DAN/RIG2] 0.15 3 [60MCA/NAN] 0.10 – 0.003 ± 0.04) (1.37 ± 0.10) (3.81 NaClO 4 ± (1.19 ± 0.10) [65BAU/SMI] 0.02) 0.51 – 0.011 (3.66 1.05 – 0.022 (3.55 0.04) (1.05 ± ± [65BOT/CIA] 0.10) ± (1.13 ± 0.10) [65NAG/UMA] 0.10) 0.10 – 0.003 (3.85 ± 0.10) [66MOO/SUT] (1.32 0.10 – 0.003 1.05 – 0.022 (3.53 0.02) ± 0.10) ± (1.06 ± (1.19 ± 0.10) 0.02) 3.50 – 0.067 (3.73 ± 0.02) (0.99 ± 0.10) [69GRE/GAR] 0.001 1.05 – 0.022 0.007 (3.54 0.10 – 0.003 0.017 (3.91 ± 0.10) 0.007 ± 0.10) [69VOR/IVA] (1.28 ± 0.02) 0.001 (1.11 ± 0.10) [72MAG/BIS] 1.05 – 0.022 0.007 (3.59 0.10) (1.31 ± 0.10) [73ARM/DUN] ± 0.10 – 0.003 (3.81 0.15 (3.64 – 0.004 – 0.037 – 0.016 (1.07 0.10) [76MAK/TOU] 0.10) ± ± 0.02) (0.99 ± 0.10) [77HED/OLI] ± 1.05 – 0.022 (3.54 ± 0.02) (1.20 ± 0.10) [83CRU/HEY] 3.50 – 0.067 (3.75 3.50 – 0.067 (3.77 0.02) (1.19 ± 0.10) ± ± [95MIR/SAD] 0.06) 0.10 – 0.003 (3.81 ± 0.06) 0.31 – 0.007 (3.63 0.51 – 0.011 (3.55 0.06) ± ± 0.06) 1.05 – 0.022 (3.52 ± 0.06) 2.21 – 0.044 (3.58 3.50 – 0.067 (3.74 0.06) ± ± 0.06) 4.95 – 0.093 (3.92 ± 0.02) [96CHO/CHE] 0.31 – 0.007 (3.72 0.51 – 0.011 (3.64 ± (1.27 ± 0.10) 0.02) ± 0.02) (1.31 ± 0.10) 1.05 – 0.022 (3.58 0.10) ± 0.02) (1.42 ± 3.50 – 0.067 (3.77 6.58 – 0.119 (4.11 0.02) (1.30 ± 0.10) ± ± 0.02) (0.99 ± 0.20) [97BAR/CEC] 0.51 – 0.011 (3.63 ± 0.02) [98FER/MAN] 3.50 – 0.067 (3.76 2.21 – 0.044 (3.53 ± 0.04) (1.10 ± 0.10) [2000FER/IUL] (Continued on next page)

172 VI Discussion of data selection for oxalate 130 Table VI-3: (continued) Electrolyte I log ∆ K Reference ∆ log K K K log ∆ K log log 10 2 2 10 1 10 10 10 (molal) (molal) (25°C) (25°C) NaClO (3.74 ± 0.04) (1.07 ± 0.10) 3.50 – 0.067 [2000FER/IUL] 4 0.04) (0.98 ± 0.10) [2000VAS/CAR] 1.05 – 0.022 (3.55 ± 0.02) (1.25 ± 0.10) [2001CIA/TOM2] 1.05 – 0.022 (3.56 ± 0.02) (1.02 ± 0.10) [2002HAV/SOT] ± 3.50 – 0.067 (3.76 ± 0.04) [70ASC/BRI] NaCl 0.51 – 0.005 (3.59 0.61 – 0.006 (3.56 (0.96 ± 0.10) [85SJO/OEH2] ± 0.02) 0.04) (0.88 ± 0.10) [86CRU/HEY] 1.02 – 0.009 (3.51 ± ± 0.02) (1.15 ± 0.10) [89FUE/REB2] 0.51 – 0.005 (3.59 0.04 0.02) (1.17 ± 0.20) [92ROB/STE] ± (3.95 (3.76 0.20) 0.02) (1.11 ± 0.17 ± (3.65 (1.08 0.02) 0.39 ± 0.20) ± 0.74 ± 0.02) (1.06 ± 0.20) (3.58 1.23 (3.54 ± 0.02) (1.04 ± 0.20) 0.15 – 0.002 (3.76 [93GLA/MAJ] ± 0.10) ± [95MIR/SAD] 0.06) 0.10 – 0.002 (3.82 0.06) ± 0.30 – 0.003 (3.64 0.06) 0.51 – 0.005 (3.57 ± ± 0.06) 1.02 – 0.009 (3.49 ± 0.06) 2.09 – 0.018 (3.49 ± 0.06) 3.20 – 0.028 (3.59 ± 4.37 – 0.039 (3.72 0.06) ± 0.020) [98KET/WES] 0.100 (3.803 0.185 (3.706 0.020) ± 0.334 ± 0.029) (3.613 0.334 (3.618 ± 0.029) 1.001 (3.541 0.029) ± (3.553 0.029) 1.001 ± (3.556 ± 0.029) 1.001 2.002 (3.467 ± 0.027) 2.002 (3.496 ± 0.027) 3.149 (3.517 0.027) ± 5.000 ± 0.027) (3.733 0.100 (1.28 ± 0.46) 0.100 ± 0.58) (1.18 0.316 (1.07 ± 0.16) 0.316 (1.06 ± 0.16) (Continued on next page)

173 VI.3 Protonation constants of oxalate 131 Table VI-3: (continued) Electrolyte I log ∆ K Reference ∆ log K K K log ∆ K log log 2 10 1 2 1 10 10 10 10 (molal) (molal) (25°C) (25°C) (0.95 ± 0.15) [98KET/WES] NaCl 0.601 ± 0.15) 0.602 (0.97 ± 1.002 0.12) (0.94 0.17) 2.001 ± (0.96 3.002 ± 0.10) (0.98 3.172 0.13) (1.10 ± (1.03 ± 0.10) 4.002 4.999 (1.20 ± 0.10) ± 0.04) (1.06 ± 0.10) 0.3 – 0.003 (3.67 [99MIZ/BON] ± 0.04) (0.89 ± 0.39) 1 – 0.009 (3.51 0.04) (0.88 ± 0.20) ± 2 – 0.018 (3.52 0.04) (0.83 ± 0.14) 3 – 0.028 (3.52 ± ± 0.04) (1.00 ± 0.10) 4 – 0.039 (3.70 ± (1.15 ± 0.39) 5 – 0.050 (3.80 0.04) ± 0.10 – 0.002 (3.84 (1.10 ± 0.20) [2004CRE/ROB] 0.02) ± 0.02) (1.02 ± 0.20) 0.25 – 0.003 (3.70 ± (0.97 ± 0.20) 0.51 – 0.005 (3.60 0.02) 1.02 – 0.009 (3.52 0.02) (0.93 ± 0.20) ± 2.09 – 0.018 (3.51 ± 0.03) (0.92 ± 0.20) 3.20 – 0.028 (3.59 ± 0.03) (0.97 ± 0.20) 4.98 – 0.044 (3.81 ± (1.11 ± 0.20) 0.04) NaNO – 0.040 (3.82 ± 0.06) – 0.003 [81DAN/RIG] 0.10 3 – 0.005 – 0.034 (3.67 ± 0.05) 0.30 0.15 – 0.003 – 0.038 (3.78 ± 0.05) [83DAN/RIG2] NaCF (1.24 0.101 SO ± 0.29) [98KET/WES] 3 3 (1.24 0.101 0.29) ± 0.101 ± 0.30) (1.23 (1.21 0.102 0.31) ± 0.316 (1.06 ± 0.12) 0.316 (1.05 ± 0.12) 0.600 (1.04 0.10) ± (1.00 ± 0.600 0.10) ± 0.11) (1.00 1.001 (1.01 ± 0.11) 1.001 (3.821 0.100 0.022) ± 0.100 (3.826 ± 0.022) 0.100 (3.819 ± 0.022) 0.100 (3.819 0.022) ± 0.273 ± 0.022) (3.675 0.316 (3.641 ± 0.022) 0.600 ± 0.022) (3.571 0.600 (3.573 ± 0.022) (Continued on next page)

174 VI Discussion of data selection for oxalate 132 Table VI-3: (continued) I log ∆ K Electrolyte ∆ log Reference K K K log ∆ K log log 10 1 10 10 2 2 10 1 10 (molal) (molal) (25°C) (25°C) NaCF 1.000 SO ± 0.022) [98KET/WES] (3.541 3 3 ± 0.04) KCl 0.04 ± 0.20) [92ROB/STE] (3.96 (1.18 (3.80 0.04) (1.12 ± 0.20) 0.17 ± ± (1.10 ± 0.20) (3.71 0.41 0.04) ± 0.04) (1.10 ± 0.20) 0.77 (3.65 (3.61 ± (1.09 ± 0.20) 1.30 0.04) ± (1.08 ± 0.18) [94KIS/SOV] 0.20 – 0.004 (3.73 0.04) ± 0.04) [95LU/MOT] 0.10 – 0.002 (3.86 (1.30 ± 0.39) [2000BUG/KIS] 0.20 – 0.004 (3.75 ± 0.04) 0.10 – 0.003 (3.81 0.04) [66LHE/MAR] ± KNO 3 0.04) (1.08 ± 0.10) [67RAJ/MAR] 1.05 – 0.019 (3.60 ± 0.10 – 0.003 (3.82 (1.26 ± 0.10) [69CON/MAR] ± 0.04) 0.04) [77BRO/PET] 0.10 – 0.003 (3.88 ± 0.04) (1.25 ± 0.10) [80GRA/MUS] 0.10 – 0.003 (3.85 ± 0.10 (3.85 ± 0.05) [81DAN/RIG] – 0.003 – 0.040 0.30 (3.69 ± 0.05) – 0.034 – 0.006 – 0.003 – 0.038 (3.80 ± 0.05) [83DAN/RIG2] 0.15 ± 0.04) (0.99 ± 0.10) [84JOH/JON] 0.15 – 0.003 (3.75 ± 0.08) [92AZA/ELN] 0.10 – 0.003 (3.87 0.05) NI – 0.009 – 0.042 (3.93 ± 0.10 [81DAN/RIG] Et 4 0.32 – 0.027 (3.91 ± 0.05) – 0.040 – 0.009 (3.94 ± 0.05) 0.10 [83DAN/RIG] – 0.042 – 0.027 – 0.040 (3.91 ± 0.05) 0.32 0.63 – 0.050 – 0.040 (3.99 ± 0.05) 1.17 – 0.042 (4.05 ± 0.05) – 0.086 ± [85ROB/STE] 0.04 – 0.004 (4.00 0.04) ± 0.04) 0.26 – 0.022 (3.91 ± 0.96 – 0.072 (4.02 0.04) ± 0.04 – 0.004 (4.04 [90ROB/STE] 0.05) ± 0.05) 0.17 – 0.014 (3.93 0.39 – 0.032 (3.94 ± 0.05) 0.73 – 0.057 (4.00 ± 0.05) ± 0.05) 1.23 – 0.090 (4.08 ± 0.02) (4.00 (1.13 ± 0.16) [92ROB/STE] 0.04 (3.90 ± 0.02) 0.17 ± 0.16) (1.05 0.39 ± 0.02) (1.16 ± 0.16) (3.91 (3.96 ± 0.020) (1.22 ± 0.16) 0.73 1.23 (4.04 ± 0.02) (1.29 ± 0.16) 0.10 – 0.009 (3.92 (1.12 ± 0.20) [2004CRE/ROB] ± 0.02) ± 0.03) (1.09 ± 0.20) 0.26 – 0.022 (3.89 0.55 – 0.044 (3.93 ± 0.02) (1.08 ± 0.20) 1.23 – 0.090 (4.06 0.03) (1.09 ± 0.20) ± Me [96XUE/TRA] ± 0.10) (1.14 ± 0.10) NI 0.10 – 0.009 (3.86 4

175 VI.3 Protonation constants of oxalate 133 Table VI-4: Accepted data on the protonati on constants for oxalate. The experimental [39HAR/FAL] emf data reported in and [48PIN/BAT] have been later re-evaluated et al. together with new measurements by Kettler [98KET/WES] . The concentrations and K values at 25 ° C included here have kindly been provided by R.M. Kettler log 10 (private communication) for this review. Uncertainties are estimated in this review. + – + 2– – log K ] [Hox [Na ] [ox ] [Cl ] ] [K Reference I m 1 10 ± [39HAR/FAL] 0.0066 0.0086 0.0045 (3.967 0.020) 0.029 0.0242 ± 0.053 0.0447 0.0121 0.0159 0.0083 (3.881 0.020) 0.0153 0.0200 0.0105 (3.845 ± 0.020) 0.067 0.0564 0.0176 0.0230 0.0121 (3.822 ± 0.020) 0.077 0.0647 0.0212 0.0278 0.0146 (3.789 0.020) ± 0.093 0.0782 0.120 0.1012 0.0275 0.0359 0.0189 (3.741 ± 0.020) 0.0323 0.0423 0.0222 (3.706 ± 0.020) 0.141 0.1190 0.0378 0.0495 0.0260 (3.674 ± 0.020) 0.165 0.1393 ± 0.020) [48PIN/BAT] 0.052 0.0312 0.0104 0.0104 0.0104 0.0104 (3.889 0.058 0.0393 0.0031 0.0079 0.0031 0.0157 (3.880 ± 0.020) ± 0.020) 0.064 0.0431 0.0034 0.0086 0.0034 0.0172 (3.864 0.020) 0.069 0.0440 0.0042 0.0021 0.0042 0.0209 (3.852 ± ± 0.020) 0.073 0.0492 0.0039 0.0098 0.0039 0.0197 (3.840 0.077 0.0520 0.0042 0.0104 0.0042 0.0208 (3.833 ± 0.020) ± 0.020) 0.084 0.0570 0.0046 0.0114 0.0046 0.0228 (3.813 ± 0.085 0.0543 0.0052 0.0026 0.0052 0.0258 (3.812 0.020) 0.100 0.0635 0.0060 0.0030 0.0060 0.0302 (3.786 0.020) ± ± 0.103 0.0620 0.0207 0.0207 0.0207 0.0207 (3.770 0.020) ± 0.020) 0.105 0.0706 0.0057 0.0141 0.0057 0.0283 (3.780 0.107 0.0723 0.0058 0.0145 0.0058 0.0289 (3.777 0.020) ± 0.112 0.0755 0.0060 0.0151 0.0060 0.0302 (3.771 ± 0.020) 0.114 0.0724 0.0069 0.0034 0.0069 0.0345 (3.769 ± 0.020) 0.119 0.0755 0.0072 0.0036 0.0072 0.0360 (3.756 ± 0.020) ± 0.128 0.0867 0.0069 0.0173 0.0069 0.0347 (3.741 0.020) ± 0.020) 0.140 0.0949 0.0076 0.0190 0.0076 0.0380 (3.727 0.145 0.0921 0.0088 0.0044 0.0088 0.0439 (3.723 0.020) ± 0.145 0.0982 0.0079 0.0196 0.0079 0.0393 (3.717 ± 0.020) 0.146 0.0986 0.0079 0.0197 0.0079 0.0394 (3.715 ± 0.020) 0.147 0.0884 0.0295 0.0295 0.0295 0.0295 (3.708 0.020) ± 0.155 0.1049 0.0084 0.0210 0.0084 0.0420 (3.706 ± 0.020) 0.160 0.1082 0.0087 0.0216 0.0087 0.0433 (3.697 ± 0.020) 0.167 0.1061 0.0101 0.0051 0.0101 0.0505 (3.694 ± 0.020) 0.177 0.1193 0.0095 0.0239 0.0095 0.0477 (3.675 ± 0.020) (Continued on next page)

176 VI Discussion of data selection for oxalate 134 Table VI-4: (continued) + + – – 2– K ] [Hox I ] [ox ] [Cl ] ] [K Reference log [Na m 1 10 0.183 0.1238 0.0099 0.0248 0.0099 0.0495 (3.670 0.020) [48PIN/BAT] ± 0.020) 0.191 0.1290 0.0103 0.0258 0.0103 0.0516 (3.661 ± ± 0.020) 0.206 0.1391 0.0111 0.0278 0.0111 0.0557 (3.644 ± 0.207 0.1315 0.0125 0.0063 0.0125 0.0626 (3.648 0.020) 0.209 0.1256 0.0419 0.0419 0.0419 0.0419 (3.638 0.020) ± ± 0.020) 0.212 0.1434 0.0115 0.0287 0.0115 0.0574 (3.641 0.020) 0.226 0.1463 0.0133 0.0133 0.0133 0.0665 (3.627 ± 0.228 0.1540 0.0123 0.0308 0.0123 0.0616 (3.623 ± 0.020) 0.262 0.1770 0.0142 0.0354 0.0142 0.0708 (3.594 ± 0.020) ± 0.274 0.1646 0.0549 0.0549 0.0549 0.0549 (3.572 0.020) 0.276 0.1866 0.0149 0.0373 0.0149 0.0747 (3.583 0.020) ± ± 0.020) 0.295 0.1771 0.0590 0.0590 0.0590 0.0590 (3.556 ± 0.020) 0.317 0.2144 0.0172 0.0429 0.0172 0.0858 (3.550 0.321 0.2166 0.0173 0.0433 0.0173 0.0867 (3.546 ± 0.020) 0.324 0.2189 0.0175 0.0438 0.0175 0.0876 (3.543 ± 0.020) 0.327 0.2078 0.0198 0.0099 0.0198 0.0990 (3.547 ± 0.020) ± 0.020) 0.362 0.2444 0.0196 0.0489 0.0196 0.0978 (3.518 0.020) 0.367 0.2202 0.0734 0.0734 0.0734 0.0734 (3.505 ± 0.374 0.2524 0.0202 0.0505 0.0202 0.1010 (3.510 ± 0.020) 0.375 0.2534 0.0203 0.0507 0.0203 0.1014 (3.508 ± 0.020) ± 0.406 0.2740 0.0219 0.0548 0.0219 0.1096 (3.490 0.020) 0.442 0.2988 0.0239 0.0598 0.0239 0.1195 (3.471 0.020) ± ± 0.020) 0.450 0.2908 0.0264 0.0264 0.0264 0.1322 (3.470 ± 0.020) 0.468 0.3159 0.0253 0.0632 0.0253 0.1264 (3.455 0.478 0.3233 0.0259 0.0647 0.0259 0.1293 (3.448 ± 0.020) 0.480 0.3053 0.0291 0.0145 0.0291 0.1454 (3.456 ± 0.020) 0.487 0.2919 0.0973 0.0973 0.0973 0.0973 (3.432 ± 0.020) ± 0.538 0.3636 0.0291 0.0727 0.0291 0.1455 (3.418 0.020) ± 0.020) 0.557 0.3763 0.0301 0.0753 0.0301 0.1505 (3.406 0.561 0.3789 0.0303 0.0758 0.0303 0.1516 (3.408 0.020) ± 0.641 0.4076 0.0388 0.0194 0.0388 0.1941 (3.379 ± 0.020) 0.689 0.4459 0.0405 0.0405 0.0405 0.2027 (3.357 ± 0.020) 0.724 0.4893 0.0391 0.0979 0.0391 0.1957 (3.333 0.020) ± 0.734 0.4958 0.0397 0.0992 0.0397 0.1983 (3.334 ± 0.020) 0.020) ± 0.747 0.4831 0.0439 0.0439 0.0439 0.2196 (3.335 0.747 0.5046 0.0404 0.1009 0.0404 0.2019 (3.329 ± 0.020)

177 VI.3 Protonation constants of oxalate 135 Analysis of K VI.3.2 1 The 131 data in Table VI-3 for the first protonation of oxalate: 2 + − − ox Hox U + H , (VI.5) were treated with the SIT methodology described in Appendix B and in Section V.3.1, using a weighted multi-linear least-squares re gression procedure. This procedure as- ο log K sumes that a common value of , should fit all the data measured in LiClO 4 10 1 LiNO , KCl, KNO , NaCl, NaNO , NaClO , NaCF NI media. SO NI and Me , Et 3 4 4 3 4 3 3 3 − + – − + + + – – – NO log )· ,I − m − ε (H ε ,Cl (H )·m D K − ε (H , , + 4 ClO (H )· )·m m ε − − − 10 1 4 3 Cl I NO ClO 3 4 ο + * * * + * + + + + + + = – log K ∆ ε (Na ∆ε )·m N (Li − − (R ∆ ε (K )·m )·m ε )·m − ∆ 4 10 1 1 1 1 1 K N Li Na R 4 − + + m (H CF SO ε )· (VI.6) , − 33 CF SO 33 + − In Eq.(VI.6) the SIT interaction parameters ε , 0.02), ClO (H ) = (0.14 ± 4 − – – + + + NO ) = 0.01), , (H ε ± ) = (0.07 ± 0.01) ( cf. Appendix B) and ε (H ) = (0.12 ,I ,Cl (H ε 3 + * (0.19 ) = Section V.3.5) have been used, and 0.01) ± (M ( cf. ∆ε 1 +2 −−+ + (ox , M ) ε − are fit parameters, where M (Hox ,M ) is the cation of the background ε + + + + electrolyte, N or R ., Li i.e , K (tetraethylammonium and tetramethylammonium , Na 4 data have been fitted by a common tetraalkylammonium parameter). In the case of + − NaCF media, ε (H SO , ) is an additional fit parameter. CF SO 3 3 33 When applying the SIT model to the ac tivity coefficients of tetraalkyl ammo- nium halides, it may be shown that the specific ion-interaction coefficient, + − ε (R N ,X ) Section V.3.2). A proper representation , depends on the ionic strength ( cf. 4 of the data is achieved by setting: − ++ + + −− (R N ,X ) = . +⋅ (R N ,X ) ε ε (RN,X)log [RN] ε 104 24 14 4 K data (Figure A pronounced ionic strength dependence is found in log 10 1 VI-2f) but not in log data (Figure VI-3e). Because of this, the protonation constant K 10 2 of oxalate in tetraalkylammonium salts was fitted to the SIT equations by setting: + + * + – + 2– 2– + (R N N (R ,Hox ) = ( ) − ε [R (R ) · log N N ,ox ] ) ) − ε ∆ε (R ,ox N ε 4 4 4 10 1 4 2 4 1 The regression plots are shown in Figure VI-2, and the results, with the values selected in this review, ar e reported in Table VI-5.

178 VI Discussion of data selection for oxalate 136 * ο ∆ K Table VI-5: Selected values of (VI.5) and ε at 25°C. log 1 1 10 ο K (VI.5) = (4.25 ± 0.01) log 1 10 * a − 1 ∆ (kg ⋅ mol ε ) Medium 1 + ± 0.03) Li (0.23 + ± 0.01) Na (0.01 + K − (0.08 ± 0.02) + + ± (0.39 ± 0.01) + (0.58 0.06) log [R N − ] N R 4 10 4 − + (H ε , CF SO ) = (0.11 ± 0.02) 33 + a: R represents tetraalkylammonium N 4 i.e ., data measured in NaCl – Na ox – NaHox The 70 data in mixed media, 2 [39HAR/FAL] media and in NaCl – Na ox – KHox media [48PIN/BAT] presented in 2 Table VI-4 have been re-evaluated together with new measurements by Kettler et al. [98KET/WES] in an iterative procedure. Hence, the data using their results for log K 10 2 results. in Table VI-4 might be biased by the [98KET/WES] If all the 201 data in Table VI-3 and Table VI-4 are used for a weighted multi- + 2– linear least-squares regression proce dure, additional fit parameters for , ox ) and ε (H + – – 2– ) are obtained, because Hox , Hox and ox ε represent a non-negligible fraction of (H other fit parameters differ from the values species contributing to the ionic medium. All ο log K in Table VI-5 by less than their assigned uncertainties, the difference i.e ., for 10 1 * + + + is 0.003, and in the case of it is 0.005, 0.002, 0.006 and 0.004 for Li and , Na ε , K ∆ 1 + R N ured in “self-media” presented in Table , respectively. Hence, the data sets meas 4 VI-4 are compatible with the other data in Table VI-3. – 2– However, the free concentrations [Hox ] and [ox ] shown in Table VI-4 can only be calculated together with the values of the protonation constants in an iterative procedure. Due to this proced ure, there is a risk for signi ficant co-variance between the – ionic medium and the calculated [Hox protonation constants calculated at each ] and 2– [ox source of systematic errors in the fitted ] values, and this co-variance might be a + 2– + – (H ) and ε (H , ox , Hox ε ) values. Hence, these fit parameters are not credited in this review.

179 VI.3 Protonation constants of oxalate 137 T regression plots for the reaction: Figure VI-2: Multi-linear least-squares SI 2 − − + ox U Hox + H . a) LiClO : 4 [68DEN/MEI] 4.4 m [70CIA/GRI] I 2 − + − ) + H U ox Hox − : LiNO 3 + ,X [83DAN/RIG2] + in Li electrolytes (H 4.2 ε − D + 4 1 4.0 K 10 log 3.8 0.4 1.4 0.8 0.6 0.2 0.0 1.2 1.0 + [Li ] / molal b) [2000VAS/CAR] NaClO : 4 [2000FER/IUL] [60MCA/NAN] − − + 2 U ox + H Hox m 4.4 [2001CIA/TOM] [65BAU/SMI] I + ) [2002HAV/SOT] − [65BOT/CIA] in Na electrolytes NaCl: ,X [65NAG/UMA] + [70ASC/BRI] [66MOO/SUT] (H [85SJO/OEH2] [69GRE/GAR] ε [86CRU/HEY] 4.2 − [69VOR/IVA] [89FUE/REB2] D [72MAG/BIS] [92ROB/STE] [73ARM/DUN] + 4 1 [93GLA/MAJ] [76MAK/TOU] K [95MIR/SAD] [77HED/OLI] 10 [98KET/WES] 4.0 [83CRU/HEY] log [99MIZ/BON] [95MIR/SAD] [2004CRE/ROB] [96CHO/CHE] : NaNO [97BAR/CEC] 3 1.2 0.4 0.8 0.6 1.4 0.2 0.0 1.0 [98FER/MAN] [81DAN/RIG] + [83DAN/RIG2] [Na ] / molal (Continued on next page)

180 VI Discussion of data selection for oxalate 138 Figure VI-2 (continued) c) 4.6 : [2000VAS/CAR] NaClO 4 − 2 − + [2000FER/IUL] [60MCA/NAN] ox U + H Hox m [2001CIA/TOM] [65BAU/SMI] + I in Na electrolytes ) [2002HAV/SOT] − [65BOT/CIA] 4.4 NaCl: [65NAG/UMA] ,X + [70ASC/BRI] [66MOO/SUT] (H [85SJO/OEH2] [69GRE/GAR] ε [86CRU/HEY] − [69VOR/IVA] [89FUE/REB2] D [72MAG/BIS] 4.2 [92ROB/STE] [73ARM/DUN] + 4 1 [93GLA/MAJ] [76MAK/TOU] K [95MIR/SAD] [77HED/OLI] 10 [98KET/WES] [83CRU/HEY] 4.0 log [99MIZ/BON] [95MIR/SAD] [2004CRE/ROB] [96CHO/CHE] : NaNO [97BAR/CEC] 3 01234567 [98FER/MAN] [81DAN/RIG] + ] / molal [Na [83DAN/RIG2] d) 4.6 2 − + − Hox ox + H U [98KET/WES] in NaCF SO 4.4 3 3 D + 4 1 K 4.2 10 log 4.0 0.00.20.40.60.81.0 [NaCF ] / molal SO 3 3 (Continued on next page)

181 VI.3 Protonation constants of oxalate 139 Figure VI-2 (continued) e) 4.6 KCl: [92ROB/STE] − 2 − + ox U Hox + H m [94KIS/SOV] I + ) [95LU/MOT] − electrolytes in K [2000BUG/KIS] ,X + 4.4 KNO : 3 (H [66LHE/MAR] ε − [67RAJ/MAR] D [69CON/MAR] [77BRO/PET] 4.2 + 4 1 [80GRA/MUS] K [81DAN/RIG] 10 [83DAN/RIG2] log [84JOH/JON] [92AZA/ELN] 4.0 1.5 1.0 0.5 0.0 + ] / molal [K f) 4.8 NI: Et m 4 I ) [81DAN/RIG] − 4.6 [83DAN/RIG] ,X + [85ROB/STE] (H [90ROB/STE] ε + 2 − − [92ROB/STE] 4.4 − Hox U + H ox [2004CRE/ROB] D NI electrolytes in R Me NI: 4 4 + 4 1 [96XUE/TRA] K 4.2 10 log 4.0 0.5 1.0 1.5 0.0 + N [R ] / molal 4

182 VI Discussion of data selection for oxalate 140 K VI.3.3 Analysis of 2 The 91 data in Table VI-3 for the second protonation of oxalate: − + H + H U ox(aq) (VI.7) Hox 2 were treated with the SIT methodology described in Appendix B and in Section V.3.1, s regression procedure. This procedure using a weighted multi-linear least-square ο K log assumes that a common value of (VI.7) should fit all the data measured in 2 10 LiClO , NaClO , NaCl, NaCF SO , Et NI and Me NI media. , KCl, KNO 3 4 4 4 3 4 3 − − + – + + + – – – ClO NO )·m K ε )· + 2 ,I m D − ε (H ,Cl − )·m (H − ε (H (H , ε , )· − m log − − 2 10 3 4 Cl I ClO NO 3 4 + * + + * + − * ο + + + K ∆ε − )· )·m ∆ ε (Li CF SO )·m ε = – (H m ∆ ε (Na , )·m (K log − – − 33 10 2 2 2 2 Na Li K CF SO 33 * + + ∆ε (R − N (VI.8) )·m 4 2 R N 4 + − In Eq.(VI.8) the SIT interaction parameters ε , ) = (0.14 ± 0.02), (H ClO 4 − + + – – + NO Appendix B), ± , ε 0.01), ) = (0.12 ) = (0.07 ± 0.01) ( cf. (H ε (H ,Cl ,I (H ) = ε 3 + − (0.19 Section V.3.5) and ε (H CF SO , ± 0.01) ( cf. ) = (0.11 ± 0.02) (selected in Section 33 + * – + VI.3.2) have been used, and ) = ε (H ) are fit parameters, ox, MX) – ε ∆ε (M , M (Hox 2 2 + + + + + where M , Na i.e , K ., Li or R (tetra- N is the cation of the background electrolyte, 4 ethylammonium and tetramethylammonium data have been fitted by a common tetraal- kylammonium parameter). The regression plots are shown in Figure VI-3, and the re- sults, with the values selected in this review, are reported in Table VI-6. ο * Table VI-6: Selected values of K (VI.7) and at 25°C. log ε ∆ 2 10 2 ο log (VI.7) = (1.40 K 0.03) ± 2 10 * a − 1 ε ∆ (kg ⋅ mol Medium ) 2 + (0.28 ± 0.09) Li + Na (0.07 ± 0.01) + K (0.01 ± 0.08) + 0.09) N (0.01 ± R 4 + a: R N represents tetraalkylammonium. 4

183 VI.3 Protonation constants of oxalate 141 SIT regression plots for the reaction: Figure VI-3: Multi-linear least-squares − + U H + H ox(aq). Hox 2 a) 2.0 LiClO : 4 − + m I 1.8 [68DEN/MEI] U H + H ox(aq) Hox ) 2 − [70CIA/GRI] + in Li ,X electrolytes 1.6 + (H ε 1.4 − D 1.2 + 2 2 K 1.0 10 log 0.8 0.6 1.4 0.4 0.2 0.6 0.8 1.0 0.0 1.2 + [Li ] / molal b) 2.0 NaClO : [97BAR/CEC] 4 [2000VAS/CAR] [60MCA/NAN] 1.8 m [2000FER/IUL] [65BAU/SMI] I ) − [2001CIA/TOM] [65BOT/CIA] 1.6 ,X [2002HAV/SOT] [65NAG/UMA] + NaCl: (H [66MOO/SUT] ε 1.4 [85SJO/OEH2] [69GRE/GAR] − [86CRU/HEY] [69VOR/IVA] D [89FUE/REB2] [72MAG/BIS] 1.2 + 2 2 [92ROB/STE] [73ARM/DUN] K [98KET/WES] [76MAK/TOU] 1.0 10 [99MIZ/BON] [77HED/OLI] + − log U H + H ox(aq) Hox [2004CRE/ROB] [83CRU/HEY] 2 0.8 + [96CHO/CHE] : SO NaCF in Na electrolytes 3 3 [98KET/WES] 0.6 1.0 0.2 0.0 0.6 0.8 0.4 1.4 1.2 + ] / molal [Na (Continued on next page)

184 VI Discussion of data selection for oxalate 142 Figure VI-3(continued) c) 2.0 NaClO [97BAR/CEC] : 4 + − [2000VAS/CAR] [60MCA/NAN] H U ox(aq) + H Hox m I 2 1.8 [2000FER/IUL] [65BAU/SMI] ) + − electrolytes in Na [2001CIA/TOM] [65BOT/CIA] ,X 1.6 + [2002HAV/SOT] [65NAG/UMA] (H NaCl: [66MOO/SUT] ε 1.4 [85SJO/OEH2] [69GRE/GAR] − [86CRU/HEY] [69VOR/IVA] D [89FUE/REB2] [72MAG/BIS] 1.2 [92ROB/STE] + 2 [73ARM/DUN] 2 [98KET/WES] K [76MAK/TOU] 1.0 10 [99MIZ/BON] [77HED/OLI] [2004CRE/ROB] log [83CRU/HEY] 0.8 [96CHO/CHE] : SO NaCF 3 3 [98KET/WES] 0.6 01234567 + ] / molal [Na d) 2.0 KCl: [92ROB/STE] m I 1.8 [94KIS/SOV] ) − [2000BUG/KIS] ,X 1.6 + KNO : 3 (H [67RAJ/MAR] ε 1.4 [69CON/MAR] − [80GRA/MUS] D [84JOH/JON] 1.2 + 2 2 K + − 1.0 Hox U + H H ox(aq) 10 2 + log 0.8 in K electrolytes 0.6 1.0 0.0 0.5 1.5 + ] / molal [K e) 2.0 m I 1.8 Et NI: 4 ) − [92ROB/STE] ,X 1.6 + [2004CRE/ROB] Me NI: (H 4 ε 1.4 [96XUE/TRA] − D 1.2 + 2 2 K 1.0 + − 10 U Hox + H ox(aq) H 2 log 0.8 NI electrolytes in R 4 0.6 1.5 0.0 0.5 1.0 + [R ] / molal N 4

185 VI.3 Protonation constants of oxalate 143 VI.3.4 Temperature effects Temperature effects on the equilibrium constants for the protonation of oxalate have been determined both calorimetrically and by the determination of the protonation con- stants at different temperatures. As expected for the dissociation of carboxylic groups, the individual protonation steps: the reported enthalpy changes for r 2) ( + ( r − 3) − H U H + H ox ox (VI.9) r − r 1) ( − 1 values range between − 7 and + 7 kJ ⋅ , depending on the ionic H ∆ mol are small: rm media. The corresponding heat capacity changes, C ∆ , are reported to be in the r,m p − 1 − 1 range 15 to 380 J K ⋅ ⋅ . For temperatures between 0 and 50°C these values corre- mol spond to quite small changes of the equilibrium constants. A number of calorimetric investigations on the protonation of oxalate have been reported [67CHR/IZA] , [71VAS/KOC2] , , [73VAS/SHE2] , [73BAR/RED] [75BAR/DUB] [76VAS/SHE] , [87LIN/GU] , [91BEI/GRA] , [98ALD/BIA] . All these , papers have been scrutinised in detail in order to extract reliable information and the uncertainties of the reported values have been estimated in this review ( Appendix A). cf. [70DAV/WAT] , [81CER/CAS] , [67MAK2] A few papers had to be rejected for various reasons. For details see Appendix A. Large uncertainties are expected in the determination of enthalpy changes from the T -variation of log K . For the studies considered in this review where this method- 10 ology has been applied [39HAR/FAL] , [48PIN/BAT] , [61MCA/NAN] , [69KUR/FAR] , [76KAL] , , [92ROB/STE] , [98KET/WES] the uncertainty was estimated [90ROB/STE] from the rules of error propagation and the relationship: log ( ) KT ∂ ⎛⎞ 2 10 HT ∆= . ln(10) R rm ⎜⎟ T ∂ ⎝⎠ P ± 0.02 in the values, log K determined uncertainty of For example, for an 10 ∆ H over the temperature range 5 to 50°C, the total uncertainty in was estimated to be rm − 1 ± 1.1 kJ mol . This method was used to increase, when necessary, the uncertainties in ⋅ Appendix H ∆ obtained from regression analyses of the T -variation of cf. log K ( 10 rm A). For this procedure the total uncertainty in the individual measurements of K log 10 was set to at least 0.02 log were -units. For studies where both H ± ∆ and ∆ C 10 rm p r,m obtained from the T -variation of gned with the method log the uncertainty assi K r 10 outlined above was decreased by one third. In addition, the uncertainty of ob- C ∆ r,m p sed, when necessary, by using the relation- tained from regression analyses was increa ship: () HT ∂∆ ⎛⎞ r C ∆= . r,m p ⎜⎟ T ∂ ⎝⎠ P

186 VI Discussion of data selection for oxalate 144 H ∆ The selected data and the corresponding uncertainties assigned by this rm review are listed in Table VI-7. These data were treated according to the SIT model, described in Appendix B and in Section V. 3.6, and a multi-linear least-squares regres- sion was performed ( cf . Figure VI-4 and Figure VI-5). H ∆ = 1) an initial 8 dimensional regression analysis r ((VI.9), In the case of rm revealed that the results ob tained for chloride, nitrate and trifluoromethanesulfonate media cannot be distinguished from a statistical point of view. Hence, all these data have been fitted subsequently with a common pa- ∆ε parameter. In addition, ∆ε L, 1 L, 1 rameters for NaClO and Et NI media have been fitted in the final 3 dimensional regres- 4 4 sion analysis. Note that the KCl data of [92ROB/STE] -variation of , obtained from the T log data, are in marked disagreement with the calorimetric results of K 10 [75BAR/DUB] . The latter KCl data perfectly agree with the overall behaviour of NaCl and KNO data (Figure VI-4), and there is no obvious reason why KCl data should 3 [92ROB/STE] behave different than all other chloride and nitrate data. The KCl data of need confirmation; they have been exclud ed from the final multi-linear least-squares large discrepancy between Et regression. There is also a I > 0.5 NI values derived at 4 from the log T K -variation of [90ROB/STE] and [92ROB/STE] . This discrep- data in 10 H ∆ ancy indicates the order of magnitude of the uncertainty of deriving ((VI.9), r = rm 1) from a few log K data over a narrow temperature range. 10 = 2) an initial multi-dimensional regression r H ∆ ((VI.9), In the case of rm analysis revealed that the re sults obtained for all media cannot be distinguished from a statistical point of view. Hence, all these data have been fitted subsequently with a common parameter (Figure VI-5). The standard enthalpy changes were found to be: ο − 1 ∆ ((VI.9), r = 1) = (7.3 ± ⋅ mol H 0.1) kJ rm ο − 1 r = 2) = (3.3 ± H ⋅ mol ∆ ((VI.9), . 0.5) kJ rm ∆ε . The From the slopes of the regressions it is possible to obtain values for L,r multi-linear least-squares regression of the protonation enthalpies H ∆ ((VI.9), r = 1) rm for oxalate in (Li,Na,K)NO SO media gave: ∆ε , (Na,K)Cl and NaCF = 1 L, 3 3 3 − 3 − 3 (2.26 = (3.3 ± , in NaClO = media: ∆ε ∆ε NI media: 0.08) ± 0.2) · 10 · 10 and in Et 4 1 1 4 L, L, − 3 – (1.5 0.6) ± 10 . The least-squares regression of the protonation enthalpies · − 3 ± ((VI.9), = 2) for oxalate gave ∆ε ∆ = (2.3 r 0.3) · 10 H in all Li, Na, K media (all 2 L, rm − 1 − 1 in units of kg ⋅ K mol ⋅ ).

187 VI.3 Protonation constants of oxalate 145 Table VI-7: Literature data on the enthal py and heat capacity changes for oxalate protonation, with the uncertainties assigned in this review. Data in italics were reported in molal units. Reference Medium t ∆ C C H ∆ I H ∆ Method ∆ r,m r,m p p rm rm ( r = 1) = 2) = 2) ( r ( r = 1) r ( –1 –1 –1 –1 –1 –1 J·K ·mol ·mol (°C) J·K kJ·mol kJ·mol (M) pK ± / ∂ T I → 0 NaCl 0 - 50 (6.7 ± 0.7) (241 ∂ 13) [39HAR/FAL] a (7.0 / ∂ T I → 0 NaCl 0 - 50 ∂ ± 0.7) (231 ± 13) [48PIN/BAT] pK a [61MCA/NAN] / ∂ T I → 0 HCl 0 - 45 (2.0 ± 1.1) pK ∂ a 25 (6.3 ± 0.8) (4.3 ± 0.8) [67CHR/IZA] I → 0 (H,Na)ClO cal 4 / ∂ T I → 0 HCl (4.0 ± 1.8) [69KUR/FAR] 25 - 55 pK ∂ a I → 0 25 (3.1 ± 0.8) [71VAS/KOC2] cal cal 2 KCl 25 (1.2 0.4) – (2.9 ± 0.8) [73BAR/RED] ± , [75BAR/DUB] 25 (3.9 0.4) [73VAS/SHE2] ± cal 0.5 NaClO 4 0.4) 1.0 (2.2 ± 0.4) 2.0 – (1.1 ± ± 0.5 NaCl (4.3 0.4) ± 0.4) 1.0 (2.9 2.0 (0.8 0.4) ± (4.7 ± 0.4) 0.5 NaNO 3 ± 0.4) 1.0 (3.5 2.0 (1.4 0.4) ± ± 0.4) 0.5 (4.6 1.0 (3.6 ± 0.4) 2.0 (1.6 ± 0.4) (4.9 ± 0.4) 0.5 KNO 3 ± 0.4) 1.0 (2.9 (4.8 ± 0.4) 0.5 LiNO 3 1.0 (3.4 ± 0.4) 2.0 (0.3 ± 0.4) ∂ T 1 NaClO [76KAL] / 40 – 55 (3.5 ± 2.9) ∂ pK 4 a 35 - 55 (2.3 ± 5.0) 1 KNO 3 (Continued on next page)

188 VI Discussion of data selection for oxalate 146 Table VI-7: (continued) Reference Medium t Method I ∆ H H ∆ C ∆ ∆ C r,m rm rm r,m p p ( = 2) r = 1) ( r ( r = 2) = 1) r ( –1 –1 –1 –1 –1 –1 ·mol ·mol (°C) J·K kJ·mol kJ·mol (M) J·K ± ± 0.3) (180 5 - 35 (4.4 17) [76VAS/SHE] cal 0.5 NaNO 3 0.3) ± 15) ± 1.0 (3.3 (183 0.3) (220 ± 14) 2.0 (1.1 ± (148 (2.8 ± 0.6) 8 - 30 ± 18) 0.15 NaNO 3 [87LIN/GU] 25 (2.7 ± 0.8) cal 1 KNO 3 (272 / T 0.04 Et NI (6.9 ± 0.7) ∂ ± 14) [90ROB/STE] ∂ pK a 4 0.7) ± 14) ± 0.16 (6.4 (269 0.7) (258 ± 14) 0.36 (6.1 ± ± 0.7) ± 14) 0.64 (5.9 (254 ± (255 ± 14) 1 (5.8 0.7) [91BEI/GRA] 25 ± 0.8) (4.7 cal 1 (H,Na)ClO 4 / ± T 0.04 NaCl 10 - 45 (6.0 ∂ 0.9) (220 ± 27) [92ROB/STE] pK ∂ a 0.17 ± 0.9) (243 ± 62) (4.9 (4.6 ± 0.40 (173 ± 32) 0.9) 0.75 (3.8 ± 0.9) (71 ± 27) 1.23 (4.4 ± 0.9) (15 ± 27) 0.04 KCl (5.4 ± 0.9) (76 ± 40) (103 ± 0.9) 0.41 ± 27) (0.8 0.77 ± 0.9) (79 ± 39) – (1.7 1.30 – (2.2 ± 0.9) (46 ± 141) NI (6.8 ± 0.9) (323 ± 27) 0.04 Et 4 (6.3 ± 0.9) (379 ± 48) 0.17 0.39 (290 ± 0.9) ± 86) (6.3 (7.1 ± 1.1) (286 ± 213) 0.73 1.23 (8.7 ± 1.3) (208 ± 251) ± 0.8) [98ALD/BIA] 25 (4.7 cal 0.5 NaClO 4 [98KET/WES] ± / ∂ T 0.10 NaCl 0 - 175 (5.1 0.8) (190 ± 12) pK ∂ a 0.7) 0.19 (4.9 (195 ± 13) ± 0.33 (4.8 ± 0.3) (159 ± 5) 1.00 (1.3 ± 1.0) (161 ± 120 0.8) (0.8 ± (122 ± 14) 2.00 3.15 – (0.6 ± 0.3) (95 ± 5) ± – (4.8 ± 0.4) (78 5.00 6) (Continued on next page)

189 VI.3 Protonation constants of oxalate 147 Table VI-7: (continued) Reference Medium t Method I ∆ H H ∆ C ∆ C ∆ rm r,m rm r,m p p ( = 2) r ( ( r r = 1) = 2) = 1) ( r –1 –1 –1 –1 –1 –1 ·mol ·mol (°C) J·K kJ·mol (M) kJ·mol J·K pK / ∂ T 0.10 ∂ NaCl 0 - 125 (0.0 ± 2.6) (122 ± 68) [98KET/WES] a 0.32 (0.3 ± 2.0) (128 ± 52) 0.60 (2.3 ± 1.9) (95 ± 31) 1.00 (1.4 ± 1.9) (98 ± 30) ± 2.0 – (2.5 2.1) (148 ± 35) 3.1 (0.4 ± 2.9) (68 ± 70) 5.0 – (7.0 ± 1.9) (152 ± 18) ± SO 0 - 125 – (2.2 25) 1.9) (187 ± 0.10 NaCF 3 3 0.32 – (0.3 ± 1.9) (171 ± 14) 0.60 – (0.3 ± 1.9) (144 ± 53) ± (1.0 1.90 (120 ± 27) 1.00 0.10 0 - 125 (6.1 ± 0.4) (177 ± 7) 0.27 (5.1 ± 0.4) (168 ± 3) (167 (5.1 ± 0.4) ± 3) 0.32 0.60 (4.3 ± 0.4) (158 ± 3) 0.4) (3.5 ± 1.00 (149 ± 5)

190 VI Discussion of data selection for oxalate 148 C and multi-linear least-squares SIT regression plots ° Figure VI-4: Enthalpy changes at 25 + − − 2 for the reaction: ox . + H Hox U 13 NaCl [39HAR/FAL] NaCl [48PIN/BAT] 11 NaCl [73VAS/SHE2] NaCl [92ROB/STE] 1 − 9 NaCl [98K ET /WES] NaClO4 [67CHR/IZA] NaClO4 [73VAS/SHE2] 7 NaClO4 [91BEI/GRA] ) / kJ·mol m NaClO4 [98ALB/BIA] I 5 ( L NaNO3 [73VAS/SHE2, 76VAS/SHE] D NaCF3SO3 [98KET/WES] 3 KCl [75BAR/DUB] + 4 1 KCl [92ROB/STE] H 1 r KNO3 [76VAS/SHE] ∆ KNO3 [87LIN/GU] -1 LiNO3 [76VAS/SHE] Et4NI [90ROB/STE, 92ROB/STE] -3 Series 18 012345 Series 19 Series 20 / molal I m ° C and least-squares SIT regression plot for the Figure VI-5: Enthalpy changes at 25 − + reaction: Hox + H ox(aq). H U 2 8 6 1 − HCl [61M CA/NAN] 4 HCl [67KUR/FAR] var [71VAS/KOC2] 2 NaCl [98K ET /WES] ) / kJ·mol NaClO4 [67KUR/FAR] m I 0 ( NaClO4 [76KAL] L NaNO3 [76VAS/SHE] D -2 NaCF3SO3 [98KET/WES] + 2 2 KCl [75BAR/DUB] -4 H KNO3 [76KAL] r ∆ Series 11 -6 Series 12 Series 13 -8 012345 / molal I m

191 VI.3 Protonation constants of oxalate 149 Selected protonation constants for oxalate VI.3.5 Summarising the results from the previous sections, the recommended standard values tants of oxalate are: for the protonation cons ο ± ((VI.9), r = 1, 298.15 K) = (4.25 log 0.01) K 1 10 ο − 1 H = 1) = (7.3 ± r ⋅ mol ∆ ((VI.9), 0.1) kJ rm ο log K ((VI.9), r = 2, 298.15 K) = (1.40 ± 0.03) 2 10 ο − 1 ((VI.9), r = 2) = (3.3 ± 0.5) kJ ⋅ mol H . ∆ rm These values corresponds to the following overall standard protonation constants: ο − + − 2 log β ox + H Hox 0.01) (298.15 K) = (4.25 ± U 10 1 − + 2 ο U H (298.15 K) = (5.65 + 2H ± 0.03). log β ox(aq) ox 2 2 10 in their detailed evaluation of temperature and ionic [98KET/WES] Note that strength effects, using an empirical multi-parameter fitting function and including the data of [39HAR/FAL] [48PIN/BAT] to their own extensive measurements, derived , ο constants consistent with the present review: log K (298.15 K) = (4.264 ± 0.014), 10 1 ο ο ο − 1 K ∆ log H H ∆ ⋅ mol (298.15 K) = (1.401 ± and = (7.3 ± 0.052), = 0.5) kJ r2 r1 10 2 − 1 (0.7 mol ± 0.7) kJ . ⋅ ο ο H ∆ for oxalate protonation and H ox, ∆ (H Using the selected values of 2 rm fm − 1 aq, 298.15 K) = ± 1.5) kJ·mol − (820.1 selected in Section VI.2.3 leads to: – ο − 1 (823.4 − ∆ ± (Hox 1.6) kJ·mol , 298.15 K) = H fm 2– ο − 1 (ox . , 298.15 K) = H (830.7 ± 1.6) kJ·mol − ∆ fm Protonation constants for oxalate for different ionic media may be calculated using the SIT model described in Appendix B, and selected standard protonation con- stants and specific ion interaction coefficien ts. In Table VI-8 the results from such cal- culations are reported for some ionic media commonly used in chemical equilibrium studies.

192 VI Discussion of data selection for oxalate 150 2– + + / H / Table VI-8: Calculated equilibrium constants in Molar units for some ox / Na + K cients has been used with the ∆ε ° systems at 25 C. The SIT model for activity coeffi n obtained in this review. Other parameters from Appendix B were used as appropriate. using values at the highest Care should be exercised when ionic strengths, because they are confirmed by few experimental data only, as indicated in the figures of Sections VI.3.2 and VI.3.3. I (M) log I K log K (molal) 1 2 m 10 10 NaClO 4 ± (1.40 ± 0.03) 0 0.000 (4.25 0.01) ± 0.01) (1.19 ± 0.1 0.101 (3.83 0.03) 0.25 0.254 (3.71 0.01) (1.13 ± 0.03) ± ± 0.02) ± 0.03) 0.5 0.513 (3.62 (1.10 0.02) 0.03) ± ± 0.75 0.779 (3.59 (1.08 ± (1.08 ± 0.04) 1 1.05 (3.59 0.03) ± 0.05) (1.13 2 2.21 (3.64 0.06) ± 3 3.50 (3.77 0.08) (1.21 ± 0.08) ± ± (1.32 4 4.95 (3.94 ± 0.11) 0.11) 5 6.58 (4.15 0.165 (1.44 ± 0.15) ± NaNO 3 ± 0.01) (1.40 0.03) 0 0.000 (4.25 ± 0.01) (1.18 0.03) ± 0.1 0.101 (3.82 ± 0.01) (1.11 ± 0.03) 0.25 0.253 (3.69 ± ± ± 0.03) 0.5 0.509 (3.59 0.01) (1.06 0.75 0.769 (3.54 ± 0.01) (1.03 ± 0.03) 1 1.03 (3.51 0.02) (1.00 ± 0.03) ± 2 2.14 (3.47 0.03) (0.96 ± 0.04) ± ± ± 0.06) 3 3.33 (3.50 0.05) (0.95 ± 0.07) (0.94 4 4.61 (3.55 0.07) ± 5 6.02 (3.62 ± 0.09) (0.95 ± 0.09) NaCl 0 0.000 (4.25 ± ± 0.03) 0.01) (1.40 ± ± 0.03) 0.1 0.100 (3.83 0.01) (1.20 ± 0.01) (1.14 ± 0.03) 0.25 0.252 (3.70 0.5 0.506 (3.61 0.01) (1.12 ± 0.03) ± 0.75 0.762 (3.57 ± 0.01) (1.11 ± 0.03) 1 1.02 (3.55 ± 0.02) (1.12 ± 0.03) 2 2.09 (3.57 0.03) (1.20 ± 0.04) ± 3 3.20 (3.64 ± 0.05) (1.32 ± 0.05) 4 4.37 (3.74 ± 0.06) (1.45 ± 0.07) 5 5.61 (3.86 ± 0.08) (1.59 ± 0.08) (Continued on next page)

193 VI.3 Protonation constants of oxalate 151 Table VI-8: (continued) I (M) (molal) log I K log K 10 2 m 10 1 KCl ± 0 0.000 (4.25 ± 0.03) 0.01) (1.40 0.01) (1.19 ± 0.03) 0.1 0.101 (3.83 ± 0.01) (1.14 ± 0.04) 0.25 0.252 (3.72 ± 0.02) (1.11 0.05) ± ± 0.5 0.509 (3.66 0.02) (1.11 ± 0.75 0.769 (3.64 ± 0.07) ± 0.03) (1.12 ± 0.09) 1 1.03 (3.65 ± 0.05) (1.20 ± 2 2.13 (3.77 0.17) 3 3.31 (3.96 0.07) (1.31 ± 0.27) ± ± 0.10) (1.45 0.37) 4 4.58 (4.19 ± KNO 3 0.01) (1.40 ± 0.03) 0 0.000 (4.25 ± 0.01) (1.19 ± 0.03) ± 0.1 0.101 (3.83 ± 0.01) (1.13 ± 0.25 0.253 (3.71 0.04) 0.5 0.512 (3.63 0.02) (1.09 ± 0.05) ± ± ± 0.07) 0.75 0.776 (3.61 0.02) (1.08 ± 0.03) (1.07 1 1.05 (3.60 0.09) ± 2 2.19 (3.68 ± 0.05) (1.10 ± 0.18) 3 3.44 (3.83 ± 0.08) (1.17 ± 0.28) ε (H ox, MX) will be small. In this review the It is expected that the value of 2 –1 approximation is made that ε ± 0.01) kg·mol ox, MX) = (0.00 . This approximation (H 2 is corroborated by emf measurements [41LAR/TOM] from which the authors concluded that the activity coefficient of the undissociated oxalic acid does not change in value over an ionic strength range of 0.02 to 0.33 M. Table VI-9 contains selected specific ion interaction coefficients based on this approximation and on the selected in Sections ∆ε n VI.3.2 (Table VI-5) and VI.3.3 (Table VI-6). –1 Table VI-9: Selected specific ion interaction coefficients (kg·mol ) for oxalate and its + + + protonated forms in Li , K and tetraethylammonium electrolytes. , Na + + + + K R N Na Li 4 + 2– ε , ox (M ) – (0.51 ± 0.09) – (0.08 ± 0.01) (0.07 ± 0.08) (0.38 ± 0.09) – (0.58 ± 0.06) log I m 10 + – 0.09) , Hox ε ) – (0.28 ± 0.09) – (0.07 ± 0.01) – (0.01 ± 0.08) – (0.01 ± (M ± 0.01) ox, MX) (0.00 ± 0.01) (0.00 0.01) (0.00 ± 0.01) (0.00 ± (H ε 2

194 VI Discussion of data selection for oxalate 152 VI.4 Alkali metal oxalate compounds and complexes Sodium and potassium oxalate compounds VI.4.1 The solids formed at equilibrium in the system sodium oxalate – oxalic acid – water are, besides H ox·2H cf. Section VI.2.2), NaHox·H O(cr) and Na O(cr) ( ox(cr). The crystal 2 2 2 2 structure of NaHox·H [70FOL/KAN] , whereas et al. O has been determined by Follner 2 the crystal structure of Na ox(cr) has first been reported by Jeffrey and Parry 2 [81REE/OLM] . Crystal and was later refined by Reed and Olmstead [54JEF/PAR] structure – crystal morphology relations of Na [95STR/GRI] , ox(cr) are discussed in 2 and for details on crystal growth effects of Na ox(cr) from aqueous solutions see 2 [2002LOW/OGD] . ox(cr) in pure water has been studied extensively (Table The solubility of Na 2 –1 VI-10). The following values have been reported: 0.279 mol·kg (25°C) –1 –1 [05FOO/AND] (50°C) (15°C) and 0.334 mol·kg [17COL] , , 0.238 mol·kg [16COL2] –1 –1 –1 0.289 mol·kg (50°C) , 0.241 mol·kg (25°C) and 0.355 mol·kg [19RIV/OCO] –1 –1 –1 (15°C), 0.255 mol·kg , 0.201 mol·kg (20°C) 0.269 mol·kg (0°C), (25°C) [28FLO] –1 –1 –1 0.275 mol·kg (25°C), 0.318 mol·kg (40°C), 0.341 mol·kg , [33FOO/VAN] (50°C) –1 –1 –3 0.31 mol·kg C) (39°C) and 0.34 mol·kg , 0.24 mol·dm (52 (18°C) ° [36BOU2] –1 –1 –1 [36BRI/JAR] (0°C), 0.28 mol·kg , 0.21 mol·kg (60°C) 0.42 (25°C), 0.36 mol·kg –1 –1 –1 mol·kg [46HIL/GOU] , 0.205 mol·kg (0°C), 0.261 mol·kg (20°C), 0.292 (80°C) –1 –1 –1 –1 mol·kg (50°C), 0.360 mol·kg (40°C), 0.339 mol·kg (60°C), (30°C), 0.314 mol·kg –1 –1 –1 –1 0.385 mol·kg (80°C), 0.439 mol·kg (90°C), 0.464 mol·kg (70°C), 0.413 mol·kg –1 –3 [51NOR] (99.6°C) [68MAT/KRO] , 0.252 mol·kg , 0.278 mol·dm (20°C) (25°C) –1 –1 [79ZHI/KOL] C) and 0.411 mol·kg , 0.339 mol·kg (75 ° C) [80KOL/ZHI] , 0.270 ° (50 –1 –1 –1 –1 mol·kg (25°C), 0.270 mol·kg (0°C), (25°C), 0.270 mol·kg (25°C), 0.210 mol·kg –1 –1 –1 –1 0.219 mol·kg (20°C), 0.270 mol·kg (15°C), 0.263 mol·kg (5°C), 0.243 mol·kg –1 –1 –1 (25°C), 0.284 mol·kg (30°C), 0.301 mol·kg (40°C), 0.326 (35°C), 0.316 mol·kg –1 –1 –1 –1 mol·kg (50°C), 0.355 mol·kg (55°C), 0.367 mol·kg (60°C) (45°C), 0.345 mol·kg [2004MEN/APE] . ⋅ O(cr) in pure water the following values have H For the solubility of NaHox 2 –1 –1 –1 been reported: 0.083 mol·kg (0°C), 0.208 mol·kg (30°C), 0.317 mol·kg (40°C), –1 –1 0.440 mol·kg . (101.0°C) [51NOR] (50°C) and 1.73 mol·kg ox – In addition, solubilities in a variety of systems have been reported: Na 2 H O [46HIL/GOU] , NaOH – H ox ox – H , [05FOO/AND] [51NOR] [91BOU/PHI] 2 2 2 Na ox – (NH ox ) ox – UO ox [19RIV/OCO] , Na , Na ox – Mg(ox) [36BRI/JAR] 2 2 2 2 4 2 [17COL] , Na ox – ZrO(ox) [36BOU2] , Na ox – NaNO ox – NaCl [16COL2] , Na 2 2 3 2 [16COL2] [79ZHI/KOL] [80KOL/ZHI] , ox – Na SO [16COL2] [46HIL/GOU] , Na 4 2 2 Na ox – NaIO [33FOO/VAN] . 2 3

195 VI.4 Alkali metal oxalate compounds and complexes 153 Table VI-10: Sodium and potassium oxalate compounds. References reporting rences reporting X–ray or neutron crystal solubility data are marked with (sol.). Refe structure data are marked with (str.). ° C Reference Solubility in H Compound O at 25 2 − 1 ± ox (0.28 Na [05FOO/AND] (sol.), [16COL2] (sol.) 0.01) mol·kg (sol.) [17COL] 2 [19RIV/OCO] (sol.), [28FLO] (sol.), [33FOO/VAN] [36BRI/JAR] (sol.), (sol.), [36BOU2] (sol.), (sol.), [51NOR] (sol.), [46HIL/GOU] [54JEF/PAR] (str.), (sol.), [68MAT/KRO] [80KOL/ZHI] (sol.), (sol.), [79ZHI/KOL] (str.), [95STR/GRI] , [81REE/OLM] [2002LOW/OGD] , [2004MEN/APE] (sol.) − 1 ≈ 0.16 mol·kg NaHox·H O (str.) [51NOR] (sol.), [70FOL/KAN] 2 1 − ox ⋅ H O (2.2 ± 0.2) mol·kg (sol.), [08KOP/CAH] [05FOO/AND] (sol.), K 2 2 (sol.), [16COL] (sol.), [13HAR/DRU] (sol.), [19RIV/OCO] (sol.), [16COL2] [26WOS] [35HEN] (str.), [36BOU] (sol.), (sol.), (sol.), (str.), [72KUP] , [69HOD/IBE] [42BEN] (sol.) [2004MEN/APE] [72KUP] H O (ox) 2H ⋅ K 3 2 2 4 − 1 KHox 0.5 mol·kg (sol.), (str.), [42BEN] (sol.), [1864ALL] [35HEN] (str.), [71MOO/POW] (str.), [72KUP] [68PED] 1 − [08KOP/CAH] ⋅ O 0.11 mol·kg (ox) [72KUP] 2H (sol.), [64HAA] (str.), , KH 2 2 3 (sol.), [81EMS/JON] (str.) [79BAU]

196 VI Discussion of data selection for oxalate 154 ox(cr) and NaHox O(cr) in water as a function of Figure VI-6: Solubility of Na ⋅ H 2 2 temperature. The line is give n as visual guideline only. 0.55 0.50 0.45 [05FOO/AND] [16COL2] [17COL] 1 0.40 − [19RIV/OCO] [28FLO] cr ( ox Na ) 0.35 2 [36BOU2] [33FOO/VAN] 0.30 [36BRI/JAR] [46HIL/GOU] 0.25 [51NOR] [68MAT/KRO] 0.20 Solubility / mol·kg [79ZHI/K O L] [80K O L/ZHI] [2004MEN/APE] 0.15 cr ( O NaHox·H ) 2 0.10 0.05 0 20406080100 t / °C ox(cr) in water as function of temperature, published The solubility data of Na 2 over a century, reveal a very consistent pict ure (Figure VI-6), and a mean value of all − 1 data at 25°C is (0.28 . However, it is outside the scope of this review to ± 0.01) mol·kg develop a thermodynamic model of such highly soluble salts like Na ox(cr). 2 The solids formed at equilibrium in the system potassium oxalate – oxalic acid – water are, in the order of decreas ing concentration of oxalic acid, H O(cr) ( ox·2H cf. 2 2 Section VI.2.2), KH ⋅ 2H O(cr), and finally (ox) (ox) H O(cr), KHox(cr), K 2H ⋅ 2 2 2 3 2 3 4 K ox·H O(cr) [08KOP/CAH] . Large crystals in the centimet er scale can be grown from 2 2 [72KUP] aqueous solutions of all these compounds . O(cr) was first determined by Haas (ox) ⋅ 2H The crystal structure of KH 2 2 3 [64HAA] , although not exactly in positioning the protons, and later re-determined by Emsley ox(cr) has already been inves- [81EMS/JON] et al. . The crystal structure of KH tigated in the early days of X–ray crystallography by Hendricks [35HEN] , was later determined more precisely by single crystal X–ray diffraction [68PED] , and by neutron diffraction in order to locate the hydrogen atoms unambiguously [71MOO/POW] . The crystal structure of K ox·H O(cr) has been investigated many times, from the advent of 2 2

197 VI.4 Alkali metal oxalate compounds and complexes 155 X–ray crystallography [35HEN] until the precise refinement of its crystal and molecular [69HOD/IBE] structure . ⋅ H O(cr) in pure water has been studied extensively ox The solubility of K 2 2 –1 (Table VI-10). The following values have been reported: 2.27 mol·kg (25°C) –1 –1 –1 (0°C), 1.92 mol·kg (10°C), 2.19 mol·kg , 1.54 mol·kg (20°C), 2.40 [05FOO/AND] –1 –1 –1 –1 mol·kg (30°C), 2.63 mol·kg (40°C), 2.91 mol·kg (50°C), 3.20 mol·kg (60°C), –1 –1 –1 –1 3.50 mol·kg (90.2°C), 4.76 mol·kg (70°C), 3.83 mol·kg (80°C), 4.16 mol·kg –1 –1 –1 [08KOP/CAH] (106.2°C) (0.0°C), 1.68 mol·kg , 1.53 mol·kg (5.1°C), 1.81 mol·kg –1 –1 –1 (9.8°C), 2.10 mol·kg (25.2°C), 2.38 mol·kg (30.0°C) (20.1°C), 2.27 mol·kg –1 –1 [13HAR/DRU] (50°C) [16COL] [16COL2] , (15°C) and 2.93 mol·kg , 1.91 mol·kg –1 –1 –1 2.25 mol·kg [19RIV/OCO] , 1.52 mol·kg (0°C), (25°C) and 2.98 mol·kg (50°C) –1 –1 –1 –1 1.82 mol·kg (10°C), 2.42 mol·kg (30°C), 2.72 mol·kg (40°C), 3.03 mol·kg –1 –1 –1 (70°C), 4.04 mol·kg (80°C), 4.40 (60°C), 3.66 mol·kg (50°C), 3.33 mol·kg –1 − –1 1 mol·kg (107°C) [26WOS] , 1.95 (90°C), 4.83 mol·kg (100°C), 5.10 mol·kg − 1 –1 –1 –1 mol·kg (35°C) and 3.13 mol·kg , 6.0 mol·kg (19°C), 2.39 mol·kg (52°C) [36BOU] –1 –1 –1 (130°C) [42BEN] (5°C), 1.739 mol·kg (10°C), (0°C), 1.598 mol·kg , 1.483 mol·kg –1 –1 –1 –1 1.811 mol·kg (20°C), 2.06 mol·kg (25°C), 2.17 mol·kg (15°C), 1.957 mol·kg –1 –1 –1 (30°C), 2.30 mol·kg (40°C), 2.51 mol·kg (45°C), 2.62 (35°C), 2.39 mol·kg 1 –1 –1 − mol·kg (55°C), 2.89 mol·kg . (50°C), 2.73 mol·kg (60°C) [2004MEN/APE] For the solubility of KHox(cr) in water the following values have been re- 1 –1 − –1 –1 ported: 0.17 mol·kg (0°C), 0.24 mol·kg (20°C), 0.59 mol·kg (10°C), 0.41 mol·kg –1 –1 –1 (30°C), 0.82 mol·kg (50°C), 1.60 mol·kg (60°C), 2.12 (40°C), 1.16 mol·kg − 1 –1 –1 –1 mol·kg (80°C), 3.35 mol·kg (100°C) (90.°C), 4.02 mol·kg (70°C), 2.71 mol·kg –1 –1 [1864ALL] , 5.2 mol·kg (131°C) [42BEN] . (116°C), 6.4 mol·kg O(cr), is used for preparation of one of (ox) 2H ⋅ Potassium tetraoxalate, KH 2 3 2 the “operational standard reference solutions” for pH calibration [83COV/BAT] . In this context, the solubility of KH ⋅ 2H (ox) O(cr) in pure water has been measured as a func- 2 2 3 –1 tion of temperature [79BAU] (see Appendix A). Older data reported are: 0.07 mol·kg –1 –1 (0°C), 0.23 mol·kg [08KOP/CAH] (60°C), (30°C), 0.64 mol·kg . ox – In addition, solubilities in a variety of systems have been reported: K 2 H ox , [08KOP/CAH] [13HAR/DRU] , K [19RIV/OCO] ox – (NH ox ) [05FOO/AND] 2 4 2 2 K ox – KCl ox – Ni(ox) ox – ZrO(ox) [36BOU] , K , K [16COL2] , K [36VOS/ISR] ox – 2 2 2 2 KNO [16COL2] , K ox – K SO [16COL2] . 4 3 2 2

198 VI Discussion of data selection for oxalate 156 ⋅ Figure VI-7: Solubility of K ox O(cr), KHox(cr) and KH ⋅ (ox) O(cr) in water as a H 2H 2 2 3 2 2 function of temperature. The lines are given as visual guidelines only. 6 [1864ALL] 5 1 [05FOO/AND] − [08KOP/CAH] O(cr) ox·H K 2 2 4 [13HAR/DRU] [16COL] [16COL2] [19RIV/OCO] 3 [26WOS] [36BOU] KHox(cr) [42BEN] 2 Solubility / mol·kg [79BAU] [2004MEN/AP E] 1 O(cr) ·2H (ox) KH 2 3 2 0 0 20406080100120140 t / °C ox ⋅ H O(cr) in water as function of temperature, pub- The solubility data of K 2 2 lished over a century, are fairly consistent (Fig ure VI-7), and a mean value of all data at 1 − 25°C is (2.2 . The most recent data set [2004MEN/APE] deviates with ± 0.2) mol·kg increasing temperature from the older data sets. The reason for this deviation is not sets available for KHox(cr) clear. Remarkably, the two data [1864ALL] [42BEN] and [79BAU] show perfect agreement. Also, the very precise measurements of Baucke for KH (ox) [08KOP/CAH] . However, ⋅ 2H O(cr) corroborate data published a century ago 2 2 3 it is outside the scope of this review to develop a thermodynamic model for these highly soluble salts. + + and K Complexes with Na VI.4.2 In the literature a number of publications has been found reporting stability constants for sodium and potassium oxalate complexes (Table VI-11). The stability constants were mainly evaluated from two types of experiments. The first type of experiments comprises measurements of oxalate protonation constants in various ionic media. The differences found for oxalate protonation constants in different electrolyte solutions of [81DAN/RIG] the same ionic strength were interpreted in terms of complex formation ,

199 VI.4 Alkali metal oxalate compounds and complexes 157 [92ROB/STE] , [2004CRE/ROB] . The second type of experiments [83DAN/RIG2] , deals with solubility measurements of oxalate solids. These measurements have been interpreted with thermodynamic models taking Na and K oxalate complex formation into account explicitly [73FIN/ROT] [81BUR/FIN] . [91BOU/PHI] , [79TOM/NAN] , , Table VI-11: Experimental equilibrium da ta for Na and K oxalate systems. The uncertainties are given as reported in the references. a t ( ° Method C) log K Reference Ionic medium 10 + 2– – U Na(ox) Na + ox [73FIN/ROT] → (1.12 ± 0.01) 38 , [81BUR/FIN] sol 0 ∆ pK 0.03) 0.10 M Et NI / NaNO [81DAN/RIG] 37 (0.53 ± gl, 3 4 a + + N 37 (0.46 / Na [83DAN/RIG2] 0.15 M Et ± 0.03) ∆ gl, pK 4 a → 0 25 1.10 [91BOU/PHI] sol (2 – 10 M NaOH) 40 0.44 60 – 0.3 Et [92ROB/STE] NI / NaCl → 0 25 0.88 gl, ∆ pK 4 a 37 0.92 [2004CRE/ROB] 0.02) Et ± NI / NaCl → 0 25 (0.92 pK ∆ gl, 4 a + – NaHox(aq) + Hox U Na gl, pK ∆ Et NI / NaCl → 0 25 (0.02 ± 0.02) [2004CRE/ROB] 4 a 2– – + U Kox + ox K 0.07 M K ox 18 (?) – 0.42 [31BAN/RIG] con 2 0.15 M K ox – 0.59 2 ox – 0.62 0.29 M K 2 → 0 sol 1.0 [79TOM/NAN] (0.1 – 0.3 M KCl) 37 gl, ∆ pK 0.10 M Et [81DAN/RIG] NI / KNO 0.05) 37 (0.43 ± 3 4 a + + [83DAN/RIG2] 0.15 M Et N / K 37 (0.41 ± 0.03) pK ∆ gl, 4 a [92ROB/STE] 0.71 Et 25 NI / KCl → 0 pK ∆ gl, 4 a 37 0.83 a: Methods: con = conductivity measurements; gl = pH–glass electrode; sol = the formation constants for = the for- it parameters from evaluating solubility data; pK alkali-metal complexation were obtained as f ∆ a mation constants for alkali–metal complexation were ob tained from differences in protonation constants determined in different backgr ound electrolytes (alkali electrolytes versus tetra-alkylammonium salts). The earliest attempt of deriving potassium oxalate stability constants [31BAN/RIG] et al. interpreted tabulated conduc- does not fit in either category. Banks tivity data for several salts in water in te rms of dissociation constants of these salts. Conductivity measuremen ts of unsymmetrical electrolytes like K ox are difficult to 2 interpret, and as discussed in Appendix A, the results of Banks et al. are not credited by this review.

200 VI Discussion of data selection for oxalate 158 The method of deriving Na and K oxalate complexes from protonation data, [81DAN/RIG] proposed by Sammartano and co-workers [83DAN/RIG2] , , [92ROB/STE] [2004CRE/ROB] , works as follows. The protonation constants of ox- , trically in different media, alate were determined potentiome ., in NaCl, KCl and i.e Et NI, and the data obtained for the different ionic strengths were simultaneously ana- 4 lysed in two different ways: (1) Taking into account the changes in ionic strength but neglecting any complex formation of oxalate with alkali-metal ions. The protonation constants as a function of the ionic strength are expressed by three semi-empirical equa- tions for the three ionic media. (2) Considering both, the changes in ionic strength and a possible complex formation of oxalate with alkali-metal ions. In this case it is assumed + that oxalate does not form complexes with the Et N cation and thus, the measurements 4 in Et NI are regarded as the reference case (“ baseline”) where no complexation occurs. 4 This also results in three semi-empirical equations describing stability constants as a function of ionic strength: one common oxalate protonation function for NaCl, KCl and NI, which is identical with the Et NI function derived in (1); two equations for the Et 4 4 – – stability of Kox and Na(ox) as a function of ionic strength calculated from the differ- ences KCl – Et NI, based on the functions derived in (1). NI and NaCl – Et 4 4 However, as indicated in Figure V–1 (Section V.3.2) the activity coefficients in tetraalkylammonium salts are quite different from those in alkali metal salts, although sarily implicate alkali metal ion complex formation, the differences do not neces ., e.g + between Na and chloride. Therefore, different ionic medium effects on protonation constants when comparing Na/K electrolytes with tetraalkylammonium salts are not + + necessarily an indication of complex formation between the ligand and Na . In or K this review oxalate protonation data were successfully modelled using SIT with the 2– + 2– + interaction parameters (ox ε (ox ε , K ) and ) without considering a complex , Na – – Na(ox) cf. in the speciation model ( or Kox Section VI.3). The solubility of sodium oxalate has been investigated with the aim of develop- ing a thermodynamic model for sodium oxalate solubility in “Bayer liquor” [91BOU/PHI] , [93BEC/GRO] . Processing of bauxites (aluminium ores) containing organic matter by the so-called “Bayer process” generates soluble organic sodium salts These organics accumulate in Bayer liquors and degrade of various molecular weights. to low molecular weight products, the ultim ate species being sodium oxalate and car- bonate. The Bayer process is based on the amphoteric nature of aluminium, i.e ., on the dissolution of aluminium hydroxide in strong sodium hydroxide solution. Thus, Bayer liquor is essentially a rather concentrated sodium aluminate solution, and sodium ox- alate is an impurity, which limits alumina production from most Bayer refineries. Ox- alate removal represents a significant fracti on of the operating costs of many alumina refineries. Sodium oxalate solubility in Bayer liquor depends on various factors [87THE/BUS] . The and Bush [87THE/BUS] correlated laboratory data on the apparent solubility of sodium oxalate in Bayer plant liquors with the pertinent variables by multi-

201 VI.4 Alkali metal oxalate compounds and complexes 159 [91BOU/PHI] ple regression analysis. Bouzat and Philipponneau reported an attempt to develop a thermodynamic model of sodium oxalate solubility in concentrated sodium – aluminate solutions. The model comprises the sodium complexes Na(ox) , NaOH(aq), − − NaCl(aq), N aSO aCO N . The results of this model fit are not credited by this and 4 3 review ( [93BEC/GRO] investigated the same cf. Appendix A). Beckham and Grocott chemical system and criticise the approach of [91BOU/PHI] . Beckham and Grocott [93BEC/GRO] propose a solubility model based on the solubility product of sodium oxalate and a set of purely empirical parameters. It is outside the scope of the present review to develop a thermodynamic model valid for highly concentrated sodium alumi- nate solutions like the Bayer liquor. Several investigations of calcium oxalate solubility have been published where, – in addition to the complex Ca(ox)(aq), the complex Na(ox) [73FIN/ROT] – – – , [85DAN/SON] , or both Na(ox) [89SIN] and Kox , the complex Kox [81BUR/FIN] [79TOM/NAN] have been included into the speciation models. The ionic strength de- [73FIN/ROT] pendence has been calculated using the Davies equation , [79TOM/NAN] , [81BUR/FIN] or the SIT [85DAN/SON] . As discussed in Section VI.5.1, , [89SIN] calcium oxalate solubility data can be modelled successfully applying SIT with the 2– + 2– + interaction parameters ) and ε (ox , Na , K ε ) derived from oxalate protonation (ox – – reactions but without considering a complex Na(ox) in the speciation model or Kox ( cf. Section VI.3). VI.5 Magnesium and calcium oxalate compounds and complexes Magnesium and calcium oxalate compounds VI.5.1 O(cr), is the magnesium oxalate compound Magnesium oxalate dihydrate, Mg(ox)·2H 2 forming at ambient conditions. It is found in nature as the mineral glushinskite [80WIL/JON] . Glushinskite is formed by lichens, fungi and other plants, and it may be termed a biomineral since it is found as a waste product in litter decomposition (see references in [2004FRO/ADE] ). revealed Thermal analysis, Raman and infrared spectroscopy [2004FRO/ADE] that glushinskite is the dihydrate phase in the temperature range up to 148 C. The phase ° change at 148 ° C involves dehydration and results in the formation of anhydrous magne- sium oxalate, Mg(ox)·2H O(cr) C con- Mg(ox)(cr) + 2 H ° O(g), which above 397 U 2 2 verts to magnesium oxide, MgC (g). The dehydra- O MgO(cr) + CO(g) + CO (cr) U 4 2 2 –1 ο tion enthalpy is reported as , determined by a tran- ∆ = 110.5 and 100.1 kJ·mol H m dehyd spiration method and by differential s canning calorimetry (DSC), respectively [2001TAN/FUR] . A number of studies have been published reporting solubility data for Mg(ox)·2H , O(cr) in pure water [03KOH] , [08KOH] , [24CHA/DHA] , [25WAL] 2 [34FRE] , [36BRI/JAR] , (Table VI-12) , [51BAR/ARG] and [77FED/KHO] [42KAR]

202 VI Discussion of data selection for oxalate 160 ox , [36BRI/JAR] , Na ox [36BRI/JAR] , and in solutions containing H [27BOB/MAL] 2 2 K SO , (NH ox , MgSO ox [27BOB/MAL] , H [36BRI/JAR] [51BAR/ARG] ) 4 2 2 4 2 4 [25WAL] . Cl [27BOB/MAL] , NH [77FED/KHO] Cl – NH [34FRE] , NH , and NaClO 4 4 4 3 In general, the solid phase has not been characterised in detail in these studies, but in the case of magnesium oxalate this omission poses no problem, as O(cr) is the only phase forming in the temperature range of these investiga- Mg(ox)·2H 2 [2004FRO/ADE] tions . However, a general problem affecting all these solubility stud- ies is the slow dissolution and even slower precipitation kinetics of Mg(ox)·2H O(cr). 2 This peculiar property of magnesium oxalate has been described already one hundred compound needed a lot of ti me because the solution satu- years ago, “working with this [03KOH] rates slowly” , and highly oversaturated solutions used in conductivity meas- urements approached equilibrium saturation only after two weeks [08KOH] . In a de- investigated this slow kinetics and came tailed analytical study Karaoglanov [42KAR] to the conclusion that the results of his Mg(ox)·2H O(cr) precipitation experiments were 2 not completely reproducible. Solubility studies showed a slightly decreasing concentra- tion of dissolved magnesium oxalate, approaching constant values after 40 days [42KAR] . Table VI-12: Solubility data of magnesium and calcium oxalate compounds in pure water as reported in the literature. t ( ° C) Equilibrium time Concentrati on(M) Reference Remarks Mg(ox)·2H O(cr) 2 2 – 4 hours 0.0027 , [08KOH] Conductivity measurements 18 [03KOH] [27SCH3] 0.00264 Titration of oxalate with KMnO 18 24 hours 4 36 ? 0.00301 [24CHA/DHA] Mg determined gravimetrically as MgSO 4 Salt precipitated in the cold 0.00269 Salt precipitated from boiling solu- tions 92 ? 0.00362 Salt precipitated in the cold 0.00350 Salt precipitated from boiling solu- tions 25 30 hours [25WAL] 0.00307 Titration of oxalate with KMnO 4 18 “sufficiently long” 0.00331 [34FRE] Mg determined with 8-oxy- chinolinate 0.00230 [36BRI/JAR] 18 one month 21 40 days 0.00327 [42KAR] Titration of oxalate with KMnO , Mg 4 determined gravimetrically as Mg P O 2 2 7 25 4 days 0.00320 [51BAR/ARG] Mg determined colorimetrically 25 ? 0.00924 [77FED/KHO] Mg determined complexometrically (Continnued on next page)

203 VI.5 Magnesium and calcium oxalate compounds and complexes 161 Table VI-12 (continued) t C) Equilibrium time Concentrati on(M) Reference Remarks ( ° “Ca(ox)” –5 5.3 [01RIC/MCC] Titration of oxalate with KMnO × 25 15 minutes 10 4 –5 × No characterisation of “Ca(ox)” 10 7.5 50 –4 10 × 95 1.1 -5 30 minutes × 10 18 4.3 O Conductivity measurements. H [08KOH] [03KOH] 2 -5 25 not determined, assumed to be 10 × 4.8 Ca(ox)·H O 2 -5 × 10 7.1 25 “over night” [16HEN/TAY] No characterisation of “Ca(ox)” -5 5.2 × 10 20 1 – 2 hours [27AUM2] ox and CaCl Prepared from H 2 2 -5 × 10 5.7 ) Prepared from (NH ox and CaCl 2 2 4 “According to my experience Ca(ox)·H O” 2 Ca(ox)·H O(cr) 2 –4 26–27 “until saturation” 10 × [03HER/MUH] Ca in Ca(ox)·H 2.3 O analysed. Solubil- 2 ity: Residue determined gravimetri- cally at 70ºC –5 ± 0.2) × 10 18 13 hours (4.7 Oxalate in Ca(ox)·H O analysed. [27SCH3] 2 Solubility: Titration of oxalate with KMnO 4 –5 6.7 25 “until equilibrium × 10 [33KOL/SAN] “Saturated solutions were analyzed was obtained” for oxalate or calcium by micro methods” –5 25 4 weeks 10 6.2 × Ca and oxalate in Ca(ox)·H [34BAS] O ana- 2 lysed –5 2 – 6 hours 0.06) × 10 (4.84 25 ± Oxalate in Ca(ox)·H O analysed [39PED2] 2 –5 × 10 30 10 minutes 6.46 [40SHE/PAL] Ca(ox)·H O dried at 105 ° C 2 –4 × 10 95 1.13 Titration of oxalate with KMnO 4 –5 IV 25 1 hour × 10 4.55 Titration of oxalate with Ce (SO [42MCC/RIE] ) 2 4 –5 25 × 10 1 – 2 hours 4.90 [51NYD] Ca determined by flame spectrogra- phy. -5 37 6.2 × 10 [71CHU/REE] -5 38 10 6.1 × [73FIN/ROT] -5 37 5.02 × 10 [75PAK/OHA] -5 37 10 5.4 × [79TOM/NAN] -5 (4.7 ± 0.5) × 10 37 [80HOD] -5 37 50 hours × 10 5.64 [89SIN] Ca and oxalate determined by ion chromatography (Continnued on next page)

204 VI Discussion of data selection for oxalate 162 Table VI-12 (continued) ( Concentrati on(M) Reference Remarks ° t C) Equilibrium time Ca(ox)·2H O(cr) 2 -5 [80TOM/NAN] × 10 37 8.82 O(cr) Ca(ox)·3H 2 -4 × [79TOM/NAN] 10 1.15 37 [24CHA/DHA] The data reported in are not credited in this review because of the almost complete lack of details concerning the solubility experiments. The values of [77FED/KHO] medium, are considered unreliable and are rejected , obtained in NaClO 4 by this review (see discussion in Appendix A). No other study in NaClO medium could 4 be identified by this review. Results obtained in solutions containing H ox 2 [27BOB/MAL] , H SO , [36BRI/JAR] , MgSO [25WAL] [36BRI/JAR] , and 4 2 4 NH are not further evaluated in this NH Cl review because of the lack of [34FRE] − 4 3 appropriate SIT interaction parameters for NH , and ambiguities concerning 3 + H − ox and H oxalate interactions in concentrated H SO solutions. In addition, poten- 4 2 2 tial Mg sulphate complexation in mixed sulphate – oxalate pose additional ambiguities. In all other cases, [03KOH] , , solubility of Mg(ox)·2H i.e. O(cr) in pure water 2 [08KOH] [42KAR] , , [34FRE] , [36BRI/JAR] , [25WAL] , [51BAR/ARG] and , [27SCH3] in solutions containing Na ox [36BRI/JAR] , K ox ox [51BAR/ARG] , (NH ) 4 2 2 2 [27BOB/MAL] Cl [27BOB/MAL] , and NH , the analytical data reported were [34FRE] 4 using the Mg complexation constants for used to calculate the aqueous speciation 2 − Mg(ox) Mg(ox)(aq) and selected in this review ( cf. Section VI.5.2), and using the SIT 2 2+ – 2– + interaction parameters ε ) = (0.19 ± 0.02) ( cf. Appendix B), ε (Mg (ox ,Na ,Cl ) = 2– + 2 − + − ,K 0.01), ) = (0.07 ± 0.08) and ε ( ε ± Mg(ox) (ox ,Na (0.08 ) = – (0.15 ± 0.03) se- 2 + − 2 2– 2– + + Mg(ox) NH lected in this review, and assuming ,K , (ox ) ), ε ( ≈ ≈ ε ,Na (ox ) ε 2 4 2 − + 2 − + ,K ( ) = ε ( (Mg(ox)(aq),NaCl) (Mg(ox)(aq),NH Ca(ox) Cl) = , Mg(ox) ε N H), ε ε ≈ 4 4 2 2 (Mg(ox)(aq),(NH ± ) 0.1), ox) = ε (Mg(ox)(aq),K ε ox) = ε (Mg(ox)(aq),Na ox) = (0.0 2 2 4 2 2– 2+ and ,ox ) ε ≈ 0.15. From each aqueous speciation the solubility product (Mg log K s 10 ,0 for reaction: 2+ 2– Mg(ox)·2H U Mg + ox + 2 H O(cr) O (VI.10) 2 2 has been calculated and extrapolated to zero ionic strength using the same SIT interac- tion parameters (Figure VI-8). The actual value of the SIT interaction parameters has no influence on the re- sults in pure water, and also in the other cases at I ≤ 0.6 M the results do not depend in any significant way on the estimated SIT pa rameters. The complex Mg(ox)(aq) is the dominating species ( > 60%) in pure water, whereas with increasing K ox ox and Na 2 2 2 − Mg(ox) concentrations the complex becomes the dominating species. The numerical 2

205 VI.5 Magnesium and calcium oxalate compounds and complexes 163 ο K depend on the selected complexation constants, but values of the calculated log 10 s ,0 the scatter of the values revealed in Figure VI-8 is independent of the complexation constants and SIT interaction parameters used in the calculations. The above discussed slow kinetics of magnesium oxalate dissolution and precipitation could be the major cause of the observed scatter. For scoping cal culations in modelling exercises related to ο radioactive waste disposal a value 0.2) could be used, en- (VI.10) = − (6.4 ± log K s ,0 10 compassing the available literature data as shown in Figure VI-8, but none of the above ect of slow kinetics on equilibrium solubility discussed studies properly addresses the eff of Mg(ox)·2H O(cr) and therefore no value is recommended for its solubility product. 2 O(cr) according to Reaction (VI.10), Figure VI-8: Solubility product of Mg(ox)·2H 2 calculated from analytical solubility data and extrapolated to zero ionic strength using parameters discussed in the text. Data close to zero ionic strength refer to solubility in pure water, in all other cases the added salt is indicated. -5.9 [51BAR/ARG] [36BRI/JAR] -6.0 [27BOB/MAL] [03KOH] [08KOH] -6.1 [25WAL] Cl NH 4 [27SCH3] -6.2 [34FRE] [42K AR] -6.3 NH Cl 4 ,0 s ° ox K 2 K -6.4 10 (NH ) ox 4 2 log -6.5 ox Na 2 -6.6 -6.7 -6.8 -6.9 0.3 0.4 0.6 0 0.1 0.2 0.5 / molal I m

206 VI Discussion of data selection for oxalate 164 Calcium oxalate forms three different hydrates at ambient conditions, Ca(ox)·H O(cr), and Ca(ox)·3H O(cr). The monoclinic monohydrate, O(cr), Ca(ox)·2H 2 2 2 Ca(ox)·H O(cr), is found in nature as the mineral whewellite, its crystal structure data 2 are reported in [80TAZ/DOM] . The tetragonal dihydrate, [81DEG/PIR] Ca(ox)·2H O(cr), occurs in nature as the mine ral weddellite. Single crystal X–ray dif- 2 fraction [80TAZ/DOM] revealed that the structure contains a channel in which zeolitic water occurs and thus, the formula of weddellite actually is Ca(ox)·(2+x)H O(cr) with x 2 its crystal structure data are given by 0.5. The trihydrate, Ca(ox)·3H ≤ O(cr), is triclinic; 2 [81BLO/KAN] . Whewellite and weddellite have been found in the litter layer of several differ- ent soils, indicating that oxalate is a major metabolic product of fungi in natural envi- ronments [77GRA/CRO] . Not only whewellite and weddelite but also the trihydrate is . Urolithiasis, ., the formation of kidney or bladder i.e found in urinary stones [82HEI] stones, constitutes a serious health problem, and about 70% of these stones have cal- cium oxalate as a major component [97KON/TRA] . as the low-temperature and high- O(cr) exists in two phases, known Ca(ox)·H 2 temperatures phases [96KOC/GAL] . According to thermo-analytical, structural and spectroscopic measurements this order - disorder phase transition occurs between 0 and ο 100ºC. The enthalpy of this transition is very small. It has been reported as H ∆ ≈ m trs –1 0.4 kJ·mol [96KOC/GAL] . et al. temperature calorimetric report a careful low- [33LAT/SCH] Latimer study for determining the heat capacity and entropy of Ca(ox)·H O(cr) in the range 19 2 to 300 K. The experimental data, reported in tabular form in [33LAT/SCH] , have been re-evaluated in this review ( cf. of this numerical fit agree Appendix A), and the results [33LAT/SCH] well with the results originally obtained by graphical integration . The heat capacity data show some scatter above – 50ºC which might be an indication of the onset of the above discussed order - disorder phase transition of Ca(ox)·H O(cr). How- 2 ever, besides this “roughness” no systematic effect is reveal ed in the data (see Figures A-2 and A-4 in Appendix A) and this review selects: –1 –1 ο O, cr, 298.15 K) = (153.34 ± (Ca(ox)·H 2.00) J·K ·mol C 2 ,m p ο –1 –1 (Ca(ox)·H S O, cr, 298.15 K) = (156.37 ± 2.00) J·K . ·mol 2 m ο 2– ∆ H (ox , In an early stage of this review it was intended to derive fm 298.15 K) via the enthalpy values of thermal decomposition of Ca(ox)·H O(cr) and 2 hence, an in-depth literature search and review concerning thermal decomposition of calcium oxalates was started. The result of this effort is summarised here. Early thermogravimetric studies [47PEL/DUV] [59PET/WIE] revealed that , whewellite Ca(ox)·H O(cr) dehydrates in the temperat ure range 100 to 240ºC forming 2 anhydrous calcium oxalate, Ca(ox)·H O(cr) U Ca(ox)(cr) + H O(g), which converts to 2 2 calcium carbonate above 400 ° (cr) + CO(g), which finally CaCO O U (cr) C, CaC 4 2 3

207 VI.5 Magnesium and calcium oxalate compounds and complexes 165 mperature range 660 to 900ºC, CaCO U converts to calcium oxide in the te (cr) 3 CaO(cr) + CO (g). Whereas the dehydration of calcium oxalate monohydrate and the 2 decarbonation of calcium tions, the formation of calcium carbonate are reversible reac [59PET/WIE] carbonate from anhydrous calcium oxalate is irreversible . A recent study involving high resolution thermogravimetry combined with Raman spectroscopy [2004FRO/WEI] corroborates the early findings: Whewellite is stable up to around 161ºC, above which temperature the anhydrous calcium oxalate is formed; at 479ºC the oxalate transforms to calcium carbonate, and above 684ºC calcium oxide is formed. Weddellite, Ca(ox)·2H O(cr), dehydrates above 100ºC first to the monohydrate 2 Ca(ox)·H O(cr) and then to the an [64WAL/LAN] . The results hydrous calcium oxalate 2 of this study have recently been corroborated by high resolution thermogravimetry [2003FRO/WEI] combined with Raman and infrared emission spectroscopy : Weddellite e range up to the pre-dehydra tion temperature of 97°C. At is the phase in the temperatur this temperature, the phase formed is whewellite, Ca(ox)·H O(cr), and above 114ºC the 2 phase is the anhydrous calcium oxalate. The dehydration of Ca(ox)·3H ° C at ambient pressure O(cr) commences at 50 2 and the monohydrate is formed, Ca(ox)·3H O(cr) Ca(ox)·H O(cr) + 2 H O(g). This U 2 2 2 nder elevated vapour pressure [64HOC/GER] reaction is irreversible even u , [68GER/WAT] . The thermal and related structural effe cts of the above dehydration and decar- bonation reactions, especially of Ca(ox)·H O(cr), have been studied and discussed in 2 [47PEL/DUV] detail in numerous studies [59PET/WIE] , [64HOC/GER] , , [64SIM/NEW] , [64WAL/LAN] , [65GER/WAT] , [65HOC/WAT] , [68GER/WAT] , [68PRI/FAZ] , , [87DOL] , [95RAK/SKU] , [2001MIA] , [2003FRO/WEI] [84BRE/SKR] [2004FRO/WEI] and . Table VI-13: Enthalpies of dehydration of calcium oxalate monohydrate reported in the literature. ο –1 Method H ∆ (kJ·mol ) Reference m dehyd (VI.11) O(cr) U Ca(ox)(cr) + H Ca(ox)·H O(g) 2 2 TGA 65.7 [64HOC/GER] TGA (65.7 ± 3.8) [68GER/WAT] transpiration 69.5 [77SHI/TAN] 74.8 DTA (Pt pan) [80TAN/MOR] 72.8 DTA (Ni block) ± 0.17) [80TAN/NEG] DSC (63.64 TGA (69 ± 3) [88NER/VIT] [90BIA/BUC] DSC 52.6 ± 3) [91NER/PRO] TGA (69.8 DSC (52.8 ± 2) [92BIA/BUC]

208 VI Discussion of data selection for oxalate 166 ο The dehydration enthalpy ∆ (VI.11) has been reported in several studies H m dehyd [64HOC/GER] [68GER/WAT] , [80TAN/NEG] , [88NER/VIT] , , , [80TAN/MOR] [90BIA/BUC] , [91NER/PRO] (Table VI-13). In a recent study , [92BIA/BUC] ο [2001TAN/FUR] ∆ determined by a transpira- systematic differences between H m dehyd tion method and differential scanning calorimetry (DSC) for various salts have been reported and an empirical “correction” function is proposed. However, the large dis- I-13) are not remedied by this empirical “cor- crepancies in the reported results (Table V rection” function, as for example DSC data reported by different authors vary consid- [80TAN/NEG] erably , , [92BIA/BUC] . None of the studies listed in [90BIA/BUC] Table VI-13 can be singled out as reliable, and an overall variation of more than 20 ο –1 kJ·mol in the reported ∆ values precludes any recommendation. H dehyd m ο 2– − 1 ∆ H The value − (830.7 ± 1.6) kJ ⋅ mol (ox selected in this re- , 298.15 K) = fm the heat of combustion of oxalic acid via view has been derived on an alternative route ( cf. Section VI.3). Section VI.2) and the heat of protonation of oxalate ( cf. The field of urinary stone research triggered a series of detailed investigations of the monohydrate whewellite [74NAN/GAR] on the kinetics of crystal growth , [92MIL/GRA] [75GAR/NAN] , on the growth , its dissolution kinetics and characterisa- tion of dihydrate (weddellite) crystals [82LEP/TAW] , and on the nucleation and crystal growth of the trihydrate [75GAR] . Although Ca(ox)·H , O(cr) is the thermody- [82HEI] 2 namically stable form at ambient conditions, and the dihydrate and trihydrate are me- tastable [97KON/TRA] , calcium oxalate crystallises from aqueous solutions in one or more of its hydrated forms depending upon pH, temperature, supersaturation, calcium to oxalate ratio and stirring dynamics of the system during nucleation and growth. Fur- thermore, soluble impurities can be the predominating factor in controlling the nuclea- tion and growth rates of these hydrates and thus the eventual form of the solid phase [75GAR] . A large number of studies has been published within the last 100 years report- ing solubility data for calcium oxalate hydrates in pure water [01RIC/MCC] , [03HER/MUH] [03KOH] , [08KOH] , [16HEN/TAY] , [27AUM2] , [27SCH3] , , [29HAM] , , [34BAS] , [39PED2] , [33KOL/SAN] , [49PED] , [51NYD] , [40SHE/PAL] [71CHU/REE] , [73FIN/ROT] , [79TOM/NAN] , [80HOD] , [80TOM/NAN] , [89SIN] , (Table VI-12), and in solutions containing NaCl [01GER] [29HAM] , , , [33MAL/GLU] [42MCC/RIE] [73FIN/ROT] [73KNA/MAT] , [79TOM/NAN] , [79TOM/NAN2] , , , [80TOM/NAN] , [98STR/TRA] , KCl [29HAM] , [42MCC/RIE] , , [89SIN] [79TOM/NAN] [85DAN/SON] , LiCl [29HAM] , HCl, HCl – KCl and HCl – KNO , 3 [16HEN/TAY] , CaCl Cl , [33MAL/GLU] , [40SHE/PAL] [29HAM] , NH and CaCl – 2 2 4 NaCl [73FIN/ROT] MgCl [29HAM] , [33MAL/GLU] , [71CHU/REE] , [73KNA/MAT] , 2 [89SIN] , MgCl , – NaCl [29HAM] , [73KNA/MAT] , [89SIN] , NH [33MAL/GLU] NO 3 4 2 [40SHE/PAL] [33MAL/GLU] SO , [29HAM] , Na [42MCC/RIE] , [40SHE/PAL] , , 2 4 K SO , MgSO [29HAM] , (NH [40SHE/PAL] ) , SO [33MAL/GLU] [33KOL/SAN] , 4 4 4 4 2 2 [33MAL/GLU] , H PO PO , Na – NaOH [45HOO/WIJ] [29HAM] HPO , [01GER] , Na 4 2 4 3 3 4

209 VI.5 Magnesium and calcium oxalate compounds and complexes 167 [71CHU/REE] ) , (NH ox [51NYD] , Na , acetic acid cit [71CHU/REE] , [73KNA/MAT] 3 2 4 [03HER/MUH] [34BAS] , urea [39PED2] , NH [71CHU/REE] , , synthetic urine 3 [82LEP/TAW] . The solid, especially the number of h ydration waters, has not been character- , , [01RIC/MCC] , [03KOH] , [08KOH] [01GER] ised in a number of older studies [16HEN/TAY] [27AUM2] [30RUF] , [33MAL/GLU] , [45HOO/WIJ] , , . The results of been accepted in this review. these investigations have not Inspection of the solubility values reported for Ca(ox)·H O in pure water 2 (Table VI-12) reveals some scatter, part of which might be caused by significant amounts of other hydrates present in the experiments despite the claims of the authors to have used pure monohydrate, and some of which might be caused by inappropriate experimental procedures, e.g [03HER/MUH] and [33KOL/SAN] . Solubility data ob- ., tained in pure water as listed in Table VI-12 are not considered in the final data analysis of this review. Also a number of solubility products have been reported in the literature as summarised in Table VI-14. Table VI-14: Solubility products of magnesium and calcium oxalate compounds reported in the literature and re-evaluated in this review. If not stated otherwise in the remarks, the solubility products are given for zero ionic strength. b a ° C) Reference Remarks t log K K ( log ,0 10 s ,0 10 s 2– 2+ U Mg “Mg(ox)” + ox Number given without any details ? – 4.07 [30RUF] 2+ 2– U Mg + ox Mg(ox)·2H O + 2H O 2 2 Recalculated by this review, see 18 – 5.5 [36BRI/JAR] Figure VI-8. [77FED/KHO] 25 – 5.68 Rejected, see Appendix A 2+ 2– “Ca(ox)” U Ca + ox ? – 8.75 [30RUF] Number given without any details 2+ 2– O Ca Ca(ox)·H + ox U + H O 2 2 37 – 8.51 – (8.53 ± 0.05) [29HAM] Data in NaCl and KCl re-evaluated 25 – 8.68 – (8.72 ± 0.06) [42MCC/RIE] Data in NaCl re-evaluated 25 – 8.67 [51NYD] Calculated from solubility in water ignoring Ca(ox)(aq) complexation Valid for 0.1 M KClO . Rejected, see 20 – 7.9 [63STA] 4 Appendix A. (Continued on next page)

210 VI Discussion of data selection for oxalate 168 Table VI-14 (continued) a b t log C) K ( ° K log Reference Remarks s 10 s 10 ,0 ,0 2+ 2– O + ox U Ca Ca(ox)·H O + H 2 2 [71PAZ/STE] 20 – 8.87 Rejected, see Appendix A – (8.54 0.08) [73FIN/ROT] Data in NaCl re-evaluated 38 ± ± [73KNA/MAT] – (8.58 37 0.06) I > 0.1 M NaCl re-evaluated Data in No experimental raw data reported 15 – 8.81 [74NAN/GAR] ± 25 – (8.70 0.02) 35 – 8.61 45 – 8.55 No experimental raw data reported 0.03) 37 – (8.60 ± [75PAK/OHA] 0.09) [79TOM/NAN] – (8.53 37 ± Data in NaCl and KCl re-evaluated Data in 0.15 M NaCl re-evaluated – 8.79 [79TOM/NAN2] – 8.92 15 25 – 8.78 – 8.63 – 8.66 – 8.49 37 – 8.57 50 – 8.38 Calculated from solubility in water 0.10) ± 37 – (8.65 [80HOD] ignoring Ca(ox)(aq) complexation [85DAN/SON] 37 – 8.55 Data in NaCl re-evaluated – 8.66 – (8.49 ± 0.08) [89SIN] 37 Data in NaCl re-evaluated, results ± 0.02) – (8.82 ± 0.11) [98STR/TRA] 20 – (8.84 from ISE and AAS data averaged ± 0.01) ± 0.07) 25 – (8.77 – (8.74 0.01) ± 0.04) ± 30 – (8.71 – (8.66 ± – (8.58 ± 0.08) 37 – (8.65 0.03) ± 0.02) – (8.56 40 – (8.62 0.08) ± 2– 2+ Ca(ox)·2H Ca + ox O + 2H O U 2 2 [75GAR] 37 – 8.3 Number given without any details 15 – 8.49 – 8.37 [80TOM/NAN] Data in 0.15 M NaCl re-evaluated 25 – 8.38 – 8.24 Data in 0.15 M NaCl re-evaluated 37 – (8.18 0.12) Data re-evaluated, 0.03–0.3 M NaCl ± – 8.20 Data in 0.15 M NaCl re-evaluated – 8.01 50 Data in NaCl re-evaluated, results – (8.39 ± 0.09) [98STR/TRA] ± 20 – (8.42 0.02) from ISE and AAS data averaged 25 – (8.34 0.02) – (8.30 ± 0.06) ± 30 – (8.26 ± 0.03) – (8.21 ± 0.08) 37 – (8.17 0.03) – (8.12 ± 0.07) ± 40 – (8.13 ± 0.04) – (8.08 ± 0.09) (Continued on next page)

211 VI.5 Magnesium and calcium oxalate compounds and complexes 169 Table VI-14 (continued) a b t ° log Reference Remarks K C) ( K log s 10 s 10 ,0 ,0 2+ 2– Ca(ox)·3H Ca + ox O + 3H U O 2 2 [75GAR] Number given without any details 37 – 8.08 ± 37 [79TOM/NAN] Data in NaCl and KCl re-evaluated – (8.00 0.07) – 8.48 15 [79TOM/NAN2] – 8.36 Data in 0.15 M NaCl re-evaluated – 8.32 25 – 8.17 37 – 8.14 – 7.98 – 7.81 50 – 7.99 Data in NaCl re-evaluated, results – (8.29 ± 0.03) 20 – (8.33 ± 0.01) [98STR/TRA] 25 – (8.24 0.01) – (8.20 ± 0.04) from ISE and AAS data averaged ± ± 0.02) – (8.07 ± 0.06) 30 – (8.12 37 – (8.02 ± 0.02) – (7.96 ± 0.07) 40 – (7.97 ± – (7.92 ± 0.11) 0.02) n as reported in the literature. a: Values and uncertainties give (see text for details). b: Values and uncertainties re-evaluated and accepted by this review [75PAK/OHA] fits well to the et al. The solubility product reported by Pak The solid has been char acterised in detail, other data reported at 37ºC (Table VI-14). and series of experiments varying pH, ionic strength and total concentrations of calcium and oxalate all resulted in consistent solubility products [75PAK/OHA] . However, the Ca(ox)(aq) stability constant = 3.19, valid for 25ºC, has been used in all calcu- log K 10 s ,0 lations, and no experimental data are reported in [75PAK/OHA] , preventing any recal- culation. Hence, the result of this study has not been included in the final data analysis. Further results listed Table VI-14 but rejected by this review are [30RUF] , [51NYD] , [75GAR] , [80HOD] (see remarks in Table VI-14) and [63STA] , [71PAZ/STE] (see discussion in Appendix A). For the final data analysis this review considered only studies carried out in NaCl and KCl media with proper characterisation of the solids in equilibrium with the [29HAM] solutions: [42MCC/RIE] , [73FIN/ROT] , [73KNA/MAT] , [79TOM/NAN] , , [79TOM/NAN2] , [80TOM/NAN] , [85DAN/SON] , [89SIN] , [98STR/TRA] . In all the above cases, when analytical equilibrium concentration data of dis- solved calcium, oxalate, NaCl or KCl have been reported, the aqueous speciation was recalculated in this review as follows: 2 − The complexes Ca(ox)(aq) and Ca(ox) e speciation calcula- are included in th 2 tions. For Ca(ox)(aq), 2+ 2– Ca + ox U Ca(ox)(aq) (VI.12)

212 VI Discussion of data selection for oxalate 170 ο the values (VI.12) = (3.19 ± 0.06) valid at 25ºC and selected in this review ( cf. K log 10 1 ο Section VI.5.2), and log K (VI.12) = (3.27 ± 0.07) valid at 37ºC were used. For inter- 1 10 polation and extrapolation to other temperatures (15 to 50ºC) a constant enthalpy of 2 − reaction has been calculated from these two values. For , Ca(ox) 2 2– 2 − Ca(ox) U (VI.13) Ca(ox)(aq) + ox 2 ο log K cf. (VI.13) = (0.83 ± 0.19) valid at 25ºC and se lected in this review ( the value 2 10 Section VI.5.2) was used. It has been assumed that the iso-coulombic Reaction (VI.13) is temperature independent. In order to adjust the above constants to the actual ionic strength of a particular equilibrium experiment the SIT has been used (see Appendix B) with the following 2+ – 2– + interaction coefficients: ± (Ca 0.01) ( Appendix B), ε (ox , Cl , Na ε ) = ) = (0.14 cf. 2– + 0.01) and ε (ox − , K (0.08 ) = (0.07 ± 0.08) selected in this review ( cf. Section ± VI.3.5), and it has been assumed that ε ≈ ε (Ca(ox)(aq), NaCl) = (Mg(ox)(aq), NaCl) − − + 2 + 2 , Na ) = Mg(ox) ε (Ca(ox)(aq), KCl) = (0.0 ± ) ≈ ε ( 0.1), and ε Ca(ox) ( , Na 2 2 2 − + Ca(ox) , K ( ) = − ε ± 0.10) ( cf. Section VI.5.2). (0.15 2 − 2 Ca(ox) is unimportant ns showed, the complex As the speciation calculatio 2 (< 0.04 % of total oxalate in all cases) in the calcium oxalate equilibrium experiments considered in this review, and hence the uncertainties of its stability constant and related SIT interaction coefficient have no influence on the results. However, the complex stability of Ca(ox)(aq) turned out to be a cr itical parameter in a consistent re-evaluation of calcium oxalate solubility experiments (1 to 10 % of total oxalate, depending on ionic strength and temperature). Using the results of the speciation calculations, ., the calculated concentra- i.e 2+ 2– tions of [Ca ] and [ox ], the solubility products of the reactions: 2– 2+ O(cr) U Ca + ox O(l) (VI.14) + H Ca(ox)·H 2 2 2+ 2– U Ca + ox + 2 H O(l) (VI.15) O(cr) Ca(ox)·2H 2 2 2– 2+ O(l) (VI.16) O(cr) Ca + ox + 3 H U Ca(ox)·3H 2 2 were calculated and extrapolated to zero ionic strength with the same SIT interaction parameters as used in the speciation calculations. The very detailed and careful early study of precipitation and solubility of cal- cium oxalate hydrates by Hammarsten [29HAM] contains solubility data for Ca(ox)·H O in various electrolytes. The solubility data in NaCl and KCl have been used 2 to recalculate solubility products in this review. McComas and Rieman [42MCC/RIE] prepared Ca(ox)·H O by two different 2 methods and measured its solubility at (25.0 ± 0.2)ºC in NaCl (0.1 – 1.0 M). The meas- ured equilibrium concentrations are reported in tabular form and solubility products have been recalculated there from in this review.

213 VI.5 Magnesium and calcium oxalate compounds and complexes 171 et al. Finlayson report that X-ray diffraction was used to verify [73FIN/ROT] that the powder they used was whewellite, Ca(ox)·H O, before and after the solubility 2 studies. The solubility data at 38ºC in NaCl are reported in graphical form only [73FIN/ROT] . They have been digitised and solubility products have been recalculated in this review. The Ca(ox)·H [73KNA/MAT] O solubility data of Knappwost and Matouschek 2 at 37ºC in NaCl were reported in graphical form only. They have been digitised and solubility products have been recalculated in this review from measurements at I > 0.1 M NaCl. Solubility data for Ca(ox)·H O and Ca(ox)·3H O in 0.15 M NaCl at various 2 2 . Both hydrates and Nancollas [79TOM/NAN2] temperatures were reported by Tomaži č diffraction, scanning electron were characterised by X-ray microscopy, chemical analy- sis and by determining the water content of the solids. The transformation of the tri- hydrate to the monohydrate during measurements has been followed by X-ray diffrac- tion. The reported solubility data have been used to recalculate solubility products in this review. Solubility data for Ca(ox)·H O and Ca(ox)·3H O at 37ºC in NaCl (0.03 – 0.3 2 2 M) and KCl (0.1 – 0.3 M) were reported by Tomaži č and Nancollas [79TOM/NAN] including results already published in [79TOM/NAN2] . The same procedures have been applied as described in [79TOM/NAN2] , and the reported solubility data in NaCl and KCl have been used to recalculate solubility products in this review. Solubility data for Ca(ox)·2H O in 0.03 – 0.3 M NaCl at 37ºC and 0.15 M 2 and Nancollas [80TOM/NAN] NaCl at 15, 25 and 50ºC were reported by Tomaži č . The [79TOM/NAN2] same procedures have been applied as described in , and the reported solubility data have been used to recalculate solubility products in this review. The Ca(ox)·H O solubility data of Singh at 37ºC in NaCl have been [89SIN] 2 used to recalculate solubility products in this review. Ca(ox)·H O and Ca(ox)·3H O solubilities were studied at 20, 25, O, Ca(ox)·2H 2 2 2 30, 37, and 40ºC in 0.02, 0.05, 0.10 and 0.20 M NaCl by Streit et al. . All [98STR/TRA] calcium oxalate hydrates were characterised by a combination of X-ray diffraction, scanning electron microscopy and thermogravimetric analysis. After the solubility measurements, the Ca(ox)·2H O and Ca(ox)·3H O crystals were analysed again to con- 2 2 firm that no transformation to the thermodynamically stable Ca(ox)·H O had occurred. 2 2+ During the solubility experiments the concentration of Ca was continuously measured by a calcium ion selective electrode (ISE in Figure VI-9 and Figure VI-10). In addition, the total dissolved Ca concentrations were also measured by atomic adsorption spec- troscopy (AAS in Figure VI-9 and Figure VI-10). The latter data have been used to recalculate a full specia tion in this review, whereas the ISE data are assumed to repre- 2+ sent free Ca concentrations.

214 VI Discussion of data selection for oxalate 172 summarised in Table VI-14 and shown The results of these recalculations are in Figure VI-9 and Figure VI-10. No attempt has been made in this review to assess individual uncertainties of the calculated solubility products. In the case of series of data derived for a certain tem- perature but varying ionic strength unweighted averages were calculated with 95% confidence intervals (Table VI-14). Figure VI-9 and Figure VI-10 reveal that the re-evaluated solubility products, extrapolated to zero ionic strength with para meters selected and estimated in this re- view, do not show any systematic ionic strength dependence, which is an indication of the internal consistency of the speciation model derived in this review. Especially the 2– + applicability of SIT using the interaction parameter , Na ε (ox ) derived from oxalate – protonation reactions but without considering a complex Na(ox) in the speciation model ( cf. Section VI.3) is corroborated by the consistent results obtained at I > 0.1 M m NaCl. Figure VI-9: Solubility products of calcium oxalate hydrates at 25ºC, calculated from analytical solubility data and extrapolated to zero ionic strength using parameters discussed in the text. -8.0 m m I I ) 25°C ) -8.1 + + [42MCC/RIE] [79TOM/NAN2] 2- 2+ ,Na , Na Ca(ox)·3H Ca + 3 H O [80TOM/NAN] + ox O 2 2 U -8.2 2- 2– [98STR/TRA] ISE (ox 2+ 2- [98STR/TRA] AAS (ox -8.3 Ca(ox)·2H O Ca + ox + 2 H O 2 2 U ε ) + ) + - – -8.4 ,Cl , Cl 2+ -8.5 2+ (Ca (Ca ε -8.6 2+ 2- O + H + ox Ca O Ca(ox)·H 2 2 U + ( + ( -8.7 D D – 8 - 8 -8.8 s,0 s,0 K K 10 -8.9 10 log log -9.0 0.0 0.1 1.0 0.2 0.3 0.4 0.9 0.6 0.7 0.5 0.8 / molal I m

215 VI.5 Magnesium and calcium oxalate compounds and complexes 173 Figure VI-10: Solubility products of calcium oxalate hydrates at 37ºC, calculated from analytical solubility data and extrapolated to zero ionic strength using parameters discussed in the text. -7.8 m m I I 37°C ) [29HAM] ) + -7.9 + [73FIN/ROT] 2+ 2- , X ,X [73KNA/MAT] Ca(ox)·3H + 3 H Ca O O + ox 2 2 U 2– 2- -8.0 [79TOM/NAN] [79TOM/NAN2] (ox (ox ε [80TOM/NAN] 2+ 2- -8.1 ox ) ·2H O Ca + ox + 2 ( Ca 2 U [89SIN] ) + ) + – - [98STR/TRA] ISE -8.2 [98STR/TRA] AAS ,Cl , Cl 2+ 2+ -8.3 (Ca (Ca ε 2+ 2- -8.4 + ox H H · ) ox ( Ca Ca O + O 2 2 U + ( + ( D D -8.5 – 8 - 8 -8.6 s,0 s,0 K 10 K -8.7 10 log log -8.8 0.1 0.3 0.4 0.5 0.6 0.7 0.2 0.0 / molal I m The data shown in Figure VI-9 and Figure VI-10 have been used to calculate global unweighted averages of solubility products at 25 and 37ºC. The results obtained for 25ºC have been selected: ο ± 0.06) K (VI.14) = – (8.73 log ,0 10 s ο 0.06) ± (VI.15) = – (8.30 log K s 10 ,0 ο K (VI.16) = – (8.19 ± log 0.04) s ,0 10 K log (VI.14)= – (8.54 ± 0.08), whereas the results obtained for 37ºC, s 10 ,0 0.09), and 0.09) are used for (VI.15) = – (8.13 ± (VI.16) = – (7.98 ± log K K log ,0 ,0 10 10 s s consistency checks of the selected reaction enthalpies (Figure VI-11). The solubility of Ca(ox)·H O in 0.03 to 0.50 M KCl at 37ºC has been measured 2 by Daniele . The analytical solubility data are not reported, only [85DAN/SON] et al. calculated solubility products are given. However, the speciation model applied and the constants used by Daniele selected by this review, and are very similar to the ones et al. the solubility products were extrapolated to zero ionic strength using SIT in [85DAN/SON] K . The reported result, log (VI.14) = – 8.55, is in good agreement 10 ,0 s with the value obtained from data mainly in NaCl at 37ºC, log K (VI.14) = 10 ,0 s 0.08), in this review. − (8.54 ±

216 VI Discussion of data selection for oxalate 174 Enthalpies of dissolution or precipitation of calcium oxalate hydrates reported in the literature (Table VI-15) have either been derived from the temperature depend- [74NAN/GAR] ence of solubility products [80TOM/NAN] , , [79TOM/NAN2] , [98STR/TRA] ∂ T” in Table VI-15), or have been measured calorimetri- pK ∂ (method “ / a [74NAN/GAR] cally [93SOH/COS] [92VEL/SOH] [97SOH/KRO] , [98STR/TRA] . , , , Table VI-15: Enthalpies of dissolution of calcium oxalate hydrates reported in the literature and re-evaluated in this review. If not stated otherwise, the enthalpy values refer to 25ºC. a b ο ο –1 –1 ) H (kJ·mol ∆ (kJ·mol H ) Reference Method ∆ m sol sol m 2+ 2– Ca O(cr) + ox Ca(ox)·H + H U O(l) 2 2 c ± (22.3 ± 0.8) [74NAN/GAR] cal (22.3 0.4) pK ∂ / ∂ T (23.0 ± 0.4) a ∂ pK 3.0) ∂ 1.3) (20.8 ± ± [79TOM/NAN2] / T (17.3 a c 0.4) cal (18.1 ± [92VEL/SOH] c cal 17.5 [93SOH/COS] c cal (19.83 ± 1.24) ± 2.9) [97SOH/KRO] (19.8 c, d ± (16.69 0.58) ∂ pK / ∂ T (18.9 ± 0.9) (23.0 ± 2.7) [98STR/TRA] a 0.8) 0.8) (20.8 ± ± cal (20.8 2+ 2– Ca(ox)·2H U Ca O(l) + ox O(cr) + 2 H 2 2 pK ∂ T (14.1 ± 1.3) [80TOM/NAN] ∂ / a 2.0) T (25.6 ∂ pK ± 0.5) (27.1 ± / [98STR/TRA] ∂ a c ± 1.4) cal (24.3 (24.3 1.4) ± 2– 2+ Ca(ox)·3H Ca + ox O(cr) + 3 H O(l) U 2 2 pK ∂ ∂ T (25.1 ± 0.7) (28.0 ± 1.9) [79TOM/NAN2] / a ∂ pK / ∂ T (32.1 ± 1.2) (33.2 ± 3.3) [98STR/TRA] a ± (30.5 ± 2.3) cal (30.5 2.3) n as reported in the literature. a: Values and uncertainties give b: Values and uncertainties re-evaluated and (see text for details). accepted by this review itation, multiplied by (–1). c: Value obtained from precip d: Enthalpy value refers to 37ºC. The temperature dependence of solubility products , recalculated in log K s ,0 10 [79TOM/NAN2] this review from analytical solubility data reported in and [98STR/TRA] , (Table VI-14) has been described by a constant enthalpy model, i.e. ο assuming = 0. The results of least-squares fits of unweighted values, C ∆ log K 10 ,0 ,m s p sol ο ed in Table VI-15. ∆ with 95% confidence intervals are summaris H sol m are Solubility products reported by Nancollas and Gardner [74NAN/GAR] given as numbers only without any further details. No analytical equilibrium solubility data are provided which would allow consis tency checks and recalculations. Attempts –1 ο by this review to reproduce the reported enthalpy H , 0.4) kJ·mol = (23.0 ± ∆ sol m

217 VI.5 Magnesium and calcium oxalate compounds and complexes 175 , failed and [74NAN/GAR] claimed to be derived from the reported solubility products hence, this value has not been included in the final data analysis. Figure VI-11: Temperature dependence of solubility products of calcium oxalate ο ο hydrates. The lines are calculated using H K log and ∆ values selected in this sol m 10 s ,0 review with their assigned uncertainties. -7.6 [74NAN/GAR] -7.8 Ca(o x)·3H O U 2 [79TOM/NAN2] 2- 2+ [80TOM/NAN] O + ox + 3 H Ca 2 [98STR/TRA] -8.0 25°C selected values 37°C mean values O Ca(o x)·2H 2 U ° -8.2 s,0 2+ 2- O + 2 H + ox Ca 2 K 10 -8.4 log -8.6 O Ca(o x)·H U 2 2- 2+ Ca O + ox + H 2 -8.8 -9.0 0.0030 0.0031 0.0032 0.0033 0.0034 0.0035 0.0036 (K) T 1 / The solubility products recalculated in this review from analytical solubility seem to belong to two discrepant data sets (Figure [80TOM/NAN] data reported in VI-11). Enthalpy values derived from data at 15 and 25ºC, as well as from 37 and 50ºC agree well with results obtained from [98STR/TRA] data and with the temperature variation of average solubility products at 25 and 37ºC calculated in this review. How- ο –1 H ∆ [80TOM/NAN] ever, an overall fit of all , data results in 4.3) kJ·mol ± = (17.3 sol m a value clearly discrepant with all other enthalpies listed in Table VI-15, which has not been included in the final data analysis. Velich and co-workers published three pa pers describing an isoperibolic reac- tion twin calorimeter and measuremen ts of the enthalpy of Ca(ox)·H O(cr) precipitation 2 [92VEL/SOH] . In the first paper [97SOH/KRO] , [93SOH/COS] , [92VEL/SOH] no lids is reported. In the second paper characterisation of the precipitating so [93SOH/COS] they state that “in all experi ments crystals of characteristic Ca(ox)·H [97SOH/KRO] O(cr) shape were formed”. In the third paper they state that 2 “Ca(ox)·H O(cr) precipitated as a single solid phase under studied conditions, as deter- 2 mined by X-ray diffraction”. The somewhat discrepant results obtained over time by the

218 VI Discussion of data selection for oxalate 176 or same method are not discussed in [97SOH/KRO] [93SOH/COS] . Taking the im- of the solid as an indicator of improvement of the quality provements in characterisation review considers the results of in experimental procedures, this [97SOH/KRO] as reli- able and an unweighted average of reported results (Table 1 in [97SOH/KRO] ) with a 95% confidence interval is included in the final data analysis. Weighted means of the accepted valu es (Table VI-15) were selected: ο –1 ± 0.5) kJ·mol ∆ H (VI.14) = (21.5 m sol ο –1 (VI.15) = (25.2 ± 1.1) kJ·mol H ∆ sol m ο –1 H ∆ 1.3) kJ·mol . (VI.16) = (29.7 ± m sol [74NAN/GAR] The temperature dependence of solubility products reported by and recalculated in this review from data given in [79TOM/NAN2] , [80TOM/NAN] [98STR/TRA] and is consistent, within the estimated confidence intervals, with the values calculated using solubility products and solution enthalpies selected in this re- view (Figure VI-11). ο G ∆ review for Reaction (VI.14), The values selected in this (298.15 K) = m sol ο –1 –1 ο 0.5) kJ·mol log K − = (49.83 ± 0.34) kJ·mol , and °·ln(10)· R· H T ∆ = (21.5 ± sol m 10 s ,0 ο ο ( = ( T ∆ 8.15 K), are used to calculate H ° is the reference temperature, 29 ∆ – S m sol m sol ο ο ο –1 2+ –1 S ∆ ∆ S G 2.0) J·K . From Reaction (VI.14) ·mol ) / T ) + ° = − = (Ca ( 95.0 ± m m sol sol m 2+ ο ο 2– ο ο S S ) = O, l) – S S ) + (Ca(ox)·H (Ca O, cr), using CODATA values (ox (H 2 2 m m m m –1 –1 ο –1 –1 − and (56.2 ± S 1.0) J·K (H , and O, l) = (69.95 ± 0.03) J·K ·mol ·mol 2 m ο –1 –1 S (Ca(ox)·H ± 2.0) J·K selected in this review, the selected ·mol O, cr) = (156.4 2 m value of the standard molar entropy of the oxalate anion is calculated: ο 2– –1 –1 ± 3.0) J· K S ·mol , 298.15 K) = (47.6 . (ox m ο ο ο 2– ∆ S S S The standard molar entropy of formation, ) – 2· = (C, cr) – (ox fm m m ο ο ο 2· (C, cr) = , g) + ( n /2)· S S S (H , g), is calculated using the CODATA values (O 2 2 m m m –1 –1 ο –1 –1 ο (5.74 ± ·mol (O (H , g) = (205.152 ± 0.005) J·K ·mol , , g) = , and 0.10) J·K S S 2 2 m m –1 –1 (103.680 , where ± 0.003) J·K n = − 2 is the charge of the oxalate anion, ·mol –1 –1 ο 2– (ox (504.9 ± 3.0) J·K S ·mol − . ∆ , 298.15 K) = fm ο ο ο ∆ , using ∆ H G ∆ – T · (298.15 K) = S According to the equation, fm fm fm ο 2– –1 ∆ H , 298.15 K) = − (830.7 ± 1.6) kJ ⋅ mol (ox selected in this review ( cf . Section fm VI.2.3), the standard molar Gibbs energy of formation of the oxalate anion is selected: ο 2– − 1 (ox . , 298.15 K) = − (680.1 ± G ⋅ mol 1.8) kJ ∆ fm ο 2– S (ox , 298.15 K) and with the selected This value may be combined with m for reactions (VI.14), (VI.15) and (VI.16). equilibrium constants and enthalpy changes 2+ Using auxiliary data for H O(l) and Ca it is possible to calculate the standard molar 2 free energies of formation, enthalpies of formation and entropies for Ca(ox) ⋅ H O(cr), 2

219 VI.5 Magnesium and calcium oxalate compounds and complexes 177 O(cr) at 298.15 K. The results are given in Table ⋅ ⋅ 3H Ca(ox) O(cr) and Ca(ox) 2H 2 2 VI-16. Table VI-16: Selected formation data for Ca(ox) ⋅ O(cr), Ca(ox) ⋅ 2H H O(cr) and 2 2 Ca(ox) ⋅ 3H O(cr). 2 ο ο ο –1 –1 –1 –1 S ) ∆ (kJ·mol (kJ·mol ∆ ) G (J·K ·mol ) H fm fm m ⋅ ± H O(cr), – (1519.9 Ca(ox) 2.1) – (1681.0 ± 1.9) (156.4 ± 2.0) 2 Ca(ox) 2H 5.0) O(cr) – (1754.6 ± 2.1) ⋅ ± 2.2) (205.7 ± – (1970.5 2 5.5) ± (258.4 2.3) O(cr) – (1991.1 ± 2.1) – (2260.8 ± 3H Ca(ox) ⋅ 2 2– and the selected reaction data for Reaction The selected formation data for ox r = 1 (see Section VI.3.4, page143) yield: (VI.9), ο –1 – G ∆ (Hox ± 1.8) kJ·mol , 298.15 K) = – (704.4 fm ο – –1 –1 ± S 3.0) J·K (Hox ·mol , 298.15 K) = (153.4 . m and [27SCH/GAD] (ox) O (M = K, Rb, Cs) ·H Compounds of the type CaM 2 2 2 CaK(ox)(HCOO)·H [27SCH2] have been prepared from boiling solutions in crystal- O 2 line form, and their stoichiometry has been confirmed by elemental analysis. However, all these compounds do not exhibit congruent equilibrium solubility but rapidly decom- pose in water by releasing alkali oxalate or formiate, respectively, and hence no ther- modynamic data are available for any of these compounds. Synthesis and thermal decomposition studies of Ca[Ca(ox) ]·2H O(cr) have 2 2 been reported [96DEB/BAR] . However, the authors failed to show, ., by X-ray pow- e.g der diffraction, that the compound they worked with was structurally different from whewellite, Ca(ox)·H O(cr). 2 VI.5.2 Magnesium and calcium oxalate complexes Complex formation in Mg and Ca oxalate systems has been studied by several investi- gators by a variety of experimental methods. The equilibrium data found in the litera- ture are summarised in Table VI-17.

220 VI Discussion of data selection for oxalate 178 Table VI-17: Experimental equilibrium data for the Mg and Ca oxalate systems. The uncertainties are given as reported in the references. ( ° C) log Method Ionic medium K Reference t 10 2+ 2– U Mg(ox)(aq) Mg + ox 18 3.43 con , [32MON/DAV] → 0 [27DAV] 25 2.55 [38CAN/KIB] ise-H 0.2 M (KCl) 37 sol [39PED] 0.32 M NaCl 2.28 → 18 3.41 con 0 pot 1 M KNO [57LEF] 27 1.99 3 ise-Hg 0.10 M NaNO 20 (2.76 0.04) [57SCH/AND] ± 3 0.10 M NaCl 20 ± 0.10) [60RAA] sp (2.76 0.2 M KCl + 0.1 M Tris 2.61 [62ASA/MOR] (a) 23 0.1 M KClO dis 20 (2.39 ± 0.05) [63STA] 4 ≈ 0.1 M Tris buffer sp 2.88 [63WAT/TRO] 25 ≈ 2.90 0.1 M Tea buffer gl 0.10 M NaClO 25 (3.10 ± [75PRA/JON] 0.05) 4 sol 0.5 M NaClO 25 1.62 [77FED/KHO] 4 1 1.52 2 1.44 1.40 3 4 1.37 1.34 6 8 1.32 gl 0.03 M Et NI 37 (3.02 ± 0.03) [82DAN/MAR] 4 (2.92 ± 0.03) 0.05 (2.75 0.10 0.04) ± (2.54 0.04) 0.30 ± (2.52 ± 0.04) 0.50 → 0 3.58 pot 0.10 M KNO 35 4.65 [85RED/RAO] 3 pot 1.0 M NaClO 30 ? 2.65 [88GHA/MAN] 4 cou 0.15 M NaCl 25 2.18 [93GLA/MAJ] 37 2.39 pot 0.3 m NaCl 25 (2.33 ± 0.04) [2001CHO/BON] 1 m (2.00 ± 0.01) 2 m ± 0.02) (1.94 3 m (1.77 ± 0.08) 4 m (1.99 ± 0.01) 5 m ± 0.06) (2.00 (Continued on next page)

221 VI.5 Magnesium and calcium oxalate compounds and complexes 179 Table VI-17 (continued) ( ° C) log Method Ionic medium t Reference K 10 2 − 2– 2+ + 2 ox Mg Mg(ox) U 2 0 4.37 [51BAR/ARG] → 25 sol [59TEK/VIN2] ise-ox 0.089 M 25 4.24 0.049 M 4.54 4.54 0.031 M ± 25 (4.00 0.01) pot 0.3 m NaCl [2001CHO/BON] (3.71 0.01) 1 m ± ± 0.06) 2 m (3.73 (3.79 0.09) 3 m ± ± 0.08) 4 m (4.07 (3.99 ± 0.06) 5 m − 2 2– Mg(ox)(aq) + ox U Mg(ox) 2 [57SCH/AND] 0.1 M (NaNO ) 20 < 1 ise-Hg 3 2+ – + U Mg(Hox) + Hox Mg 0.2 M (KCl) 25 ≈ 0.5 [38CAN/KIB] ise–H 2+ 2– + ox Ca U Ca(ox)(aq) con 0 18 3.00 [32MON/DAV] → 1 M NaClO 25 1.64 [67HAS/MAK] dis 4 0.05 M NaCl 25 3.07 [68GOR/FIL] cix em 0.1 M (K,H)NO 25 (2.30 ± 0.07) [70STE/PAZ] 3 cix 0.02 NaClO –Na [72ARM/DUN] ox 25 (2.08 ± 0.04) 2 4 (2.08 ± 0.04 0.04) ± (2.11 0.06 0.03) ± 0.03) 0.08 (2.20 (2.10 0.03) 0.10 ± 0.20 (2.06 ± 0.04) → 0 38 (3.37 ± sp [73FIN/SMI] 0.01) [74NAN/GAR] → (3.19 ± 0.01) 25 con 0 gl 0.10 M NaClO 25 (3.5 ± [75PRA/JON] 0.05) 4 0.5 M NH dis NO ± 0.01) [76MCD/KEL] 25 (1.95 4 3 ise-Ca 0.05 M NaCl 25 (2.70 ± 0.03) [79CRA/MOO] 0.10 ± 0.09) (2.54 (1.99 0.04) ± 0.15 ise-Ca 0.03 M Et NI 37 (2.72 ± 0.04) [82DAN/MAR] 4 (2.61 ± 0.04) 0.05 (2.46 ± 0.05) 0.10 0.30 (2.23 ± 0.05) 0.50 (2.22 0.06) ± 0 3.27 → gl, ise-Ca 0.0015 M 37 (2.24 ± 0.04) [82RAO/AGA] 0.10 M KNO pot 35 4.85 [85RED/RAO] 3 cou 0.15 M NaCl 25 1.46 [93GLA/MAJ] (Continued on next page)

222 VI Discussion of data selection for oxalate 180 Table VI-17 (continued) ° ( C) log Method Ionic medium t K Reference 10 2 − 2+ 2– + 2 ox Ca(ox) Ca U 2 1 M NaClO 25 2.68 [67HAS/MAK] dis 4 0 20 3.49 [71PAZ/STE] sol → – 2+ + + Hox Ca(Hox) U Ca [73ARM/DUN] 0.1 M HClO cix ox 25 (1.38 ± 0.09) – H 2 4 2+ – Ca U Ca(Hox) (aq) + 2 Hox 2 cix – H 0.1 M HClO ox 25 ≈ 1.84 [73ARM/DUN] 2 4 a: Refractive index measurements with an interferometer. The stabilities of Mg oxalate 1:1 and 1:2 complexes are in a range suitable for direct determination by alkalimetric titr ation (Figure VI-12 and Figure VI-13). Figure VI-12: Simulated titration curves in 0.3 molal NaCl (left) and 3 molal NaCl –4 –4 –4 (right) of 2 ox(aq) (upper curve) and 2 × 10 × molal H molal H × 10 10 ox(aq) and 2 2 2 molal MgCl (lower curve). The do tted lines are calculated neglecting the complex 2 2 − Mg(ox) denotes the moles of base added per moles of a in the latter case. The symbol 2 ligand present in solution. Stability constants selected in this review have been used to calculate the titration curves. 6 6 5 5 ] ] ] + + ] + + [H [H H [H 10 10 4 [ 4 10 10 log log g – log – lo 3 3 2 2 00.511.52 00.511.52 a a

223 VI.5 Magnesium and calcium oxalate compounds and complexes 181 Figure VI-13: Distribution of complex species in the simulated titration of the Mg oxalate system (Figure VI-12) in 0.3 m NaCl (left) and 3 m NaCl (right). 100 100 90 90 80 2+ 80 [Mg ] 70 70 2+ ] [Mg 60 60 − [Hox ] − 50 50 [Hox ] 40 40 2 − a g( q)] )( M [ ox [ox ] 30 30 [M ox )( a ] g( q) % Concentration % Concentration 20 20 2 − 2 − 2 [Mg(ox) ] − 2 ] [Mg(ox) 2 ] [ox ox(aq)] [H 2 10 10 [H ox(aq)] 2 0 0 2345 2345 + + − − [H ] log ] [H log 10 10 –3 M and low ionic strength the In the concentration range [Mg], [ox] < 10 − 2 complex is a minor species and its impact on the titration curve is very small Mg(ox) 2 (Figure VI-12 and Figure VI-13 left), whereas its relative influence increases with in- creasing ionic strength (Figure VI-12 and Figure VI-13 right). This behaviour explains why generally in studies carried out at low i onic strength the results were interpreted in terms of Mg(ox)(aq) only. The literature data collected in Table VI-17 have been scrutinised in order to select reliable studies, summarised in Table VI-18, on which the evaluation of recom- mended values is based. experimental procedures or the report- Because of various shortcomings in the ing of the results, the values of [27DAV] , [32MON/DAV] , [59TEK/VIN2] , [62ASA/MOR] , , [63STA] , [63WAT/TRO] , [75PRA/JON] , [77FED/KHO] [85RED/RAO] , have been discarded (see detailed [93GLA/MAJ] , [88GHA/MAN] discussions in Appendix A). The values for Mg oxalate complexation derived from Mg(ox)·2H O(cr) solubility measurements [51BAR/ARG] are not used in the final data 2 analysis because of the uncertainties due to the slow reaction kinetics of magnesium oxalate (see discussion in Section VI.5.1). + The species Mg(Hox) has been postulated in a single titration study [38CAN/KIB] , but with the remark that the reported formation constant is “highly tenta- tive”. The existence of protonated Mg oxalate complexes remains to be shown by other + techniques, and thus Mg(Hox) is disregarded by this review.

224 VI Discussion of data selection for oxalate 182 r the Mg and Ca oxalate systems. The Table VI-18: Accepted equilibrium data fo uncertainties are estimated in this review. ο ( ° log K Ionic medium log K t Reference C) 10 10 2+ 2– Mg Mg(ox)(aq) U + ox a (3.59 ± 0.20) ± (2.55 [38CAN/KIB] 25 0.2 M (KCl) 0.20) 0.32 M NaCl ± 0.10) (3.52 ± 0.10) [39PED] 37 (2.28 a 1 M KNO 0.20) (3.38 ± 0.20) ± [57LEF] 27 (1.99 3 a 0.10 M NaNO 0.08) (3.62 ± 0.08) ± [57SCH/AND] 20 (2.76 3 20 (2.76 ± 0.20) (3.61 ± 0.20) [60RAA] 0.10 M NaCl 0.03 M Et NI 37 (3.02 ± [82DAN/MAR] 0.10) 4 ± 0.10) 0.05 (2.92 ± 0.10 (2.75 0.10) 0.10) 0.30 (2.54 ± b ± ± 0.07) 0.10) 0.50 (2.52 (3.58 25 0.3 m NaCl ± 0.10) [2001CHO/BON] (2.33 1 m (2.00 ± 0.10) 2 m (1.94 ± 0.10) ± 0.16) 3 m (1.77 (1.99 ± 0.10) 4 m c (2.00 ± 0.12) ± 0.08) 5 m (3.52 (3.56 ± 0.04) weighted mean − 2 2+ 2– Mg Mg(ox) U + 2 ox 2 (4.00 ± 0.10) [2001CHO/BON] 0.3 m NaCl 25 (3.71 0.10) 1 m ± ± 0.12) 2 m (3.73 (3.79 ± 0.18) 3 m 0.16) ± (4.07 4 m d ± 0.12) (5.17 ± 0.08) 5 m (3.99 2+ 2– + ox U Ca(ox)(aq) Ca e 0.20) ± 0.20) (3.06 ± 1 M NaClO 25 (1.64 [67HAS/MAK] 4 e 25 (2.35 ± 0.10) (3.21 ± 0.10) 0.1 M (K,H)NO [70STE/PAZ] 3 → 25 (3.19 ± 0.10) (3.19 ± 0.10) [74NAN/GAR] 0 e, f 0.5 M NH 25 (1.83 ± 0.08) (3.18 ± 0.13) NO [76MCD/KEL] 4 3 (3.19 ± 0.06) weighted mean 0.03 M Et NI 37 (2.72 ± 0.10) [82DAN/MAR] 4 0.05 (2.61 0.10) ± ± 0.10) 0.10 (2.46 0.30 (2.23 ± 0.10) g 0.50 (2.22 0.12) (3.27 ± 0.07) ± (Continued on next page)

225 VI.5 Magnesium and calcium oxalate compounds and complexes 183 Table VI-18 (continued) ο ( Ionic medium log K t log K ° Reference C) 10 10 2 − 2– U Ca(ox)(aq) + ox Ca(ox) 2 e, h ± 0.25) (0.95 ± 1 M NaClO 0.29) 25 (1.04 [67HAS/MAK] 4 e, f, h 0.5 M NH 25 (0.86 ± 0.22) (0.74 ± 0.25) NO [76MCD/KEL] 4 3 ± 0.19) weighted mean (0.83 ε ≈ ε (Mg(ox)(aq),KCl) = a: Extrapolated to zero ionic strength assuming that (Mg(ox)(aq),NaCl) 0.1). ± ) = ε (Mg(ox)(aq),KNO ) = (0.0 ε (Mg(ox)(aq),NaNO 3 3 b: Weighted least squares SIT-regr ession according to Figure VI-14. ssion according to Figure VI-15. c: Weighted least squares SIT-regre ession according toFigure VI-16. d: Weighted least squares SIT-regr ) = e: Extrapolated to zero ionic strength assuming that ε (Ca(ox)(aq),NaClO ε (Mg(ox)(aq),NaCl) ≈ 4 ε (Ca(ox)(aq),KNO 0.1). ε ) = NO (Ca(ox)(aq),NH ) = (0.0 ± 3 4 3 + 2– + 2– ± ) ≈ ε (ox , K , 0.20). ) = (0.07 NH ε f: Extrapolated to zero ionic strength assuming that (ox 4 g: Weighted least squares SIT-regr ession according to Figure VI-17. 2 − − 2 + + ) ≈ ε ( , Na Mg(ox) Ca(ox) ) = , Na ε ( h: Extrapolated to zero ionic strength assuming that 2 2 2 − + , ) = – (0.15 0.10). NH Ca(ox) ± ( ε 4 2 A weighted least squares SIT regression plot of the data reported in [2001CHO/BON] for the reaction 2– 2+ U + ox (VI.17) Mg(ox)(aq) Mg t for Mg(ox)(aq) actually is zero (Figure reveals that the SIT interaction coefficien VI-14). Hence, it has been assumed that (Mg(ox)(aq),NaCl) ε (Mg(ox)(aq),KCl) = ε ≈ 2+ – ) = ε (Mg(ox)(aq),KNO ) = ) = (0.0 ε 0.1), and using ε (Mg (Mg(ox)(aq),NaNO ,Cl ± 3 3 − + 2– 2+ NO (0.19 Appendix B), and , ε ) = (0.17 ± 0.01) ( cf. (Mg ε (ox 0.02) and ,Na ± ) = 3 + 2– − ± (ox (0.08 ,K ε ) = (0.07 ± 0.08) selected in this review, the values of 0.01) and [38CAN/KIB] , [39PED] , [57LEF] , [57SCH/AND] and [60RAA] were extrapolated to zero ionic strength (Table VI-18).

226 VI Discussion of data selection for oxalate 184 Figure VI-14: Weighted least squares SIT -regression plot of equilibrium data from according to Reaction (VI.17) in for the formation of Mg(ox)(aq) [2001CHO/BON] ο NaCl at 25 (Mg(ox)(aq),NaCl) ε log ° C. The results are (VI.17) = (3.52 ± 0.08) and K 1 10 = (0.00 ± 0.03). 4.0 m I 3.9 )) + 3.8 ,Na 2 3.7 (ox ε )+ 3.6 ,Cl 2+ 3.5 (Mg 3.4 ε ( − 3.3 D 3.2 + 8 K 3.1 10 log 3.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 / molal I m

227 VI.5 Magnesium and calcium oxalate compounds and complexes 185 Figure VI-15: Weighted least squares SIT–regression plot of equilibrium data from [82DAN/MAR] according to Reaction (VI.17) in Et NI for the formation of Mg(ox)(aq) 4 ο log K at 37 ± ° C. The results are (VI.17) = (3.58 0.22). ± ∆ε = – (0.66 0.07) and 1 10 4.2 4.1 4.0 3.9 D 3.8 + 8 3.7 K 10 3.6 log 3.5 3.4 3.3 3.2 0.6 0.5 0.4 0.3 0.2 0.1 0.0 I / molal m be detected in Table VI-18 for the No systematic temperature effect can Mg(ox)(aq) values extrapolated to zero ionic strength and hence, a weighted mean is calculated from all data leading to the selected value: ο log K (VI.17) = (3.56 ± 0.04). 1 10 2– ο (ox ∆ G , 298.15 K) selected in this re- With NEA-TDB selected data and fm view (Section VI.2.3), this selection yields: ο –1 ∆ G 2.28) kJ·mol ± . (Mg(ox), aq, 298.15 K) = – (1155.83 fm

228 VI Discussion of data selection for oxalate 186 A weighted least squares SIT regression plot of the data reported in [2001CHO/BON] for the reaction: − 2 2– 2+ Mg(ox) U + 2 ox (VI.18) Mg 2 gives the value selected in this review: ο log β (VI.18) = (5.17 ± 0.08). 10 2 ο 2– G ∆ , 298.15 K) selected in this re- (ox With NEA-TDB selected data and fm view (Section VI.2.3), this selection yields: ο 2 –1 − G 3.92) kJ·mol Mg(ox) ∆ , 298.15 K) = – (1845.15 ± ( 2 fm -regression plot of equilibrium data from Figure VI-16: Weighted least squares SIT − 2 [2001CHO/BON] Mg(ox) for the formation of according to Reaction (VI.18) in NaCl 2 + ο 2 − at 25 log ° β (VI.18) = (5.17 ± 0.08) and ε ( C. The results are ,Na ) = Mg(ox) 2 10 2 − (0.15 ± 0.03). 6.3 m I )) + 6.1 ,Na 2 5.9 (ox ε 5.7 )+2 ,Cl 5.5 2+ (Mg ε 5.3 ( − D 5.1 + 8 2 4.9 β 10 log 4.7 4.0 1.0 0.0 6.0 3.0 5.0 2.0 / molal I m The experimental determination of the formation of calcium oxalate complexes O (see Section VI.5.1), and as a is severely hampered by the low solubility of Ca(ox)·H 2 reported in the literature concerning the consequence of these difficulties the values formation constant of Ca(ox)(aq) vary by three orders of magnitude (Table VI-17).

229 VI.5 Magnesium and calcium oxalate compounds and complexes 187 be accepted; they we re obtained by ra- In this review only a few results could diochemical measurements of trace concen trations of Ca in solvent extraction [67HAS/MAK] [76MCD/KEL] and electromigration studies [70STE/PAZ] , by using a , Ca selective electrode at sufficiently low calcium and oxalate concentrations [82DAN/MAR] , or by deriving a complexation constant from conductivity measure- [74NAN/GAR] ments . All other reported va lues were rejected [32MON/DAV] , [68GOR/FIL] [71PAZ/STE] , [72ARM/DUN] , [73FIN/SMI] , [75PRA/JON] , , [79CRA/MOO] [82RAO/AGA] , , [93GLA/MAJ] . For a detailed discus- [85RED/RAO] , sion of all papers see Appendix A. + otonated species Ca(Hox) The existence of the pr (aq) has been and Ca(Hox) 2 postulated in a single study [73ARM/DUN] . However, as discussed in Appendix A, the experiments of [73ARM/DUN] are inconclusive and have been rejected by this review. Some papers referenced in databases or other papers are erroneous citations, such as [58GEL/HAY] reporting data for oxaloacetic acid, not oxalic acid, and [85KIL] dealing with cadmium oxalate, and not calcium oxalate. Figure VI-17: Weighted least squares SIT–regression plot of equilibrium data from [82DAN/MAR] according to Reaction (VI.19) in Et NI for the formation of Ca(ox)(aq) 4 at 37 C. The results are 0.25). ± log K (VI.19) = (3.27 ± 0.07) and ° = – (0.66 ∆ε 1 10 4.0 3.9 3.8 3.7 D 3.6 + 8 3.5 K 10 3.4 log 3.3 3.2 3.1 3.0 0.4 0.5 0.6 0.0 0.1 0.2 0.3 / molal I m

230 VI Discussion of data selection for oxalate 188 For the reaction: 2– 2+ U Ca(ox)(aq) (VI.19) + ox Ca (Mg(ox)(aq),NaCl) ε (Ca(ox)(aq),NaClO ) = ≈ ε it has been assumed in this review that 4 2+ − ε (Ca ε NO (Ca(ox)(aq),NH ) = (0.0 ± 0.1), and using ε ) = ) = , (Ca(ox)(aq),KNO ClO 3 4 3 4 2+ − 2– + (0.27 0.03) and NO ε (Ca ± 0.01) ( cf. Appendix B), and ε (ox ± ,Na , ) = ) = (0.02 3 + 2– 0.01) and ) = (0.07 (ox ± ,K (0.08 − ± 0.08) selected in this review, as well as assum- ε + 2– + 2– ing N ε (ox ,K , ε ) ≈ H) = (0.07 ± 0.20), the values of [67HAS/MAK] , (ox 4 [70STE/PAZ] and [76MCD/KEL] were extrapolated to zero ionic strength (Table VI-18). A weighted mean is calculated from all data at 25 C leading to the selected ° value: ο log K (VI.19) = (3.19 ± 0.06). 10 1 This selection yields: ο –1 . ∆ (Ca(ox),aq, 298.15 K) = – (1251.1 ± 2.1) kJ·mol G fm The value obtained from the SIT regression analysis of equilibrium data from ο log K [82DAN/MAR] 0.07) seems to indicate a ± NI at 37 ° C, (VI.19) = (3.27 in Et 4 10 1 non-zero temperature effect fo r this reaction. However, c onsidering that the difference of 0.08 log ° -units is in the same order of magnitude as the uncertainties of the 25 C and 10 37 C values, no enthalpy of Ca ° oxalate complex formation is calculated in this review. For the reaction: 2– − 2 Ca(ox) (VI.20) U Ca(ox)(aq) + ox 2 , [76MCD/KEL] could be identified by this only two reliable studies [67HAS/MAK] review. It has been assumed that ≈ ε (Ca(ox)(aq),NaClO ) = ε (Mg(ox)(aq),NaCl) 4 2 + − + 2 − 0.1), and ε ( (Ca(ox)(aq),NH ε Mg(ox) NO ,Na ) = ) ≈ ε ( ) = (0.0 ,Na Ca(ox) ± 3 4 2 2 + − 2 2– + Ca(ox) NH 0.01), as well as as- ± ε ) = – (0.15 ± 0.10), and ε (ox ( ,Na , ) = – (0.08 4 2 + 2– + 2– suming ) ≈ ε (ox ,K ε (ox , and N H ) = (0.07 ± 0.20), the values of [67HAS/MAK] , 4 [76MCD/KEL] were extrapolated to zero ionic st rength (Table VI-18). A weighted mean is calculated leading to the selected value: ο K (VI.20)= (0.83 ± 0.19). log 10 2 This selection yields: ο 2 − –1 ∆ Ca(ox) G ( . , 298.15 K) = – (1936.0 ± 3.0) kJ·mol 2 fm

231 VI.5 Magnesium and calcium oxalate compounds and complexes 189 VI.5.3 Selected values for Mg and Ca oxalate compounds and complexes The values for magnesium and calcium compounds and complexes selected in this re- view are summarised in Table VI-19 (see also Table III-1 and III-2 in Chapter III). Table VI-19: Selected values for Mg and Ca oxalate compounds and complexes. ο ο –1 log ∆ (kJ·mol G K ) 10 fm Mg Ca Mg Ca ± 2.3) M(ox), aq ± 2.1) – (1155.8 – (1251.1 − 2 – (1845.2 ± 3.9) – (1936.0 ± 3.0) Mox 2 2+ 2– ± + ox Mox(aq) (3.56 U 0.04) (3.19 ± 0.06) M 2 − 2+ 2– U + 2 ox Mox ± (5.17 ± 0.08) (4.02 0.19) M 2 Ca(ox) H ⋅ O(cr), – (1519.9 ± 2.1) 2 Ca(ox) ⋅ 2H O(cr) – (1754.6 ± 2.1) 2 Ca(ox) 3H ⋅ O(cr) ± 2.1) – (1991.1 2 2+ 2– U Ca 0.06) + ox ± + H O(cr) O(l) – (8.73 Ca(ox)·H 2 2 2– 2+ Ca 0.06) + ox ± + 2 H U O(l) – (8.30 O(cr) Ca(ox)·2H 2 2 2– 2+ O(l) – (8.19 O(cr) + ox Ca + 3 H 0.04) ± U Ca(ox)·3H 2 2 Selenium oxalate compounds and complexes VI.6 Solid selenium oxalates VI.6.1 SeOox was prepared from SeOCl and Ag ox (1:1.1) in tetrahydrofuran and dioxane 2 2 [66PAE] . No reaction was found with Na ox. The compound crystallises as 2 SeOox·tetrahydrofuran or SeOox·2dioxane. There are two studies describing the properties of mixed uranium(VI) selenate oxalates in the solid state, (NH and ) [96MIK/GOR2] [UO O (SeO )ox]·1.5H 4 2 2 4 2 Cs . [UO [2000MIK/GOR2] (SeO )ox] 4 2 2 No thermodynamic data are available for any of these compounds.

232 VI Discussion of data selection for oxalate 190 VI.6.2 Aqueous selenium oxalate complexes [98OBR/MIT] report pH metric and conductivity measurements in the Obradovi et al. ć system H − Na , and the authors state that their SeO experimental results indicate the ox 3 2 2 2– 2 − formation of complexes in which the ox SeO : ratio is 2:1. However, a series of 3 cf. Appendix A) reveals that mo st of the effect reported simple speciation calculations ( in the figure of this paper is not due to oxalate - selenite complex formation but can be inferred from the acid-base titration data reported by [98OBR/MIT] . The slight devia- alysed because the original data are not tions from the calculated curve cannot be an available. No further experimental studies involving the thermodynamic properties of se- lenium oxalates in solution could be identified by this review . 1 Nickel oxalate compounds and complexes VI.7 Solid nickel oxalates VI.7.1 ns is nickel oxalate dihydrate. Two crys- The solid that precipitates from aqueous solutio tallographic forms of Ni(ox)·2H α -form and an O have been reported: a monoclinic 2 orthorhombic β -form . The structure of the anhydrous nickel(II) oxalate, [73DEY/BER] Ni(ox)(cr), has been described in [82NIK/SHA] . This solid may be obtained by heating the dihydrate. The equilibrium water vapour pressure for the dissociation of Ni(ox)·2H . The thermal de- O into Ni(ox)(s) at 125-177 ° C is given in [53ALL/SCA] 2 composition of nickel oxalate has been investigated extensively, see for example [99MAJ/SAR] , [2000LVO] and references cited therein. O is meta-stable and it precipita -Ni(ox)·2H β tes initially when solutions of 2 nickel(II) and oxalic acid (or oxalate) are mixed [70DEY/PEN] [73DEY/BER] . This , solid transforms into α -Ni(ox)·2H O when standing in a boiling solution with excess of 2 oxalate ions [73DEY/BER] , although the transformation is inhibited by sulphate ions [70DEY/PEN] , [75DEY/CAN] . In the presence of potassium oxalate, a solid solution is apparently formed [09DEA/SCO] , [36VOS/ISR] between K O. In solutions containing ox and Ni(ox)·2H 2 2 > 12.5 weight-% potassium oxalate the solid K ⋅ (H O) Ni(ox) 4H O is formed 2 2 2 2 2 [36VOS/ISR] . The structure of this solid has been analysed by Raman spectroscopy [91BIC/EDW] and by single-crystal X-ray diffraction [93ROM/GUZ] . The solid Na [11DOD] Ni(ox) . ·8H O has also been described 2 2 2 1 Prof. J. Havel and Dr. P. Lubal (both from the Department of Analytical Chemistry, Masaryk University, Brno, Czech Republic) contributed to this section during the early stages of the review process.

233 VI.7 Nickel oxalate compounds and complexes 191 2H ⋅ The solubility of Ni(ox) O may be used to derive the solubility product of 2 the solid, and such data has been reported in several papers. Literature data is available [1894NAS] for the solubilities in water [51BAR/ARG] , [27SCH3] , , , [37LED/HAU] [89FEL/TOP] SO (Table VI-21) (Table VI-20), in H , [75DEY/CAN] [70DEY/PEN] 4 2 and in K ox media . The crystallographic form of the solid studied is not [51BAR/ARG] 2 reported in any of the references published before 1970. As discussed in the following, none of the solubility studies in the literature is suitable to recommend a solubility product. [27SCH3] Scholder observed very slow kinetics for the equilibration, and they had to wait for 100 hours, and therefore it is not surprising that Ledrut and Hauss [37LED/HAU] obtained higher solubilities from their measurements only after 6 hours in initially oversaturated solutions. Table VI-20: Reported values for the solubility of nickel oxalate in water. ( ° C) Solubility Comments Reference Solid phase t 2 − 10 M See Appendix A [1894NAS] × Ni(ox) ? 1.8 − 5 × 10 O 18 2.0 Ni(ox)·2H M [27SCH3] . From under- Titration of oxalate with KMnO 4 2 4.69 − saturation after 100 hours. See Appendix A. (10 M) − 5 O 20 27.7 × 10 [37LED/HAU] M . From over- Ni(ox)·2H Titration of oxalate with KMnO 4 2 − 3.56 (10 saturation after 6 hours. M) 5 − [51BAR/ARG] M 10 × Ni(ox) 25 7.15 Ni: spectrophotometric with dimethylglyoxime. − 4.15 (10 M) From undersaturation after 96 hours. See also Appendix A. − 5 Ni(ox)·2H 10 O 20 1.6 × M Not clear how it was measured, or even if it was [89FEL/TOP] 2 − 4.79 (10 measured. M) Solubility data for the two crystallographic forms, and O in , of Ni(ox) ⋅ 2H β α 2 H SO solutions were used to determine the solubility products in [70DEY/PEN] , 2 4 [75DEY/CAN] , and the corresponding enthalpy and entropy changes were obtained [75DEY/CAN] from their temperature variation in . These values, summarised on Table VI-21, were obtained by Deyrieux et al . assuming the formation of only the following two complexes in solution: Ni(ox)(aq) and NiH (ox)SO (aq). However, as discussed in 4 2 Appendix A, these solubility data are not well suited to unequivocally describe the stoichiometry of the protonated complexes, and the values listed in Table VI-21 are considered unreliable.

234 VI Discussion of data selection for oxalate 192 Table VI-21: Literature values for the solubility products and thermodynamic data for − 2 2+ the reaction: Ni(ox)·2H + ox + 2H O(l). Ni O U 2 2 ⋅ ⋅ -Ni(ox) O α -Ni(ox) 2H 2H O Reference β 2 2 ± 0.01) − (9.65 [70DEY/PEN] ο K log ,0 10 s [75DEY/CAN] (9.43 (9.68 ± 0.10) 0.14) ± − − ο − 1 ± (18.3 [75DEY/CAN] ± ) (22.2 0.1) 0.2) mol (kJ ⋅ H ∆ rm ο − 1 − 1 ) (J mol ⋅ K S − ∆ ± 3) − (124 ± 2) [75DEY/CAN] (106 rm has serious prob- ox in [51BAR/ARG] The solubility data as a function of K 2 lid phase properly, and the solubility curve lems: the authors did not characterise the so differs substantially with the shape theoretically expected, especially at low values of [K ox] . added 2 In aqueous solutions in equilibrium with nickel oxalate dihydrate, the follow- ing reaction takes place: ⋅ O(l) Ni(ox)(aq) + 2 H O” U 2H “Ni(ox) 2 2 and the concentration of Ni(ox)(aq) is almost constant for any medium, as its activity coefficient will be close to unity. The concentration of Ni(ox)(aq) corresponds to the minimum possible solubility: any other Ni-species formed will increase the solubility above this concentration. The minimum reported solubility at 25 ° C is [Ni] = TOT 4.77 − 4 − 10 [51BAR/ARG] M in [K 10 ox] = 5.6 M × . If it is assumed that this solubil- 2 TOT ity corresponds to the concentration of Ni(ox )(aq), and this is combined with the value − 2 2+ ο of K ± 0.04) selected by this review for Ni log + ox = (5.19 Ni(ox)(aq), then U 10 1 ο ≤ log K the maximum value for the solubility product may be estimated to be s 10 ,0 ( 4.77 − 5.19) = − − 9.96. Figure VI-18 shows a comparison between the experimental data from [51BAR/ARG] and [70DEY/PEN] for the solubility of nickel oxalate and curves calculated using this solubility product and the selected values for Ni(ox)(aq) − 2 Ni(ox) and . The figure shows that the two data sets are discrepant: the value at 2 [K ox] = 0 from [51BAR/ARG] seems to agree with the general trend in the data 2 added from [70DEY/PEN] , while the other data from [51BAR/ARG] appear to be too low, perhaps because of the formation of a solid solution (see discussion in Appendix A). Although the data from [70DEY/PEN] are insufficient to estab- are more reliable, they lish the protonated species formed in the acid H SO solutions, and therefore no value is 2 4 recommended for the solubility product for nickel oxalate. Calorimetric measurements of the heat of dissolution of “Ni(ox)” in hydrochlo- ric acid [69VAN/PER] or in water [71VAN] have been used to calculate a value of − 1 ο (“Ni(ox)”, 298.15 K) = − (854 ± 19) kJ·mol − ± [69VAN/PER] and (882 1) H ∆ fm − 1 kJ·mol [71VAN] . These two references contain however almost no details on the experimental and on the data treatment procedures. The enthalpy of formation of Ni(ox) is also reported in [56KOR/PET] but this reference is not available to this review.

235 VI.7 Nickel oxalate compounds and complexes 193 1 − ο ( α -Ni(ox)·2H [75DEY/CAN] O, 298.15 K) = − reports ± 0.1) kJ·mol (1478.3 ∆ H 2 fm 1 − ο and ( β -Ni(ox)·2H 0.2) kJ·mol O, 298.15 K) = − (1482.1 ± calculated from the ∆ H 2 fm temperature dependence of the solubility products of these solids in H SO . However, as 2 4 mentioned above, these solubility products ar e considered unreliable by this review, see also the discussion of [75DEY/CAN] in Appendix A. In summary, no reliable value for the enthalpy of formation of any nickel oxalate is available in the literature. Figure VI-18: Comparison of experimental and calculated values for the solubility of C in: (a) potassium oxalate; and (b) sulphuric aci d solutions. The ° nickel oxalate at 25 lected values for th e complex formation curves have been calculated using the se between Ni(II) and oxalate, and with the following two constants: 2+ 2 − ο U Ni − + ox β -Ni(ox)·2H + 2 H 9.96 O(l) = log O K 2 2 10 s 2 − 2+ ο + + NiHSO U + H = 5.46 SO + K log Ni 4 10 4 −3 −3 10 10 (a) (b) / M −4 −4 TOT 10 10 [70DEY/PEN] [51BAR/ARG] [Ni(II)] −5 −5 10 10 0.000 0.002 0.004 0.004 0.000 0.002 [K ox] / M [H / M SO ] added 2 2 4 added

236 VI Discussion of data selection for oxalate 194 Aqueous nickel oxalate complexes VI.7.2 2+ with The experimental equilibrium data available on the complex formation of Ni 2+ oxalate are listed in Table VI-22. Studies on the complex formation between Ni and − 2 ox cf . Section VI.7.1, have the difficulty of the rather low solubility of nickel oxalate, especially if there is no excess of oxalate present in the solutions. Table VI-22: The stability constants of Ni(II) oxalate complexes reported in literature ° C) log K Reference ( Method Ionic medium t 10 2+ 2 − + ox U Ni(ox)(aq) Ni [32MON/DAV] con 18 > 5.3 0.10) ) 25 (4.10 ± gl, sp [60WAT/DEW] 1 M (KNO 3 , 0 (5.179 ± 0.014) [61MCA/NAN] → 0 M NiCl pot 2 ox 15 (5.137 ± 0.013) (Na,H) 2 ± 0.009) 25 (5.158 0.017) ± 35 (5.173 45 (5.185 ± 0.014) ) 20 (3.83 ± 0.05) [69MAN/BHA] 0.1 M (NaClO dis 4 SO ± 0.03) [70DEY/PEN] 25 (5.39 0 M H → sol 4 2 ± SO 25 (4.8 0.3) [75DEY/CAN] → 0 M H sol 2 4 1.0 M Na(ClO dis ) 25 3.7 [76MUR/KUR] 4 ) 20 4.255 [81URB/BIE2] pol 0.1 M (NaClO 4 ) 35 5.23 [85RED/RAO] pot 0.1 M (KNO 3 pol 0.1 M (NaClO [88CRO] ) 25 4.16 4 (4.05 pot 0.01) [89FUE/REB2] 0.5 M (NaCl) 25 ± [90RED/SAT] 0.05) ) 35 (4.43 ± 0.1 M (KNO pot 3 0.11 M (NaClO cix ± 0.05) [91LOO/KOP] ) 25 (4.27 4 ± ) 25 (4.432 0.03) [93AZA/HAS] pot 0.1 M (KNO 3 0.69) ) 22 (4.69 ± [93JAN] chr 0.1 M (NaClO 4 pot 0.1 M (KNO ) 30 5.28 [94RED/SHI] 3 ) 25 (3.46 ± 0.08) [98KHA/RAD] pot 0.1 M (NaNO 3 [2003BAE/BRA] ) 25 (4.4 ± 0.1) cix 0.1 M (NaClO 4 0.5 M (NaClO ) (3.9 ± 0.1) 4 20 (3.98 ± 0.05) [2003BOR/CHO] dis 0.3 m (NaCl) 1 m ± 0.04) (3.72 2 m (3.71 ± 0.05) 3 m (3.76 ± 0.07) 4 m ± 0.07) (3.87 5 m (3.97 ± 0.04) (Continued on next page)

237 VI.7 Nickel oxalate compounds and complexes 195 Table VI-22: (continued) Method Ionic medium ° C) log t K Reference ( 10 2 − 2+ − 2 + 2 ox Ni(ox) U Ni 2 6.2 M LiCl ? 13.7 [34SAR] pol 0.012) 25 6.51 [51BAR/ARG] sol var. ( ≤ 2 − 25 7.2 [54YAT/ZOL] (0.25 – 0.8) M ox cal ) 25 (7.15 0.14) [60WAT/DEW] ± 1 M (KNO gl, sp 3 [63FRI/VER] ? ? 25? 7.64 ) 20 (7.88 ± 0.04) [63STA] dis 0.1 M (NaClO 4 0.13) ) 20 (7.06 ± [69MAN/BHA] dis 0.1 M (NaClO 4 1.0 M Na(ClO dis [76MUR/KUR] ) 25 6.6 4 6.5 [76YAD/GHO] 30 em 0.1 M (Na,H)ClO 4 ) 20 7.544 [81URB/BIE2] 0.1 M (NaClO pol 4 pol 0.1 M (NaClO ) 25 7.58 [88CRO] 4 25 (6.01 ± 0.03) [89FUE/REB2] pot 0.5 M (NaCl) 20 (6.70 0.05) [2003BOR/CHO] dis 0.3 m (NaCl) ± (6.49 0.05) 1 m ± 2 m (6.44 ± 0.05) (6.51 ± 3 m 0.05) 4 m ± 0.06) (6.72 (6.95 ± 0.06) 5 m − 2 − 2 U Ni(ox)(aq) + ox Ni(ox) 2 0.25 M (NaClO ) 25 (1.8 ± 0.2) [69SUB/COR] dis 4 [92TRI/TRI] 30 2.2 em 0.1 M (Na,H)ClO 4 pot 0.1 M (NaNO ± 0.05) ) 25 (2.96 [98KHA/RAD] 3 − 4 2+ − 2 Ni + 3 ox Ni(ox) U 3 gl, sp ) 25 (8.51 ± 0.17) [60WAT/DEW] 1 M (KNO 3 1 M (KNO ) 25 7.2 pot [68FRI/VER] 3 2+ − + Ni + Hox U Ni(Hox) em 0.1 M (Na,H)ClO 30 2.3 [76YAD/GHO] 4 em 0.1 M (Na,H)ClO 30 2.3 [92TRI/TRI] 4 − − 2+ )(aq) + Hox HSO U + Ni(Hox)(HSO Ni 4 4 25 (6.183 0 M H SO sol → ± 0.001) [70DEY/PEN] 2 4 sol → 0 M H SO [75DEY/CAN] 25 (6.11 ± 0.21) 4 2 Most of the papers deal with the determination of stability constants of com- 2 n − Ni(ox) plex species with n =1 and 2, according to: n − 2+ n 2 − 2 N ox U n i(ox) + (VI.21) Ni n

238 VI Discussion of data selection for oxalate 196 n = 3 was reported only in two papers: The formation of complex species with [60WAT/DEW] [68FRI/VER] . The reason is that such a and complex is formed only at high oxalate concentrations, perhaps at [ox] > 0.1 M. As discussed in Appendix A TOT these two papers are not recommended because the equilibrium constants were obtained t (ethylenediamine or glycine), in experi- from mixtures containing a fourth componen ments where the ionic medium was not constant. + The protonated complex Ni(Hox) has been postulated from electromigration [76YAD/GHO] data, , [92TRI/TRI] , but these studies are considered unreliable, see Appendix A. Another protonated species that has been proposed in sulphuric acid solu- tions is Ni(Hox)(HSO )(aq), [70DEY/PEN] , [75DEY/CAN] , but it is possible to fit 4 these solubility measurements equally well with several other protonated complexes, see Appendix A. Even if a rather high number of refe rences with data for the formation of 2 − Ni(ox)(aq) and Ni(ox) has been found (Table VI-22), only a limited number of stud- 2 ies, listed in Table VI-23, are accepted in this review. All references containing data not considered [32MON/DAV] , [51BAR/ARG] , [54YAT/ZOL] , [34SAR] , [60WAT/DEW] , [63STA] , [63FRI/VER] , [69MAN/BHA] , , [68FRI/VER] [69SUB/COR] [70DEY/PEN] , [75DEY/CAN] , [76YAD/GHO] , [81URB/BIE2] , , [85RED/RAO] , [90RED/SAT] , [92TRI/TRI] , [93JAN] , [94RED/SHI] , [88CRO] , [98KHA/RAD] are discussed in Appendix A. The interpretation of the selected litera- ture values listed in Table VI-23 according to the SIT mode l must take into account the different ionic media and temperatures. The data at 20 ° C from [2003BOR/CHO] were corrected to 25 C using the equations given in Section V.3.6 and the selected ° ο H ∆ However, the interpretation of the calo- (VI.21) values for complex formation. rm rimetric data requires selected complex formation constants to obtain the enthalpy changes (see below). This interdependency was resolved in an iterative manner. The ο H ∆ ionic medium dependence of (VI.21) is not known, and Equation (V.58) was used rm 1 1 − − assuming 0.010) kg·K ± ·mol = (0 . The temperature corrections to the data in ∆ε L [2003BOR/CHO] were small, between − 0.011 and − 0.018 for ; and between K log 1 10 β log − 0.042 for − 0.034 and . The uncertainties in the equilibrium constants were 2 10 also increased to reflect the uncertainty in the temperature corr ection, which arises mainly from the uncertainty in ∆ε . The corrected values are listed in Table VI-23. L The data in Table VI-23, converted to molal units, was treated using a edure with an expression similar to (V.48): weighted linear least-squares regression proc 2 ο – + 2– * 2+ D – ( ε (Ni log , X ∆ ) + n ε(Μ K , ox (VI.21) – ) I I = ε log ∆ K z (VI.21) – m m n n n 10 10 2 z = 1 and 2. As the first complex is not electri- = − 8 for both n In this case ∆ * cally charged, the specific ion interaction parameter is expected to be small: ε ∆ = 1 − 2 * + 2 − : because only data Ni(ox) ε( Ni(ox)(aq), MX ) ) ε ∆ = ε( Na , ≈ 0; while for Ni(ox) 2 2 2 2+ – + 2– in sodium electrolytes are available. The values of ε( Ni and ε( M ) , ox , Cl ) listed in Tables B-4 and B-5 were used for the SIT evaluation. Given the spread of the data, a

239 VI.7 Nickel oxalate compounds and complexes 197 0.1 log minimum uncertainty of -units was assigned to the data sets, cf . Table VI-23. ± 10 The results of the regressions are displayed in Figure VI-19 and Figure VI-20 and the selected stability constants are: ο K n = 1) = (5.19 ± 0.04) log ((VI.21), 1 10 ο log β ((VI.21), n = 2) = (7.64 ± 0.07) 2 10 Ni(ox)(aq), MX ) = ε( with the following specific ion interaction coefficients: − 1 + 2 − − 1 i(ox) , 0.03) kg·mol , and N ± ε( ) = − (0.26 ± 0.03) kg·mol Na (0.07 . − 2 ° C with Table VI-23: Accepted stability constants of nickel(II) oxalate complexes at 25 the uncertainties assigned in this review. ( Method Ionic medium ° log K Reference t C) 10 2+ 2 − U Ni(ox)(aq) Ni + ox a 25 (5.158 ± 0.10) [61MCA/NAN] pot 0 → 1.0 M Na(ClO dis ) 25 (3.7 ± 0.2) [76MUR/KUR] 4 0.5 M (NaCl) ± 0.10) [89FUE/REB2] (4.05 pot 25 a 0.11 M (NaClO cix 0.10) [91LOO/KOP] ) 25 (4.27 ± 4 pot 0.1 M (KNO ) 25 (4.432 ± 0.10) [93AZA/HAS] 3 cix 0.1 M (NaClO ) 25 (4.4 ± [2003BAE/BRA] 0.1) 4 0.5 M (NaClO ± 0.1) ) (3.9 4 b → ± 0.10) [2003BOR/CHO] dis 0.3 m (NaCl) 25 (3.969 (3.706 ± 1 m 0.10) 2 m ± 0.10) (3.694 (3.743 0.10) 3 m ± (3.852 ± 4 m 0.11) 5 m (3.952 ± 0.12) − 2 2 2+ − Ni + 2 ox Ni(ox) U 2 dis 1.0 M Na(ClO ) 25 (6.6 ± 0.2) [76MUR/KUR] 4 0.5 M (NaCl) 25 (6.01 ± 0.10) [89FUE/REB2] pot b 25 (6.666 ± [2003BOR/CHO] → dis 0.3 m (NaCl) 0.10) 1 m ± 0.10) (6.452 (6.401 ± 0.10) 2 m 3 m (6.470 ± 0.10) 4 m ± 0.11) (6.679 5 m (6.908 ± 0.13) a: see comments in Appendix A. were corrected to 25 ° C as described in the text. b: the data from [2003BOR/CHO]

240 VI Discussion of data selection for oxalate 198 Figure VI-19: Weighted least squares SIT-regression plot of equilibrium data for the formation of Ni(ox)(aq) according to Reaction (VI.21) with n = 1. 6.0 2+ − 2 Ni + ox Ni(ox)(aq) U m I )) − 2 5.5 ,ox + (M ε + ) − ,X 5.0 2+ (Ni ε [61MCA/NAN] ( [76MUR/KUR] − D [89FUE/REB2] 4.5 8 [91LOO/KOP] + 1 [93AZA/HAS] K [2003BAE/BRA] 10 [2003BOR/CHO] log 4.0 012345 I / molal m Figure VI-20: Weighted least squares SIT-regression plot of equilibrium data for the − 2 = 2. Ni(ox) according to Reaction (VI.21) with n formation of 2 m − − 2 2 2+ I 9.0 U ox Ni(ox) Ni + 2 )) 2 − 2 ,ox + (M ε 8.5 2 + ) − ,X 2+ (Ni 8.0 ε ( − D 8 [76MUR/KUR] + 2 [89FUE/REB2] β 7.5 [2003BOR/CHO] 10 log 012345 I / molal m

241 VI.7 Nickel oxalate compounds and complexes 199 Temperature effects VI.7.2.1 Enthalpy changes for the complex formation between Ni(II) and oxalate have been ed values are listed in Table VI-24. determined in four studies. The report Table VI-24: Literature values for the enthal ith the formation of py changes associated w nickel(II) oxalate complexes at 25 C. ° 1 − ) Reference Method Ionic medium (kJ ⋅ mol H ∆ rm 2 − 2+ U Ni(ox)(aq) + ox Ni 0 M (0.6 ± pot [61MCA/NAN] → 0.4) a SO 0 M (H ) (51.6 ± 0.2) [75DEY/CAN] → sol 2 4 b cal 1 M (KNO (5.4 ± − [90LIN/GU] ) 0.2) 3 − 2 2 2+ − Ni U + 2 ox Ni(ox) 2 − 2 − 5.15 cal (0.25-0.8) M ox [54YAT/ZOL] b ) − cal ± 0.3) [90LIN/GU] 1 M (KNO (11.0 3 4 − 2+ 2 − Ni + 3 ox Ni(ox) U 3 b ) − (17.1 ± 0.6) [90LIN/GU] cal 1 M (KNO 3 − 2+ − + + Hox U Ni(Hox)(HSO )(aq) Ni HSO 4 4 a → 0 M (H SO ) − (40.8 ± 0.1) [75DEY/CAN] sol 4 2 - and β -Ni(ox) ⋅ 2H a: from the temperature variation of the solubility of O in H α SO solutions. 4 2 2 . b: the authors used equilibrium constants from [60WAT/DEW] [54YAT/ZOL] The calorimetric study in was performed in varying ionic me- dia, and the data cannot be used in this review, see Appendix A. enthalpy data were obtained from the temperature variation In [75DEY/CAN] of the solubility of - and β -Ni(ox) ⋅ 2H solutions. However, these values are α SO O in H 2 2 4 (aq) and the speciation glected the form ation of NiSO unreliable because the authors ne 4 in the solutions studied cannot be pinpointed from the type of measurements performed, see also the discussion in Appendix A. The calorimetric data in [90LIN/GU] was interpreted by the authors using the , which are considered unreliable by equilibrium constants reported in [60WAT/DEW] ree calorimetric measurements reported in this review (Appendix A). Therefore, the th [90LIN/GU] are re-evaluated in this review, as explained in Appendix A, using selected = 1) = n ∆ ((VI.21), 0.3) H − (5.0 ± values for the equilibrium constants, obtaining rm 1 − 1 − H ∆ kJ·mol n − (12.6 ± 0.3) kJ·mol ((VI.21), = 2) = at I = 1 M KNO and and 3 rm ° I = 0 using Eq. (V.58) with the same parameters C. These values are extrapolated to 25 2 − ∆ 8 and, lacking additional information, z = = 1 and 2 in Reaction (VI.7.1): for both n ο H ∆ ± ((VI.21), ∆ε 0.3) = 0, resulting in the following values: n = 1) = − (0.2 L rm − 1 ο − 1 and ° H ⋅ ∆ ((VI.21), n = 2) = − (7.8 mol 0.3) kJ ⋅ mol ± C. , both at 25 kJ rm Table VI-25 lists the re-e valuated values from [61MCA/NAN] , [90LIN/GU] . For Ni(ox)(aq) the selected en thalpy of reaction is taken as the weighted average of the

242 VI Discussion of data selection for oxalate 200 [61MCA/NAN] value from extrapolated to zero ionic and that from [90LIN/GU] − 2 only the value from [90LIN/GU] is Ni(ox) strength as indicated above, while for 2 available. The selected values become: ο 1 − n 0.3) kJ ⋅ mol (298.15 K, (VI.21), ∆ ± H = 1) = (0.0 rm ο − 1 n H − (7.8 ± 0.3) kJ ⋅ mol = 2) = ∆ . (298.15 K, (VI.21), rm With these selections, the following formation values and molar entropies are obtained: ο –1 ∆ (Ni(ox), aq, 298.15 K) = – (755.5 ± 2.0) kJ·mol G fm ο –1 ∆ (Ni(ox), aq, 298.15 K) = – (885.7 ± 1.8) kJ·mol H fm ο –1 –1 S (Ni(ox), aq, 298.15 K) = (15.2 ± 3.6) J·K ·mol . m − ο –1 2 ( i(ox) N ∆ G , 298.15 K) = – (1450.0 ± 3.8) kJ·mol 2 fm –1 ο − 2 H ( Ni(ox) ∆ ± 3.3) kJ·mol , 298.15 K) = – (1724.1 2 fm ο 2 − –1 –1 S Ni(ox) , 298.15 K) = (83.5 ( ·mol ± . 6.4) J·K 2 m Table VI-25: Selected literature data for the enthalpy changes associated with the ° C. formation of nickel(II) oxalate complexes at 25 1 − Method Ionic medium ) Reference mol ⋅ (kJ H ∆ rm − 2 2+ Ni U Ni(ox)(aq) + ox a (1.4 ± 0.8) pot → [61MCA/NAN] 0 M a − (5.0 ± 0.3) 1 M (KNO ) [90LIN/GU] cal 3 − 2 2 − 2+ U + 2 ox Ni Ni(ox) 2 a [90LIN/GU] − (12.6 ± 0.3) 1 M (KNO ) cal 3 a: re-evaluated by this review, cf . Appendix A VI.7.2.2 Ternary complexes Only ternary complexes between Ni(II), oxalate and either inorganic ligands, or citrate or edta are included in this review. The formation of ternary complexes between Ni(II), edta and oxalate is discussed in Section VIII.7.2. Equilibrium constants for the forma- )(aq) have been reported in [70DEY/PEN] , [75DEY/CAN] . These tion of Ni(Hox)(HSO 4 (aq) and the se the authors neglected the formation of NiSO values are unreliable becau 4 speciation in the solutions studied cannot be identified from the type of measurements performed, see also the discussion in Appendix A. VI.8 Technetium oxalate compounds and complexes There is very little information in the literature concerning the compounds and aqueous complexes of technetium, and no thermodynamic data has been published.

243 VI.8 Technetium oxalat e compounds and complexes 201 A few solid phases have been synthe sised and characterised, and these are There is no information available on the thermodynamic summarised in Table VI-26. stability of these compounds. Table VI-26: Technetium oxalate compounds reported in the literature. Formula Name Reference Tc(II): H P [92SEI/MUE] ) Tc(C (ox) ⋅ KPF lphosphino)ethane)-Tc(II) ⋅ 1.5H O oxalato-bis(1,2-bis(dipheny 2 6 2 24 26 2 potassium hexafluorophosphate sesquihydrate Tc(IV): H bis(tetraphenylarsonium) tris(oxalato-O,O')-Tc(IV) As) (C Tc(ox) [87COL/WIL] 3 20 24 2 3H μ -O) Tc(ox) ] ⋅ Tc( O tetra-potassium bis(( μ -oxo)-bis(oxalato-O,O')-Tc(IV)) K [90ALB/AND] [(ox) 4 2 2 2 2 2 trihydrate Tc(V): 3H As) TcO(Hox)(ox) (C ⋅ H O bis(tetraphenylarsonium ) (hydrogen oxalato-O)- [88BAL/COL] 24 2 20 2 2 bis(oxalato-O,O')-oxo-Tc(V) trihydrate Tc(VI): [88BAL/COL] -oxalato- As) H [(TcN(ox)) O(ox)] (C tetrakis(tetraphenylarsonium) cyclo-bis(( μ 2 24 4 2 2 20 -oxo)-bis(nitrido -(oxalato-O,O')- μ O,O',O'',O''')-( 2 Tc(VI))) Tc(VII): ) [90BAL/COL] [N(O ) Tc(ox)TcN(O H As) ] bis(tetraphenylarsonium) ( μ -oxalato-O,O',O'',O''')- (C 2 2 24 2 2 2 20 2 [91BAL/COL] bis(nitrido-bisperoxo-Tc(VII)) Some qualitative information about the interaction between technetium and ox- alate in aqueous solutions has also been published. The stable oxidation state of Tc in − VII Tc O contact with air is pertechnetate, , and its reduction appears to be facilitated by 4 coordinate technetium in lowe the presence of ligands that r oxidation states, such as Tc(IV) or perhaps Tc(III). However, in the absence of reducing agents Tc(VII) is rela- . In the presence of nitric tively stable towards reduction by oxalate [2001BER/BAU] and oxalic acids as well as re ducing agents, for ex ample Sn(II) or hydrazine, the techne- tium-oxalate complex formed may be extracted into tributylphosphate (TBP) in do- , [2002SHA] . Although the oxidation state of Tc in the complex(es) has decane [85FRI] not been established, it is known that Tc(IV) is the most stable oxidation state in reduc- ained when reducing Tc(VII) . Spectra of the products obt [99RAR/RAN] ing conditions by Sn(II) in oxalate solutions are given in [85GRA/FAR] . None of these publications and patents report data that may be used to obtain thermodynamic information for the technetium-oxalate system in aqueous solutions.

244 VI Discussion of data selection for oxalate 202 Zirconium oxalate compounds and complexes VI.9 Zirconium oxalate compounds VI.9.1 Since the first report of insoluble zirconium oxalate [1820DUB/SIL] , it was generally that precipitates formed on adding oxalic acid or oxalate salts to observed [58BLU] solutions of zirconium salts, and that the precipitates dissolved and very stable soluble s of oxalate were added. In an excess complexes were formed when excess amount oxalate has been studied in solution con- amount of oxalate, the solubility of zirconium ium to caesium and ammonium) oxalate taining variable amounts of alkali (sod , ). Although the solubilities and the solid salts compositions were [36BOU2] [36BOU] ( reported, the results are not credited in this review due to the lack of experimental de- tails. The compositions, structures and preparative conditions for the insoluble com- pounds formed with relatively small amounts of oxalate are uncertain due to the strong tendency of zirconium to undergo hydrolytic polymerisation even in acidic solutions , [55MOH/SIN] , ). By using two , [58BLU] , [58BOB/BEN] [57ZAI/SHU2] [31GAB] ( [49CON/MCV] phase distribution with TTA, Connick and coworkers ( [51CON/REA] , , 4+ 3+ ) have shown that monomeric Zr can only exist at zirconium or ZrOH [56ZIE/CON] –4 –4 M in 2 M HClO × and lower than 10 10 M in 1 M concentrations lower than 5 4 HClO . Thus, although it was once thought that the simple monomeric “zirconyl” ion, 4 2+ , is present in acidic solutions, inferred from the stoichiometry of basic salts such ZrO O, structural determinations have shown that the solids do not contain ·8H as ZrOCl 2 2 8+ such “zirconyl” ions but tetranuclear [Zr (OH moieties, as shown in Figure (OH) ) ] 16 2 4 8 and the references cited therein). [97RIC] , [76BAE/MES] VI-21 ( Figure VI-21: Structure of the μ –hydroxy bridged tetranuclear cation, 8+ [Zr ·8H (OH , present in solid ZrOCl ) ] O. (OH) 2 2 8 2 4 16 H H H H 2 2 2 2 O O O O OH OH O H 2 2 Zr Zr OH OH H O 2 2 OH OH OH OH OH O H OH 2 2 Zr Zr OH OH O H 2 2 O O O O H H H H 2 2 2 2

245 VI.9 Zirconium oxalate compounds and complexes 203 To obtain information about the degree of polymerisation (change of hydroxide into hydroxo- or oxo-bridged species) in “zirconium hydroxide”, Zaitsev and co- workers ( [62ZAI/BOC3] , [64ZAI3] , [64ZAI2] , [66ZAI2] ) examined by , [65ZAI/BOC] titration in the presence of fluoride the reactivity of variously aged or ignited precipi- and obtained infrared absorption spectra of tates for the displacement of OH by fluoride sult indicated that there were three types of zirconium hydrox- these compounds. The re α β ) precipitate from concentrated “zirconyl chlo- ) freshly prepared precipitate, ( ides ( γ ) precipitate aged by boiling for a long time, each precipitate ride” in methanol, and ( corresponding to the structure shown in Figure VI-22. Figure VI-22: Structures of the va riously aged “zirc onium hydroxides” ) α ( HO OH OH OH HO Zr Zr OH OH OH OH OH OH Zr Zr OH OH HO OH HO γ ( β ) ) ( HO OH OH HO OH HO OH Zr Zr Zr O O Zr HO O O OH OH OH OH Zr Zr O Zr Zr O HO OH HO OH OH HO HO OH [66ZAI2] Zaitsev has also shown that, in the reaction with oxalate, the reactiv- → α -ring is the most β γ → , whereas the β ity of the hydroxide decreases in the order stable, and a small amount of oxalate generally does not alter the tetrameric unit and

246 VI Discussion of data selection for oxalate 204 alate brings about an increase in the hydroxides, and ox only exchange with peripheral β -ring into -ring. This explains why oxalate - or α γ stability of the ring converting the forms precipitates (mixed hydroxy-oxalate complexes) when oxalate is deficient and [64KHA/BOC2] oxalate is in excess ( soluble complexes when [64KHA/ZAI2] ). Based , [67ZAI/SHU2] ) measured the , [64ZAI/BOC2] on this fact, Zaitsev and coworkers ( solubilities of these zirconium hydroxides and estimated the apparent solubility prod- ucts, K , assuming the composition of “zirconyl oxalates” ( [67ZAI/SHU2] in Table sp , values of K were calculated by the VI-27). As discussed in the paper [67ZAI/SHU2] sp authors only for comparing the behavior of these solid compounds. Although these values give some qualitative insight about the solubility of these compounds, they can- not be considered as thermodynamic constants since the solids slowly transform into other forms. At present, no thermodynamic value can be selected for zirconium oxalate compounds. Crystal structure data are reported for Na , ] [ZrOH(ox) ·7H [98MOR/ALM] O 2 3 6 2 [K ox·H O] {Zr(ox) }H , and K [{Zr(ox) } ( μ –ox)]·4H O [97BAG/GAR] 3 2 2 2 n 2 2 6 3 , but no thermodynamic data are available for these compounds. [97BAG/GAR2] Table VI-27: Literature data on the solubility product of the solid compounds of zirconium with oxalate. 2– 2+ 2+ K ] where “ZrO ”][ox ” corresponds to the entity given in this table. = [“ZrO sp t ( ° C) Solid Definition of Reference Ionic medium K K sp sp 2+ 2 + 2– –10 [ 25 ] × Zr(OH) ][ox ] [64ZAI/BOC2] f 0.98 0.01 – 0.06 M [H 10 ± 2 )ClO I = 0.1 M (H, NH 4 4 2+ 2– –10 + ][ox 25 ] (5.0 ± 1) × 10 α [ [67ZAI/SHU2] ] Zr (OH) 0.02 – 1.0 M [H 26 2+ 2– –12 I = 1 M (H, NH )ClO 4 4 ] (9.0 Zr O(OH) ][ox β [ ± 1) × 10 24 2+ 2– –11 Zr O (OH) ][ox ] (4.9 ± 1) × 10 γ [ 2 22 VI.9.2 Zirconium oxalate complexes 4+ Although it is certain that Zr s with oxalate, reliable de- forms very stable complexe termination of their stability constants is quite difficult. As discussed in the previous 4+ can be predominant only in strongly acidic solutions at low zirconium section, Zr ). Even at low concen- [56ZIE/CON] , [51CON/REA] , [49CON/MCV] concentrations ( trations, zirconium undergoes hydrolytic polymerisation at a higher pH. Thus, the ex- perimental determination of the stability of zirconium oxalate complexes can be carried 4+ usually forms very stable out only in strongly acidic solutions. Moreover, since Zr 4+ complexes with many substances, the concentration of free Zr is difficult to be reliably determined. These restrictions, not only concerning the adaptable pH or acidity but also concerning suitable ionic media, are serious obstacles for obtaining reliable speciation and stability constants.

247 VI.9 Zirconium oxalate compounds and complexes 205 In Table VI-28, experimental equilibrium data found in the literature are pre- sented. Several papers report qualitative information only about zirconium oxalate com- plexation [59KIR/TAN] , [68POS/ZAI2] [86MAT/PAS] , or review original [68KOZ] , , [64RYA/MAR] , , and those are discussed in Appendix A but not literature [64CAL] included in Table VI-28. As discussed in detail in Appendix A, no paper can be ac- cepted in this review for the selection of the stabilities of zirconium oxalate complexes. [60BAB/DUB] , , [58ERM/BEL] Most papers had to be rejected for various reasons [61GRI/AST] , [62MAR/RYA2] , [62SHV/MAS2] , [66KOR/SHE2] , , [61BAB/SHT] [71PYA/KRA2] , , and only one study [64CAL/KYR] , [76TRI/SCH] [71SOL/IVA2] contains reliable experimental data. However, unlike the other metal ions considered in this review, no NEA selected values for Zr h ydrolysis were available at the time of the preparation of this section. Thus, a consistent re-evaluation of Zr oxalate complexation has to be postponed until selected hydrolysis data [64CAL/KYR] based on the data of are available. Table VI-28: Literature data on the form ation constants of oxalate complexes of zirconium t ( ° Method Medium Reference K C) log 10 4+ 2+ 2– U Zr(ox) + ox Zr [60BAB/DUB] 1, 0.5 M HClO sp ? 9.80 4 1 M HCl 9.80 [61BAB/SHT] 1) ? − + + − 2 4+ = 0.04 M (H I em 11.13 ,Zr ,ox ClO ) → 0 20 )( [62SHV/MAS2] ,Na 4 2) 1 N HNO ? (11.07 ± 0.2) [64CAL/KYR] 3 [64CAL/KYR] 0.2) ? (11.28 ± 5 N HNO 2) 3 1 N HCl ± 0.12) [66KOR/SHE2] ? 1) (10.26 2+ 4+ + U Zr(ox) + H + 2H ox Zr 2 0.2 M HCl ? 3.85 [58ERM/BEL] aix cix 2 M HClO ? (5.47 ± 0.10) [62MAR/RYA2] 4 cix 4 M HClO ? (5.60 ± 0.12) [62MAR/RYA2] 4 2– 2+ Zr(ox) + ox U Zr(ox) (aq) 2 sp 1, 0.5 M HClO ? 7.34 [60BAB/DUB] 4 − + 4+ + 2 − , ox )( ClO I = 0.04 M (H ,Zr ,Na ) → 0 20 9.18 [62SHV/MAS2] em 4 [64CAL/KYR] 0.3) ? (9.21 ± 1 N HNO 2) 3 2+ + ox(aq) U + H (aq) + 2H Zr(ox) Zr(ox) 2 2 aix 0.2 M HCl ? 4.15 [58ERM/BEL] cix 4 M HClO ? (4.21 ± 0.33) [62MAR/RYA2] 4 (Continued on next page)

248 VI Discussion of data selection for oxalate 206 Table VI-28 (continued) Method Medium ° C) log t K Reference ( 10 2 − 2– (aq) + ox Zr(ox) U Zr(ox) 2 3 ? 3.72 [60BAB/DUB] sp 1, 0.5 M HClO 4 2 + − + − 4+ 0 )(ClO ,Zr ,ox = 0.04 M (H ,Na ) → em 20 6.66 [62SHV/MAS2] I 4 + − 2 Zr(ox) ox(aq) U +2H (aq) + H Zr(ox) 2 2 3 0.2 M HCl 2.0 [58ERM/BEL] aix ? 2– − 2 4 − U + ox Zr(ox) Zr(ox) 4 3 1, 0.5 M HClO ? 0.30 sp [60BAB/DUB] 4 pot ? 3.89 − 4.10 [61GRI/AST] 0.1 M HCl 4+ + − + 2 − em = 0.04 M (H I 0 ,ox ,Zr ,Na ) → )(ClO 20 6.07 [62SHV/MAS2] 4 + − 2 4 − + H ox(aq) U + 2H Zr(ox) Zr(ox) 2 4 3 ? 1.7 [58ERM/BEL] aix 0.2 M HCl − 2+ 2 U + ox ZrO ZrO(ox)(aq) − 4 + 10 sp [H [60BAB/DUB] M ? 6.68 ] = 0.001, 0.5 × 2 − − 2 U ZrO(ox)(aq) + ox ZrO(ox) 2 + − 4 10 sp [H M ] = 0.001, 0.5 ? 3.34 [60BAB/DUB] × 2+ 2 − 2 − U + 2ox ZrO ZrO(ox) 2 sol pH 5.2 ? ± 0.10) [71PYA/KRA2] (8.68 + 2+ + H ox(aq) q Zr(OH) (ox)(aq) + 2H Zr(OH) 2 2 2 − dis 2 ? 6.6 [71SOL/IVA2] 4 M HClO 4 + 3+ 2+ ox(aq) Zr(ox) + H + H ZrOH U 2 ± I dis + NaClO ) ? (4.3 = 4 M (HClO 0.1) [76TRI/SCH] 4 4 1) Metal indicator method , 2) Sorption on silica gel Uranium oxalate compounds and complexes VI.10 VI.10.1 Solid uranium oxalates No structural or thermodynamic data for solid U(III) or U(V) oxalates could be identi- fied in the literature. Thus, only the information on solid U(IV) and U(VI) oxalates is discussed in this review. VI.10.1.1 Solid uranium(IV) oxalates Seekamp was probably the first to report the formation of a “green crystalline powder” of U(IV) oxalate when a solution of U(VI) nitrate and oxalic acid was exposed to “the sunlight of a July morning” [1862SEE] , presumably due to the reduction of U(VI) through photolysis. The first preparation and identification of the chemical composition

249 VI.10 Uranium oxalate compounds and complexes 207 [01KOH/ROS] . of solid U(IV) oxalate compounds were reported in 1901 [01KOH] , Since then, numerous papers have been published on the preparation of solid U(IV) oxalates, ranging from binary, ternary, quaternary and more complex compounds. Ex- H O(s) ( n = 1, 2, 6), U n · ·12H O(s), (ox) amples of the solid U(IV) oxalates are: U(ox) 2 2 2 2 4 + + + + ]· n H = 3, 4, 5, 7, 8; O (M = H k , Na [U , K = 1, , (ox) n NH ; i = 2, 4, 6, 8; j = 1, 2; M i j k 2 4 2+ 2+ 2+ 2+ 4, 5, 8, 10), M = 8, O(s) (M = Ca [U(ox) , Ba k , Cd ]· , Pb n ; i = 2, 4; H = 4, 6; n 2 i k + + + O(s) ( n ]·8H , K = 3, 22), , Sr[U(ox) O(s) (M = Na N H), La H [U(ox) n ] · 12), M 3 2 2 4 2 4 4 4 + + 3+ 3+ 3+ 3+ 3+ , O(s) (M = K , Ce , Pr ), , Nd NH , Tb ; Ln = La ]·8H MLn[U(ox) 4 2 4 U (ox) (SO ) ·6H O(s), U (ox) SO ·12H O(s) and other mixed ligand U(IV) oxalate 2 2 3 4 2 2 2 4 2 – – 2 − – 3 – − [87DOL] CO , , SCN , OH and PO , , , F [66CHE] compounds containing Cl 4 3 . [91MAT/KAR] emical analysis, Some of the solid U(IV) oxalates were characterised by ch X-ray diffraction and thermal gravimetry and, for a much smaller number of com- pounds, by solubility measurements. Since this review mainly concerns about the ther- modynamic data, Table VI-29 lists only the solid U(IV) oxalates of which the solubili- ties are reported in the literature. Information on the preparation, structure and proper- ties of thermal decomposition of solid U(IV) oxalates is available in the literature in- , [87DOL] , [87MEF/KRO] and [91MAT/KAR] . [66CHE] cluding the review articles Table VI-29: Solid U(IV) oxalates of which the solubilities are reported in the literature. Compound Reference [56GRI/PET2] [60ZAK/MOS] , , , [58GRI/PET2] O(s) ·6H U(ox) 2 2 [63GRI/PET] [70PET/STE2] , [79NIK] , U(ox) ·H O(s) or U(ox) [58DOR] ·6H O(s) 2 2 2 2 U(ox) ·8H O(s) [02ORL] K 4 2 4 ·8H (ox) [70PET/STE2] U O(s) K 2 5 2 2 [59GRI/PET] ]·8H O(s) KCe[U(ox) 2 4 U(ox) O(s) is the most frequently studied solid U(IV) oxalate. Upon heat- ·6H 2 2 (s) in nitrogen and U (s) in air O ing, it dehydrates and eventually decomposes to UO 8 3 2 , [87DOL] . The stepwise enthalpies of th e first and the second dehydra- [63BRE/CLA2] –1 tion and the decomposition were determined to be 96, 21, and – 238 kJ·mol , by differ- O(s) was . The standard entropy of U(ox) ·6H [69SUB/SIN] ential thermal analysis 2 2 ο –1 –1 . (298.15 K) = 385 J· K [73MOS] ·mol S estimated to be m It is generally agreed in the literature that U(ox) O(s) is sparingly soluble ·6H 2 2 [63GRI/PET] , [58DOR] , [58GRI/PET2] , [60ZAK/MOS] , , [56GRI/PET2] in water . The solubility increases as the concentrations of mineral acids , [79NIK] [70PET/STE2] (HClO , , HCl, HNO [58DOR] , and H , SO [56GRI/PET2] ) are increased 4 2 3 4 , [60ZAK/MOS] , [63GRI/PET] . Figure VI-23 shows the solubility of [58GRI/PET2]

250 VI Discussion of data selection for oxalate 208 U(ox) O(s) in HCl and water ( c = 0) from [56GRI/PET2] , [58DOR] and ·6H 2 HCl 2 [60ZAK/MOS] is unknown, but . The temperature for the data from [60ZAK/MOS] there is a good agreement between the data from [58DOR] and [60ZAK/MOS] . How- –1 ·6H (0.097 mmol·kg O(s) in water from [56GRI/PET2] ever, the solubility of U(ox) 2 2 − 1 (0.043 mmol·kg U). U) is more than twice as much as that from [60ZAK/MOS] t ·6H ( O(s) in HCl. Symbols: data from [58DOR] = Figure VI-23: Solubility of U(ox) 2 2 ( t [60ZAK/MOS] ( t unknown); curve: unweighted = 25°C), and [56GRI/PET2] 25°C), polynomial fit of the data. 2.5 [56GRI/PET2] [58DOR] [60ZAK/MOS] 2.0 -3 1.5 /mmol·dm 1.0 U C 0.5 0.0 01234567 -3 C /mol·dm HCl In the solutions of oxalic acid or the salt of oxalate (sodium, ammonium, potas- sium), the solubility of U(ox) due to the formation of ·6H O(s) increases significantly 2 2 , [63GRI/PET] , [79NIK] . , [60ZAK/MOS] aqueous U(IV) oxalate complexes [58DOR] ·6H ature is increased. Figure O(s) increases as the temper Also, the solubility of U(ox) 2 2 VI-24 shows the data reported in [79NIK] .

251 VI.10 Uranium oxalate compounds and complexes 209 ox at different tempera- ·6H O(s) in HClO Figure VI-24: Solubility of U(ox) and Na 2 2 4 2 . Symbols: experimental data; curves: unweighted polynomial tures reported in [79NIK] fit of the data. 8.0 2.5 0.01m Na ox 2 0.735 m HClO 4 2.0 7.5 0.1005 m HClO 4 -1 1.5 7.0 /mmol·kg 6.5 1.0 U C 0.5 6.0 0.0 5.5 60 40 100 20 80 o C / t The solubility experiments of U(ox) O(s) are usually complicated by the ·6H 2 2 potential oxidation of U(IV) during the experiments and the strong tendency of U(IV) towards hydrolysis. For example, in a study of the solubility of U(ox) O(s) in ·6H 2 2 , and H , the total concentration of uranium was [63GRI/PET] SO , HCl, HNO HClO 3 2 4 4 found to be much higher than the concentration of U(IV) in all the solutions except HCl, despite that great care was taken to prevent U(IV) from being oxidised by saturating the solutions with hydrogen and passsing nitrogen through the solutions. The difference between the concentrations of total uranium and U(IV) was attributed to the partial -radiolysis α oxidation of U(IV) to U(VI) by tracer amounts of oxygen or the product of . This casts serious doubts on the reliability of the solubility data of water [63GRI/PET] obtained from the experiments where the analysis of oxidation states of uranium is not performed and/or the equilibrium solid phase is not identified. It should be pointed out that the composition of the equilibrium solid phase was not characterised in any of these , O(s) was given [56GRI/PET2] ·6H solubility experiments, though the formula of U(ox) 2 2 [58GRI/PET2] [79NIK] , [60ZAK/MOS] , [63GRI/PET] , [70PET/STE2] , .

252 VI Discussion of data selection for oxalate 210 Solubility data from the references where experimental conditions are appro- priately described are considered in this review and summarised in Table VI-30. For the references in which the uncertainties of data were not given, or were too optimistic, a 0.2 is assigned to the ± K by this review. The results from log value of 10 s are included despite that the temperature of the experiment is unknown, [60ZAK/MOS] with the results at 25°C from [58DOR] because the data appear to agree (see Figure within the uncertainties (Table VI-30). VI-23) and with the data at 25°C from [79NIK] without essential experimental Solubility data from the refere nces with serious faults or conditions ( e.g. , temperature, ionic strength) are no t accepted in this review, including ·6H , O(s) [56GRI/PET2] , [58GRI/PET2] , [63GRI/PET] the data for U(ox) 2 2 ·8H [70PET/STE2] [02ORL] , for K U (ox) (s) , for K O(s) U(ox) , and [70PET/STE2] 4 2 4 5 2 2 for KCe[U(ox) . The reasons for rejection are provided in Ap- O(s) [59GRI/PET] ].8H 2 4 pendix A. Table VI-30: Solubility constants of solid uranium(IV) oxalates reported in the litera- ture. t Method Ionic medium ° C) Reference log K ( s 10 3+ 2 − + ·H + 2 ox O(s) U + H U(OH) U(ox) 2 2 25 − (21.91 0.118-6.18 M HCl 0.20) [58DOR] sol ± ox/3.08 N HCl 25 − (21.03 ± sol 0.26% H [58DOR] 0.20) 2 [58DOR] − (22.00 ± 0.20) 25 ox/6.05 N HCl sol 0.26% H 2 + 4+ O(l) O(s) + 4 H ox(aq) + 6 H U U ·6H + 2 H U(ox) 2 2 2 2 sol (0-3) M HCl ? − (10.82 ± 0.20) [60ZAK/MOS] + 2+ U(ox) U O(s) + 2 H U(ox) + H ox(aq) + 6 H O(l) ·6H 2 2 2 2 (0-3) M HCl − (7.60 ± 0.20) [60ZAK/MOS] ? sol O(l) ·6H U(ox) O(s) U U(ox) (aq) + 6 H 2 2 2 2 (0-3) M HCl ? − (4.52 ± 0.20) [60ZAK/MOS] sol 0.20) − (4.82 ± sol 0.1005 m HClO [79NIK] 25 4 40 (4.74 ± 0.20) − − (4.70 ± 0.20) 50 60 − (4.47 ± 0.20) 70 (4.34 ± 0.20) − 80 ± (4.24 0.20) − − (4.03 ± 0.20) 90 2– 4+ + 2 ox ·6H U U O(l) O(s) + 6 H U(ox) 2 2 2 sol (0-3) M HCl ? − (21.37 ± 0.19) [60ZAK/MOS] The solubility of U(ox) ·6H O(s) in 0.1005 m HClO was studied at variable 2 2 4 . The solubility constant °C) (Figure VI-24) [79NIK] temperatures (25-90 log K (VI.22) was found to increase as the te mperature was increased (Table VI-30), 10 s

253 VI.10 Uranium oxalate compounds and complexes 211 ·6H O(s) U U(ox) (aq) + 6 H O(l). (VI.22) U(ox) 2 2 2 2 From these data, the values of ∆ (VI.22) at variable temperatures were H rm calculated in [79NIK] 10%) are assigned by this (Table VI-31). The uncertainties ( ± review. . [79NIK] Table VI-31: E xperimental enthalpy of R eaction (VI.22) reported in –1 ∆ t Method Ionic medium H ° ( kJ·mol ( ) C) rm O(l) ·6H (aq) + 6 H O(s) U U(ox) U(ox) 2 2 2 2 25 ± 1.0) sol 0.1005 m HClO (10.4 4 2.0) ± 40 (16.7 2.0) ± 50 (21.5 ± 60 (26.9 3.0) 70 (34.4 ± 3.0) ± 4.0) 80 (39.0 ± 90 (46.0 5.0) As pointed out previously and discussed in Appendix A, the equilibrium solid ·6H O(s) was not characterised to confirm phase in the solubility experiments of U(ox) 2 2 [79NIK] the composition in any of the references, including . This raises questions on the reliability of the solubility constants in Table VI-30 and the enthalpy of dissolution in Table VI-31. However, the data in Tabl e VI-30 reasonably agree between different references and are consistent with the related stability constants of aqueous U(IV) ox- alate complexes from different references (Tab le VI-36, page 226). The enthalpy values in Table VI-31 are also consistent with the trend of the solubility constants in Table K log (VI.22) as a function of 1/ T (Figure VI-25-a) indicates that VI-30. The plot of 10 s / H d ∆ log K (VI.22) should be positive in the temperature range because d (1/ T ) < 0, 10 rm (VI.22) probably does not remain constant because H ∆ that rm H ∆ d ) d (1/ T log ≠ constant, and that (VI.22) should become increasingly posi- K / rm 10 in- T ) becomes larger as T log K / d (1/ d tive as the temperature is increased because 10 ∆ (VI.22) as a function of the te mperature is shown in Figure H creases. A plot of rm VI-25-b. A weighted linear regression results in a value of ± (VI.22) = (513 41) ∆ C r,m p –1 –1 . ·mol J·K The above discussions show that some of the solubility data reported in the lit- erature agree within experimental uncertainty and appear to be of good quality. Since there are no ionic species invo lved in Reaction (VI.22), th e medium effect on the solu- bility constant and enthalpy is expected to be small and the correction for ionic strength is probably insignificant compared to the uncertainty of the data. Thus, the values of in Table VI-31 may be taken as log K (VI.22) in Table VI-30 and (VI.22) H ∆ rm 10 s ο ο K log and H ∆ (VI.22). However, taking into consideration various shortcomings s rm 10

254 VI Discussion of data selection for oxalate 212 in the experiments, including the lack of characterisation of the equilibrium solid phase and confirmation of the oxidation state of U( IV) and the varied ionic strength, no ther- modynamic solubility constants or enthalpy are recommended for U(ox) ·6H O(s) by 2 2 this review. Figure VI-25: Experimental solubility constants (a) and enthalpy (b) for the Reaction (VI.22), original data from [79NIK] . 60 -3.6 (b) (a) 50 -4.0 40 -1 K 10 30 -4.4 /kJ.mol log m H r ∆ 20 -4.8 10 0 -5.2 300 320 340 360 380 0.32 0.34 0.26 0.28 0.30 280 (E-2) -1 -1 T T /K /K VI.10.1.2 Solid uranium(VI) oxalates VI.10.1.2.1 General comments Preparation and characterisation of many solid U(VI) oxalate compounds were reported . [66CHE] in the literature, ranging from binary, ternary and more complex compounds Examples of the binary and tern ary solid U(VI) oxalates are: UO (ox)· n = 0, 1, n H O ( 2 2 2+ + 2+ , Ba ; ]· n H [UO O (M = Li = 4, 5, 10), M , (ox) ]· n H n (ox) O (M = Sr 3), M[UO 2 2 2 2 2 2 2 2+ + + + + + + , Tl n ; (ox) , K [UO , Rb = 0, 2, 3, 4, 5, 6), M n ; NH O (M = Ba , H ]· , Cs Na 2 3 2 2 4 + + + NH n = 7), M H ]· O (M = K , Tl , n ; n = ?), M H [(UO ]· ) O (M = (ox) (ox) [UO n 2 2 3 2 2 2 4 3 2 4 + + + + + + + H; n = 0, 3, 4) and M [(UO ) (ox) ]· n H O (M = Na , Tl , K , , Cs N , N H, Cs 2 5 2 2 6 4 4 + [66CHE] NH . The quaternary or more complex solid U(VI) oxala- ; n = 2, 6, 10, 13) 25 – – – tes usually contain alkali metals and one or two other ligands such as OH , , F , Cl 2 2 2 – − 2 − − 2 − − , , O SO , , CO and acetate. Detailed descriptions of the com- , SeO SO SCN 4 3 4 2 3 , pounds are available in the literature including the review articles [66CHE] [87MEF/KRO] [91MAT/KAR] . ,

255 VI.10 Uranium oxalate compounds and complexes 213 characterised by chemical analysis, Some of the solid U(VI) oxalates were X-ray powder diffraction and thermal gravim etry and, for a much smaller number of compounds, by structures from X-ray single crystal data or solubility measurements. Table VI-32 lists only the solid U(VI) oxalates of which either the structural data by X-ray crystallography or the solubilities are reported in the literature. Table VI-32: Solid U(VI) oxalates of which the structures by X-ray crystallography or the solubilities are reported in the literature. Reference Compound Structure by X-ray Solubility Crystallography , O [56STA/CRO3] ox·3H [1842EBE] UO 2 2 [72JAY/CHA] [16COL] , [17COL] [17COL2] , [25COL] [34COL] [52AMP/DAV] [57BOL/KOR3] [57BOL/KOR4] [58BOL/KOR] [59MOS/ZAK] ox·H UO O, UO [64BRE/CLA] ox 2 2 2 (NH ] [UO ) [73ALC2] (ox) 2 2 4 2 Li [UO (ox) [97DAH/CHA2] ]·5H O O, Na ]·4H [UO (ox) 2 2 2 2 2 2 2 2 Na ]·3H ]·5H O, K [UO (ox) (ox) O [UO [67SHC/BEL2] 2 2 2 2 2 2 2 2 M [69SHC/BEL] ]· n H NH O (M = Rb, Cs, NH (ox) , (C H = ) n [UO ; 2 2 4 2 2 2 2 2 5 2 and 4) [(C (ox) (NH)] ]·4H O [UO ) H 2 3 2 5 2 2 2 (CN ) (ox) H CO(NH [UO ) ] 2 2 2 2 2 6 3 a M ) ) (ox) ]·H [UO O CO(NH (M = K, Rb, Cs, NH 4 2 2 2 2 2 2 Tl [73JAY/DIA] [UO ]·2H O (ox) 2 2 2 2 (NH [UO [73ALC] (ox) ) ] 3 2 4 4 (NH ) [(UO ) (ox) ] [73ALC3] 2 4 2 2 3 K [(UO ) [75JAY/SIN] (ox) O ]·4H 2 2 3 2 2 [2002SZA/FIS] F (ox)], K [UO F (ox) ], K [UO [(UO ) F (ox)] [78CHA/BHA] K 4 2 2 2 2 2 4 2 2 2 2 K F(ox) [UO (ox)]·2H O, K [UO F ]·3H O [68SHC/BEL] 2 2 2 2 3 3 3 2 (CN [UO F (ox)]·H O (CN H ) [UO F(ox) ]·H O H ) 2 2 2 6 3 3 2 3 6 3 2 3 ) [UO (ox)(SO )]· n H O [67ZAK/ORL2] (NH 2 2 3 2 4 (NH ) [UO [64BAS/KRI] (O O )ox]·3H 2 2 2 4 2 (N H [(UO ) (ox) ]·2H O [86GOV/PAT] ) 6 5 2 2 2 5 2 (Continued on next page)

256 VI Discussion of data selection for oxalate 214 Table VI-32: (continued) Reference Compound Structure by X-ray Solubility Crystallography K (ox) [(UO ]·10H ) [76LEG/JEA] [67SHC/BEL] O 2 6 2 2 5 [67SHC/BEL] (ox) [(UO ]·2H Rb O, Cs ) [(UO O ) ]·10H (ox) 2 5 2 2 6 6 2 5 2 2 b (CN ) ) H (ox) ]·4H [(UO O 2 6 3 6 2 2 5 c [C ].3H ] H [(UO (NH ) (ox) O ) 2 5 2 2 4 3 2 2 3 Cs [UO [2000MIK/GOR2] (ox)(SeO )] 4 2 2 (NH ) [UO (ox)(SeO )]·1.5H O [96MIK/GOR2] 4 4 2 2 2 Rb [UO (ox)(SO )]·4H [93MIS/MIK] O 2 4 2 2 a: CO(NH ) , urea. 22 + b: CN H , guanidinium cation. 36 2 + [C H (NH ) ] , ethylenediaminium cation. c: 323 24 The crystallographic data from the studies listed in Table VI-32 are not the subject of this review. The solubility data from the studies listed in Table VI-32 for the U(VI) oxalates are not consider ternary or more complex solid ed in this review, includ- H [(UO O from ) n (ox) ]· ]· n H (ox) O from [67SHC/BEL] , M [UO ing M 2 2 2 5 2 6 2 2 2 and [UO , M O from H n (ox) }]· {CO(NH [69SHC/BEL] ) [67SHC/BEL2] 2 2 2 2 2 2 , , M [68SHC/BEL] [UO O from F H (ox)]· n H n O and K ]· [UO F(ox) [67SHC/BEL2] 2 3 2 2 3 3 2 2 (NH H ) (SO . The reasons for rejection are given in )ox]· n [UO [67ZAK/ORL2] O from 2 3 4 2 2 Appendix A. UO ox·3H O(cr) VI.10.1.2.2 2 2 ox·3H O is the most frequently studied solid U(VI) oxalate. Upon heating, it dehy- UO 2 2 (s) in nitrogen and U O (s) in air drates and eventually decomposes to UO 2 8 3 , [87DOL] . The standard enthalpy of th e dehydration Reactions (VI.23), [63BRE/CLA2] ο ο –1 ∆ H H ∆ (VI.24) are determined to be ± 3.4) kJ·mol and (VI.23) = (119.9 (VI.24) rm rm –1 3.5) kJ·mol : at 298.15 K by solution calorimetry [77OHA] ± = (73.3 UO ox·3H O(cr) + 2 H O(cr) U UO O(g) (VI.23) ox·H 2 2 2 2 2 UO O(g) (VI.24) O(cr) U UO ox(cr) + H ox·H 2 2 2 2 By determining the enthalpies of dissolution of H ox·2H O(cr), 2 2 ox·H ox·3H O(cr), UO ox(cr) in a solution of nitric acid and using a O(cr) and UO UO 2 2 2 2 2 ο H ∆ thermochemical cycle upon which (UO ox·3H O, cr) is based, the standard en- 2 2 fm ο H O(cr) and UO ∆ , of UO ox·3H O(cr), UO ox·H ox(cr) at thalpies of formation, 2 2 2 2 2 fm 2.5) 2.4), – (2711.8 ± 2.4), and – (1796.7 ± ± 298.15 K were calculated to be – (2715.3 ο –1 H ∆ kJ·mol O, cr) was also calcu- (UO , respectively [77OHA] . The value of ox·3H 2 2 fm –1 − ± 4.1) kJ·mol from the heat of combustion determined by bomb (2688 lated to be . The standard [77OHA] , comparable to th at reported in [92THA/KUM] calorimetry

257 VI.10 Uranium oxalate compounds and complexes 215 ο –1 –1 ox·3H entropy of UO S (298.15 K) = 343 J·K ·mol O was estimated to be 2 2 m ox·3H O(cr), . This value is consistent with the entropy of formation of UO [73MOS] 2 2 –1 –1 (347.517 63.351) J·K ·mol , calculated from reaction data selected by this review ± Table III-1). ( cf. ο O, cr) from (UO H ox·3H and ∆ [77OHA] Based on the values of 2 2 fm [92THA/KUM] ox·3H O(cr) is selected by , the standard enthalpy of formation of UO 2 2 this review: ο –1 . ∆ (UO H ox·3H kJ·mol O, cr, 298.15 K) = − ( 2702 ± 1 8 ) 2 2 fm The solubility of UO ox·3H O has been studied in water, mineral acids and so- 2 2 , , [17COL] , [16COL] lutions of oxalates under various conditions [1842EBE] [25COL] , [34COL] , [52AMP/DAV] , , , [57BOL/KOR4] , [57BOL/KOR3] [17COL2] [58BOL/KOR] , [59MOS/ZAK] . Some of the results from these papers are not conside- red by this review due to various shortcomings associated with the experimental condi- ox tions or the lack of information on the experiments. These include the solubility in K 2 [16COL] , in Na [17COL] , in (NH ) ox [17COL2] , in Mox (M = Ca, Sr and Ba) ox 2 2 4 , and the solubility in mixed electrolyte solutions . Detailed [57BOL/KOR4] [34COL] discussions on the exclusion of the results are provided in Appendix A. Table VI-33 summarises only the studies of which the solubility data of UO ox·3H O are considered 2 2 by this review. ox·3H O(s) considered by this review. Table VI-33: Studies of the solubility of UO 2 2 o System ( C) Solid phase Reference t water 14, 100 UO ox·3H [1842EBE] O (?) 2 2 11 – 100 water ox·3H UO O [25COL] 2 2 ox·3H HNO 20 UO O /water 3 2 2 (1.8–31% by mass) HNO /water 20 UO [52AMP/DAV] ox·3H ox O, UO 2 2 3 2 (15–70% by mass) /water HNO 25, 50 UO ox·H ox·3H O, UO O, [58BOL/KOR] 2 2 3 2 2 (0–77% by mass) UO (NO ) 2 2 3 /water O(?) ox·3H HNO [59MOS/ZAK] 20 UO 2 3 2 –3 (0.5–3.0 mol·dm ) HClO /water O(?) ox·3H 20 UO 2 2 4 –3 (0.5–3.0 mol·dm ) H ox/water O(?) ox·3H 0-70 UO [57BOL/KOR3] 2 2 2 H (0–38% by mass) O(?) ox·2H 2 2 H ox/HClO [59MOS/ZAK] /water O(?) ox·3H 20 UO 2 2 2 4 (0.5 and 1.0 M HClO ox) , 0.08-0.6 M H 4 2 H ox/HNO /water O(?) ox·3H 20 UO 2 2 2 3 (0.5, 2.0 and 3.0 M HNO ox) , 0.08-0.6 M H 2 3 (Continued on next page)

258 VI Discussion of data selection for oxalate 216 Table VI-33: (continued) o System C) Solid phase Reference t ( ) O(?) /water ox·3H (NH ox/HNO [59MOS/ZAK] 20 UO 4 2 2 2 3 (0.5 M HNO ) , 0.07-0.3 M (NH ox) 4 3 2 ) ox/HClO (NH O(?) /water ox·3H 20 UO 2 2 2 4 4 (0.5 and 1.0 M HClO ) ox) , 0.07-0.3 M (NH 2 4 4 O(s) in water Solubilities of UO ox·3H VI.10.1.2.2.1 2 2 ox·3H O(s) in water at different temperatures (0-100°C) were The solubilities of UO 2 2 . As shown in [59MOS/ZAK] , [25COL] , [57BOL/KOR3] and [1842EBE] reported in and at 0°C from Figure VI-26, all the results, except the data at 14°C from [1842EBE] [57BOL/KOR3] ggesting that the values are repro- , agree with each other very well, su seems erroneously ducible and probably reliable. The data at 14°C from [1842EBE] ) and the data at 0°C from [57BOL/KOR3] high (> 40% higher than that from [25COL] ith the trend demonstrated by all the other data. seems too low compared w Figure VI-26: Solubility of UO O(s) in water at different temperatures reported ox·3H 2 2 in the literature. Symbols: experimental data; curve: unweighted polynomial fit of the data. 100 [1842EBE] [25COL] 80 [57BOL/KOR3] [59MOS/ZAK] -1 60 /mmol·kg 40 U C 20 0 80 0 20 40 120 100 60 o C t /

259 VI.10 Uranium oxalate compounds and complexes 217 Data in Figure VI-26 show that the solubility of UO O(s) in water ox·3H 2 2 –1 ) to is increased from 25°C (~ 0.015 mol·kg increases by six times as the temperature –1 100°C (~ 0.09 mol·kg ). By comparison of the solubility in water at the same tempera- O is a much more soluble solid than U(ox) ·6H O. For example, at ox·3H ture, UO 2 2 2 2 –1 ox·3H O in water (~ 15 mmol·kg ) is about 300 times as 25°C, the solubility of UO 2 2 –3 high as that of U(ox) O in water (~ 0.05 mmol·dm ·6H ) (see Figure VI-23, the data 2 2 = 0). point at c HCl VI.10.1.2.2.2 Solubilities of UO ox·3H O(s) in HNO 2 2 3 , ox·3H and O in HNO [52AMP/DAV] was reported in [25COL] The solubilities of UO 3 2 2 at 20°C and in [58BOL/KOR] at 25 and 50°C. The data obtained from [59MOS/ZAK] and [58BOL/KOR] are ic acid solutions from [52AMP/DAV] highly concentrated nitr rejected by this review due to the change of solid phase(s) under these conditions. De- tailed discussions are given in Appendix A (Figure A-7 and Figure A-9). The data in –1 [59MOS/ZAK] ] < 5 mol·kg from [25COL] , [52AMP/DAV] , solutions with [HNO 3 and [58BOL/KOR] data indicate that, at each tempera- are shown in Figure VI-27. The ture, the concentration of U(VI) increases as the concentration of nitric acid is increa- ox·3H 25 and 20°C, and that the O is higher at 50°C than sed, that the solubility of UO 2 2 data at 20°C from different studies are in good agreement. This review considered that [59MOS/ZAK] the data at 20°C are reproducible an d reliable. The data at 20°C from O. ox·3H are included in the analysis to calculate the solubility constants for UO 2 2 O(s) in nitric acid at different temperatures ox·3H Figure VI-27: Solubility of UO 2 2 reported in the literature. Symbols: experimental data; curves: unweighted polynomial fit of the data at 20°C and 50°C. 0.13 o [58BOL/KOR] C 50 o C 25 [58BOL/KOR] o C 20 [25COL] 0.10 o C 20 [52AMP/DAV] o o 50 C C 20 [59MOS/ZAK] -1 0.07 /mol·kg U o 20 C C 0.04 0.01 012345 -1 /mol·kg C HNO3

260 VI Discussion of data selection for oxalate 218 ox·3H VI.10.1.2.2.3 O(s) Solubility constants of UO 2 2 ox·3H ± 1)°C in a O at (20 Based on the comprehensive studies of the solubility of UO 2 2 variety of solutions, solubility constants of UO O for the following reactions ox·3H 2 2 by solving sets of equations with the solubility data: were calculated in [59MOS/ZAK] UO O(l) (VI.25) U UO ox(aq) + 3 H O(s) ox·3H 2 2 2 2 2+ + UO UO U ox·3H O(l) (VI.26) O(s) + 2 H + H ox(aq) + 3 H 2 2 2 2 2 2 − + UO U O(s) + H O(l) (VI.27) + 3 H ox·3H + 2 H ox(aq) UO (ox) 2 2 2 2 22 4 + − UO + 3 H ox(aq) U ox·3H O(l) (VI.28) UO (ox) O(s) + 2 H + 4 H 2 2 2 2 23 K K K (VI.26), K (VI.28) are the solubility constants “directly” (VI.27), and (VI.25), s s s s calculated from the experimental solubility data. Using the dissociation constants of + – –5 – + 2– ][ox ox] = 0.108), the au- ]/[Hox ox ( ] = 6.4 × 10 K and K ]/[H = [H = [H ][Hox H 2 2 1 2 thors [59MOS/ZAK] calculated the solubility product of UO ox·3H O for Reaction 2 2 (VI.29): 2+ 2– + 3 H O(s) U O(l) (VI.29) ox·3H UO + ox UO 2 2 2 2 2+ 2– UO where K K (VI.29) = [ ] [ox ). The solubility constants ] = K (VI.26) × ( K 1 s 2 s 2 [1842EBE] K , (VI.26), K [25COL] (VI.27), K , (VI.28) and K (VI.25), (VI.29) from K s s s s s [59MOS/ZAK] and VI-34. Because the un- are summarised in Table [57BOL/KOR3] certainties of data were unavailable or too optimistic in the references, a value of ± 0.2 K log by this review. Data for Reactions (VI.25), (VI.26), (VI.27), is assigned to the 10 (VI.28) and (VI.29) are discussed as follows. (ox)·3H O(s) in water, According to Reaction (VI.25), the solubility of UO 2 2 –1 [57BOL/KOR3] ), obtained in [1842EBE] , [25COL] or would be equal to or (mol·kg s (VI.25) if UO ox(aq) is the only species or the dominant species in solution K close to 2 s in equilibrium with the solid. To determine if these data can be accepted as K (VI.25), s –1 speciation calculations were performed by this review for a solution of 0.015 mol·kg O(s)), (ox)·3H ox(aq) at 25°C (this is the experimentally observed solubility of UO UO 2 2 2 using the formation constants of U(VI) oxalate complexes selected by this review (Sec- tion VI.10.2.4.1) and the hydrolysis constants of U(VI) in the literature . The results indicate that the solution would contain 91% UO ox(aq), [2003GUI/FAN] 2 2+ 2 − + 2 + UO UO (ox) (UO ) (OH) 4% and 0.3% , 4.5% , 0.1% UO , suggesting OH 2 2 22 2 22 shows that the pcH of eciation calculation also s ≈ log (VI.25) at 25°C. The sp K log s 10 10 the solution would be around 4, which is consistent with the observations in the litera- ture [56GRI/PET2] . It is not clear how the speciation of the solution of (ox)·3H O(s) would change at other temperat ures because the U(VI) oxalate com- UO 2 2 ox(aq) plex formation constants at other temperatures are unknown. However, UO 2 could still be the dominant species in that solution. Therefore, as an approximation, the values of K (VI.25) at variable temperatures are directly calculated from the solubility s [57BOL/KOR3] and , [25COL] O(s) in water reported in [1842EBE] , (ox)·3H of UO 2 2

261 VI.10 Uranium oxalate compounds and complexes 219 0.2 is assigned to the values of ± and included in Table VI-34. An uncertainty of K (VI.25) calculated directly from the solubility in water. log s 10 Table VI-34: Solubility constants of UO (ox)·3H O(s) reported in the literature 2 2 ° C) log K Method Ionic medium t Reference ( 10 s ox(aq) + 3 H O(s) U UO UO ox·3H (VI.25) O(l) 2 2 2 2 (1.92 (20 ± 1) − sol (0.5-3.0) M HClO 0.20) [59MOS/ZAK] ± 4 (20 ± 1) − (1.92 ± 0.20) (0.5-3.0) M HNO 3 1.0 M HClO sol − ox (20 ± 1) /0.08-0.636 M H (1.92 ± 0.20) 2 4 2.0 M HNO (1.89 1) ± /0.08-0.636 M H − ox (20 ± 0.20) 3 2 I → 0 14 − (1.72 ± 0.20) [1842EBE] sol (1.09 ± 0.20) 100 − 0.20) 15 − (1.88 ± 0 [25COL] → I sol − (1.85 ± 0.20) 20 50 (1.55 ± 0.20) − − (1.33 ± 0.20) 75 100 − (1.05 ± 0.20) I sol 0 0 − (2.23 ± 0.20) [57BOL/KOR3] → 25 − (1.80 ± 0.20) 40 − (1.64 ± 0.20) 50 (1.55 ± 0.20) − 70 − (1.35 ± 0.20) 2+ + UO ox(aq) + 3 H O(s) + 2 H U O(l) (VI.26) ox·3H + H UO 2 2 2 2 2 ± /0.08-0.636 M H ox (20 ± 1) − (3.52 sol 0.20) [59MOS/ZAK] 1.0 M HClO 2 4 2.0 M HNO ox (20 ± 1) − (3.66 ± 0.20) /0.08-0.636 M H 2 3 a sol 0.5 M HClO (20 ± 1) − (3.39 ± 0.20) [59MOS/ZAK] 4 a 1.0 M HClO ± 1) − (3.65 ± 0.20) (20 4 a 1.5 M HClO (20 ± 1) − (3.39 ± 0.20) 4 a 2.0 M HClO (20 ± 1) − (3.54 ± 0.20) 4 a 2.5 M HClO (20 ± 1) − (3.60 ± 0.20) 4 a 3.0 M HClO ± 1) − (3.84 ± 0.20) (20 4 a 0.5 M HNO sol (20 ± 1) − (3.15 ± 0.20) [59MOS/ZAK] 3 a 1.0 M HNO ± 1) − (3.31 ± 0.20) (20 3 a 1.5 M HNO (20 ± 1) − (3.44 ± 0.20) 3 a 2.0 M HNO (20 ± 1) − (3.31 ± 0.20) 3 a 2.5 M HNO (20 ± 1) − (3.38 ± 0.20) 3 a 3.0 M HNO (20 ± 1) − (3.48 ± 0.20) 3 (Continued on next page)

262 VI Discussion of data selection for oxalate 220 Table VI-34: (continued) Method Ionic medium ° C) log t K ( Reference s 10 2 − + ox(aq) U UO O(l) UO ox O(s) + H + 2 H ox·3H + 3 H (VI.27) 2 2 2 2 22 (1.96 sol ox (20 ± 1) − /0.08-0.636 M H ± 0.20) [59MOS/ZAK] 0.5 M HClO 2 4 − and 0.07-0.28 M (NH ) 0.20) ox (20 ± 1) 0.5 M HClO (1.92 ± 2 4 4 1.0 M HClO (1.92 and 0.07-0.28 M (NH ox (20 ± 1) − ) ± 0.20) 4 2 4 0.5 M HNO ) 0.20) ox (20 ± and 0.07-0.28 M (NH − (1.85 ± 1) 2 3 4 1.0 M HNO 1) ) ox (20 ± and 0.07-0.28 M (NH − (1.85 ± 0.20) 2 4 3 4 − + (VI.28) U ox·3H ox(aq) UO (ox) O(l) + 4 H O(s) + 2 H + 3 H UO 2 2 2 2 23 0.5 M HClO sol /0.08-0.636 M H ox (20 ± 1) − (1.96 ± 0.20) [59MOS/ZAK] 4 2 1) ) 0.20) ox (20 ± 0.5 M HClO − (1.85 ± and 0.07-0.28 M (NH 2 4 4 1.0 M HClO and 0.07-0.28 M (NH ) 0.20) ox (20 ± 1) − (1.80 ± 2 4 4 0.5 M HNO and 0.07-0.28 M (NH ) 0.20) ox (20 ± 1) − (1.85 ± 2 4 3 1.0 M HNO and 0.07-0.28 M (NH ) 0.20) ox (20 ± 1) − (1.70 ± 2 4 3 2+ 2– + 3 H O(s) U O(l) (VI.29) ox·3H UO + ox UO 2 2 2 2 − 0.5 M HClO (20 sol 1) (8.55 ± 0.20) [59MOS/ZAK] ± 4 1.0 M HClO (20 ± 1) − (8.81 ± 0.20) 4 1.5 M HClO (20 1) − (8.55 ± 0.20) ± 4 2.0 M HClO ± 1) − (8.70 0.20) ± (20 4 2.5 M HClO (20 ± 1) − ± 0.20) (8.85 4 3.0 M HClO 0.20) (20 1) − ± ± (9.00 4 1.0 M HClO and 0.08-0.636 M H 0.20) ox (20 ± 1) − (8.68 ± 2 4 3.0 M HClO (8.74 − ox (20 ± 1) and 0.07-0.28 M (NH ) ± 0.20) 2 4 4 sol 0.5 M HNO (20 ± 1) − (8.31 ± 0.20) [59MOS/ZAK] 3 1.0 M HNO ± 1) − (20 ± 0.20) (8.47 3 1.5 M HNO ± (20 1) − (8.60 ± 0.20) 3 2.0 M HNO (20 ± 1) − (8.47 ± 0.20) 3 2.5 M HNO (20 (8.54 ± 0.20) − 1) ± 3 3.0 M HNO ± 1) − (8.64 ± 0.20) (20 3 2.0 M HNO ox (20 ± and 0.08-0.636 M H − (8.82 ± 0.20) 1) 3 2 3.0 M HNO − and 0.08-0.476 M H ox (20 ± 0.20) (8.55 ± 1) 2 3 a: Calculated from K (VI.29) and the “stepwise” dissociati on constants of oxalic acid used in [59MOS/ZAK] –5 ( K K = 6.4 × 10 = 0.108 and ). 1 2 Reaction (VI.25) does not involve ionic sp ecies so that the effect of the ionic medium on the equilibrium is expected to be small and probably negligible. As a result, all the values of log K (VI.25) from different ionic me dia in Table VI-34 are treated 10 s in Figure VI-28. From the linear regression T → 0 and plotted as a function of 1/ I as for ο log K in Figure VI-28, the value of (VI.25) at 298.15 K is calculated to be 10 s − (1.80 ± 0.27), identical to that from [57BOL/KOR3] . This value is selected by this review: ο log K (VI.25) = − (1.80 ± 0.27). s 10

263 VI.10 Uranium oxalate compounds and complexes 221 , the selected enthalpy of vs . 1/ T (VI.25) K From the linear regression of log 10 s Reaction (VI.25) is calculated to be ο –1 ∆ H (VI.25) = (20.2 ± 3.5) kJ·mol . rm ο Using the value of 2702 ∆ H ox·3H 18) O, cr, 298.15 K) = − ( (UO ± 2 2 fm − 1 ο O, l, 298.15 Section VI.10.1.2.2), and the (H H , selected in this review ( ∆ cf. kJ·mol 2 fm 1 − K) = – (285.83 ± from NEA-TDB auxiliary values, the enthalpy of for- 0.04) kJ·mol ox(aq) can be calculated: mation of UO 2 ο –1 . (UO H (ox), aq, 298.15 K) = − (1824.3 ± 18.3) kJ·mol ∆ 2 fm From the selections for Reaction (VI.25), the following selected values are cal- culated: ο –1 ± ∆ (UO ox·3 H 3.1) kJ·mol O, aq, 298.15 K) = − (2395.1 G 2 2 fm ο –1 –1 S . O, aq, 298.15 K) = (UO ± 63.4) J·K (347.5 ·mol ox·3 H 2 2 m Figure VI-28: K (VI.25) as a funtion of 1/ T . log s 10 -0.8 [1842EBE] [25COL] [57BOL/KOR3] -1.2 [59MOS/ZAK] s K 10 -1.6 log -2.0 -2.4 0.30 0.38 0.36 0.34 0.32 0.28 0.26 (E-2) -1 -1 T K / K log The values of or (VI.27) in Table VI-34 a ppear constant in HClO 4 10 s solutions of different concentrations (0.5 or 1.0 M). Weighted average values HNO 3 and HNO with the were calculated by this review for the media of 0.5-1.0 M HClO 3 4 uncertainties assigned according to the method outlined in Appendix C. The values are shown in Table VI-35.

264 VI Discussion of data selection for oxalate 222 Subtraction of Reaction (VI.28) by Reaction (VI.27) results in the reaction of 4 − complex: UO (ox) the stepwise formation of the 23 2 − − 4 + UO (ox) UO (ox) U + 2 H ox(aq) (VI.30) + H 2 23 22 From the values of (VI.28) in Table VI-34, the K (VI.27) and log log K s 10 10 s K log ± 0.28), (0.07 ± 0.28), (VI.30) are calculated by this review to be: (0 values of 10 (0.12 ± 0.28) and (0.15 ± 0.28) corresponding to the 5 different ionic media ± 0.28), (0 for Reactions (VI.27) and (VI.28) shown in Table VI-34. t compared to the uncertainty of the These values are small and insignifican data, suggesting that the formation of the third U(VI) oxalate complex is negligible under the conditions in [59MOS/ZAK] solution acidity was too , probably because the high and/or the concentration of oxalate was too low. Consequently, the values of log K (VI.28) from [59MOS/ZAK] are considered to be qu estionable by this review. 10 s log K (VI.28) in Table VI-34 are not included in further As a result, the values of 10 s evaluation. K log (VI.29), in Table VI-34, are not the primary data di- The values of 10 s rectly calculated from the experiments. Instead, they were obtained from log K (VI.26) in [59MOS/ZAK] by using the dissociation constants of oxalic acid ( K 1 10 s + – 2– + – –5 = [H ]/[H ] = 6.4 and K × = [H ]/[Hox ][Hox ][ox 10 ox] = 0.108) that differ from the 2 2 selected values by this review (Section VI.3). Besides, the dissociation constants should be dependent on the ionic media, but it appears that the same values were erroneously K log log K (VI.29) from to calculate (VI.26) in different [59MOS/ZAK] used in s 10 s 10 , the oxalate [59MOS/ZAK] ionic media. Under the strongly acidic conditions used in ects on the dissociation constants of oxalic ligand is fully protonated, but the medium eff acid are not well known. Therefore, the values of solubility product for Reaction 2– , must be regarded with suspicion. The (VI.29) that involves the free oxalate ion, ox log (VI.29) are rejected by this review, but re-evaluated to retrieve two K values of s 10 log K sets of primary data of (VI.26) for 0.5 – 3.0 M HClO at and 0.5 – 3.0 M HNO 3 4 s 10 20°C (Table VI-34). log K (VI.26) in Table VI-34 obtained in mixed ionic media (1.0 The data of s 10 M HClO ox or 2.0 M HNO /0.08-0.636 M H ox) are rejected, but the /0.08-0.636 M H 4 3 2 2 = 0.5 – 3.0 M HClO are accepted K (VI.26) at I log and 0.5 – 3.0 M HNO values of 4 3 s 10 by this review. These values are for 20°C and are corrected to 25°C with the following method. –1 The enthalpy of Reaction (VI.26) is estimated to be 5.5 kJ·mol by this review, ο –1 (ox)·3H ∆ (UO (selected H 2702 kJ·mol O, s) = − using the enthalpy of formation 2 2 fm 2+ ο 1 − ∆ UO H by this review, Section VI.10.1.2.2), ) = – 1019 kJ·mol , ( [2003GUI/FAN] fm 2 ο –1 ο ∆ ∆ ox, aq) = O, l) = – 285.83 kJ·mol H [89COX/WAG] and (H H (H 2 2 fm fm –1 (820.1 . Section VI.2.3). Therefore the 1.5) kJ·mol − selected in this review ( cf ±

265 VI.10 Uranium oxalate compounds and complexes 223 log K (VI.26) from 20°C to 25°C is estimated to be ~ 0.02 by this review change of 10 s using the van’t Hoff equation, ⎛⎞ ∆ H 11 rm KK −= ×− log log ⎜⎟ 10 10 TT 21 ⋅ TT 2.3023 R ⎝⎠ 12 log K (VI.26) in This value is much smaller than the uncertainty of the 10 s K log (VI.26) at 20°C is taken as those at 25°C e, the values of Table VI-34. Therefor s 10 0.20 to ± by this review without correction, but the uncertainties are increased from ± 0.30 to account for the additional uncertainty the change in temperature could arise. Some of the log (VI.26) are for nitric acid medi a and need to be corrected K s 10 by the complexation of U(VI) by nitrate. in [59MOS/ZAK] mmarised under the entry of The correction processes are su log K (VI.26) are shown in Table VI-35. Appendix A. The corrected values of 10 s Table VI-35: Solubility constants of UO O obtained by re-evaluation of the (ox)·3H 2 2 data in Table VI-34. t Method Ionic medium log Reference K (°C) s 10 2+ + + H U UO (ox)·3H UO O(s) + 2 H O(l) (VI.26) (ox)(aq) + 3 H 2 2 2 2 2 a 25 − (3.39 ± 0.30) 0.513 m HClO [59MOS/ZAK] sol 4 a 25 − (3.65 ± 0.30) 1.051 m HClO 4 a 0.30) 25 − (3.39 ± 1.614 m HClO 4 a 2.204 m HClO 25 − (3.54 ± 0.30) 4 a − 0.30) 25 (3.69 ± 2.824 m HClO 4 a 25 − (3.84 ± 0.30) 3.471 m HClO 4 a,b 0.508 m HNO sol − (3.25 ± 25 [59MOS/ZAK] 0.30) 3 a,b 25 (3.49 ± 0.30 − 1.032 m HNO 3 a,b 25 − (3.70 ± 0.30) 1.572 m HNO 3 a,b 2.129 m HNO − 25 ± 0.31) (3.68 3 a,b ± 25 − (3.86 0.31) 2.707 m HNO 3 a,b − 0.32) 25 (4.09 ± 3.304 m HNO 3 2 − + O(l) (VI.27) (ox)(aq) U + 3 H (ox)·3H + 2 H O(s) + H UO UO (ox) 2 2 2 2 22 0.5-1.0 M HClO sol (20 ± 1) − (1.93 ± 0.12) [59MOS/ZAK] 4 0.5-1.0 M HNO (20 ± 1) − (1.85 ± 0.14) 3 a: The uncertainties are increased by this review from 0.20 to ± 0.30 to account for the adjustment of the ± temperature from 20 to 25°C. b: The constants are corrected fo r the complexation of nitrate (details are provided in Appendix A). The values of (VI.26) in Table VI-35 are used to obtain log K s 10 ο log K (VI.26) at 298.15 K by the SIT approach outlined in Appendix B. With s 10 2 ∆ z (VI.26) = 2,

266 VI Discussion of data selection for oxalate 224 ο D = log K log − – 2 ∆ε (VI.26) × I (VI.31) K m 10 s s 10 = where D ) and I I / (1 + 1.5 I 0.509 is the ionic strength in molality. A plot of m mm log K ( – 2 D ) vs . I is shown in Figure VI-29. A weighted linear regression results in m s 10 the following values: ο K log − (3.62 ± 0.19), (VI.26) = s 10 –1 ± . 0.09) kg·mol (VI.26) = (0.24 ∆ε The ysis (Figure VI-29) is in good ∆ε (VI.26) calculated from the SIT anal 2+ − ε ( ) and ∆ε UO , ClO (VI.26) calculated fr om the values of agreement with the 24 − + ClO , (H ) selected by [2003GUI/FAN] , if the interaction of the neutral species ε 4 2+ − [2003GUI/FAN] , ε ( ox(aq) with the ionic medium is ignored. From ) UO , ClO H 2 24 –1 + − –1 ε (H (VI.26) = , ± ClO 0.03) kg·mol ) = (0.14 ± 0.02) kg·mol and , thus ∆ε (0.46 4 2+ − − + –1 UO , ClO ClO ε (H ) = (0.18 , . ε ( ) – 2 ± 0.05) kg·mol 24 4 ο Because the value of K (VI.26) can also be calculated from other ther- log s 10 ο K log (VI.26) obtained from the modynamic constants selected by this review, the s 10 SIT analysis of the solubility data in Table VI-35, (3.62 − 0.19), can be used for a ± consistency check of the thermodynamic co nstants selected by this review. From 2– + ο ο + ox (VI.25) = – (1.80 ± log K K = (5.65 ± 0.03) for the reaction, 2H 0.27), log s 10 10 ο 2+ log K UO + U H ± = (7.13 ox(aq), ( 0.06) for the reaction, cf . Section VI.3.5), and 2 2 10 2– ox(aq), ( U UO cf . section VI.10.2.4.1), the solubility constant of Reaction (VI.26) ox 2 ο 0 and 298.15 K is calculated to be: 0.28). This value log ± K (3.28 (VI.26) = − → I at s 10 is somewhat higher than that obtained from the SIT analysis of the solubility data in Table VI-35, but the two values overla p within the range of expectancies. Figure VI-29: SIT plot for Reaction (VI.26). ( Ο ) HNO ; ( □ ) HClO 3 4 -2.2 0 = - (3.62 ± 0.19) Κ log 10 s ∆ = 0.24 ± 0.09 ε -2.8 -3.4 -2D s Κ 10 -4.0 log -4.6 -5.2 1.0 3.0 4.0 2.0 0.0 -1 mol·kg / I

267 VI.10 Uranium oxalate compounds and complexes 225 VI.10.2 Aqueous uranium oxalate complexes Aqueous uranium(III) oxalate complexes VI.10.2.1 U(III) is unstable in aqueous solution and is thermodynamically capable of being oxi- , [70PER/KRO] . The complexa- [60ALE/ZHD] , [63JEZ] dised to U(IV) by many anions tion of the anions with U(III) is expected to be weaker than that with U(IV), which IV III could shift the potential of the couple U and facilitate the oxidation of U(III) to /U . Peretrukhin et al. observed that, when ammonium oxalate was [70PER/KRO] U(IV) tion of U(III), U(III) was almost instantane- introduced into a weakly acidic acetate solu ously oxidised to U(IV) while oxalate was reduced to glyoxylic acid (OCHCOOH) . No thermodynamic data for aqueous U(III) oxalates could be identified [70PER/KRO] in the literature by this review. VI.10.2.2 Aqueous uranium(IV) oxalate complexes 2+ 2 − 4 − , are and A few binary aqueous U(IV) oxalates, U(ox) (aq), U(ox) , U(ox) U(ox) 2 3 4 reported in the literature. The stability constants are usually determined in conjunction with the solubility measurements [58GRI/PET2] , , , [60ZAK/MOS] [60GRI/PET] . As discussed previously, the solubility experiments for solid U(IV) oxalates [79NIK] are usually associated with shortcomings including the oxidation of U(IV) during the experiment by oxygen or the product of the radiolysis of water, the strong tendency of e solid phase. Conse- U(IV) toward hydrolysis, and lack of the characterisation of th quently, the assignment of the aqueous U(IV) oxalate complexes and their stability constants based on solubility measurements must be regarded with suspicion, particu- larly if essential experimental conditions are not provided. The stability constants re- and some of the stability constants reported in [60GRI/PET] ported in [58GRI/PET2] for rejection are given in Appendix A. are rejected by this review. The reasons As discussed in Section VI.10.1.1, a variety of stable polynuclear solid U(IV) + + + oxalates have been prepared, ., U e.g , (ox) (ox) n H O (M = H ·12H , Na O, M , K [U ]· 2 2 k j i 4 2 + NH i = 2, 4, 6, 8; j = 2; k = 3, 4, 5, 7, 8; ; = 1, 4, 5, 8, 10), U (ox) (SO ) ·6H O and n 2 2 2 2 4 4 [66CHE] (ox) . SO [91MAT/KAR] ·12H , O [87DOL] , U 2 2 4 3 This may imply that polynuclear U(IV) oxalate complexes could also form in aqueous solutions. For example, in the solubility experiments with K U (ox) ·8H O, the 2 2 5 2 charge on the anion in solution was found to be – 2 by ion exchange and the author − 2 U(ox) , was present in the solution assumed that a dimeric U(IV) oxalate species, 25 [70PET/STE2] . However, this experiment is at fault due to the change in the solid phase during the experiment so that the results are considered unreliable by this review. The − 2 remains to be verified. No other experimental evidence is identi- U(ox) presence of 25 fied in the literature on the formation of polynuclear U(IV) oxalate complexes in aque- ous solution. 2– A ternary complex, [U(ox) ] (OH) , was assumed to form when 2 2 U(ox) ·6H O(s) was dissolved in solutions of ammonium bicarbonate, but no stability 2 2

268 VI Discussion of data selection for oxalate 226 . The stability constant of a ternary complex, [72GRI/PET2] constant was determined − 2 , was reported in [83PER/MIS] but not accepted by this review due to the Uedta(ox) scarce experimental information (Appendix A). Stability constants considered by this review are summarised in Table VI-36. Table VI-36: Stability constants of aqueous U(IV) oxalates reported in the literature. Method Ionic medium C) log Reference K ( ° t 10 − 4+ 2+ 2 U(ox) + ox U U 0.5 M HCl and sol 0.2) ? (8.6 [60ZAK/MOS] ± 0.07-0.241 M (NH ox ) 2 4 9.01 ? ? ? [67KUM/SER] 2+ 2 − U U(ox) (VI.32) (aq) + ox U(ox) 2 a (25 ± 0.1) (8.28 0.735 m HClO 0.20) [79NIK] sol ± 4 (40 ± 0.1) (8.30 ± 0.20) (50 ± ± 0.20) 0.1) (8.36 ± (8.45 ± 0.20) (60 0.1) ± 0.1) (8.48 ± 0.20) (70 ± (8.55 ± 0.20) 0.1) (80 ± 0.1) (8.81 ± 0.20) (90 4+ 2 − U + 2 ox U (aq) U(ox) 2 b 0.5 M HCl and sol ? (16.85 ± 0.20) [60ZAK/MOS] 0.07-0.241 M (NH (17.40 ± ) 0.20) ox 2 4 2 − 2 − U(ox) (VI.33) U(ox) (aq) + ox U 2 3 sol ? [60GRI/PET ? 5.4 a [79NIK] (5.05 ± 0.20) 0.1) ox (25 ± sol 0.01 m Na 2 ± 0.1) ± 0.20) (40 (5.14 (50 ± 0.1) (5.28 ± 0.20) (60 ± 0.1) (5.36 ± 0.20) (70 (5.52 ± 0.20) ± 0.1) (80 0.1) (5.79 ± 0.20) ± (90 ± 0.1) (6.22 ± 0.20) 2 − 4+ 2 − U U + 3 ox U(ox) 3 b sol 0.5 M HCl and ± 0.20) [60ZAK/MOS] ? (22.77 0.07-0.241 M (NH ) (22.58 ± 0.20) ox 2 4 4 − 2 − 4+ U + 4 ox U U(ox) 4 b 0.5 M HCl and sol ? (27.24 ± 0.20) [60ZAK/MOS] 0.07-0.241 M (NH ± ox 0.20) ) (27.38 4 2 pot ? ? 26 [60GRI/PET] a: The uncertainties of 0.10 given in the literature are conisdered to be too optimistic, thus increased ± to ± 0.20 by this review. b: The two values from [60ZAK/MOS] are the results calculated with two different methods from the same experimental data. Details are given in Appendix A.

269 VI.10 Uranium oxalate compounds and complexes 227 Despite the shortcomings associated with the solubility experiments, the stabil- ity constants in Table VI-36 show general agreement between different references and − 2 2+ U(ox) are internally consistent among the series of species: U(ox) (aq), , U(ox) and 2 3 − 4 . For example, the stepwise stability constants at 25°C follow the order: log U(ox) K 1 10 4 K K (8.3) > log (4.6). The overall stability constants K (5) > log (8.6 – 9) > log 3 2 4 10 10 10 also appear to agree with the sum of stepwise constants from different references within the experimental uncertainty. The stability constants, log (VI.32) and log K K (VI.33), determined in 0.735 10 10 m HClO and 0.01 m Na ox at variable temperatures (2 5-90°C), increased as the tem- 4 2 H ∆ (VI.32) and perature was increased (Table VI-36). From these data, the values of rm H ∆ (VI.33) at variable temper atures were calculated in and shown in Table [79NIK] rm –1 VI-37. The uncertainties ( ± ) are assigned by this re- 10 %, but no less than 1 kJ·mol view. 2+ 2 − U U(ox) + ox (aq) (VI.32) U(ox) 2 − 2 2 − U(ox) U(ox) U (VI.33) (aq) + ox 2 3 xperimental enthalpy of Reacti ons (VI.32) and (VI.33) from [79NIK] Table VI-37: E . –1 ( ° C) ) H Method Ionic medium ( kJ·mol t ∆ rm 2+ 2 − + ox U(ox) U U(ox) (aq) (VI.32) 2 sol 0.735 m HClO 1.0) (25 ± 0.1) (2.8 ± 4 (40 0.1) (8.54 ± 1.00) ± ± 0.1) ± 1.0) (50 (13.0 0.1) ± 2.0) ± (17.9 (60 0.1) (25.1 ± (70 ± 2.0) ± 0.1) (29.4 ± 3.0) (80 ± 0.1) (36.0 ± (90 4.0) − 2 − 2 (aq) + ox (VI.33) U(ox) U U(ox) 2 3 1.0) ox (25 ± 0.1) (6.5 ± sol 0.01 m Na 2 (40 ± 0.1) (18.7 ± 2.0) (50 ± 0.1) (28.0 ± 3.0) (60 ± (38.4 ± 4.0) 0.1) ± (53.6 ± 5.0) (70 0.1) ± 0.1) (62.6 ± (80 6.0) (90 ± 0.1) (76.4 ± 8.0) The enthalpy values in Table VI-37 are consistent with the trend of the stability constants in Table VI-36 and Figure VI-30. The plot of log (VI.33) K (VI.32) and log K 10 10 (VI.33) ∆ H H ∆ (VI.32) and (Figure VI-30-a) indicates that as a function of 1/ T rm rm (1/ ) < 0, that T K / d log d should be positive in the temperature range because 10 H ∆ H ∆ (VI.32) and (VI.33) probably do not remain constant because rm rm

270 VI Discussion of data selection for oxalate 228 (VI.33) should become d / d (1/ T ) ≠ constant, and that log H ∆ (VI.32) and K H ∆ rm 10 rm increasingly more positive as the temperature is increased because d T / d (1/ log ) becomes larger as T increases. K 10 The plots of H ∆ (VI.32) and ∆ H (VI.33) as a function of temperature are rm rm shown in Figure VI-30-b. A weighted linear regression results in the values of –1 − 1 –1 –1 56) J·K . ± ·mol (VI.32) = (480 and ·mol ± 33) J·K (VI.33) = (976 ∆ ∆ C C r,m r,m p p In spite of the self-consistency and the apparent reasonableness of the stability constants in Table VI-36 and the enthalpy of complexation in Table VI-37, further evaluation of the data to derive thermodynamic values with SIT for aqueous U(IV) oxalates is not pursued in this review, due to the previously discussed shortcomings associated with the solubility experiments. No thermodynamic data for aqueous U(IV) oxalates are recommended by this review. Figure VI-30: Stability constants (a) and enthalpy (b) for Reactions (VI.32): (O) and (VI.33): ( ∇ ), original data from [79NIK] . 80 10.0 (b) (a) 9.0 60 8.0 -1 K 10 7.0 40 /kJ·mol log m H r ∆ 6.0 20 5.0 0 4.0 300 280 0.28 380 360 340 0.26 320 0.30 0.32 0.34 (E-2) -1 -1 , K T , K T Aqueous uranium(V) oxalate complexes VI.10.2.3 Oxalate complexes of U(V) were identified only as reaction intermediates in the electro- [46KOL/HAR] lytic reduction of U(VI) , [56GRA/GRA] , [60ALE/ZHD] , [2002MAK/MER] , [72MAN/VAR] , , in the reduction of U(VI) by hy- [69PER/KRO] , or in the photolysis of [81NAS/MUL] drated electrons generated with pulse radiolysis U(VI) oxalates [76BRI/ELD2] , [76BRI/ELD3] , [77BRI/ELD] . The presence of oxalate

271 VI.10 Uranium oxalate compounds and complexes 229 VI V IV V /U and U and facilitates the dispro- /U shifts the potentials of the redox couples U , [72MAN/VAR] [2002MAK/MER] , suggesting , [60ALE/ZHD] portionation of U(V) an the U(IV) oxalate that the U(V) oxalate complex is weaker th or U(VI) oxalate com- + 4+ 2+ plexes due to the lower charge density of UO and U . No ex- UO than those of 2 2 perimental stability constants of aqueous U(V) oxalate complexes are identified in the literature. The value of log K of the reaction: 1 10 – + 2– + ox U UO ox UO 2 2 was estimated to be 4.0 in [71MOS6] er the temperature nor by “extrapolation”. Neith the ionic strength was sp ecified. The value is rejected in this review. VI.10.2.4 Aqueous uranium(VI) oxalate complexes Binary U(VI) oxalate complexes VI.10.2.4.1 (2 2 ) j − UO (ox) It is generally accepted that U(VI) forms the complexes of the type: 2 j where j = 1 – 3, in aqueous solutions. Figure VI-31 shows three postulated binding modes of oxalate in U(VI) oxalate complexes, including “side-on” bidentate, “end-on” bidentate and unidentate. Different binding modes are observed in solid U(VI) oxalate ox(aq) and compounds, but not confirmed in solution. While the oxalate ligands in UO 2 2 − UO (ox) are believed to be “side-on” bidentate with both carboxylate groups 22 coordinating to uranium in th ear O=U=O moiety, there are e equatorial plane of the lin − 4 UO (ox) . For of the third oxalate ligand in disagreements about the binding mode 23 − 4 − 2 UO (ox) UO (ox) example, the stepwise formation constants of UO and ox(aq), 2 22 23 = 2.0 M ( I = 3.69 at log K = 6.20, K log log K = 5.01 and [2000FER/IUL] from 10 1 10 3 2 10 ) suggest that all the three oxalates are identically co-ordinated to uranium in the NaClO 4 “side-on” bidentate mode. However, the much smaller third-step constant of − 4 K UO (ox) [69HAV] ( ) could suggest = 1.0 M NaClO log from = 0.36 at I 4 23 3 10 indicate that the most echanical calculations [2003VAL/MOL] otherwise. Quantum m − 4 UO (ox) stable structure of is a five-coordinate isomer with the third oxalate bonded 23 to uranium with a single carboxylate oxygen (unidentate). The theo retical results seem and, are consistent with the results [69HAV] log K from to agree with the small 3 10 from EXAFS [2003VAL/MOL] . However, since the accuracy of some of the EXAFS and a recent re-investigation of the results is not particularly high [2003VAL/MOL] U(VI) oxalate complexation yields a higher value of the third stepwise constant K log ( = 2.59 at I = 3.0 M NaClO ) [2002HAV/SOT] , the question of the binding 4 3 10 − 4 UO (ox) in solution remains to be answered. The enthalpy and modes of oxalate in 23 entropy of complexation that can be determined by calorimetry should help to answer this question. Oxalate forms the strongest complexes with U(VI) among the unsubstituted al- kyldicarboxylates including oxalate, malonate , succinate, glutarate and adipate, mainly

272 VI Discussion of data selection for oxalate 230 n length increases, the ligand suffers greater due to an entropy effect. As the carbon chai loss of rotational and translational freedom upon chelation with U(VI), resulting in a greater entropy loss. A Raman spectroscopic study indicates that the coordination of oxalate and other ligands with O=U=O weakens the axial U–O bond(s) and shifts its . From the magnitude of the symmetrical stretching vibration frequency [92NGU/BEG] Raman shifts, the binding strength of oxalat e with U(VI) was place d between carbonate – – 2 − − − 2 2– – – and fluoride in this order: OH , > ox SO CO > > F CH CO > > Cl > > Br 3 4 32 − − − , which is in good agreement with the order of the formation con- ClO NO , , HSO 4 4 3 stants of the complexes obtained by thermodynamic experiments. Figure VI-31: Postulated binding modes of oxalate in U(VI) complexes. (a) “side-on” bidentate; (b) “end-on” bidentate; (c) unidentate. The large atom represents uranium and the oxalates coordinate to uranium in the equatorial plane. The two axial U=O bonds are perpendicular to the plane. (a) (b) (c) U(VI) oxalate complexes in solution undergo photochemical reactions when exposed to UV light or the sunlight [1862SEE] , [42HEI] , [76BRI/ELD2] , . The photolysis leads to the reduction of U(VI) to U(IV), , [77BRI/ELD] [76BRI/ELD3] accompanied by the release of CO and CO [76BRI/ELD2] , [76BRI/ELD3] , 2 2+ . The photosensitive species were found to be UO , UO H ox(aq) and ox [77BRI/ELD] 2 2 2

273 VI.10 Uranium oxalate compounds and complexes 231 2 − [42HEI] , UO (Hox) in acidic solutions (aq) in solutions with pH ≤ 1.7 UO (ox) 2 2 22 − 2 [76BRI/ELD3] UO (ox) , and in solutions with pH = 4.5 , [76BRI/ELD2] 22 [77BRI/ELD] . The kinetics and mechanism of the photochemical reactions were re- [76BRI/ELD3] , [77BRI/ELD] , . The possibility of photolysis ported [76BRI/ELD2] certainly raises concerns on the reliability of the experimental data if the solutions of U(VI) oxalate were exposed to light for a prolonged time. Precautions were taken in some studies to avoid or reduce the exposure of the U(VI) oxalate solutions to sunlight , . [69HAV] or UV light [67MIY/NUR] The polynuclear and ternary or quaterna ry U(VI) oxalate complexes are dis- cussed in Sections VI.10.2.4.2 and VI.10.2.4.3. Table VI-38 lists only the stability con- stants of the mononuclear binary U(VI) oxalate complexes reported in the literature. Table VI-38: Stability constants of aqueous U(VI) oxalates reported in the literature. ( ° C) log K Method Ionic medium Reference t 10 2+ 2 − + ox UO UO U ox(aq) 2 2 2 − 2+ – sp, pot / ≈ I SO UO /Hox 0.05, 0.08 M 25–26 5.82 [42HEI] 4 2 I /HNO sol (20 ± 1) (6.8 ± 0.1) [59MOS/ZAK] = 0.5–1.0 M HClO 4 3 → 0 (25.0 ise-Ag 0.1) 6.0 [59PTI/TEK] I ± I = 1.0 M NaClO sp (4.63 (20 ± 0.2) ± 0.02) [67MIY/NUR] 4 em = 0.1 N KCl 25 (6.7 ± 0.2) [67STE/MAK2] I [69HAV] 0.07) 20 (6.36 ± sp I = 0.1 M NaClO 4 I = 1.0 M NaClO 20 (5.99 ± 0.03) 4 [72MAN/VAR] 20 5.0 = 0.2 M NaClO pol I 4 [79KUM/CHA] ± 1) (3.97 ± 0.27) (25 I pot = 0.10 M KNO 3 = 4.0 M (1 M HNO I dis ) 25 (6.20 ± 0.06) [83CHO/BOK] /3 M NaNO 3 3 ± /2 M NaNO ) 25 (6.22 0.07) = 4.0 M (2 M HNO I 3 3 ± /1 M NaNO 0.09) ) 25 (6.26 = 4.0 M (3 M HNO I 3 3 = 4.0 M HNO I 25 (6.36 ± 0.06) 3 25 4.48 [85VEN/SWA] = 0.1 M KNO I pot 3 /NaClO [89ABD/ALI] (20 ± 1) 3.22 I = 0.05 M H ClO sp 4 4 I = 0.11 M NaClO (including 0.01 M 4 ix 25? (6.10 0.11) [91LOO/KOP] ± acetate) (6.00 25 [94ERT/MOH] ± 0.05) I = 3.0 M NaClO dis 4 0.28) ± (6.55 I = 5.0 M NaClO 4 I = 7.0 M NaClO (6.55 ± 0.33) 4 (7.37 ± 0.05) I = 9.0 M NaClO 4 = 5.0 m NaCl 25 (5.50 ± I [96BOR/LIS] dis 0.02) pot (gl. I = 2.0 M NaClO [2000FER/IUL] (25.00 ± 0.02) (6.20 ± 0.02) 4 ise-ox) I = 3.0 M NaClO (6.39 ± 0.01) 4 (Continued on next page)

274 VI Discussion of data selection for oxalate 232 Table VI-38: (continued) t C) log ( K Reference Method Ionic medium ° 10 2+ − 2 U UO UO ox(aq) + ox 2 2 = 1.0 M NaClO pot (gl. (25.00 ± 0.02) (6.03 ± 0.03) [2000VAS/CAR] I 4 ise-ox) 25 (5.94 ± = 0.3 m NaCl [2001BOR/MOO] dis I 0.01) = 1.0 m NaCl (5.92 ± 0.01) I I (5.89 ± 0.01) = 2.0 m NaCl = 3.0 m NaCl (6.61 ± 0.02) I = 4.0 m NaCl (6.70 ± 0.01) I = 5.0 m NaCl (5.82 I ± 0.02) ± (25.0 ± 0.2) (6.31 [2002HAV/SOT] 0.02) I sp = 3.0 M NaClO 4 2+ − 2 2 − U + 2 ox UO (ox) UO 22 2 2+ 2 − – ≈ UO / 0.05, 0.08 M SO /Hox I sp, pot 25–26 10.56 [42HEI] 2 4 1) sol /HNO (20 ± = 0.5–1.0 M HClO (12.0 ± 0.2) [59MOS/ZAK] I 3 4 → 0 (25.0 ± 0.1) (10.8 ± 0.5) [59PTI/TEK] ise-Ag I I dis = 0.1 M NaClO ? (11.08 ± 0.03) [60STA3] 4 [67MIY/NUR] (20 ± 0.2) (8.7 ± 0.05) I sp = 1.0 M NaClO 4 0.2) ± [67RAJ/MAR] ± 0.05) (9.1 (25.0 = 1.0 M KNO I pot 3 em I 25 (11.8 ± 0.3) [67STE/MAK2] = 0.1 N KCl 20 (10.59 ± 0.07) [69HAV] I sp = 0.1 M NaClO 4 20 (10.64 ± 0.06) I = 1.0 M NaClO 4 I = 0.2 M NaClO pol 20 9.8 [72MAN/VAR] 4 [76BRI/ELD] 0.25) 25 (9.36 ± = 0.5 M (H/K)NO I sp, pol 3 (9.08 ± (25 1) [79KUM/CHA] ± 0.25) I pot = 0.10 M KNO 3 I em = 0.1 M NaClO 40 6.47 [84SIR] 4 25 8.43 [85VEN/SWA] I = 0.1 M KNO pot 3 ± 25 (11.21 0.18) [94ERT/MOH] = 3.0 M NaClO dis I 4 I = 5.0 M NaClO (12.28 ± 0.11) 4 (12.94 ± 0.40) = 7.0 M NaClO I 4 (14.08 ± 0.11) = 9.0 M NaClO I 4 I = 5.0 m NaCl 25 (10.60 ± 0.02) [96BOR/LIS] dis [2000FER/IUL] 0.02) ± (11.21 (25.00 ± 0.02) = 2.0 M NaClO I pot (gl. 4 ise-ox) (11.52 ± 0.02) I = 3.0 M NaClO 4 [2000VAS/CAR] 0.07) ± (10.87 (25.00 ± 0.02) I pot (gl. = 1.0 M NaClO 4 ise-ox) I = 0.3 m NaCl 25 (10.1 ± 0.06) [2001BOR/MOO] dis I = 2.0 m NaCl (10.2 ± 0.08) (Continued on next page)

275 VI.10 Uranium oxalate compounds and complexes 233 Table VI-38: (continued) Method Ionic medium ° C) log t K Reference ( 10 2 2+ − 2 − U UO (ox) UO + 2 ox 2 22 0.07) ± = 3.0 m NaCl [2001BOR/MOO] dis I (11.0 0.07) (25.0 ± 0.2) (11.21 ± = 3.0 M NaClO [2002HAV/SOT] I sp 4 2 − 2 − U (ox)(aq) + ox UO UO (ox) 2 22 –1 con [56GRI/PTI] I < 0.1 mol·kg 25 4.46 − 4 2+ 2 − U UO (ox) + 3 ox UO 23 2 0.2) I sp ± = 1.0 M NaClO (11.99 ± 0.11) [67MIY/NUR] (20 4 sp = 1.0 M NaClO 20 (11.0 ± 0.2) [69HAV] I 4 (25 ± 1) (16.40 ± 0.12) [79KUM/CHA] I pot = 0.10 M KNO 3 [2000FER/IUL] 0.01) (25.00 ± 0.02) (14.9 ± I pot (gl. = 2.0 M NaClO 4 ise-ox) (15.2 ± 0.01) = 3.0 M NaClO I 4 [2000VAS/CAR] 0.3) (14.0 (25.00 ± 0.02) ± I = 1.0 M NaClO pot (gl. 4 ise-ox) ± (25.0 ± 0.2) (13.8 [2002HAV/SOT] 0.04) = 3.0 M NaClO I sp 4 2+ + 2– + + ox UO U UO + H (Hox) 2 2 I = 3.0 M NaClO dis (7.57 ± 0.06) [94ERT/MOH] 25 4 I = 5.0 M NaClO (9.22 ± 0.17) 4 I = 7.0 M NaClO 0.07) (10.43 ± 4 (10.06 ± 0.02) = 9.0 M NaClO I 4 2+ – + + Hox U UO (Hox) UO 2 2 1) I (25 ix = 0.16 M HClO 3.40 [57LI/WES] ± 4 I = 1.0 M HClO 2.83 4 I = 2.0 M HClO 2.89 4 2+ – UO + 2 Hox U UO (aq) (Hox) 2 2 2 I = 0.5 M (H/K)NO sp, pol 25 (6.00 ± 0.08) [76BRI/ELD] 3 5.96 = 0.16 M HClO ix (25 ± 1) I [57LI/WES] 4 4.67 I = 1.0 M HClO 4 = 2.0 M HClO I 4.73 4 2+ + 2– UO + 2 ox (aq) U UO + 2 H (Hox) 2 2 2 (18.86 = 5.0 M NaClO dis 25 I ± 0.15) [94ERT/MOH] 4 I = 7.0 M NaClO (20.04 ± 0.42) 4 (19.48 0.59) ± I = 9.0 M NaClO 4 2+ 2+ ox) (H (ox)(aq) U UO + H UO 2 2 2 2 2+ 2 − – sp, pot ≈ [42HEI] I 25–26 2.57 0.05, 0.08 M /Hox / SO UO 4 2

276 VI Discussion of data selection for oxalate 234 As the data in Table VI-38 show, the stability constants of UO ox(aq), 2 4 − − 2 and UO (ox) are quite scattered: from 3.22 to 7.37, β log β UO (ox) log 2 1 10 10 22 23 from 6.47 to 14.08 and from 11.0 to 16.4. Critical reviews of the papers where log β 3 10 the data are reported lead to the following findings: Systematic errors may be associat ed with some data, including the from i. β log 1 10 [79KUM/CHA] [85VEN/SWA] , [89ABD/ALI] , and the , [84SIR] from β log 2 10 . These data are significantly lower than the majority of the data and [67RAJ/MAR] for the same complexes in the literature. They also seem too low in comparison with the data of relevant systems such as lanthanide oxalate complexes or U(VI) complexes with other dicarboxylates in the literature. ii. There are various shortcomings in some studies so that the data (or a portion of the hortcomings are: a) unknown or errone- data) are questionable. Examples of the s ous dissociation constants of oxalic acid [67MIY/NUR] , , [67STE/MAK2] [85VEN/SWA] , [89ABD/ALI] ; b) lack of experimental details , [72MAN/VAR] such as the electrode calibration for pH/pcH calibration, [67RAJ/MAR] , , [76BRI/ELD] , [79KUM/CHA] , [85VEN/SWA] , [89ABD/ALI] ; [72MAN/VAR] [56GRI/PTI] , , c) varying ionic strengths or mixed electrolyte media [42HEI] [59MOS/ZAK] , [59PTI/TEK] [72MAN/VAR] , [76BRI/ELD] , [83CHO/BOK] , , includes questionable species or ex- ; d) incorrect models that either [89ABD/ALI] [76BRI/ELD] . , [72MAN/VAR] cludes reasonable species iii. The existence of the protonated U(VI) oxalate complexes that were assumed to + 2+ , UO , (Hox) (Hox) (aq) and UO (H ox) have formed in some experiments, UO 2 2 2 2 2 is still open for debate. These species, difficult to identify by physical methods, were usually postulated to improve the fitting of potentiometric or spectropho- tometric data. If they form, it is most likely to occur in strongly acidic solutions. There are contradictory results in the literature on the protonated U(VI) oxalate complexes. For example, while Heidt and others found the photosensitive species 2+ [42HEI] ox) (H . implies , the work by Pitzer et al in acidic solutions included UO 2 2 − 2 . ox(aq) and UO (ox) [36PIT/GOR] complexes that no hydrogen entered the UO 2 22 The results from solvent extraction with TBP indicate that no protonated U(VI) oxalate complexes formed in 1 – 4 M HNO [83CHO/BOK] , but the data from 3 solvent extraction with TTA and HEDHP were fitted with four complexes includ- + ing UO (Hox) (Hox) (aq) in a pH range of 1.6 to 3.5 [94ERT/MOH] . As and UO 2 2 2 the detailed discussions in Appendix A indicate, the stability constants of + (Hox) , UO are not reliable due to the scar- (Hox) [94ERT/MOH] (aq) from UO 2 2 2 city and scatter of the experimental data. Based on these findings, the data in Ta obably associated ble VI-38 that are pr with systematic errors or various shortcom ings are not accepted by this review. These include the data from [42HEI] , [56GRI/PTI] , [59MOS/ZAK] , [59PTI/TEK] , , [67RAJ/MAR] , [67STE/MAK2] , [72MAN/VAR] , [79KUM/CHA] , [67MIY/NUR]

277 VI.10 Uranium oxalate compounds and complexes 235 , [85VEN/SWA] [89ABD/ALI] , and a portion of data from [83CHO/BOK] , [84SIR] , . Detailed comments on the individual papers are given in Appendix A. [94ERT/MOH] The reasonableness of the stability constants of the protonated U(VI) oxalate complexes in Table VI-38 can be qualitatively tested by using the combination of the ox(aq) and the dissociation constants of oxalic acid. Such tests stability constants of UO 2 indicate that the stability constants for UO (Hox) [94ERT/MOH] seem too (aq) from 2 2 + 2+ [57LI/WES] , UO (Hox) (Hox) (aq) and UO , (H from ox) high, but the data for UO 2 2 2 2 2 [42HEI] and appear reasonable and self-consistent. However, taking into [76BRI/ELD] consideration of the contradictory information in the literature on the presence of these + 2+ complexes, the stability constants of UO , UO in (Hox) (Hox) (aq) and UO ox) (H 2 2 2 2 2 Table VI-38 are not accepted by this review. In addition, the stability constants in Table VI-38 that are determined at ionic –1 are not included in the further analysis with SIT by this strengths higher than 5 mol·kg review. Table VI-39 shows the stability constants of aqueous U(VI) oxalate complexes that are accepted and used for the SIT analys is by this review. The ionic strengths in molarity are converted into molality based on the information in Appendix B. The sta- bility constants on the molarity scale are also converted to the values on the molality New uncertainties are assigned to the data scale according to Eq. ( II.37) in Section II.2. e given in the original paper, rules: 1) If the uncertainties ar according to the following σ . However, if the uncertainties given in the they are usually doubled to represent 2 are and , paper appear too optimistic, the uncertainties of β log β log β log 10 10 1 2 10 3 ± ± 0.2, and ± 0.3, respectively. 2) If the uncertainties are not given 0.1, adjusted to be in the original paper, ± 0.2, ± 0.3, and ± 0.4 are assigned to and , log β β log 2 10 1 10 , respectively. log β 10 3 2 − The stability constants of UO ox(aq) and from [60STA3] and UO (ox) 2 22 [69HAV] were originally obtained at temperatures other than 25°C (Table VI-38). In the SIT analysis, they order to include these data in should be converted to the values at 25°C, for example, using the van’t Hoff equation with the enthalpy of complexation ( . cf Section VI.10.1.2.2.3). The enthalpy of complexation for UO ox(aq), 2 − 2 2+ UO U UO ox(aq) (VI.34) + ox 2 2 H ∆ ox(aq), (VI.34) can be estimated from the enthalpy of formation of UO 2 rm –1 (1824.3 18.3) kJ·mol ( cf . Section VI.10.1.2.2.3), the enthalpy of formation of − ± 2+ –1 , and the enthalpy of formation − (1019.000 ± 1.5000) kJ·mol UO [2003GUI/FAN] , 2 –1 2 − of ox , (830.7 ± 1.6) kJ·mol ( cf . Section VI.3.5): − –1 H ∆ (VI.34) = (25.4 ± 18.4) kJ·mol . rm 2 − Because of the absence of the enthalpy of complexation for UO (ox) , 22

278 VI Discussion of data selection for oxalate 236 2 2+ − − 2 U UO (ox) (VI.35) UO + 2 ox 22 2 it is assumed by this review to be equal to the enthalpy of complexation for the analo- gous U(VI) malonate complex as an approximation, − 2+ − 2 2 UO + 2 mal UO (mal) U (VI.36) 2 22 –1 Thus, [2002RAO/JIA] (VI.35) ∆ H ∆ (VI.36) = (11 ± 1) kJ·mol H ≈ . rm rm 2 − UO (ox) Using the enthalpies of complexation for UO ox(aq) and , the val- 2 22 (VI.35) at other temperatur log log β (VI.34) and es are converted to those β ues of 1 10 2 10 at 25°C. The converted values are shown in Table VI-39. Detailed information on the conversion is provided under the entries of [60STA3] in Appendix A. [69HAV] and The value of log was determined in 4 M (VI.34) from [83CHO/BOK] β 1 10 and and the values of [96BOR/LIS] (VI.35) from log β (VI.34) and β log HNO 3 2 1 10 10 [2001BOR/MOO] were determined in NaCl of variable ionic strengths. Thus, the values should be corrected for the complexation by nitrate or chloride. Detailed information on , [96BOR/LIS] and the corrections was given under the entries of [83CHO/BOK] [2001BOR/MOO] in Appendix A. The corrected stability constants are shown in Table VI-39. (VI.34) and [94ERT/MOH] log β are (VI.35) from β log The values of 2 10 1 10 + also re-valuated using the model that excludes the protonated complexes UO (Hox) , 2 (aq), because the latter complexes are unlikely to form in the experi- (Hox) and UO 2 2 are not sufficient to allow the [94ERT/MOH] ments at pH 1.6 to 3.5 and the data in calculation of such complexes. Details on the re-evaluation are given in Appendix A. The values after the re-evaluation are shown in Table VI-39. Table VI-39: Stability constants of aqueous uranium(VI) oxalates at 25°C accepted by this review. a K Reference log Method Ionic medium 10 2+ 2 − + ox U UO UO ox(aq) 2 2 b,c [69HAV] sp ± 0.19) I = 0.101 m NaClO (6.42 4 b,c = 1.05 m NaClO I (6.03 ± 0.11) 4 b,d dis I = 4.56 m HNO ± (7.25 [83CHO/BOK] 0.30) 3 e b I ix = 0.111 m NaClO (6.10 [91LOO/KOP] 0.20) ± 4 b,f = 3.50 m NaClO I dis (6.06 0.22) [94ERT/MOH] ± 4 g dis I = 5.0 m NaCl (6.92 ± 0.76) [96BOR/LIS] b pot (gl. ise-ox) I = 2.21 m NaClO ± 0.10) (6.16 [2000FER/IUL] 4 b I = 3.50 m NaClO (6.32 ± 0.10) 4 b I = 1.05 m NaClO pot (gl. ise-ox) (6.01 ± 0.10) [2000VAS/CAR] 4 (Continued on next page)

279 VI.10 Uranium oxalate compounds and complexes 237 Table VI-39 (continued) a K Reference Method Ionic medium log 10 2+ − 2 U UO UO ox(aq) + ox 2 2 g ± (5.99 = 0.3 m NaCl [2001BOR/MOO] dis I 0.10) g = 1.0 m NaCl ± 0.10) I (6.05 g (6.16 ± 0.10) I = 2.0 m NaCl g I 0.18) ± = 3.0 m NaCl (7.13 g = 4.0 m NaCl I ± 0.45) (7.60 g (7.24 ± 0.76) I = 5.0 m NaCl b I = 3.50 m NaClO sp ± 0.10) (6.24 [2002HAV/SOT] 4 2+ − 2 2 − U + 2 ox UO (ox) UO 22 2 b,c (11.11 ± 0.11) I dis [60STA3] = 0.101 m NaClO 4 b,c = 0.101 m NaClO (10.62 ± 0.19) sp I [69HAV] 4 b,c I = 1.05 m NaClO ± 0.17) (10.65 4 b,f I = 3.50 m NaClO dis ± 0.18) [94ERT/MOH] (10.95 4 g dis I = 5.0 m NaCl (12.02 ± 0.76) [96BOR/LIS] b pot (gl. ise-ox) = 2.21 m NaClO I ± 0.20) [2000FER/IUL] (11.12 4 b I = 3.50 m NaClO 0.20) ± (11.39 4 b pot (gl. ise-ox) I = 1.05 m NaClO 0.14) ± [2000VAS/CAR] (10.83 4 g = 0.3 m NaCl (10.15 ± 0.12) dis I [2001BOR/MOO] g I (10.47 ± 0.17) = 2.0 m NaCl g = 3.0 m NaCl ± 0.22) I (11.52 b = 3.50 m NaClO sp I ± 0.14) [2002HAV/SOT] (11.08 4 − 4 2+ − 2 + 3 ox UO U UO (ox) 23 2 b = 2.21 m NaClO pot (gl. ise-ox) (14.8 ± 0.3) I [2000FER/IUL] 4 b I = 3.50 m NaClO ± 0.3) (15.0 4 b I = 1.05 m NaClO pot (gl. ise-ox) ± (13.9 [2000VAS/CAR] 0.6) 4 b = 3.50 m NaClO sp I ± 0.3) (13.6 [2002HAV/SOT] 4 a: Converted from molarity to molalit y following the method in Chapter II. b: Converted from the units of molarity to the units of molality, following the method in Chapter II. The uncertainties are adju sted (Appendix C). c: Corrected from 20°C to 25°C, see Appendix A. d: Corrected for the complexation of nitrate, see Appendix A. e: The medium contains ~ 10% oxalate. by excluding the protonated U(VI) lation of the values in [94ERT/MOH] f: Obtained from the recalcu oxalate complexes. g: Corrected for the complexation of chloride, see Appendix A. The re-evaluated values in Table VI-39 are used to obtain the stability con- stants of U(VI) oxalates at I = 0 and 298.15 K with the SIT approach outlined in Ap- pendix B.

280 VI Discussion of data selection for oxalate 238 The formation of the U(VI) oxalates is represented by the following equilibria: − 2 2 2+ (VI.37) UO + ox ox(aq) ∆ z UO = – 8 U 2 2 2 − 2+ 2 − 2 UO UO (ox) (VI.38) + 2 ox U z = – 8 ∆ 2 22 − − 4 2+ 2 2 UO + 3 ox (VI.39) UO (ox) U z = 0 ∆ 23 2 For Reactions (VI.37) and (VI.38), data in Table VI-39 are fitted with Eq.(VI.40) and (VI.41): ο log β + 8 D = I log β − (VI.40) ∆ε × 1 m 1 10 1 10 ο β + 8 D = (VI.41) I log β log − ∆ε × 2 m 10 2 10 2 β log = where D 0.509 I I and I is the ionic strength in molality. / (1 + 1.5 ) m 1 10 mm ο –1 ) while log I and (0.1 – 5 mol·kg are the stability constants at β log β and m 10 10 2 1 2+ ο + – β are the stability constants at I , = 0. UO ( = ε (UO ε ox(aq), Na log + X ∆ε ) – 2 m 1 2 2 10 − 2 2+ 2– + + – 2– – + UO (ox) UO X (Na ) where (Na , ox , Na ) – ) – ε ( ) and ∆ε ε , X = ) – 2 ε ε ( , ox 2 2 22 – − – − , Cl ClO , or . N O represents X 4 3 certainties as the weighing factors) and unweighted Both weighted (using the un linear regressions were performed on the data with Eq.(VI.40) and Eq.(VI.41). Slightly different results are obtained as shown in Table VI-40: ο ο β log β log Table VI-40: Calculation of and by weighted and unweighted linear 1 2 10 10 regression. Weighted Unweighted ο ± log β (VI.37) (7.16 ± 0.02) (7.13 0.16) 10 1 ∆ε – (0.41 ± 0.02) – (0.43 ± 0.05) 1 ο ± log (11.65 β (VI.38) (11.68 ± 0.07) 0.15) 2 10 0.06) ± – (0.44 ± 0.03) – (0.48 ∆ε 2 The results from the unweighted linear re gression are selected by this review because of two reasons: 1) the weighted regr ession is based on the uncertainties of the individual data points. However, the uncertainties of the data in Table VI-39 are assigned by this review based on very limited or no information in the papers. They do not necessarily reflect the actual uncertainties of the original data. 2) the values of ο and β obtained by the weighted and unweighted agree within the error limits. log ∆ε 10 The smaller error limits obtained from the weighted regression might be too optimistic. Figure VI-32 and Figure VI-33 show the unweighted regression and the propagation of the error limits from I = 0.

281 VI.10 Uranium oxalate compounds and complexes 239 Figure VI-32: SIT plot of Reaction (VI.37) ( ) ( ̈ ) [69HAV] , ( ­ ) log β 10 1 ) ) z U ) [94ERT/MOH] , ( S [91LOO/KOP] [96BOR/LIS] , ( { ) , ( , ( [83CHO/BOK] [2000FER/IUL] ) [2000VAS/CAR] , ( ‘ ) , ( , ( … ) [2002HAV/SOT] . V [2001BOR/MOO] 12.0 0 β log = 7.13 ± 0.16 1 11.0 ε ∆ = -(0.43 ± 0.05) 1 10.0 + 8D 1 9.0 β log 8.0 7.0 6.0 3.0 2.0 6.0 0.0 4.0 5.0 1.0 -1 I / mol·kg Figure VI-33: SIT plot of Reaction (VI.38) ( ̈ ) ­ log β ) ( [60STA3] ) [69HAV] , ( , 2 10 , z ) [94ERT/MOH] , ( U ) [96BOR/LIS] , ( S ) [2000FER/IUL] , ( { ) [2000VAS/CAR] ( ( ) [2001BOR/MOO] , ( ‘ V ) [2002HAV/SOT] . 16.0 0 β = 11.65 ± 0.15 log 2 = -(0.48 ± 0.06) ε ∆ 15.0 2 14.0 + 8D 2 β 13.0 log 12.0 11.0 4.0 5.0 6.0 0.0 1.0 2.0 3.0 -1 mol·kg / I

282 VI Discussion of data selection for oxalate 240 include those obtained in different ionic Since the data used in the SIT analysis the above analysis implies that the ion media: perchlorate, chloride and nitrate, 2+ – + – – UO interaction constants ( ox(aq), Na , X + X ) are the same when X ) and ε is ε (UO 2 2 – − − 2+ – , Cl ClO . This implication should be correct for ε ( , or O UO , X N ) because the 2 3 4 – 2+ − with Cl UO NO has been corrected (see Appendix A) so that and complexation of 2 3 2+ − 2+ − 2+ – –1 UO NO UO UO ClO ) = , Cl ) = ε ( ( , , ( ε ε ) = (0.46 ± 0.03) kg·mol 2 2 2 3 4 . It should also be noted that one constant from the HNO medium [2003GUI/FAN] 3 + is included in the SIT analysis together with other data from the Na [83CHO/BOK] 2+ 2+ – UO UO medium. From the values of ( and ∆ε ∆ε , Cl in Table VI-40, using ) = ε , ε ( 2 1 2 2 − 2+ 2 − − 1 + − UO , N ClO ) = ) = (0.46 ± 0.03) kg·mol ε ( [2003GUI/FAN] and ε (Na O , ox ) = 2 4 3 − 1 (0.08 ± ( cf . Section VI.3.5), two new ion interaction parameters can be − 0.01) kg·mol derived and selected by this review: 1 + – − ox(aq), Na ) = − (0.05 ± 0.06) kg·mol (UO ε , + X 2 – – − − where X NO or ClO = Cl , 3 4 2 − − + 1 UO (ox) , Na − (0.18 ± 0.07) kg·mol ( ε . ) = 22 4 − Only four values of the stability constants of from the literature are UO (ox) 23 –1 . The values seem I accepted in Table VI-39, with a range of from 1.05 to 3.50 mol·kg to be in discrepancy and belong to two parent distributions ((14.8 0.3) and ± ± ± 0.3)), ((13.9 0.6) and (13.6 ± 0.3)). Such data do not allow a meaningful re- (15.0 ο and log for Reaction (VI.39). β ∆ε gression of Eq.(VI.42) by SIT to derive the 3 10 3 ο log β β log D = (VI.42) − ∆ε I + 0 × 3 m 10 10 3 3 2– + – 4 2+ + − (Na , Na ) – ε ( ∆ε ) UO , X ( ) – 3 ε UO (ox) ε , ox = 3 2 23 As a result, an analogous approach described below is taken in this review to − 4 ο + β log UO (ox) estimate ≈ ε . Assuming and calculate the , Na ∆ε ) ( 3 23 10 3 4 − + − 1 ) = ε − (0.01 ± 0.11) kg·mol UO (CO ) ( [2003GUI/FAN] and using the values , Na 233 2+ – − 1 – − − – ± 0.03) kg·mol ε where X UO = , X ClO ( , Cl ) = (0.46 , or NO of 3 4 2 + 2 − − 1 [2003GUI/FAN] ε (Na ) = − (0.08 ± 0.01) kg·mol , ox and ( cf . Section VI.3.5), the − 1 ο are (0.23 ± 0.12) kg·mol ∆ε is calculated to be . Thus, values of β − log value of 3 3 10 and summarised in Table VI-41: log β obtained by Eq. (VI.42) from the 10 3 ο log β at 298.25 K. Table VI-41: Calculation of 3 10 ο − 1 ) Reference β log β ∆ε I × I (mol·kg log 3 3 3 10 10 (14.3 ± − 2.21 (14.8 ± 0.22) 0.3) ± 0.4) [2000FER/IUL] (0.51 3.50 (15.0 ± 0.3) − (0.80 ± 0.35) (14.2 ± 0.5) (13.7 − (0.24 ± 0.10) 0.6) ± 0.6) [2000VAS/CAR] 1.05 (13.9 ± 3.50 (13.6 ± 0.3) − (0.80 ± 0.35) (12.8 ± 0.5) [2002HAV/SOT]

283 VI.10 Uranium oxalate compounds and complexes 241 ο log β is (13.8 ± 1.5). A large uncertainty ( ± 1.5) An unweighted average of 10 3 was assigned to cover the whole range of expectancy of all the four values. from the literature and the analysis Based on the accepted stability constants with the SIT approach, the following stability constants of U(VI) oxalates are selected by this review (Table VI-42). Table VI-42: Selected stability constants of aqueous U(VI) oxalates at 298.15 K. ο − 1 ⋅ mol β log , kg ∆ε Reaction 10 2+ 2 − UO + ox UO U ox(aq) (7.13 ± 0.16) – (0.43 ± 0.05) 2 2 2+ 2 − 2 − UO + 2 ox U ± UO (ox) 0.06) (11.65 ± 0.15) − (0.48 22 2 − 4 2+ 2 − U UO UO (ox) + 3 ox (13.8 ± 1.5) − (0.23 ± 0.10) 2 23 From the selected reaction data, the fo llowing Gibbs energies of formation are calculated: –1 ο ∆ (UO ox, aq, 298.15 K) = – (1673.4 ± 2.7) kJ·mol G 2 fm ο 2 − –1 G ∆ UO (ox) ( , 298.15 K) = – (2379.3 ± 4.1) kJ·mol 22 fm ο 4 − –1 ∆ UO (ox) ( G ± 10.3) kJ·mol , 298.15 K) = – (3071.7 . 23 fm ο ο From the selected ∆ (UO ox, aq, ox, aq, 298.15 K) and (UO G H ∆ 2 2 fm fm 298.15 K), the following standard molar entropy of this species is calculated: –1 ο –1 ± 62.0) J·K (UO ·mol S . ox, aq, 298.15 K) = (171.0 2 m Polynuclear U(VI) oxalate complexes VI.10.2.4.2 A few dinuclear U(VI) oxalate complexes were assumed to exist in solutions, including 2 − 6 − , and [59TEK/VIN2] , [2002HAV/SOT] , [56GRI/PTI] (UO ) (ox) (UO ) (ox) 5 22 22 3 , . The presence of such complexes seems to be [2002HAV/SOT] [59TEK/VIN2] [56GRI/PTI] supported by the data of electric conductance , and the [2002HAV/SOT] . Also, the stability constants of data of vapor pressure osmometry [2002HAV/SOT] 2 − 6 − (UO ) (ox) (UO ) (ox) reported in the literature appear reasonable (Table and 22 5 22 3 VI-43). However, the conditions in the vapour pressure and conductance measurements were not identical to those in the spectrophotometric experiments based on which the stability constants of dinuclear species were calculated. The concentrations of oxalate in the spectrophotometric experiments were relatively low. At such concentrations, both 2 − − 6 (UO ) (ox) species are minor (< 5%). The inclusion of such and (UO ) (ox) the 22 3 22 5 species only slightly improved the overall fit for the spectrophotometric data [2002HAV/SOT] . It is evident that more studies ar e needed to confirm the presence of such complexes in solution and obtain reliable stability constants. Therefore, the stabil- ity constants of dinuclear U(V I) oxalates in Table VI-43 are not accepted by this review.

284 VI Discussion of data selection for oxalate 242 Table VI-43: Stability constants of dinuclear U(VI) oxalates reported in the literature. Method Ionic medium ( K Reference t ° C) log 10 2 − 2+ 2– UO + 3 ox (UO ) ox U 2 22 3 2 sp (25.0 ± 0.2) (18.5 ± 0.2) [2002HAV/SOT] I = 3.0 M NaClO 4 2 − 2 − 2– UO ox 2 (UO ) ox U + ox 22 3 22 ise = 0.022 M 25 1.3 [59TEK/VIN2] I = 0.008 M 25 – 0.5 [59TEK/VIN2] I − 6 2+ 2– U (UO ) (ox) 2 + 5 ox UO 5 22 2 [2002HAV/SOT] = 3.0 M NaClO (25.0 ± 0.2) (28.5 I 0.1) sp ± 4 − 6 2 − 2– 2 U (UO ) (ox) + ox UO (ox) 22 5 22 –1 con 25 6.30 [56GRI/PTI] < 0.1 mol·kg I 6 − − 2 2– (UO ) (ox) U (UO ) (ox) + 2 ox 5 22 3 22 ise I = 0.069 M 25 4.4 [59TEK/VIN2] Ternary U(VI) oxalate complexes VI.10.2.4.3 A variety of ternary U(VI) oxalate complexes were assumed to exist in solution, includ- – 6 − 4 − , (OH)ox (UO ) (OH) (ox) (UO ) (OH) (ox) [56GRI/PTI] and ing UO 2 2 22 4 22 4 2 3 − − 2 − 4 UO ox(SO ) UO oxSO UO F ox [2000PAL] [68SHC/BEL] and , , 232 23 23 2 4– − 2– UO CO ox where , [61CHE/GOL] , UO (L) ox(L) , and UO (ox) [67ZAK/ORL2] 2 2 2 23 − 4 2– where L = aminodiacetate and ox(L) UO ox(L) , UO [79KUM/CHA] L = malonate 2 22 3– 2– [79KUM/CHA] , UO where L = nitrilotriacetate [84SIR] , UO where L = ox(L) ox(L) 2 2 . However, only a few papers report stability constants for phthalate [85VEN/SWA] some of these complexes as summarised in Table VI-44. 4– where L’ = malonate ox(L’L’’) alate complexes, UO Quaternary U(VI) ox 2 , in conjunc- [79KUM/CHA] and L’’ = iminodiacetate, are also reported in the literature U(VI) oxalate complexes. Th ese data are not accepted by tion with the data for binary this review because essential experimental conditions are not provided in and discussions on quaternary complexes containing other organic [79KUM/CHA] ligands than ox, cit, edta or isa are beyond the scope of this review. Detailed reasons for the rejection are provided in Appendix A. As shown in Table VI-44, data on the ternary U(VI) oxalate complexes are rare. Their presence is not confirmed by independent studies or other methods [2000PAL] . Besides, there are a few shortcomings associated with these data. For ex- ample, some of the stability constants were determined in conjunction with the prepara- , such as those described in [67ZAK/ORL2] tion of the corresponding solid compounds Section VI.10.1.2.1. As a result, the calculated stability constants are correlated to the solubility constants of the solid phases and are associated with high uncertainty if the solid phase is not well defined. The solubility experiments were not conducted in a single ionic medium with constant ionic strength [67ZAK/ORL2] . Experimental condi-

285 VI.10 Uranium oxalate compounds and complexes 243 [56GRI/PTI] , [79KUM/CHA] tions are either incorrect or missing , [84SIR] , . Therefore, the stability constants for the ternary and quaternary U(VI) [85VEN/SWA] oxalate complexes are not accepted in this review. Detailed comments on the shortcom- ings of the references are given in Appendix A. Table VI-44: Stability constants of ternary U(VI) oxalate complexes reported in the literature. C) log Method Ionic medium K Reference t ( ° 10 – + U UO ox·H ox(OH) UO + H O(aq) 2 2 2 –1 25 ~ – 6.30 [56GRI/PTI] I < 0.1 mol·kg pot 2+ 6 − 2– + UO U + 2 H 2 + 2 H (UO ) (OH) (ox) O + 4 ox 2 4 2 2 22 0.05) 0.02) ± = 1.0 M NaClO (12.48 ± (25.00 [2000PAL] I pot, cou 4 − 2+ 4 + 2– + 4 H U O + 2 ox UO (UO ) (OH) (ox) 2 + 4 H 2 2 2 22 4 pot,cou = 1.0 M NaClO I (25.00 ± 0.02) – (7.15 ± 0.08) [2000PAL] 4 2 − 2 − UO U SO ox + UO oxSO 2 23 3 = ? I ? 4.53 [67ZAK/ORL2] sp 4 2 − − 2– )ox + SO U (SO UO UO ox(SO ) 3 2 232 3 I = 2.5 M NH ± Cl (23 sol 1) 3.72 [67ZAK/ORL2] 4 2+ 2– 2– 2– + L UO + ox UO L = malonate oxL U , 2 2 pot I = 0.10 M KNO (25 ± 1) (7.67 ± 0.11) [79KUM/CHA] 3 2+ 2– 4– 2– L UO (ox) + L UO , L = malonate U + 2 ox 2 2 2 = 0.10 M KNO pot (25 ± 1) (14.32 ± 0.24) [79KUM/CHA] I 3 2+ 2– 2– 2– + L L = iminodiacetate UO + ox oxL U , UO 2 2 0.28) I (25 ± 1) (10.07 ± = 0.10 M KNO [79KUM/CHA] pot 3 2+ 2- 2– 4– + L L = iminodiacetate U UO (ox) , L UO + 2 ox 2 2 2 pot I = 0.10 M KNO (25 ± 1) (13.65 ± 0.11) [79KUM/CHA] 3 2+ 2– 3– 3– + L UO U UO + ox oxL , L = nitrilotriacetate 2 2 I = 0.1 M NaClO [84SIR] 40 8.98 em 4 2+ 2– 2– 2– + L L = phthalate U UO UO oxL + ox , 2 2 = 0.1 M KNO 25 7.24 I pot [85VEN/SWA] 3 Among the reported ternary complexes, the U(VI)-hydroxyl-oxalato, U(VI)- VI)-fluorido-oxalato complexes are of importance in predict- carbonato-oxalato and U( ing the chemical behavior of U(VI) in geological environments where the pH, the con- centration of carbonate and/or fluoride are high. For example, formation of 6 − 4 − could be significant in solu- [2000PAL] and (UO ) (OH) (ox) (UO ) (OH) (ox) 2 22 4 2 22 4 tions with pH > 7 and change the onset of the hydrolysis of U(VI). Using the data on the and [2000PAL] , the stability [56GRI/PTI] U(VI)-hydroxyl-oxalato complexes from constants of U(VI) oxalate complexes selected by this review (Table VI-42), and the

286 VI Discussion of data selection for oxalate 244 [2003GUI/FAN] hydrolysis constants of U(VI) , a speciation diagram (Figure VI-34) is –5 –5 10 M U(VI) and 1 × 10 M oxalate at 25°C. As shown in × drawn for a solution of 1 – 6 − Figure VI-34, assuming the stability constants of UO (OH)ox , (UO ) (OH) (ox) 2 2 22 4 4 − are correct, the ternary U(VI)-hydroxyl-oxalato complexes, and (UO ) (OH) (ox) 22 4 2 – 4 − (OH)ox , and could amount to 20 – 30% of the total U(VI) in UO (UO ) (OH) (ox) 2 4 22 2 the pH regions between 6 and 8, even at such low concentration of oxalate. These two 6 − complexes, as well as that is insignificant when [oxalate] is low, (UO ) (OH) (ox) 2 4 22 could become even more important if the concentration of oxalate is higher. Obviously, it is of great value to conduct further studies and obtain reliable data on ternary U(VI)- hydroxyl-oxalato complexes, so that the effect of oxalate on the hydrolysis of U(VI) can be better evaluated. Figure VI-34: Speciation of U(VI) at 25°C, using the stability constants of ternary U(VI)-hydroxyl-oxalato complexes in Table VI-44. Formation of solid phases is suppressed. The formation of U(VI)-carbonato-ox alato complexes was postulated in [61CHE/GOL] , but no stability constants were determined. Potentiometric titrations of (ox)·3H O(s) with (NH ) CO showed pH “plateaus” th at suggested the formation UO 2 3 4 2 2 ) CO ]/[UO (ox)] = 1:1 of U(VI)-carbonato-oxalato complexes at molar ratios of [(NH 3 4 2 2 and 2:1. Eventually, a precipitate that was assumed to be ammonium tricarbonatouranate formed at a molar ratio of [(NH ) CO ]/[UO (ox)] = 3:1. Similar 4 2 3 2 CO (ox)·3H or K O(s) with Na CO observations were made in the titrations of UO 2 3 2 2 2 3 [61CHE/GOL] . The first pH plateau appears to correspond to the formation of

287 VI.10 Uranium oxalate compounds and complexes 245 2– CO O . Solids with the compositions corresponding to M UO oxH CO oxH O (M UO 3 2 2 2 2 3 2 + + + , Na ) were isolated and the fresh aqueous solutions of such compounds and K = NH 4 ties characteristic of three ions, supporting the formula showed electrical conductivi 2– [61CHE/GOL] direct identification of the UO ox . Though there is a lack of CO 2 3 complex in solution, its formation could be expected solely based on structural 2+ information. The coordination number of in its equatorial plane would be five UO 2 carbonate and one water. The formation of with one bidentate oxalate, one bidentate ffect the concentration of uranium in U(VI)-carbonato-oxalato complexes could a p is high. To evaluate th e importance of the U(VI)- groundwater where the CO 2 carbonato-oxalato complexes under such conditions, approximation approaches can be taken, in the absence of reliable data, to estimate the stability constants from 2 − 2 − and ) (see Chapter 3 of corresponding binary complexes ( UO (CO ) UO (ox) 232 22 ). Further discussions on this subject is beyond the scope of this review. [97ALL/BAN] The formation of U(VI)-fluorido-oxalato complexes was postulated, based on + + + the studies of the solubility of M (M = F ). Stepwise re- , K UO and CN H NH 5 2 3 4 36 placements of fluoride by oxalate, as suggested by the change in solubility as a function 3– 3 − es with the formula: UO ox ox], could result in speci and F of [M UO F(ox) 2 2 3 22 [68SHC/BEL] . Solids isolated from solutions with corresponding molar ratios of UO F ox appear to be consistent with the formula of ternary U(VI)-fluorido- and M M 2 5 3 2 oxalato compounds, though no evidence was provided to support the existence of such complexes in solution. Similar to the U(VI)-carbonato-oxalato complexes, formation of ternary U(VI)-fluorido-oxalato complexes could also be expected solely based on the structural information of binary U(VI) oxalate and U(VI) fluoride complexes. An equa- 3– 2+ 3 − ox F is always satisfied in , UO torial coordination number of five for UO F UO 2 3 2 25 − 3 . Again, to evaluate the importan ce of the U(VI)-fl uorido-oxalato com- and UO F(ox) 22 – ] is high, approximation approaches can be taken, in the plexes in groundwater where [F absence of reliable data, to estimate the stability constants from corresponding binary ). U(VI) fluoride and U(VI) oxalat [97ALL/BAN] e complexes (see Chapter 3 of 1 VI.11 Neptunium oxalate compounds and complexes VI.11.1 Solid neptunium oxalates VI.11.1.1 Solid neptunium(III) oxalates Solid Np(III) oxalate, which is unstable to oxidation, was prepared by [71MEF/GEL2] , and the composition of the solid was found to correspond to Np (ox) ≈ · n H O, with n 2 2 3 11. The standard entropy of this compound was estimated by Moskvin using a Latimer- –1 1 − ο . S (298.15 K) = 709 J·K : ·mol [73MOS] like method m 1 Prof. J. Havel and Dr. P. Lubal (both from the Department of Analytical Chemistry, Masaryk University, Brno, Czech Republic) contributed to this section during the early stages of the review process.

288 VI Discussion of data selection for oxalate 246 Solid neptunium(IV) oxalates VI.11.1.2 ·6H O, was characterised by single crystal (IV) oxalate, Np(ox) The structure of Np 2 2 [97GRI/CHA] X-ray diffraction . A modified Latimer methodology was used by Mosk- –1 ο 1 − (298.15 K) = 368 J·K S ·mol vin to estimate the standard entropy of this compound: m [73MOS] . Np(IV) oxalate is sparingl y soluble in water but in excess of oxalic acid it dis- solves forming complexes. In acid media a me tastable solid may precipitate when solu- O) ·7H Np O (equivalent to (ox) tions of Np(IV) and oxalic acid are mixed: (H 5 2 2 3 2 H Np ·9H O). With time this solid is however transformed into Np(ox) ·6H O (ox) 2 2 2 2 2 5 , [94ZHI/MAT2] , [98CHA/KRO] . when standing in the mother liquor [88BYK/KUZ2] O was reported in [60KON/GEL] , [64BAN/SHA] , ·6H The solubility of Np(ox) 2 2 [64POR] , [83LUE] . Unfortunately the solid phase was not characterised in any of these O) Np (ox) ·7H O cannot be ruled out. Furthermore the studies, and the presence of (H 3 5 2 2 2 ≈ ° C reported in [64POR] , [83LUE] are in disagreement 22 experimental solubilities at cf with each other for solutions with acidities < 4 M, . Figure VI-35. The solubilities published in [64POR] are not considered in this review, cf . Appendix A. Figure VI-35: Comparison of some of the experimental solubility data for Np(IV) ° solutions reported in [64POR] , [83LUE] at 23 and 22 C, respectively. oxalate in HNO 3 have been plotted in the figure, to Only the data in the range 1.6 to 2.9 M HNO 3 illustrate the discrepancies between the two sets of data. For these acidities the disagreement is largest at [H ox] < 0.1 M. 2 25 Ref. HNO 3 [64POR] 1.64 M 20 [64POR] 2.88 M [83LUE] 2 M 15 10 Np (IV) / (mg /L) 5 0 0.20 0.30 0.40 0.00 0.10 0.50 ox] / M [H 2

289 VI.11 Neptunium oxalat e compounds and complexes 247 Solubility constants have been obtained from solubility measurements of , [83LUE] , cf . Table [64BAN/SHA] , Np(IV) oxalate and reported in [60KON/GEL] [64POR] VI-45. The data from were also re-evaluated in , but as already [68MOS2] are not considered in this re- mentioned above, the solubilities published in [64POR] , [64BAN/SHA] , [64POR] , [60KON/GEL] view. Appendix A contains comments on [83LUE] is not considered . The earlier study by Kondratov and Gel'man [60KON/GEL] ortcomings described in Appendix A. in this review because of diverse sh 4+ ions in High acidities are usually used in studies on aqueous chemistry of M order to avoid the hydrolysis of the metal ion. Under these conditions the oxalate ligand is fully protonated, but the medium effects on the dissociation constants of oxalic acid are not well known in these strongly acidic solutions. Therefore, any published equilib- − 2 rium constant involving the oxalate ligand, ox , and Np(IV) should be regarded with , [83LUE] have been suspicion. The experimental solubilities reported in [64BAN/SHA] ox(aq) as a ligand. Details of these re-evaluations are analysed by this review using H 2 given in the entries of Appendix A. The resulting complex formation constants are dis- cussed in Section VI.11.2.2, while the solubility constants are listed in Table VI-46. There is large disagreement between the complex formation constants derived from . Sec- these solubilities and those from spectrophotometric or solvent extraction data, cf tion VI.11.2.2. Taking this into account, and considering that the solid phase was not O. ·6H characterised in these studies, no solubility constant is recommended for Np(ox) 2 2 Table VI-45: Solubility constants of solid Np(IV) oxalate reported in the literature. ( ° C) log Method Ionic medium K t Reference s 10 + 4+ 6H O(l) O(cr) + 4 H ⋅ U Np Np(ox) + 2 H ox(aq) + 6 H 2 2 2 2 (1.56 and 2.09) M HCl 16 sol (11.32 ± 0.03) [60KON/GEL] − (0.41 and 1.08) M HCl 19 − (11.22 ± 0.09) 4+ 2 − + 2 ox Np ⋅ 6H O(cr) U + 6 H O(l) Np(ox) 2 2 2 (1.56 and 2.09) M HCl 0.03) − (22.16 ± sol [60KON/GEL] 16 (0.41 and 1.08) M HCl ± − (22.07 0.09) 19 16 − (22.30 ± 0.06) 4.1 M HClO 4 [64BAN/SHA] 0.03) (26 ± 2) − (18.89 ± 1 M HClO sol 4 1 M (Na, H)ClO − (18.89 ± 0.03) 4 Np(ox) ⋅ 6H O(l) O(cr) U Np(ox) (aq) + 6 H 2 2 2 2 sol (0.4–10) M HNO 22 − (4.75 ± 0.07) [83LUE] 3 45 − (4.2 ± 0.2) 60 − (3.75 ± 0.05)

290 VI Discussion of data selection for oxalate 248 Ternary neptunium(IV) oxalates with alkali metal ions in an outer coordination sphere are quite soluble in water but they may be precipitated by adding alcohol. In addition to (H Np O) (ox) ·7H O mentioned above, the following solids have been 2 2 5 2 3 Np(ox) O ·3H ·9H O and K (ox) Np(ox) Np ·4H , M O [67MEF/GEL2] reported: Na 2 2 4 5 4 2 4 4 2 2 ) . The preparation of an ammo- Np [88BYK/KUZ2] (ox) O ·7H (M = Na, K) and (NH 2 2 2 4 5 nium salt reported to be (NH n Np(ox) ) ·(NH O was described in H ox· ) 2 4 2 4 4 4 [67GEL/MOS] . , [60KON/GEL] 4+ + U Np + O(cr) + 4 H ·6H Table VI-46: Solubility constants for reaction: Np(ox) 2 2 2 H ox(aq) + 6 H O(l), obtained from the re-evaluations described in Appendix A of 2 2 25 ° C. ≈ experimental literature solubilities of Np(IV) oxalate at K ° C) log t Ionic medium Reference to the original data ( s 10 → 0 (26 ± 2) − (13.1 ± 1.0) [64BAN/SHA] I a − (10.8 ± 0.9) 1 M HClO 4 → 0 22 − (12.2 ± 0.2) [83LUE] I a: The equilibrium constant listed in , but was obtained by this line refers to 1 M HClO 4 simultaneously analysing solubility data obtained both in 1 M NaClO and in (0.5 M HClO + 4 4 0.5 M NaClO ), cf . Appendix A. 4 VI.11.1.3 Solid neptunium(V) oxalates Binary Np(V) oxalate may be prepared by mixing oxalic acid dissolved in tertiary butyl- alcohol with a solution of Np(V) in 1 M HCl. The product was analysed and found to be NpO [54GRU/HUT] , (Hox)·2H . This solid is also mentioned in an- O [52GIB/GRU] 2 2 reporting its IR absorption spectra, but in this case the other study [73KHA/MOS2] authors did not give any details on the analysis of the solid. The same synthetic proce- [72JON/STO] dure was attempted by but a compound with a composition correspond- ) O was isolated instead [72JON/STO] . A similar solid reported to ox·4H ing to (NpO 2 2 2 ) ox·H O was prepared by [81ZUB/KRO] and the authors be the monohydrate (NpO 2 2 2 reported that this solid has a slow precipitation rate and a solubility in water at room 3 − 3 − temperature 10 mol·dm 2.6 . Apparently the number of water molecules in the ≈ × stoichiometry of Np(V) oxalate depends on the conditions for its synthesis. The molecu- lar structure of (NpO ox·6H O, obtained from a single-crystal X-ray diffraction study ) 2 2 2 ++ N bonds. , shows clearly the presence of pO NpO − [96GRI/CHA2] 22 Several ternary and quaternary Np(V) oxalate compounds have been reported: MNpO ox· n , O, where M = Na, K, Cs or NH [70GEL/BLO2] , [81ZUB/KRO] H 4 2 2 n and Co(NH ) . The sodium, ammonium NpO [91GRI/BAT] (ox) O · H [87MEF/KRO] 3 2 2 2 6 and hexamminecobalt(III) salts have been st udied by single-crystal X-ray diffraction [84TOM/VOL] , [91GRI/BAT] , [91GRI/BES] . MNpO O compounds have solu- ox· n H 2 2 ≈ (0.02 to 0.04) M, while Co(NH ) NpO O is “almost insoluble” in ⋅ n H (ox) bilities 6 3 2 2 2 , [76MEF/BLO] . It has been observed that the Np(V) concentra- water [70GEL/BLO2] tion decreases with time in solutions equilibrated with MNpO O (M = Na, K or ox ⋅ n H 2 2

291 VI.11 Neptunium oxalat e compounds and complexes 249 NH ), and this has been explained by assuming that a less soluble compound, probably 4 O, is formed ox ⋅ H ) . For the least soluble caesium salt [81ZUB/KRO] (NpO 2 2 2 CsNpO ⋅ n H ox O the decrease in Np(V)-concentration with time was not observed. 2 2 No thermodynamic data is available for any neptunium(V) compound with ox- alate. Solid neptunium(VI) oxalates VI.11.1.4 The redox reactions of neptunyl(VI) with oxalate at room temperature are slow and olate Np(VI) oxalate compounds from aqueous solutions for therefore it is possible to is subsequent X-ray analyses or solubility experiments, e.g. , NpO ox ⋅ 3H O, as well as a 2 2 series of ternary oxalates of neptunyl(VI) and alkali metals. . In order ⋅ 3H O was described in [69MEF/KRO2] ox The preparation of NpO 2 2 to avoid reduction of neptunium(VI) to neptunium(V) the synthesis of Np(VI) oxalate has to be performed from cold solutions in the presence of an oxidising agent ( e.g. , 0.05 [81MEF/GRI] ). The unit cell parameters for NpO O were reported in ox ⋅ 3H . M KBrO 2 2 3 NpO O is slightly soluble in water and in diluted inorganic acid solutions. The ⋅ 3H ox 2 2 solubility of this solid was studied as a function of oxalic acid concentration (up to 0.15 at 14 ° C [69MEF/KRO2] , but the reported solubility constant is only M) in 1 M HNO 3 of qualitative use, cf . Appendix A. The standard entropy of NpO ox O was esti- ⋅ 3H 2 2 ο using a modified Latimer method: S (298.15 K) = 272 mated by Moskvin [73MOS] m –1 − 1 K ·mol ⋅ . J Ternary oxalates, K (ox) ) ⋅ (NpO H O ( n = 2 or 4), were precipitated from n 2 2 6 2 5 cold, slightly acidic, concentrated solutions of neptunium(VI) mixed with saturated ox solutions [81MEF/GRI] . An analogous way of preparation was used for the syn- K 2 thesis of Cs (NpO ) . These compounds and of Cs NpO (ox) ⋅ 2H O [81MEF/GRI] (ox) 2 2 3 2 2 2 2 2 are highly soluble in water and they are stable for several days when dried and kept at ° C low temperatures, but they decompose rapidly at temperatures above 100 [81MEF/GRI] . VI.11.1.5 Solid neptunium(VII) oxalates Neptunium(VII) is readily reduced by oxalate and therefore it is not possible to prepare Np(VII) oxalate compound s in the solid state. VI.11.2 Aqueous neptunium oxalate complexes VI.11.2.1 Neptunium(III) oxalate complexes Neptunium(III) is oxidised by oxalate [71MEF/GEL2] and there is no evidence in the literature on the formation of Np(III)-oxalate complexes in aqueous solutions.

292 VI Discussion of data selection for oxalate 250 VI.11.2.2 Neptunium(IV) oxalate complexes The complexation of neptunium(IV) with oxalate has been studied mainly by solubility [60KON/GEL] , and by solvent extraction measurements [64BAN/SHA] [83LUE] , . In , , [64BAN/SHA] , [79KUS/GAN] [76BAG/RAM3] , [77RAM/RAM] [62ISH/NAK] . Table VI-47 lists the stability [67MEF/GEL2] addition, spectrophotometry was used in constants reported in the literature. Values of stability constants at temperatures differ- [77RAM/RAM] ent than 25ºC or room temperature were also obtained in . , [83LUE] − (4 2 ) n Species of stoichiometry Np(ox) n = 1 to 4) were reported in these refer- ( n [62ISH/NAK] ences, except for , but this paper is not credited by this review, cf . Appen- 2+ − 2 (aq) and were as- , Np(ox) N p(ox) dix A. In most cases only the species Np(ox) 2 3 sumed to be present in solution. A question not always addressed in these publications is the redox state of the neptunium ions. All the studies on Np(IV) complex formation have to be performed in strongly acidic solution in order to avoid the formation of hy- 3+ . In these conditions oxalate is fully protonated, but drolysis products such as NpOH the medium effects on the acidity constants of oxalic acid are not known in e.g. , HClO 4 solutions. It is then necessary to evaluate experimental data using oxalic acid or HNO 3 as a ligand: − n (4 2 ) + 4+ Np(ox) ox(aq) U H n (VI.43) + + 2 n H Np 2 n Except for the rejected [62ISH/NAK] cf . Appendix A), the solvent extraction ( that also [64BAN/SHA] studies may be grouped into two sets: an early investigation reported solubility data; and a series of publications by the same research group [76BAG/RAM3] , [77RAM/RAM] [77RAM/RAM2] , [79KUS/GAN] . All these reports , are discussed in Appendix A, where the re-analysis made by this review of the experi- mental data at 25 ° C is described. The equilibrium constants obtained for Reactions (VI.43) are listed in Table VI-48. Table VI-47: Literature stability constants for neptunium(IV) oxalate complexes. ° Method Ionic medium t K Reference ( C) log 10 2 − 2+ 4+ U Np(ox) Np + ox 1.0 M HCl (25 ± 2) (8.5 ± 0.3) [60KON/GEL] sol 1 M HClO ± 2) (7.47 ± 0.01) [64BAN/SHA] (26 sol 4 (7.40 ± 0.05) 1 M (H, Na)ClO 4 (8.19 ± 0.06) dis 1 M HClO 4 sp 0.3 M HClO 20 (8.8 ± 0.1) [67MEF/GEL2] ≈ 4 25.0 (9.220 ± 0.005) [76BAG/RAM3] dis 1 M HClO 4 [77RAM/RAM] 0.007) 10 (9.277 ± dis 1 M HClO 4 40 (8.935 ± 0.005) [77RAM/RAM2] 0.004) 25.0 (9.318 ± dis 1 M HClO 4 + 25 9.3 [79KUS/GAN] 1 M H dis (Continued on next page)

293 VI.11 Neptunium oxalat e compounds and complexes 251 Table VI-47: (continued) ( ° t K Reference Method Ionic medium C) log 10 − 4+ 2 + 2 ox Np(ox) (aq) Np U 2 0.01) 2) (17.54 ± ± [60KON/GEL] sol 1.0 M HCl (25 (13.69 (26 ± 2) sol 1 M HClO 0.02) [64BAN/SHA] ± 4 1 M (H, Na)ClO (13.62 0.04) ± 4 (16.2 ± 0.1) 1 M HClO dis 4 ± ≈ 20 (17.2 0.2) [67MEF/GEL2] sp 0.3 M HClO 4 dis 1 M HClO 25.0 (16.63 ± 0.02) [76BAG/RAM3] 4 0.03) [77RAM/RAM] 10 (16.75 ± 1 M HClO dis 4 ± 0.01) 40 (16.23 1 M HClO 4 dis 1 M HClO ± 0.05) [77RAM/RAM2] 25.0 (16.43 4 + 25 16.4 [79KUS/GAN] 1 M H dis 2 − − 2 4+ Np + 3 ox Np(ox) U 3 1.0 M HCl ± 2) (23.96 ± 0.04) [60KON/GEL] sol (25 1 M HClO (26 ± sol (19.37 ± 0.07) [64BAN/SHA] 2) 4 1 M (Na, H)ClO (19.45 ± 0.05) 4 [67MEF/GEL2] 0.2) ≈ 20 (24.4 ± 0.3 M HClO sp 4 ± 0.02) [77RAM/RAM] 10 (16.64 1 M HClO dis 4 dis 1 M HClO ± 0.2) [77RAM/RAM2] 25.0 (23.2 4 + 25 23.2 [79KUS/GAN] dis 1 M H − 4 4+ 2 − Np U Np(ox) + 4 ox 4 1.0 M HCl (25 sol 2) 27.4 [60KON/GEL] ± sp 0.3 M HClO ≈ 20 28.4 [67MEF/GEL2] 4 − 4+ (aq) + 4 Hox U Np(Hox) Np 4 dis [62ISH/NAK] (0.5 & 0.9) M HCl room temperature 3.5 2+ + Np(ox) U + H ox(aq) (aq) + 2 H Np(ox) 2 2 sol (0.4 - 10) M HNO 22 (2.15 ± 0.08) [83LUE] 3 (1.6 ± 0.2) 45 60 ± 0.06) (2.09 2 − + ox(aq) U (aq) + H + 2 H Np(ox) Np(ox) 2 2 3 sol (0.4 - 10) M HNO 22 (0.29 ± 0.08) [83LUE] 3 (0.6 ± 0.2) 45 60 ± 0.07) (0.02 The solubility studies [60KON/GEL] , [64BAN/SHA] , [83LUE] are also dis- cussed in Section VI.11.1.2. The earlier study by Kondratov and Gel'man [60KON/GEL] is not considered because of diverse shortcomings described in Appen- and the spectrophotometric [83LUE] , [64BAN/SHA] dix A. The solubility data in measurements in [67MEF/GEL2] were re-analysed by this review using H ox(aq) as the 2 ligand, see Appendix A, and the resulting equilibrium constants are listed in Table VI-48.

294 VI Discussion of data selection for oxalate 252 2+ for the formation constant of Np(ox) While there is a general good agreement between the different experimental techniques, this is not so for the second and third complex. In general the solubility values are lower and the spectrophotometric value higher than the two values obtained by solvent extraction, cf . Table VI-48. This dis- agreement is illustrated in the graphs of Figure VI-36. The discrepancies between the different experimental techniques indicate th e presence of unknown systematic errors. In addition to the experimental uncertainties in the measurements, the equilibrium con- stants are not very well defined because of lack of solubility data at low oxalic acid concentrations, and lack of solvent extraction and spectrophotometric data at high ox- select any values for the oxalate complex alic acid concentrations. This review does not formation of Np(IV). and lack of information on the com- Taking into account the large uncertainty ° C it is of no value to discuss temperature ≈ 25 plexation of Np(IV) by oxalic acid at , [79KUS/GAN] , [77RAM/RAM] fects are however small effects. The reported ef [83LUE] . described in Appendix A of experimental Table VI-48: Results from the re-evaluations 25ºC. ≈ literature data on the complex formation between Np(IV) and oxalic acid at ( ° C) log t Reference to the original data Method Ionic medium K 10 4+ + 2+ U Np(ox) ox(aq) + 2 H + H Np 2 dis (26 ± 2) (4.1 ± 0.2) [64BAN/SHA] 1 M HClO 4 (3.7 1.0) sol (a) ± 0.3 M HClO sp [67MEF/GEL2] 20 (5.5 ± 1.5) ≈ 4 1 M HClO dis 25.0 (4.6 ± 0.1) [76BAG/RAM3] , , [77RAM/RAM2] 4 [79KUS/GAN] → 0 M 22 (6.5 ± 0.5) [83LUE] sol 4+ + Np + 2 H ox(aq) U Np(ox) (aq) + 4 H 2 2 dis 1 M HClO (26 ± 2) (7.4 ± 0.4) [64BAN/SHA] 4 sol (a) (5.5 ± 1.0) 0.3 M HClO sp ≈ 20 (9.5 ± 1.0) [67MEF/GEL2] 4 dis 1 M HClO 25.0 (6.9 ± 1.0) [76BAG/RAM3] , [77RAM/RAM2] , 4 [79KUS/GAN] → 0 M 22 (7.1 ± 0.3) [83LUE] sol 2 − + 4+ Np U Np(ox) ox(aq) + 6 H + 3 H 2 3 dis 1 M HClO (26 ± 2) (10.8 ± 0.4) [64BAN/SHA] 4 sol (a) (7.1 ± 1.0) 0.3 M HClO sp ≈ 20 (11.7 ± [67MEF/GEL2] 1.0) 4 dis 1 M HClO , 25.0 (9.9 ± 0.7) [76BAG/RAM3] , [77RAM/RAM2] 4 [79KUS/GAN] sol → 0 M 22 ± 0.3) [83LUE] (7.1 a: The equilibrium constants listed in these lines refer to 1 M HClO , but were obtained by 4 data obtained both in 1 M NaClO simultaneously analysing solubility + and in (0.5 M HClO 4 4 0.5 M NaClO ), cf . Appendix A. 4

295 VI.11 Neptunium oxalat e compounds and complexes 253 Figure VI-36: Comparison of equilibrium constants for the complex formation between ≈ 25ºC. These values were obtained by the re-evaluations of Np(IV) and oxalic acid at experimental literature data described in Appendix A. The background colour of the symbols indicates the experimental method: black for solubility, white for two phase distribution (solvent extraction) and grey for spectrophotometry. The lines correspond to a weighted linear fit, added to the graphs for illustrative purposes only. 13.0 4+ ox(aq) U Np + H 2 References: 2+ + 11.0 Np(ox) + 2 H D [64BA N/SHA ] s o l 9.0 [64BA N/SHA ] d is + 10 1 7.0 K [67M EF/GEL2] 10 [76BA G/RA M 3], [77RA M /RA M 2] log 5.0 [83LUE] 3.0 1.0 0.5 0.0 I m 16.0 15.0 4+ 4+ ox(aq) + 3H Np U Np ox(aq) U + 2H 2 2 + + − 2 13.0 14.0 (aq)+ 4H + 6H Np(ox) Np(ox) 2 3 D D 11.0 12.0 + 6 + 12 3 2 β 9.0 10.0 β 10 10 log log 7.0 8.0 5.0 6.0 1.0 0.0 0.5 0.0 0.5 1.0 I m I m

296 VI Discussion of data selection for oxalate 254 VI.11.2.3 Neptunium(V) oxalate complexes + + + 4 H NpO Although Np(V) is considered to be stable towards disproportionation (2 2 2+ 4+ NpO + 2 H O(l)) in most chemical systems, it has been reported that dis- U Np + 2 2 ≈ 1.3 proportionation has been observed in the presence of 0.1 M oxalic acid at pH . The rate of disproportionation of Np(V) in oxalate solutions increases [53GRU/KAT] with temperature [67KRO/MEF2] . Furthermore, in the presence of , [73SHI/RUM2] ., from air, Np(V) decomposes catalytically oxalate, edta and e.g molecular oxygen, [79SHI/STE2] ° C . The decomposition proceeds through complex other ligands at 60 formation with the incorporation of molecular oxygen, the oxidation of Np(V) to Np(VI) by the coordinated oxygen, and subsequent reduction of Np(VI) by decomposi- tion of the ligand. At room temperature and moderate acidities the disproportionation of Np(V) is unlikely to be a problem for the experimental studies reviewed below. Neptunyl(V) complexation by oxalate has been studied by a wide range of ex- perimental techniques: potentiometry, spectrophotometry, solvent extraction, electromi- gration, ionic exchange and co-precipitation. The reported stability constants are sum- e studies. The exis- marised in Table VI-49. Appendix A contains comments on all thes (1 2 ) n − NpO (ox) tence of the species ( n = 1 and 2) is indicated in these references: n 2 + − 2 − NpO ox pO + ox (VI.44) N U 2 2 − 3 + − 2 U + 2 ox NpO NpO (ox) (VI.45) 2 22 There is no evidence for the formation of higher complexes. A protonated Hox(aq), has also been postulated from ion exchange or solvent extrac- complex, NpO 2 . As discussed in the corre- , [63ZOL/ALI] , [97POK/CHO] [61ZOL/MAR2] tion data sponding entries of Appendix A these claims have been rejected by this review. Fur- thermore, a re-analysis by this review of the spectrophotometric data at acidities up to gave no indication of such a complex, although there were clear pH = 1.1 from [72STO] − − 2 indications for the formation of mixed Np(V)-OH -ox ≥ complexes at pH 9. Table VI-49: The stability constants of neptunium(V) oxalate complexes reported in the literature. K ° C) log Method Ionic medium t Reference ( 10 + 2– – ox U NpO NpO + ox 2 2 0.5 M (NaClO ) room temperature 3.29 [53GRU/KAT] sp 4 0.05 M (NH ClO cix ) (20 ± 2) 4.0 [61ZOL/MAR2] , [63ZOL/ALI] 4 4 0.1 M (?) (25.0 ± 0.3) 3.56 [67KRO/MEF2] sp kin 0.23 M (HNO ?) 57.5 3.48 3 [72MAG/BIS] ) 20 (3.74 ± 0.05) gl 1 M (NaClO 4 [72STO] ) (25.0 ± 0.2) (4.54 ± 0.01) 1 M (NaClO sp 4 (a) 0.5 M (NH [78MOS/POZ] 2) (3.38 ± 0.11) ± , [79MOS/POZ4] Cl) (20 4 1 M (NaCl) (25 ± 2) (3.42 ± 0.05) [82INO/TOC] dis (Continued on next page)

297 VI.11 Neptunium oxalat e compounds and complexes 255 Table VI-49: (continued) t C) log Method Ionic medium K Reference ( ° 10 + 2– – U NpO ox + ox NpO 2 2 0.03) ) (25 ± 1) (3.44 ± dis 1 M (NaClO [83INO/TOC] 4 0.01) ± (3.67 0.01) ± (3.59 ± ) (23 ± 2) (3.52 1 M (NaClO 0.06) sp [87CAC/NEC] 4 sp 1 M (H,K)NO 25 4.39 [89NIK/ION2] 3 [89ROS/DIT] , ox) (25.0 ± 0.1) (3.77 ± 0.02) em 0.05 M Na(ClO 4 , ox) (3.90 ± 0.04) 0.1 M Na(ClO 4 0.3 M Na(ClO ± 0.04) , ox) (3.57 4 , ox) (3.75 ± 0.04) 0.1 M Na(NO 3 [89STO/CAC] ) (23 ± 2) (3.53 ± 0.04) sp 1 M (NaClO 4 0.1 M (NaClO dis ) (25 1) (3.7 ± 0.2) [92TOC/INO] ± 4 (3.73 ± ) 0.02) 1 M (NaClO 4 0.01) (3.70 ± ± 0.08) [96BOR/LIS] 5 m (NaCl) 25 (3.04 dis [97POK/CHO] 0.04) ± ) 25 (3.40 (NaClO dis 0.1 m 4 (3.40 0.04) 0.3 m ± (3.48 0.05) 0.5 m ± (3.35 ± 0.06) 1 m (3.61 0.13) 3 m ± ± 0.04) 5 m (3.96 (4.23 ± 0.05) 7 m 9 m (4.66 ± 0.05) ± 0.03) [2001BOR/MOO] dis 0.3 m (NaCl) 25 (3.62 (3.80 ± 0.02) [2001CHO/BON] 1 m 2 m ± 0.02) (3.89 3 m (4.05 ± 0.02) 4 m (4.18 ± 0.02) 5 m ± 0.05) (4.63 3 − +2 − NpO (ox) NpO + U 2ox 222 sp ) room temperature 7.06 [53GRU/KAT] 0.5 M (NaClO 4 cix 0.05 M (NH ClO [63ZOL/ALI] ) (20 ± 2) 7.4 [61ZOL/MAR2] , 4 4 0.1 M (?) ± 0.3) 5.89 [67KRO/MEF2] sp (25.0 ?) 57.5 6.02 0.23 M (HNO kin 3 gl 1 M (NaClO ) 20 (6.31 ± 0.10) [72MAG/BIS] 4 (7.86 [72STO] ) (25.0 ± 0.2) 0.04) ± sp 1 M (NaClO 4 (Continued on next page)

298 VI Discussion of data selection for oxalate 256 Table VI-49: (continued) Method Ionic medium C) log t K Reference ( ° 10 − − 3 +2 + U NpO (ox) 2ox NpO 222 [78MOS/POZ] Cl) (20 2) (5.65 ± 0.12) ± , [79MOS/POZ4] 0.5 M (NH (a) 4 (25 ± 2) (5.64 ± 0.05) [82INO/TOC] dis 1 M (NaCl) dis 1 M (NaClO 1) (5.83 ± 0.03) ± ) (25 [83INO/TOC] 4 0.01) (6.29 ± ± 0.01) (6.16 ) (23 ± 2) (6.09 ± 0.06) [87CAC/NEC] 1 M (NaClO sp 4 , ox) (25.0 ± 0.1) (5.98 ± 0.20) [89ROS/DIT] 0.05 M Na(ClO em 4 0.05) , ox) (6.27 ± 0.1 M Na(ClO 4 0.3 M Na(ClO , ox) (5.95 ± 0.13) 4 , ox) (6.13 ± 0.06) 0.1 M Na(NO 3 ± ) (23 ± 2) (6.12 0.05) [89STO/CAC] sp 1 M (NaClO 4 dis 0.1 M (NaClO ) (25 ± 1) (5.84 ± 0.2) [92TOC/INO] 4 ± 0.02) 1 M (6.16 0.01) (6.14 ± , ± 0.01) (NaCl) 25 (6.96 dis 2 m [2001BOR/MOO] 3 m 0.06) [2001CHO/BON] (7.07 ± (6.99 ± 0.05) 4 m − + Hox NpO + U NpO (Hox)(aq) 22 0.05 M (NH , ClO ) (20 ± 2) (2.7 ± 0.2) [61ZOL/MAR2] cix [63ZOL/ALI] 4 4 +− +2 ox NpO (Hox)(aq) ++ NpO U H 22 dis 3 m (NaClO ) 25 < 7.3 [97POK/CHO] 4 a: co-precipitation method; (?) ionic medium not reported. Some of the equilibrium constants reported in Table VI-49 were not included in this review for diverse reasons discussed in the corresponding Appendix A entries: [53GRU/KAT] , [63ZOL/ALI] , , , [78MOS/POZ] , [61ZOL/MAR2] [67KRO/MEF2] , , , [87CAC/NEC] , [89NIK/ION2] , [89ROS/DIT] [82INO/TOC] [79MOS/POZ4] . The remaining data were consider ed acceptable for further scrutiny, [96BOR/LIS] except that the experimental spectrophotometric data in [53GRU/KAT] , [72STO] were re-analysed by this review as described in Appendix A. The acceptable stability con- stants, with their assigned uncertainties, are listed in Table VI-50. The investigations ° C and 25 ° C. The tempera- summarised in that table have been performed between 20 − ° ox 20, 29 and 47 was studied at 10, ≈ C ture dependence for the formation of NpO 2 and the authors concluded that the heat of complexation was (0 ± 1.3) [53GRU/KAT] –1 kJ·mol . Although this value is not selected by this review, it indicates that all data in Table VI-50 can be included in the final data analysis without explicit temperature cor- ° C. rections to 25

299 VI.11 Neptunium oxalat e compounds and complexes 257 Table VI-50: Accepted stability constants for neptunium(V) oxalate complexation, with the uncertainties assigned by this review. ( ° Method Ionic medium K Reference t C) log 10 − − +2 NpO U NpO ox + ox 22 0.5 M (NaClO ) room temperature (3.29 ± 0.15) [53GRU/KAT] sp 4 ± ) 20 (3.74 gl 0.10) 1.0 M (NaClO [72MAG/BIS] 4 b ± 0.2) (3.84 0.10) ) (25.0 [72STO] ± sp 1.0 M (NaClO 4 [78MOS/POZ] ± 2) (3.4 Cl) (20 0.4) , [79MOS/POZ4] ± 0.5 M (NH (a) 4 1.0 M (NaClO dis ) (25 ± 1) (3.62 ± 0.20) [83INO/TOC] 4 ) (23 ± 2) (3.53 ± 0.10) [89STO/CAC] sp 1.0 M (NaClO 4 ) (25 ± 1) (3.71 ± 0.20) [92TOC/INO] dis 0.1 M (NaClO 4 1.0 M (3.71 ± 0.10) [97POK/CHO] 0.20) ) 25 (3.40 ± dis 0.1 m (NaClO 4 (3.40 0.20) 0.3 m ± (3.48 ± 0.20) 0.5 m 1.0 m ± 0.20) (3.35 3.0 m (3.61 ± 0.20) 5.0 m (3.96 ± 0.20) , dis 0.3 m ± 0.15) [2001BOR/MOO] (NaCl) 25 (3.62 0.15) [2001CHO/BON] (3.80 1.0 m ± (3.89 0.15) 2.0 m ± ± 0.15) 3.0 m (4.05 (4.18 ± 0.15) 4.0 m 5.0 m ± 0.15) (4.63 3 − − +2 NpO (ox) + NpO 2 ox U 222 1.0 M (NaClO ) 20 (6.31 ± 0.15) [72MAG/BIS] gl 4 b 1.0 M (NaClO ) (25.0 ± 0.2) (6.36 ± 0.15) [72STO] sp 4 ) (25 ± 1) (6.20 ± 0.20) [83INO/TOC] dis 1.0 M (NaClO 4 sp 1.0 M (NaClO ) (23 ± 2) (6.12 ± 0.10) [89STO/CAC] 4 (5.84 ) (25 ± 1) [92TOC/INO] ± 0.20) 0.1 M (NaClO dis 4 1.0 M (6.15 ± 0.10) dis 2 m ± 0.15) [2001BOR/MOO] (NaCl) 25 (6.96 , 3 m (7.07 ± 0.15) [2001CHO/BON] 4 m (6.99 ± 0.15) a: co-precipitation method. These values , which were obtained in ammoni um chloride, were not included in the SIT regression. b: values obtained from a re-analysis of the original data, see Appendix A.

300 VI Discussion of data selection for oxalate 258 Most of the studies listed in Table VI-50 were performed in sodium electro- . As this review has not , [79MOS/POZ4] lytes, with the exception of [78MOS/POZ] been able to examine the effect of ammoni um electrolytes on the protonation constants [78MOS/POZ] of oxalate, the data from [79MOS/POZ4] is not considered any further. , Only two studies report data at I > 1 M, namely [97POK/CHO] , and in NaClO 4 [2001BOR/MOO] in NaCl. The medium effect might be interpreted , [2001CHO/BON] − + pO and Cl , or as a difference in activity N as a weak complex formation between 2 coefficients. According to the SIT model described in Appendix B the activity coeffi- + NpO cients of in chloride media are smaller than in perchlorate media: Equation (B.4) 2 + − − 1 + − ( ) = (0.09 ± 0.05) kg ⋅ mol ε 0.05) and ε ( N ± N pO , ) = (0.25 ClO pO ,Cl with 2 2 4 1 − ⋅ mol [2001LEM/FUG] . This requires that the stability constants in perchlorate me- kg dia should be larger than in chloride media: ο ββ γγγ . =+ + − log log log log log − +2 − 1 10 10 10 1 10 10 NpO NpO ox ox 22 If the alternative model of weak complex formation with anions from the background electrolyte is analysed, the same effect should be observed, a smaller con- stant should be observed in chloride media. Following a similar discussion to that given in Section V.4, one may write: † − =−+ ⋅ log K (NpO Cl) [Cl ]) ββ log log (1 2 1 10 1 10 10 † where β is the equilibrium constant in chloride media; β is the same constant in the 1 1 (NpO Cl) K is the equilibrium absence of chloride, but at the same ionic strength; and 2 + − NpO constant for the weak complex formation between , also at the same ionic and Cl 2 strength. It must be noted in this context that the NEA-TDB reviews have chosen to de solutions as arising from changes in ac- interpret medium effects of Np(V) in chlori complex formation which is difficult to dem- tivity coefficients, rather than from weak onstrate experimentally at moderate concentrations of chloride ions. In any case it is clear that equilibrium constants for complex formation be- tween Np(V) and any ligand should be larger in perchlorate media than in chloride in NaClO and [97POK/CHO] media. However, the two sets of data for oxalate, 4 , in NaCl, show a reverse trend. [97POK/CHO] [2001CHO/BON] [2001BOR/MOO] nd a comparison with the results listed in contains data also on acetate and citrate a [2001CHO/BON] shows that for acetate the equilibrium constants in NaCl media are es at background electro- (with one exception: the valu also larger than those in NaClO 4 lyte 0.3 molal). On the other hand the citrate data show the expected trend: the values of β in NaClO are larger than those in NaCl medi a. There is no clear reason for the dis- 1 4 agreement between the Np(V)-oxalate (and -acetate) equilibrium constants in [97POK/CHO] on one hand and in [2001BOR/MOO] , [2001CHO/BON] on the other. A possible explanation might be that the interpretation of the data in [97POK/CHO] 3 − NpO (ox) , although Pokrovsky and Choppin should have included the formation of 22 reported not to have found evidence for it. Nevertheless, because of this unresolved

301 VI.11 Neptunium oxalat e compounds and complexes 259 disagreement the data from these two studies at > 1 molal are not considered accept- I m ≤ 1 molal are included in the selection procedure, with an I able, while the data at m increased uncertainty of 0.2 log -units. ± 10 Therefore, all the values liste d in Table VI-50, except the values from β 1 [97POK/CHO] I , > 1 molal from , and all values at , [79MOS/POZ4] [78MOS/POZ] m [2001BOR/MOO] , [2001CHO/BON] were converted to molal units and they were ex- trapolated to zero ionic strength using the SIT methodology: 2 zI ∆⋅ 0.509 m −ο + ′ log (NpO , X ) II log ββ −−ε=−∆ε 10 m 10 2 m nnn I 11.5 + m − − − 2 + + 2 − (2) nn − ′ ; z , (Na ∆ ); and = is either Cl ε − n ε (Na or ,ox where X ClO NpO (ox) ∆ε 4 n 2 n 4 and 0 for n = 1 and 2, respectively. The SIT extrapolations are illustrated in Figure − = VI-37 and Figure VI-38, and they resulted in the following selected values: 1 − ο ′ ε ∆ 0.1), (VI.44) = (3.9 ⋅ mol ± (VI.44) = − (0.3 ± 0.1) kg log β 1 10 1 1 − ο ′ ε ∆ ± (VI.45) = (5.8 mol 0.2), ⋅ − (0.1 ± 0.2) kg (VI.45) = β log 2 10 2 1 − + 2 − Using the interaction coefficient ε ± 0.01) kg ⋅ mol (0.08 , Na se- ) = (ox − − + .3), the interaction coefficients lected by this review (Section VI (Na ox ε ) = , NpO 2 + − − 1 1 − 3 and ε (Na (0.4 , ± 0.1) kg are obtained. ⋅ ) = − mol ± 0.2) kg ⋅ mol (0.3 − NpO (ox) 22 the data for Np(V)- carbonate complexes These values may be compared with + − ο − [2001LEM/FUG] , ( ) = (Na : ) = (4.96 ± 0.06), ε NpO CO β NpO CO log 10 23 1 23 − 1 + ο 3 − − (0.18 ; and , ± (Na 0.15) kg ε ( ⋅ 0.10), ± mol ) = (6.53 NpO (CO ) log β 2 10 232 − 1 3 − − (0.33 ± 0.17) kg ⋅ mol . It can be seen that the stability constants for ) = NpO (CO ) 232 neptunium(V)-oxalate at zero ionic strength are lower than analogous complexes of Np(V)-carbonate. The selected equilibrium constants yield: ο –1 − ∆ G 5.9) kJ·mol , 298.15 K) = – (1610.2 ± ( NpO ox fm 2 –1 ο 3 − 6.8) kJ·mol . ∆ ( , 298.15 K) = – (2301.1 ± G NpO (ox) fm 22

302 VI Discussion of data selection for oxalate 260 2 − − + ox U NpO pO + ox N listed in Figure VI-37: Extrapolation of data for reaction 2 2 Table VI-50 according to the SIT model described in Appendix B. Symbols with white background correspond to NaCl ionic medium, all other data were obtained in NaClO 4 media. 5.00 [53GRU/KAT ] 2 + − − + ox ox NpO U NpO 2 2 [72MAG/BIS] 4.75 [72ST O] 4.50 D [83INO/T OC] + 4 [89ST O/CAC] 4.25 1 K [92T OC/INO] 10 4.00 log [97POK/CHO] [2001BOR/MOO], 3.75 [2001CHO/BON] 3.50 0.00 0.75 0.50 1.00 0.25 I m 2 − 3 − + Figure VI-38: Extrapolation of data for reaction listed + 2 ox NpO NpO (ox) U 22 2 in Table VI-50 according to the SIT model described in Appendix B. 6.50 + 3 − − 2 NpO NpO (ox) U + 2 ox 22 2 [72MAG/BIS] 6.25 [72ST O] [83INO/T OC] 2 β 6.00 10 [89ST O/CAC] log [92T OC/INO] 5.75 5.50 1.00 0.75 0.50 0.00 0.25 I m

303 VI.11 Neptunium oxalat e compounds and complexes 261 VI.11.2.4 Neptunium(VI) oxalate complexes + 2 complexation with oxalate is complicated by redox reactions NpO The study of 2 where Np(VI) is reduced by oxalate to Np(V) which is somewhat stabilised by oxalate complexation [67KRO/MEF2] , [69SHA/AMI] , [84RAO/CHO] , [69MEF/KRO2] , . At acid concentrations above 3 M (HClO , HCl, HNO ) the reduction [98REE/WYG] 4 3 of Np(VI) is fast. et al Mefod’eva [69MEF/KRO2] . studied the Np(VI)-oxalate system at tem- peratures below 20ºC. These authors determined approximate values of complex forma- tion using spectrophotometry and solubility measurements. The complexes NpO ox(aq) 2 2 − N pO (ox) . Ap- were proposed, and their formation constants were reported, cf and 22 pendix A. These values are only of qualitative significance for laboratory studies. For e.g any application modelling the long term behaviour of neptunium, ., for nuclear waste disposal, the unstable Np(VI) oxalate complexes are unimportant as was pointed out in . [98REE/WYG] VI.11.2.5 Neptunium(VII) oxalate complexes In acidic medium Np(VII) is rapidly reduced both by water [2001LEM/FUG] and by . For example the half-life of Np(VII) is about 0.01 seconds at oxalate [70SHI/KRO] + − 4 25 C in solutions with [Np] = 9.2 × 10 ° M, [H = 0.08 M, [ox] ] = 0.059 M, and I TOT TOT . In basic medium Np(VII) is more stable, and the reduction by [78COO/WOO] = 1.0 M water to Np(VI) takes place over a period of hours to weeks at room temperature [2001LEM/FUG] . Because of the instability of Np(VII) no equilibrium constants for Np(VII)- oxalate species have been determined. Howe ver, for nuclear waste disposal applica- tions, often involving long term perspectives, equilibrium constants for highly unstable neptunium(VII)-oxalate species are not needed anyway. VI.12 Plutonium oxalate compounds and complexes Solid plutonium oxalates VI.12.1 Solid plutonium oxalates have been widely used and investigated because of their use- fulness, e.g. , in Pu-separation from other actinides. The solids reported in the literature are listed in Table VI-51. Although there is a large amount of information concerning the temperature stability of several of these oxalates and their conversion into plutonium oxide, these properties are not discussed in this review. Information concerning the thermal stabili- [54CUN] ties may be found in several monographs , [67GEL/MOS] , [79CLE] , , [87MEF/KRO] , [91MAT/KAR] . [86WEI/KAT]

304 VI Discussion of data selection for oxalate 262 It should be kept in mind when discussing the properties of plutonium com- pounds and solutions that there is, as an additional difficulty, the problem of radiation 238 239 damages and water radiolysis, especially for Pu Pu, but also to a lesser extent for , [58GEL/SOK] , , [86KAN/KIM] . [60SOK/GEL] [56FOM/KAR] VI.12.1.1 Solid plutonium(III) oxalates I) oxalate was reported quite The precipitation of solid Pu(II early in the history of plu- , [49PAT] . The solubility of Pu(III)-oxalate in water is [49PAT2] tonium chemistry − 6 − 5 and 1.88 × 10 10 M at 20 and 70 × ° C, respectively [57GEL/MAT2] , given to be 4.32 [57GEL/MAT3] [67GEL/MOS] . The extent of the hydration of this compound has , , but most probably the correct formula is been subject to discussion [79CLE] . In some earlier papers the formula is given (ox) [65JEN/MOO] ·10H , O [63CHA] Pu 2 2 3 with nine water molecules, e.g. , in [57GEL/MOS] . The standard entropy of ο (ox) ⋅ 9H O was estimated by Moskvin using a Latimer-like method: (298.15 K) S Pu 2 2 3 m − − 1 1 mol K ⋅ [73MOS] . ⋅ = 651 J The references reporting the solubility of Pu(III)-oxalate are summarised in Table VI-52. These studies fall into two categories: those in near-neutral solutions, where the total concentration of oxalate was varied, and those in acid media, where either the total acidity or total oxalic acid was varied. In acid media oxalic acid remains protonated and the extent of complex forma- tion is expected to be limited or negligible. Under such conditions a solubility product may be determined from the solubility data. On the other hand, solubility determina- tions in near-neutral solutions with varying ligand concentration would allow the deter- mination of equilibrium constants for complex formation. In both cases several re- quirements must be fulfilled: proper phase separation after the equilibration time; checks that the solutions are not over- or un isation of the solid dersaturated; character phase and verification that it does not change with varying conditions such as acidity and ligand concentration. All studies on the Pu(III)-oxalate system are deficient in one or several of these requirements. In addition, the instability of Pu(III) towards oxidation in oxalate solutions, especially in nitric acid media, requires the presence of reductants increasing the uncertainty of these studies. , ons of oxalate are reported in [57GEL/MAT2] The data in near-neutral soluti [67GEL/MOS] . Data at several temperatures have also been reported , [57GEL/MAT3] in these publications. In addition solubilities in K 4 are given in ox solutions at pH ≈ 2 . The authors of these studies derived equilibrium constants for the com- [58FOM/VOR] 3+ and oxalate, see Table VI-57. All these values are rejected plex formation between Pu in this review, see details for these publications in Appendix A. Likewise the enthalpy changes reported in [57GEL/MAT2] , [57GEL/MAT3] are not recommended by this review.

305 VI.12 Plutonium oxalate compounds and complexes 263 Table VI-51: Pu-oxalate compounds reported in the literature. Compound Reference(s) Pu(III) (ox) 10H O [63CHA] , [65JEN/MOO] ⋅ [67BUR/POR] , [69SUB/SIN] , Pu , 2 2 3 [94ZHI/MAT4] Pu ⋅ 9H O [57GEL/MOS] (ox) 2 3 2 LiPu(ox) 3.5H O [84ZUB/KRO] ⋅ 2 2 NaPu(ox) ⋅ O [84ZUB/KRO] 3.5H 2 2 ⋅ O [84ZUB/KRO] 3.5H KPu(ox) 2 2 CsPu(ox) 0.5H [84ZUB/KRO] O [83ZUB/KRO3] , ⋅ 2 2 NH [83ZUB/KRO3] ⋅ H Pu(ox) [84ZUB/KRO] , O 2 2 4 Pu(IV) , ⋅ 6H , O [61MAN/FRA] , [65JEN/MOO2] , [69SUB/SIN] Pu(ox) [80NIS] 2 2 [94ZHI/MAT3] ⋅ 2H Pu(ox) O [65JEN/MOO2] 2 2 H (ox) Pu ⋅ 9H [84KHA/AND] O 2 2 2 5 ) H Pu(ox) [91MAT/KAR] ⋅ n p.51 , O [58GEL/SOK] (NH 2 5 4 6 Na Pu(ox) [91MAT/KAR] ⋅ 5H , p.51 [58GEL/SOK] O 2 4 4 K [58GEL/SOK] ⋅ 4H Pu(ox) O p.52 , [91MAT/KAR] 4 2 4 [58GEL/SOK] Pu(ox) 4H ⋅ O K 6 2 5 K n ) [58GEL/ZAI3] ox ⋅ Pu(CO H O 2 2 3 3 K Pu(CO [58GEL/ZAI3] ) O ox ⋅ n H 2 3 4 3 Pu(CO ) (ox) ⋅ 3H O [58GEL/ZAI3] Na 2 2 4 2 3 Na H ) Pu(CO ox ⋅ n [58GEL/ZAI3] O 2 3 3 4 Pu(ox) ⋅ 2CO(NH [84AND/KIR] ) O ⋅ 2H 2 2 2 2 (Pu(ox) ) ⋅ n H O [77HOS/UEN2] ) (Co(en) 2 3 3 4 4 (Cr(en) n (Pu(ox) [77HOS/UEN2] ) O ⋅ ) H 2 3 4 4 6 (Cr(CO(NH ) [72HOS/UEN] ) O ) H (Pu(ox) n ) ⋅ 3 4 6 2 2 4 2 Pu(V) [64GEL/ZAI] ⋅ n H O ox , NH [73ZAI/ALE] PuO 2 2 4 Pu(VI) [65JEN/MOO] 3H ox O [58GEL/DRA3] , ⋅ , [69SUB/SIN] , [96BES/KRO] PuO 2 2 ⋅ H O [96BES/KRO] ox PuO 2 2 BaPuO (ox) ⋅ n H O [91MAT/KAR] p.74 2 2 2 (NH (PuO ) ) (ox) [96BES/KRO] 3 2 4 2 2 [96BES/KRO] ) PuO (ox) (NH 2 4 2 2 K (PuO ) (ox) ⋅ 4H O [96BES/KRO] 2 2 2 3 2

306 VI Discussion of data selection for oxalate 264 Table VI-52: Literature studies reporting the solubility of Pu(III)-oxalate. t Comments Ref. Reported Ionic Media (ºC) solid phase + a Pu O [H (ox) ] = 0.22 to 3.7 M; ·9H Experimental details not available. 3 2 2 [H ox] = 0.14 to 0.5 M 2 + Experimental details not available. b Pu(III) oxalate ] = 0.22 to 3.7 M [H 3+ 2 3 [Pu ] [H ox] 10 − 2 =≈× 210 K +6 [H ] (ox) 3 M K ox, I not constant; and s Sodium formaldehyde sulfoxylate used a 20 ≤ O ·9H c Pu 2 3 2 2 ≤ 2.4 M K ox at = constant (KCl) t atm. Ionic strength no reductant in N I 2 2 t specified. 4-6 hours equilibration. Solid no ox ) up to 0.7 M (NH 70 4 2 characterised. Experimental details not available. Presuma- 0.3 M K d Pu ox and 0.1-2 M of either 20 O ·9H (ox) • 2 2 3 2 r ly the same conditions as fo SO . , HCl or H b HNO 3 2 4 0.1-2 M HNO • [57GEL/MAT2] and 0.28 M . 3 ox. (NH ) 4 2 and 0.64 M H ox • 0.5-2 M HNO 3 2 ox] = 1.8 and 2.4 M 20-90 [K 2 3+ (ox) [Pu ·9H ] measured polarographically “immedi- O 25 0.17 to 0.87 M K e Pu ox, 2 2 3 2 pH 2.7 to 4.2. ately” after precipitation. Solid not character- = 3 (KCl), 0.35 to I Some data also at ised. ox, pH ≈ 4. 0.95 K 2 Ascorbic acid as reductant. Equilibration ox and O 21 0.09 to 0.3 M H (ox) ·10H f Pu 3 2 2 2 from under-saturation. Filtered samples. 0.5 to 3.1 HNO 3 Solid not characterised. 23 and Hydrazine or ascorbic acid as reductant. O ·10H g Pu (ox) 0.6 to 2 M HNO 2 3 2 3 ox d Precipitation, 0.5 to 3 hour equilibration, an 0.12 and 0.18 M H 2 paper filtration. Solid not characterised. , 0.2 M H 30-68 1.45 M HNO ox and 0.18 M 2 3 N H 4 2 s ox and 21 h Pu(III) oxalate 0.02 to 0.45 M H Hydrazine and hydroxylamine nitrate a 2 , with NaNO reductants. Precipitation and 50 hou r 0.2-2 M HNO added to 3 3 equilibration followed by centrifugation. I = 2.3 M; constant Solid not characterised. ox and 3 M also 0.02 to 0.45 M H 2 HNO 3 ox and 0.02 to 0.45 M H 50 2 0.5 & 1 M HNO added , with NaNO 3 3 to constant I = 2.3 M Ascorbic acid used as reductant. Both pre- O ·10H i Pu ≈ 25 0.02 to 0.2 M H ox and (ox) 2 3 2 2 0.5-2 M cipitation experiments (0.5 hour equilibra- HCl or HNO 3 tion) and equilibration from under-saturation (1 hour shaking and 0.5 hour settling). Separation by centrifuge. Solid not character- ised. a: Thomas & Warner, eds., 1944 (ref.5, p.421 in [54CUN] ) e: [58FOM/VOR] b: Reas & Connick 1946 (ref.156, p.335 in [54HIN] ) f: [67BUR/POR] [71CHE2] , [57GEL/MAT3] , [67GEL/MOS] g: [57GEL/MAT2] c: [87SUE/HU] ) h: [67GEL/MOS] d: Moskvin 1957 (ref.34, p.31 in i: [94HAS/KHE] , [95HAS/KHE]

307 VI.12 Plutonium oxalate compounds and complexes 265 Two values for the solubility constant of Pu(III)-oxalate have been reported. K = − 9.7 for: log A value of 10 s + 3+ Pu ox H n O(s) + 6 H ⋅ U 2 Pu n + 3 H O(l) (VI.46) ox(aq) + H 2 2 3 2 2 was reported in [54HIN] originating from a technical report by Reas and Connick cf (1946), . Table VI-52. It is probable that this value was calculated from the four solu- bilities tabulated in p.421 of [54CUN] . It has not been possible for this review to re- trieve the original reports, and it is not clear if this constant refers to standard conditions log K (VI.46) , [67GEL/MOS] (zero ionic strength). The other value was reported in 10 s = 9.44 for the reaction given above, which was recalculated in − , [57GEL/MAT2] − , [67GEL/MOS] log 24.79 for reaction: K (VI.47) = to be [57GEL/MAT3] s 10 3+ 2 − Pu n H n O(s) U 2 Pu O(l) (VI.47) ox H ⋅ + + 3 ox 2 2 2 3 Corresponding values at 70 ° C were also reported. The authors used the data at all ionic strengths without applying activity coefficient corrections when evaluating these constants. All literature values for the solubility constants (Reactions (VI.46) and (VI.47)) and the derived enthalpy changes are rejected by this review. The original solubility data available in the literature was examined by this re- view in order to try to find a qualitative chemical model. A simple model would consist 32 n − Pu(ox) of the complexes with = 0 to 4. There are no other reliable studies based n n on independent experimental techniques to establish the composition of the complexes Section VI.12.2.1. The formati on of oxalate complexes is con- cf. in the aqueous phase, − 2 , that is, at low acidities. strained by the solubility data at varying concentration of ox The model is compared to some of the experimental data in Figure VI-39 and Figure VI-40, which show that although there is a relatively large spread in the solubilities, the data display a consistent pattern, suggesting that probably the same solid phase was present in all these studies. No difference is apparent in systems containing either HCl or different reducing agents ). etc (ascorbic acid, hydrazine, or HNO 3 By varying the equilibrium constants of the model it is possible to see that, 5 − Pu(ox) in the given the spread of the experimental data, it is not necessary to include 4 model. Furthermore, the calculations show that protonated complexes are not required. The dotted curves in Figure VI-39 and Figure VI-40 were obtained using the following equilibrium reactions: + 3+ 2+ Pu (VI.48) O(l) + H U PuOH + H 2 + 3+ Pu H ox O(s) + 6 H ⋅ U 2 Pu O(l) (VI.46) + 3 H ox(aq) + n H n 3 2 2 2 2 + 2 − 3+ Pu U Pu(ox) + ox (VI.49) 3+ 2 − − Pu U (VI.50) Pu(ox) + 2 ox 2 3 − 3+ 2 − Pu(ox) Pu + 3 ox (VI.51) U 3

308 VI Discussion of data selection for oxalate 266 lue for Reaction (VI.48) [2003GUI/FAN] with the NEA-selected va , and the following ο ο ο log (VI.46) = − 7.5; (VI.50) = β log K β (VI.49) = 8.5; log approximate values: s 10 1 10 2 10 ο log β 12.7; and (VI.51) = 12.1. Furthermore, the SIT model described in Appendix B 10 3 was used with parameters that included the possible weak complex formation between 3+ 1 . and either chloride or nitrate Pu Figure VI-39: Comparison of experimental Pu(III) concentrations in solutions in contact with Pu(III)-oxalate and varying concentrations of the oxalate ligand. All data were ≥ 2. The dotted line has been calculated using the model described in the obtained at pH text. The arrow shows the value corresponding to the solubility of Pu(III)-oxalate in [57GEL/MAT2] water . -1.0 -2.0 TOT -3.0 [Pu (III)] -4.0 10 [57GEL/M AT2] log [67GEL/M OS] -5.0 [58FOM /VOR] -6.0 1 0.001 0.01 0.1 10 [o x] / M ribed above: the experimental data it There are a few caveats to the model desc tries to describe has several deficiencies; the solid phase was not characterised in any of the investigations; and (VI.50). The unexpected (VI.51) < β β ly weak formation con- 3 2 3 − Pu(ox) might indicate that either there are serious systematic deviations in the stant of 3 experimental data, or the model is inade quate: for example other species are perhaps formed that have not been considered, or the assumptions in the values of the 3+ parameters are erroneous. A study on the Yb -oxalate system [2000GAM/WOO] ε− showed that the possibility of formation of ternary solid oxalates with alkali metals 3+ cannot be excluded, adding further uncertainty to the interpretation of the Pu -data. 1 − 3+ 3+ − ) = action coefficients were used: NO The following specific ion-inter ε (Pu ,Cl ε (Pu , ) = 0.23 3 − − 2+ 2+ 1 − − 1 + 2 − − 1 + − ⋅ mol kg ε (K ) = 0.39 kg ,Hox ⋅ ) mol ε (PuOH , ε (K ,Cl , , ,ox (PuOH ) = 0.07 kg ⋅ mol NO ) = , and ε 3 Z Z 1 − M X = – 0.01 kg mol ⋅ ( . For all other ions the fo llowing approximation was used: ) = 0.15 + 0.15 M , ε X − − 1 + (( Z × ε mol ) = 0.15 kg − 1) + ( , (H ⋅ + 1)), giving for example Z . Pu(ox) X M 2

309 VI.12 Plutonium oxalate compounds and complexes 267 In summary, the available studies on the solubility of Pu(III)-oxalate have sev- eral shortcomings, and although a simple model may be used to describe the data quali- tatively, no thermodynamic values are recommended by this review. Solids of composition MPu(ox) ⋅ n H O where M is an alkali metal or ammo- 2 2 cf . Table VI-51. No thermodynam nium ion have been reported, ic data are available for these solids. Figure VI-40: Comparison of the experimental solubility data for Pu(III) oxalate: a) in solutions containing ammonium oxalate and nitric acid; and in solutions containing b) oxalic and nitric acids. The dotted lines have been calculated using the model described in the text. −1.0 [67BUR/POR] in HNO3 3 ≈ 2 M [HX] [67GEL/M OS] in HNO3 3 −2.0 TOT [67GEL/M OS] in HCl [71CHE2] in HNO3 3 −3.0 [87SUE/HU] in HNO3 3 [Pu(III)] 10 [94HAS/KHE] in HNO3 3 −4.0 log [94HAS/KHE] in HCl −5.0 0.0 0.6 0.4 0.2 / M ox] [H 2 TOT −1.0 ≈ [H 0.2 M ox] 2 −2.0 TOT −3.0 [Pu(III)] 10 log −4.0 −5.0 3.0 0.0 2.0 1.0 [HNO ] / M 3

310 VI Discussion of data selection for oxalate 268 Solid plutonium(IV) oxalates VI.12.1.2 Pu(IV) oxalate is relatively insoluble and may be used to separate plutonium from in- dustrial aqueous streams or as a precursor for the preparation of plutonium oxide, see for example [86WEI/KAT] 6H O, and X-ray . The precipitate is a hexahydrate, Pu(ox) ⋅ 2 2 . A modified Latimer method- [65JEN/MOO2] powder data for this solid is reported in ology was used by Moskvin to estimate the standard entropy of Pu(ox) 6H ⋅ O: 2 2 1 ο − 1 − ·mol (298.15 K) = 372 J·K [73MOS] . S m The solubility of Pu(ox) ⋅ 6H O has been studied and reported by many groups, 2 2 cf. Table VI-53. Although an effort has been made by this review to retrieve all the literature on this subject, it is quite possible that some technical reports have been missed. Table VI-53: Literature studies reporting the solubility of Pu(IV)-oxalate. Reference t Ionic media Comments Reported solid phase (ºC) 3 Precipitation of the solid in 0.06 cm [44OCO2] . Pu(IV) oxalate 25 [HNO ] = 0.1 to 6 M; 3 ox] = 0.05 to 0.2 M [H 2 days equilibration. 2 Sampling and analytical method not reported. 10 days equilibration. ⋅ 6H [49REA] O 25 [HNO ] = 0.75 M; Pu(ox) 3 2 2 ox] = 0 to 0.4 M Sampling method not reported. [H 2 Data analysed to r etrieve equilibrium constants. See Appendix A. 4-6 hours equilibration. Pu(ox) ] = 1 M; 6H , ⋅ O 20 [HNO [58GEL/MOS2] 3 2 2 ) Sampling method not reported. ox] = 0 to 0.35 M [58MOS/GEL] [(NH 2 4 etrieve equilibrium Data analysed to r constants. See Appendix A. Pu(ox) ⋅ 6H ] = 3.8 M; O 20 [HNO [58MOS/GEL3] The same experimental method as in 3 2 2 [(NH ox] = 0 to 0.35 M [58MOS/GEL] . Data analysed to re- ) 4 2 trieve equilibrium constants. See Ap- pendix A. Pu(ox) ⋅ 6H 6 hours equilibration. O 20 [(NH ) ox] = 0 to 0.26 M 4- [58MOS/GEL] 2 2 2 4 Sampling method not reported. Data analysed to r etrieve equilibrium constants. See Appendix A. (Continued on next page)

311 VI.12 Plutonium oxalate compounds and complexes 269 Table VI-53 (continued) Reported solid t Ionic Media Comments Reference (ºC) phase Pu(ox) The same experimental method as in SO 6H , O 20 [HNO [58GEL/MOS2] ] or [H ⋅ ] or 4 3 2 2 2 1M ] [58MOS/GEL] . Data analysed to re- ≤ [HClO [58MOS/GEL3] 4 trieve equilibrium constants. See Ap- pendix A. 14-21 days equilibration. Pu(IV) oxalate 27 [HNO ] = 0 to 3.52 M; [61MAN/FRA] 3 ox] = 0 to 0.6 M Sampling method not reported. [H 2 [67ABR] 18 hours equilibration. 24 [HNO Pu(IV) oxalate ] = 0.6 to 5.3 M; ≈ 3 2 − Paper rejected, see Appendix A. ] = 0.1 and 0.25 M [ox 15 minutes equilibration. ] = 0 to 3.5 M; 6H O ≈ 21 [HNO ⋅ Pu(ox) [67RIC] 2 2 3 Data disregarded because of too short [H ox] = 0 to 0.3 M 2 reaction time. ] = 3 M; 22 to 95 [HNO 3 ox] = 0.05 to 0.3 M [H 2 20 hours equilibration. [83CHH/GOP] ] = 1 to 6 M; 6H Pu(ox) O ? [HNO ⋅ 2 3 2 ox] = 0 to 0.5 M Data reported only in a graph. [H 2 ] = 0.5 to 4 M; Pu(ox) O ? [HNO ⋅ 6H 20 hours equilibration. [86RAO/PIU] 2 3 2 [H ox] = 0, 0.1, or 0.2 M Data reported only in a graph. 2 Not clear if the data is reported as mg- –1 –1 or as mg-solid·L . Pu·L 40 hours equilibration. [2001GUO/LIA] Pu(IV) oxalate 25 [HNO ] = 0.2 to 1 M; 3 ox] = 0.05 to 0.7 M Not clear if the data is reported as mg- [H 2 –1 –1 or as mg-solid·L . Pu·L Most of the publications reviewed are oriented towards the parameterisation of nt of correct equilibrium conditions or a industrial separation processes, and attainme careful description of the experimental methodology was not prioritised. The references where the data have been analysed in terms of equilibrium constants are indicated in Table VI-53. The rest of the studies are of technical type, and in some of them the solu- –1 and it is not clear if the data are given as mass of Pu(IV)- bilities are reported in mg·L –1 oxalate per litre or as mg-Pu·L (this is indicated in Tabl e VI-53); the difference is a ≈ 2.2, i.e ., 0.34 log [67RIC] -units in Pu-concentration. References [67ABR] , factor of 10 see comments in Appendix A, and the data from the other are not considered reliable, cal characterisation of Pu-s eparation processes are of publications aiming to techni [86RAO/PIU] , [61MAN/FRA] , , , [83CHH/GOP] [44OCO2] qualitative interest only . [2001GUO/LIA] The majority of the solubility data have been obtained in strong acids (mainly 3+ 4+ ) because both disproportionation (for example: 3 Pu + + 2 H 2 Pu O(l) U HNO 2 3 2+ + 3+ 4+ + PuO U + H + H O(l) + 4 H PuOH ) and hydrolysis (Pu ) are suppressed at high 2 2

312 VI Discussion of data selection for oxalate 270 acidities. In the absence of added oxalic acid the solubility of the solid increases with acidity: 4+ + Pu(ox) 6H U Pu O(cr) + 4 H + 2 H ⋅ ox(aq) + 6 H O(l) 2 2 2 2 Already from the earliest solubility experiments [44OCO2] it could be deduced − 2 4+ is strong: even in solutions contain- and ox that the complex formation between Pu + ing 1 M H the solubility is increased by adding H ox: ≥ 2 2 + − Pu(ox) U 6H O(l) Pu(ox) ⋅ + 2 H ox(aq) + 6 H O(cr) + H 2 2 2 2 3 The references where the data have been analysed in terms of equilibrium con- [49REA] stants are considered to be most reliable , [58GEL/MOS2] , [58MOS/GEL3] , , and the corresponding solubility and complex formation constants are [58MOS/GEL] listed in Table VI-54 and Table VI-58. However, in none of these publications is the before nor after the experiments. Therefore, these solu- solid state characterised, neither bility data may not be recommended by this review. In addition it has been reported that Pu(IV) oxidises oxalic acid, at least in acid solutions. The process is relatively slow at but it is quite fast at ≈ 100 . In C [67RIC] , [88NIK/DZY] ° [49REA] room temperature nitric acid solutions the Pu(III) formed is oxidised back to Pu(IV), cf. Section VI.12.2.2. The relative decrease in oxalic acid concentration may be quite large in solutions where Pu(IV)-oxalate is allo wed to dissolve without added oxa lic acid, with increased meas- ured solubilities as a consequence. It is therefore not surprising that there is a large variety of reported solubilities in solutions where neither oxalate nor oxalic acid has been added, sometimes larger than a fact or of ten, as shown in Figure VI-41. Nevertheless, it is deemed that it may be useful for the reader to have the ex- perimental solubilities re-evaluated in order to obtain approximate values of equilibrium en planning future experiments, etc . The data tabu- constants, for example as a guide wh [58MOS/GEL3] , , and in [58GEL/MOS2] , at 0.75 M HNO [49REA] lated in 3 at 1 M HNO have therefore been reanalysed by this review using [58MOS/GEL] 3 H ox(aq) as a ligand, assuming that the ⋅ 6H solid phase in these studies was Pu(ox) O 2 2 2 and using the SIT model, see the descriptions in Appendix A. The results of these evaluations are summarised in Table VI-55. The agreement between the equilibrium constants obtained from the two sets of data is relatively good, but the agreement be- , [58MOS/GEL3] , [58GEL/MOS2] tween the model results and other data reported in is less satisfactory, see the comments in Appendix A for [58MOS/GEL] [58MOS/GEL3] . Taking into account that none of these publications reported a charac- terisation of the solid phase, the values listed in Table VI-55 should be regarded as qualitative. The other published solubilities, considered to be less accurate, are com- pared in Figure VI-41 with values calculated using averages of the re-evaluations listed in Table VI-55. The figure illustrates the spread of the solubilities obtained for technical show that applications. Figure VI-41 and the figure in Appendix A for [58MOS/GEL3] the agreement between the model and the majority of the experimental data is qualita- tively correct, but the low quality of the published results prevents any data selection.

313 VI.12 Plutonium oxalate compounds and complexes 271 Figure VI-41: Experimental Pu(IV) concentrations (converted to molal units) measured in solutions in contact with Pu(IV)-oxalate. The upper di agram shows data in strong acids (up to 2 M) without added oxalic acid or excess oxalate (black squares are data in , all other data were obtained in HNO ). The lower diagram shows data in 1 M HClO 3 4 (except for [49REA] which is at 0.75 M) as a function of added oxalic acid HNO 3 (except for [58GEL/MOS2] , [58MOS/GEL] where ammonium oxalate was added). The curves were calculated assuming the formation of complexes according to: − n 42 + O(cr) + ( − 2) H n ox(aq) U O(l); with the 6H Pu(ox) ⋅ + 6 H 4) H + (2 n − Pu(ox) 2 2 2 2 n ο ο K K log log following equilibrium constants: n = 0) = 13; 8.0; ( − = 1) = − ( n s s 10 10 ο * ο * ο = 4) = ( 4.6; = 2) = K log . Table K cf ( n = 3) = − 4.3; 8.4, log log − K − ( n n 10 10 s s 10 s VI-55. In the lower diagram the continuous and dashed lines corresponds to calculated solubilities as a function of added H ) ox, respectively. ox and (NH 2 4 2 −2.0 [49REA] [58GEL/M OS2], [58M OS/GEL3] −3.0 [58GEL/M OS2], [58M OS/GEL3] [Pu] [61M AN/FRA] 10 log [83CHH/GOP] −4.0 [86RAO/PIU] −5.0 1.0 2.0 0.0 + [H ] (molal) [44OCO2] −3.0 ) (0.75 M HNO [49REA] (0.75 M HNO3) 3 [58GEL/M OS2], [58M OS/GEL] −3.5 ((NH4)2ox) ((NH ox) ) [Pu] 4 2 10 [61M AN/FRA] log −4.0 [83CHH/GOP] [86RAO/PIU] −4.5 GUO/LIA] [2001 0.001 0.01 0.1 1 ox] (molal) ) ox] or [(NH [H 2 4 2

314 VI Discussion of data selection for oxalate 272 Table VI-54: Literature solubility constants involving solid Pu(IV) oxalate. Ionic Medium C) log K References ( t ° 10 s − 2 4+ U Pu(ox) + 2 ox ⋅ 6H + 6 H O(cr) O(l) Pu 2 2 2 , 1 M HNO [58MOS/GEL] + 20 − 21.4 [58GEL/MOS2] 3 0 to 0.35 M (NH ) ox 4 2 0.5 to 1 M HClO − 21.3 4 − 20.1 + 3.8 M HNO 3 ox ) 0 to 0.35 M (NH 2 4 + 4+ Pu(ox) O(l) O(cr) + 4 H ⋅ U Pu 6H + 2 H ox(aq) + 6 H 2 2 2 2 11 [58MOS/GEL3] 20 0.5 to 1 M HClO − 4 + 2+ Pu(ox) O(cr) + 2 H U ⋅ Pu(ox) 6H + H O(l) ox(aq) + 6 H 2 2 2 2 [49REA] 7.51 − 25 0.75 M HNO + 3 ox 0 to 0.4 H 2 [58MOS/GEL] 7.51 20 − + 1 M HNO 3 ) ox 0 to 0.35 M (NH 2 4 Pu(ox) O(l) ⋅ 6H (aq) + 6 H O(cr) U Pu(ox) 2 2 2 2 25 − 4.52 [49REA] + 0.75 M HNO 3 ox 0 to 0.4 H 2 [58MOS/GEL] 4.49 20 − 1 M HNO + 3 ox ) 0 to 0.35 M (NH 2 4 20 − 4.49 [58MOS/GEL3] 0.5 to 1 M HClO 4 2 − + Pu(ox) O(cr) + H ⋅ U O(l) 6H Pu(ox) ox(aq) + 2 H + 6 H 2 2 2 2 3 [49REA] 3.12 − 25 + 0.75 M HNO 3 ox 0 to 0.4 H 2 + 1 M HNO 20 − 3.17 [58MOS/GEL] 3 0 to 0.35 M (NH ox ) 2 4 4 − + Pu(ox) + 6 H O(cr) + 2 H O(l) ox(aq) U ⋅ + 4 H Pu(ox) 6H 2 2 2 2 4 [58MOS/GEL] 0.02) ± (4.15 − 20 1 M HNO + 3 ox ) 0 to 0.35 M (NH 4 2 Several ternary Pu(IV) oxalates have been reported, cf . Table VI-51. No ther- modynamic data are available for any of these compounds. The precipitation of pluto- nium(IV) oxalate using the ions tris(ethylenediamine) Co(III), hexamine Cr(III) and hexaurea Cr(III) has been reported in [72HOS/UEN] , [77HOS/UEN2] . Mixed oxalate-carbonate compounds have also been synthesised : these solids are reported to be readily soluble in water. [58GEL/ZAI3]

315 VI.12 Plutonium oxalate compounds and complexes 273 Table VI-55: Results from re-analysis performed by this review of literature data on Pu(IV) oxalate solubility. All equilibrium constants are in molal units, in standard conditions (zero ionic strength) and at the temperature of the original experiments. 3+ 3+ 2+ ) and nitrate complexation (PuNO ) and Pu(OH) Corrections for hydrolysis (PuOH 2 3 are included in the equilibrium constants. See Appendix A for more details on the calculations. C) log t ( ° Original references K ° Background electrolyte s 10 † used in the experiments + 4+ 6H O(cr) + 4 H ⋅ U Pu Pu(ox) + 2 H O(l) ox(aq) + 6 H 2 2 2 2 0.75 M HClO 25 [49REA] ≤ 13 − 4 (and 0.001 to 0.4 H ox) 2 − [58GEL/MOS2] ≤ 20 13 , , [58MOS/GEL3] 1 M HNO 3 ) ox) (and up to 0.35 M (NH [58MOS/GEL] 2 4 + 2+ Pu(ox) O(cr) + 2 H + H U Pu(ox) ⋅ 6H O(l) ox(aq) + 6 H 2 2 2 2 0.75 M HClO [49REA] 0.3) 25 − (8.1 ± 4 (and 0.001 to 0.4 H ox) 2 [58GEL/MOS2] 0.1) ± 20 − (7.9 [58MOS/GEL3] 1 M HNO , , 3 ) ox) (and up to 0.35 M (NH [58MOS/GEL] 4 2 Pu(ox) ⋅ 6H (aq) + 6 H O(l) U Pu(ox) O(cr) 2 2 2 2 0.75 M HClO − (4.58 ± 0.04) [49REA] 25 4 (and 0.001 to 0.4 H ox) 2 [58GEL/MOS2] 0.1) ± (4.6 − 20 , 1 M HNO , [58MOS/GEL3] 3 ) ox) [58MOS/GEL] (and up to 0.35 M (NH 2 4 2 − + Pu(ox) O(cr) + H + 6 H ox(aq) U ⋅ 6H Pu(ox) O(l) + 2 H 2 2 2 2 3 0.75 M HClO [49REA] 0.03) 25 − (4.26 ± 4 (and 0.001 to 0.4 H ox) 2 [58GEL/MOS2] 20 − (4.37 ± 0.05) 1 M HNO , , [58MOS/GEL3] 3 ox) ) [58MOS/GEL] (and up to 0.35 M (NH 4 2 4 − + Pu(ox) ⋅ ox(aq) U + 6 H 6H Pu(ox) O(cr) + 2 H + 4 H O(l) 2 2 2 2 4 0.75 M HClO − (8.2 ± 0.5) [49REA] 25 4 (and 0.001 to 0.4 H ox) 2 1) 20 − (9 ± [58GEL/MOS2] , , [58MOS/GEL3] 1 M HNO 3 ) ox) (and up to 0.35 M (NH [58MOS/GEL] 2 4 reported here correspond to I †: The equilibrium constants = 0. VI.12.1.3 Solid plutonium(V) oxalates The preparation of NH = 3 and 6, has been described in PuO n ox ⋅ n H O, with 2 4 2 quite soluble in aqueous me- , [73ZAI/ALE] . These solids are apparently [64GEL/ZAI] dia but can be prepared by adding alcohol to solutions containing 4-6 g of Pu per litre. No studies reporting solubility or thermodynamic data have been found for Pu(V) ox- alate compounds.

316 VI Discussion of data selection for oxalate 274 Solid plutonium(VI) oxalates VI.12.1.4 ox O was synthesised and analysed in [58GEL/DRA3] , where it is reported ⋅ PuO 3H 2 2 osively when h eated over 180 that this solid decomposes expl ° C. The standard entropy using a modified Latimer method: of this solid was estimated by Moskvin [73MOS] − 1 − 1 ο ⋅ mol S (298.15 K) = 290 J . ⋅ K m The solubility of Pu(VI) oxalate has only been reported in acid media [58GEL/DRA2] [58GEL/DRA3] , [59MOS] , possibly because in , , [58DRA/MOS2] cf neutral solutions the reduction of Pu(VI) by oxalate is expected to be faster, . Section VI.12.2.2. The solubility valu es were originally publishe d by Gel’man and Drabkina , and later modelled by the same authors in collaboration with Moskvin [58GEL/DRA3] 2 − , ox(aq) and . The formation of PuO [58GEL/DRA2] PuO (ox) was [58DRA/MOS2] 2 22 equilibrium consta nts involving PuO assumed. The reported values for ox 3H O are ⋅ 2 2 listed in Table VI-56. The experimental methodology is not systematically described in these papers, for example it is not stated how the solution was separated from the solid after the equilibration time (filtration or centrifugation ?). The solid phase was charac- fore the dissolution experiments, and the terised only once by chemical analysis be equilibration time used, two hours, might have been too short: the authors state that equilibrium was “almost complete” after this time. The interpretation of the data is also doubtful: solubilities at several ionic strengths are generally mixed and there is no de- scription of any activity coe fficient correction. The oxalate protonation constants used K seems too = 4.19 and log K K = 0.97, and the value of by the authors were log 2 10 1 10 1 ese deficiencies, the equilibrium con- 1 M (see Table VI-8). Because of th ≈ large for I stants in Table VI-56 are disregarded by this review. An approximate chemical equilibrium model is obtained in this review by fit- , [58DRA/MOS2] ting calculated Pu(VI) oxalate solubilities to the experimental data , [58GEL/DRA3] . The calculations involve the SIT model for activity [58GEL/DRA2] coefficients described in Appendix B and the following equilibrium reactions: − 2+ 2 ⋅ PuO + ox O(cr) (VI.52) O U PuO 3H ox + 3 H 2 2 2 2 2+ 2 − + ox PuO ox(aq) (VI.53) U PuO 2 2 2+ − 2 − 2 PuO PuO (ox) + 2 ox (VI.54) U 2 22 ο A reasonably good fit (see Figure VI-42) is found using K (VI.52) = 10, log s 10 − 2+ ο ο (VI.53) = 7.2, and log β β (VI.54) = 11.6; together with ε ( log ≈ PuO , ) NO 2 10 1 2 3 10 2+ − − 1 UO NO , ( ± 0.03) kg ⋅ mol ε (which requires neglecting the Pu(VI)-nitrate ) = (0.24 2 3 + − + 2 − + 2 − (H (H ,ox ,Hox ) = ) = ε (H ε , ) = ε PuO (ox) complex formation), and 22 − 2 − 1 ) = 0 kg ε (H (PuO ox(aq),HNO ε ox(aq),HNO ⋅ mol is ) = . In these acid solutions ox 2 2 3 3 − 2 + not an important species, and the results are not sensitive to (H ε ). However, the ,ox − ox(aq)]/[Hox ] ratio, modelled solubilities in this system depend on the evaluated [H 2 − + (H ,Hox ), as it ε and the calculated de-protonation of oxalic acid varies somewhat with is assumed that ε 0. Nevertheless, practically the same results are ≈ ox(aq),HNO ) (H 3 2

317 VI.12 Plutonium oxalate compounds and complexes 275 − − 1 + ε − 0.3) kg ⋅ mol ,Hox obtained with . The equilibrium con- (H ) in the range (+ 0.3 to stants given above may be recombined with the selected protonation constants to give: 2+ + PuO U 3H ⋅ O(cr) + 2 H + H ox(aq) + 3 H O(l) (VI.55) ox PuO 2 2 2 2 2 PuO ⋅ 3H PuO O(cr) U ox O(l) (VI.56) ox(aq) + 3 H 2 2 2 2 2 − + PuO ox O(l) (VI.57) ox(aq) U + 3 H 3H ⋅ PuO (ox) + 2 H O(cr) + H 2 2 2 2 22 ο ο ο K log K log log K − − 4.35, (VI.56) = − 2.8 and (VI.57) = 4.05. These (VI.55) = s s s 10 10 10 = 0) agree within exp approximate equilibrium constants (for ected uncertainties with I , [58GEL/DRA2] (for an unknown ionic strength) and [58DRA/MOS2] those reported in they may be used to describe qualitatively the reported solubilities for the Pu(VI)- oxalate system in nitric acid. There are how ever several reservations, expressed above, on the quality of the experimental data and therefore no equilibrium constants derived from these solubilities are recommended. However, the importance of Pu(VI) in oxalate solutions is marginal taking into account the instability of Pu(VI) towards reduction by α -radiation and the reduction of Pu(VI) by oxalate and oxalic radiolysis produced by acid in aqueous solutions. cf. Table VI-51. No ther- Other oxalate solids of Pu(VI) have been reported, modynamic data are available for these compounds. Table VI-56: Literature solubility constants involving solid Pu(VI) oxalate. Reference t C) log ( K Ionic Medium ° s 10 2+ 2 − 3H PuO O(cr) U O(l) + 3 H PuO + ox ox ⋅ 2 2 2 2 1.1 to 3.1 M HNO (20 ± 1) − 9.20 [58DRA/MOS2] 3 a [58DRA/MOS2] 9.85 − (20 ± 1) + 1 M HNO 3 ox ) 0 to 0.4 M (NH 4 2 2+ + PuO 3H O(l) O(cr) + 2H ox(aq) + 3 H U ox ⋅ + H PuO 2 2 2 2 2 b (20 ± 1) − 4.07 + [58DRA/MOS2] 1.1 to 3 M HNO 3 ox 0 to 0.75 M H 2 O(l) ox ⋅ 3H ox(aq) + 3 H PuO U PuO O(cr) 2 2 2 2 [58DRA/MOS2] 2.66 − 1) ± (20 1.1 to 3 M HNO + 3 ox 0 to 0.75 M H 2 + 1 M HNO [58GEL/DRA2] 2.57 − ± 1) (20 3 0 to 0.4 M (NH ox ) 2 4 2 − + PuO 3H + 3 H O(cr) + H + 2 H ox(aq) U O(l) ox PuO (ox) ⋅ 2 2 2 2 22 (20 2.95 − 1) ± [58GEL/DRA2] 1 M HNO + 3 ox ) 0 to 0.4 M (NH 4 2 [58GEL/DRA2] . a: using their own exp. data in b: using literature values for the protonation constants of oxalate, the authors converted 3H O(cr). K ⋅ = − 9.23 for the solubility product of PuO ox this value into log 2 s 2 10

318 VI Discussion of data selection for oxalate 276 [58DRA/MOS2] , Figure VI-42: Comparison of some of the experimental solubility data , [58GEL/DRA3] for Pu(VI) oxalate in solutions containing nitric acid. [58GEL/DRA2] The dotted line has been calculated using the model described in the text. −1.2 −1.2 [HNO ] /M: 3 ≈ ] [HNO 1 M 3 −1.6 −1.6 1.1 2 −2.0 −2.0 [P u(V I)] [P u(V I)] 3.08 10 10 lo g −2.4 −2.4 lo g −2.8 −2.8 00.20.4 00.250.50.751 ) ox] /M [(NH /M [H ox] added 2 4 added 2 Aqueous plutonium oxalate complexes VI.12.2 VI.12.2.1 Plutonium(III) oxalate complexes [57GEL/MOS] , [67GEL/MOS] , The oxidation of Pu(III) in oxalate solutions is fast . Therefore, studies on the complex formation between Pu(III) and [83ZUB/KRO3] ducing agents in the solutions, oxalate require the presence of re e.g ., hydrazine, and this introduces some uncertainty to all experimental studies of this system. Several Pu(III)-oxalate complexes have been proposed in the literature: a pro- − − 32 n = 2 to 4. The equilibrium , as well as with n Pu(ox) Pu(Hox) tonated complex, n 4 constants reported in the literature for this system are listed in Table VI-57. Except for a [58GEL/MAT] cation-exchange study , all values were obtained from solubility investi- gations, see also the discussion of the solubility of Pu(III) oxalate in Section VI.12.1.1. − Pu(Hox) The formation of reported in the cation-exchange study [58GEL/MAT] is not 4 substantiated by the experimental data, see Appendix A. None of the other equilibrium constants reported are accepted in this review , see details in Appendix A. Likewise the , [57GEL/MAT3] are not recommended by enthalpy changes derived in [57GEL/MAT2] this review either. Approximate values of equilibrium constants for qualitative model- ling of this system may be found in Section VI.12.1.1.

319 VI.12 Plutonium oxalate compounds and complexes 277 Table VI-57: Literature stability constants for Pu(III) oxalate complexes. Method Ionic medium ° C) log t K Reference ( 10 − − 3+ 2 U Pu(ox) Pu + 2 ox 2 20 9.31 [57GEL/MAT2] , [57GEL/MAT3] sol 4.5? M KCl 0-0.7 M (NH ) ox 70 7.94 2 4 cix 1 M NH ± 1 9.15 Cl 20 , [58GEL/MAT] [57GEL/MAT3] 4 − 3 2 − 3+ Pu + 3 ox Pu(ox) U 3 sol 18.7 [57GEL/MAT2] , [57GEL/MAT3] 4.5? M KCl 20 ox 70 16.2 0-0.7 M (NH ) 2 4 25 3 M KCl 1 10.66 [58FOM/VOR] pol ± − 5 2 − 3+ U + 4 ox Pu Pu(ox) 4 4.5? M KCl 20 27.6 [57GEL/MAT2] , [57GEL/MAT3] sol 0-0.7 M (NH ) ox 70 24.8 2 4 sol, pol 3 M KCl ± 1 11.62 [58FOM/VOR] 25 − 3+ − Pu U Pu(Hox) + 4 Hox 4 Cl 20 ± 1 10.95 [57GEL/MAT3] , [58GEL/MAT] cix 1 M NH 4 Plutonium(IV) oxalate complexes VI.12.2.2 In aqueous solutions not containing ligands the disproportionation of Pu(IV) is en- : hanced by low acidities, e.g. 2+ 4+ 3+ + PuO 2 Pu + 2 H + (VI.58) O U + 4H 3 Pu 2 2 4+ Due to its high electric charge, Pu is expected to fo rm stronger complexes with most ligands as compared with the other valences of plutonium. Therefore it is to be expected that the tetravalent state of plutonium is stabilised against disproportiona- tion by the presence of oxalate and other ligan ds. There is however the potential risk of oxalic acid being oxidised by Pu(IV): 4+ + 3+ O (aq) U 2 Pu C + 2 CO + H (aq) + 2 H 2 Pu 2 2 4 2 –1 ο . = − (261 ± 8) kJ·mol G ∆ rm In nitric acid solutions the Pu(III) ions produced are easily oxidised back to Pu(IV) in a complex set of reactions (see e.g ., p.269 in [54CON] or p.76 [79CLE] ): 4+ 3+ + 3+ 4+ + − Pu + 3 H U 2 Pu N + HNO (aq) + H O O(l); Pu U + HNO (aq) + H + 2 Pu 2 2 2 3 O(l) + NO(aq); etc . The oxidation of oxalic acid is relatively slow at room tempera- + H 2 ture [49REA] but it is quite fast at ≈ 100 ° C [88NIK/DZY] .

320 VI Discussion of data selection for oxalate 278 The potential risk for disproportionation (Reaction (VI.58)) and the strong hy- drolysis of this cation imposes that studies of complex formation are performed in acid itions the reactions of com- media where oxalate is fully protonated. Under these cond plex formation may be written: 4+ 42 n − + Pu H Pu(ox) + ox(aq) U + 2 n n (VI.59) H 2 n Data has been presented in the literature supporting the existence of complexes − 42 n n = 1 to 4, see Table VI-58. Pu(ox) with n Table VI-58: Literature stability constants for Pu(IV) oxalate complexes. Method Ionic medium ° C) log t K References ( 10 4+ 2 − 2+ U Pu + ox Pu(ox) sol 1 M HNO + 0 to 0.35 M (NH [58MOS/GEL] ox 20 8.74 [58GEL/MOS2] , ) 2 4 3 (18 ± 3) 10.7 sol 0.3 M HNO [69MIK] 3 10 (9.75 0.01) [77RAM/RAM] ± dis 1 M HClO 4 dis 1 M HClO 25 (9.73 0.01) [76BAG/RAM3] ± 4 25 (9.425 ± 0.008) [77RAM/RAM2] dis 1 M HClO 4 + 3.9 M NaNO 0.08) 25 (7.84 ± [83CHO/BOK] dis 0.1 M HNO 3 3 1 M HNO + 3 M NaNO (8.24 ± 0.05) 3 3 + 2 M NaNO (8.32 ± 0.05) 2 M HNO 3 3 0.04) ± + 1 M NaNO (8.33 3 M HNO 3 3 3.9 M HNO (8.30 ± 0.09) + 0.1 M NaNO 3 3 0.05) ± (8.29 4 M HNO 3 − 4+ 2 Pu + 2 ox U Pu(ox) (aq) 2 + 0 to 0.35 M (NH ) sol 1 M HNO ox 20 16.92 [58GEL/MOS2] , [58MOS/GEL] 2 3 4 10 (17.64 0.01) [77RAM/RAM] dis 1 M HClO ± 4 dis 1 M HClO ± 0.03) [76BAG/RAM3] 25 (17.37 4 25 (16.5 ± 0.2) [77RAM/RAM2] dis 1 M HClO 4 + 3 M HaNO 0.09) 25 (15.02 ± [83CHO/BOK] dis 1 M HNO 3 3 2 M HNO + 2 M HaNO (14.71 ± 0.08) 3 3 0.05) + 1 M HaNO ± (14.56 3 M HNO 3 3 (15.17 ± 0.08) 4 M HNO 3 2 − 4+ − 2 + 3 ox Pu(ox) U Pu 3 [58MOS/GEL] + 0 to 0.35 M (NH sol 1 M HNO ) ox 20 23.40 [58GEL/MOS2] , 3 2 4 sol 0.02 to 0.14 M (NH ) ox 20 20.6 [58MOS/GEL] 4 2 [77RAM/RAM2] 25 (24.1 ± 0.2) dis 1 M HClO 4 4 − 4+ 2 − Pu Pu(ox) + 4 ox U 4 sol 1 M HNO + 0 to 0.35 M (NH ) ox 20 27.5 [58GEL/MOS2] , [58MOS/GEL] 2 4 3 pol 3 M KCl (25 ± 1) 27.48 [58FOM/VOR] (Continued on next page)

321 VI.12 Plutonium oxalate compounds and complexes 279 Table VI-58 (continued) Method Ionic medium C) log ( K References t ° 10 2+ + ox(aq) Pu(ox) + 2 H Pu(IV) + H U 2 98 (2.34 ± 0.09) [88NIK/DZY] kin 5 M (H,Na)NO 3 5 M ? HCl (2.0 ± 0.2) 2+ + + H U Pu(ox) (aq) + 2 H ox(aq) Pu(ox) 2 2 [49REA] 25 2.98 + sol 0.75 M HNO 3 ox 0 to 0.4 M H 2 2 − + U Pu(ox) (aq) + H Pu(ox) + 2 H ox(aq) 2 2 3 [49REA] 25 1.40 sol 0.75 M HNO + 3 ox 0 to 0.4 M H 2 3+ + + ox(aq) PuOH PuOH(ox) + H + 2 H U 2 dis 0.1, 2 and 4 M HNO 20-25 7.38 [72SOL/IVA] , [74SOL] 3 No evidence has been found in the literature for the formation of protonated 6 − 3+ Pu(ox) complexes such as PuHox is suggested by the solid . The existence of 5 Pu(ox) O (Table VI-51). Data from ion-exchange experiments for solutions at ⋅ 4H K 2 6 5 6 with K ox concentration varying between 0.065 and 0.74 M suggested the ≈ pH 2 6 − Pu(ox) stoichiometry . Ionic strength effects were however not taken [59ERM/BEL2] 5 into account, and the results are therefore inconclusive. Measurements of solubilities of on VI.12.1.2), and of sodium ⋅ 6H O (references listed in Secti Pu(IV) oxalate, Pu(ox) 2 2 oxalate, Na ox [68MAT/KRO] , give no evidence for such a complex, although these 2 studies extend only to oxalate concentrations below 0.4 M, and the solid phase was not + was proposed in characterised in any of the solubility studies. The complex PuOH(ox) [72SOL/IVA] [74SOL] , but these references are considered unreliable, see the discus- , sion in Appendix A. 42 − n Pu(ox) complexes (Table VI-58) are The studies on the stability of the n based on either solubility or solvent extraction data. The only exceptions are the inves- tigation on the rate of oxidation of oxalic acid at ≈ 100 ° C [88NIK/DZY] which is of no which is not credited by this interest here, and a polarographic study [58FOM/VOR] review. See comments in Appendix A on these two references. The references using solvent extraction techniques have been discussed in Ap- [72SOL/IVA] pendix A. The data from , [74SOL] cannot be accepted. The results at 25 ° C from [76BAG/RAM3] , [77RAM/RAM2] have been re-evaluated in this review are considered less reliable. The ( cf . Appendix A), but the data from [77RAM/RAM2] stepwise equilibrium constants obtained by this re-evaluation are listed in Table VI-59 together with the stepwise equilibrium constants from [83CHO/BOK] , which have been converted to molal units and corrected for Pu(IV)-nitrate complexation.

322 VI Discussion of data selection for oxalate 280 The papers where the solubility of Pu(IV) oxalate has been studied are dis- cussed in Section VI.12.1.2. The reported solubility data are considered to be only of qualitative nature because the solid phase was not characterised in the original refer- , [58MOS/GEL3] , [58MOS/GEL] . The data has been [58GEL/MOS2] , [49REA] ences re-evaluated in this review to provide the readers with qualitative values that support the solvent extraction data, see Table VI-59. These qualitative data may also be useful for the solubility of arylarsonates was used [69MIK] provisional modelling calculations. In to estimate the complex formation between Pu(IV) and several ligands, but as discussed in Appendix A, the reported data is not considered to be reliable. Figure VI-43 illustrates the ionic strength dependence of the equilibrium con- − 2+ 2 stants for Pu(ox) (aq) and Pu(ox) listed in Table VI-59. Given the large , Pu(ox) 2 3 uncertainties in most of the data, and the qualitative nature of the equilibrium constants derived from the solubility studies, no value is recommended. Mixed oxalate-carbonate compounds of Pu(IV) have been synthesised [58GEL/ZAI3] but there is no information available on the stability of mixed Pu(IV) 2 − 2 − and ox CO . complexes in aqueous solutions with both 3 Table VI-59: Stepwise stability constants for Pu(IV) oxalate complexes converted to 3+ molal units and corrected for the formation of PuOH , and when necessary for 3 + PuNO . 3 t ( ° Method Ionic medium Reference K C) log 10 + 4+ 2+ Pu(ox) U + 2 H ox(aq) Pu + H 2 a,b 0 25 ≥ 4.9 sol I [49REA] → b,c sol 0 20 ≥ 5.1 I [58GEL/MOS2] , [58MOS/GEL3] , → [58MOS/GEL] d 25 (5.3 ± 0.3) [76BAG/RAM3] 1 M HClO dis 4 [83CHO/BOK] 0.8) ± 25 (4.7 = 4 M (H,Na)NO dis I 3 + 2+ U Pu(ox) (aq) + 2 H ox(aq) + H Pu(ox) 2 2 a,b → 0 25 (3.5 ± 0.3) sol [49REA] I b,c → sol 20 (3.3 ± 0.1) I [58GEL/MOS2] , [58MOS/GEL3] , 0 [58MOS/GEL] d 25 (2.8 ± 0.7) [76BAG/RAM3] dis 1 M HClO 4 dis I = 4 M (H,Na)NO 25 (1.7 ± 1.1) [83CHO/BOK] 3 − 2 + Pu(ox) + 2 H (aq) + H ox(aq) U Pu(ox) 2 2 3 a,b → 0 25 (0.32 ± 0.05) sol [49REA] I b,c I → 0 20 (0.2 ± 0.1) sol [58GEL/MOS2] , [58MOS/GEL3] , [58MOS/GEL] d dis 1 M HClO 25 (2.9 ± 0.9) [76BAG/RAM3] 4 (Continued on next page)

323 VI.12 Plutonium oxalate compounds and complexes 281 Table VI-59 (continued) ° C) log Method Ionic medium K t Reference ( 10 2 − − 4 + + H U Pu(ox) Pu(ox) + 2 H ox(aq) 2 3 4 a,b → 0 25 − (3.9 ± 0.5) sol [49REA] I b b,c [58MOS/GEL3] , , [58GEL/MOS2] − sol 1.0) 20 0 → I (4.6 ± [58MOS/GEL] a: Equilibrium constants obtained , see Table al solubility data at 0.75 M HClO by re-evaluating the origin 4 VI-55 and Appendix A. b: As the solid phase was not characterised in the orig um constants are consid- inal reference these equilibri ered to be of qualitative nature. ed by re-evaluating the orig c: Equilibrium constants obtain inal solubility data at 1 M HNO , see Table 3 VI-55 and Appendix A. d: Results from a reanalysis of the orig inal solvent-extraction data, see Appendix A. Figure VI-43: Overview of the equilibrium constants at ≈ 25ºC listed in Table VI-59 for 42 − n n = 1 to 3. The background colour of the Pu(ox) with the Pu(IV) complexes n symbols indicates the experimental method: black for two phase distribution (solvent extraction) and grey for solubility. The solubility data are considered to be qualitative t characterised. The lines corre spond to a weighted linear because the solid phase was no fit, added to the graphs for illustrative purposes only. 9.0 4+ U ox(aq) + H Pu 2 References: 2+ + Pu(ox) + 2H 8.0 D [49REA ] 7.0 [58M OS/GEL3] + 10 1 [76BAG/RAM3] K 6.0 10 [83CHO/BOK] log 5.0 4.0 −1 0 1 2 3 4 5 I m 5.0 5.0 2+ (aq) + H + H U U ox(aq) ox(aq) Pu(ox) Pu(ox) 2 2 2 + + 4.0 2 − (aq)+ 2H + 2H Pu(ox) Pu(ox) 2 3 4.0 D D 3.0 - 6 + 2 3 2.0 3.0 2 K K 10 1.0 10 log 2.0 log 0.0 −1.0 1.0 −0.5 0.0 0.5 1.0 1.5 2.0 −1012345 I m I m

324 VI Discussion of data selection for oxalate 282 Plutonium(V) oxalate complexes VI.12.2.3 The disproportionation of Pu(V) is enhanced in oxalate solutions because of the 2+ + 4+ PuO PuO stronger complexes formed with and oxalate as compared with and Pu 2 2 . Pu(III) is not formed as a disproportionation product , [67ERM/KRO2] [58COO/FOR] able in oxalate me- because it is quickly oxidised in oxalate media. Even Pu(VI) is unst dia, but the reduction proceeds more slowly. Two Pu(V)-oxalate complexes with 1:1 and 1:2 stoichiometry have been pro- Hox(aq). Table VI-60 lists posed in this system, as well as a protonated complex: PuO 2 the equilibrium constants reported in the literature, which correspond to four independ- ent studies, The data from Kondrashova ., cited in the book of Gel’man et al . • et al [67GEL/MOS] , has not been published elsewhere and will not be dis- cussed any further. The coprecipitation (or sorption) method is judged to have too many un- • certainties and the data obtained using this method [78MOS/POZ] , , are not credited in this review, see the [79MOS/POZ] [79MOS/POZ4] in Appendix A. [79MOS/POZ] discussion of The remaining studies , [73ZAI/ALE] were performed in • [67ERM/KRO2] conditions where the disproportionation reactions of Pu(V) competed with the oxalate complexa tion, see also the comments on these two papers in Appendix A. In summary, Pu(V) disproportionates strongly in oxalate solution, and there are indications on the formation of Pu(V) oxalate complexes. However, the stoichiometry and stability of these complexes has not been acceptably determined. Table VI-60: Literature stability constants for Pu(V) oxalate complexes. t Method Ionic medium ° C) log References K ( 10 + 2 − − U PuO PuO ox + ox 2 2 kin (NH ) [67ERM/KRO2] ox; I ≈ 0.1 25 3.88 2 4 3.85 (NH ox; I ≈ 0.2 ) 2 4 b ix 0.05 ? 4.5 Kondrashova et al. [73ZAI/ALE] ? 3.7 ClO ≈ sp 0.15 NH 4 4 [78MOS/POZ] (3.95 Cl (20 ± 2) ± 0.12) (a) 0.5 M NH 4 + − 3 2 − U + 2 ox PuO PuO (ox) 2 22 ≈ I ox; kin (NH ) 0.1 25 6.7 [67ERM/KRO2] 2 4 (NH ) 6.8 ox; I ≈ 0.2 4 2 b ? 7.4 Kondrashova ix ? et al. 0.25) [78MOS/POZ] ± 2) (6.43 ± Cl (20 0.5 M NH (a) 4 (Continued on next page)

325 VI.12 Plutonium oxalate compounds and complexes 283 Table VI-60 (continued) Method Ionic medium ° C) log t K References ( 10 + 2 − + + ox U + H PuO Hox(aq) PuO 2 2 0.1 ) ox; I ≈ kin (NH 25 2.32 [67ERM/KRO2] 2 4 ) (NH ox; I ≈ 0.2 2.27 2 4 a: co-precipitation method. b: cited in section “Formation of acido complexes by Pu(V)” and in Table 22 of [67GEL/MOS] . VI.12.2.4 Plutonium(VI) oxalate complexes Pu(VI) in solution is prone to reduction cause d by the radiolysis products of water pro- . In addition to this instability, the presence of an [86WEI/KAT] duced by self-radiation excess of oxalic acid in acid solutions induces the reduction of Pu(VI) to Pu(IV) [59DRA/GEL] . The stability of Pu(VI) increases with the concentration of HNO , but it 3 decreases with increasing HCl concentration. In neutral ammonium oxalate solutions (of . The kinetics of unreported concentration) this process is quite fast [59DRA/GEL] Pu(VI) reduction in oxalate solutions in 1 M (Na,H)ClO have also been studied at tem- 4 . However, at room temperature the reduction in- ° C [73ZAK/ORL] peratures 70 to 90 [96BES/KRO] , , duced by oxalate is relatively slow: in the order of days [59DRA/GEL] [98REE/WYG] . − 2 The complexes PuO PuO (ox) have been proposed in two stud- ox(aq) and 2 22 ies, one based on solubility and the other using potentiometry. The equilibrium con- stants reported are summarised in Table VI-61. The solubility of plutonium(VI) oxalate, PuO ox ⋅ 3H O, was reported by Drab- 2 2 , [58GEL/DRA3] , [58GEL/MOS2] , [58GEL/DRA2] kina, Gel’man and Moskvin [67GEL/MOS] . These papers are discussed in Section VI.12.1.4, where approximate equilibrium constants at I = 0 are proposed to describe qualitatively the reported solu- bilities. appears to be well performed. The [73POR/PAO] The potentiometric study of authors used freshly prepared Pu(VI) to avoid self-reduction by radiolysis, and the oxi- dation state was checked spectrophotometrically both before and after each titration. An ± -units should cover all possible errors in this redox-unstable 0.5 log uncertainty of 10 = 1 M agrees with I log the values obtained = (9.35 ± 0.50) at β system. The value 2 10 from solubility measurements, see Section VI.1 2.1.4, within their expected uncertainties (perhaps 1 log -units or more). ± 10 No equilibrium constants are recommended for the Pu(VI)-oxalate system be- cause of several factors: the instability of Pu(VI) in oxalate solutions; the uncertainties concerning the accuracy of the solubility data; and that there is only a single additional 2 − , but not for PuO PuO (ox) ox(aq). study reporting only data for 2 22

326 VI Discussion of data selection for oxalate 284 Table VI-61: Literature stability constants for Pu(VI) oxalate complexes. Method Ionic medium t log K Reference C) ( ° 10 2+ − 2 + ox PuO U ox(aq) PuO 2 2 a 1 M HNO sol [58GEL/DRA2] ± 1) 6.64 (20 3 0 to 0.4 M (NH ox ) 4 2 2+ − 2 − 2 U + 2 ox PuO (ox) PuO 22 2 a (20 ± 1) 11.5 1 M HNO sol [58GEL/DRA2] 3 ox ) 0 to 0.4 M (NH 4 2 20 (9.35 ± pot [73POR/PAO] 1 M NaClO 0.50) 4 a: value obtained by the authors fr om a least-squares fitting of the experimental solubility data, using log [58DRA/MOS2] K = − 9.23 for the solubility product of PuO O(cr) published elsewhere ox . 3H ⋅ 2 2 10 s See also comments in Section VI.12.1.4. VI.13 Americium oxalate compounds and complexes VI.13.1 Solid americium oxalates VI.13.1.1 Am(III) compounds Among possible oxidation states, Am(III), Am(IV), Am(V) and Am(VI), solid oxalate compounds of Am(III) and Am(V) have been prepared and some of their properties were investigated. For Am(III), Am n H O is known to be formed as an insoluble (ox) · 2 2 3 solid in weakly acidic or neutral solutions and was used for the isolation of americium from solution. However, TGA ( [58MAR2] ), DTA ( [89VAS/KAL] , [90VAS/KAL2] ) changes from 11 to 0 de- ) of Am n (ox) O indicate that · n H and XRD ( [67WEI/MEE] 2 3 2 pending on the experimental conditions. In air, it loses its hydrated water and changes into Am ° by 240 ° C. It begins to decompose at 270 C and the conversion into (ox) 2 3 241 completes at ∼ 420 ° C. It was observed that solid O Am (ox) (s) is self-destroyed Am 3 3 2 2 241 α− radiation of the Am isotope and the compound is transformed into by Am (CO . The ex- ) [62LEB/PIR2] ·5H within 50-60 days O liberating gaseous CO 2 3 2 3 2 periments were usually carried out under the conditions to avoid this effect of radiolysis. Although attempts were made to measure the solubility [60LEB/PIR2] α− , [67BUR/POR] , [77ZAK/KOR] , [87PAZ/KRI] and to determine the solubility product O, as shown in Table VI-62, the present review could not select any H (ox) n · of Am 3 2 2 value due to the problems in the references , as discussed in Appendix A. These are mainly the uncertainties in identifying the dissolved species and the solid phase in equi- · x H , O (M = Na, K, NH librium with the solutions. Some ternary oxalates MAm(ox) 2 4 2 , Cs) have been isolated and th eir properties were examined ( [83ZUB/KRO] [83ZUB/KRO2] ), but no thermodynamic data are available for these compounds.

327 VI.13 Americium oxalat e compounds and complexes 285 Table VI-62: Literature data on the solubility product of the solid compounds of Am(III) with oxalate. ( ° C) n log Ionic medium t K Reference s ,0 10 3+ 2– · n H O O U 2Am (ox) + 3ox Am + n H 2 2 2 3 ∼ 0.3 M HClO 9 0.2 → 0 25 – 30.66 [60LEB/PIR2] 4 0.1 → 0 14 – 29 1 M nitiric acid –31 [77ZAK/KOR] ∼ ∼ → 0 14 – 19.2 water VI.13.1.2 Am(V) compounds Although MAmO (ox)· x H O (M = K, Cs) has been isolated and their properties were 2 2 , no thermodynamic data are available for Am(V) compounds. examined [82ZUB/KRO] Aqueous americium oxalate complexes VI.13.2 VI.13.2.1 Am(III) oxalate complexes A literature search by this review on the complex formation of americium-oxalate sys- tems revealed information concerning the aqueous complexes of Am(III) and Am(V). n 32 − Am(ox) there are many papers dealing with As shown in Table VI-63, , where n = 1 n and 2. Higher values of n may be possible, but their existence is not well established since the concentration range of oxalate in the experiments is not wide enough. Several report the [87PAZ/KRI] , [65SEK4] , [66STA] , [74BYK/PET2] , [60LEB/PIR2] papers 3 − Am(ox) , and its existence is fairly probable since there is no steric or stability of 3 – may be possi- other reason to exclude this speci es. Although the complexes with Hox 2– ble, their stabilities are expected to be much smaller than those with ox . Based on the discussion of the references in Appendix A, this review does not select any values for − 3 n Am(Hox) due to the lack of reliability of all the complexes of Am(III) of the type n – papers which report values for Am(III) complexes with Hox .

328 VI Discussion of data selection for oxalate 286 ation constants of oxalate complexes of Table VI-63: Literature data on the form Am(III). Ionic medium log K Reference Method t (°C) 10 2– + 3+ U Am(ox) Am + ox cix Cl → 0 20-25 5.99 [60LEB/PIR] 0.2 M NH 4 – 2– – + + ,K ) sol 0.7 M (H ,ox ∼ ,Hox = 0.2 I → 0 25 (7.30 ± 0.06) [60LEB/PIR2] ,ClO 4 [65SEK4] 25 (4.63 ± 0.08) [64SEK] , dis 1 M NaClO 4 + – 2– + – em 25 ,Cl ,NH ,ox 0.3 M (H ,Hox I = 0.1 ) (6.15 ± 0.16) [65STE/MAK] ∼ 4 0.5 M NaClO cix 25 (4.82 ± 0.03) [68AZI/LYL] 4 em 0.1 M 25 (5.25 ± 0.10) [71STE] dis 21 (4.58 ± 0.05) [83CAC/CHO] 0.7 M NaCl 6.89 → 25 0 [87PAZ/KRI] sol 0.10 Na(H)ClO 25 (5.01 ± 0.13) [90ROS/REI] em 4 ± 0.05 Na(H)ClO (5.11 0.13) 4 (5.38 ± 0.18) 0.01 Na(H)ClO 4 1.0 M NaClO 25 (4.66 ± 0.02) dis [96CHO/CHE] 4 3.0 M NaClO (4.64 ± 0.03) 4 (4.83 ± 0.06) 5.0 M NaClO 4 dis 0.3 m NaCl 25 (4.53 ± 0.01) [2001BOR/MOO] 1 NaCl (4.17 ± 0.05) m 2 (4.40 ± 0.04) NaCl m m (4.56 ± 0.04) 3 NaCl m NaCl (4.63 ± 0.04) 4 0.06) (4.57 ± m 5 NaCl − 2– 3+ U + 2ox Am Am(ox) 2 Cl → 0 20-25 0.2 M NH 10.15 [60LEB/PIR] cix 4 – + – 2– + → 0.7 M (H ,ClO = 0.2 ,ox I ,Hox sol ) ∼ 0 25 (11.46 ± 0.10) [60LEB/PIR2] ,K 4 cix 1 M NH Cl (9.7-10.0) [60MOS/KHA] 4 dis 1 M NaClO 25 (8.35 ± 0.09) [64SEK] , [65SEK4] 4 + – + – 2– 0.3 M (H ,Cl em ,ox ∼ ,Hox I = 0.1 ) 25 (10.54 ± 0.19) [65STE/MAK] ,NH 4 dis 0.1 M NH Cl 20 8.3 [66STA] 4 0.5 M NaClO [68AZI/LYL] 25 (8.60 ± 0.04) cix 4 ± 25 em 0.1 M 0.10) [71STE] (8.85 dis 0.7 M NaCl 21 (7.91 ± 0.10) [83CAC/CHO] 9.73 sol → 0 25 [87PAZ/KRI] (Continued on next page) .

329 VI.13 Americium oxalat e compounds and complexes 287 Table VI-63 (continued) Ionic medium (°C) log K Reference Method t 10 − 3+ 2– + 2ox U Am Am(ox) 2 0.10 Na(H)ClO em (8.16 ± 0.20) [90ROS/REI] 25 4 (8.30 ± 0.19) 0.05 Na(H)ClO 4 (8.96 ± 0.34) 0.01 Na(H)ClO 4 1.0 M NaClO 25 (8.35 ± 0.02) [96CHO/CHE] dis 4 3.0 M NaClO (8.56 ± 0.09) 4 (9.24 ± 0.09) 5.0 M NaClO 4 (8.22 m 25 0.3 ± 0.02) [2001BOR/MOO] dis NaCl 1 m NaCl (7.77 ± 0.08) 2 NaCl (8.22 ± 0.03) m m NaCl 3 ± 0.07) (8.42 4 m NaCl (8.46 ± 0.02) 5 NaCl (8.6 ± 0.1) m − 3 3+ 2– + 3ox Am Am(ox) U 3 – + + 2– – ,K sol ,ClO I = 0.2 ,ox ∼ ,Hox 0.7 M (H ) → 0 25 (12.3 ± 0.2) [60LEB/PIR2] 4 dis 25 (11.15 ± 0.07) 1 M NaClO [64SEK] , [65SEK4] 4 0.1 M NH dis Cl 20 11.8 [66STA] 4 25 11.58 [87PAZ/KRI] → 0 sol 3+ – 2+ U Am(Hox) Am + Hox → [87PAZ/KRI] sol 0 25 4.64 3+ – + 3Hox Am U Am(Hox) (aq) 3 cix 1 M NH Cl 9.64 [60MOS/KHA] 4 − 3+ – q Am Am(Hox) + 4Hox 4 cix 1 M NH Cl 11 [60MOS/KHA] 4

330 VI Discussion of data selection for oxalate 288 ation constants for oxalate complexes of Am(III) at 25 Table VI-64: Accepted form ° C used to derive the selected values. a, b K Ionic medium Reference Method log 10 2– + 3+ + ox U Am(ox) (VI.60) Am 1.1 m NaClO ± (4.61 dis 0.20) [65SEK4] 4 0.7 m NaCl (4.61 ± 0.10) [83CAC/CHO] dis ± 1.1 NaClO dis (4.64 m 0.10) [96CHO/CHE] 4 3.5 m NaClO 0.10) (4.57 ± 4 6.6 m (4.71 ± 0.20) NaClO 4 m NaCl (4.55 ± dis [2001BOR/MOO] 0.3 0.10) m NaCl (4.21 ± 0.10) 1 m NaCl 2 ± 0.15) (4.44 m NaCl (4.61 ± 0.20) 3 4 ± NaCl (4.70 m 0.20) m NaCl (4.66 5 0.30) ± − 3+ 2– Am U Am(ox) (VI.61) + 2ox 2 dis 1.1 m NaClO (8.31 ± 0.20) [65SEK4] 4 dis m NaCl (7.93 0.7 0.15) [83CAC/CHO] ± 1.1 m NaClO (8.30 ± 0.20) dis [96CHO/CHE] 4 m NaClO (8.42 ± 0.20) 3.5 4 6.6 NaClO m (9.00 ± 0.30) 4 (8.24 dis m NaCl ± 0.20) [2001BOR/MOO] 0.3 1 m NaCl (7.81 ± 0.20) m (8.26 ± 0.30) 2 NaCl 3 m NaCl (8.47 0.30) ± 4 m NaCl ± 0.40) (8.53 5 m NaCl (8.7 ± 0.5) 3+ 2– 3 − U + 3ox (VI.62) Am Am(ox) 3 dis 1.1 m NaClO (11.15 ± 0.40) [65SEK4] 4 ic strength given in the table. Uncertainties a: Refers to the reactions indicated, the ion are estimated by this review. y unit, and corrected for chlo- are those converted into molalit K b: The values of log 10 ride complex formation if ionic strength is controlled by NaCl.

331 VI.13 Americium oxalat e compounds and complexes 289 log K (VI.60) at each ionic strength (Table Figure VI-44: Fitting of the values of 10 VI-63) to the SIT equation. The solid line is drawn by using the result of the fitting given below. Solid line: I 0.509 2 ο 2 m log K = – 12 log K – – ∆ε I ∆ where z z D ∆ and = = D m 10 10 + I 11.5 m ο log K = 6.51 10 –1 2– + – + 3+ – ε ) – (Am , X , X (Am(ox) ) – ε (Na ε , ox = ) = – 0.33 kg·mol ∆ε –1 dotted line: (log (from fitting) K ) = ± 0.07, ∆ ( ∆ε ) = ± 0.04 kg·mol ∆ 10 –1 K ) = ± 0.15, ∆ ( ∆ ) = ± 0.10 kg·mol ∆ε (selected) (log broken line: 10 8.0 + 2- 3+ Am(ox) U Am + ox 7.5 D 12 7.0 + K 10 log 6.5 [65SEK] [83CAC/CHO] [96CHO/CHE] [2001BOR/MOO] 6.0 01234 I / m

332 VI Discussion of data selection for oxalate 290 . The solid line is drawn by using the (VI.60) against log K I Figure VI-45: Plot of m 10 result of the fitting given below. Solid line: I 0.509 2 ο 2 m I = = – 12 log K z – ∆ε D ∆ where z – ∆ K log and D = m 10 10 + I 11.5 m ο log K = 6.51 10 –1 2– + – + 3+ – ε ) – (Am , X , X (Am(ox) ) – ε (Na ε , ox = ) = – 0.33 kg·mol ∆ε –1 dotted line: (log (from fitting) K ) = ± 0.07, ∆ ( ∆ε ) = ± 0.04 kg·mol ∆ 10 –1 K ) = ± 0.15, ∆ ( ∆ε ) = ± ∆ 0.10 kg·mol (selected) (log broken line: 10 6.0 − 2 3+ + Am Am(ox) + ox U [65SEK] [83CAC/CHO] [96CHO/CHE] 5.5 [2001BOR/MOO] Κ 10 5.0 log 4.5 4.0 01234 m I /

333 VI.13 Americium oxalat e compounds and complexes 291 K (VI.61) at each ionic strength (Table log Figure VI-46: Fitting of the values of 10 VI-63) to the SIT equation. The solid line is drawn by using the result of the fitting given below. I 0.509 2 ο 2 m log = K log K – – ∆ε I ∆ where z z D ∆ and = – 16 = D m 10 10 + I 11.5 m ο log K = 10.71 10 –1 2– + – 3+ + − ε ,Na ) – ε (Am ∆ε , X ( ) – 2 Am(ox) (Na ε , ox = ) = – 0.54 kg·mol 2 –1 ∆ (log (from fitting) K ) = ± 0.12, ∆ ( ∆ε ) = ± 0.07 kg·mol dotted line: 10 –1 (log (selected) K ) = ± 0.20, ∆ ( ∆ ) = ± 0.10 kg·mol ∆ε broken line: 10 13 3+ − − 2 Am + 2ox Am(ox) U 2 12 D 16 + K 10 log 11 [65SEK] [83CAC/CHO] [96CHO/CHE] [2001BOR/MOO] 10 01234 / m I

334 VI Discussion of data selection for oxalate 292 log I (VI.61) against . The solid line is drawn by using the K Figure VI-47: Plot of m 10 result of the fitting given below. Solid line: I 0.509 ο 2 2 m log K log = – 16 ∆ – K ∆ε I D where = z z ∆ and – D = m 10 10 I + 11.5 m ο log K = 10.71 10 –1 2– + – 3+ + − ,Na Am(ox) (Am ) – , X ( ) – 2 ε (Na ε , ox = ) = – 0.54 kg·mol ∆ε ε 2 –1 ∆ K ) = ± 0.12, dotted line: ( ∆ε ) = ± 0.07 kg·mol ∆ (from fitting) (log 10 –1 broken line: (log (selected) K ) = ± 0.20, ∆ ( ∆ε ) = ± 0.10 kg·mol ∆ 10 10.5 [65SEK] 3+ − 2 − U + 2ox Am(ox) Am [83CAC/CHO] 2 [96CHO/CHE] 10.0 [2001BOR/MOO] 9.5 Κ 10 9.0 log 8.5 8.0 7.5 01234 I / m

335 VI.13 Americium oxalat e compounds and complexes 293 Based on the discussion for the literature studies on Appendix A, the constants listed in Table VI-64 are accepted in this review. The values in Table VI-64 are those converted into molality units and corrected for chloride complex formation when neces- 3+ 3+ – –1 − ) = ε (Am ± , Cl , ) = (0.49 ε 0.03) kg·mol ClO , the analyses (Am sary. By assuming 4 < 4 m in the form of: I are conducted for the data at m 0.509 I ο 2 m D = log K K – – ∆ε I where ∆ z log = D m 10 10 I 11.5 + m For the reaction: 3+ 2– + Am + ox Am(ox) (VI.60) U − − + 2 3+ + 2– ClO ClO (Am(ox) = – 12 and ∆ε ) – ε (Am = , ), ε ∆ ) – ε (Na z , ox , 4 4 and for the reaction, 3+ 2– − Am + 2ox Am(ox) (VI.61) U 2 − − + 2 3+ + 2– Am(ox) ClO z (Am = – 16 and ) – ε ∆ε = , ). ε (Na ) – 2 ε (Na ∆ , ox , 4 2 Figure VI-44 to Figure VI-47 show the result of the fittings. The obtained val- ues are: –1 K (VI.60) = (6.51 ± 0.07), ∆ε = – (0.33 ± 0.04) kg·mol log 10 –1 log K (VI.61) = (10.71 ± 0.12), ∆ε = – (0.54 ± 0.07) kg·mol . 10 As shown in the Figures, the values at the ionic strength lower than around 1.0 m show somewhat large deviations probably due to some overlooked systematic errors. 3+ − + 2– If we use (Am ) = (0.49 ± 0.03) and ε (Na , , ox ε ) = – (0.08 ± 0.01), the ob- ClO 4 + − + − ε , ) = ClO ∆ε ) = (0.08 ± 0.05) and give (Na ε , (Am(ox) Am(ox) tained values of 2 4 − (0.21 ± 0.08), which seem reasonable. Taking these details of the results in considera- tion, this review selects values for these reactions but with larger uncertainties as given in Table VI-65. For the reaction, 3+ − 2– 3 (VI.62) U Am(ox) + 3ox Am 3 there is only one value which is considered reliable as is given in Table VI-64. If we ο log K want to obtain by using the equation, 10 ο 2 K log = D (VI.63) K – ∆ z log 10 10 3 − 2 3+ − + ο where ε ∆ε = ε (Na log , K ∆ Am(ox) ) – ) – , the value of (Am = – 12, , z ClO 3 4 10 − 3 − 3 2– + + + Am(ox) Am(CO ) 3 ε (Na , ox ) = ε (Na (Na , ε ) ) is necessary. If we assume , 33 3 ο ± (0.23 0.10), ∆ε − ± 0.11) is estimated. This would give (VI.62) = log K = – (0.48 = 10 13.04. Considering the large uncertainty coming from the assumption for the ion inter- action term and scarcity of the available data, this review selects the value of 13.0 with a large uncertainty of 1.0 as is given in Table VI-65.

336 VI Discussion of data selection for oxalate 294 3+ rmation constants for the oxalate complexes of Am ° C. Table VI-65: Selected fo at 25 ο –1 ∆ε / kg·mol log K Reaction 10 2– + 3+ U Am(ox) (6.51 ± 0.15) + ox – (0.33 ± Am 0.10) − 2– 3+ 0.20) U Am Am(ox) + 2ox ± – (0.54 ± 0.10) (10.71 2 − 3 2– 3+ U + 3ox Am(ox) Am (13.0 ± 1.0) – (0.48 ± 0.50) 3 The selected values report ed in Table VI-65 yield: –1 ο + , 298.15 K) = – (1316.0 ± 5.2) kJ·mol (Am(ox) G ∆ fm ο − –1 Am(ox) G ∆ , 298.15 K) = – (2020.1 ± 6.1) kJ·mol ( fm 2 3 ο − –1 . ∆ Am(ox) ( , 298.15 K) = – (2713.3 ± 9.2) kJ·mol G 3 fm VI.13.2.2 Am(V) oxalate complexes At room temperature, americium(V) is relatively stable in oxalate solutions (Am(V) at the concentration level of 1 mM is reduced by 0.1 M oxalate by about 2 % within 1 hour). The process is independent of oxalate concentration, but strongly depends on , [82ZUB/KRO] ). Qualitatively, [74SHI/NIK2] proton concentration or temperature ( + + AmO forms oxalate complexes with similar stabilities as pO complexes. Two N 2 2 papers were found for the reactions of Am(V) with oxalates (Table VI-66): + 2– – AmO U AmO + ox (ox) (VI.64) 2 2 3 2– − + U AmO (ox) (VI.65) AmO + 2 ox 22 2 These papers give qualitative information about the stabilities of oxalate com- plexes of Am(V). However, because of the instability of Am(V) in the presence of ox- alate, their values may only be of importance in short term laboratory studies, and could not be selected due to the experimental problems in the reports, as discussed in Appen- dix A. Table VI-66: Literature data on the formation constants of Am(V) oxalate complexes t ( ° log K Reference Method Ionic medium C) 10 + 2– – (ox) U AmO + ox AmO 2 2 − + 2– – ,ox I ,Hox = 0.25 M (K , NO sp ) 25 (3.27 ± 0.05) [74SHI/NIK2] 3 I = 0.5 M NH Cl (3.08 ± 0.08) [79MOS/POZ4] (a) 4 − 3 + 2– + 2ox U AmO (ox) AmO 2 22 − – 2– + ) ,ox ,Hox I , NO sp = 0.25 M (K 25 (5.36 ± 0.09) [74SHI/NIK2] 3 a: co-precipitation method

337 VI.13 Americium oxalat e compounds and complexes 295 VI.13.2.3 Am(VI) oxalate complexes In solutions with excess H ox, Am(VI) was found to be reduced by oxalate and mainly 2 to disproportionate into Am(III) and Am(V) ( [85SHI] ). No thermodynamic data are available for these highly unstable Am(VI) complexes.

338

339 Chapter VII VII Discussion of data selection for citrate compounds and complexes Equation Section 7 VII.1 Introduction Citric acid (2-hydroxypropane-1,2,3-tricarbox ylic acid) has the chemical formula, HOOC-CH ; molecular weight: 192.123 -COOH (C -C(OH)-(COOH)-CH H O 8 6 2 2 7 − 1 mol g · ; CAS Registry Number: 77-92-9). The three carboxylic groups may dissociate in aqueous solutions at acidities that are perhaps most common in nature. In reactions cit, and the citrate ligand in and formulae in this review citric acid will be denoted as H 3 3 − aqueous solutions will be denoted as cit . The structures of citric acid and of citrate are shown in Figure VII-1. cit ⋅ H O(cr) [72ROE/KAN] and of citrate Figure VII-1: The structures of citric acid in H 3 2 . ⋅ 5.5H cit O(cr) [86VIO/ROD] in Na 2 3 H H 3− H O H H O H H H H O O O O O H O O O O H H O H O O Citrate has four functional groups, three carboxylates and one hydroxy group, all of which are potential coordination sites. As a result, citrate is a strong complexing 297

340 VII Discussion of data selection for citrate compounds and compl exes 298 agent because it may easily form chelate complexes. The hydroxyl group in organic ligands is, in general, a very weak acid and only deprotonates at pH > 12 in the absence of metal-ions. Furthermore the configuration of these four functional groups imposes steric constraints on the coordination and more than three groups cannot bind to the same metal ion. The non-bonded functional group(s) can be used as a bridge to a second metal ion, and polynuclear complexes are ther efore abundant in metal – citrate systems. , [78STR/KAR] Especially for citrate complexes, there is some evidence ( [83BAK/BAK] , [98MAT/RAP] ) that, depending on the size and charge density of the metal ion, only two carboxylates and the hydroxyl group can bind to the same metal ion. In this case, it is expected that the strong interaction between the metal ion and the hy- droxyl oxygen would lead to the deprotonation of the hydroxyl group at much lower pH. The polyprotic citric acid can form a number of ternary complexes of the type M cit . Hence, the general stoichiometric equation for the complex formation in metal H p q r cit as components is: – citrate systems using M and H 3 + p M(aq) + r H cit(aq) U M q H )H cit (VII.1) (aq) + (3 r – r q p 3 where charges have been omitted for simplicity. It is in general straightforward to de- not termine the stoichiometry of the complexes formed in Equation (VII.1), but it is the case for their constitution that describes how the different functional groups are bonded. ider: one is the isomers formed which There are two types of problem to cons , e.g. relate to the number of functional groups that are bonded in a certain complex, M(cit). The magnitude of the equilibrium constant provides an indication on this num- ber: the strength of the complex increases with the number of functional groups coordi- nated. Comparison of the equilibrium constants in the citrate system with those found in ligands with fewer functional groups (acetate, glycolate, oxalate, etc .) provide useful hints on the actual coordination. However, such comparisons are outside the scope of this review. The second type of problem has its origin in the so-called “proton ambiguity” that arises in many “standard” solution chemical methods such as potentiometry. With these methods it is in general not possible to determine the origin of the protons released 3– n n cit , = 3,2,1,0, – 1 or from in Equation (VII.1), they may come from the ligand H n coordinated water. Hence the thermodynamic equilibrium constants are the same for complexes such as M(cit) and M(OH)(Hcit), because they result in the same number of protons on the right hand side of Equation (VII.1). A second example is the complexes 4– cit is the fully deprotonated ligand. In this re- cit), where H M(OH)(cit) and M(H –1 –1 cit) unless we have clear view, we generally use the simpler notations M(cit) and M(H –1 evidence for a different constitution. The deduction of the constitution is facilitated if one has information from and other spectroscopic methods, but the structural data obtained NMR [78STR/KAR]

341 VII.1 Introduction 299 useful information. Some examples are from metal citrate compounds provide also shown in Figure VII-2 and Figure VII-3. The second one shows the structure of a com- plex involving two different metal ions. In the case of sodium, citrate coordination oc- curs with a protonated carboxylic group, and two bridging oxygen atoms belonging to deprotonated carboxylic and hydroxy groups. − 2 5 − [83BAK/BAK] [Ni(cit)(H O) ] , and (b) Fe(cit) Figure VII-2: The structures of: (a) 2 222 [98MAT/RAP] . a) O H O H 2 2 O 2− O Ni O O O O O H H O O O O O Ni O O O H H O 2 2 b) O O 5− O O O O O Fe O O O O O O O

342 VII Discussion of data selection for citrate compounds and compl exes 300 O) (H Figure VII-3: The structure of SbNa(cit) [91HAR/SMI] . 2 2 2 H O O O O O H O O Sb O O H O O O O H O 2 O Na H H O 2 VII.1.1 Metal ion citrates Several metal-ion salts with citrate have b een reported in the literature. The solubilities of these citrates might be of interest when modelling natural systems containing metal ions and citrate. The available information in the literature concerning the solubility of Ni, Zr, Tc, U, Np, Pu and Am is discussed in this section. No information was found for selenium citrate compounds. The thermodynamic properties of magnesium and calcium citrates are reviewed in Section VII.5. Information on the solubility of other metal-ion citrates is also briefly mentioned in this s ection to help the reader in estimating the pos- sible importance of citrate salts in any given system. It must be borne in mind that the solubility of all citrates is pH-dependent. The protonation of the citrate ligand at pH ≤ 6 will increase the solubility of these salts, see for example Figure VII-4. In the alkaline region mixed metal-hydroxo-citrate complexes may be formed, and the possibility of the precipitation of other solids must be taken into account, such as oxides, hydroxides, hydroxocitrates, etc . The formation of metal- citrates with different numbers of water molecules must also be considered when inter- preting solubility data, and the fact that hydration might be dependent on factors such as ionic strength ( i.e. etc . , activity of water), temperature, + ions is probably quite high, for example the re- The solubility of citrates of M . However, the [91AUK] cit(cr) in water at 25 ° C is 0.94 M ported solubility for KH 2

343 VII.1 Introduction 301 1 − ⋅ , i.e ., ≈ 0.6 mM ( cf . solubility of Ag(I) citrate at 25ºC is 0.284 g (1000 g water) [59VIN/GET] and references therein). 2+ -citrates have solubilities that vary by several orders of magnitude, depend- M ing on the metal cation. The solubility of Mg, Ca, Zn and Cd citrates is reported in [69SKO/KUM] as a function of pH. The solubility is found to follow the trend Mg > Ca 4.2 the solubility of Mg-citrate is 0.04 M (corresponding to ≈ > Cd > Zn. At pH [Mg] = 0.12 M), while for Ca and Zn the solubilities at the same pH are only 0.005 TOT and 0.0006 M, respectively. In pure water the solubility of magnesium and calcium cit- 0.05 M and ≈ 0.0015 M, respectively [93APE] ≈ [95ROB/GIA2] , , rates is found to be 2+ [2001CIA/TOM] . There is no information in the literature about the solubility of Ni 2+ the preparation procedure for UO citrate compounds. However, and 2 [Ni(cit)(H O)] ⋅ 4H O suggests that the citrates of nickel are quite soluble K 2 2 2 2 [83BAK/BAK] . Similarly the studies on the complex formation between citrate and 2+ 2+ and UO in aqueous solution indicate that it is possible to prepare solutions con- Ni 2 –1 . of these ions, ., [78KER/CHU] and [80VAN/KUC] e.g taining several mmol · L 3 − Smaller solubility limits for U(VI)-citrate may be achieved by using Co(NH ) as pre- 36 . cipitating agent [77HOS/UEN] 3+ -citrate compounds have been obtained with rare-earth elements, Al(III), M Fe(III), etc . Citrates of lanthanides have been prepared by Skorik et al . [65SKO/SER] , , . The solubilities of the rare-earth citrates are reported [69SKO/MAM] [66SKO/SER] I (3.0 ± 0.2) [65SKO/SER] , ≈ = 0.1 M to be in the range (0.3 – 3) mM at pH for [66SKO/SER] [69SKO/MAM] . At pH ≈ 6 these solubilities decrease slightly (to 0.2 – , ≈ 0.025 M [93APE] . The solu- 1 mM). For Fe(III)-citrate the solubility is reported to be ugh millimolar concentrations bility of actinide(III) citrates has not been reported, altho the studies concerning the complexation of of reactants have been used in several of Am(III) and Pu(III) with citrate, ., [72EBE/MOA] , e.g and [89POC] . [84BOU/GUI] 4+ -citrate. The solubility of Thorium citrate may be used as an example of a M Th(IV) citrate at I = 0.1 M has been reported to be ≈ 0.03 mM [67SKO/KUM] . The experimental studies on the complex formation between Zr(IV) and citrate have also been conducted with solutions < 0.1 mM, , and [75ZAI/NIK] . On [66KOR/SHE2] e.g. 4+ the other hand, for Pu [66NEB] , solutions in the millimolar range have been used [66NEB2] . This is also the case for the Tc(IV)-citrate system [77MUN2] . Somewhat lower solubility limits for the Th(IV)- and Pu(IV)-citrate systems may be achieved by 3 − using Co(NH ) as precipitating agent [77HOS/UEN] . 36 In conclusion, for most applications related to modelling the fate of metal ions in natural systems it is unlikely that the solubility limits of the corresponding citrate salts are reached. This is due to the fact that the concentrations of metal ions in such systems are in general too low, with the exception, perhaps, of calcium and to a lesser extent magnesium. In particular, for judging the safety of nuclear waste disposal facili- ties, the solubilities of radionuclide citrate salts appear to be too high to be used as a

344 VII Discussion of data selection for citrate compounds and compl exes 302 source-term limitation of either citrate or of radionuclides. However, high levels of cal- cium might impose limits on the maximum levels of citrate. ncentrations needed to reach the solu- As an example, the calculated citrate co bility limit of Ca O in CaCl (cit) solutions are presented in Figure VII-4. The fig- 4H ⋅ 3 2 2 2 ure shows: 2+ The common-ion effect: an increased Ca concentration decreases the solu- • bility of calcium citrate; + -concentrations (decreasing pH) in- The effect of side reactions: rising H • 2 − − cit(aq)) and the cit , H and H crease the protonation of the ligand (Hcit 3 2 + formation of protonated complexes (Ca(Hcit)(aq) and Ca(H cit) ). 2 There is no evidence for side reactions in alkaline conditions such as the for- , and therefore the solubilities etc. mation of hydroxocitrate complexes with calcium, shown in Figure VII-4 remain constant at pH > 6. Figure VII-4 was drawn using equilibrium constants selected by this review for protonation, complex formation and solubility. The activity of water was estimated with − 2+ (Ca ε ) in Table B.4. The ac- ,Cl equations (B.9), (B.10) and (B.11) using the value of tivity coefficients of ions were calculated using equation (B.4) and the specific ion in- 2+ − + − ε (Ca ε (H ) and ,Cl ,Cl )). As there is no information teraction parameters in Table B.4 ( 2+ − about the e.g ., for ε (Ca ε ,Ca(cit) -values for other ion pairs, ), a crude relationship be- ε ε (M,X) = 0.15 + and the charge of the interacting ions was used: tween the values of 1 − Z (( − 1) + ( Z . The following specific ion +1)) kg ⋅ mol × interaction coefficients 0.15 M X − 1 2+ 3 − (in kg (Ca ⋅ ,cit mol ) were obtained using this rough estimation method: ) = 0.00, ε − 2+ 2+ − 2 − 2+ (Ca ε ,H (Ca cit ,Hcit ) = 0.30, ε ) = 0.15, (Ca ,Ca(cit) ε ) = 0.30, and 2 + − cit(aq) and Ca(Hcit)(aq), were assumed to cit) ε ,Cl (Ca(H ) = 0.15. Neutral species, H 3 2 behave ideally. The results shown in Figure VII-4 indicate that even in saline Ca-rich waters substantial concentrations of citrate may occur without the precipitation of calcium cit- rate. The citrate levels appear to be high enough to be able to affect the transport of trace metals in natural waters. The results shown in Figure VII-4 also illustrate the ca- pabilities of the SIT model.

345 VII.1 Introduction 303 O in ⋅ 4H Figure VII-4: The total concentration of citrate at equilibrium with Ca cit 2 3 2 calcium chloride solutions as a function of pH at 25 ° C. 0.0 ] /molal [CaCl 2 0.00 −1.0 0.01 0.10 TOT 0.50 −2.0 [cit] 10 log −3.0 −4.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 pH VII.2 Citric acid At temperatures below 36°C the solid formed at equilibrium in the system citric acid- water is the monohydrate, H cit cit(cr) is formed ⋅ H O(cr). Above this temperature H 2 3 3 instead [37DAL] . The crystal structure for anhydrous citric acid has been reported by [69GLU/MIN] . et al . Glusker O(cr) cit·H H VII.2.1 2 3 , [62EVA/HOA] The standard heat capacity has been determined in . [82KRU/MIL] From these data, the following values are selected: ο 1 − − 1 O, cr, 298.15 K) = (268.05 , mol ⋅ K ⋅ 0.10) J ± C (H H cit ⋅ 2 3 p ,m 1 − 1 − ο O, cr, 298.15 K) = (283.6 ± 0.2) J ⋅ K (H S ⋅ mol cit ⋅ . H 3 2 m The selected heat of combustion of citric acid monohydrate has been deter- [58CHA/HOA] mined by Chappel and Hoare and the value re-evaluated by Domalski [72DOM] is: 1 − ο cit − H ∆ (1838.46 H ± 2.00) kJ ⋅ mol ⋅ O, cr, 298.15 K) = . (H 3 2 fm

346 exes VII Discussion of data selection for citrate compounds and compl 304 This value agrees with the calculation made in this review using the auxiliary data in Table IV-1. The standard Gibbs energy of formation is calculated to be: − 1 ο (H ⋅ H G O, cr, 298.15 K) = − (1473.3 ± 2.0) kJ ⋅ mol ∆ cit . 3 2 fm cit(cr) VII.2.2 H 3 et al The standard molar heat capacity and entropies determined by de Kruif . are adopted in this review, with the following assigned uncertainties, [82KRU/MIL] − 1 ο 1 − cit, cr, 298.15 K) = (225.4 (H 0.2) J ⋅ K ± C ⋅ mol , 3 ,m p − 1 − 1 ο S 0.2) J ⋅ K (H cit, cr, 298.15 K) = (252.1 ⋅ mol ± . 3 m cit(cr) may be calculated using the enthalpy The enthalpy of formation of H 3 change for the reaction: O(l) cit(cr) + H O(cr) (VII.2) H U H H cit ⋅ 3 3 2 2 ο H (VII.2) = − (11.3 ± 1.0) ∆ which has been determined calorimetrically to be rm − 1 − 1 mol kJ ⋅ (10.85 ± 0.19) kJ ⋅ mol [82KRU/MIL] [86APE] . Using the weighted − and − 1 ± 0.2) kJ ⋅ mol average, − (10.9 O(l), and , with the auxiliary data for H 2 ο cit (H H O, cr, 298.15 K) selected in Section VII.2.1, this review obtains the fol- ⋅ ∆ H 2 3 fm lowing selected value for the standard enthalpy of formation of H cit(cr) at 25 ° C: 3 1 − ο (H . cit, cr, 298.15 K) = H (1541.7 ± 2.0) kJ ⋅ mol ∆ − 3 fm from early determi- This agrees with value calculated by Domalski [72DOM] ο [64WIL/SHI] nations of the heat of combustion of this solid ∆ (H H , namely cit, cr, 3 fm − 1 . ⋅ mol 298.15 K) = − (1543.8 ± 8.0) kJ The standard Gibbs energy of formation is calculated from the values of the standard entropy and of the enthalpy of formation selected above: 1 − ο (H . cit, cr, 298.15 K) = − (1236.7 G 2.0) kJ ± mol ∆ ⋅ 3 fm cit(cr) has been evaluated calorimetrically The enthalpy of solution of H 3 − 1 ο ∆ = (19.3 ± 0.3) kJ ⋅ mol H [93BAL/BAR] by . sol m H cit(aq) VII.2.3 3 ο cit, aq, G ∆ (H The solubility of citric acid may be used to calculate the value of 3 fm 298.15 K) from the following reaction, H cit ⋅ H O(l) (VII.3) O(cr) U H cit(aq) + H 3 2 2 3 ο ο ο ο ∆ (VII.3) = ⋅ G O(cr)) ∆ (H cit cit(aq)) + (H G H ∆ (H ∆ O(l)) G − G 3 3 2 2 rm fm fm fm ο ο log K . (VII.3) = − =+ G a ∆ (VII.3) T a log /(R ln (10)) log 10 H O H cit 10 10 rm 32

347 VII.2 Citric acid 305 At 25°C, the solubility of the monohydrate, H O(cr), is quite high: cit ⋅ H 3 2 molal ± (8.45 [37DAL] 0.03) , , , [76LAG/AUB] , [55LEV] [82KRU/MIL] . [76LAG/AUB] are discarded: these au- . The results of Laguerie et al [87APE/MAN] thors determined the solubility of the monohydrate as a function of temperature, and their data give 207.5 g of monohydrate in 100 g of H O, which corresponds to 9.88 mo- 2 lal, in disagreement with all other studies. The osmotic coefficient of the saturated solution (8.45 m ) at 25°C is φ = log ± [95APE/DOV2] , i.e. , , a = − (0.108 0.001). Apelblat 0.005) (1.625 [55LEV] ± 10 H O 2 cit(aq)- et al . obtained expressions for the osmotic and activity coefficients of H 3 solutions [95APE/DOV2] , using a correction for the limited dissociation of the acid into + 2 − cit = 1.615 and H . From these equations, the values for the saturated solution are φ H 2 log γ = (0.509 ± 0.004). and 10 H cit(aq) 3 = log log + log am γ = Under equilibrium conditions then, H cit(aq) H cit 10 H cit 10 10 33 3 ± (8.45 ± 0.03) + (0.509 ± 0.004). Therefore: 0.004) = (1.436 log 10 ο K (VII.3) = (1.328 ± 0.004). log 10 1 − ο This gives ⋅ − (7.58 ± 0.02) kJ G mol , and results in the following ∆ (VII.3) = rm selected value: 1 − ο G cit, aq, 298.15 K) = − (1243.7 ± 2.0) kJ ⋅ mol ∆ (H . 3 fm The enthalpy change for reaction (VII.3) may be determined either calorimetri- cally or from the temperature dependence of the solubility of the saturated solutions. The values found in the literature are given in Table VII-1. ο ∆ (VII.3). H Table VII-1: Literature values for rm − 1 ⋅ ) Reference mol Method Enthalpy changes at 298.15 K (kJ cit(cr) H 3 a H ∆ ( m ≈ cal ± 0.09) 0.01) = (18.47 [82KRU/MIL] m sol [86APE] 0.07) ∆ ( m = 0.0200) = (18.21 ± cal H sol m H cit ⋅ H O(cr) 2 3 H m / ∂ T ∂ [55LEV] ∆ ( m ) = 29.8 sat sat sol m a 0.20) ∆ ( m ≈ 0.01) = (29.25 ± H cal [82KRU/MIL] sol m cal H ∆ ( m = 0.0203) = (29.06 ± 0.12) [86APE] sol m m ∂ / ∂ T [87APE/MAN] ∆ ( m ) = 26.0 H sat sat sol m m / ∂ T [95APE/DOV] ) = 28.98 H ∆ ( ∂ m sat sat sol m a: These values are given with negative sign in Table 6 of [82KRU/MIL] , apparently a misprint .

348 exes VII Discussion of data selection for citrate compounds and compl 306 The calorimetric values in dilute solutions should be more accurate than the values derived from the temperature dependency of concentration of saturated solutions. Selecting the average of the two values determined from solution calorimetry for H cit H O(cr) leads to the selected values: ⋅ 3 2 − 1 ο ο ∆ H mol ∆ (VII.3) = (29.11 ± 0.11) kJ ⋅ = H . m sol rm have determined the enthalpy . et al In addition, Dobrogowska [90DOB/HEP] of dilution of aqueous citric acid. Their data may be extrapolated to a saturated solution, 1 − ∆ . Therefore, m (298.15 K, ) = (2.8 ± 0.2) kJ ⋅ mol H giving an enthalpy of dilution sat m dil 1 − the average of the data at saturated conditions ( H m ∆ ) = (28.3 ± 2.5) kJ ⋅ mol ( ) sat sol m − 1 ∆ ± H 2.5) kJ ⋅ mol = (31.0 , which may be extrapolated to infinite dilution to give m sol agrees with the calorimetric value within the uncertainties. ο H ∆ (VII.3) determined calorimetrically and Using the value of rm ο ⋅ ∆ (H O, cr, 298.15 K) selected above leads to: cit H H 2 3 fm 1 − ο ± . cit, aq, 298.15 K) = − H ∆ 2.0) kJ ⋅ mol (H (1523.5 3 fm The standard entropy for aqueous citric acid may be calculated from the changes in entropy for Reaction (VII.3) and the standard entropies of H ⋅ H O(cr) (se- cit 3 2 . Table IV-1): O(l) ( cf lected in Section VII.2.1) and H 2 1 − − 1 ο S cit, aq, 298.15 K) = (336.7 ± 0.4) J ⋅ K . (H ⋅ mol 3 m As a consistency check, the enthalpy change for the reaction: cit(aq) (VII.4) U H cit(cr) H 3 3 1 − ο mol ∆ ± 3) kJ ⋅ (VII.4) = (18 , in agreement with the value H is calculated to be sol m 1 − ο . ± 0.3) kJ ⋅ mol H ∆ obtained calorimetrically [93BAL/BAR] (VII.4) = (19.3 sol m The partial molar heat capacity at infinite dilution for undissociated citric acid ο 1 − 1 − has been reported to be C ± 1.6) J ⋅ K cit, aq, 298.15 K) = (322.5 (H ⋅ mol 3 ,m p [89SIJ/ROS] . ο V (H cit, The partial molar volume of the undissociated acid in water is 3 m –1 3 , ·mol [55LEV] [89MAN/APE] , [89SIJ/ROS] , 298.15 K) = (113.5 1.5) cm ± [90APE/MAN] [90APE/MAN] determined . For the citrate ions Apelblat and Manzurola 2 − − –1 3 3 ο ο 0.5) cm and ·mol (Hcit 1) ± V V (cit , 298.15 K) = (88.5 ± , 298.15 K) = (72 m m 3 3 –1 − –1 ο , 298.15 K) = (99.7 ·mol (uncertainty cit , and estimated V ± 1.0) cm ·mol (H cm 2 m assigned by this review).

349 VII.3 Protonation constants for citrate 307 VII.3 Protonation constants for citrate Introduction VII.3.1 Citric acid normally behaves as a triprotic acid. The hydroxy-group has been found to dissociate only at high pH-values (> 12), although it might dissociate at substantially lower pH values to participate in the co-ordination of metal cations. In this review citric acid is denoted as H cit, and the citrate ion with the hydroxy-group dissociated is de- 3 4 − cit . noted as H − 1 The standard TDB nomenclature is used for the protonation of a ligand: − r (3) ] [H cit 4) ( + − r 3) − ( r r U cit cit + H H H = K − ( r 1) r r (4) − + r ] [H ] [H cit ⋅ (1) − r − r (3) [H cit ] ( r − 3) + 3 − r + U b H = cit r H cit r r r −+ 3 [cit ] [H ] ⋅ Using this nomenclature, the equilibrium constant for the deprotonation of the hydroxy-group is denoted as follows, 4 − + ⋅ [H cit ] [H ] + 4 − 3 − 1 − cit cit H U H b − = 1 − − 1 − 3 [cit ] − 4 cit is expressed as: and the protonation of H − 1 − 3 [cit ] + 3 − 4 − H U cit = + H cit K 1 − 0 + − 4 ⋅ [H ] [H cit ] − 1 1 − β . = () K Therefore, − 1 0 A large number of references ( ≈ 130) were found from a literature search for the acid-base equilibria of citrate. The majority of these references contain studies on metal complexation, where the authors needed values for the dissociation constants of citric acid under the same experimental conditions as the metal-complexation study. Because of the large number of references, it was advantageous to judge the quality of the experimental details quite rigorously. The following criteria were consid- ered when discarding references within the screening process: Clear indication must be given that the acid-base constants where determined ex- • perimentally in the actual study, and not taken from another publication. • The calibration method for the pH-electrodes must be indicated. They must have been calibrated in the concentration scale, and not with standard pH-buffers. That + is, “pH” must refer to log [H ] − . References were discarded when they reported 10 mixed equilibrium constants, i.e. , involving both proton activities and ligand con- centrations:

350 VII Discussion of data selection for citrate compounds and compl exes 308 (3) − r [H cit ] mixed r ⋅ = K r − r (4) a ] cit [H + r − (1) H In some cases it is reported that the glass electrodes were calibrated with stan- + dard buffers, and [H ] calculated from pH, for example with the Davies equation. This procedure is not accepted in this review. The total ligand concentration, the temperature, the ionic strength, and the nature • of the background electrolyte must be given. tant ionic medium must be used. For A background electrolyte providing a cons • ≤ 0.1 M this is quite difficult to achieve. For example, pH values I studies where I < 0.1 M without disturbing substantially the com- below 2 can not be reached at position of the ionic medium. In the case of citrate, with a charge of − 3, the total concentration that can be used is limited, otherwise the ionic strength is substan- 3 − tially increased. For example, if the back ground electrolyte is 0.1 M then [cit ] TOT ≤ 4 mM. In this review equilibrium constants at should be < 0.1 M were not con- I sidered because of the large junction potentials and the difficulties in keeping con- stant both the ionic strength and the nature of the electrolyte at such low concentrations. Studies were also rejected if they had been performed with − 3 [cit that was clearly too high for the constant ionic media concept. ] TOT The dissociation constants of citric acid reported in the following references were discarded from the review procedure because they did not fulfil one or more of the criteria indicated above: [28KOL/BOS] , [29BJE/UNM] , [49BAT/PIN] , [28SIM] , [51HEI] [52ELL2] , [53WAR/WEB] , [57LEF4] , [59LI/LIN] , [61LI/TAN] , , [62RYA/MAR] [66NEB] [63FUR/GIU] , [65TAT/GRZ] , [63MAT] , [68SPI/MAR] , , , [69LIT/PUR] [70BAR/BRI] , [70GRZ/TAT] , [71SKO/KUM] , [72BEL/KAZ] , , [73GOR/KHU] , [73HUB/HUS] , [73KHA/PET] , [73RAM/MAN2] , [74MEY] , [74PET/KHA] , , [75DAN/OST] , [75PEA/CRE] , [76NOW/CAN] , [75BRI/STU] [76VAN/KUC] [78USS/BOS] [77WIL] , [78FLY/KOR] , , , [81BOU] , , [77ROO/WIL] [81MAL/SEN] [82GAR/RAM] , [82INO/TOC] , [84BOU/GUI] , [84BOU/PET] , , [85RIZ/ANT] , , [86CAP/ROB] , [86SAL/ZHU] , [87FIN/DUF] , [86BAR/HAV] [87HYN/ODO] , , [87RAY/DUF] , [87KIT/ITO] , [88MEL/BAR] , [88GHA/MAN] [89MAN/APE] , [89PAP/ZIO] , [89YAD/GHO] , [90FIN/DUF] , [90WAN/YAN] , [91APE/BAR] , , [92DAN/ROB] , [92MIL/DOB] , [92POG/KAP] , [91CHR/CUM] [95PAP/ZIO] [2001JAN/HAR] , [96SCH/CON] , [2000YOS/OKA] , [96SAE/KHA] . , These references are not discussed in Appendix A. For other references a more detailed discussion is given in Appendix A [80ARE/CAL] , [80DAN/RIG] , [81CUC/DAN] , [83DAN/RIG] , [84DAN/OST] , [85DAN/ROB] [86CRU/WAT] , [90DAN/ROB] , [90DAN/ROB2] , [91BAP] , , [95LIS/CHO] , [96BOR/LIS] , [96XUE/TRA] , [99ROB/STE] , [2001SAR] .

351 VII.3 Protonation constants for citrate 309 + + + , Na The data reported in the remaining references for Li and tetraal- , K kylammonium ionic media are listed in Table VII-2. A number of other ionic media have been used in literature studies: – NH [81AMI/DAN] Cl 4 RbCl and CsCl – [90DAN/ROB2] – Sea water [99STE/GIA] , [2000ROB/STE] However, the number of studies performed with these salts is scarce, and the corresponding protonation constants were not included in this review. The uncertainties reported in the original publications were multiplied by a fac- ., including i.e tor (1.96) to obtain error limits closer to a 95% total uncertainty level, random and possible systematic deviations. In cases where no uncertainty limits were reported in the original publication, a value of ± 0.1 log -units was used in the least- 10 squares regression. Reported uncertainties below 0.01 in ± log K in the original pa- 10 ± 0.02 ( ≈ ± 0.01 × 1.96) in the data processing. pers were increased to When applying the SIT model described in Appendix B to the activity coeffi- cients of tetraalkylammonium halides, it may be shown that the specific ion-interaction + − coefficient, (R N ,X ) ε , depends on the ionic strength. A proper representation of the 4 + − ++ − − + data is achieved by setting ε⋅ (RN,X) ε ε + (R N ,X ) = . (RN,X)log [RN] 24 4 10 4 14 Because of this, the protonation constants of citrate in tetraalkylammonium salts was +(3) r − +(3) − r fitted to the SIT equations by setting (R N ,H cit ε = ) + (R N ,H cit ) ε r r 4 14 + +(3) − r ·log (R N ,H cit ) [R N ε ]. 4 10 r 24 Several references report data at other temperatures than 25°C. Those refer- ences that reported only data at another temperature in the range (18 to 37)°C were also included in the review procedure, and these log values were extrapolated to 25°C . K 10 enthalpies selected in section VII.3.7 . The corrections were obtained from reaction Ionic media corrections to the reaction en thalpies consist of two parts ( cf . Section V.3.6): a Debye-Hückel expression, and a specific ion interaction ( ∆ε In cases where no ) term . L − 1 − 3 data was available, ∆ε 10 5) ± × was set equal to (0 . The resulting reaction kg · mol L enthalpies are all relatively small, and because of the limited temperature interval in- volved ( ≤ ± 12°C), the calculated corrections for - K were always ≤ 0.07 log log 10 10 r units. If needed the uncertainties in the adjusted protonation constants were also in- creased following the error-propagation rules described in Appendix C for the additional uncertainty in the value of H ∆ . rm

352 VII Discussion of data selection for citrate compounds and compl exes 310 Table VII-2: Literature data on the protonation constants for citrate considered in this review. Data in were reported in molal units. italics K Reference Electrolyte t Method log I log K K log log K 10 3 1 10 2 10 0 10 (°C) (M) a 4.092 2.778 pot 1 Na(Cl) 18 5.135 [40ADE] 4.077 2.788 1.5 5.082 5.077 2.811 2 4.091 4.139 2.857 2.5 5.101 5.124 4.178 3 2.91 pot 0.1 (KCl) 19 10.82 5.67 2.95 [59OKA/KOL] 4.40 ) 25 4.10 2.74 [62FUR/CER] pot 0.5 (NaClO 5.26 4 5.68 4.35 2.87 [64CAM/OST] pot 0.1 (NaClO ) 20 4 ) 20 (5.68 ± 0.02) (4.38 ± 0.02) pot 0.1 (NaClO ± 0.03) [64TIM] (2.96 4 5.65 4.30 2.79 [65RAJ/MAR] pot 0.1 (KNO ) 25 3 5.34 1 2.63 4.11 sp 1 NaCl 25 4.168 [67FIS2] 2 4.188 3 4.284 4.420 4 4.223 sp 1 NaBr 25 2 4.276 3 4.396 4.542 4 sp 1 NaNO 25 4.173 3 2 4.168 3 4.231 4 4.328 sp 1 NaClO 25 4.237 4 2 4.314 3 4.434 4 4.601 4.240 sp 1 KCl 25 2 4.307 3 4.435 sp 1 KBr 25 4.282 2 4.400 3 4.536 sp 1 KNO 25 4.261 3 2 4.297 (Continued on next page)

353 VII.3 Protonation constants for citrate 311 Table VII-2: (continued) Method t I Electrolyte Reference K log log K K log log K 1 10 10 2 0 10 3 10 (°C) (M) 4.7 pot 3 LiClO 2.8 [68GUI] 25 4.2 4 (5.685 ± 0.017) (4.375 ± 0.013) (2.892 ± 0.013) [72KAN] pot 0.1 KCl 25 ) 25 2.70 3.90 pot 1 (LiClO [72MET/GUI3] 5.00 4 , [73BOT/VIC] 0.03) (4.16 ± 0.05) (5.18 ± 0.07) ) 25 >13 pot 2 Na(ClO ± (2.90 4 [73BOT/VIC2] 25 (5.70 ± 0.02) (4.36 ± pot 0.1 KNO (2.81 ± 0.08) [74FIE/MCC] 0.03) 3 ) 25 (5.33 ± 0.05) (4.13 ± 0.03) (2.83 ± pot 1 (KNO [74VAN/GEN] 0.04) 3 25 (5.646 0.024) (4.387 ± 0.006) (2.871 ± 0.003) [75MAT/HIR] pot 0.1 KCl ± ) 25 ± 0.01) (4.36 ± 0.01) (2.91 ± 0.01) [76HAR/MAR] pot 0.1 (KNO (5.74 3 (5.539 ± pot 0.15 Na(ClO ± 0.004) (2.869 ± 0.006) [78BER/MAY] ) 37 0.003) (4.236 4 25 5.05 4.05 2.71 [78KER/CHU] pot 1 NaClO 4 ± pot 0.12 (NaCl) 25 (13.8 ± 0.05) (4.32 ± 0.01) (2.94 ± 0.05) [78RAJ/MAI] 0.3) (5.70 pot 0.15 (KNO ) 37 (5.62 ± 0.02) (4.29 ± 0.04) (2.87 ± 0.06) [79AMI/DAN] 3 [79EKS/OLI] ) 25 pot 1 Na(ClO ± 0.002) (4.097 ± 0.003) (2.811 ± 0.003) (5.171 4 pot 0.1 (Me NBr) 25 5.78 4.32 2.89 [79HEU/POP] 4 pot 0.14 KCl 37 (5.652 ± 0.005) (4.302 ± 0.007) (2.88 ± 0.01) [80DAN/RIG] 0.3 ± 0.005) (4.190 ± 0.007) (2.78 ± 0.01) (5.472 (5.292 0.007) (4.098 ± 0.008) (2.84 ± 0.01) 1 ± ± ± (4.37 ± 0.01) (2.94 (5.72 0.01) [80HED/LID] pot 0.1 (KCl) 25 0.02) ± 0.3) (5.70 ± 0.06) (4.32 ± 0.01) (2.92 ± 0.06) [81RAJ/MAI] pot 0.12 (NaCl) 25 (13.0 0.02) ) 25 ± 0.02) (4.38 ± (5.69 (2.92 ± 0.03) [82AVD/KEA] pot 0.2 (KNO 3 (5.468 ± 0.002) (4.187 ± 0.003) (2.775 pot 0.15 (NaCl) 37 0.005) [82JAC] ± 0.02) (5.86 pot 0.1 (Et ± 0.02) (4.40 ± NI) 37 (2.92 ± 0.02) [83DAN/RIG] 4 0.3 (5.82 ± 0.02) (4.37 ± 0.02) (2.915 ± 0.002) 1 ± 0.01) (4.505 ± 0.008) (3.065 ± 0.015) (6.07 (5.217 [83OHM/SJO] 0.005) (4.081 ± 0.004) (2.769 ± 0.003) pot 0.6 Na(Cl) 25 , ± [83OHM/SJO2] NBr) 25 (5.82 ± 0.01) (4.38 pot 0.25 (Et 0.02) (2.95 ± 0.04) [84DAN/OST] ± 4 pot 2 Na(ClO ) 25 (5.02 ± 0.02) (4.03 ± 0.03) (2.75 ± 0.03) [84GRE/WIK] 4 pot 0.1 (KNO ) 25 (5.710 ± 0.002) (4.389 0.002) (3.025 ± 0.002) [84MOT/MAR] ± 3 pot 1 (Na)Cl 25 (5.12 ± 0.01) (4.05 ± 0.01) (2.77 ± 0.01) [86CRU/WAT] [86GRE/POW] pot 0.1 (KCl) 25 ± 0.02) (4.35 ± 0.01) (2.91 ± 0.02) (5.70 (Continued on next page)

354 VII Discussion of data selection for citrate compounds and compl exes 312 Table VII-2: (continued) Method I Reference log K K log log K Electrolyte t log K 2 10 3 1 10 0 10 10 (M) (°C) 25 (5.32 ± 0.01) (4.16 ± 0.01) (2.81 ± pot 0.5 NaNO 0.01) , [89RIB/SAL] 3 [89RIB/SAL2] ± ± 0.001) (2.856 ± 0.002) [89VEN/BER] pot 0.15 NaCl 37 (5.539 0.001) (4.248 ) 25 (5.69 ± 0.02) (4.35 pot 0.1 (KNO ± (2.90 ± 0.07) [90ARE/CON] 0.04) 3 b 2.879 [90DAN/ROB] 5.545 , 4.267 pot 0.16 (NaCl) 25 [90DAN/ROB2] 4.162 2.824 0.36 5.315 4.090 0.64 5.169 2.783 4.017 1 5.093 2.739 0.16 (KCl) 25 5.595 4.297 2.900 4.219 2.857 0.36 5.433 0.64 5.349 4.167 2.804 2.723 4.097 1 5.311 2.889 0.16 (LiCl) 25 5.487 4.247 2.846 4.127 0.36 5.210 0.64 5.033 4.047 2.824 2.806 1 4.963 3.981 b pot 0.16 Et ± 0.015) (4.396 ± 0.01) (2.953 ± 0.011) [90ROB/STE] (5.848 N(I) 25 4 0.36 (5.858 0.015) (4.396 ± 0.01) (2.973 ± 0.011) ± ± (3.029 ± 0.01) 0.64 (5.958 ± 0.011) 0.015) (4.454 1 (6.099 0.015) (4.536 ± 0.01) (3.106 ± 0.011) ± pot 0.15 (NaCl) 25 (5.540 ± 0.005) (4.240 ± 0.008) (2.867 ± 0.010) [91JAC/TOI] pot 0.1 (NaClO 5.70 [92GLA/HUL] ) 25 4 c 5.71 pot 0.15 (NaCl) 25 [93GLA/MAJ] 4.38 2.92 [95KIS/BUG] , 0.03) ± (2.87 ± (4.27 0.02) ± pot 0.2 (KCl) 25 (5.57 0.02) [96KIS/JEZ] ± ) 25 (5.317 ± 0.007) (4.147 ± 0.010) (2.832 pot 0.5 (NaClO 0.011) [95PII/LAJ] 4 d , [96CHO/ERT] 25 pot ± 0.01) (4.42 ± 0.01) (2.99 ± 0.01) (5.70 NaClO 0.1 4 [2001CHO/BON] 0.01) 0.3 ± 0.01) ± (5.41 (2.72 ± 0.01) (4.19 0.5 (5.33 ± 0.01) (4.26 ± 0.01) (2.92 ± 0.02) 1 ± 0.01) (4.24 ± 0.01) (2.95 ± 0.01) (5.25 3.5 0.02 ± 0.01) (4.29 ± (2.96 ± 0.01) (5.22 0.01) ± 0.01) (4.63 ± (5.56 (3.27 ± 0.01) 6.5 0.1 NaCl 25 (5.63 ± 0.01) (4.36 ± 0.01) (2.93 ± 0.01) 0.01) ± 0.01) (4.25 ± (5.38 (2.88 ± 0.02) 0.3 0.5 (5.27 ± 0.01) (4.23 ± 0.01) (2.90 ± 0.02) (Continued on next page)

355 VII.3 Protonation constants for citrate 313 Table VII-2: (continued) Method I Electrolyte t Reference K log log K K log log K 0 10 10 2 10 3 1 10 (°C) (M) d , [96CHO/ERT] (5.20 0.02) 0.01) (2.88 ± 0.05) ± ± NaCl 25 pot 1 (4.18 ] [2001CHO/BON (5.12 ± 0.01) (4.19 ± 0.01) (2.87 ± 0.01) 2 (5.17 3 0.01) (4.28 ± 0.01) (2.98 ± 0.03) ± (3.13 5 0.01) (4.49 ± 0.01) ± ± 0.02) (5.35 NCl 25 5.84 4.40 2.94 pot 0.1 Me [96XUE/TRA] 4 e 0.1 0.02) (4.280 ± 0.015) (2.901 ± 0.002) [97BEN/PAL] (NaCl) 25 ± pot (5.643 (5.373 ± 0.02) (4.176 0.3 0.022) (2.822 ± 0.002) ± 0.024) (2.752 0.6 0.02) (4.081 ± ± ± 0.006) (5.251 1.0 (5.132 ± 0.02) (4.066 0.024) (2.742 ± 0.008) ± pot Me 0.300 N(Cl) 25 (5.741 ± 0.01) [97FOT/GIA] 4 (5.737 ± 0.01) 0.301 0.523 0.01) ± (5.678 (5.679 ± 0.01) 1.014 1.014 (5.674 ± 0.01) 1.475 ± 0.01) (5.687 1.479 (5.717 0.01) ± (5.73 0.01) 2.139 ± (5.76 ± 0.01) 2.141 2.982 (5.803 ± 0.01) 2.988 (5.856 ± 0.01) (4.314 ± 0.01) 0.271 0.271 (4.312 ± 0.01) 0.506 (4.265 ± 0.01) 1.02 (4.251 ± 0.01) 1.019 (4.249 ± 0.01) 1.509 (4.254 ± 0.01) 1.512 (4.276 ± 0.01) 2.214 (4.294 ± 0.01) 2.215 (4.313 ± 0.01) (4.352 ± 0.01) 3.114 3.113 (4.388 ± 0.01) 0.259 ± 0.01) (2.894 0.259 (2.892 ± 0.01) 0.505 (2.876 ± 0.01) ± (2.885 1.051 0.01) (Continued on next page)

356 VII Discussion of data selection for citrate compounds and compl exes 314 Table VII-2: (continued) Method Electrolyte t Reference K log log K K log log I K 10 1 10 10 10 2 0 3 (°C) (M) Me (2.888 N(Cl) 25 pot ± 1.051 [97FOT/GIA] 0.01) 4 (2.928 ± 0.01) 1.571 (2.926 ± 0.01) 1.576 (2.987 ± 2.325 0.01) 2.328 (2.991 ± 0.01) 3.291 3.069 ± 0.01) 3.298 3.082 ± 0.01) Na(Cl) 25 (5.464 0.04) 0.008) (4.25 ± 0.01) (2.92 ± 0.3 [99MIZ/BON] pot ± (2.87 (5.179 0.005) (4.11 ± 0.02) ± ± 0.02) 1 2 (5.076 ± 0.006) (4.09 ± 0.01) (2.84 ± 0.04) 3 ± 0.006) (4.116 ± 0.002) (2.79 ± 0.01) (5.084 4 (5.20 ± 0.04) (4.24 ± 0.02) (2.93 ± 0.06) (2.96 5 ± 0.04) (4.32 ± 0.03) (5.22 ± 0.07) f Na(Cl) 25 (5.412 ± 0.003) (4.137 ± 0.003) (2.817 ± 0.005) [99ROB/STE] pot 0.5 1 (5.275 ± 0.003) (4.079 ± 0.003) (2.794 ± 0.005) 2 ± 0.003) (4.093 ± 0.003) (2.818 ± 0.005) (5.190 0.003) (2.880 3 ± 0.003) (4.167 ± (5.189 ± 0.005) 5 (5.287 ± 0.003) (4.397 ± 0.003) 3.065 ± 0.005) 0.5 (5.439 ± 0.003) (4.151 ± 0.003) (2.822 ± 0.005) K(Cl) 25 1 ± (5.327 ± 0.003) (4.106 0.003) (2.803 ± 0.005) ± (5.288 ± 0.003) (4.142 0.003) (2.833 ± 0.005) 2 3 (5.324 ± 0.003) (4.233 ± 0.003) (2.898 ± 0.005) 4.5 (5.432 ± 0.003) (4.415 ± 0.003) (3.030 ± 0.005) 0.007) (2.799 pot 1 NaClO (5.162 ± 0.003) (4.089 ± 25 ± 0.012) [2001CIA/TOM2] 4 pot 0.2 (KCl) 25 (5.57 ± 0.02) (4.27 ± 0.02) (2.87 ± 0.03) [2001LAK/BAN] a: Large citrate concentrations were used in [40ADE] I ≥ 1 M are considered in this re- , and only data at view. b: Protonation constants were also determined at other temperatures (10, 20, 30, 40 and 50°C) in [90DAN/ROB2] and [90ROB/STE] . . c: Results at 37°C were also reported in [93GLA/MAJ] d: Protonation constants were also determined at higher ionic strengths in [96CHO/ERT] . e: Protonation constants were also determined at other temperatures (between 5 and 150°C) in [97BEN/PAL] . f: Large citrate concentrations were used in [99ROB/STE] , and only data at I ≥ 0.5 M are considered in this review.

357 VII.3 Protonation constants for citrate 315 K Analysis of VII.3.2 3 The data in Table VII-2 for the third protonation of citrate, − + H U H + H cit cit(aq) (VII.5) 2 3 were treated with the SIT methodology described in Appendix B, using a weighted multi-linear least-squares regression procedure. This procedure assumes that a common ο * value of log K should fit all the data. For this reaction ∆ε (VII.5) 3 10 3 − + −+ = , where X and M ε−ε are the anion and cation, respec- (H cit , M ) (H cit, MX) 32 2 ∆ tively, of the background electrolyte, and − z 2. The regression plots are shown in = Figure VII-5, and the results are reported in Table VII-3. − Figure VII-5: Multi-linear least-squares SIT regression plots for the reaction: H + cit 2 + H cit(aq). Data from Table VII-2 were extrapolated to 25°C and converted to U H 3 molal units in the plots. 4.0 [68GUI] [72MET/GUI3] – + − + H cit cit(aq) + H H U 2 3 m [90DAN/ROB], z H cit(aq) H + H cit I + ) 3 2 − electrolytes in Li [90DAN/ROB2] 3.5 + ,X + in Li electrolytes (H ε − 3.0 D + 2 3 K 10 2.5 log 2.0 01234 + [Li ] / molal 4.0 [89RIB/SAL], [40ADE] [89RIB/SAL2] [62FUR/CER] + – − [64CAM/OST] + [89VEN/BER] H cit + H H cit(aq) U 2 3 m H cit cit(aq) + H H z I [64TIM] [90DAN/ROB], ) + 2 3 − electrolytes in Na 3.5 [73BOT/VIC], [90DAN/ROB2] + ,X + in Na electrolytes [91JAC/TOI] [73BOT/VIC2] (H [78BER/MAY] [93GLA/MAJ] ε [95PII/LAJ] [78KER/CHU] − 3.0 ) [78RAJ/MAI] [96CHO/ERT] (NaClO D 4 [79EKS/OLI] [96CHO/ERT] (NaCl), + 2 3 [81RAJ/MAI] [2001CHO/BON] K [82JAC] [97BEN/PAL] 10 [83OHM/SJO], [99MIZ/BON] 2.5 log [83OHM/SJO2] [99ROB/STE] [84GRE/WIK] [2001CIA/TOM2] [86CRU/WAT] 2.0 0123456 + ] / molal [Na (Continued on next page)

358 exes VII Discussion of data selection for citrate compounds and compl 316 Figure VII-5 (Continued) 4.0 [82AVD/KEA] [59OKA/KOL] m I − + ) [65RAJ/MAR] [84MOT/MAR] − + H cit H U H cit(aq) 3 2 [72KAN] [86GRE/POW] ,X + + in K electrolytes [74FIE/MCC] [90ARE/CON] 3.5 (H ε [90DAN/ROB], [74VAN/GEN] − [90DAN/ROB2] [75MAT/HIR] ) m [76HAR/MAR] [95KIS/BUG], I ( 3.0 [96KIS/JEZ] [79AMI/DAN] D [80DAN/RIG] [99ROB/STE] + 2 [2001LAK/BAN] [80HED/LID] 3 K 2.5 10 log 2.0 012345 + [K ] / molal 4.0 m I ) + − calc. Me N 4 ,X + [79HEU/POP] 3.5 [96XUE/TRA] (H ε [97FOT/GIA] − + ) calc. Et N m 4 I ( 3.0 [83DAN/RIG] D [84DAN/OST] [90ROB/STE] + 2 3 + − K H + H cit U H cit(aq) 2.5 10 3 2 + log in R N electrolytes 4 2.0 0123 + ] / molal N [R 4 * ο Table VII-3: Selected values of (VII.5) and log at 25°C. ε K ∆ 3 3 10 ο ± log K (VII.5) 0.01) = (3.13 3 10 a * 1 − ∆ ) (VII.5) / (kg · mol ε Medium 3 + (0.11 ± 0.03) Li + (0.05 ± 0.01) Na + 0.01) (0.04 ± K + + ± N − (0.05 N 0.02) + (0.07 0.04) log [Me ] ± Me 4 10 4 + + 0.02) + (0.2 ] − (0.29 ± N ± 0.1) log N [Et Et 10 4 4 + + : Me represent tetramethyl- and tetraethylammonium, respectively a N and Et N 4 4

359 VII.3 Protonation constants for citrate 317 VII.3.3 Analysis of K 2 The data in Table VII-2 for the second protonation of citrate: 2 − + − Hcit cit (VII.6) + H U H 2 were treated with the SIT methodology described in Appendix B, using a weighted multi-linear least-squares regression procedure. This procedure assumes that a common ο * value of K ε (VII.6) = log should fit all the data. For this reaction ∆ 10 2 2 + −+ 2 −+ , where M (H cit , M ) is the cation of the background electrolyte, (Hcit , M ) ε−ε 2 2 ∆ and z = − 4. The regression plots are shown in Figure VII-6, and the results, with the values selected in this review, are reported in Table VII-4. 2 − Figure VII-6: Multi-linear least-squares SIT regression plots for the reaction: Hcit + + − H . Data from Table VII-2 were extrapolated to 25°C and converted to molal H U cit 2 units in the plots. 6.0 [68GUI] 2– – + − − 2 + [72MET/GUI3] + H cit Hcit H U 2 z cit Hcit + H H m 2 [90DAN/ROB], + I ) electrolytes in Li + − [90DAN/ROB2] 5.5 in Li electrolytes ,X + (H ε − 5.0 D + 4 2 K 10 4.5 log 4.0 01234 + ] / molal [Li 6.0 [86CRU/WAT] [40ADE] [89RIB/SAL], [62FUR/CER] 2– + – 2 − + − U cit H + H Hcit 2 [64CAM/OST] [89RIB/SAL2] + H H Hcit z cit m + 2 I [89VEN/BER] [64TIM] electrolytes in Na ) + − 5.5 [90DAN/ROB], [67FIS2] in Na electrolytes ,X + [73BOT/VIC], [90DAN/ROB2] (H [73BOT/VIC2] [91JAC/TOI] ε [78BER/MAY] [93GLA/MAJ] − 5.0 [78KER/CHU] [95PII/LAJ] D ) [96CHO/ERT] (NaClO [78RAJ/MAI] 4 + 4 2 [79EKS/OLI] [96CHO/ERT] (NaCl), K [81RAJ/MAI] [2001CHO/BON] 10 [82JAC] [97BEN/PAL] 4.5 log [83OHM/SJO], [99MIZ/BON] [83OHM/SJO2] [99ROB/STE] [84GRE/WIK] [2001CIA/TOM2] 4.0 0123456 + ] / molal [Na (Continued on next page.)

360 VII Discussion of data selection for citrate compounds and compl exes 318 Figure VII-6 (Continued) 6.0 [82AVD/KEA] [59OKA/KOL] 2 − + − m [84MOT/MAR] [65RAJ/MAR] H Hcit U + H cit I 2 ) − [86GRE/POW] [67FIS2] + in K electrolytes ,X [90ARE/CON] [72KAN] 5.5 + [74FIE/MCC] [90DAN/ROB], (H ε [90DAN/ROB2] [74VAN/GEN] − [75MAT/HIR] [95KIS/BUG], 5.0 D [96KIS/JEZ] [76HAR/MAR] [99ROB/STE] [79AMI/DAN] + 4 2 [2001LAK/BAN] [80DAN/RIG] K 10 [80HED/LID] 4.5 log 4.0 012345 + ] / molal [K 6.0 + m I N calc. Me ) 4 − [79HEU/POP] ,X 5.5 + [96XUE/TRA] (H [97FOT/GIA] ε + calc. Et N − 4 5.0 D [83DAN/RIG] [84DAN/OST] + 4 2 [90ROB/STE] K + − 2 − 10 4.5 U cit Hcit H + H 2 log + in R N electrolytes 4 4.0 0123 + [R N ] / molal 4 ο * Table VII-4: Selected values of (VII.6) and log K ε ∆ at 25°C. 2 2 10 ο (VII.6)= (4.78 log K 0.01) ± 10 2 a * − 1 ∆ ε (VII.6) / (kg · mol ) Medium + Li (0.06 0.03) ± + Na ± 0.01) – (0.01 + K (0.03 ± 0.01) − + + Me N (0.16 ± − ] 0.02) + (0.21 N ± [Me 0.05) log 10 4 4 + + N 0.1) log ± 0.02) + (0.4 ± (0.46 − N ] [Et Et 4 10 4 + + N represent tetramethyl- and tetraethylammonium, respectively Me : and Et N a 4 4

361 VII.3 Protonation constants for citrate 319 Analysis of K VII.3.4 1 The values listed in Table VII-2 for the first protonation of citrate: 3 + − − 2 cit Hcit (VII.7) U + H were plotted using the SIT methodology described in Appendix B. For the first protona- + * +3 − − +2 tion constant of citrate, (M , cit ) ∆ε ε−ε , where M (M , Hcit ) is the cation of = 1 2 ∆ the background electrolyte, and 6. Regressions of the data for individual back- − z = + + + ground cations (Li , and tetraalkylammonium ions) resulted in slightly differ- , K , Na ο ent values for log K . Table VII-5. , cf 1 10 perimental data obtained with each of the Table VII-5: Results from regressions of ex ° C. background cations at 25 ο * 1 − Background cation ∆ = 0 value obtained at K log ε / (kg ⋅ mol I ) m 1 10 1 + Li ± (6.22 (0.16 ± 0.07) 0.11) + (6.28 ± 0.02) (0.02 ± 0.01) Na + K ± 0.02) − (0.01 (6.37 0.02) ± + + + Me and Et ] N N (6.41 ± 0.06) − (0.37 ± 0.05) + (0.43 ± 0.08) log N [Me 4 10 4 4 + N 0.2) log − [Et (0.77 ± ± ] 0.05) + (0.5 4 10 These would be the expected results if alkali-cation complexes were formed with citrate, . Section V.3 and Section VII.4. The values extrapolated to cf = 0 in the I m + different ionic media indicated that weak complex formation perhaps takes place in Li - + media, and to a lesser extent in Na electrolytes. Further data on the complex formation between citrate and alkali cations is discussed in section VII.4. + The results in potassium and R N electrolytes agreed within the uncertainties 4 ο and they indicated that the value of K should be between 6.35 and 6.47. A log 1 10 + + weighted multi-linear least-squares regression of the data in K N and R media gave as 4 ο a result ± log K 0.01). = (6.36 1 10 A similar weighted multi-linear least-squares regression but with the whole ο data set, assuming a common value of log K for all ionic media, could be performed 1 10 + + * successfully by assuming that was not independent of ionic strength in Li ∆ε , Na as 1 + well as R N The weights of the data in lithium and sodium media were electrolytes . 4 decreased by a factor or 4 (uncertainties multiplied by 4 ) because of the larger num- + + + + ber of data in Li and Na and R N media (62 values) as compared with those in K 4 tance to data where complex formation with salts (50 values). This also gave less impor background cations could introduce an unknown bias. This modified weighting is indi- cated in the error bars of Figure VII-7 that shows a comparison of the data with the re- sults of the regression. Table VII-6 reports the values selected in this review, where the ο limits in log K have been increased to reflect the uncertainties in this system. 1 10

362 VII Discussion of data selection for citrate compounds and compl exes 320 ο * ∆ (VII.7) and K ε at 25°C. Table VII-6: Selected values of log 10 1 1 ο K = (6.36 ± 0.02) log 10 1 a * − 1 ∆ /(kg · mol ) Medium ε 1 + + [Li (0.38 ± 0.2) log Li 0.10) ± ] − (0.3 10 + + 0.02) Na (0.13 ± 0.03) log − [Na (0.11 ] ± 10 + K (0.03 ± 0.01) − + + Me − (0.43 ± 0.02) + (0.52 ± 0.05) log N [Me ] N 4 4 10 + + Et − (0.82 ± 0.02) + (0.8 ± N 0.1) log [Et ] N 4 10 4 + + represent tetramethyl- and tetraethylammoium, respectively N a and Et : N Me 4 4 Figure VII-7: Equilibrium constants for Reaction (VII.7). Data from Table VII-2 were lal units and plotted according to the SIT extrapolated to 25°C and converted to mo methodology (see text for details). 7.5 [68GUI] 2– + 3– U Hcit + H cit [72MET/GUI3] 3 + 2 − − + m [90DAN/ROB], Hcit cit z + H electrolytes in Li I ) − [90DAN/ROB2] 7.0 + ,X electrolytes in Li + (H ε − 6.5 D + 6 1 K 10 6.0 log 5.5 01234 + ] / molal [Li 7.5 [40ADE] [89RIB/SAL], [89RIB/SAL2] [62FUR/CER] 3– 2– + − 3 + 2 − [89VEN/BER] [64CAM/OST] U Hcit + H cit Hcit z cit + H m I [64TIM] [90DAN/ROB], + ) − electrolytes in Na + 7.0 [90DAN/ROB2] [73BOT/VIC], electrolytes in Na ,X + [73BOT/VIC2] [91JAC/TOI] (H [78BER/MAY] [92GLA/HUL] ε [93GLA/MAJ] [78KER/CHU] − 6.5 [78RAJ/MAI] [95PII/LAJ] D [79EKS/OLI] [96CHO/ERT] (NaClO ) 4 + 6 1 [81RAJ/MAI] [96CHO/ERT] (NaCl), K [82JAC] [2001CHO/BON] 10 [83OHM/SJO], [97BEN/PAL] 6.0 log [83OHM/SJO2] [99MIZ/BON] [84GRE/WIK] [99ROB/STE] [86CRU/WAT] [2001CIA/TOM2] 5.5 0123456 + [Na ] / molal (Continued on next page)

363 VII.3 Protonation constants for citrate 321 Figure VII-7 (continued) 7.5 [59OKA/KOL] [82AVD/KEA] + − 3 − 2 m [84MOT/MAR] [65RAJ/MAR] cit Hcit + H U I ) − [86GRE/POW] [72KAN] + in K electrolytes ,X [74FIE/MCC] [90ARE/CON] 7.0 + [90DAN/ROB], [74VAN/GEN] (H ε [75MAT/HIR] [90DAN/ROB2] − [95KIS/BUG], [76HAR/MAR] 6.5 D [79AMI/DAN] [96KIS/JEZ] [80DAN/RIG] [99ROB/STE] + 6 1 [2001LAK/BAN] [80HED/LID] K 10 6.0 log 5.5 012345 + ] / molal [K 7.5 + m I calc. Me N ) 4 − [79HEU/POP] ,X 7.0 + [96XUE/TRA] (H [97FOT/GIA] ε + calc. Et N − 4 6.5 [83DAN/RIG] D [84DAN/OST] + 6 1 [90ROB/STE] 2 3 − + − K + H cit Hcit U 10 6.0 + in R N electrolytes log 4 5.5 0123 + N [R ] / molal 4 VII.3.5 Analysis of K 0 The protonation of the hydroxy-group, O O O O + O O O O H + O O O OH O O may be schematically written as: − 3 4 − + H U cit . cit + H 1 −

364 VII Discussion of data selection for citrate compounds and compl exes 322 The equilibrium constant for this reaction, , has been reported in a few pa- K 0 + pers. Most of these studies have been made in Na media (Table VII-7). In contradiction with these determinations Bottari and Vicedomini found no evidence for this dissocia- − tion in [OH media ] ≤ 0.2 M using a hydrogen electrode in 2 M NaClO 4 [73BOT/VIC2] . 4 − Table VII-7: Literature data on the protonation constant for H cit included in the 1 − review process. Reference t (°C) log Method K I Electrolyte 10 0 pot 0.1 KCl 19 10.82 [59OKA/KOL] pot 2 25 >13 [73BOT/VIC2] , [73BOT/VIC] NaClO 4 [78RAJ/MAI] 25 pot 0.12 NaCl 0.3) ± (13.8 [81RAJ/MAI] 25 pot 0.12 NaCl 0.3) ± (13.0 The experimental determination of is associated with special difficulties, K 0 + mainly due to the high alkalinity of the solutions ( log [H ] 12). Under these con- ≤ − 10 ditions the response of the glass electrode may be affected by the sodium effect and by changes in the junction potentials. The solutions might also be unnoticeably contami- nated by atmospheric CO . Substantial changes in the composition of the ionic media 2 also take place for experiments conducted at I 1 1, when a substantial amount of sodium hydroxide replaces the background electrolyte. Owing to all these factors, no value for ο K may be recommended, although it is exp ected to be larger than that for the log 10 0 4– − ο protonation of HO > 14). Because of this H ( , log can not be used as a cit i.e. K –1 10 0 [58HEI/FRI] component when calculating equilibrium constants, as done in . The equi- librium constant, from [57PAT/PAN2] and [65PAT/PAN2] for the reaction: 2– + – Ni(H 7.87 U Ni(cit) + H log cit) K ≈ 10 –1 indicates a very large inductive effect on the dissociation constant of the OH-group in 3– cit upon coordination to Ni(II); the dissociation constant increases more than six or- ders of magnitude. The complex formation between Ni(II) and citrate is discussed in detail in Section VII.7. Selected protonat ion constants for citrate VII.3.6 Summarising the results from sections VII.3.2 to VII.3.5, the selected standard values for the protonation constants of citrate are: ο ± log K 0.02), ((VII.7), 298.15 K) = (6.36 1 10 ο ± log K 0.01), ((VII.6), 298.15 K) = (4.78 2 10 ο ((VII.5), 298.15 K) = (3.13 0.01). K ± log 3 10

365 VII.3 Protonation constants for citrate 323 These values correspond to the following global standard protonation constants: + 3 − 2 − ο * cit 0.02), + H Hcit log U b (298.15 K) = (6.36 ± 10 1 3 − + − * ο cit U H ± cit + 2 0.02), H b log (298.15 K) = (11.14 2 2 10 3 + − * ο cit U H + 3 cit(aq) (298.15 K) = (14.27 0.02). H log b ± 3 3 10 In a few studies the protonation of citrate has been investigated in solutions of weak ionic strengths, and the authors used Debye-Hückel expressions for activity coef- ο ficients to obtain values for [29BJE/UNM] K , log , [28SIM] , [28KOL/BOS] r 10 [49BAT/PIN] , [51HEI] , [69LIT/PUR] , [78FLY/KOR] , [78USS/BOS] , [89PAP/ZIO] , [89YAD/GHO] , , [96SAE/KHA] . The data from [51HEI] , [96SAE/KHA] [91APE/BAR] ο are discrepant and were not considered. The averages of the remaining data ( K = log 10 r ± 0.09), (4.77 ± 0.08) and (6.38 ± 0.11) for (3.12 r = 3 to 1, respectively) agree well with the values selected by this review. Protonation constants for citrate for different ionic media may be calculated us- ing the SIT model described in Appendix B, and the selected standard protonation con- stants and specific ion interaction coefficients. In Table VII-8 the results from such cal- culations are reported for some ionic media commonly used in chemical equilibrium onated forms of citrate as a function of pH studies. The distribution of the different prot in 1 M NaCl is shown in Figure VII-8. Figure VII-8: Calculated distribution of citrat e species as a function of pH in 1 M NaCl at 25°C. The equilibrium constants listed in Table VII-8 were used to draw the curves. 3− cit 1.0 cit H 3 0.8 − cit H 2 2− Hcit 0.6 0.4 Fraction 0.2 0.0 234567 pH

366 VII Discussion of data selection for citrate compounds and compl exes 324 3 − + + Table VII-8: Calculated equilibrium constants / Na for some cit / / H in Molar units + K systems at 25°C. The SIT model for activity coefficients has been used with the ∆ε n obtained in this review. Other parameters from Appendix B were used as appropriate. Care should be exercised when using values at highest ionic strengths, because they are confirmed by few experimental data only, as indicated in the figures of Sections VII.3.2 to VII.3.4. NaClO 4 I (M) (molal) log K K log I K log 10 2 1 10 10 3 m ± 0.02) (4.78 ± 0.01) (3.13 ± 0 0.000 (6.36 0.01) 0.1 0.101 (5.70 0.01) (4.36 ± 0.01) (2.92 ± 0.01) ± 0.25 0.254 (5.48 ± 0.01) (2.87 ± 0.01) 0.01) (4.24 ± 0.5 0.513 (5.31 ± ± 0.02) (2.84 ± 0.02) 0.02) (4.17 ± 0.03) (4.14 ± 0.02) (2.83 ± 0.02) 0.75 0.779 (5.23 1 1.05 (5.18 0.03) (4.14 ± 0.03) (2.84 ± 0.03) ± 2 2.21 (5.16 ± ± 0.05) (2.90 ± 0.05) 0.07) (4.22 0.11) (4.37 ± 0.08) (3.01 ± 0.08) ± 3 3.50 (5.28 4 4.95 (5.48 ± 0.18) (4.57 0.11) (3.15 ± 0.11) ± 5 6.58 (5.76 ± ± 0.15) (3.30 ± 0.15) 0.25) (4.81 NaNO 3 (M) I K (molal) log log K log I K 10 2 10 10 m 3 1 ± ± ± 0.01) (3.13 0 0.000 (6.36 0.01) 0.02) (4.78 0.1 0.101 (5.69 0.02) (4.35 ± 0.01) (2.92 ± 0.01) ± 0.25 0.253 (5.46 ± 0.02) (4.22 ± 0.01) (2.85 ± 0.01) ± ± 0.01) (2.80 ± 0.01) 0.02) (4.13 0.5 0.509 (5.28 0.75 0.769 (5.17 ± ± 0.01) (2.77 ± 0.01) 0.03) (4.08 1 1.03 (5.11 0.03) (4.06 ± 0.02) (2.76 ± 0.02) ± 0.03) ± 0.06) (4.05 ± 0.03) (2.74 ± 2 2.14 (5.00 3 3.33 (5.01 ± 0.09) (4.10 ± 0.05) (2.74 ± 0.05) 4 4.61 (5.08 ± 0.14) (4.18 ± 0.07) (2.77 ± 0.07) ± 0.20) (4.27 ± 0.09) (2.80 ± 0.09) 5 6.02 (5.21 (Continued on next page)

367 VII.3 Protonation constants for citrate 325 Table VII-8 (continued) NaCl I I (molal) log (M) K K log log K 2 10 1 3 10 m 10 ± 0.01) ± 0.02) (4.78 0 0.000 (6.36 ± 0.01) (3.13 0.02) (4.36 0.01) (2.92 ± 0.01) ± 0.1 0.100 (5.69 ± ± 0.02) (4.23 ± 0.01) (2.86 ± 0.01) (5.47 0.25 0.252 0.5 0.506 (5.30 ± 0.01) (2.82 ± 0.01) 0.02) (4.15 ± 0.75 0.762 (5.21 ± ± 0.01) (2.81 ± 0.01) 0.03) (4.12 0.03) (4.10 ± 0.02) (2.80 ± 0.02) ± 1 1.02 (5.15 2 2.09 (5.09 ± 0.06) (4.14 ± 0.03) (2.83 ± 0.03) 3 3.20 (5.15 ± ± 0.05) (2.89 ± 0.05) 0.09) (4.24 0.13) (4.36 ± 0.06) (2.96 ± 0.06) ± 4 4.37 (5.26 5 5.61 (5.42 ± 0.08) (3.04 ± 0.08) 0.18) (4.50 ± KCl (M) log I K (molal) I K log K log 2 10 10 m 3 1 10 ± 0.02) (4.78 ± 0 0.000 (6.36 ± 0.01) 0.01) (3.13 0.1 0.101 (5.72 0.02) (4.36 ± 0.01) (2.92 ± 0.01) ± (5.53 0.25 0.252 0.02) (4.24 ± 0.01) (2.86 ± 0.01) ± 0.5 0.509 (5.39 ± 0.02) (4.16 ± 0.01) (2.83 ± 0.01) 0.01) ± ± 0.02) (4.13 ± 0.01) (2.82 (5.33 0.75 0.769 1 1.03 (5.30 ± ± 0.02) (2.82 ± 0.02) 0.02) (4.13 2 2.13 (5.31 0.04) ± 0.03) (2.86 ± 0.03) ± (4.20 0.05) (4.33 ± 0.05) (2.94 ± 0.05) ± 3 3.31 (5.41 4 4.58 (5.55 (4.49 ± ± ± 0.07) 0.07) 0.07) (3.04 KNO 3 I I (M) (molal) log K K log log K 10 1 10 10 3 m 2 ± 0.02) (4.78 ± 0.01) (3.13 0 0.000 (6.36 0.01) ± 0.1 0.101 (5.72 0.02) (4.36 ± ± ± 0.01) 0.01) (2.92 0.25 0.253 (5.51 ± 0.02) (4.23 ± 0.01) (2.85 ± 0.01) ± (4.14 ± 0.01) (2.80 ± 0.01) 0.02) 0.5 0.512 (5.37 (5.29 0.75 0.776 0.02) (4.10 ± 0.01) (2.78 ± 0.01) ± 1 1.05 (5.25 0.02) (4.08 ± ± ± 0.02) 0.02) (2.77 ± 0.04) (4.10 ± 0.03) (2.77 ± 0.03) 2 2.19 (5.21 3 3.44 (5.27 ± 0.05) (4.19 ± 0.05) (2.79 ± 0.05) It is expected that the value of (H cit, MX) ε will be small. In this review the 3 − 1 approximation is made that ε = (0.00 ± 0.01) kg ⋅ mol . Table VII-9 con- (H cit, MX) 3 tains selected specific ion interaction coefficients based on this approximation and on * the ∆ε listed in previous subsections. These ε -values agree within the uncertainties n with experimental water activities in Na cit solutions and in mixtures of H cit-NaCl: see 3 3 the discussion in Appendix A for [2004SCH/MAU] .

368 VII Discussion of data selection for citrate compounds and compl exes 326 − 1 for citrate and its mol Table VII-9: Selected specific ion interaction coefficients (kg ⋅ ) + + + protonated forms in Li , Na , K and tetramethylammonium electrolytes. + + + + K Li Me Na N 4 +− 3 (0.15 (0.55 0.11) ± − 0.03) ± ± − 0.04) (0.64 (M ,cit ) ε (0.02 ± 0.02) + (0.13 + (0.3 I I I ± ± 0.03) log − (0.80 ± 0.08) log 0.2) log 10 m 10 10 m m +− 2 ± 0.03) (0.21 ± 0.04) (M , Hcit ) (0.04 ± 0.02) − (0.01 ± 0.02) − ε − (0.17 ± 0.07)) log I (0.28 − 10 m +− ± 0.02) (0.05 0.01) 0.03) (M , H cit ) (0.05 ± ε − (0.04 ± 0.01) − (0.11 ± − 2 ± 0.04) log I − (0.07) m 10 (0.00 ± 0.01) (0.00 ± 0.01) (0.00 ± 0.01) (0.00 ± 0.01) (H cit, MX) ε 3 VII.3.7 Temperature effects Temperature effects on the equilibrium constants for the protonation of citrate have been determined both calorimetrically and by the determination of the protonation constants at different temperatures. The data found in the literature are listed in Table VII-10. In [29RIC/MAI] addition some heats of dilution and neutralisation were presented in . Table VII-10: Literature data on the enthalpy changes for citrate protonation, with as- signed uncertainties . a b Medium t ((VII.8),) ∆ H Reference Method rm –1 (kJ·mol ) (°C) r =1 r =3 =2 r c c c [29BJE/UNM] 18 - 37 0 NaCl ∂ (2.51 ± 2.5) p − (2.18 ± 2.5) K − (4.27 ± 2.5) / ∂ T → a c c c [49BAT/PIN] 0 - 50 0 KCl K (3.36 ± 0.64) ∂ − p ± 0.64) / − (4.17 ± 0.64) ∂ T → (2.44 a d,e [61BAR/BEC] 23.6 3.3 cal cit − 5.0 − 1 M (H,Na) ≈ 2.9 3 d,e 25 [80ARE/CAL] cal − 0.13) (2.01 ± 0.13) 0 NaClO (3.97 ± 0.13) → (3.35 ± − 4 0.05 M (2.22 − (3.01 ± 0.13) − (4.44 ± 0.13) ± 0.13) 0.10 − (3.14 ± 0.13) − ± ± 0.13) (4.52 0.13) (1.92 0.15 0.13) − (3.26 ± 0.13) − (4.73 ± 0.13) ± (1.72 d,e [80DAN/RIG] 37 ∂ T → and 0 (KNO ∂ p / K ± 0.21) (3.14 a 3 Et NBr) 4 d,e + + 10 - 45 /Et N [84DAN/OST] 0.25 M K (0.5 T ∂ p ± 0.8) − (4.3 ± 1.9) − (4.6 ± K 3.1) / ∂ 4 a d,f 0.1 M KNO [86CAP/ROB] 10 - 45 − / ∂ T ± 0.7) − (5 ± 0.9) K (4 ± 1.6) ∂ p (3 3 a e c c c [86CRU/WAT] 1 M (Na)Cl 25 cal (4.45 ± 2) − − (1.32 ± 2) ± 2) (4.29 − c 10 - 50 [89YAD/GHO] − p / ∂ ∂ (4.23 ± 1.2) T → 0 HCl K a c c c , 10 - 50 [90DAN/ROB] 0 Na(Cl) / (1.6 ± 0.8) ∂ − (4.1 ± 0.8) K − (5.4 ± 0.8) p → T ∂ a e c c c [90DAN/ROB2] 0.16 M 0.8) − − ± 0.8) 0.8) (5.1 ± ± (5.6 (0.4 c c c 0.36 − − (6 ± 0.8) (0.6 − (6.4 ± 0.8) ± 0.8) c c c 0.64 0.8) ± 0.8) 0.8) − (6.9 ± (6.2 ± (1.2 − − c c c 1 0.8) − (5 ± 0.8) (1.2 − (6.9 ± − 0.8) ± e c c c 10 - 50 [90ROB/STE] − p N(I) / (3 ± 0.8) ∂ K (3.1 ± 0.8) T − (5.0 ± 0.8) → ∂ 0 Et 4 a c c c 0.16 M − ± 0.8) ± (6.1 ± 0.8) − (4.9 0.8) (0.5 c c c 0.36 ± − (5.7 ± 0.8) − − (6.8 0.8) 0.8) ± (0.5 c c c 0.64 − − (6.6 ± 0.8) (1.5 − (7.7 ± 0.8) ± 0.8) c c c 1 0.8) − (7.7 ± − (2.6 − (8.8 ± 0.8) ± 0.8) (Continued on next page)

369 VII.3 Protonation constants for citrate 327 Table VII-10 (continued) b a t Method Medium H ∆ ((VII.8),) Reference rm –1 (kJ·mol ) (°C) c c 5 - 35 [91APE/BAR] 0 (Li,Na,K) ∂ cit / T − (1.88 ± 1.1) ∂ − (3.47 ± K p → 1.1) a 3 d,e ?? 25 - 45 [91BAP] 38.7 / − ∂ − p − 33.3 ∂ T K 44.1 a d,g [96SAE/KHA] 25 - 50 ∂ 68.7 → 0 (H,Na) T cit / − 24.8 − 31.5 − ∂ K p a 3 h c c c m (NaCl) 5 - 150 0.1 [97BEN/PAL] (1.84 K − (1.10 ± 1.0) ∂ / − (5.53 ± 1.0) T ± p 1.0) ∂ a c c c 0.3 (1.28 (3.32 ± 1.0) ± − (5.24 ± 1.0) 1.0) − c c c 0.6 − (4.03 1.0) 1.0) ± − (3.87 ± 1.0) (0.64 − ± c c c 1.0 − (4.90 ± 1.0) (1.07 − (4.93 ± 1.0) − ± 1.0) NaCl 25 1.03 m [2001STE/FOT] cal (0.94 0.6) ± − (4.60 (3.80 0.85) − 0.85) ± ± 1.55 (0.00 − (4.60 ± 0.85) − (5.49 ± 0.85) ± 0.6) 2.08 − − (5.42 ± − ± (5.77 ± 0.85) (0.81 0.6) 0.85) 3.17 − 0.6) − (7.04 ± 0.85) (3.00 (6.24 ± 0.85) − ± 5.41 − 0.6) − (10.21 ± 0.85) − (7.54 ± 0.85) (4.16 ± ∂ K / ∂ T = temperature dependence of protonation constants. p a: Methods: cal = calorimetry; a + are reported in molal units in the original publication. Et italics stands for tetraethylammo- N b: Values in 4 nium. c: Uncertainties assigned in this review. d: Reference not included in the review procedure. e: See comments in Appendix A. were too large (up to 0.01 M) to ensure a constant ionic f: The total citrate concentrations in [86CAP/ROB] medium. + , were g: “Mixed” protonation constants, defined using concentrations for citrate species but activity for H . [96SAE/KHA] obtained in h: The data from [97BEN/PAL] were recalculated in this review as described in Appendix A. Most of the calorimetric investigations on the protonation of citrate are associ- ated with several problems [61BAR/BEC] , [80ARE/CAL] , [86CRU/WAT] , ( cf. discus- sions in Appendix A) . For example, the ionic strength was not kept constant in [61BAR/BEC] , . Because of this, the study by [61BAR/BEC] is not con- [86CRU/WAT] sidered in this review, and a large uncertainty is assigned to the values determined in [86CRU/WAT] et al. [80ARE/CAL] . The study by Arena is rejected because apparently the enthalpy changes were corrected for al kali cation complexation. The most extensive [2001STE/FOT] calorimetric study is and in the other extreme contains [84REK/GAL] only qualitative information on the heat of neutralisation of pH-buffers. Reaction enthalpies were obtained in several studies by fitting the temperature variation of protonation constants to some model function. Some of these values were not included in this review because of different kinds of shortcomings: • protonation constants mixing concentrations and activities ( i.e. , “pH” does not + refer to – log H , ) : [96SAE/KHA] 10 equilibrium constants corrected for alkali cation complexation: [80DAN/RIG] • , • mixed ionic media: [84DAN/OST] , [90DAN/ROB] , [90DAN/ROB2] ,

370 VII Discussion of data selection for citrate compounds and compl exes 328 • citrate concentration high enough to affect substantially the nature of the ionic [86CAP/ROB] medium: , [91BAP] no details given or no original experimental data: • [92TAN/NOM] . , Most of these rejected references are discussed in Appendix A. An overview of the remaining literature values reveals that the reported enthalpy changes for the indi- vidual protonation steps: r ( r − 4) + ( 3) − H cit H + H cit (VII.8) U r − r ( 1) − 1 are small: − H ⋅ mol ∆ (VII.8) values range between , depending on the 11 and + 5 kJ rm ionic media, as expected for the dissocia tion of carboxylic groups. The corresponding heat capacity changes, (VII.8), are reported to be in the range 80 to C ∆ p r,m 1 − 1 − 260 J mol ⋅ . For temperatures between 0 and 50°C these values correspond to quite ⋅ K small changes of the equilibrium constants. Large uncertainties are expected in the determination of ∆ from the T - H rm variation of log K . For the studies considered in this review that used this methodol- 10 ogy [29BJE/UNM] [49BAT/PIN] , [89YAD/GHO] , , , [90DAN/ROB2] , [90DAN/ROB] [90ROB/STE] [91APE/BAR] [97BEN/PAL] the uncertainty was estimated from the , , rules of error propagation and the relationship: KT ( ) log ∂  2 10 HT Rln(10) ∆= rm  T ∂  P ± log K values, determined over the 0.02 in the for example, for an uncertainty of 10 temperature range 5 to 50°C, the total uncertainty in ∆ was estimated to be H rm − 1 1.1 kJ ⋅ . This method was used to estimate or to increase the uncertainties in re- ± mol ported H ∆ data obtained from the T -variation of K log . For this procedure the total 10 rm uncertainty in the individual measurements of K was set to at least ± 0.02 log - log 10 10 units. For studies where both H ∆ and -variation of T C ∆ were obtained from the p rm r,m log K the uncertainty assigned with the method outlined above was decreased by one 10 third. The selected H (VII.8) data and the corresponding uncertainties assigned ∆ rm by this review, listed in Table VII-10 were treated according to the SIT model ( cf. Sec- tion V.3.6), and a multi-linear least-squares regression was performed ( cf . Figure VII-9 to Figure VII-11). The standard enthalpy changes of reaction were found to be: 1 − ο ((VII.8), H = 1) = (3.3 ± 0.3) kJ ⋅ mol ∆ , r rm 1 − ο ⋅ = 2) = – (2.4 ± 0.3) kJ r mol ((VII.8), H , ∆ rm − 1 ο ∆ ((VII.8), r = 3) = – (4.5 ± 0.3) kJ ⋅ mol H . rm

371 VII.3 Protonation constants for citrate 329 + 2 − − 3 + H Figure VII-9: Enthalpy changes at 25°C for the reaction: cit U Hcit plotted according to the SIT methodology. 6 NaCl: [29BJE/UNM] − − + 3 2 − + − 3 2 + H Hci U t ci t [86CRU/WAT] + H z Hcit cit [90DAN/ROB], 1 [90DAN/ROB2] 4 − [97BEN/PAL] [2001STE/FOT] Et NI: 4 [90ROB/STE] / kJ·mol L 2 Other: D [49BAT/PIN] + 6 1 H r ∆ 0 −2 0123456 / molal I m 2 − + − Figure VII-10: Enthalpy changes at 25°C for the reaction: Hcit U plot- + H H cit 2 ted according to the SIT methodology. 2 NaCl: − 2 + − 2 + − − + H Hcit U H cit [29BJE/UNM] 2 cit H + H Hcit z [86CRU/WAT] 2 [90DAN/ROB, 0 1 90DAN/ROB2] − [97BEN/PAL] [2001STE/FOT] −2 Et NI: 4 [90ROB/STE] / kJ·mol L Other: D [49BAT/PIN] −4 + 4 [91APE/BAR] 2 H r ∆ −6 −8 0123456 I / molal m

372 exes VII Discussion of data selection for citrate compounds and compl 330 + − Figure VII-11: Enthalpy changes at 25°C for the reaction: H cit(aq) cit U H + H 3 2 plotted according to the SIT methodology. 0 NaCl: − + [29BJE/UNM] + − H z + H cit H cit(aq) t q) H ci ( + H cit U a H 2 3 [86CRU/WAT] 2 3 [90DAN/ROB, −2 1 90DAN/ROB2] − [97BEN/PAL] Et NI: 4 −4 [90ROB/STE] Other: / kJ·mol L [49BAT/PIN] D [89YAD/GHO] −6 + 2 [91APE/BAR] 3 H r ∆ −8 −10 0123456 / molal I m

373 VII.3 Protonation constants for citrate 331 From the slopes of the regressions it should be possible to obtain values for . The regressions of the protonation enthalpies for citrate in NaCl solutions gave: ∆ε L,r − 3 − 3 3 − ∆ε 10 = (0.9 ± 0.2) × 10 ∆ε × , and ∆ε , = (0.2 ± 0.2) × 10 0.2) ± ; and in = (0.6 2 L, 3 L, 1 L, 3 − 3 − Et 0.8) ± 10 = (2.5 , ∆ε = (3.4 ± × × 10 0.8) ∆ε , and ∆ε = NI media: 2 4 L, 3 1 L, L, − 3 − 1 − 1 (3.6 ⋅ K × 10 ⋅ mol ± 0.8) ). However, because of the large uncertain- (all in units of kg ties that are associated with most of the data, and because the ionic strength range stud- ied is so limited, the values are not recommended in this review. ∆ε L [49BAT/PIN] In several cases , [90DAN/ROB2] , , [90ROB/STE] [91APE/BAR] the constant heat capacity model was used, and the au- , [97BEN/PAL] thors reported ∆ values. Heat capacity changes have also been determined calori- C r,m p metrically at a pressure of 3.5 bar and in the temperature range 5 to 120 C ° [2001PAT/WOO] . From the data at zero ionic strength the following weighted averages are selected: 1 ο − − 1 = 1) = (222 ± 14) J K ∆ C ⋅ mol ((VII.8), r ⋅ p r,m 1 ο − 1 − ((VII.8), r = 2) = (167 ± 8) J ⋅ K C ⋅ mol ∆ p r,m − 1 1 − ο ± ⋅ K ((VII.8), ∆ 6) J mol = 3) = (116 r . C ⋅ p r,m The selections made in Section VII.3 lead to the following formation and en- tropy values: – –1 ο G cit ± , 298.15 K) = – (1225.8 ∆ 2.0) kJ·mol (H 2 fm –1 – ο H ∆ cit , 298.15 K) = – (1519.0 ± 2.0) kJ·mol (H 2 fm –1 – –1 ο cit , , 298.15 K) = (291.9 ± 1.1) J·K (H ·mol S 2 m 2– –1 ο ± G ∆ (Hcit 2.0) kJ·mol , 298.15 K) = – (1198.6 fm –1 2– ο ∆ (Hcit , 298.15 K) = – (1516.6 ± 2.0) kJ·mol H fm –1 2– –1 ο , 298.15 K) = (208.4 ± (Hcit 1.5) J·K ·mol S , m 3– –1 ο ∆ (cit G , 298.15 K) = – (1162.3 ± 2.0) kJ·mol fm –1 3– ο H , 298.15 K) = – (1519.9 ± 2.0) kJ·mol (cit ∆ fm 3– –1 –1 ο . , 298.15 K) = (75.6 ± 1.9) J·K S ·mol (cit m VII.4 Alkali metal citrate compounds and complexes + + VII.4.1 Complexes with Na and K + + or K Equilibrium constants for the formation of citrate complexes with either Na are reported in several studies, cf. Table VII-11.

374 VII Discussion of data selection for citrate compounds and compl exes 332 Table VII-11: Literature data on the formation of citrate complexes with alkali cations + + + + (M or K ). The equilibrium constants correspond to reactions: + = either Na m M 3) − ( r − 3) r ( m ((3)) + q mqr +− +− mqr ((3)) ] /[M β [M (H L) ] , with q ] [M (H L) = [H L U L ] H . , r , m r q r mrq mrq a + Method Medium t (°C) M I Reference log b (M) 10 1,0,1 2– M(cit) + 0.70 Na NCl / NaCl gl 0.16 Me − 25 [61WAL] 4 + − 0.43 K b + b ≈ Na 0.16 (0.70 ± 0.01) [64REC/ZAM] / NaCl THAM 25 ise + 0.01) ± (0.59 K c d + , 0 p NCl / NaClO 25 ? Na K (1.35 ± 0.05) [80ARE/CAL] ∆ → Me gl, 4 a 4 e f e , + ∆ Et p NBr / KNO K 37 K (0.56 ± 0.05) [80DAN/RIG] 0.15 gl, 3 a 4 d + + + + (0.68 ± / K / Na K 37 Na p 0.15 Et N 0.05) [81CUC/DAN] ∆ gl, a 4 e e c , + 0.25 ∆ (Et p NBr / K(Cl,NO K )) 25 K (0.57 ± 0.03) [84DAN/OST] gl, 3 4 a e e + + ∆ (Et N K / NaNO 0.25 ) 25 Na p (0.68 ± 0.05) [85DAN/ROB] gl, 3 a 4 + K , 0.1 Et gl, NI / (Na,K)Cl 25 Na ∆ 1.03 [90DAN/ROB] p a 4 c f g , , [90DAN/ROB2] 1.03 0.5 1 1.21 + 0.90 K 0.1 0.91 0.5 1 1.08 f h + , ± (Me → 0.02) [95ROB/GIA] (1.43 N)Cl 25 Na 0 ise 4 0.02) ± (0.93 0.1 0.02) ± (0.88 0.16 + 0.05) 0 → (1.42 ± K ± 0.05) 0.1 (0.92 0.05) ± (0.87 0.16 a + Method Medium t (°C) M log (M) I b Reference 1, 0, 2 10 – (cit) M 2 + p K 0.1 (Et gl, , NI / (Na,K)Cl) 25 Na ∆ 1.48 [90DAN/ROB] 4 a c f g , , [90DAN/ROB2] 1.50 0.5 1.91 1 + 1.03 0.1 K 0.5 1.05 1 1.35 f h + , N)Cl 25 Na → (2.31 ± 0.02) (Me 0 [95ROB/GIA] ise 4 ± 0.02) 0.1 (1.47 0.02) ± (1.39 0.16 + 0 K (1.95 → ± 0.06) ± 0.06) (1.11 0.1 ± 0.06) 0.16 (1.04 (Continued on next page)

375 VII.4 Alkali metal citrate compounds and compl exes 333 Table VII-11: (continued) a + b Medium t (°C) M Method (M) log I Reference 1,1,1 10 – M(Hcit) c d + , → 0 Me gl, NCl / NaClO 25 ? Na ∆ (0.6 ± p [80ARE/CAL] K 0.1) 4 a 4 e f e + , Et p NBr / KNO 37 K K gl, (0.30 ± 0.02) [80DAN/RIG] − 0.15 ∆ 3 4 a d + + + + [81CUC/DAN] N ∆ p K / K 37 Na 0.15 Et (0.10 ± 0.15) / Na gl, 4 a e e c + , ± NBr / K(Cl,NO p )) 25 K 0.25 (0.11 K 0.09) [84DAN/OST] ∆ (Et gl, 3 4 a e e + + 0.25 (Et [85DAN/ROB] ∆ p K 0.07)) ) 25 Na N (0.06 ± / NaNO gl, a 3 4 [90DAN/ROB] , + f,g c, K NI / (Na,K)Cl) 25 ∆ 0.1 (Et 0.53 p Na gl, 4 a [90DAN/ROB2] 0.5 0.54 1 0.68 + 0.1 0.33 K 0.34 0.5 1 0.47 a + Reference log Medium (°C) M (M) I b Method t 10 1,2,1 cit)aq M(H 2 d c, + → 0 Me gl, NCl / NaClO ∆ 25 ? Na p (0.25 ± 0.1) [80ARE/CAL] K 4 4 a e e c, + ∆ (Et K NBr / K(Cl,NO gl, K p − (0.20 ± 0.15) [84DAN/OST] 0.25 )) 25 3 4 a a: Methods: gl = pH-glass electrode; ise = ion selective electrode; ∆ p K = the formation constants for al- a kali-metal complexation were obtained from differences in protonation constants determined in differ- tetraalkylammonium salts). versus ent background electrolytes (alkali electrolytes + b: . The possibility of Na CNH THAM = tris(hydroxymethyl)aminomethane, (HOCH complexes with ) 2 2 3 THAM was not considered. values are also reported. ∆∆ and H S c: rm rm d: Combined measurements in NaClO with previous literature data in tetraalkylammonium electrolytes, 4 cf . Appendix A. e: A broader range of ionic strengths was studied, but only values at one ionic strength were reported. f: Included the formation of ion pairs in the background electrolyte (NaCl(aq), KCl(aq), or KNO (aq)). 3 g: Reported also the following equilibrium constants: 0.08) for: = (0.64 ± 0.03) and (0.73 ± log K 10 − − − (cit) 2 KNa(Hcit)(aq), respectively. U (cit) + Na and Na (Hcit)(aq) + K U (Hcit)(aq) 2 KNa(cit) K 2 2 2 2 h: The ionic strength was not kept constant during the experiments. Two experimental methods have been used: emf measurements with ion- selective electrodes, and comparison of prot onation constants obtained in different ionic media. The latter approach is based on the assumption that tetraalkylammonium ions do not form ion pairs with citrate anions, and that any differences in the values for the protonation constants in different media with the same ionic strength are due to the for- mation of alkali-metal complexes. However, a comparison between the following two cf . Table VII-2): values (

376 VII Discussion of data selection for citrate compounds and compl exes 334 1 in 1.21 m Et ± 0.02) log = (6.02 K NI at 25°C, [90ROB/STE] 4 10 1 log K = (5.67 ± 0.01) in 1.014 m Me NCl at 25°C, [97FOT/GIA] 4 10 1 reveals that specific ion interactions (activity coefficient effects) also may induce differ- ences in protonation constants. There is ther efore a large uncertainty associated with all ” in Table VII-11). With this K p ∆ the references that used this methodology (labelled “ a − and method the existence of protonated complexes has also been postulated (M(Hcit) M(H cit)(aq)), but these complexes must be regarded with scepticism when considering 2 the uncertainties associated with this technique. Two studies used ion-selective electrodes to determine the stability of citrate + + and K , [95ROB/GIA] . Both studies have an im- [64REC/ZAM] complexes with Na portant limitation in their experimental methodology: the ionic strength was not accu- “adjusted” the ionic [64REC/ZAM] rately controlled. Rechnitz and Zamochnick strength with THAM (tris(hydroxymethyl)aminomethane, (HOCH CNH ), a com- ) 2 3 2 pound usually used as pH-buffer, without investigating the possibility of formation of ). Furthermore, the concentration of sodium complexes with THAM (see [82SIG/SCH] 3 − cit in these experiments (10 mM) had a large contribution to the total ionic strength, as compared with the concentrations of the “background” electrolyte (0.1 M + ) CNH ). (HOCH ) CNH and 0.2 M (HOCH 2 3 2 23 3 De Robertis et al. [95ROB/GIA] also used a Na-sensitive glass electrode, and they described in their paper that “our measurements were carried out at different ionic strength values, and vary substantially during the experiments”. The equilibrium con- stants reported by these authors are therefore dependent on their model for activity coef- ficients. Furthermore, the authors also consid ered the formation of weak ion pairs, such as NaCl(aq) and KCl(aq), in disagreement with the SIT methodology used in this re- et al. view. Calculations using the equilibrium constants reported by De Robertis − − [95ROB/GIA] (cit) and K (cit) indicate that the complexes Na were always below 8% 2 2 of the total alkali metal. Therefore the existence of these complexes should not be con- sidered as completely proven. In conclusion, although there is some evidence suggesting the formation of cit- + + and K , there are many uncertainties concerning their rate complexes with Na –1 stoichiometry and stability, and the complexes, if formed, are weak ( K 10 kg·mol ). 1 In this review the interactions between alkali-metal ions and citrate at 25°C are instead cf . Section treated as specific ion interaction effect s included in the activity coefficients, VII.3. + + Enthalpy changes for citrate complexation with Na have been reported and K in three studies (Table VII-11), but as in the case of the equilibrium constants, no en- , [90DAN/ROB] , thalpy changes can be recommended by this review. In [84DAN/OST] 1 The equilibrium constant and ionic strength have been converted to molal units. The original value was 0.02) in 1 M Et ± log NI [90ROB/STE]. = (6.099 K 4 1 10

377 VII.4 Alkali metal citrate compounds and compl exes 335 2 − ο , values of ∆ for the formation of citrate complex, (M(cit) ) , [90DAN/ROB2] H rm were obtained from the temperature variation of formation constants, which were ob- obtained in different ionic media. Because tained by comparing protonation constants the temperature variations are small and the uncertainties with this methodology large, , enthalpy changes for these data are not considered by this review. In [80ARE/CAL] + - citrate complexation were determined calorimetrically, without specifying how Na this was performed, and therefore these data can not be included in this review. VII.5 Magnesium and calcium citrate compounds and complexes lcium citrate complexes VII.5.1 Magnesium and ca The experimental equilibrium data found in the literature on the complex formation of 2+ 2+ Mg with citrate are listed in Table VII-12. and Ca 2+ Table VII-12: Literature data on the formation constants for citrate complexes of Mg 2+ and Ca The uncertainties are given as reported in the references . . a ( ° C) Reference Method Ionic medium t b log 10 2+ 3 − − U Mg(cit) Mg + cit − 0.163 M (NaCl) 3.22 [34HAS/MCL] (b) 0.155 22 0.08 M (NaCl + K SO ) 37 3.58 [39NOR] sol 4 2 25 3.29 [59LI/LIN] pol 0.09 M ) NCl, NaCl or KCl) 25 3.55 [61WAL] pot 0.16 M ((CH 3 4 cix 0.1 M NH Cl 25 3.16 [63TOB/MIL] 4 sp 25 3.63 [63WAT/TRO] 0.1 M Tris and Tea [64CAM/OST] ) 20 3.40 pot 0.1 M (NaClO 4 cix 0.01 0.09 M (NaCl) 25 3.96 [64TOB/MIL] 0 → − [65PAT/PAN] 32.5 3.46 pot ) NCl 25 3.73 [65TAT/GRZ] 0.1 M (CH pot 3 4 pot 0.1 M ((CH ) NCl) 25 3.85 [70GRZ/TAT] 4 3 0.15 M NaCl 37 3.34 [74MEY] pot [75FIE/COB] 25 (3.38 ± 0.07) 0.1 M KNO pot 3 pot 0.1 M KNO 25 3.63 [80PEA] 3 0.03 M (KNO pot ) 37 (3.940 ± 0.005) [82AMI/DAN] 3 ) 37 (3.451 ± 0.005) [82AMI/DAN] pot 0.1 M (KNO 3 pot 0.3 M (KNO ) 37 (3.068 ± 0.005) [82AMI/DAN] 3 (c) → 0 25 (4.92 ± 0.05) [82HIR/KIS] ± 0.06) [85RIZ/ANT] pot 25 (3.46 0.1 M (NaCl) pot 0.15 M NaClO 37 (3.33 ± 0.01) [87BLA/BER] 4 pot 1.0 M NaClO 30 ? 3.62 [88GHA/MAN] 4 (Continued on next page)

378 VII Discussion of data selection for citrate compounds and compl exes 336 Table VII-12: (continued) a ( Method Ionic medium t C) ° Reference log b 10 3 − − 2+ Mg U + cit Mg(cit) cou [93GLA/MAJ] 0.15 M NaCl 25 3.27 37 3.24 [93GLA/MAJ] 0.15 M NaCl cou pot 0.5 M NaClO ± 0.01) 25 (2.71 [95PII/LAJ] 4 ± 0.01) [2001CHO/BON] (2.97 pot 25 0.3 m NaCl 25 1 m NaCl pot (2.40 ± 0.03) [2001CHO/BON] 25 (1.97 pot 2 m NaCl 0.05) [2001CHO/BON] ± 0.04) [2001CHO/BON] ± 25 (2.02 3 m NaCl pot 4 m NaCl pot (2.08 25 ± [2001CHO/BON] 0.02) 5 m NaCl pot 25 (2.07 [2001CHO/BON] ± 0.02) [2001SAR] 0.02) ± (3.31 pot 0.1 M NaCl 25 2+ 2 − Mg U Mg(Hcit)(aq) + Hcit pol 0.09 M 25 1.60 [59LI/LIN] 0.1 M (NaClO ) 20 1.84 [64CAM/OST] pot 4 pot 0.1 M (CH ) NCl 25 1.85 [65TAT/GRZ] 3 4 ) NCl) 25 1.92 [70GRZ/TAT] 0.1 M ((CH pot 4 3 0.15 M NaCl 37 1.62 [74MEY] pot pot 0.1 M KNO 25 (1.96 0.19) [75FIE/COB] ± 3 25 1.78 [80PEA] pot 0.1 M KNO 3 pot 0.03 M (KNO ) 37 (1.79 0.05) [82AMI/DAN] ± 3 0.1 M (KNO pot ± 0.07) [82AMI/DAN] ) 37 (1.51 3 ± 0.03) [82AMI/DAN] ) 37 (1.60 pot 0.3 M (KNO 3 (c) → ± 1.20) [82HIR/KIS] 25 0 (1.67 0.1 M (NaCl) 25 (2.13 pot 0.10) [85RIZ/ANT] ± 37 (1.94 ± 0.02) [87BLA/BER] 0.15 M NaClO pot 4 0.5 M NaClO pot 25 (1.23 ± 0.12) [95PII/LAJ] 4 0.3 m NaCl 25 (1.7 pot 0.2) ± [2001CHO/BON] 0.1) [2001CHO/BON] ± pot (1.1 25 1 m NaCl pot 2 m NaCl 25 (0.77 ± 0.01) [2001CHO/BON] pot 3 m NaCl (0.98 25 0.02) [2001CHO/BON] ± 0.1) [2001CHO/BON] ± 25 (1.2 4 m NaCl pot 5 m NaCl 25 pot (1.2 ± [2001CHO/BON] 0.1) 0.1 M NaCl pot (2.18 25 ± 0.03) [2001SAR] + 2+ − Mg U Mg(H + H cit) cit 2 2 pot 0.1 M NaClO 20 0.84 [64CAM/OST] 4 pot 0.1 M KNO 25 0.60 [80PEA] 3 0.15 M NaClO pot 37 (1.23 ± 0.06) [87BLA/BER] 4 pot 0.1 M NaCl 25 (1.66 ± 0.4) [2001SAR] (Continued on next page)

379 VII.5 Magnesium and calcium citrate compounds and compl exes 337 Table VII-12: (continued) a t ° ( Method Ionic medium Reference C) b log 10 3 − 2+ − + 3 Mg Mg H(cit) + H + 2cit U 22 ± 0.07) [87BLA/BER] 37 (10.41 0.15 M NaClO pot 4 (11.00 ± 0.02) [2001SAR] pot 0.1 M NaCl 25 − 4 − 3 2+ Mg + 2cit Mg(cit) U 2 0.15 M NaClO 37 (5.13 ± 0.04) [87BLA/BER] pot 4 25 (5.17 ± 0.02) [2001SAR] pot 0.1 M NaCl 2+ − + 3 2 − + cit MgH (cit) − H U Mg 1 − 0.1 M NaCl 25 (8.51 ± 0.03) [2001SAR] pot − 4 − − + 2+ 3 + 2cit U − 2Mg Mg H (cit) 2H − 22 2 pot 37 − (12.64 0.15 M NaClO 0.03) [87BLA/BER] ± 4 2+ − 3 − 3 + (cit) 2H U MgH Mg + cit − 2 − 0.15 M NaClO 0.02) 37 − (18.47 ± pot [87BLA/BER] 4 [2001SAR] 25 − (21.13 ± 0.07) 0.1 M NaCl pot 2+ 3 − − Ca U Ca(cit) + cit (b) − 0.163 M (NaCl) 22 (3.23 ± 0.03) [34HAS/MCL] 0.155 − 0.163 M (NaCl + KIO ) 25 3.21 [36MUU/LEB] sol 0.155 3 0.08 M (NaCl + K sol ) 37 3.50 [39NOR] SO 2 4 2+ − 0.04 0.31 m (NaCl) 25 3.17 [46JOS] ise–Ca → 0.15 m (NaCl) 3.46 [48PYN] dialysis 0.03 M 2 → (4.84 r. t. 0 ± 0.01) [51HEI] pot cix 25 (3.15 0.16 M NaCl ± 0.01) [52SCH/LIN] 0.15 M NaCl 28 3.12 [57LEF] pot [61PAT/PAN] pot ? 33 3.42 0.16 M ((CH pot NCl, NaCl or KCl) 25 3.15 [61WAL] ) 3 4 cix 0.16 M NaCl 25 3.16 [63MAT] 20 3.55 [64CAM/OST] 0.1 M NaClO pot 4 pot 0.1 M (KNO ) 20 3.24 [69BOS/MAR] 3 2+ ise–Ca 0.1 M NaClO [69REC/HSE] 25 3.67 4 Cl 19.6 (3.15 ± 0.015) [71RUM] cix 0.2 M NH 4 0.2 M NH cix Cl 28.2 (3.10 ± 0.015) [71RUM] 4 cix 0.2 M NH Cl 38 (3.04 0.015) [71RUM] ± 4 ± 0.015) [71RUM] Cl 51.4 (2.97 0.2 M NH cix 4 cix 0.2 M NH Cl 65.0 (2.82 ± 0.015) [71RUM] 4 pot 37 3.27 [74MEY] 0.15 M NaCl 25 (3.50 ± 0.06) [75FIE/COB] 0.1 M KNO pot 3 (Continued on next page)

380 VII Discussion of data selection for citrate compounds and compl exes 338 Table VII-12: (continued) a ( Method Ionic medium t C) ° Reference b log 10 − 3 − 2+ Ca(cit) Ca + cit U ) 25 3.54 [75RAM/MAN] pot 0.1 M (NaClO 4 ? ? ? 2.95 [79BUL/SAC] 2+ ise–Ca 0.1 M (NaCl) (3.42 ± 0.06) [79CRA/MOO] 25 [80PEA] 25 3.64 0.1 M KNO pot 3 pot 25 4.87 [80PEA] → 0 0.03 M (KNO pot ) 37 (3.909 ± 0.003) [82AMI/DAN] 3 [82AMI/DAN] ± ) 37 (3.485 0.006) 0.1 M (KNO pot 3 0.3 M (KNO pot ± 0.007) [82AMI/DAN] ) 37 (2.936 3 (c) → (4.85 ± 0.05) [82HIR/KIS] 25 0 ± 0.04) [85RIZ/ANT] 0.1 M (NaCl) (3.43 pot 25 pot 0.15 M NaClO 37 (3.36 ± 0.01) [87BLA/BER] 4 1.0 M NaClO pot 30 ? 3.02 [88GHA/MAN] 4 2+ 0.0198 M NaCl 37 (3.23 ± 0.02) [91SIN/YEB] ise–Ca 2+ ise–Ca 0.056 M NaCl 37 (3.27 ± [91SIN/YEB] 0.04) 2+ ise–Ca 0.15 M NaCl 37 (3.69 ± 0.01) [91SIN/YEB] 2+ 37 (4.08 ± 0.05) [91SIN/YEB] 0.17 M NaCl ise–Ca 2+ ise–Ca 18 3.27 [91SIN/YEB] 0.15 M NaCl 2+ ise–Ca 25 3.26 [91SIN/YEB] 0.15 M NaCl 2+ 0.15 M NaCl 37 3.27 [91SIN/YEB] ise–Ca 2+ ise–Ca 0.15 M NaCl 45 3.25 [91SIN/YEB] 0.15 M NaCl sol [91SIN/YEB] 37 3.29 25 3.5 [92GLA/HUL] 0.1 M NaClO cou 4 0.15 M NaCl 25 3.17 [93GLA/MAJ] cou cou 37 3.28 [93GLA/MAJ] 0.15 M NaCl [95PII/LAJ] 25 (2.71 ± 0.01) pot 0.5 M NaClO 4 cal 0.1 M (NaCl) (3.42 25 0.02) ± [96MIR/PAS] cal 0.2 M (NaCl) 25 (3.11 0.02) [96MIR/PAS] ± ± 0.02) [96MIR/PAS] (2.83 cal 25 0.3 M (NaCl) cal 0.4 M (NaCl) 25 (2.70 ± 0.02) [96MIR/PAS] cal 25 (2.38 0.5 M (NaCl) ± 0.02) [96MIR/PAS] → 0 25 (4.80 ± 0.03) [96MIR/PAS] cal 0.1 M NaCl 25 (3.38 pot ± 0.1) [2001SAR] (Continued on next page)

381 VII.5 Magnesium and calcium citrate compounds and compl 339 exes Table VII-12: (continued) a t ( Method Ionic medium ° C) Reference b log 10 − 2 2+ U Ca(Hcit)(aq) Ca + Hcit ) 25 2.3 [36MUU/LEB] 0.110 M (NaCl + KIO sol 3 0 r. t. (3.29 ± 0.01) pot → [51HEI] sol 0.03 → 0 3.09 [53DAV/HOY] − 0.06 M 25 [63MAT] 0.16 M NaCl cix 2.67 pot 0.1 M (NaClO [64CAM/OST] ) 20 2.10 4 pot 0.15 M NaCl 37 1.82 [74MEY] [75FIE/COB] 25 (2.32 ± 0.15) 0.1 M KNO pot 3 pot 0.1 M KNO 25 2.03 [80PEA] 3 pot 0 3.03 [80PEA] → 25 [82AMI/DAN] ) 37 (2.23 0.02) ± pot 0.03 M (KNO 3 0.1 M (KNO pot ) 37 (2.09 ± [82AMI/DAN] 0.02) 3 pot 0.3 M (KNO ) 37 (1.65 ± 0.05) [82AMI/DAN] 3 25 (2.79 ± 0.13) [82HIR/KIS] 0 → (c) 0.1 M (NaCl) 25 (2.80 pot 0.20) [85RIZ/ANT] ± 0.15 M NaClO pot 37 (2.08 ± 0.02) [87BLA/BER] 4 25 (1.38 ± 0.09) [95PII/LAJ] 0.5 M NaClO pot 4 pot (2.07 0.1 M NaCl 25 [2001SAR] ± 0.02) 2+ − + Ca U Ca(H cit) cit + H 2 2 , [53DAV/HOY] sol 0.03 1.10 − 0.06 M → 0 [55DAV/HOY] pot 0.1 M (NaClO ) 20 1.05 [64CAM/OST] 4 0.1 M KNO pot [80PEA] 25 1.04 3 pot 0.15 M NaClO ± 0.03) 37 (1.23 [87BLA/BER] 4 ± 0.04) [2001SAR] 0.1 M NaCl 25 (1.18 pot 3 − 2+ 3 − + Ca U + H CaH(cit) + 2cit 2 pot 25 (10.60 ± 0.04) [2001SAR] 0.1 M NaCl − 4 2+ − 3 U + 2cit Ca(cit) Ca 2 cix 0.2 M NH Cl 25 (4.33 ± 0.05) [71RUM] 4 0.15 M NaClO 37 (4.97 ± 0.04) [87BLA/BER] pot 4 pot 0.1 M NaCl 25 (5.44 ± 0.02) [2001SAR] (Continued on next page)

382 VII Discussion of data selection for citrate compounds and compl exes 340 Table VII-12: (continued) a C) Method Ionic medium ( t ° Reference b log 10 3 − + 2+ 2 − U Ca CaH − H (cit) + cit 1 − [87BLA/BER] 37 − (8.94 ± 0.03) pot 0.15 M NaClO 4 25 − (9.06 ± 0.01) [2001SAR] pot 0.1 M NaCl − 6 − + 2+ 3 Ca − 2H CaH (cit) + 2cit U − 22 0.15 M NaClO 37 − pot ± 0.02) [87BLA/BER] (16.81 4 a: Refers to the reactions indicated, the ionic strength and temperature given in the table. Uncertainties are those given in the references. ). b: Frog heart method (see Appendix A [34HAS/MCL] c: Isotachophoresis. Most of these experimental equilibrium constants were obtained by potenti- ometric titration. Probable reactions are: 3 − − 2+ U Mg(cit) (VII.9) + cit Mg 2+ 2 − U Mg(Hcit)(aq) + Hcit (VII.10) Mg 2+ − + Mg cit U Mg(H (VII.11) cit) + H 2 2 3 − − 2+ Ca Ca(cit) (VII.12) + cit U 2+ − 2 + Hcit U Ca(Hcit)(aq) (VII.13) Ca 2+ + − Ca cit U Ca(H cit) + H . (VII.14) 2 2 Also the formation of a number of other species has been postulated to fit ex- − 3 − − 2 3 4 − − 4 , Ca(cit) Mg(cit) , Mg(H , Ca(Hcit)(cit) , cit) , perimental data: Mg(Hcit)(cit) − 1 2 2 − 3 − 2 − 4 6 − Ca(H , , , Ca(H cit) Mg (H cit) cit) [71RUM] , Mg(H , cit) [87BLA/BER] 1 − − 2 − 212 − 12 [2001SAR] . However, the existence of these species is doubtful and the numerical re- cf. Appendix A). sults are not credited by this review ( − − The formation of Mg(cit) , Mg(Hcit)(aq), Ca(cit) and Ca(Hcit)(aq) is com- − ture. Although very weak, the formation of Mg(H monly accepted in the litera cit) and 2 − cit) ). To show the is also indicated ( [64CAM/OST] , [80PEA] , [87BLA/BER] Ca(H 2 2+ 2+ extent of the complex species of Mg citrates formed under ordinary condi- or Ca − 3 tions, the simulated titration curves of 2 10 × M citric acid in the absence and presence 2+ 2+ 3 − M Ca at I = 0.1 M, and the distribution of the Ca -citrates as a function of of 2 10 × + − in this titration procedure are given in Figure VII-12 and Figure VII-13, log [H ] 10 respectively. For the simulation, both the values accepted in this review (at I = 0.1 M) for the protonation constants and the formation constants of Ca − citrate complexes (dis- cussed later in this section) are used. Figure VII-13 indicates that the predominant spe- − and Ca(Hcit). As shown by cies in the course of the potentiometric titration are Ca(cit) + + cit) or Ca(H cit) is significant only at a lower this example, the contribution of Mg(H 2 2 pH region. Since the titration curves of citric acid in the absence and presence of the

383 VII.5 Magnesium and calcium citrate compounds and compl exes 341 metal ion do not differ much in this region, very careful measurements of pH (and the + corresponding conversion to ) should be done for the determination of these log [H ] − 10 formation constants. 3 − 3 − M citric M citric acid, (2) 2 × 10 Figure VII-12: Simulated titration curves of (1) 2 × 10 3 − 2+ × 10 acid and 2 M Ca = 0.1 M with the following assumptions: at I =2) = 4.36 and log log K ( r =1) = 5.69, log K ( r K =3) = 2.92 r ( 10 3 10 1 10 2 43 +− − rr for Hcit (VII.12) = 3.48, b log HHcit + U , and rr 10 1 − 1 (VII.14) = 1.09 b b log log (VII.13) = 2.05, 10 1 10 1 8 3 − M, I = 2·10 = 0.1 M c cit H 3 3 − 7 c = (1) 0, (2) 2·10 M Ca 6 (1) ] + [H 5 10 (2) log − 4 3 2.5 0.0 3.0 0.5 1.0 1.5 2.0 per ( added c c = c ) NaOH H Ca cit 3 3 − 10 Figure VII-13: Distribution of complex species in the simulated titration of 2 × M 2+ 3 − citric acid and 2 × 10 M Ca at I = 0.1 M (shown in Figure VII-12). -2 2+ Ca − Ca(cit) -3 -4 Ca(Hcit) c -5 10 -log + -6 Ca(H cit) 2 -7 3 − I = c = 0.1 m M, c = 2·10 cit H Ca 3 -8 34567 + − log [H ] 1 0

384 VII Discussion of data selection for citrate compounds and compl exes 342 Because of various shortcomings in the experimental procedures or in the re- [34HAS/MCL] porting of the results, the values of [39NOR] , [36MUU/LEB] , , [61WAL] , , [59LI/LIN] [51HEI] [61PAT/PAN] , [53DAV/HOY] , [63MAT] , , , [48PYN] [63TOB/MIL] , [65PAT/PAN] , [69BOS/MAR] , , , [64TOB/MIL] [71RUM] [75RAM/MAN] , [82HIR/KIS] , [85RIZ/ANT] , [88GHA/MAN] , , [79BUL/SAC] [2001SAR] have been rejected by this review ( cf. Appendix A). , [96MIR/PAS] Based on the discussion of the remaining literature studies (see Appendix A), values listed in Table VII-13 are accepted as reliable in this review. Data at I ≤ 0.5 M I ≤ 0.5 m, the contribution of the ≅ ( 0.5 m) are available in various ionic media. At m ∆ε term to be smaller than the experimental to the SIT formulation can be considered I m uncertainties, except in the cases of (CH ) NCl medium where the interaction of 4 3 + with citrate cannot be neglected. N ) (CH 4 3 2+ 2+ Table VII-13: Accepted formation constants for citrate complexes of Mg and Ca used to derive the selected values. Uncertainties have been estimated in this review. a Ionic medium ° C) Reference t log b ( 1 10 2+ − 3 − Mg(cit) Mg + cit U 0.03 M KNO 37 (3.94 ± 0.1) [82AMI/DAN] 3 0.1 M KNO 25 (3.38 ± 0.3) [75FIE/COB] 3 0.1 M KNO 25 (3.50 ± [80PEA] 0.1) 3 0.1 M KNO ± 0.1) [82AMI/DAN] 37 (3.45 3 37 (3.07 ± 0.1) [82AMI/DAN] 0.3 M KNO 3 0.1 M NaClO 20 (3.40 ± [64CAM/OST] 0.15) 4 0.5 M NaClO 25 (2.71 0.1) [95PII/LAJ] ± 4 NCl 25 (3.73 ± 0.2) [65TAT/GRZ] ) 0.1 M (CH 3 4 0.1 M (CH ) NCl 25 (3.85 ± 0.2) [70GRZ/TAT] 3 4 0.15 M NaCl 25 (3.27 ± 0.2) [93GLA/MAJ] [87BLA/BER] 37 (3.33 ± 0.1) 0.15 M NaClO 4 37 0.15 M NaCl (3.24 0.2) [93GLA/MAJ] ± 0.15 M NaCl 37 (3.34 [74MEY] ± 0.5) ± 0.2) [2001CHO/BON] 0.3 m NaCl (2.97 25 1 m NaCl 25 (2.40 ± 0.2) [2001CHO/BON] 2 m NaCl 25 (1.97 ± 0.2) [2001CHO/BON] ± [2001CHO/BON] 0.2) 25 (2.02 3 m NaCl 4 m NaCl 25 (2.08 ± 0.2) [2001CHO/BON] 5 m NaCl 25 (2.07 ± 0.2) [2001CHO/BON] (Continued on next page)

385 VII.5 Magnesium and calcium citrate compounds and compl exes 343 Table VII-13: (continued) a ° C) t ( log b Ionic medium Reference 1 10 2+ 2 − U Mg(Hcit)(aq) Mg + Hcit ± 0.2) [82AMI/DAN] 37 (1.79 0.03 M KNO 3 25 (1.96 ± 0.5) [75FIE/COB] 0.1 M KNO 3 0.1 M KNO 0.2) ± [80PEA] 25 (1.78 3 37 (1.51 0.2) [82AMI/DAN] ± 0.1 M KNO 3 0.3 M KNO ± 0.2) [82AMI/DAN] 37 (1.60 3 0.1 M NaClO 20 (1.84 ± 0.15) [64CAM/OST] 4 25 (1.23 ± 0.2) [95PII/LAJ] 0.5 M NaClO 4 0.1 M (CH ) ± 0.2) [65TAT/GRZ] NCl 25 (1.85 3 4 0.1 M (CH ) NCl 25 (1.92 ± 0.2) [70GRZ/TAT] 3 4 37 (1.94 ± 0.2) [87BLA/BER] 0.15 M NaClO 4 0.15 M NaCl 37 (1.62 ± 0.5) [74MEY] 25 (1.1 1 m NaCl 0.3) ± [2001CHO/BON] ± 0.3) [2001CHO/BON] 25 2 m NaCl (0.77 3 m NaCl 25 (0.98 0.3) [2001CHO/BON] ± 4 m NaCl 25 (1.2 0.3) ± [2001CHO/BON] ± 0.3) [2001CHO/BON] 25 5 m NaCl (1.2 + 2+ − Mg(H Mg cit U + H cit) 2 2 0.1 M KNO ± 0.4) [80PEA] 25 (0.60 3 20 (0.84 [64CAM/OST] 0.2) 0.1 M NaClO ± 4 0.15 M NaClO ± 0.4) [87BLA/BER] 37 (1.23 4 2+ − 3 − Ca + cit U Ca(cit) 37 (3.91 ± 0.1) [82AMI/DAN] 0.03 M KNO 3 0.3) 25 (3.50 ± 0.1 M KNO [75FIE/COB] 3 0.1 M KNO 25 (3.50 ± 0.1) [80PEA] 3 0.1 M KNO 37 (3.49 ± 0.1) [82AMI/DAN] 3 37 (2.94 ± 0.1) [82AMI/DAN] 0.3 M KNO 3 0.1 M NaClO 20 (3.55 ± 0.15) [64CAM/OST] 4 0.1 M NaClO 25 (3.67 ± 0.2) [69REC/HSE] 4 25 (3.5 ± 0.2) [92GLA/HUL] 0.1 M NaClO 4 0.5 M NaClO 25 (2.71 ± 0.1) [95PII/LAJ] 4 0.15 M NaClO 37 (3.36 0.1) [87BLA/BER] ± 4 ± 0.2) [79CRA/MOO] 0.1 M NaCl (3.42 25 0.15 M NaCl 25 (3.17 ± [46JOS] 0.4) 0.16 M NaCl 25 (3.15 ± 0.2) [52SCH/LIN] (Continued on next page)

386 exes VII Discussion of data selection for citrate compounds and compl 344 Table VII-13: (continued) a ° t log b ( Reference Ionic medium C) 10 1 3 2+ − − Ca(cit) U Ca + cit (3.17 0.2) [93GLA/MAJ] 25 0.15 M NaCl ± (3.28 ± 0.2) [93GLA/MAJ] 0.15 M NaCl 37 37 0.15M NaCl (3.27 [74MEY] ± 0.5) − 45 (3.25 ± 0.1) [91SIN/YEB] 18 0.16 M NaCl 2+ 2 − Ca + Hcit U Ca(Hcit)(aq) 37 (2.23 0.03 M KNO ± 0.2) [82AMI/DAN] 3 0.1 M KNO ± 0.5) [75FIE/COB] 25 (2.32 3 0.1 M KNO 25 (2.03 0.2) [80PEA] ± 3 37 (2.09 ± 0.2) [82AMI/DAN] 0.1 M KNO 3 0.3 M KNO 37 (1.65 ± [82AMI/DAN] 0.2) 3 0.1 M NaClO 20 (2.10 ± 0.15) [64CAM/OST] 4 25 (1.38 ± 0.2) [95PII/LAJ] 0.5 M NaClO 4 0.15 M NaClO 37 (2.08 ± 0.2) [87BLA/BER] 4 0.15M NaCl (1.82 37 0.5) [74MEY] ± 2+ − + Ca cit U + H Ca(H cit) 2 2 25 (1.04 0.4) [80PEA] 0.1 M KNO ± 3 20 (1.05 0.2) [64CAM/OST] 0.1 M NaClO ± 4 0.15 M NaClO ± 0.4) [87BLA/BER] 37 (1.23 4 a: Refers to the reactions indicated, the ionic strength and temperature given in the table. b for Reaction (VII.9), listed in The values of the equilibrium constant, log 10 1 Table VII-13 (except the cases of (CH ) NCl medium) were fitted to the following 4 3 equation: 2 ∆ zI 0.5091 m ο (VII.15) bb I −∆ε log =+ log m 1 10 10 1 + 11.5 I m 23 +− +−+− where: (Mg , Cl ) (Na , cit ) ∆ε=ε (Na , Mg(cit) ) −ε −ε –1 2 +− ε=± · mol (0.19 0.02) kg (Mg , Cl ) –1 3 +− . I (Na , cit ) ε =−±+± kg · mol (0.15 0.03) (0.13 0.03) log m 10 The result of the fitting is shown in Figure VII-14 and Figure VII-16 with the fitted and selected values of: –1 ο +− log . b (VII.9) = (4.81 ± 0.03) and (Na , Mg(cit) ) (0.03 0.03) · ε=± kg mol 10 1 +− is consistent with the generally observed values for the ε The value of (Na , Mg(cit) ) + with complex anions of charge − 1. interaction of Na

387 VII.5 Magnesium and calcium citrate compounds and compl exes 345 − 2+ 3 − Mg cit Mg(cit) log of + U at each b Figure VII-14: Fitting of the values of 10 ionic strength (given in Table VII-13) to the SIT equation. Solid line is drawn by using the result of the fitting equation (VII.15) with: 2 ο z (VII.9) = (4.81 ± 0.03), ∆ log b = – 12, 1 10 2 +− +− (Mg , Cl ) ∆ε = ε ' − ε (Na , Mg(cit) ) –1 +− 2 mol · (Mg , Cl ) ε=± (0.19 0.02) kg –1 3 +− I (Na , cit ) ε =−±+± kg · mol (0.15 0.03) (0.13 0.03) log m 10 –1 +− (Na , Mg(cit) ) (0.03 0.03) kg · mol ε=± . [64CAM/OST] [74MEY] 6.0 [75FIE/COB] [80PEA] [82AMI/DAN] m I [87BLA/BER] ) − [93GLA/MAJ] 3 [95PII/LAJ] , cit [2001CHO/BON] + 5.5 (Na ε − D + 12 β 10 5.0 log 4.5 012345 I /m

388 exes VII Discussion of data selection for citrate compounds and compl 346 For Reaction (VII.10), 2+ 2 − q Mg(Hcit)(aq), + Hcit Mg the values listed in Table VII-13 were used to fit equation: 2 0.5091 ∆ zI m ο bb I (VII.16) log =+ −∆ε log 1 1 10 10 m I + 11.5 m ο where b and ∆ε are treated as fitting parameters. The result of the fitting is log 10 1 shown in Figure VII-15 and Figure VII-16 with the values of: –1 ο . b (VII.10) = (2.60 ± 0.07) and (0.13 0.05) ∆ε = − ± kg · mol log 1 10 The value of ∆ε is consistent with the selected values of ε : 2+ 2 −+− (Na , Hcit ) (0.19 0.02) (0.04 0.02) (0.15 0.03) (Mg , Cl ) ∆ε=−ε −ε =−±+±=−± . ο b with previously selected log Thus, this review selects the above value of 10 1 . ε values of Figure VII-15: Fitting of the values of log b (VII.10) at each ionic strength (given in 1 10 Table VII-13) to the SIT equation. Solid line is drawn by using the result of the fitting equation (VII.16). –1 ο log . b (VII.10) = (2.60 ± 0.07); (0.13 0.05) ∆ε = − ± kg · mol 1 10 [64CAM/OST] [74MEY] 4.0 [75FIE/COB] [80PEA] [82AMI/DAN] [87BLA/BER] [95PII/LAJ] 3.5 [2001CHO/BON] D + 8 3.0 β 10 g lo 2.5 2.0 012345 I /m

389 VII.5 Magnesium and calcium citrate compounds and compl 347 exes For Reaction (VII.11), 2+ + − , Mg(H + H q cit cit) Mg 2 2 the values listed in Table VII-13 were used in the equation: 2 0.5091 z I ∆ m ο bb (VII.17) log log =+ 10 10 11.5 I + m ∆ε I where the I ≤ 0.5 M ( ≅ 0.5 m). The term is neglected since the data are limited to m result of the fitting is shown in Figure VII-16. The selected values for magnesium com- plexes of citrate are summarised in Table VII-14. 2+ log -citrate at each ionic strength to b of Mg Figure VII-16: Fitting of the values of 1 10 2 0.5091 ∆ zI m ο I bb log log =+ −∆ε 1 10 1 10 m I 11.5 + m where ο 2 (VII.11) = (1.31 ± 0.16), z ∆ = − 4, log 0 I ∆ ε= b 10 1 m 1 − ο 2 b (VII.10) = (2.60 ± 0.07), log z ∆ = − 8, (0.13 0.05) ∆ε = − ± kg·mol 1 10 ο 2 3 + − 0.03) log z ∆ = − 12, b '(Na,cit) (VII.9) = (4.81 ± ∆ε = ∆ε − ε 1 10 1 − ∆ε = − , kg·mol ± (0.16 0.03) ' 1 − 3 −+ (cit , Na ) ε=−±+± , kg·mol I (0.15 0.03) (0.13 0.03) log 10 m 5 2+ -citrate Mg 4 3 − β β for Mg(cit ) log 10 1 10 log 2 1 (aq) for Mg(Hcit) β log 10 1 + for Mg(H cit) β log 10 2 1 0 012345 /m I

390 exes VII Discussion of data selection for citrate compounds and compl 348 − , Ca(Hcit)(aq) and Selected values for the calcium complexes, Ca(cit) + Ca(H were obtained in a manner similar to that used for magnesium species, by cit) 2 I ∆ε term is negligible. The results of the fittings are assuming that the contribution of m shown in Figure VII-17, and the selected values are given in Table VII-14. 2+ of Ca b -citrate at each ionic strength to log Figure VII-17: Fitting of the values of 10 1 2 ∆ I z 0.5091 m ο =+ bb log log 10 1 10 1 I + 11.5 m where 2 ο ∆ log ± 0.16) 4 z (VII.14) = (1.53 = − b 1 10 2 ο ∆ log (VII.13) = (2.92 ± 0.07) 8 z b = − 1 10 ο 2 log b 12. (VII.12) = (4.80 ± 0.03) − z ∆ = 1 10 5 2+ Ca -citrate [46JOS] [52SCH/LIN] 4 [64CAM/OST] [69REC/HSE] − [74MEY] log for Ca(cit) β 10 1 [75FIE/COB] 3 [79CRA/MOO] β [80PEA] 10 [82AMI/DAN] log β log for Ca(Hcit)(aq) [87BLA/BER] 1 10 2 [91SIN/YEB] [93GLA/MAJ] [95PII/LAJ] 1 + log for Ca(H cit) β 2 1 10 0 0.5 0.2 0.0 0.1 0.3 0.4 I /m

391 VII.5 Magnesium and calcium citrate compounds and compl exes 349 2+ 2+ and Ca Table VII-14: Selected formation constants for the citrate complexes of Mg at 25 ° C. ο b log Reaction 10 2+ − + ± U Mg(H + H cit) Mg (1.31 cit 0.16) 2 2 2+ 2 − U + Hcit (2.60 ± 0.07) Mg Mg(Hcit)(aq) 3 2+ − − Mg 0.03) (4.81 ± Mg(cit) + cit U − 2+ + cit ± U Ca(H + H cit) 0.16) (1.53 Ca 2 2 2+ − 2 Ca Ca(Hcit)(aq) (2.92 ± 0.07) + Hcit U − 3 2+ − Ca + cit (4.80 ± 0.03) U Ca(cit) ° C. Data at 25 ° C and at 37 ° C were ob- Table VII-13 contains several data at 37 tained by different authors and the differences between those at 25 C and at 37 ° C are ° et al [91SIN/YEB] also obtained the values of . within the estimated uncertainties. Singh − ο I b (VII.12) for Ca(cit) = 0.15 – from the measurements at 18, 25, 37, 45 ° C and log 1 10 ο 0.17 M NaCl. The temperature dependence of log (VII.12) was smaller than their b 1 10 experimental error. From these observations, we can estimate that the variation of − − ο ο b (VII.9) and and Ca(cit) log log are less than 0.1 for the b (VII.12) for Mg(cit) 1 10 10 1 temperature difference of 45 and 18 ° C. From the relations: 2 ∆ / ∆ 0.1/(45 18) T <− b and , log dln /d ∆ / THRT =− β 10 1 1 − − 1 we can roughly estimate that for Ca(cit) ∆ (VII.12) = (0 , 6) kJ·mol = 0.1 H at I ± m rm + 2+ 2+ cit) (M = Mg or Ca ), the uncertain- m. For the formation of M(Hcit)(aq) and M(H 2 ure dependences. By using isothermal calo- ties are too large to discuss their temperat ο directly obtained (VII.12) and log b et al rimeter, Mironov [96MIR/PAS] . 10 1 (VII.12) H ∆ (VII.12) at I = 0.1 − 0.5 M NaCl, pH = ∆ 12 and 25 ° C. The values of H rm rm − 1 − 1 − 1 ( ( I = 0.2 M), 3.0 kJ·mol = 0.1 M), 2.1 kJ·mol I ( I = 0.3 M), 3.4 are 1.2 kJ·mol − 1 − 1 I = 0.4 M) and 3.7 kJ·mol 0.3. ( ( I = 0.5 M) with the uncertainties of ± 0.2 to ± kJ·mol They are small and do not contradict the general observation of the temperature depend- ο b log (VII.12). However these data may contain much larger errors than ence of 1 10 given by the authors since there is a possibility of the formation of hydroxide and chlo- 2+ under the condition adopted in this work (see Appendix A). In ride complexes of Ca ο ∆ H (VII.12) with a large uncertainty as: conclusion, this review selects the value of rm − 1 ο ∆ ((VII.12), 298.15 K) = (0 ± . H 6) kJ mol rm The above selections yield the formation and molar entropy values summarised in Table VII-15:

392 exes VII Discussion of data selection for citrate compounds and compl 350 2+ Table VII-15: Selected Gibbs energy of and formation of the citrate complexes of Mg – 2+ and selected molar entropy of Ca(cit) . Ca ο ο ο –1 –1 –1 –1 ∆ ∆ (kJ·mol H ) ) S ) (J·K G ·mol Species (kJ·mol m fm fm + Mg(H – (1688.7 ± 2.6) cit) 2 Mg(Hcit)(aq) – (1668.8 ± 2.4) − Mg(cit) – (1645.1 ± 2.4) + cit) ± 2.4) – (1787.4 Ca(H 2 2.3) ± Ca(Hcit)(aq) – (1768.0 − Ca(cit) 20) 2.3) ± 6.4) ± ± – (2062.9 – (1742.5 (111 VII.5.2 Magnesium and calcium citrate compounds In Table VII-16, there are several papers reporting the solubility products of the solid 2+ 2+ or Ca with citrate. All the experimental points reported in these papers species of Mg are summarised in Figure VII-18. 2+ Table VII-16: Literature data on the solubility product of the solid compounds of Mg 2+ with citrate. and Ca a ° C) log Reference K Ionic medium t ( s,0 10 2+ 3 − 15H O(l) O(cr) U + 15H (cit) + 2cit Mg ⋅ 3Mg 3 2 2 2 − 0.22 M (H, Na)ClO → 0.17 0.1 25 − (11.10 ± 0.04) [69SKO/KUM] 4 2+ 3 − 9H + 9H O(cr) U 3Mg Mg (cit) O(l) ⋅ + 2cit 2 2 2 3 c –1 58 0.0467 → 0.0705 mol·kg not described 13 [93APE] → − 2+ 3 + 14H ⋅ 14H Mg O(cr) U 3Mg O(l) + 2cit (cit) 2 2 2 3 c –1 [93APE] 0.1077 mol kg → → 0.0295 not described 13 53 2+ 3 − 4H + 4H O(cr) U 3Ca Ca + 2cit O(l) (cit) ⋅ 2 2 2 3 − 0.5 M (Na,Ca)cit 21 − (17.63 ± 0.08) + (10.84 ± 0.23) I [60BOU/MAR] 0.004 m (14.68 0.12 → 0.1 25 − 0.17 M (H, Na)ClO ± 0.09) [69SKO/KUM] − 4 c –1 10 not described 0.0014 mol kg → 0.0017 [93APE] 58 → b 0.54 M (NaCl) or I − [95ROB/GIA2] 8.52 − 16.95 + 30 g ( I ) − 25 0.01 ((CH ) NCl) → 0 4 3 3.50 m NaClO − → 0 25 − (17.81 ± 0.03) [2001CIA/TOM] 0.10 4 − 3 2+ + Ca(Hcit)(cr) + H Ca U + cit b 0.54 M (NaCl) or ((CH ) NCl) → 0 0.01 − 11.35 + 14 g ( I ) − 5.95 I − [95ROB/GIA2] 25 3 4 a: Refers to the reactions indicated, the ionic strength and temperature given in the table. Uncertainties are those given in the references. 3/2 I () /(2 3 ) 0.1 gI . I I =++ b: c: Only total solubilities are reported.

393 VII.5 Magnesium and calcium citrate compounds and compl 351 exes (cit) 4H O(cr) and the ⋅ Figure VII-18: The experimental solubility product data of Ca 2 3 2 fitting of the selected values using equation (VII.19), including and neglecting the ∆ε terms. log na and H O 10 2 -9 -10 2 -11 1 -12 -13 s,0 -14 K 10 [60BOU/MAR] log -15 [69SKO/KUM] [95ROB/GIA2] -16 [2001CIA/TOM] − 4log − a Ι ∆ε 1: − D 17.9 + 30 H O 10 m -17 2 17.9 + 30 D 2: − -18 -19 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 I / m begin to decrease with at a fairly low ionic I [95ROB/GIA2] The data by strength of around 0.1 m. Since it is unlikely for the ionic interaction term to have such [95ROB/GIA2] a large contribution, this review does not accept the values given by were obtained by treating the data with [2001CIA/TOM] (see Appendix A). The data by −+ 3 (cit , Na ) m ε=−±−± . This value is different (0.08 0.03) (0.06 0.01) the value of a N from that selected by NEA-TDB review. Based on the discussion in the Appendix A, only the data at lower ionic strength were accepted by this review. The values accepted as reliable are given in Table VII-17.

394 VII Discussion of data selection for citrate compounds and compl exes 352 Table VII-17: Accepted literature values of the solubility product of the solid 2+ ° with citrate to estimate the selected values at 25 C. compounds of Ca a Ionic medium K log s,0 10 Reference 2+ 3 − (cit) O(cr) U 3Ca Ca + 2cit ⋅ 4H + 4H O(l) 2 2 3 2 (17.8 ± 0.2) [2001CIA/TOM] 0 m − − (14.5 ± 0.2) [2001CIA/TOM] 0.10 m NaClO 4 ± (13.6 − 0.2) [2001CIA/TOM] 0.25 m NaClO 4 − (12.8 ± 0.2) [2001CIA/TOM] 0.51 m NaClO 4 a: Refers to the reaction indicated and the ionic strength given in the table. Uncertainties are estimated by this review. 2+ For the Ca ered the values for the following -citrate system, this review consid reaction: 2+ 3 − Ca U O(cr) (cit) 3Ca O(l). (VII.18) + 2cit ⋅ 4H + 4H 2 2 3 2 According to the SIT , the equilibrium constant should be expressed by: 2 ο log log log K KzDIna =+∆−∆ε− (VII.19) ss m 10 ,0 10 10 H O ,0 2 0.5091 I m where: , D = I + 11.5 m 2+ 3 −−+ 3(Ca ,ClO) 2(cit ,Na ) ∆ε=ε , +ε 4 2+ 3 − + − , , (cit , Na ) (Ca , ClO ) 0.27 ε = – (0.15 ± 0.03) + (0.13 ± 0.03)log ε= I m 10 4 222 32 23 30 z ∆=×+×= , n = and 4 . ≤ 0.5 m, a SIT analysis neglecting the Since the accepted data are limited to I m terms has been conducted in this review. This resulted in the se- na log and ∆ε 10 H O 2 2+ citrate at 25 ° C. lected constant for the solubility product of Ca ο ± log K 0.10) (VII.18) = − (17.90 ,0 10 s Accordingly, the selected value for the Gibbs energy of formation is: –1 ο (cit) ∆ (Ca . G 5.1) kJ·mol ·4H ± O, cr, 298.15 K) = – (5033.7 2 3 2 fm The solid lines marked (1) and (2) in Figure VII-18 are the calculated values using this value in Equation (VII.19) including and neglecting ∆ε and na log 10 H O 2 terms, respectively. Since the data in [2001CIA/TOM] were obtained by assuming a 3 −+ (cit , Na ) ε , the points at higher ionic strengths do not agree with different value of the estimation by using the value selected by this review.

395 VII.7 Nickel citrate com pounds and compl 353 exes VII.6 Selenium citrate compounds and complexes No experimental data could be found for the selenium citrate system. VII.7 Nickel citrate compounds and complexes No thermodynamic data for Ni citrate compounds could be identified in this review (see also Section VII.1.1). 2+ The experimental equilibrium data found on the complex formation of Ni tiometric titrations were used in most stud- with citrate are listed in Table VII-18. Poten which are rejected by this re- ies. Except for the species proposed by Sari [2001SAR] + , Ni(Hcit)(aq) and cit) view (see comments in Appendix A), the formation of Ni(H 2 − in the lower pH region is commonly accepted in the literature. When the ratio of Ni(cit) 4 − N can be detected [80HED/LID] . In i(cit) citrate to nickel is higher, the formation of 2 − − ) is in excess of 3 C , Ni(cit) is the higher pH region where the amount of base (OH Hcit 3 , [45BOB/JOR] reported to react with one equivalent of base in many papers ( , [58HEI/FRI] , [58MIG/SYC] , [65PAT/PAN2] , [70BES/CHA] , [57PAT/PAN2] , [80STI/WIK] , [84SAL/DEV] ), suggesting that the alcoholic OH of the [76DAN/OST] 3 − − is much easier to be deprotonat ed as compared to that in cit . citrate ligand in Ni(cit) , [58MIG/SYC] , [57PAT/PAN2] , [58HEI/FRI] , Many papers ( [45BOB/JOR] [70BES/CHA] [84SAL/DEV] , [89ISH/ENO] ) consider the formation , , [65PAT/PAN2] 2 − 4 4 − − cit) i (cit) (OH) N ) by i(H cit) N (or . Other proposed species are of Ni(H 1 − − 22 2 212 − 5 . At the moment, there is no N i (OH)(H cit) and by [80STI/WIK] [76DAN/OST] − 413 that is, no examinations have been done well-grounded argument about the speciation, for the speciation other than the potentiometric titration at a certain fixed concentrations 2+ and citrate. Moreover, the experimental da ta are considered to be affected by the of Ni 2+ since the formation of the proposed species proceeds in the hydrolysis reactions of Ni . This is illustrated in Figure VII-19 which pH region where hydrolysis of nickel occurs – , cit) shows the simulated titration curve considering both the formation of Ni(H 2 2+ – − 4 , but excluding any other spe- i(cit) , and the hydrolysis of Ni N Ni(Hcit)(aq), Ni(cit) 2 cies. It is clear that the data should be an alysed considering the simultaneous hydrolysis 2+ . This review could not judge the exact species nor its formation con- reactions of Ni stant.

396 VII Discussion of data selection for citrate compounds and compl exes 354 2+ Table VII-18: Literature data on the formation constants for citrate complexes of Ni . a ( β ° t Reference Method Ionic medium C) log 10 − 2+ + (VII.20) U cit + H cit) Ni Ni(H 2 2 pot 0.1 M NaClO 20 1.75 [64CAM/OST] 4 0.04) 25 (?) (1.55 ± dis [72KER/CHU] 0.5 M NaNO 3 pot 1 M NaClO [78KER/CHU] 25 1.45 4 25 (1.50 ± 0.16) [80HED/LID] pot 0.1 M KNO 3 2+ + + U + H cit) (VII.21) + H cit Ni(H Ni 2 3 zone electrophoresis 0.7 M KNO 25 (?) − 1.4 3 [70BES/CHA] 2+ − Ni cit + 2H U Ni(H (VII.22) cit) (aq) 2 2 2 [84SAL/DEV] 1.8 NMR (4.989 0.105) [2001SAR] pot 25 ± 0.1 M NaCl + 3 − 2+ + cit U Ni(Hcit)(aq) (VII.23) + H Ni 25 (8.75 pot 0.1 KNO 0.01) [80STI/WIK] ± 3 b pot 0.25 KNO (?) 25 (8.75 ± 0.05) [84DAN/OST] 3 b (?) 25 (9.13 ± 0.06) pot 0.1 KNO [88DAN/OST] 3 2 − 2+ Hcit + U Ni(Hcit)(aq) Ni (VII.24) pot 0.25 M KNO 32.5 3.37 [57PAT/PAN2] 3 − 2+ , pot 0.15 M (Ni ) 25 (3.19 ± 0.15) [59LI/LIN] NO 3 3.30 [64CAM/OST] 20 0.1 M NaClO pot 4 0.5 M NaNO dis ± 0.04) [72KER/CHU] 25 (?) (2.90 3 pot 0.1 M KNO ± 0.07) [75FIE/COB] 25 (3.34 3 25 (3.23 ± 0.03) [76DAN/OST] 0.1 M KNO pot 3 1 M NaClO pot 25 2.90 [78KER/CHU] 4 0.1 M KCl 25 (3.35 pot 0.06) ± [80HED/LID] NMR 2.6 [84SAL/DEV] − − 2 2 Ni(OH) U Ni(H cit) (VII.25) (s) + Hcit + 2 H O 2 − 2 1 sol 35 3.21 [65PAT/PAN2] − 2 2 2+ − + 2Hcit Ni(Hcit) Ni U (VII.26) 2 NMR [84SAL/DEV] 4.3 0.1 M NaCl 25 (6.721 ± 0.04) [2001SAR] pot 2+ 3 − − U Ni(cit) + cit (VII.27) Ni amperometric titration 25 3.39 [45BOB/JOR] pot 0.25 M KNO 32.5 6.12 [57PAT/PAN2] 3 0.12) ± [58HEI/FRI] pot, cond, sp 25 (?) (4.40 (4.54 0.08) ± 25 4.99 [58MIG/SYC] pot 2 M KNO 3 2+ − [59LI/LIN] NO , ) 25 (5.11 ± 0.15) pot 0.15 M (Ni 3 pol 0.1 M NaCl r. t. 4.05 [61ISH/YOK] r. t. 3.20, 3.62 sp (Continued no next page)

397 VII.7 Nickel citrate com pounds and compl 355 exes Table VII-18: (continued) a Method Ionic medium t β ° C) log Reference ( 10 0.1 M NaClO 5.40 [64CAM/OST] pot 20 4 ± 0.12) [72KER/CHU] 0.5 M NaNO dis 25 (?) (4.25 3 0.1 M KNO pot 25 (5.40 ± 0.04) [75FIE/COB] 3 25 (5.30 ± 0.02) [76DAN/OST] 0.1 M KNO pot 3 1 M NaClO pot 25 4.30 [78KER/CHU] 4 pot 0.1 M KNO ± 0.04) [80HED/LID] 25 (5.49 3 b (?) 25 (5.09 0.05) [84DAN/OST] ± pot 0.25 KNO 3 NMR [84SAL/DEV] 4.6 b pot 0.1 KNO 0.06) [88DAN/OST] ± (?) 25 (5.35 3 0.1 M KNO 25 (4.54 ± pot [93AZA/HAS] 0.04) 3 0.3 - 5 m NaCl dis I ± 0.10) [2000BOR/CHO] (6.82 → = 0 20 − − cit (VII.28) (s) + H Ni(cit) + 2H O U Ni(OH) 2 2 2 sol 35 6.835 [65PAT/PAN2] − 4 3 − 2+ + 2cit Ni(cit) Ni U (VII.29) 2 pot 25 7.76 [58MIG/SYC] 2 M KNO 3 0.1 M KNO [80HED/LID] 25 (7.82 ± 0.20) pot 3 b pot 0.25 KNO (7.93 ± 0.09) [84DAN/OST] (?) 25 3 NMR [84SAL/DEV] 8.0 b ± 0.06) [88DAN/OST] (?) 25 (8.11 pot 0.1 KNO 3 pot 0.1 M NaCl 25 (7.747 ± 0.015) [2001SAR] 4 − − − 3 Ni(cit) Ni(cit) U (VII.30) + cit 2 [80STI/WIK] 25 (2.85 ± 0.02) pot, sp 0.1 M KNO 3 7 − 2+ 3 − + 3cit Ni U (VII.31) Ni(cit) 3 8.8 [84SAL/DEV] NMR 2+ 4 − 2 − + H U Ni(H (VII.32) cit) cit Ni 1 − 1 − 25 (?) 11.22 [58HEI/FRI] pot, cond, sp 2 M KNO 25 5.27 [58MIG/SYC] pot 3 − + 2 − Ni(H (VII.33) cit) U + H Ni(cit) 1 − 0.25 M KNO 7.87 [57PAT/PAN2] 32.5 pot − 3 7.89 − 35 sol [65PAT/PAN2] + + − 2 Ni(H cit) Ni(H U + 3H (VII.34) cit) 1 − 2 zone electrophoresis 0.7 M KNO 25 (?) − 12.9 [70BES/CHA] 3 3 − 2+ + 3 − Ni + 2cit (VII.35) NiH(cit) + H U 2 b (?) 25 (12.94 ± 0.11) [84DAN/OST] pot 0.25 KNO 3 b 0.2) (?) 25 (13.5 ± pot 0.1 KNO [88DAN/OST] 3 0.028) 0.1 M NaCl 25 (13.549 ± pot [2001SAR] (Continued on next page)

398 VII Discussion of data selection for citrate compounds and compl exes 356 Table VII-18: (continued) a Method Ionic medium β ( ° t C) log Reference 10 − + 3 − 2+ Ni NiH (cit) + 2cit U (VII.36) + 3H 32 0.1 M NaCl (21.55 ± 0.06) [2001SAR] pot 25 − 2+ 3 5 − + cit)(cit) (VII.37) + 2cit U Ni(H + H Ni 1 − 0.1 M NaCl 25 − (2.34 ± 0.01) [2001SAR] pot 6 − 2+ − 3 + Ni(H cit) Ni U + 2H + 2cit (VII.38) − 12 b pot 0.25 KNO (?) 25 – (4.76 ± 0.09) [84DAN/OST] 3 b (?) 25 – (4.64 ± 0.06) [88DAN/OST] pot 0.1 KNO 3 0.1 M NaCl − (13.55 ± 0.02) [2001SAR] 25 pot 5 − 2+ 3 − + (VII.39) O + 3cit + 4H Ni (OH)(H cit) + H U 4Ni 2 413 − 0.1 M KNO − 7.1 [80STI/WIK] pot, sp 25 3 − − 4 − 3 2+ 3 cit 2Ni (cit)(H U Ni + H (VII.40) cit) + cit − 1 1 − 2 14.8 [84SAL/DEV] NMR 4 − 3 − + 2+ Ni (H cit) + 2cit U + 2H (VII.41) 2Ni − 212 pot 0.1 M KNO 25 − (4.71 ± 0.03) [76DAN/OST] 3 4 − 2+ (VII.42) cit + 2H U Ni (H cit) 2Ni 1 − − 212 (17.7 ± 0.02) [84SAL/DEV] NMR a: Refers to the reactions indicated, the ionic strength and temperature given in the table. Uncertainties are those given in the references. b: The values at 10, 35, 45°C and H ∆ are also reported but not listed in this table. rm r.t.: room temperature Figure VII-19 and Figure VII-20 show the simulated titration curve and the dis- tribution of the species in the titration calculated by this review using the values of the (cr) recommended by Plya- hydrolysis constants and solubility products of Ni(OH) 2 . The results indicate that the precipitation of Ni(OH) (cr) sunova et al . [98PLY/ZHA] 2 C 3 ) exceeds . This review consid- C occurs immediately after the amount of base ( OH Hcit 3 3 CC ≥ and the assigned stability constants for the rele- ers that the speciation at OH H cit 3 vant species are affected by the slowly progressing precipitation. Thus, this review does not consider these species and values.

399 VII.7 Nickel citrate com exes pounds and compl 357 − − 3 3 10 10 M citric acid, (2) 5 × Figure VII-19: Simulated titration curves of (1) 5 × M cit- 3 − 2+ I M Ni = 0.1 M, with the following assumptions: at 10 ric acid and 5 × rr 43 +− − For Hcit HHcit + U , − rr 1 =3) = 2.92 log K ( r =1) = 5.69, r ( log K ( r =2) = 4.36, K log 3 10 1 10 2 10 For Ni-citrate ccomplexation: + log ((VII.20), Ni(H ) = 1.63, b log b ((VII.24), cit) 2 1 10 10 1 − log b ((VII.27), Ni(cit) ) = 5.46, Ni(Hcit)(aq)) = 3.32, 1 10 4 − b ((VII.29), ) = 7.82 log N i(cit) 1 10 2 2 − mn 2 ++ For i N mn Ni (OH) n H O H ++ U [98PLY/ZHA] 2 mn + − log K (Ni(OH) ) = (cr)) = − 10.52, 9.50, (NiOH log K 2 1,1 2,1 10 10 − ( N i(OH) (aq)) = − 18.0, log K ( K N i(OH) log ) = − 29.7, 3 3,1 2,1 2 10 10 − 2 + 3 N ) = 9.8, i(OH) − log − 44.96, K iOH log K ( ( N ) = 10 4,1 10 1,2 4 2 4 + i (OH) 27.9 ( − N log K ) = 10 4,4 44 12 3 − 11 I = 5·10 C M, = 0.1 M (1) H cit 3 3 − = (1) 0, (2) 5·10 M C Ni 10 9 8 ] + (2) 7 [H 10 6 log − 5 4 3 2 0123456 per C C added cit H NaOH 3

400 VII Discussion of data selection for citrate compounds and compl exes 358 − 3 × Figure VII-20: Distribution of complex species in the simulated titration of 5 M 10 2+ − 3 = 0.1 M (shown in Figure VII-19). Concentration of I M Ni at citric acid and 5 × 10 (cr) is expressed by the amount of precipitate divided by the volume of the Ni(OH) 2 solution. -2 - Ni(cit) (cr) Ni(OH) 2 2+ Ni -4 Ni(Hcit) 4- Ni(cit) -6 2 + Ni(H cit) 2 C 10 -8 -log + NiOH -10 Ni(OH) 3 Ni(OH) 2 -12 34567891011 + -log [H ] 10 Because of various shortcomings in the experimental procedures or in the re- porting of the results, the values of [45BOB/JOR] , [57PAT/PAN2] , [58HEI/FRI] , , [59LI/LIN] , [61ISH/YOK] , [65PAT/PAN2] , [70BES/CHA] , [58MIG/SYC] [80STI/WIK] , [84SAL/DEV] , [93AZA/HAS] , [2001SAR] have been , [72KER/CHU] rejected by this review ( cf. Appendix A). Based on the discussion of the remaining literature studies (see Appendix A) the constants listed in Table VII-19 are accepted in this review, where the data at 20 ° C and 25 ° C are found to be identical within the limit of the given uncertainties.

401 VII.7 Nickel citrate com exes pounds and compl 359 2+ at 25 ° Table VII-19: Accepted formation constants for citrate complexes of Ni C used to derive the selected values. a Ionic medium Reference log β 10 − 2+ + + H U Ni cit cit) (VII.20) Ni(H 2 2 0.1 M NaClO (1.75 ± 0.30) [64CAM/OST] 4 (1.50 0.1 M KNO 0.32) [80HED/LID] ± 3 ± 0.40) [78KER/CHU] (1.45 1 M NaClO 4 − 2+ 2 + Hcit U Ni(Hcit)(aq) (VII.24) Ni (3.30 ± 0.1 M NaClO [64CAM/OST] 0.20) 4 0.1 M KNO (3.34 ± 0.20) [75FIE/COB] 3 (3.23 ± 0.20) [76DAN/OST] 0.1 M KNO 3 0.1 M KCl (3.35 [80HED/LID] 0.12) ± 1 M NaClO ± [78KER/CHU] (2.90 0.40) 4 − 3 − 2+ Ni + cit (VII.27) U Ni(cit) I → (6.82 ± 0.10) [2000BOR/CHO] 0.3 - 5 m NaCl = 0 (5.40 0.1 M NaClO ± 0.20) [64CAM/OST] 4 ± 0.10) [75FIE/COB] (5.40 0.1 M KNO 3 0.1 M KNO ± 0.20) [76DAN/OST] (5.30 3 0.1 M KNO (5.49 ± 0.08) [80HED/LID] 3 (4.30 ± 0.40) [78KER/CHU] 1 M NaClO 4 4 − 2+ 3 − Ni + 2cit U (VII.29) Ni(cit) 2 (7.82 ± [80HED/LID] 0.1 M KNO 0.40) 3 a: Refers to the reactions indicated, the ionic strength and temperature given in the table. Uncertainties are estimated in this review. I = 0.10 M (= 0.10 In Table VII-19, the estimated uncertainties of the values at m) in NaClO ∆ε I and KNO . Thus, SIT analyses were media are larger than the term m 3 4 conducted based on the assumption that the data in Table VII-19 are identical to those in + cit) and Ni(Hcit)(aq), the data were fitted to the media. In the cases of Ni(H NaClO 2 4 following equation, ο 2 z DI log −∆ = −∆ε bb (VII.43) log m 1 10 1 10 2 +− +− + − 2 where (Ni(H cit) , ClO ) (Ni , ClO ) (Na , H cit ) and ∆ε=ε −ε −ε =− 4 z ∆ 24 2 4 22 +− −+ 2 (Hcit , Na ) (Ni ,ClO ) ∆=− ∆ε = − ε −ε + ε (Ni(Hcit)(aq),NaClO z ) and 8 and 4 4 respectively. − In the case of Ni(cit) , data were fitted to the following equation: 23 +− ο = log ( Na , cit ) log ε (VII.44) −∆ +ε z DI I bb −∆ ' mm 1 10 1 10 −+− +2 (Ni , ClO ) ∆ε = ε −ε (Na , Ni(cit) ) ' where 4 − 1 −+ 3 2 – (0.15 ± . ± 0.03)log z I 12 (kg·mol (cit , Na ) ∆=− ) and ε= 0.03) + (0.13 m 10

402 exes VII Discussion of data selection for citrate compounds and compl 360 The results of the fittings are given in Figure VII-21. For the formation of − 4 ο , the value of N log i(cit) b was obtained by simply neglecting the ∆ε I term at low m 10 2 ionic strength of 0.1 M and a larger uncertainty was given. Table VII-20 lists the equi- librium constants selected in this review for the nickel-citrate system. 2+ b log − citrate system at each ionic for the Ni Figure VII-21: Fitting of the values of 10 strength (given in Table VII-19) to equations (VII.43) and (VII.44). Solid lines are drawn by using the result of the fitting (given in Table VII-20). I 0.5091 m ο 2 bb DI = =+∆−∆ε log where log z D Solid lines : 10 1 10 1 m I + 11.5 m − 1 ο 2 (VII.20) = (2.05 ± 0.25), z ∆ = log 4, (0.2 0.5) ∆ε = − ± kg · mol − b 1 10 − 1 ο 2 (VII.24) = (4.16 ± 0.10), log z ∆ = 8, (0.4 0.5) ∆ε = − ± kg · mol − b 10 1 ο 2 3 − + (VII.27) = (6.76 ± 0.08), ∆ε = ∆ε − ε z ∆ = − 12, log '(cit,Na) b , 1 10 − 1 , ± kg · mol ' (0.05 0.41) ∆ε = − − 1 −+ 3 (cit , Na ) 0.03)log I – (0.15 kg · mol ε= 0.03) + (0.13 ± ± m 10 8 [64CAM/OST] [75FIE/COB] 7 [76DAN/OST] [80HED/LID] [78KER/CHU] 6 − [2000BOR/CHO] for Ni(cit) β log 10 1 5 1 β 4 10 β log for Ni(Hcit)(aq) 10 1 log 3 + for Ni(H β cit) log 2 10 1 2 1 0.4 1.2 0.6 0.2 0.0 1.0 0.8 / m I

403 VII.7 Nickel citrate com pounds and compl exes 361 2+ at 25 ° Table VII-20: Selected formation constants for the citrate complexes of Ni C. ο K Reaction log 10 2+ − + + H U Ni(H Ni cit) cit (2.05 ± 0.25) 2 2 2 − 2+ + Hcit U Ni (4.16 ± 0.10) Ni(Hcit)(aq) 2+ 3 − − Ni U (6.76 ± 0.08) + cit Ni(cit) − 4 2+ − 3 Ni(cit) U (8.5 ± 0.4) + 2cit Ni 2 and ∆ε ’ given in the caption of Figure VII-21 From the obtained values of ∆ε 1 − 2 +− and using the values of = (0.370 0.032) kg·mol (Ni ,ClO ) [2005GAM/BUG] , ε ± 4 1 − 1 − − + 2 −+ − ε , = (Hcit , Na ) (H cit , Na ) (0.05 ε = − (0.04 ± 0.02) kg·mol ± 0.01) kg·mol 2 (obtained in this review), the specific ion interaction parameters for the Ni-citrate com- 1 − + − ε (Ni(H cit) ,ClO ) = (0.12 ± 0.5) kg·mol , plexes can be estimated to be 24 − 1 − + = (0.22 0.5) kg·mol ε − and (Ni(cit) , Na ) (0.07 ± (Ni(Hcit)(aq), NaClO ) = ± 0.5) ε 4 − 1 . It is not possible to evaluate the reliability of these values due to their large kg·mol uncertainties . The above selections yield: + –1 ο ± cit) (Ni(H , 298.15 K) = – (1283.3 ∆ 2.6) kJ·mol G 2 fm –1 ο ∆ (Ni(Hcit), aq, 298.15 K) = – (1268.1 ± 2.2) kJ·mol G fm – –1 ο ∆ (Ni(cit) , 298.15 K) = – (1246.6 ± G 2.2) kJ·mol fm –1 ο 4 − . G N i(cit) ∆ , 298.15 K) = – (2418.8 ± 4.7) kJ·mol ( 2 fm VII.8 Technetium citrate compounds and complexes Despite the various oxidation states of technetium, only a limited number of papers ), all published by the same author, , [77MUN2] , [77MUN/GRO] , [81MUE] [77MUN] ( dealing with the complex formation of technetium(IV) with citrate could be found. Al- suggests a very strong Tc citrate complex formation which sup- [77MUN/GRO] though presses the precipitation of technetium hydroxides similarly to other tetravalent metal ), U(IV) and Th(IV) ( [60ADA/SMI] , [78STR/KAR] ions such as Zr(IV) ( , [74SHA/KII] ), the stoichiometries of the species formed are not well [66NEB/URB] in Appendix A). No thermodynamic data [77MUN/GRO] established (see discussion of ). [81MUE] , [77MUN2] , [77MUN/GRO] , [77MUN] are reported by Münze ( VII.9 Zirconium citrate compounds and complexes No thermodynamic data for Zr citrate compounds could be identified in this review (see also Section VII.1.1).

404 VII Discussion of data selection for citrate compounds and compl exes 362 As shown in Table VII-21, there is only a limited number of papers dealing with the complex formation of Zr(IV) with citrate. Altho ugh these papers suggest a very strong complex formation which suppresses the precipitation of zirconium hydroxides up to a fairly high pH, similarly to other tetravalent metal ions such as Tc(IV) ), [60ADA/SMI] , [66NEB/URB] , [74SHA/KII] ), U(IV) and Th(IV) ( ( [77MUN/GRO] even the stoichiometries of the species formed are not well established. Among the pa- , [75ZAI/NIK] ) claim the forma- [60RYA/ERM] pers listed in Table VII-21, two papers ( 3+ 2+ , one ( [66KOR/SHE2] ) claims the formation of Zr(Hcit) cit) , and the tion of Zr(H 2 + − 2 , Zr(cit) ) claims the formation of Zr(cit) and the species com- [78STR/KAR] other ( 2 4– 4+ cit . Due to the strong tendency of Zr to be hydrolysed and the strong plexed with H –1 interaction with the ligand, the complex formation of Zr(IV) with citrate can be studied mainly in highly acidic solutions, where the activity coefficients of the relevant species are very difficult to be estimated. Unfortunately, the conditions studied in the papers listed in Table VII-21 are limited and their speciation cannot be relied on. So far, the present review cannot select any value for the complex formation of Zr(IV) with citrate. 4+ . Table VII-21: Literature data on the formation constants for citrate complexes of Zr a ( Method Ionic medium C) K t Reference ° 4+ 3+ + cit(aq) U Zr(H cit) + H + H Zr 2 3 b + ] = (1300 ± 216) cix 2 M HClO [60RYA/ERM] , ? K/[H 4 [64RYA/MAR] [75ZAI/NIK] 25 (310 , 0.1 – 2 M HNO dis 20) 2 M LiNO ± 3 3 (121 9) ± 4 M LiNO , 0.1 – 2 M HNO 3 3 6 M LiNO (90 ± 6) , 0.1 – 2 M HNO 3 3 2+ 4+ − 2 + Hcit U Zr(Hcit) Zr 10 1 M HCl 20 (6.0 ± 0.5) × 10 sp [66KOR/SHE2] 18 − − 3 + 4+ Zr(cit) + cit U Zr 6 × 10 (c) not controlled (at pH 0) [68KOZ] ? 1.4 8 [78STR/KAR] ? NMR not controlled (at pH 3.0) × 0.7 10 8 1 (at pH 3.1) 10 × 8 0.4 (at pH 3.2) 10 × 8 1 (at pH 3.4) × 10 2 − 4+ 3 − U Zr(cit) + 2cit Zr 2 14 [78STR/KAR] NMR not controlled 1 × (at pH 3.0) 10 14 6 10 (at pH 3.1) × ? 14 1 10 (at pH 3.2) × 14 1 × (at pH 3.4) 10 a: Refers to the reactions indicated, the ionic strength and temperature given in the table. Uncertainties are those given in the references. b: This is a conditional constant given at 2 M HClO . . See the discussion for [60RYA/ERM] 4 . c: Metal indicator method [68KOZ]

405 VII.10 Uranium citrate compounds and compl exes 363 VII.10 Uranium citrate compounds and complexes No thermodynamic data for uranium citrate compounds could be identified in this re- view (see also Section VII.1.1). VII.10.1 U(IV) citrate complexes Concerning citrate complexes of the tetravalent uranium, there is a limited number of papers (Table VII-22). Similarly to other tetravalent metal ions such as Tc(IV) ) and Zr(IV) ( ), the papers listed in Table VII-22 suggest [78STR/KAR] ( [77MUN/GRO] a very strong complex formation of U(IV) with citrate which suppresses the precipita- tion of hydroxides up to a fairly high pH. However, the stoichiometries of the species 4+ to hydrolyse and too formed are not well established due to the strong tendency of U far, the present review cannot select any strong complex formation with citrate. So value for the complex formation of U(IV) with citrate. 4+ Table VII-22: Literature data on the formation constants for citrate complexes of U . a ( ° C) log Method Ionic medium K Reference t 10 4+ − 3 + + cit Ucit U U 0.1 M (H, Na)ClO r.t. 9.72 (at pH 1.50) to sp [60ADA/SMI] 4 b 3.93 (at pH 4.50) 0.5 M HNO ? 11.53 pot [66NEB/URB] 3 2 − 4+ 3 − + 2cit U U(cit) U 2 pot 0.5 M HNO 19.46 [66NEB/URB] 3 4+ + 2+ U(Hcit) U + 2H cit U + H 3 22 0.85 pot ± 0.03 [74SHA/KII] 4 M (H, Na)ClO 4 a: Refers to the reactions indicated, the ionic strength and temperature given in the table. Uncertainties are those given in the references. – K on pH, the actual complex was considered as U(OH) . cit b: From the dependence of log 2 10 r. t.: Room temperature. VII.10.2 U(VI) citrate complexes − 2 − − 3 2+ UO ), H , Hcit may form complexes. How- and cit With hexavalent uranium ( cit 2 2 2+ UO has a linear dioxo structure and the ligand should coordinate in the equato- ever, 2 rial plane perpendicular to the O − U − O axis, and this sets some constraints on the stoichiometry of the complexes due to the steric hindrance. Because of this, only two carboxylate groups may be simultaneously coordinated to the central uranium, resulting in the strong interaction of the alcoholic hydroxyl and the uranium. Thus, it is to be ex- − cit may not be much larger than that of UO (Hcit)(aq). pected that the stability of UO 2 2 3– as ligand may form polynuclear complexes Furthermore, the species containing cit bridged by non-bonded carboxylate groups. A summary of the experimental data available for the citrate complexes of 2+ UO is presented in Table VII-23. 2

406 VII Discussion of data selection for citrate compounds and compl exes 364 Table VII-23: Literature data on the formation constants for the citrate complexes of 2+ UO . 2 a Method Ionic medium ° β Reference C) log t ( 10 2+ 3 − − U + cit UO (cit) UO 2 2 ? 3.165 sp not controlled [54HEI/BOB] 25 [59LI/LIN] 0.15 M NaCl pot 8.5 pot 0.1 M KNO ± [65RAJ/MAR] 25 (7.40 0.21) 3 1.0 M KNO 25 (6.87 0.11) ± 3 1.0 M KNO pot 25 5.78 [72MAR/KLO] 3 cix 0.1 M Na(H [75OHY/ODA] cit) 25 7.22 2 1 M (H,Na, (UO sp ) ) cit 25 6.20 [80VAN/KUC] 2 0.5 3 25 (7.17 ± 0.16) 0.05 M NaClO sp [89ABD/ALI] 20 3.93 4 dis 5 NaCl (6.04 ± 0.01) [96BOR/LIS] m 25 dis 0.3 25 (7.30 ± m [99BRO/POK] NaCl 0.04) 1.0 m NaCl 25 (7.08 ± 0.01) 2.0 NaCl 25 (7.22 ± m 0.02) 3.0 NaCl 25 ± 0.01) m (7.10 4.0 NaCl m (7.02 ± 0.01) 25 5.0 NaCl m (7.03 ± 0.02) 25 cix 0.1 M NaClO 21 (6.69 ± 0.03) [2000LEN/CAB] 4 2+ − 4 − 3 + 2cit UO U UO (cit) 2 22 dis 5 NaCl ± 0.03) [96BOR/LIS] (10.98 25 m − 2 2+ 3 − + 2cit U UO 2 (UO ) (cit) 2 2 22 b 25 19.26 pot 0.136 M KNO [60FEL/NOR] 3 pot 25 (18.87 ± 0.06) [65RAJ/MAR] 0.1 M KNO 3 1.0 M KNO 25 (17.70 ± 0.04) 3 0.15 M NaCl 25 18.90 [59LI/LIN] pot pot 1.0 M KNO 25 17.57 [72MAR/KLO] 3 ) ) cit 25 15.25 [80VAN/KUC] 1 M (H,Na, (UO sp 0.5 3 2 25 (17.00 ± 0.14) 2+ 5 − 3 − 2 U + 3 cit UO (UO ) (cit) 2 3 22 [54HEI/BOB] sp not controlled ? 6.095 2 − − 2 UO U (cit) (UO ) (cit) 2 2 22 dis 0.15 M NaNO 20 ≥ 6.0 [70ADI/KLO] 3 2+ 2 − (Hcit)(aq) UO U UO + Hcit 2 2 cix 0.1 M Na(H cit) 25 4.23 [75OHY/ODA] , 2 [75OHY] sp 1 M (H,Na, (UO ) [80VAN/KUC] ) cit 25 3.82 3 2 0.5 25 (4.56 ± 0.26) (Continued on next page)

407 VII.10 Uranium citrate compounds and compl exes 365 Table VII-23 (continued) a β Method Ionic medium C) log Reference ( t ° 10 2 − 2+ 2 − + 2Hcit UO (Hcit) U UO 2 22 11.2 [62STA/BAL] ? ? 25 ≅ 2+ − + UO + H U UO (H cit cit) 2 2 2 2 0.1 M Na(H [75OHY/ODA] cix cit) 25 2.79 2 ) ) (cit) 25 1.53 [80VAN/KUC] sp 1 M (H,Na, (UO 3 2 0.5 2+ 2 + − 2 cit UO U + 2H (UO ) (H cit) 2 2 22 2 2 1 M (H,Na, (UO sp (cit) 25 8.89 [80VAN/KUC] ) ) 0.5 2 3 z 6 − 2+ z+ + − UO M(H cit) cit + 2H U + M UO + 8H 3 − 2 212 M = Fe(III) 2.45 [89MAN/APE] pot 25 M = Al(III) 25 8.21 M = In(III) 25 11.30 M = Cu(II) 25 − 1.41 a: Refers to the reactions indicated, the ionic strength and temperature given in the table. Uncertainties are those given in the references. . [60FEL/NOR] b: Recalculated from the reported value. See the discussion on Among the publications on the determination of equilibrium constants for 2+ , [60FEL/NOR] UO citrate system, four papers, [65RAJ/MAR] and , [59LI/LIN] 2 2+ , report potentiometric titrations of solutions containing UO and [72MAR/KLO] 2 cit. The titration curves exhibit a first and a second inflection point at 3 and 3 + 5/3 H 3 2+ UO . Mass balance equations for these titrations are: moles of base per mole of 2 C , [H L] [H L] [HL] [L] [ML] [MHL] [MH L] 2[M L ] =+++++ + + 22 2 L3 2 [M] [ML] 2[M L ] [MHL] [MH L] Cp =+ + + + + , [M (OH) ] ∑ p q M222 + − −= + + + + 3 3[H L] 2[H L] [HL] [MHL] 2[MH L] + CC q [M (OH) ] [H ] ∑ 2 3 pq 2 LOH 3 − 2+ where L and M denote cit UO , respectively, and charges are omitted for brevity. and 2 2+ C UO and and C are the total added concentrations of citric acid (H , cit), C 3 OH M L 2 NaOH, respectively. The formation constants for the complexes are expressed by: + r [H ] [L] = b , [H L] (H) rr p q , b [M] [H L] = [M (H L) ] r prq qp , p [M] * . b = [M (OH) ] pq qp (OH), q + [H ] C If , C and C and the formation constants are known, these equations can L OH M be solved to give the equilibrium concentrations of all species. 2+ reported that equimolar concentrations of UO Although Li et al. [59LI/LIN] 2

408 VII Discussion of data selection for citrate compounds and compl exes 366 − , Feldman et al. and citric acid react to form only a mononuclear chelate UO (cit) 2 showed that their data could be interpreted in terms of the formation of [60FEL/NOR] − 2 , (UO ) (cit) . Posterior and more elaborate studies [60FEL/NOR] the dimer 22 2 and [72MAR/KLO] also revealed that the predominating species should [65RAJ/MAR] − 2 (UO ) (cit) under these conditions. This is also supported by the obser- be the dimer 22 2 , [72MAR/KLO] of the similar formation of the dimer [60FEL/NOR] vation 3 − 2 − where L is cit , malate and the mixed-metal complex UO (UO ) (cit) InL 2 2 22 2 2– 2– ) and tartrate (OOCCH(OH)CH(OH)COO CH(OH)COO ). Assuming the (OOCCH 2 − 2 − , the above equations and the dimer (UO ) (cit) cit formation of the monomer UO 2 22 2 are simplified to: [H L] [H L] [HL] [L] [ML] 2[M L ] C =+++++ , L3 2 22 C , [M] [ML] 2[M L ] =+ + M22 + CC −= + + + . 33[HL]2[HL][HL][H] 3 2 LOH + ′ [H L], [H C and [H ]: L], [HL] and [L] can be expressed in terms of 2 3 L ′ C =α , [H L] L rr r + [H ] b r (H) ′ b ( 1) and C α == =+++ [H L] [H L] [HL] [L] where: 0(H) L3 2 r 3 r + [H ] b ∑ (H) r 0 = r − C can be transformed into: C Since the equation for 3 L OH + ′ CC C −− =α+α+α , 3[H](32) 3 21L LOH ′ C can be estimated from the measured pH, which in turn enables to estimate [M] and L [L] by: ′ [L] =α , C L0 ′ ( ) . [M] =−− CCC MLL Therefore, if the monomer ML is the predominant species in the titration proc- ess, the following β for the formation of ML is obtained as a constant. 1 ′ CC − [ML] LL == . b 1 ′′ [M] [L] ⋅ CCCC ()() − +α MLLL0 Alternatively, if the dimer M L is the predominant species, with the following 2 2 equilibrium constant, ′ CC − ) 2 ( [M L ] L 22 L b == 2,2 2 22 2 ′′ ( ) ( ) [M] [L] ⋅−+α CCC C MLL L0 , Rajan log b values were obtained by Feldman et al. [60FEL/NOR] fairly constant 10 2,2 and Markovits et al. [72MAR/KLO] , indicating a good reproduci- et al. [65RAJ/MAR] bility of the potentiometric titration. From the second inflection point in the titration curves, the authors of [60FEL/NOR] and [65RAJ/MAR] inferred the formation of a trimeric species

409 VII.10 Uranium citrate compounds and compl exes 367 − 8 . However, taking into account the possible formation of hydrox- (UO ) (cit) (OH) 3 23 5 2+ UO , this second inflection point can be expected even without citric acid. ides of 2 Figure VII-22 shows the simulated titration curve which is calculated by assuming the formation of hydroxide complexes without precipitation. Although the reported obser- vations, showing no precipitation in this pH region, suggest the formation of com- 4– cit or of mixed hydroxide-citrate complex(es), their formation con- plex(es) with H –1 ecting the simultaneously proceeding hydroly- stant(s) should be estimated without negl sis reactions, since the proposed reaction(s) occurs at the similar pH as shown by Figure VII-23. At present, the trimerization of the citrate complex is not considered as proven by this review. Figure VII-22: Simulated titration curves of 0.01 M citric acid (upper dashed line), 0.01 2+ UO (solid line), and 0.01 M strong acid (lower dashed line) M citric acid and 0.01 M 2 at ed in this review have been used: I = 0.1 M. The protonation constants select 43 +− − rr Hcit HHcit + U . For 1 − rr K ( r log log K ( r =2) = 4.36, log K ( r =3) = 2.92. =1) = 5.70, 10 10 2 1 10 3 2+ 2 ( 3) rpqr − +− 3 For the U(VI)-citrate complexes, + , the pq UO H cit (UO ) (H cit) U 22 rprq equilibrium constants used are: − − 2 (cit) (Hcit)(aq): 4.23. : 19.02, UO (UO ) (cit) : 7.67, UO 2 2 2 22 2 pq 2+ − + For the hydrolysis of U(VI), H , the UO p qq (H O) (UO ) (OH) ++ U 2 22 pq equilibrium constants used are: + 2+ : − 5.43, UO − (aq): − 10.52, OH (UO ) (OH) : 5.81, (OH) UO 2 2 2 2 22 2+ + − (UO ) (OH) : − 12.30 and 16.15. (UO ) (OH) : 5 4 23 23 12 10 8 6 pH cit H 3 4 2+ H cit + UO 3 2 2 strong acid 0 0123456 or c added c c per NaOH H cit strong acid 3

410 VII Discussion of data selection for citrate compounds and compl exes 368 Figure VII-23: Changes in the concentrations of the relevant species in the titration of 2+ UO shown in Figure VII-22. The same equilibrium 0.01 M citric acid and 0.01 M 2 constants were used as in Figure VII-22. -2 − 2 (cit) ) (UO 2 2 2 2+ UO 2 UO (OH) (aq) 2 2 − UO (cit) + 2 -3 (OH) (UO ) 2 5 3 [X] 10 UO (Hcit)(aq) 2 log -4 -5 0123456 per ( c added ) c = c H UO NaOH cit 3 2 As compared to the fact that potentiome try gives a relatively reliable value of − 2 b [75OHY] , ion exchange [99BRO/POK] , solvent extraction log (UO ) (cit) , for 2 22 2,2 10 2+ and spectrophotometry [80VAN/KUC] at lower UO concentrations [75OHY/ODA] 2 − for UO b et al log . Especially, Bronikowski . (cit) give a reliable value of 2 1 10 − b log at different ionic strengths which give values of (cit) for UO [99BRO/POK] 2 1 10 can be used for SIT analysis. Although the stabilities of the complexes with protonated + 2+ (Hcit)(aq), UO (H (UO ) (H cit) are reported, these species and cit) ligands, UO 2 2 2 22 2 2 would be formed in appreciable amounts only at lower pH. Among these species, only (Hcit)(aq) is considered to be plausible in this review. UO 2 Because of various shortcomings in the experimental procedures or in the re- , , [62STA/BAL] , [70ADI/KLO] porting of the results, the values of [54HEI/BOB] , [89MAN/APE] , [2000LEN/CAB] have been rejected by this review ( cf. [89ABD/ALI] Appendix A). . Appendix A), constants Based on the discussion of the remaining literature ( cf listed in Table VII-24 for the following reactions: − − 3 2+ U UO + cit (VII.45) cit UO 2 2 − 3 2 2+ − 2 U (UO ) (cit) UO + 2cit (VII.46) 2 22 2 2 − 2+ UO + Hcit U UO (Hcit)(aq) (VII.47) 2 2 . are accepted to derive the selected values in this review

411 VII.10 Uranium citrate compounds and compl exes 369 2+ Table VII-24: Accepted formation constants for citrate complexes of UO used to de- 2 C. ° rive the selected values at 25 a b Ionic medium log Reference b b log app 10 10 2+ 3 − − U UO + cit cit UO 2 2 0.3) (7.22 ± ± [75OHY/ODA] 0.3) 0.10 M NaH cit (7.22 2 [65RAJ/MAR] ± (7.43 ± 0.5) 0.5) 0.1 M KNO (7.4 3 1.0 M KNO ± 0.5) (7.23 (6.84 0.5) [65RAJ/MAR] ± 3 1 M (H,Na,(UO ) 1.0) (7. 17 ± 1.0) [80VAN/KUC] )(cit) (7. 17 ± 2 0.5 0.3 m NaCl (7.30 0.3) [99BRO/POK] (7.35 ± ± 0.3) (7.07 1.0 m NaCl (7.20 ± 0.3) 0.3) ± (7.20 2.0 m NaCl (7.46 0.3) ± ± 0.3) 3.0 m NaCl (7.07 ± ± 0.3) 0.3) (7.51 (6.98 4.0 m NaCl 0.3) ± ± 0.4) (7.67 5.0 m NaCl (6.98 0.3) (7.98 0.5) ± ± 2+ 2 − − 3 2 (UO ) (cit) U UO + 2cit 2 2 22 (18.87 ± 0.1) (18.90 0.1 M KNO 0.10) [65RAJ/MAR] ± 3 (19.18 ± 0.3) (19.22 ± 0.30) [60FEL/NOR] 0.136 M KNO 3 (18.89 0.15 M NaCl ± 0.2) (18.92 ± 0.20) [59LI/LIN] 1.0 M KNO (17.62 0.1) (17.79 ± 0.19) [65RAJ/MAR] ± 3 1 M (H,Na,(UO ± )(cit) (17.00 ± 1.0) (17.00 ) 1.0) [80VAN/KUC] 0.5 2 2 M KNO (17.45 ± 0.2) (17.78 ± 0.45) [72MAR/KLO] 3 2+ 2 − UO U UO (Hcit)(aq) + Hcit 2 2 0.10 M NaH (cit) (4.23 ± 0.5) (4.23 ± 0.5) [75OHY/ODA] 2 1 M (H,Na,(UO ) [80VAN/KUC] )(cit) (4.56 ± 1.0) (4.56 ± 1.0) 0.5 2 a: Values of log b are converted into molality units. 10 app b: Values of are corrected for chloride or nitrate complex formation. b log 10 log b in mo- In the Table VII-24, the reported values were converted into app. 10 were corrected for the complex formation of log b lality units. Then the values of app. 10 − 2+ − . Grenthe N UO with Cl or select the following formation et al. [92GRE/FUG] O 3 2 + − 2+ ο ο , log ± UO + Cl b U UO = Cl b = (0.17 0.02) for log constants: 2 2 2 10 1 10 − ο 2+ 2+ UO + 2Cl = (0.3 U UO Cl b (aq), and UO + 0.15) for log ± 0.4) for ± (1.1 − 2 2 2 2 1 10 + − − 2+ − 2+ ε (NO (UO ,ClO ) N O with the assumptions = (UO , Cl ) U ε = UO ) 3 2 3 2 24 –1 2+ − + − + − and ε (UO ,NO ) (UO Cl ,Cl ) = (0.46 ε = ± ε (UO Cl , ClO ) 0.03) kg · mol 24 2 23 –1 − 2+ 0.04) kg · mol = (0.33 ± [2001LEM/FUG] ε (Section B.2 in (UO ,NO ) ). From = 23 − 1 − + − + 0.01) kg cf. (Na , NO ) (Na , Cl ) ε = (0.03 ± · mol ε = ( these values, and using 3 [99BRO/POK] were corrected at log b and in b log Appendix B), the values of 2 1 10 10 each ionic strength by the following equation: 2 −− (corrected) = log (reported) + log (1 + [X ] + [X ] ) (VII.48) bb bb log 1,X 2, X 10 10 10 − − − where X is Cl or N O . 3

412 exes VII Discussion of data selection for citrate compounds and compl 370 The magnitude of the correction ( cf. Table VII-24) is not large as compared to − b (VII.45) for UO since (cit) log the estimated uncertainties of the values. For 2 10 1 − +3 (Na , cit ) is dependent on ionic strength, the analysis was conducted in the form of: ε 2 0.5091 ∆ zI m 3 ο +− = bb I I −∆ε (VII.49) −ε − ' (Na , cit ) log log 10 10 mm + I 11.5 m +3 − where (Na , cit ) ± 0.03) + (0.13 ± 0.03) log = – (0.15 I ε , (VII.50) m 10 2+ − +− ε (UO (cit) , Na ) (VII.51) (UO ,Cl ) ε ' ∆= − ε and 22 media, the calculation of the left-hand side of For the data obtained in KNO 3 3 − + − 1 0.02) kg ± ) = (0.02 ⋅ mol , cit in ε equation (VII.49) was carried out with using (K − 3 + , cit ). place of (Na ε The result of fitting to equation (VII.49) is shown in Figure VII-24 and a plot log against I is shown in Figure VII-25. b of m 10 Figure VII-24: Fitting of the values of b (VII.45) at each ionic strength (given in log 10 Table VII-24) to the SIT equation. Solid line is drawn by using the result of the fitting Equation (VII.49). ο 2 3 − + (VII.45) = (8.96 ± 0.17), ∆ε = ∆ε − ε z ∆ = − 12, log '(cit,M) , b 1 10 − 1 , ∆ε = − ' kg·mol ± (0.57 0.08) + 1 3 − − −+ 3 , M ε – (0.15 ± 0.03) + (0.13 ± 0.03) log ) = I kg·mol (cit , Na ) ε (cit m 10 − + 3 − 1 0.02 ) kg , K mol ) = (0.02 ± . ⋅ = ε (cit or 12.5 2+ 3 − − U + cit cit UO UO 2 2 12.0 11.5 m I ) − 3 11.0 , cit + (M 10.5 ε - D 10.0 + 12 1 β 10 9.5 log 9.0 [75OHY/ODA] [65RAJ/MAR] 8.5 [80VAN/KUC] [99BRO/POK] 8.0 0123456 I / m

413 VII.10 Uranium citrate compounds and compl exes 371 I Figure VII-25: A plot log b . The solid line is drawn using the SIT (VII.45) against m 1 10 1 − ο (VII.45) = (8.96 log , and kg·mol ± ∆ε = − (0.57 0.08) ' 0.17), ± b model with: 10 1 − 1 3 −+ (0.15 0.03) (0.13 0.03) log kg·mol I ε=−±+± (cit , Na ) 10 m 9.0 − − 3 2+ cit + cit U UO UO 2 2 8.5 8.0 1 β 7.5 10 log 7.0 [75OHY/ODA] [65RAJ/MAR] 6.5 [80VAN/KUC] [99BRO/POK] 6.0 0123456 / m I The fitting results in the following values: –1 ο log . b (VII.45) = (8.96 ± 0.17) and ' ∆ ε = − (0.57 ± 0.08) kg · mol 1 10 –1 2+ − ' From the obtained ∆ε and using ε ± 0.03) kg · mol (UO ,Cl ) , it is = (0.46 2 –1 − + (0.11 equal to: − . Due to ± 0.09) kg · mol ε (UO (cit) ,Na ) possible to estimate that 2 + − (UO (cit) ,Na ) , we cannot judge whether this ε the large uncertainty in the estimated 2 + with a complex for the interaction of Na value is consistent with the other values of ε ο log b (VII.45) , cf. Table − 1. Thus, this review selects only the value of with charge 10 VII-25: ο log (VII.45) = (8.96 ± 0.17). b 10 1 2 − For the formation of and UO (UO ) (cit) (Hcit)(aq), the analysis was done 2 2 22 simply using the expression: 2 ∆ 0.5091 zI m ο I ε (VII.52) bb −=−∆ log log 10 10 m I 11.5 + m since the available data are obtained at rather lower ionic strength. The result for − 2 (UO ) (cit) of the fitting of equation (VII.52) is shown in Figure VII-26 and in the 22 2 in Figure VII-27. I log b (VII.46) against form of the plot of m 10

414 VII Discussion of data selection for citrate compounds and compl exes 372 Table cf. (VII.46) at each ionic strength ( b log Figure VII-26: Fitting of the values of 10 VII-24) to the SIT equation (VII.52). The solid line is drawn by using the result of the fitting given below. 1 ο − (VII.46) = (21.34 ± 0.5), log (0.8 0.2) ∆ε = − ± kg·mol b . 2,2 10 24 − 2+ 3 − 2 2UO (cit) ) + 2cit (UO U 2 2 2 2 23 D 22 + 22 2,2 β 10 log [65RAJ/MAR] 21 [60FEL/NOR] [59LI/LIN] [80VAN/KUC] [72MAR/KLO] 20 0.00.51.01.52.02.53.0 / m I Figure VII-27: Values of (VII.46) b log versus ionic strength. The solid line is drawn 10 ο (VII.46) = (21.34 ± 0.5), log b by using the result of the fitting given below, 10 2,2 1 − kg·mol ∆ε = − ± (0.8 0.2) 21 − − 3 2+ 2 [65RAJ/MAR] ) (UO (cit) + 2cit 2UO U 2 2 2 2 [60FEL/NOR] [59LI/LIN] 20 [80VAN/KUC] [72MAR/KLO] 19 2,2 β 10 18 log 17 16 2.5 0.0 0.5 1.0 1.5 2.0 I / m

415 VII.11 Neptunium citrate compounds and compl exes 373 2 − (UO ) (cit) results in: The fitting of the accepted values for 22 2 –1 ο = = (21.34 ± 0.5) and ∆ ε b − (0.8 ± 0.2) kg · mol log , (VII.46) 10 2,2 2 − 2+ UO (Hcit)(aq) U , the two accepted literature val- UO + Hcit and for the reaction 22 ues (Table VII-24) give: –1 ο (VII.47) = (5.0 ± 0.6) and ∆ ε log − (1 ± 1) kg · mol = . b 10 ε ∆ cannot be selected since the number of experimental results The values of is very limited. Due to the scarcity of reliable data, this review assigns larger uncertain- ties to the values obtained from the fitting procedures. In conclusion, Table VII-25 lists 1 the selected equilibrium constants in the uranium(VI)-citrate system. 2+ Table VII-25: Selected formation constants for the citrate complexes of UO at 25 C. ° 2 ο b Reaction log 1 10 2+ 3 − − + cit U UO (cit) UO (8.96 ± 0.17) 2 2 2+ 2 − 3 − UO U + 2cit 0.5) ± (UO ) (cit) 2 (21.3 2 2 22 2+ 2 − U + Hcit UO UO 1.0) Hcit(aq) (5.0 ± 2 2 The selection in Table VII-25 yields: – –1 ο (cit) ∆ , 298.15 K) = – (2166.0 ± (UO 2.8) kJ·mol G 2 fm –1 ο 2 − G (UO ) (cit) ( , 298.15 K) = – (4351.2 ± 6.0) kJ·mol ∆ 2 22 fm –1 ο ∆ (UO . Hcit, aq, 298.15 K) = – (2180.0 ± 6.3) kJ·mol G 2 fm VII.11 Neptunium citrate compounds and complexes A literature search by this review on the thermodynamics of neptunium-citrate systems revealed only information concerning the aqueous complexes of neptunium(V). Simi- 2+ + UO , pO has a linear dioxo structure and the ligand should co- N larly to the case of 2 2 − Np − O axis. This sets some con- ordinate in the equatorial plane perpendicular to the O straints on the stoichiometry of the complexes due to the steric hindrance. Due to the + 2+ + N pO forms only one-to-one as compared to pO N UO , smaller formal charge of 2 2 2 − 3 , whose stability is relatively small. A summary of the equilibrium complexes with cit + N pO citrate system is presented in Table VII-26. data available for the 2 1 et al. [2005KAN/GIL], available to the reviewers only in the final stage of A paper by Kantar preparation of this book, reports ion-exchange experiments which were performed to evaluate the formation of U(VI) – citrate and U(VI) – Fe(III) – citrate complexes. The stability con- 2+ ο 3– – stant for the reaction = 0), determined by U UO I cit UO ( 0.6 at ± log K + cit = 8.7 2 2 1 10 the authors from ion-exchange data, is in agreement with the value selected in this review ο ( ± log K 0.17)). For a consistent evaluation of the mixed complexes in the U(VI) – = (8.96 1 10 Fe(III) – citrate system NEA selected data on Fe(III) hydrolysis are needed.

416 VII Discussion of data selection for citrate compounds and compl exes 374 + N pO . Table VII-26: Literature data on the formation constants of citrate complexes of 2 a t β ( C) log Reference Method Ionic medium ° 10 + 2 − − U NpO + Hcit (Hcit) NpO 2 2 0.12) ClO4 20 (2.69 ± cix [61MOS/MAR] 0.05 M NH 4 ~ 0.05 M NH (CH dis COO) 25 2.37 [82INO/TOC] 4 3 sp 0.05 M ? (2.49 ± 0.05) [85SEV] + − 3 − 2 U NpO (cit) + cit NpO 2 2 0.05 M NH ClO [61MOS/MAR] 20 (3.67 cix 0.09) ± 4 4 1 M (NaClO ± 0.2) (4.42 ± 0.11) [72STO] sp ) (25.0 4 dis 0.05 M NH (CH COO) 25 3.94 [82INO/TOC] ∼ 3 4 cix 0.1 M NaClO ± 0.53) [84REE/DAN] 10 (4.70 4 22 (4.84 ± 0.72) 35 (4.93 0.11) ± ? (2.87 sp 0.05 M [85SEV] ± 0.05) 2.0 M NaClO sp ± 0.02) [90RIZ/NEC] 25 (2.49 4 5.0 m NaCl dis (2.40 25 ± [96BOR/LIS] 0.06) 0.10 m NaClO dis 25 (3.03 ± 0.17) [97POK/CHO] 4 0.30 m NaClO (2.89 0.03) ± 4 0.50 m NaClO ± 0.01) (2.73 4 1.0 m NaClO ± 0.03) (2.74 4 2.0 m NaClO ± 0.02) (2.81 4 3.0 m NaClO ± (2.89 0.01) 4 4.0 m NaClO (2.95 ± 0.02) 4 5.0 m NaClO (3.03 ± 0.03) 4 7.0 m NaClO ± 0.05) (3.20 4 9.0 m NaClO (3.11 0.02) ± 4 0.1 m NaCl dis (2.97 25 ± 0.01) [99BRO/POK] 0.3 m NaCl (2.62 ± 0.05) 0.5 m NaCl (2.60 0.02) ± 1.0 m NaCl (2.39 ± 0.01) 2.0 m NaCl (2.50 ± 0.07) 3.0 m NaCl (2.52 0.01) ± 4.0 m NaCl (2.56 ± 0.05) (2.56 5.0 m NaCl ± 0.03) + − − 3 − 3 + OH NpO + cit NpO U (cit)(OH) 2 2 sp 0.05 M ? (7.40 ± 0.05) [85SEV] a: Refers to the reactions indicated, the ionic strength and temperature given in the table. Uncertainties are those given in the references.

417 VII.11 Neptunium citrate compounds and compl exes 375 Based on the discussion of the literature studies in Appendix A, the constants does not listed in Table VII-27 are accepted in this review. Note that [92TOC/INO3] report thermodynamic constants. + Table VII-27: Accepted formation constants for citrate complexes of N pO used to 2 C. derive the selected values at 25 ° a β Ionic Medium Reference log 1 10 + 3 − 2 − U NpO NpO cit + cit 2 2 [90RIZ/NEC] ± 0.05) 2.2 m NaClO (2.44 4 [97POK/CHO] ± 0.20) 0.10 m NaClO (3.03 4 (2.88 0.10) ± 0.30 m NaClO 4 0.10) (2.72 ± 0.50 m NaClO 4 0.10) ± (2.74 1.0 m NaClO 4 (2.81 ± 0.10) 2.00 m NaClO 4 ± 0.10) (2.89 3.00 m NaClO 4 ± 0.10) (2.95 4.00 m NaClO 4 (3.03 0.10) ± 5.00 m NaClO 4 [99BRO/POK] ± 0.10) 0.1 m NaCl (2.97 0.10) ± 0.3 m NaCl (2.62 0.10) ± (2.59 0.5 m NaCl ± 0.10) 1.0 m NaCl (2.38 0.10) ± (2.48 2.0 m NaCl ± 0.10) (2.49 3.0 m NaCl 0.10) ± (2.52 4.0 m NaCl ± 0.10) (2.51 5.0 m NaCl are converted into molality units. a: The values of log β 10 − Although the weak formation of NpO (Hcit) is probable ( [61MOS/MAR] , 2 ), the values of the stability constant reported are not enough reliable to be ac- [85SEV] cepted. Thus, this review only considers the reaction: 3 − 2 − + U NpO (VII.53) N (cit) pO + cit 2 2 For the SIT analysis, the data obtained in NaClO ( [97POK/CHO] and 4 ) are used. Since there , 5 m ≤ I ≤ ) and in NaCl ( [99BRO/POK] , 5 m I [90RIZ/NEC] m m + 3 − (cit , Na ) ε is dependant on I , the analysis is conducted in are two series of data and m the form of:

418 exes VII Discussion of data selection for citrate compounds and compl 376 2 ∆× zI 0.5091 m ++3 −− I b ε (Na , cit )) log ( ε −−+ (NpO , X ) 1 10 2 m I + 11.5 m 2 − ο+ = (VII.54) −ε b log (Na , NpO (cit) ) I 2 1 10 m 2 2 ο − + . The values of (Na , NpO (cit) ) (VII.53) and ∆ z = − 6 and b log ε to obtain 2 10 1 –1 –1 + − − + (NpO ,ClO ) 0.05) kg · mol = (0.09 ± ε 0.05) kg · mol and = (0.25 ± ε (NpO ,Cl ) 2 24 –1 +3 − and 0.03) log kg · mol = – (0.15 ± (Na , cit ) I 0.03) + (0.13 ± ε [2001LEM/FUG] m 10 (selected by this review) were used. The result of the SIT analysis is shown in Figure b log (VII.53) against VII-28. In Figure VII-29, the result is also shown as the plot of 10 –1 + + − − = ε = 0.25 kg · mol and . As shown in Figure VII-29, ε (NpO ,Cl ) (NpO ,ClO ) I m 2 24 –1 ο (VII.53) between b log approximately describe the difference in 0.09 kg · mol 1 10 aClO and NaCl media. N 4 Figure VII-28: Fitting of the values of b (VII.53) at each ionic strength (Table log 10 VII-27) to the SIT equation. Solid line is drawn using the result of the fitting Eq. (VII.54), with: 2 ο log b (VII.53) = (3.68 ± 0.05), , 6 z ∆ =− 10 1 –1 −+ 3 I 0.03) log ± (0.13 ) + 0.03 ± – (0.15 ε= , (cit , Na ) kg·mol m 10 –1 −+ 2 ε= (0.06 ± 0.03 ) kg·mol (NpO cit , Na ) − 2 –1 +− (NpO ,ClO ) ε = (0.25 ± 0.05) kg·mol 24 –1 + − . ± (NpO ,Cl ) ε = (0.09 0.05) kg·mol 2 3 − + − + The uncertainties in the data include those from and from (cit , Na ) (NpO ,X ) ε ε . 2 − + − 3 2 cit NpO + cit U NpO 2 2 4.5 m I )) − 3 , cit + (Na ε 4.0 , X-) + + 2 (NpO ε - ( D 3.5 + 6 − β [90RIZ/NEC], ClO 10 4 − log [97POK/CHO], ClO 4 − [99BRO/POK], Cl 3.0 0123456 I / m

419 VII.11 Neptunium citrate compounds and compl exes 377 ionic strength. The uncertainties in the log Figure VII-29: Plot of b (VII.53) versus 1 10 + +− − 3 ε and from . Solid lines: (NpO ,X ) (Na ,cit ) ε data include those from 2 2 zI ∆× 0.5091 m ο bb I =+ log log −∆ε m 1 10 1 10 I + 11.5 m − 2+3 +−−+ (cit , Na ) −ε , −ε ∆ε=ε (NpO cit , Na ) (NpO , X ) 22 1 − 3 −+ kg·mol ε=−±+± I (0.15 0.03) (0.13 0.03) log (cit , Na ) 10 m − 1 2 −+ kg·mol ε=−± (NpO cit , Na ) (0.06 0.03) 2 1 − + − = (0.25 (NpO ,ClO ) 0.05) kg·mol , ε ± 24 − 1 + − (NpO ,Cl ) 0.05) kg·mol ± ε = (0.09 2 4.0 3 + 2 − − NpO NpO cit + cit U 2 2 3.5 3.0 1 β 10 log 2.5 − 2.0 [90RIZ/NEC], ClO 4 − [97POK/CHO], ClO 4 − [99BRO/POK], Cl 1.5 0123456 / m I + N pO at 25 ° C is: The selected formation constant for the citrate complex of 2 ο log b (VII.53) = (3.68 ± 0.05), 1 10 2 + − with 0.03). (Na , NpO (cit) ) ε = − (0.06 ± 2 The selection above yields: –1 2– ο ∆ (NpO G (cit) . , 298.15 K) = – (2091.0 ± 6.0) kJ·mol 2 fm

420 VII Discussion of data selection for citrate compounds and compl exes 378 VII.12 Plutonium citrate compounds and complexes A literature search by this review on the thermodynamics of plutonium-citrate systems revealed information concerning the aqueous complexes of plutonium(III), pluto- nium(IV) and plutonium(VI). No thermodynamic data for Pu citrate compounds could be identified (see also Section VII.1.1). As shown in Table VII-28 and Table VII-29, there is only one paper dealing 3+ 2+ PuO with citrate, respectively. Based on the and with the complex formation of Pu 2 and [75NEB/AND] in Appendix A, these values are not consid- discussion of [89POC] ered reliable enough to derive selected values. 3+ Table VII-28: Literature data for the formation constants of citrate complexes of Pu . a Method Ionic medium Reference C) ° ( t β log 10 1 3+ + 2 − Pu Pu(Hcit) + Hcit U 0.1 M KCl (4.82 ± 0.27) [89POC] 20-22 pot 3+ − 3 + cit Pu(cit)(aq) U Pu 0.1 M KCl (6.71 pot 0.25) [89POC] 20-22 ± 3+ − 4 − Pu cit cit) + H U Pu(H 1 1 − − 20-22 [89POC] ± 0.30) 0.1 M KCl (15.33 pot a: Refers to the reactions indicated, the ionic strength and temperature given in the table. Un- certainties are estimated in this review. 2+ Table VII-29: Literature data for the formation constants of citrate complexes of PuO . 2 a Method Ionic medium Reference C) ° ( t β log 10 2+ 3 − – PuO U PuO + cit cit 2 2 25 sp 9.00 0.5 M HClO [75NEB/AND] 4 8.90 35 2+ 4 − 3 − U + 2cit PuO (cit) PuO 22 2 pot 25 14.98 [75NEB/AND] 0.5 M HClO 4 35 14.85 a: Refers to the reactions indicated, the ionic strength and temperature given in the table. 4+ The information on the complex formation of Pu with citrate is also limited as shown in Table VII-30.

421 exes VII.12 Plutonium citrate compounds and compl 379 4+ citrate system at 25 ° Table VII-30: Experimental equilibrium data for Pu C. a Reference Method Ionic medium Reaction K log 10 − 4+ + 3 + cit U Pu(cit) gl Pu [66NEB2] 0.5 M NaClO 15.8 4 2 − 3 4+ − U + 2cit Pu(cit) Pu 29.0 2 4+ 3 − + 0.5 M NaClO sp U Pu(cit) + cit Pu − 15.43 [66NEB] 14.72 4 2 − 3 − 4+ Pu U + 2cit 30.20 − 29.83 Pu(cit) 2 + − 4+ 3 redox 0.5 M NaClO + cit U Pu(cit) Pu 15.5 [66NEB2] 4 2 − 3 − 4+ Pu 30 + 2cit Pu(cit) U 2 4+ 3+ + 1.0 M LiClO dis cit U Pu(H cit) + H + H Pu 2.85 [72MET/GUI3] 4 3 2 + 4+ 2+ Pu + H + 2H U 1.85 cit Pu(Hcit) 3 4+ + + Pu cit U Pu(cit) + H + 3H 0.1 3 4+ + Pu 5.63 cit + H O U Pu(cit)(OH)(aq) + 4H + H − 2 3 − 4+ + Pu cit + 2H 9.85 O U − Pu(cit)(OH) + H + 5H 3 2 2 3+ 4+ + Pu cit cit)(H + 2H cit) U + H Pu(H 4 3 2 3 2+ 4+ + Pu 3.5 U Pu(H cit) + 2H cit + 2H 3 22 4+ + Pu cit U Pu(Hcit) (aq) + 4H + 2H − 1.53 2 3 − 2 4+ + Pu 7.5 + 2H Pu(cit) U + 6H cit − 3 2 4 − 4+ + Pu O U + 2H 16.6 Pu(cit) (OH) cit + 2H + 8H − 2 3 22 4+ 2+ + dis 1 M LiNO [75ZAI/NIK] cit U Pu(Hcit) Pu + 2H + H 1.63 3 3 2 M LiNO 1.04 3 4 M LiNO 0.46 3 6 M LiNO − 0.07 3 a: Refers to the reactions indicated, the ionic strength and temperature given in the table. Although very strong complex formation is expected, the stoichiometry of the e interaction is very strong, it can be stud- species formed is not well established. As th ied only in highly acidic solutions (pH less than 0.5), where the activity coefficients of the relevant species are very difficult to be estimated. As a result, the values in Table VII-30 are dependent on the assumption of the complex species formed under the corre- sponding experimental condition. Since any study on plutonium complex formation involves the difficulties of controlling the oxidation state and of avoiding hydrolysis, studied instead the [96CHO/ERT] and Choppin et al. [87RAY/DUF] et al. Raymond 4+ 4+ 4+ with citrate, where Pu and Th are considered to be simi- complex formation of Th 4+ cf. is known not to hydrolyse up to pH around 4 ( lar in complex formation and Th Appendix A). 3+ 2+ Pu(H cit) , , cit) Considering the analogy between Th(IV) and Pu(IV), Pu(H 2 22 2+ + − 2 , Pu(Hcit) and are deemed most possible among the Pu(cit) (aq), Pu(cit) Pu(Hcit) 2 2 various species considered. To compare the values of the formation constants obtained log b using the protona- for these species, the data in Table VII-30 are converted into 10 tion constants of citric acid selected in this review, and the resulting equilibrium con-

422 VII Discussion of data selection for citrate compounds and compl exes 380 4+ stants are shown in Table VII-31. Some data for Th are also listed for comparison ο is used for this conversion). K = 6.33 obtained from the data in [96CHO/ERT] log ( 1(H) 10 r, there are not enough reliable values to Although the data roughly agree with each othe 4+ citrates. select any values for Pu 4+ Table VII-31: Literature data on the formation constants for citrate complexes of Pu or 4+ expressed as C. at 25 ° Th b log 10 Reaction Ionic medium Reference b log 10 4+ − 3+ cit Pu U Pu(H cit) + H 1.0 M LiClO [72MET/GUI3] 5.61 4 2 2 4+ − 2+ 2 U Pu(Hcit) Pu + Hcit 8.71 [72MET/GUI3] 1.0 M LiClO 4 1 M NaNO 8.44 [75ZAI/NIK] 3 2 M NaNO 7.78 3 4 M NaNO 7.23 3 6 M NaNO 6.79 3 4+ 3 − + Pu U Pu(cit) (15.3 0.5 M NaClO [66NEB2] + cit ± 1) [66NEB] , 4 1.0 M LiClO 11.9 [72MET/GUI3] 4 2 − 3 − 4+ Pu [66NEB] + 2cit Pu(cit) 0.5 M NaClO [66NEB2] (30 ± 1) U , 4 2 2 − 4+ 2+ [96CHO/ERT] 10.27 I → = 0 1~14 m NaClO 4 Th(Hcit) U + Hcit Th − 4+ 2 19.24 1~14 m NaClO → I = 0 4 Th(Hcit) + 2Hcit (aq) U Th 2 + 4+ − 3 13.7 1~14 m NaClO → I = 0 4 Th Th(cit) + cit U VII.13 Americium citrate compounds and complexes A literature search by this review on the thermodynamics of americium-citrate systems revealed only information concerning th e aqueous complexes of americium(III). No thermodynamic data for Am citrate compounds could be identified (see also Section 3+ VII.1.1). The experimental equilibrium data found on the complex formation of Am with citrate are listed in Table VII-32.

423 VII.13 Americium citrate compounds and compl exes 381 3+ Table VII-32: Literature data on the formation constants for citrate complexes of Am . a ( β t ° Method Ionic medium Reference C) log 10 3+ 2 − + Am(Hcit) + Hcit Am U 25 4.57 [84BOU/GUI] 0.1 M LiClO dis 4 [84BOU/GUI] 25 5.83 pot 0.1 M LiClO 4 ± 0.02) [71OHY/OHY] (5.31 cix 25 0.1 M NaCl ± 0.04) [72EBE/MOA] 25 (4.53 1.0 M NaClO sp 4 − − 3+ 2 U + 2Hcit Am(Hcit) Am 2 0.1 M LiClO [84BOU/GUI] 25 8.94 dis 4 0.1 M NaCl 25 ± 0.02) [71OHY/OHY] cix (8.23 − − 3 − 3+ 2 2 U Am(Hcit)(cit) + cit + Hcit Am dis 0.1 M LiClO 25 10.6 [71GUI/BOU] 4 0.1 M ? 25 (10.2 em ± [71STE] 0.1) 25 (10.76 ± 0.20) [74HUB/HUS] , 0.1 M LiClO dis 4 [76HUB/HUS] [84BOU/GUI] 25 10.77 0.1 M LiClO dis 4 [84BOU/GUI] 25 13.44 pot 0.1 M LiClO 4 3+ − 3 + cit U Am(cit)(aq) Am dis 0.1 M LiClO 25 7.7 [71GUI/BOU] 4 25 (7.74 ± 0.08) [71STE] em 0.1 M ? 25 8.0 [84BOU/GUI] 0.1 M LiClO dis 4 [84BOU/GUI] 25 8.69 0.1 M LiClO pot 4 [71OHY/OHY] ± 0.08) (6.74 0.1 M NaCl cix 25 [72EBE/MOA] ± 0.06) 25 (6.96 sp 1.0 M NaClO 4 ± 0.1) [2001CHO/BON] 0.3 m NaCl 25 (5.9 dis ± 0.1) [2001CHO/BON] 1 m NaCl (5.2 dis 25 [2001CHO/BON] 0.1) ± (5.0 2 m NaCl dis 25 [2001CHO/BON] 0.02) ± (4.84 25 3 m NaCl dis ± 0.06) [2001CHO/BON] dis 4 m NaCl 25 (5.38 ± 0.2) [2001CHO/BON] dis 5 m NaCl (5.1 25 3 − 3 − 3+ U + 2cit Am(cit) Am 2 0.1 M ? 25 em ± 0.2) [71STE] (10.9 dis 0.1 M LiClO 25 (12.16 ± 0.20) [74HUB/HUS] 4 25 12.1 [84BOU/GUI] dis 0.1 M LiClO 4 25 14.29 [84BOU/GUI] pot 0.1 M LiClO 4 ± 0.08) [71OHY/OHY] cix 0.1 M NaCl 25 (11.55 25 (10.3 ± 0.2) [72EBE/MOA] sp 1.0 M NaClO 4 (Continued on next page)

424 VII Discussion of data selection for citrate compounds and compl exes 382 Table VII-32: (continued) a ( β t ° C) log Reference Method Ionic medium 10 − 3 − 3+ − Am(cit)(OH) + OH + cit Am U [84BOU/GUI] 0.1 M LiClO pot 25 10.53 4 − + O U + H Am(cit)(aq) + H Am(cit)(OH) 2 1.0 M NaClO 25 (5.61 ± 0.02) [72EBE/MOA] sp 4 2 − 3 − − 3+ U + 2cit + 2OH (Am(cit)(OH)) 2Am 2 pot 25 22.80 [84BOU/GUI] 0.1 M LiClO 4 a: Refers to the reactions indicated, the ionic strength and temperature given in the table. Uncertainties are those given in the references. Due to the fairly strong complex formation and strong dependence of the inter- among research groups as to the speci- action on pH, agreement has not been reached + − , Am(Hcit) , ation and stability constants. Among the species proposed, Am(Hcit) 2 3 − Am(cit) are considered most plausible by this review. Based on the Am(cit)(aq) and 2 discussion of the experimental studies in Appendix A, the equilibrium constants listed in Table VII-33 are accepted to derive the selected values in this review. Since , [74HUB/HUS] and [84BOU/GUI] are the reports from the same group, [71GUI/BOU] [71OHY/OHY] are accepted. The values from [84BOU/GUI] only the values listed in − 3+ and Cl . are corrected for the complex formation between Am [2001CHO/BON] and 3+ Table VII-33: Accepted formation constants for citrate complexes of Am at 25 ° C used to derive the selected values. b a Ionic medium Reference log β log β 10 app 10 3+ 2 − + + Hcit U Am(Hcit) Am 0.1 m LiClO (4.57 ± 0.5) ± 0.5) [84BOU/GUI] (4.57 4 0.1 m NaCl (5.31 ± 0.5) (5.33 ± 0.5) [71OHY/OHY] 1.05 m NaClO ± 0.5) (4.51 ± 0.5) [72EBE/MOA] (4.51 4 − 3+ 2 − Am + 2Hcit Am(Hcit) U 2 0.1 m LiClO (8.93 ± 0.5) (8.93 ± 0.5) [84BOU/GUI] 4 [71OHY/OHY] (8.23 ± 0.5) (8.24 ± 0.5) 0.1 m NaCl (Continued on next page)

425 VII.13 Americium citrate compounds and compl exes 383 Table VII-33: (continued) a b Ionic medium Reference log log β β 10 10 app − 3+ 3 U Am(cit)(aq) Am + cit , ± ± 0.5) [71GUI/BOU] 0.5) (8.0 [84BOU/GUI] (8.0 0.1 m LiClO 4 ± 0.5) (6.8 ± 0.5) [71OHY/OHY] 0.1 m NaCl (6.74 1.05 m NaClO (6.94 0.5) (6.94 ± 0.5) [72EBE/MOA] ± 4 (5.90 0.3 m NaCl ± 0.21) [2001CHO/BON] 0.2) (5.93 ± 1 m NaCl (5.19 (5.25 ± 0.21) ± 0.2) [2001CHO/BON] (4.98 2 m NaCl (5.09 ± 0.22) ± 0.2) [2001CHO/BON] 3 m NaCl (4.81 0.2) (4.99 ± 0.24) [2001CHO/BON] ± (5.34 4 m NaCl (5.63 [2001CHO/BON] 0.25) ± ± 0.2) 5 m NaCl (5.05 ± 0.26) [2001CHO/BON] ± 0.2) (5.52 − 3 3+ − 3 Am + 2 cit U Am(cit) 2 [74HUB/HUS] (12.1 0.1 m LiClO ± 0.5) ± , [84BOU/GUI] 0.5) (12.1 4 (11.57 ± 0.5) (11.59 ± 0.5) [71OHY/OHY] 0.1 m NaCl 1.0 m NaClO (10.3 0.5) (10.3 ± 0.5) [72EBE/MOA] ± 4 β are those converted into molality unit. a: The values of log app 10 b: The values of log β are those corrected for chloride complex formation. 10 The NEA-TDB reviews [95SIL/BID] and [2003GUI/FAN] select the formation ο ± log = (0.24 b 0.03), for: constants, 1( Cl) 10 − 3+ 2+ + Cl U (VII.55) AmCl Am ο and ± = − (0.74 0.05) [2003GUI/FAN] for b log 2(Cl) 10 3+ − + Am + 2 Cl AmCl (VII.56) U 2 –1 − 3+ − 3+ (Am ,Cl ) = ε = (0.49 ± 0.02) kg · mol ε and (Am ,ClO ) with the assumptions of 4 –1 − − − 2+ + 2+ = ε = (0.39 ± 0.04) kg · mol (AmCl ,Cl ) , and (AmCl ,ClO ) = ε (AmCl ,Cl ) ε 2 4 –1 + − (AmF ,ClO ) = (0.17 . From these values with ± 0.04) kg · mol ε [95SIL/BID] 24 –1 + − ± 0.01) kg · mol and (Appendix B), the values of log (Na ,Cl ) β ε = (0.03 10 1(Cl) d they were used to apply corrections β at each ionic strength were estimated an log 10 2(Cl) 3 − − and Am(cit) Am(Hcit) to equilibrium constants for the formation of Am(cit)(aq), 2 2 according to the fo llowing equation: − − 2 log (reported) + log β β [Cl (corrected) = log ] + β [Cl ) (VII.57) ] β (1 + 10 10 10 1(Cl) 2(Cl) The corrections were less than 0.5 log -units at all ionic strengths. To conduct 10 1 M is ≤ I at nsidered that the contribution of the SIT analyses, this review has co ε I ∆ within the uncertainties of the equilibrium constants, and it can be neglected. For log b of the reaction: 10 1

426 VII Discussion of data selection for citrate compounds and compl exes 384 − 3+ 3 Am + cit Am(cit)(aq) U , (VII.58) the SIT analysis was conducted with the values obtained in NaCl medium and with − +3 1 M listed in Table VII-33. Since ≤ those obtained at I (Na ,cit ) ε is dependent on ionic strength, the analysis was conducted using the following equation: 2 ∆ 0.5091 zI m 33+ +− ο − (VII.59) b I I b (M , cit ) log ε (Am ,Cl ) log + + = −ε m 10 10 m I + 11.5 m 2 z ∆ = − 18 where –1 3+ 3+ −− ε (Am ,Cl ) (0.49 0.02) ε ==± kg · mol (Am ,ClO ) (VII.60) 4 +− 3 ε (M ,cit ) is and –1 +3 − (Na ,cit ) (0.15 0.03) (0.13 0.03) log kg · mol I =−±+± ε (VII.61) 10 m [71OHY/OHY] for and and [2001CHO/BON] –1 +3 − (VII.62) (Li ,cit ) (0.55 0.11) (0.3 0.2) log ε=−±+± kg · mol I 10 m . for [71GUI/BOU] As shown in Figure VII-30 and Figure VII-31, the data fits well the model with ο the estimated value of log 0.09). b (VII.58) = (8.55 ± 1 10 3+ − When the fitting was carried out with (Am ,Cl ) ε as a variable parameter as ο b log , the analysis gave: well as 10 –1 ο 3+ − =± b and , (Am ,Cl ) log (8.86 0.15) ε=± kg · mol (0.34 0.05) 1 10 ε in fair agreement with the recommended values of . and [84BOU/GUI] considerably deviate from the [72EBE/MOA] The data of fitted line, presumably due to the differences in their ionic media or to overlooked sys- tematic errors in these works. Thus, this review selects the value of: ο ± log 0.09) b (VII.58) = (8.55 10 1 obtained from the SIT regression, but with an increased uncertainty, and the already recommended values of ε .

427 VII.13 Americium citrate compounds and compl 385 exes log b (VII.58) Figure VII-30: Fitting of the values of at each ionic strength (Table 10 1 VII-33) to the SIT equation. Solid line is drawn using the result of fitting Eq. (VII.59) with: 2 ο z b ± 0.20), log 18 (VII.58) = (8.55 ∆ =− , 1 10 –1 3+ 3+ −− ε (Am ,Cl ) ==± kg · mol ε (Am ,ClO ) (0.49 0.02) 4 –1 +3 − ε (Na ,cit ) =−±+± kg · mol I (0.15 0.03) (0.13 0.03) log 10 m –1 +3 − (0.55 0.11) (0.3 0.2) log . I (Li ,cit ) ε=−±+± kg · mol 10 m 12 11 m Ι ) 3− cit , + ε(Μ 10 − D + 18 β 9 10 log [71OHY/OHY] [84BOU/GUI] [72EBE/MOA] 8 [2001CHO\BON] 012345 m / I

428 VII Discussion of data selection for citrate compounds and compl exes 386 log I for the reactions, (VII.58) and : against Figure VII-31: A plot of b m 10 − 33 +− 3 2 cit Am + U (VII.63) Am(cit) 2 Solid lines are drawn using the result of fitting equation (VII.59). ο 2 18, b (VII.58) = (8.55 ± 0.20), ∆ z log = − 1 10 + −−+ 33 (Am , Cl ) ∆ε = −ε −ε , (cit , Na ) − 1 +− 3 ± (Am , Cl ) = (0.49 ε 0.02) kg·mol –1 +3 − (Na ,cit ) = – (0.115 0.03) + (0.13 ± 0.03)log ε I kg · mol ± m 10 2 ο ∆ε log z ∆ = − 18, (VII.63) = 13.9, I = 0 b m 10 2 14 [71OHY/OHY] 13 [84BOU/GUI] [72EBE/MOA] 12 [2001CHO/BON] 11 3 − 10 Am(cit) 2 9 β 10 log 8 7 Am(cit)(aq) 6 5 4 012345 I / m

429 VII.13 Americium citrate compounds and compl exes 387 For the other reactions, the accepted literature values are quite limited. Since ε the uncertainties are larger than the contribution of I ≤ 1 m, the SIT analyses ∆ I at m ∆ : I ε have been carried out neglecting 2 I z 0.5091 ∆ m ο (VII.64) bb log log =+ 10 10 + I 11.5 m The results of the fittings are shown in and Figure VII-32. The estimated values ο log b and their uncertainties are: of 10 3+ 2 − + Am U Am(Hcit) + Hcit (VII.65) ο log b = (6.5 ± 0.3), 10 1 − 3+ 2 − U + 2Hcit Am(Hcit) (VII.66) Am 2 ο log b 0.4), = (10.8 ± 2 10 − 3+ 3 3 − Am U Am(cit) + 2cit (VII.67) 2 ο log b = (13.9 ± 0.3). 10 2 Figure VII-32: A Plot of log I . Solid lines are drawn by using the results of against β m 10 fitting equation (VII.64). ο 2 log (VII.65) = 6.5 b z ∆ = − 12, 1 10 ο 2 16. b (VII.66) = 10.8 log z ∆ = − 10 1 11 [84BOU/GUI] (LiClO ) 4 [71OHY/OHY] (NaCl) 10 ) [72EBE/MOA] (NaClO 4 9 − Am(Hcit) 8 2 7 β 10 log 6 5 + 4 Am(Hcit) 3 0.8 0.6 0.4 0.2 0.0 1.2 1.0 / m I

430 VII Discussion of data selection for citrate compounds and compl exes 388 As seen in Figure VII-31 and in Figure VII-32, the amounts of available data are quite limited and have relatively large uncertainties. The estimated value of ο may contain still larger uncertainties as compared to the uncertainties in the log b 10 individual experimental values, which are estimated by this review from an evaluation of the original publications. Therefore, this review selects the values of the formation constants obtained from the SIT fittings with increased uncertainties as listed in Table 3+ VII-34. To show the outline of the complex species of Am citrates formed, the distri- 3+ + bution of the Am log [ H ] − is given in Figure VII-33. For citrates as a function of 10 I = 0.1 M) were used the simulation, the values selected in this review (extrapolated to for the protonation constants and formation constants as listed in the Figure caption. 3+ Table VII-34: Selected formation constants for the citrate complexes of Am ° C. at 25 ο b log Reaction 10 + 3+ − 2 U Am(Hcit) (6.5 ± 1.0) Am + Hcit − 2 3+ − ± 1.0) Am(Hcit) + 2Hcit (10.8 U Am 2 3+ 3 − Am + cit U Am(cit)(aq) (8.55 ± 0.20) − 3 3+ 3 − Am U Am(cit) (13.9 ± 1.0) + 2cit 2 The selections in Table VII-34 yield: + –1 ο , 298.15 K) = – (1834.4 G ∆ ± 7.7) kJ·mol (Am(Hcit) fm –1 − ο 8.4) kJ·mol Am(Hcit) G , 298.15 K) = – (3057.5 ± ( ∆ fm 2 –1 ο ∆ (Am(cit), aq, 298.15 K) = – (1809.8 ± 5.3) kJ·mol G fm –1 ο 3 − . G ( Am(cit) ∆ , 298.15 K) = – (3002.6 ± 8.5) kJ·mol 2 fm

431 VII.13 Americium citrate compounds and compl 389 exes complex species in solutions containing Figure VII-33: Distribution of Am(III)-citrate 6 − 3 − + 10 M Am(III) and 10 log [H ] − at I M citric acid as a function of = 0.1 M. The 10 following equilibrium constants have been used: +− − 43 rr Hcit HHcit + U citrate protonation: 1 rr − = 3) = 2.92 log K ( r = 1) = 5.70, r ( log K ( r = 2) = 4.36 and K log 3 2 10 1 10 10 Am(III)citrate complexes: 3 − ((VII.58), Am(cit)(aq)) = 6.92, = 11.84, log b ((VII.63), log b Am(cit) 2 10 10 + − b ((VII.65), Am(Hcit) ) = 5.18, ) = 9.09. log log b ((VII.66), Am(Hcit) 2 10 10 -5 + − Am(Hcit) Am(Hcit) 2 -6 3+ − 3 Am Am(cit) 2 -7 Am(cit) -8 c 10 -9 log − -10 − Am(cit) Am(Hcit) -11 2 -12 0123456 + − log ] [H 10

432

433 Chapter VIII Discussion of data selection for VIII ethylenediaminetetraacetate (edta) compounds and complexes VIII.1 Introduction 1 − ; CAS Regis- H N O This acid has the formula C (molecular weight: 292.242 g·mol 10 8 2 16 try Number: 60-00-4), and alth ough it is usually referred to as ethylenediaminetetraace- tic acid, about 40 other names are known, for example: EDTA, Titriplex, Trilon B, Cheelox, Complexon II, etc. Ethylenediaminetetraacetic acid may releas e up to four protons, but it may also + act as a base, accepting up to two H ions. Following the rules by the IUPAC, in reac- 4) ( n − tions and formulae the abbreviation H edta is used for the acid, and H for the edta 4 n lenediaminetetraacetate the deprotonated ethy different protonation steps. For example, 4 − ligand in aqueous solutions is denoted as: edta . In formulae where the abbreviation is 3 − preceded by a lower case it is enclosed in parenthesis, as in Na(edta) . Also protonated – forms of the edta ligand is enclosed in parenthesis, as in Ni(Hedta) . It should however be noted that what one obtains from most experiments is the global composition of the – , and that using most techniques this complex is complexes, in this example NiHedta – indistinguishable with e.g . Ni(OH)(H edta) etraacetate group in . The ethylenediaminet 2 metal ion complexes is simply denoted as edta. Ethylenediaminetetraacetic acid behave s like most aminoacids forming zwit- terions [47SCH/ACK] . This is reflected for example in the crystal structure of H edta·H O(cr), ( cf . Figure VIII-1). 4 2 Owing to its ability to form several chelate rings, the ethylenediaminetetra- acetate ligand is a strong complexing agent. Metal-edta complexes have important practical applications in analytical and se parations chemistry. The 1:1 complexes are 2 − − most common, for example: Ni(edta) etc , Ni(Hedta) . The structures , Ni(H edta)(aq), 2 391

434 VIII Discussion of data selection for edta ligand 392 for two of these complexes, obtained from single-crystal X-ray investigations, are shown in Figure VIII-2. Mixed-ligand comple xes are also quite common, especially for larger cations, such as lanthanides, actinides, etc . Two of these complexes are shown in mplexes with edta are also known. In some of these Figure VIII-3. Polynuclear co inated to two metal atoms. Examples of complexes edta behaves as a “bridge”, co-ord polynuclear complexes are sh own in Figure VIII-4. Figure VIII-1: The structure of H edta as a double zwitterion [74LAD/POV] . 4 H + N + O N H H H O O O O O O O − 2 − Figure VIII-2: The structure of: (a) Ni(edta) , and (b) Ni(Hedta)(H O) [84NES/POR2] 2 [86POL/FIL] . a) b) 2− H − O O O H H O O O O O N Ni N Ni O O O OO O O N N O O

435 VIII.1 Introduction 393 2 − [95MIS/SER] edta Figure VIII-3: Structures of two mixed-ligand complexes: ZrCO 3 − 3 and ThF edta [85MIK/LOB] . 3 a) b) 2− O F 3− F F O O O Th O O O O O N O Zr O O O O O O O N N O O N Figure VIII-4: Structures of polynuclear edta complexes: Tc (H O edta) 2 2 2 2 − 4 [81BUE/AND] edta , and (UO . [85SHC/ORL] F ) 2 2 2 a) b) O O O 4− O O O O O O O F O F O O N N N O O Tc Tc U U O N O N O N O F O F O O O O O O O O O O

436 VIII Discussion of data selection for edta ligand 394 VIII.2 edta(cr) H 4 [74LAD/POV] and β [73LAD/POV] , α . Two crystalline modifications of H edta exist: 4 form is anhydrous H β− The common α -H edta(cr), while edta is a hydrate, perhaps 4 4 with variable amounts of water of crystallization and unstable to exposure in air. The α crystal structure of a sample of -edta with composition H ⋅ 0.39H O(cr) was re- edta 4 2 ported in [74LAD/POV] . All thermodynamic informat ion available on ethylenediaminetetraacetic acid edta(cr). β− the stable corresponds to form: H 4 Thermodynamic properties VIII.2.1 [67ADA/CAR] , edta(cr) has been determined The heat of combustion of H 4 . From these data the enthalpy of formation is calculated to be: [88VAS/BOR] ο 1 − ∆ H 1.5) kJ·mol (1759.8 ± edta, cr, 298.15 K) = (H − 4 fm which is selected by this review. There is no information on the stan dard entropy or heat capacity of H edta(cr). 4 Solubility of H VIII.2.2 edta(cr) 4 The solubility of H edta(cr) in aqueous solutions has been determined in several studies 4 [58MEC/SCH] [58YOS/IGU] , , , [59BEC/GOR2] , [59IGU/YOS] , [59BEC/GOR] [59KLY/SMI4] , [60BEC/GOR] , [62KRO/ERM] , [63TAN/TER] , , [60PAL/UDA] [67AND] [74TER/NIK] , , [74BUL] , , , [75LAG/LAG] , [67MES/VIN] [72SMY] [82TER/IVA] [83KRA/DEC] , [89SAL/BOO] , [99KAR/HAR] . , From solubility data of H edta(cr) in solutions of varying acidity, it is possible 4 to determine the equilibrium constant for: edta(aq) (VIII.1) edta(cr) U H H 4 4 as well as some of the values for the protonation constants: , K ., depend- K K , etc , 6 5 4 edta(aq)” is an equilibrium ing on the acidity range studied. It should be noted that “H 4 mixture of two isomeric zwitterions, as indicated in Figure VIII-6. Some of the investigations that reported the solubility of H edta(cr) as a function of either pH or 4 + [H were not included in this review, eith ] er because they were not performed using TOT , [59BEC/GOR2] , [59BEC/GOR2] , [58MEC/SCH] a constant ionic medium, [60BEC/GOR] , [60PAL/UDA] , [89SAL/BOO] , [99KAR/HAR] , or for other reasons discussed in the corresponding entries of Appendix A, [59KLY/SMI4] , [63TAN/TER] , [72SMY] , [74BUL] , [74TER/NIK] . The studies considered in this review are listed in Table VIII-1.

437 VIII.2 H edta(cr) 4 395 Table VIII-1: Literature studies considered by this review where the solubility of + + H ] (or [H ) at constant ionic edta(cr) has been reported as a function of [H ] TOT 4 strength. t (M) (°C) Medium log K (VIII.1) Reference I s 10 b b 0.08 , [58YOS/IGU] 30 (H,K)Cl 0.1 [59IGU/YOS] − (3.64 ± b b (3.57 [58YOS/IGU] , [59IGU/YOS] 0.04) 30 (H,K)Cl 0.5 − ± b b ± 0.5) 2 [62KRO/ERM] − (3.4 25 (H,Na)NO 3 a 1 − (3.32 ± 0.20) [67AND] 20 (H,Na)ClO 4 a [67MES/VIN] (3.75 0.20) 25 (H,K)Cl 1 ± − a − (2.88 ± 0.20) 3 [75LAG/LAG] 25 (H,Na)ClO 4 a b − (3.95 ± 0.20) [82TER/IVA] 0.1 20 (H,K)NO 3 b − (3.7 ± 0.5) 2 20 a (3.83 ± 0.20) 25 0.1 − a 1 − (3.45 ± 0.20) [83KRA/DEC] 21 (H,Na)ClO 4 a: uncertainty assigned in this review. b: data recalculated in this review, cf . Appendix A. In order to correlate studies performed at both 20 and 25°C, the enthalpy change for reaction (VIII.1) is needed. Two studies have reported ∆ (VIII.1) H rm [58YOS/IGU] [59IGU/YOS] and [79VAS/KOC] . The calorimetric study reported by , ο ∆ H Vasil’ev [79VAS/KOC] . in et al resulted in two different values of (VIII.1) in rm nitrate and perchlorate media as discussed in Appendix A. The data may be modelled ο setting a common value for ., ∆ (VIII.1) for all ionic media studied by Vasil’ev et al H rm resulting in the selected value: ο − 1 (VIII.1) = (29 ± H . ∆ 3) kJ·mol rm This selection leads to: ο –1 ∆ H (H ± 3.4) kJ·mol edta, aq) = – (1730.8 . 4 fm III.1) listed in Table VIII-1 for tem- The equilibrium constants for reaction (V ≠ 25°C were converted to 25°C values using the enthalpy change given above. peratures This value agrees with the enthalpy change obtained by Yoshino [58YOS/IGU] , et al. and Iguchi [59IGU/YOS] from the temperature variation of the solubility of et al. ο ± edta(cr) in “pure” water in the range 20 to 45°C, namely (VIII.1) = (28 H ∆ 4) H 4 rm − 1 kJ·mol , with the uncertainty assigned by this review. Although the speciation was not [58YOS/IGU] established in , [59IGU/YOS] , the error introduced is limited because of the small heats of protona tion and dissociation of H cf . Section VIII.3.8. edta(aq), 4 Because of inadequacies in the original papers, the da ta from a few references has been reinterpreted in this review (see footnotes in Table VIII-1) as described in Appendix A. It should be pointed out that H edta(aq) is not a dominant species. Because 4 of this, the total uncertainties in the individual values in Table VIII-1 were set to a

438 VIII Discussion of data selection for edta ligand 396 minimum value of ± -units. The data in Table VIII-1, converted to 25°C and 0.2 log 10 ties, is plotted in Figure VIII-5. molal units, and with increased uncertain Reaction (VIII.1) should have a negligible ionic strength dependence, because H ectric charge. Nevertheless, edta(aq) nominally has no el both anion and cation effects 4 cannot be ruled out because H edta(aq) is a double zwitterion. Although the ionic 4 + + strength dependence for data obtained in (H ) media differs somewhat with that of , Na + + the data in (H , K erences are not significant. The ) media (Figure VIII-5), the diff weighted least-squares regression gives: ο log K (VIII.1) = − (3.80 ± 0.19), 10 s − 1 − (0.29 ± 0.14) kg·mol edta) = (H . ∆ε 4 Figure VIII-5: Equilibrium constants for reaction: H edta(cr) U H edta(aq). All data 4 4 converted to molal units and to 25°C when necessary. Symbols with white background + + + + correspond to (Na , H , H ) media; grey background to (K ) media. [58YOS/IGU], [59IGU/YOS] −2.0 [62KRO/ERM] [67AND] [67MES/VIN] −2.5 [75LAG/LAG] [82TER/IVA] [83KRA/DEC] (VIII.1) s −3.0 K 10 −3.5 log edta(aq) edta(cr) U H H 4 4 H z H edta(aq) edta(cr) 4 4 C ° at (20 to 30) −4.0 at (20 to 30)°C 01234 / molal I m ta in aqueous solutions VIII.3 Acid-base equilibria of ed VIII.3.1 Introduction edta forms a double zwitterion which may act both as an When dissolved in water H 4 + ions. The acid by releasing up to four protons, and as a base accepting up to two H schematic structures of the different edta -species are shown in Figure VIII-6. The 4 − following nomenclature is used for the protonation reactions of edta : r (4) − ⎡⎤ Hedta r ⎣⎦ 4) − r ( r − 5) ( + H + H edta edta H K U = r − ( r 1) r (5) −+ r ⎤⎡ ⎤ ⎡ ⋅ Hedta H − (1) r ⎣ ⎦⎣ ⎦ − r (4) ⎤ ⎡ Hedta r ⎦ ⎣ − r ( + 4) 4 − edta = r H edta U β H + r r r −+ 4 ⎤⎡ ⎤ ⎡ edta H ⋅ ⎦⎣ ⎦ ⎣

439 VIII.3 Acid-base equilibria of edta in aqueous solutions 397 4 − 3 − anions may form su- and/or Hedta 6. It has been reported that edta ≤ r where 1 ≤ pramolecular aggregates in alkaline solutions (pH 6) with a molecular weight between 2 [94MUL/HAE] 8000 and 14000 ggest that aggregates are . The dialysis experiments su ) x ( p − 4 formed with a stoichiometry (H x ) , with 37 ≤ edta ≤ 66, but the interpretation of x p the observed alkalimetric titration curve is wholly against the claimed polymerisation. − 4) ( r Figure VIII-6: Schematic structures of H . edta r O O O H O O H O OH O + H + H + N + N N N H H O H O H OH O OH O O O + 2+ edta H H edta 5 6 O O O O H O O H O O + + H H + + N N N N H H HO O OH O O O O O edta H 4 O O O O O O O O + H + H + N + N N N H H O O O O O OH O O − − 2 H edta edta H 3 2 O O O O O O O O H + N N N N O O O O O O O O − − 4 3 Hedta edta ≈ A large number of references ( 120) were found from a literature search for the acid-base equilibria of edta. The majority of these references contain studies on metal complexation, where the authors either needed values for the dissociation con- stants of H edta under the same experimental conditions as the metal-complexation 4 study, or used the edta system to test the experimental set-up.

440 VIII Discussion of data selection for edta ligand 398 Most of these references contain data for relatively dilute ionic media, mostly for = 0.1 M. Only few studies have dealt with ionic media 1 M. With only one I I 2 + exception, all studies at > 1.2 M have been performed in Na media. I Because of the large number of references, it was a dvantageous to judge the quality of the experimental details quite rigorously. The following criteria were consid- during the screening process: ered when discarding references • Clear indication must be given that the acid-base constants were determined ex- perimentally in the actual study, and not taken from another publication. • The calibration method for the pH-electrodes must be indicated. They must have been calibrated in the concentration scale, and not with standard pH-buffers. + That is, “pH” must refer to − log [H ] . References were discarded when they 10 reported mixed acid-base constants, i.e. involving both proton activities and ligand concentrations: − (4 r) [H edta ] r ′ = K r − 5 (r ) [H edta ] a + − 1 r H It should be noted here that it is possible to correct reported mixed protonation + . constants with the estimated value for the single ion activity coefficient of H This has not been done when reviewing the protonation constants of edta. In few cases it is reported that the glass electro des were calibrated with standard buffers, + and [H ] calculated from pH, for example with the Davies equation. This proce- dure was not accepted either. A background electrolyte providing a constant ionic strength must be used. The • ligand must not contribute significantly to the total ionic strength. Therefore in 4 − the case of edta ed that for example [edta] 1 mM ≤ it must be clearly indicat TOT if = 0.1 M. I Temperature, ionic strength, and the nature of the background electrolyte must • be given. edta reported in the following references The acid dissociation constants of H 4 procedure because they did not fulfil one or more of the were discarded from the review [51CAB] criteria indicated above: , , [53CAR/MAR] , [56CAR/STA] , [52CAR/MAR] [57RIN/SII] , [59BEC/GOR] , [59BEC/GOR2] , [59KLY/SMI3] , [60BEC/GOR] , [61SIM] , , [63FUR/GIU] , [63PAL] , [65BOT/CHA] , [66KUL/RAB] , [62BAE/BEN] , , , [67TIK] , [68KUE/SCH] , [68NAA/POD] , [68WAT/SCH] [67TIK2] [66KUL] [68WIK/RIN] [70BAR2] , [71ROR/MAC] , [73CAR/SWA] , [73CHA/RAO] , , [75LIT/NIK] , , [77CHR/BOR] , [75VOT/BAR] , [77KOS/SHE] , [77HOJ/SUG] [77KUB/NIK] , [80MAC/CHI] , [80MIK/HAV] , [81RAJ/MAI] , [81SIR/KAL] , [82AVD/KEA] , , [83WIL/WIL] , [84VAS/GRE] , [85SIN/YAD] , [82HUA/ALF] [86SAL/BOO] , , [88ALL/BOL] , [87THU/KUP] , [89MIR] , [88EVS/SMI] [89NAG/SHA] , [89NES/HOF] , [89SAL/BOO] , [89SVE/DOB] , [90AHR/DAH] , [91DUF/MAR] [91SAL/BOO] , [92TUR/SAN] , [93MAZ/DAN] , [94NAG] , , , , [96SUN/AND] , [97DEL/FIG] , [97VAZ/ATB] , [97VAZ/ATB2] [95SUN/MOT]

441 VIII.3 Acid-base equilibria of edta in aqueous solutions 399 , [98SAL/BOO2] [98SAL/BOO] . These references are in general not discussed in Ap- pendix A. Data from a few other references c ould not be accepted for other reasons. [63TAN/TER] , , These references have been discussed in Appendix A [59KLY/SMI4] [77GRI/GOE] [80JAW] , [85MAR/EVA] , , , [95LIS/CHO] , [78JAW] , [86MAR/EVA] [96BOR/LIS] [96XUE/TRA] . , + + + , K and tetraal- , Na The data reported in the remaining references for Li kylammonium ionic media are listed in Table VIII-2. Some of the publications listed in [58YOS/IGU] Table VIII-2 are further discussed in Appendix A [59IGU/YOS] , , [62KRO/ERM] , , , [78ARE/MUS] , [78MER/GAT] , [75CAR/SWA] [77OYA/MAT] [82OVE/LUN] [82TER/IVA] , [83DAN/RIG] , , , [92GLA/HUL] , [85DAN/RIG] [94KUM/CHA] . contains results for a lithium background Only one publication [85DAN/RIG] electrolyte, cf. Table VIII-2, and these few data were not included in the regressions to obtain selected values of protonation constants. Table VIII-2: Literature data considered by this review on the protonation constants, K log K , log and log K , K log log K (Table VIII-2-a), , log (Table K 10 4 5 3 10 2 10 6 10 1 10 10 4 − VIII-2-b), for edta . Ionic strength values and equilibrium constants are expressed in − 1 mol·L italics are reported in molal units. If the ionic medium is shown in . Data in parenthesis, the contribution of the reacting species to the total ionic strength has been considered. Method I Electrolyte t (°C) Reference K log K log log K 2 10 10 1 10 3 20 2.672 [47SCH/ACK] 10.262 6.161 ise-H 0.1 (KCl) ) 20 10.23 6.16 gl 0.1 (KNO [57SCH/AND] 3 20 10.260 6.170 2.674 [58IRV/SHE] ise-H 0.1 KCl a 25.3 (10.07 ± gl 0.1 KNO (6.13 ± 0.005) [60BOH/MAR] 0.01) 3 gl 0.1 KCl 10.25 6.22 2.86 [63GRI/HUG] 30 b gl 0.1 (KNO ) 20 10.31 6.21 2.66 [66MOE/CHU] 3 30 10.12 6.14 gl 0.1 (Me 10.44 6.16 [67AND] NCl) 20 4 6.16 10.23 gl 0.1 (KNO 2.7 ) 20 3 ise-H 1 (Me NCl) 20 10.12 6.07 2.7 4 ise-H 1 (KCl) 20 9.95 6.26 gl 1 (NaClO ) 20 8.85 6.28 2.3 4 gl 0.1 KCl 20 (10.23 ± 0.02) (6.16 ± 0.01) (2.67 ± 0.02) [67IRV/MIL] (Continued on next page)

442 VIII Discussion of data selection for edta ligand 400 Table VIII-2-a: (continued) t (°C) I Method log K Electrolyte Reference log K K log 3 2 10 10 1 10 2.54 [67MES/VIN] sol 0.1 (K,H)Cl 25 2.57 gl 0.1 (K,H)Cl 25 ± 0.01) (6.18 ± gl 0.1 (KNO (2.69 ± 0.02) [68SIL/SIM] 0.01) ) 25 (10.15 3 c gl 0.1 (KNO 6.16 [69SIL/SIM] 10.20 ) 25 3 6.16 2.7 [70AND/MAL2] gl 0.1 (NaClO ) 20 4 ± 25 (9.04 (7.00 ± 0.05) (2.51 0.05) 0.04) [72LAG/LAG] gl 3 NaClO ± 4 25 10.27 6.19 2.79 [74BAU] gl 0.1 KNO 3 ) 25 (10.25 ± 0.02) (6.16 ± gl 0.1 (KNO (2.77 ± 0.05) [75AND/POD] 0.02) 3 9.78 6.18 [75BRU/KIR] gl 1 (KCl) 25 gl 0.5 (Me 6.12 2.89 [75CAR/SWA] NCl) 25 10.22 4 ± 25 (9.04 (7.00 ± 0.05) (2.51 0.05) 0.04) [75LAG/LAG] gl 3 NaClO ± 4 6.26 2.3 [76AND/MAL] gl 1 (KBr) 20 9.95 6.21 8.85 gl 1 (NaBr) 20 2.36 ± ) 25 ± gl 3 (NaClO ± 0.009) (2.580 (9.060 0.018) [76COR/WAL] 0.005) (7.040 4 25 (9.79 ± 0.01) (6.21 ± 0.01) (2.45 ± 0.01) [76GAT/MER] ise-H 1 (KCl) 0.03) 0.03) (2.63 ± ± [76MAK/STE] gl 0.1 (KCl) 25 (6.05 d 25 (8.63 ± 0.02) (6.36 ± 0.03) (2.64 ± gl 1 NaClO [77OYA/MAT] 0.03) 4 d (2.79 (6.00 ± 0.02) gl 0.15 (NaClO ± 0.02) [78ARE/MUS] ) 37 4 d 25 (9.79 ± 0.01) (6.21 ± 0.01) (2.45 ± 0.01) [78MER/GAT] ise-H 1 (KCl) gl 0.1 KNO [79LET/MAR] 25 10.21 6.11 2.60 3 0.01) ) 35 0.02) (6.16 ± ± [80TAQ/HUS] gl 0.1 (KNO (9.72 3 d 25 (10.26 0.02) (6.14 ± 0.02) (2.64 gl 0.1 KNO 0.02) [82OVE/LUN] ± ± 3 d 0.04) 0.03) (5.94 ± NI 37 (10.07 [83DAN/RIG] gl 0.3 Et ± 4 (10.03 ± 0.04) 0.56 ± 0.03) (5.98 1 ± 0.04) (6.04 ± 0.02) (10.10 (2.50 ± 0.03) [83KRA/DEC] sol 1 ((H,Na)ClO ) 21±1 4 37 (9.120 ± 0.002) (5.883 gl 0.15 NaCl 0.010) [84DUF/MAY] ± ) 25 9.93 6.23 2.58 [84GON/MOT] gl 1 (KNO 3 gl 0.5 (NaClO ) 25 (8.90 ± 0.05) (6.05 ± 0.04)0 (2.48 ± 0.03) [84MED/DOM] 4 d,e ) 25 (9.50 ± 0.03) (6.18 ± 0.03) [85DAN/RIG] gl 0.1 (NaNO 3 0.3 (9.08 ± 0.03) (6.08 ± 0.03) (6.08 ± 0.03) 0.6 ± 0.03) (8.83 1 (8.68 ± 0.03) (6.08 ± 0.03) gl 0.1 (KNO ) 25 (10.18 ± 0.03) (6.19 ± 0.03) 3 0.3 (9.89 ± 0.03) (6.12 ± 0.03) (9.69 ± 0.03) (6.16 ± 0.03) 0.6 1 (9.52 ± 0.03) (6.23 ± 0.03) 0.03) gl 0.1 (Me NBr) 25 (10.35 ± 4 (Continued on next page)

443 VIII.3 Acid-base equilibria of edta in aqueous solutions 401 Table VIII-2-a: (continued) Method t (°C) I Electrolyte log K Reference K log K log 3 10 10 1 10 2 0.03) ± 0.3 (10.21 0.03) 0.6 (10.16 ± ± 0.03) (10.11 1 NBr) 25 (10.37 ± 0.03) gl 0.1 (Et 4 (10.25 0.03) 0.3 ± ± 0.03) (10.24 0.6 ± (10.23 1 0.03) gl 0.1 (Pr ± 0.03) (6.19 NBr) 25 (10.39 0.03) (2.68 ± 0.06) ± 4 ± 0.03) (6.12 ± (10.28 (2.65 ± 0.06) 0.3 0.03) (10.32 ± 0.03) (6.16 ± 0.03) (2.69 ± 0.07) 0.6 1 ± 0.03) (6.22 ± 0.03) (2.77 ± 0.09) (10.37 gl 0.1 (KNO [85SMI/MOT] ) 25 10.19 6.13 3 [88TAQ/HUS] gl 0.1 (KCl) 25 6.240 10.40 gl 1 (KNO 6.17 [92AND] 9.80 ) 25 3 d ) 25 6.00 [92GLA/HUL] gl 0.1 (NaClO 4 ± NCl 25 (10.35 0.01) (6.11 ± 0.01) (2.83 ± 0.01) [93CHE/REI] gl 0.15 Me 4 d NCl) 25 (10.11 ± 0.04) (6.19 gl 0.1 (Me 0.04) (2.87 ± 0.04) [94KUM/CHA] ± 4 (2.48 ) 25 (8.90 0.01) (6.05 ± 0.01) ± ± 0.02) [95CHI/DOM] gl 0.5 (NaClO 4 0.1 (NaCl) 25 (9.40 ± ise-H [95PAL/NGU2] 0.03) 25 (8.684 ± 0.004) (5.98 1 0.01) (2.38 ± 0.03) ± ise-H 1 Me N(F 0.05) CSO ) 25 (9.73 ± 0.02) (6.07 ± 0.02) (2.97 ± 3 3 4 ) 25 (9.498 0.004) (6.161 ± 0.006) (2.94 ± 0.01)1 [96AIZ/NAT] gl 0.1 (NaNO ± 3 NNO [97DEL/QUI] ) 25 (10.22 ± 0.02) (6.05 ± 0.01) (2.71 ± 0.02) gl 0.1 (Me 3 4 gl 0.1 (KCl) 25 2.68 [98SUN/MAR] 6.05 9.93 (2.56 NaCl 25 (9.18 0.3 (6.02 ± 0.02) ± ± 0.02) [99MIZ/BON] gl 0.02) 1 ± 0.02) (6.04 ± 0.02) (2.34 ± 0.07 (8.76 (8.68 ± 0.02) (6.29 ± 0.02) (2.42 ± 0.02) 2 3 ± ± 0.02) (6.51 0.02) (2.55 ± 0.02) (8.68 0.01) ± 0.02) (6.76 ± (8.82 (2.50 ± 0.01) 4 5 (8.96 ± 0.02) (7.01 ± 0.02) (2.625 ± 0.009) d gl 0.01) (5.99 ± 0.03) (2.62 ± 0.04) [2001CHO/BON] 0.1 ± NaCl 25 (9.11 0.3 ± ± 0.04) (6.08 0.01) (2.44 ± 0.2) (9.21 0.01) ± 0.06) (5.89 ± (8.83 (2.38 ± 0.22) 0.5 1 (8.69 ± 0.01) (5.99 ± 0.01) (2.27 ± 0.03) 0.01) ± 0.01) (6.19 ± (8.62 (2.34 ± 0.03) 2 3 (8.60 ± 0.01) (6.41 ± 0.01) (2.23 ± 0.04) 0.01) (9.01 ± 0.03) (6.95 ± 5 (2.46 ± 0.01) (Continued on next page)

444 VIII Discussion of data selection for edta ligand 402 Table VIII-2-b: (Continued) Electrolyte t (°C) log Method K I log Reference K K log 10 5 10 10 4 6 [47SCH/ACK] 1.996 20 ise-H 0.1 (KCl) ise-H 0.1 KCl [58IRV/SHE] 20 1.994 (1.55 ± 0.13) [58TIL/STA] ise-H 0.1 (K)Cl 20 (f) (f) ± (1.3 ± 0.1) 0.1) [58YOS/IGU] , sol 0.1 (K,H)Cl) 30 (2.0 d,f [59IGU/YOS] (f) (f) (1.82 (1.27 ± 0.04) ± 0.5 0.05) 25 2.70 [60OLS/MAR] gl 2 (KCl) 0.96 0.26 sp 2 (Na,H)Cl) 25 d,f (f) − (0.14 ± 0.08) sol 1 (Na,H)NO [62KRO/ERM] ) 25 3 (f) (f) 2 (0.21 ± 0.14) ± (1.4 0.5) (f) (0.26 ± 0.18) 3 30 2.4 [63GRI/HUG] gl 0.1 KCl b [66MOE/CHU] gl 0.1 (KNO ) 20 2.02 3 2.2 1.4 − 0.12 [67AND] ) 20 sol 1 (Na,H)ClO 4 ) 20 2.0 gl 0.1 (KNO 3 ise-H 1 (Me NCl) 20 2.2 4 gl 0.1 KCl ± 0.1) [67IRV/MIL] 20 (1.9 sol 0.1 (K,H)Cl 25 2.14 1.34 [67MES/VIN] gl 0.1 (K,H)Cl 25 2.11 gl 0.1 (KNO 0.1) [68SIL/SIM] ) 25 (2.0 ± 3 ) 20 2 [70AND/MAL2] gl 0.1 (NaClO 4 0.03) ± gl 3 NaClO (1.70 ± 25 (2.13 [72LAG/LAG] 0.03) 4 ± 25 (0.43 sp 3 NaClO 0.05) 4 25 1.99 [74BAU] gl 0.1 KNO 3 ) 25 (2.2 0.1) [75AND/POD] gl 0.1 (KNO ± 3 NCl) 25 1.8 gl 0.5 (Me [75CAR/SWA] 4 gl 3 NaClO 25 (2.13 ± 0.03) (1.70 0.03) [75LAG/LAG] ± 4 sp 3 NaClO 25 (1.2 ± 0.7) (0.43 ± 0.15) 4 sol 3 NaClO 25 (1.75 ± 0.03) (0.15 ± 0.15) 4 gl 1 (KBr) [76AND/MAL] 20 2.2 20 2.04 gl 1 (NaBr) ) 25 (2.215 gl 3 (NaClO 0.022) [76COR/WAL] ± 4 ise-H 1 (KCl) 25 (1.95 ± 0.01) (1.50 ± 0.02) [76GAT/MER] gl 0.1 (KCl) ± 0.04) [76MAK/STE] 25 (1.88 d 0.04) (1.55 ± 0.15) ± [78ARE/MUS] gl 0.15 (NaClO ) 37 (2.10 4 d 25 [78MER/GAT] ise-H 1 (KCl) gl 0.1 KNO 25 2.00 [79LET/MAR] 3 (Continued on next page)

445 VIII.3 Acid-base equilibria of edta in aqueous solutions 403 Table VIII-2-b: (continued) Method t (°C) log I K Electrolyte log Reference K K log 10 5 10 10 4 6 d,f (f) (f) (1.7 ) 20 ± 0.2) ± [82TER/IVA] 0.5) sol 2 (K,H)NO (0.1 3 ) (21 ± 1) (1.95 ± 0.02) (1.45 sol 1 (H,Na)ClO 0.02) (0.10 ± 0.05) [83KRA/DEC] ± 4 gl 1 (KNO ) 25 2.01 [84GON/MOT] 3 ) 25 (1.68 ± 0.02) [84MED/DOM] gl 0.5 (NaClO 4 d,e NBr) 25 (1.96 0.10) [85DAN/RIG] ± gl 0.1 (Pr 4 (1.97 0.3 ± 0.09) ± (2.02 0.6 0.11) (2.09 0.13) 1 ± gl 0.15 Me NCl 25 (2.08 ± 0.02) [93CHE/REI] 4 d NCl) 25 (2.26 ± 0.04) [94KUM/CHA] gl 0.1 (Me 4 gl 0.5 (NaClO ) 25 (1.68 ± 0.01) [95CHI/DOM] 4 ) 25 (1.8 ± 0.2) [96AIZ/NAT] gl 0.1 (NaNO 3 gl 0.1 (Me NNO [97DEL/QUI] ) 25 (2.0 ± 0.1) 3 4 25 2.07 [98SUN/MAR] gl 0.1 (KCl) NaCl 25 (2.14 ± 0.05) [99MIZ/BON] gl 0.3 1 (2.1 ± 0.1) 2 (1.9 ± 0.1) 3 ± 0.00 (2.036 4 (2.1 ± 0.1) 5 (2.26 ± 0.06) d NaCl 25 (2.18 ± [2001CHO/BON] 0.1 gl 0.13) 0.3 ± 0.04) (2.00 0.5 (2.03 ± 0.05) 1 (1.92 ± 0.01) 2 (1.95 ± 0.04) 3 ± 0.08) (1.87 (1.97 ± 0.01) 5 a: Data reported also at 0.1, 13.4 and 42.4°C. b: Data reported also at 40°C. c: Data also reported at 20, 30, 35 and 40°C. d: See comments in Appendix A. were also reported. e: Data in LiNO 3 f: Original experimental data recalculated in this review. Experimental difficulties to consider in the case of edta arise from the fact that the range of acidities that has to be studied is quite wide, ranging from pH ≈ 11 to 1. These are some of the major problems:

446 VIII Discussion of data selection for edta ligand 404 • if glass electrodes are used, they must be specially suited for alkaline measure- ments to avoid alkaline-metal effects; junction potentials may affect the measurements in the extremes of the pH scale • I ≤ 0.1 M); (particularly if I ≤ the ionic strength must be constant. For studies where 0.1 M this is not easy • ≤ 0.1 M. Fur- I es below 2 cannot be reached at to achieve. For example, pH valu − 4 thermore, in the case of edta ≤ 1 mM , it must be clearly indicated that [edta] TOT I if = 0.1 M. Based on the last two items, no equilibrium constant was considered if < 0.1 M. Furthermore, the uncertainty in reported equilibrium constants was increased I during least-squares regressions within the re view process in order to reflect these vary- ing experimental difficulties. Otherwise the uncertainties reported in the original publi- cations were multiplied by a factor (1.96) to obtain error limits closer to a 95% total , including random and possible systematic deviations. The as- uncertainty level, i.e. sed in each of the corresponding Sections signment of uncertainties is further discus below. When applying the SIT model described in Appendix B to the activity coeffi- cients of tetraalkylammonium halides, it may be shown that the specific ion-interaction − + coefficient, ε (R ,X ), depends on the ionic strength ( cf. Section V.3.2). A proper rep- N 4 ++ − − resentation of the data is achieved by setting: ε=ε (RN,X) (RN,X) 414 ++ − ε⋅ + (R N ,X ) log [R N ] . Because of this, the ionic strength dependence of the 24 104 − 4 protonation constants of edta in tetraalkylammonium salts was in general modelled +(4) +(4) −− nn using the SIT model setting: ) (R N ,H edta εε ) = (R N ,H edta 4 14 nn + +(4) n − ) log [R N ] ε⋅ + (R N ,H edta . 24 10 4 n A large number of references report da ta only at 20°C, and in a few cases only to 25°C. The corrections values were extrapolated log K at 30, 35 or 37°C, and these n 10 ected in Section VIII.3 .8. The only available were obtained from reaction enthalpies sel ο 2+ + ο 2+ + information for H was ∆ H (H edta ) and H ∆ H = edta + (H edta ) edta 5 6 rm 5 rm 6 − 1 − ± . For the small temperature corrections performed in this review it (2.7 1.2) kJ·mol + ο (H edta ) H ∆ = was assumed that the enthalpy change could be divided according to rm 5 1 2+ ο − − 1 and (0.2 . The exact parti- (H edta ) H ± ∆ 2.5) kJ·mol − (2.5 ± 2.5) kJ·mol = − rm 6 tioning between these two enthalpies has howev er only a minor effect on the calculated corrections. cf . Sec- Ionic media corrections to the reaction enthalpies consist of two parts ( ∆ε ) term. In tion V.3.6): a Debye-Hückel expressi on, and a specific ion interaction ( L 1 − 3 − it was set equal to (0 ∆ ε kg ⋅ mol ± 5)·10 . The cases where no data was available for L resulting reaction enthalpies are all relatively small, and because of the limited tempera- log K were ture interval involved (from 20 to 37°C), the calculated corrections for n 10 ± 0.17 log -units. The uncertainties in the extrapolated protonation constants ≤ always 10 were also increased following the error-propagation rules described in Appendix C for H ∆ . the additional uncertainty in the value of rm

447 VIII.3 Acid-base equilibria of edta in aqueous solutions 405 Analysis of VIII.3.2 and K K 5 6 The protonation reactions of H edta(aq) are conveniently studied by measuring the 4 rature references are discussed in edta(cr) at different acidities. The lite solubility of H 4 ected data are listed in Tabl e VIII-1 and Table VIII-2. A Section VIII.2.2, and the sel few spectrophotometric and potentiometric determinations of the protonation of cf . Table VIII-2. Potentiometric measurements are edta(aq) have also been reported, H 4 they require low ligand con centrations imposed by the in this case difficult because 3 − 3 − × M) and relatively high acidities (3 edta(cr) (< 10 to 1 M). Some 10 solubility of H 4 [78MER/GAT] were performed , [77OYA/MAT] of the potentiometric investigations ese references were under conditions where precipitation could have taken place, and th not considered in this review (see also Appendix A). Because of the inherent experimental difficulties in these systems, the uncer- tainties for the 95% confidence level were assigned as follows: when no uncertainty ± 0.3 log -units was used; value was reported in the individual publication, a value of 10 ± 0.14 log -units were increased to that level. all reported uncertainties below 10 edta(aq), For the first protonation constant of H 4 + + , (VIII.2) U edta(aq) + H edta H H 5 4 cf . Section V.3.2): the SIT equations may be re-written as follows ( 2* ο +− I zD ε (H , X ) log ε K log K I −∆ −∆ − = 5 5 10 5 m m 10 − +− * 2 = 0 and z ε ∆ = is the anion of ε (H edta , X ) ε (H edta(aq), MX) ∆ − , and X where 5 54 the background electrolyte. The data were treated using a weighted least squares regres- sion procedure. The regression plot is shown in Figure VIII-7. There appears to be no significant differences between data in the different media, mainly perchlorates and chlorides. The results of the linear regression were: − ο * 1 K ε ± 0.1) and log . (VIII.2) = (1.3 ∆ = (0.06 ± 0.04) kg·mol 5 10 5 Figure VIII-7: Linear least squares SIT-regression plot for the Reaction (VIII.2). Data from Table VIII-2 have been converted to molal units in the plot, and extrapolated to 25°C when necessary. 2.0 [58TIL/STA] [59IGU/YOS] [60OLS/MAR] 1.5 m [62KRO/ERM] I ) [67AND] − [67MES/VIN] ,X 1.0 + [ 72LAG/ LAG] (H [ 75LAG/ LAG] ε − [76GAT/MER] 0.5 5 K [78ARE/MUS] + + 10 edta H H edta(aq) + H U 5 4 [82TER/IVA] + + log 0.0 [83KRA/DEC] at (20 – 37) ° C H edta(aq) + H edta z H 5 4 at (20 - 37)°C −0.5 01234 I / molal m

448 VIII Discussion of data selection for edta ligand 406 + For the protonation of H : edta 5 + + 2+ H U H + H edta edta , (VIII.3) 6 5 * −+− + 2 − 2 ∆ − ε (H edta , X ) ε (H edta , X ) ε , where X = 2 and z ∆ is the anion of the = 6 65 ed with a weighted least squares regres- background electrolyte. The data was again treat sion procedure. The regression plot is shown in Figure VIII-8. Again there appears to be no significant differences between data in the different media, mainly perchlorates and nitrates. The results of the linear regression were: 1 − ο * (0.5 (VIII.3) = − . ± 0.2), and log K ε ∆ = (0.03 ± 0.06) kg ⋅ mol 6 6 10 Figure VIII-8: Linear least squares SIT-regression plot for Reaction (VIII.3). Data from Table VIII-2 have been converted to molal units in the plot, and extrapolated to 25°C when necessary. 0.5 2+ + + + 2+ + [60OLS/MAR] edta H edta + H U H 6 5 + H edta z H H edta [62KRO/ERM] 6 5 m ° at (20 – 25) C I [67AND] ) at (20 - 25)°C − 0.0 [72LAG/LAG] ,X + [75LAG/LAG] (H [82TER/IVA] ε [83KRA/DEC] − −0.5 D 2 − 6 K 10 −1.0 log −1.5 01234 / molal I m

449 VIII.3 Acid-base equilibria of edta in aqueous solutions 407 Analysis of VIII.3.3 K 4 − , edta For the protonation constant of H 3 − + H + H H edta edta(aq) (VIII.4) U 4 3 the SIT equation may be expressed as follows ( cf . Section V.3.2): 2* ο +− = −∆ − log ε (H , X ) log ε K zD I K I −∆ 10 m 10 m 4 4 4 * − + 2 + − ε (H edta(aq), MX) ε (M , H edta ) ∆ ε with is the cation = − = 2 and ∆ , where M z 4 43 of the background electrolyte. This reaction has been investigated in several studies by edta(cr) at varying acidities, and the corresponding litera- measuring the solubility of H 4 ture references are listed in Table VIII-1. Al l data considered in the review procedure, including potentiometric determinations, are listed in Table VIII-2. There are two major difficulties in determining this equilibrium constant: + 0.01 M), and this might introduce sys- ] ≥ i) relatively high acidities are needed ([H ii) several protonated species coexist in the tematic errors as junction potentials; and studied solutions, introducing difficulties in the interpretation of the data. In the major- + was neglected. These different effects are edta ity of the studies the presence of H 5 data, as evidenced in Figure VIII-9. There- reflected in a somewhat large spread of the fore, the uncertainties in the data had to be increased to a 95% confidence level as fol- -units ± 0.3 log lows: in cases where no uncertainty limits were reported, a value of 10 0.1 in ± was used in the weighted least-squares procedure; reported uncertainties below K log ± 0.2 log -units in the regression in the original papers were increased to 10 4 10 ± 0.1 × 1.96). ≈ analysis ( The data was treated with a weighted multi-linear least-squares regression ο should fit all the log K procedure. This procedure assumes that a common value of 4 10 * ε ∆ was independent data. For this reaction it was found that it could be assumed that 4 on ionic strength for all electrolytes, including tetraalkylammonium salts. The regres- selected values are listed in Table VIII-3. sion plots are shown in Figure VIII-9 and the ο * ∆ ε log K (VIII.4) and (VIII.4). Table VIII-3: Selected values of 4 4 10 ο K (VIII.4) = (2.23 ± 0.05) log 4 10 a * 1 − ∆ ε ⋅ mol Medium ) (kg 4 + Na 0.02) ± (0.04 + K (0.15 ± 0.10) − + ± N 0.12) − (0.03 R 4 + a: R N represents tetraalkylammonium. 4

450 VIII Discussion of data selection for edta ligand 408 − edta Figure VIII-9: Multi-lin + on plots for the reaction: H ear least squares SIT-regressi 3 + U H edta(aq). Data from Table VIII-2 have been converted to molal units in the H 4 plots, and extrapolated to 25°C when necessary. [67AND] [70AND/MAL2] + − m – + 3.0 z H edta(aq) H + H edta I [72LAG/LAG] H + H edta(aq) edta U H 4 3 3 4 ) − + [75LAG/LAG] + electrolytes at (20 - 37)°C in Na ,X C ° electrolytes at (20 – 37) in Na [76AND/MAL] + [76COR/WAL] (H ε [78ARE/MUS] 2.5 − [83KRA/DEC] D [84MED/DOM] + 2 [95CHI/DOM] 4 K [96AIZ/NAT] 2.0 10 [99MIZ/BON] log [2001CHO/BON] 1.5 012345 + [Na ] / molal – + + − [47SCH/ACK] [68SIL/SIM] + H H edta H U edta(aq) 3 4 H H z + H edta(aq) edta m [74BAU] [58IRV/SHE] 4 3 3.0 I + + ) C ° electrolytes at (20 – 30) in K [75AND/POD] [58YOS/IGU, − electrolytes at (20 - 30)°C in K 59IGU/YOS] [76AND/MAL] ,X + [76GAT/MER] [60OLS/MAR] (H ε [63GRI/HUG] [76MAK/STE] 2.5 − [79LET/MAR] [66MOE/CHU] D [67AND] [84GON/MOT] [98SUN/MAR] [67IRV/MIL] + 2 4 [67MES/VIN] K 2.0 10 log 1.5 0.0 1.5 1.0 0.5 2.0 + [K ] / molal + – + − [67AND] H edta U H + H edta(aq) m 3 4 3.0 I H edta(aq) + H edta z H [75CAR/SWA] 3 4 ) + − + C ° electrolytes at (20 – 25) N in R [85DAN/RIG] 4 ,X in R N electrolytes at (20 - 25)°C + [93CHE/REI] 4 (H [94KUM/CHA] ε 2.5 [97DEL/QUI] − D + 2 4 K 2.0 10 log 1.5 0.00.20.40.60.81.01.21.4 + N [R ] / molal 4

451 VIII.3 Acid-base equilibria of edta in aqueous solutions 409 K VIII.3.4 Analysis of 3 2 − edta For the protonation constant of H , 2 2 − + − H edta H (VIII.5) + H edta U 3 2 2 * +−+ − + ε ε (M , H edta ) ∆ (M , H edta ) = − , where M ε is the cation of the background 32 3 2 4. The data was treated with a weighted multi-linear least- ∆ = − z electrolyte, and ο K log squares regression procedure. This procedure assumes that a common value of 10 3 should fit all the data. The uncertainties were increased to obtain a consistent set of data ere no uncertainty was given and to correspond to a 95% confidence level. In cases wh 0.15 log -units was used in the weighted least- ± in the original publication, a value of 10 ο in the original pa- ± K log 0.03 in squares procedure. Reported uncertainties below 3 10 ± 1.96). × ≈ 0.06 log ± 0.03 -units in the regression analysis ( pers were increased to 10 + + between edta-anions and either Na Specific ion interaction parameters or K were assumed to be independent of ionic strength. In the case of tetraalkylammonium it ∆ε -values were dependent on ionic strength according to the rela- was assumed that the * * * ε ε ∆ ∆ + . There is no significant difference between the data in = ∆ε log I tion: 10 m 2 1 tetramethylammonium and tetrapropylammonium and they were fitted to the same * ∆ε .The regression plots are shown in Figure VI II-10 and the selected values are listed in Table VIII-4. ο * log (VIII.5). ∆ (VIII.5) and K ε Table VIII-4: Selected values of 10 3 3 ο log K (VIII.5)= (3.15 ± 0.02) 3 10 a * − 1 ∆ ε Medium (kg·mol ) 3 + (0.04 ± 0.01) Na + K ± 0.04) (0.03 + I N ± − (0.27 ± 0.06) + (1.0 0.3) log R 10 4 m + a: R N represents tetramethylammoium and tetrapropylammonium. 4

452 VIII Discussion of data selection for edta ligand 410 − 2 for the reaction: H Figure VIII-10: Multi-linear least squares edta SIT-regression plots 2 + − * ε edta , assuming U to be independent of ionic strength for sodium and H ∆ + H 3 3 * ε ∆ = ( I ) for tetraalkylammonium media, see text for f potassium electrolytes, and m 3 been converted to molal units in the plots, details. Data at 25°C from Table VIII-2 have and extrapolated to 25°C when necessary. 4.0 [67AND] 2– – + [70AND/MAL2] − 2 + − edta H U H + H edta 2 3 m I z H + H H edta edta [72LAG/LAG] 2 3 ) + − [75LAG/LAG] + ° in Na electrolytes at (20 – 37) C in Na electrolytes at (20 - 37)°C ,X [76AND/MAL] + 3.5 [76COR/WAL] (H ε [77OYA/MAT] − [78ARE/MUS] D [83KRA/DEC ] + 4 [84MED/DOM] 3 3.0 K [95CHI/DOM] 10 [95PAL/NGU2] log [96AIZ/NAT] [99MIZ/BON] [2001CHO/BON] 2.5 012345 + [Na ] / molal 4.0 [75AND/POD] [47SCH/ACK] + – 2– [76AND/MAL] [58IRV/SHE] − + 2 − H edta + H edta H U m 3 2 edta H H z edta + H I [76GAT/MER] [63GRI/HUG] 3 2 ) + − [76MAK/STE] [66MOE/CHU] + C electrolytes at (20 – 30) in K ° in K electrolytes at (20 - 30)°C ,X [78MER/GAT] [67AND] + 3.5 [67IRV/MIL] [79LET/MAR] (H ε [82OVE/LUN] [67MES/VIN] − [84GON/MOT] [68SIL/SIM] D [74BAU] [98SUN/MAR] + 4 3 3.0 K 10 log 2.5 0.2 0.4 0.0 0.8 1.0 0.6 + [K ] / molal 4.0 [67AND] m [75CAR/SWA] I ) [85DAN/RIG] − [93CHE/REI] ,X + 3.5 [94KUM/CHA] (H ε [95PAL/NGU2] − [97DEL/QUI] D + 4 3 3.0 2– – + K edta edta U H + H H 10 3 2 2 + − − edta H H z edta + H + log 2 3 C N in R ° electrolytes at (20 – 25) 4 + in R electrolytes at (20 - 25)°C N 4 2.5 0.8 1.0 1.2 1.4 0.2 0.0 0.4 0.6 + ] / molal N [R 4

453 VIII.3 Acid-base equilibria of edta in aqueous solutions 411 Analysis of K VIII.3.5 2 3 − , For the protonation constant of Hedta − − + 3 2 Hedta H + H (VIII.6) edta U 2 * 23 −+ −+ + ∆ (H edta , M ) ε (Hedta , M ) = ε − , where M is the cation of the background ε 2 2 2 z 6. The data was treated using weighted multi-linear least- = − ∆ electrolyte, and ο log K should fit squares regression. This procedure assumes that a common value of 10 2 to obtain a consistent set of data and to all the data. The uncertainties were increased ce level. In cases where no uncertainty was given in the correspond to a 95% confiden ± 0.15 log -units was used in the weighted least-squares original publication, a value of 10 ο 0.05 in log K ± in the original papers were procedure. Reported uncertainties below 2 10 ± 0.10 log 1.96). -units in the regression analysis ( ≈ ± 0.05 × increased to 10 + + between edta-anions and either Na Specific ion interaction parameters or K were assumed to be independent on ionic strength. In the case of tetraalkylammonium it ∆ε -values were dependent on ionic strength according to the rela- was assumed that the *** II ∆=∆+∆⋅ εεε log . There is no significant difference between the data in tion: m1210m + + N N , and they were fitted , and tetrapropylammonium, Pr tetramethylammonium, Me 4 4 * . The regression plots are shown in Figure VIII-11 and the selected val- ∆ε to the same ues are listed in Table VIII-5. ο * log K (VIII.6) and ε ∆ (VIII.6). Table VIII-5: Selected values of 2 10 2 ο 0.02) log K (VIII.6)= (6.80 ± 2 10 a * − 1 ) ε ∆ (kg ⋅ mol Medium 2 + − (0.27 ± 0.01) Na + − (0.48 ± 0.04) K + 0.05) + (0.3 ) − (0.35 ± I ± 0.3) log N ( R 10 m 4 + a: R N represents tetramethyl- and tetrapropyl-ammoium. 4

454 VIII Discussion of data selection for edta ligand 412 − 3 T-regression plots for the reaction: Hedta Figure VIII-11: Multi-linear least squares SI * + 2 − ∆ edta , assuming U ε H to be independent of ionic strength for sodium and + H 2 2 * ε ∆ = I ) for tetraalkyl ammonium media, see text for ( f potassium electrolytes, and m 2 ± 5)°C from Table VIII-2 have been converted to molal units in the details. Data at (25 to 25°C when necessary. plots, and extrapolated [67AND] 8.0 [70AND/MAL2] m I [72LAG/LAG] ) − [75LAG/LAG] ,X + [76AND/MAL] 7.5 (H [76COR/WAL] ε [77OYA/MAT] − [78ARE/MUS] D [84DUF/MAY] 7.0 + 6 2 3– 2– + [84MED/DOM] − 2 3 − + edta U + H Hedta H K 2 z edta H + H Hedta [85DAN/RIG] 10 2 + + [92GLA/HUL] ° in Na electrolytes at (20 – 37) C log in Na electrolytes at (20 - 37)°C [95CHI/DOM] 6.5 [95PAL/NGU2] 012345 [96AIZ/NAT] [99MIZ/BON] + [2001CHO/BON] [Na ] / molal [47SCH/ACK] [78MER/GAT] 8.0 [57SCH/AND] [79LET/MAR] 3– + 2– m − 3 − + 2 I H + H edta Hedta U 2 [80TAQ/HUS] [58IRV/SHE] z Hedta + H ) edta H − 2 + [60BOH/MAR] [82OVE/LUN] + electrolytes at (20 – 35) in K C ° ,X + [63GRI/HUG] electrolytes at (20 - 35)°C in K [84GON/MOT] 7.5 (H [66MOE/CHU] [85DAN/RIG] ε [67AND] [85SMI/MOT] − [88TAQ/HUS] [67IRV/MIL] D [92AND] [68SIL/SIM] 7.0 + 6 2 [98SUN/MAR] [69SIL/SIM] K [74BAU] 10 [75AND/POD] log [75BRU/KIR] 6.5 [76AND/MAL] 0.6 1.2 0.2 0.4 0.8 0.0 1.0 [76GAT/MER] [76MAK/STE] + ] / molal [K 3– + 2– − − 3 + 2 8.0 [67AND] U H edta Hedta + H Hedta H + H edta z 2 m 2 I [75CAR/SWA] + ) + − [83DAN/RIG2] C N ° electrolytes at (20 – 37) in R 4 N electrolytes at (20 - 37)°C in R 4 ,X [85DAN/RIG] + 7.5 [93CHE/REI] (H ε [94KUM/CHA] − [95PAL/NGU2] D [97DEL/QUI] 7.0 + 6 2 K 10 log 6.5 0.8 0.6 0.4 0.2 0.0 1.4 1.2 1.0 + [R N ] / molal 4

455 VIII.3 Acid-base equilibria of edta in aqueous solutions 413 Analysis of K VIII.3.6 1 4 − For the first protonation constant of edta , − + 4 − 3 edta (VIII.7) Hedta + H U 34 * +−+− 2 + z ε ∆ = ε ε (M , Hedta ) = (M , edta ) − 8 and , where M ∆ is the cation of the − 1 ο log K were obtained from SIT- background electrolyte. Different values of 1 10 regressions of the data obtained in background electrolytes with the different cations: + + , K , and tetraalkylammonium. This would be expected if alkali metal complexes Na − 4 V.3.3). Therefore an additional (see also Sections VIII.4 and are formed with edta K as compared with the other uncertainty exists in the determination of the value of 1 protonation constants of et hylenediaminetetraacetate. Several stoichiometries have been postulated for the complexes between + − + 2 4 − − 3 − 2 , Na(Hedta) , for example: Na(edta) , Na edta and edta / K as sodium com- Na 2 plexes (see Section VIII.4.1). The most commonly accepted stoichiometry is however the formation of complexes with ratio 1:1, that is: 3 − [Medta ] 4 − 3 − + U K Medta = + edta M M 4 + − [M ][edta ] + + + or K . The following constants are select ed by this review in Section = Na with M VIII.4.1: − 3 ο ο (Na(edta) 0.2), log K β ± = , 298.15 K) = (2.8 log 10 1 N 10 a ο ο 3 − log log K β (Kedta 0.3). = , 298.15 K) = (1.8 ± 10 10 1 K When evaluating the ionic-strength dependence of data for the first protonation 4 − constant of edta in presence of a background cation that forms complexes, the follow- ing equations are applicable ( cf. Section V.3.3): †2 o* +− + ε (H , X ) log ε log (1 [M ]) KzD log K I K + −∆ − = −∆ − I 10 1 10 1 10 M 1 mm +− o2 * K I I KzD log log εε (M , X ) =+∆−∆+ M M 10 M 10 mm † K is the protonation constant determined without taking into account the for- where 1 −+− *34 + mation of alkali-metal complexes, εε (M , edta ) ε (M , Hedta ) ∆= − , 1 *34 +−+− 2 εε (M , Medta ) ε (M , edta ) ∆= − ∆ z , and = 8 for both reactions (protonation − M and complex formation with alkali-metals). The data was treated using weighted least-squares regression procedures. In all ο cases it was assumed that a common value of should fit all the data. The un- log K 1 10 certainties were increased to obtain a consistent set of data and to correspond to a 95% confidence level. In cases wher e no uncertainty was given in the original publication, a value of ± 0.15 log -units was used in the weighted least-squares procedure. Reported 10 ο K log uncertainties below ± 0.05 in in the original papers were increased to ± 0.10 10 1 log 1.96). Specific ion interaction parame- -units in the regression analysis ( ≈ ± 0.05 × 10 4 + − + 3 − ters between Na or Hedta and either edta and K were assumed to be independent

456 VIII Discussion of data selection for edta ligand 414 on ionic strength. In the case of tetraalkylammonium media it was assumed that the *** individual log ∆=∆+∆⋅ . There is no εεε -values depended on ionic strength: I ε 1210m + significant difference between the data in tetramethylammonium, Me , tetraethylam- N 4 + + * monium, Et and tetrapropylammonium, Pr N , and they were fitted to the same ∆ε . N 4 4 ο ο 3 − − 3 (Na(edta) (Kedta ) and ) were set K K During the regression, the values of * * ε ∆ ∆ ε to those recommended in Section VI II.4.1. Because the values of are and 1 M . discussion in Section V.3.3), it was necessary to fix the values of cf highly correlated ( * ∆ to those obtained in tetraalkylammonium media ( cf . Section VIII.4.1), namely: ε M * * 1 − − 1 ε ε ∆ ∆ = − and (0.24 . The results from ± 0.47) kg = − (0.50 ± 0.64) kg ⋅ mol ⋅ mol K Na the fitting of the remaining five parameters are given in Table VIII-6. The agreement between the experimental data and this model is shown in Figure VIII-12. Table VIII-6: Results of non-linear weighed least-squares fitting of literature data for 4 − the first protonation of edta at (25 5)°C. Five parameters were adjusted in the ± ο * − 1 ∆ K ε log regression, keeping constant the following: , − 0.24 kg ⋅ mol = 2.8, = Na 10 Na 1 − * ο . = 1.8, K ε ∆ = − 0.50 kg ⋅ mol log K K 10 ο log K 0.03) (VIII.7) = (11.24 ± 1 10 a * 1 − ε (kg ⋅ mol Medium ) ∆ 1 + − (0.42 ± 0.02) Na + 0.05) − (0.76 ± K + ± − (0.40 N 0.05) + (0.2 ± 0.3) log I R 4 m 10 + a: R N represents tetramethylammonium, tetraethylammonium 4 and tetrapropylammonium.

457 VIII.3 Acid-base equilibria of edta in aqueous solutions 415 4 − + Figure VIII-12: Weighted le ast squares SIT-regression plots for the reaction: edta + − 3 − 3 − 3 H and Kedta a) the formation of Na(edta) ; b) that the U Hedta , assuming: * ∆ ε values of are independent of ionic strength for sodium and potassium electrolytes, 1 * and c) that = f ε I ∆ ) for tetraalkylammonium media (see text for details). Data from ( m 1 Table VIII-2 have been converted to molal units in the plots, and extrapolated to 25°C when necessary. [67AND] 12.0 [72LAG/LAG] [75LAG/LAG] m I 4 − 3 − + + 3– 4– [76AND/MAL] ) z edta + H Hedta Hedta edta + H U − 11.5 [76COR/WAL] + ,X + + [77OYA/MAT] electrolytes at (20 - 37)°C in Na C in Na electrolytes at (20 – 37) ° (H [84DUF/MAY] ε [84MED/DOM] − 11.0 [85DAN/RIG] D [95CHI/DOM] + 8 † [95PAL/NGU2] 1 K [96AIZ/NAT] 10.5 10 [99MIZ/BON] log [2001CHO/BON] 10.0 012345 + ] / molal [Na [47SCH/ACK] [76AND/MAL] 12.0 [57SCH/AND] [76GAT/MER] [58IRV/SHE] [78MER/GAT] m [60BOH/MAR] [79LET/MAR] I ) 11.5 [63GRI/HUG] − [80TAQ/HUS] [66MOE/CHU] [82OVE/LUN] ,X + [84GON/MOT] [67AND] (H ε [85DAN/RIG] [67IRV/MIL] 11.0 − [68SIL/SIM] [85SMI/MOT] D [69SIL/SIM] [88TAQ/HUS] [92AND] [74BAU] + 8 † 1 4 + 3 − − [75AND/POD] [98SUN/MAR] 10.5 3– 4– + K + H z Hedta edta edta Hedta U + H 10 [75BRU/KIR] + + in K electrolytes at (20 - 35)°C log C ° in K electrolytes at (20 – 35) 10.0 1.0 1.2 0.4 0.2 0.6 0.8 0.0 + ] / molal [K 12.0 [67AND] [75CAR/SWA] [83DAN/RIG2] m I NCl) [85DAN/RIG] (Me 11.5 ) 4 − NBr) [85DAN/RIG] (Et 4 ,X + [85DAN/RIG] (Pr NBr) 4 (H ε [93CHE/REI] 11.0 − [94KUM/CHA] D [95PAL/NGU2] 4– + 3– [97DEL/QUI] 4 − + − 3 edta U + H Hedta + 8 1 Hedta + H z edta 10.5 K + + 10 N in R ° electrolytes at (20 – 37) C 4 electrolytes at (20 - 37)°C N in R 4 log 10.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.4 + ] / molal [R N 4

458 VIII Discussion of data selection for edta ligand 416 4 − VIII.3.7 Selected protonation constants for edta From the discussions in previous Sections, it follows that the selected values for the 4 − protonation of edta , 4) − r ( r − 5) ( + U edta + H edta H (VIII.8) H − r r ( 1) obtained in this review are: ο ο K log log β ± 0.03) 0.03) ± ((VIII.8), r = 1, 298.15 K) = (11.24 = (11.24 1 10 10 1 ο ο β r ± K log ((VIII.8), = 2, 298.15 K) = (6.80 = (18.04 ± 0.04) log 0.02) 10 2 10 2 ο ο log ((VIII.8), r K ± 0.02) 0.04) ± log = 3, 298.15 K) = (3.15 β = (21.19 3 3 10 10 ο ο K log log β ± = 4, 298.15 K) = (2.23 r ((VIII.8), = (23.42 ± 0.07) 0.05) 4 4 10 10 ο ο β r ± 0.1) log ((VIII.8), = 5, 298.15 K) = (1.3 = (24.72 ± 0.12) log K 5 10 5 10 ο ο (0.5 ((VIII.8), r = 6, 298.15 K) = − log ± 0.2) 0.23). ± log K β = (24.22 6 10 10 6 It must be recalled that these constants should be used together with the forma- tion constants for alkali metal complexation: − 3 [Na(edta) ] o 3 − 4 + − Na U Na(edta) + edta ==± log log (2.8 0.2) K 10 10 Na 4 +− [Na ][edta ] − 3 [Kedta ] o + − 4 3 − K ==± . + edta Kedta log (1.8 0.3) K log U 10 K 10 4 +− [K ][edta ] The selected specific ion interaction co efficients are listed in Table VIII-7. * They were obtained using the values listed in previous subsections and ε ∆ n 1 − (H edta, MX) = 0.14) kg·mol − (0.29 , obtained in Section VIII.2.2. ε ± 4 − 4 − 1 ) for edta and its mol ⋅ teraction coefficients (kg Table VIII-7: Specific ion in + protonated species. R N represents tetraalkylammonium. 4 − − + + − − + + Na N M X K = Cl R , = or NO ClO 4 3 4 4 +− ε (0.32 ± 0.14) (1.07 ± 0.19) (0.76 ± (M , edta ) − (1.5 ± 0.5) log ) ( I 0.21) m 10 3 +− − ε − (0.10 ± 0.14) (0.31 ± 0.18) (0.36 ± 0.20) (M , Hedta ) (1.3 ± 0.4) log ) ( I m 10 +− 2 0.18) ε − (0.37 ± 0.14) − (0.17 ± (M , H edta ) (0.01 ± 0.19) − (1.0 ± 0.43) log ) ( I m 10 2 +− (M , H edta ) ε − (0.33 ± 0.14) − (0.14 ± 0.17) − (0.26 ± 0.18) 3 0.14) ε − (0.29 ± 0.14) − (0.29 ± (H edta, MX) − (0.29 ± 0.14) − (0.29 ± 0.14) 4 +− (H edta , X ) 0.15) ε − (0.23 ± 5 +− 2 (H edta , X ) ε 0.16) − (0.20 ± 6

459 VIII.3 Acid-base equilibria of edta in aqueous solutions 417 Table VIII-8-a reports the values for the protonation constants and for the so- dium or potassium formation constants for some ionic media commonly used in chemi- cal equilibrium studies. Values for the “apparent” first protonation constant are given in Table VIII-8-b for the same set of background electrolytes. The calculated distribution of the different protonated forms of ethyle nediaminetetraacetate as a function of pH in 1 M NaCl is shown in Figure VIII-13. Figure VIII-13: Calculated distribution of dissolved ethy lenediaminetetraacetate species as a function of pH in 1 M NaCl at 25°C.

460 VIII Discussion of data selection for edta ligand 418 4 − Table VIII-8-a Calculated values for the protonation constants of edta and for alkali- + + in some Na metal complex formation in Molar units electrolytes at 25 and 20°C. / K ents has been used with the The SIT model for activity coeffici values obtained in ∆ε this review. Care should be exercised when using these values: (a) data in sodium media at the highest ionic strengths are confirmed by few experimental data only, as indicated in the figures of Sections VIII.3.2 to VIII.3.6; (b) there is no experimental support for K the values in potassium media at ionic strengths above 1 molal; and (c) values of and 5 K have only been measured at I ≤ 3.5 molal. 6 NaClO ° C at 25 4 log I K I log K K log log K 1 10 Na 2 10 10 m 3 10 molal M ± 0 0.000 (2.80 ± 0.03) (6.80 ± 0.02) (3.15 ± 0.02) 0.20) (11.24 0.1 0.101 (1.95 0.21) (10.42 ± 0.03) (6.19 ± 0.02) (2.73 ± 0.02) ± ± 0.23) (10.22 0.03) (6.03 ± 0.02) (2.60 ± 0.02) 0.25 0.254 (1.70 ± ± ± 0.03) (5.97 ± 0.02) (2.51 ± 0.02) 0.5 0.513 (1.53 0.31) (10.13 ± 0.42) (10.15 ± 0.04) (5.98 0.75 0.779 (1.46 0.03) (2.47 ± 0.03) ± 1 1.05 (1.44 0.53) (10.21 ± 0.04) (6.02 ± 0.03) (2.45 ± 0.03) ± ± 1.1) (10.65 ± 0.07) (6.34 ± 0.05) (2.48 ± 0.05) 2 2.21 (1.5 3 3.50 (1.7 ± 1.7) (11.27 ± 0.10) (6.80 ± 0.08) (2.57 ± 0.08) 4 4.95 (2.0 ± ± 0.14) (7.36 ± 0.11) (2.69 ± 0.11) 2.3) (12.02 ± ± 0.19) (8.00 ± 0.15) (2.85 ± 0.15) 3.1) (12.89 5 6.58 (2.4 I K I log log K K log 10 5 4 10 m 10 6 M molal ± 0.05) (1.30 ± 0.10) − (0.50 ± 0 0.000 (2.23 0.20) 0.1 0.101 (2.02 ± 0.05) (1.31 ± 0.10) − (0.27 ± 0.20) 0.25 0.254 (1.97 ± 0.05) (1.33 ± 0.10) − (0.17 ± 0.20) 0.5 0.513 (1.94 ± ± 0.10) − (0.08 ± 0.20) 0.05) (1.35 ± 0.11) ± 0.75 0.779 (1.94 − (0.01 ± 0.21) 0.05) (1.38 ± 0.06) (1.41 ± 0.11) (0.05 ± 0.21) 1 1.05 (1.95 2 2.21 (2.03 ± 0.08) (1.52 ± 0.14) (0.26 ± 0.24) 3 3.50 (2.15 ± ± 0.19) (0.45 ± 0.30) 0.11) (1.65 4 4.95 (2.30 0.15) (1.79 ± 0.24) (0.66 ± 0.37) ± 5 6.58 (2.47 ± 0.19) (1.95 ± 0.31) (0.88 ± 0.46) (Continued on next page)

461 VIII.3 Acid-base equilibria of edta in aqueous solutions 419 Table VIII-8-a: (continued) NaNO ° C at 25 3 log I I log K log K K log K 10 10 2 1 10 m 3 Na 10 molal M 0.20) (11.24 0 0.000 (2.80 0.03) (6.80 ± 0.02) (3.15 ± 0.02) ± ± ± ± 0.03) (6.18 ± 0.02) (2.72 ± 0.02) 0.1 0.101 (1.95 0.21) (10.42 0.23) (10.20 ± 0.03) (6.02 ± 0.02) (2.58 ± 0.02) 0.25 0.253 (1.69 ± ± 0.31) (10.09 ± 0.5 0.509 (1.51 ± 0.02) (2.47 ± 0.02) 0.03) (5.93 0.75 0.769 (1.42 0.41) (10.09 ± 0.03) (5.92 ± 0.02) (2.41 ± 0.02) ± ± 0.53) (10.12 0.04) (5.94 ± 0.02) (2.38 ± 0.02) 1 1.03 (1.38 ± ± ± 0.06) (6.16 ± 0.04) (2.31 ± 0.04) 2 2.14 (1.4 1.0) (10.45 ± 1.6) (10.93 ± 0.08) (6.48 3 3.33 (1.5 0.05) (2.30 ± 0.05) ± 4 4.61 (1.7 2.2) (11.49 ± 0.11) (6.88 ± 0.07) (2.31 ± 0.07) ± ± 2.8) (12.13 ± 0.14) (7.32 ± 0.09) (2.34 ± 0.09) 5 6.02 (1.9 C ° at 25 NaNO 3 I I log K log K log K 5 10 10 m 10 6 4 molal M ± 0.05) (1.30 ± 0.10) − (0.50 0 0.000 (2.23 0.20) ± 0.1 0.101 (2.02 0.05) (1.30 ± 0.10) − (0.27 ± 0.20) ± ± − ± 0.10) 0.25 0.253 (1.95 (0.19 ± 0.20) 0.05) (1.31 0.5 0.509 (1.90 0.05) (1.31 ± 0.10) − (0.12 ± 0.20) ± 0.75 0.769 (1.88 ± 0.05) (1.32 ± 0.10) − (0.07 ± 0.21) 1 1.03 (1.87 ± ± 0.11) − (0.03 ± 0.21) 0.06) (1.32 ± ± 2 2.14 (1.86 0.13) (0.08 ± 0.24) 0.07) (1.35 3 3.33 (1.88 0.09) (1.38 ± 0.17) (0.18 ± 0.28) ± 4 4.61 (1.91 ± 0.11) (1.41 ± 0.21) (0.26 ± 0.34) 5 6.02 (1.96 ± ± 0.27) (0.35 ± 0.42) 0.14) (1.44 (Continued on next page)

462 VIII Discussion of data selection for edta ligand 420 Table VIII-8-a: (continued) C NaCl at 25 ° log I K log K I K log K log 10 10 2 10 m 10 3 1 Na M molal ± 0.03) (6.80 ± 0.20) (11.24 ± 0.02) 0 0.000 (2.80 ± 0.02) (3.15 0.21) (10.42 ± 0.1 0.100 (1.95 ± 0.02) (2.72 ± 0.02) ± 0.03) (6.18 ± ± 0.03) (6.03 ± 0.02) (2.59 ± 0.02) 0.25 0.252 (1.71 0.23) (10.21 0.31) (10.12 0.02) (2.49 0.03) (5.95 ± ± ± 0.02) 0.5 0.506 (1.54 ± ± ± 0.03) (5.95 ± 0.02) (2.45 ± 0.02) 0.75 0.762 (1.47 0.41) (10.12 ± 0.52) (10.16 ± 0.04) (5.98 ± 0.02) (2.42 1 1.02 (1.45 0.02) ± 2 2.09 (1.5 1.0) (10.53 ± 0.06) (6.24 ± 0.04) (2.41 ± 0.04) ± ± 1.5) (11.02 0.08) (6.59 ± 0.05) (2.45 ± 0.05) 3 3.20 (1.7 ± 2.1) (11.58 0.06) (2.51 0.10) (7.00 ± ± ± 0.06) 4 4.37 (2.0 ± ± ± 0.13) (7.45 ± 0.08) (2.59 ± 0.08) 5 5.61 (2.2 2.6) (12.20 I K log I K log log K 10 10 4 10 5 m 6 M molal ± 0.05) (1.30 ± 0.10) − (0.50 ± 0.20) 0 0.000 (2.23 ± 0.05) (1.31 ± 0.10) − (0.27 ± 0.20) 0.1 0.100 (2.02 0.25 0.252 (1.96 ± 0.05) (1.32 ± 0.10) − (0.18 ± 0.20) 0.5 0.506 (1.93 ± ± 0.10) − (0.10 ± 0.20) 0.05) (1.34 ± 0.05) (1.35 0.10) − (0.04 ± 0.21) 0.75 0.762 (1.91 ± 0.05) (1.37 0.11) (0.01 ± 0.21) ± ± 1 1.02 (1.91 0.07) (1.44 ± 0.13) (0.17 2 2.09 (1.95 0.24) ± ± ± 0.09) (1.52 ± 0.17) (0.31 ± 0.28) 3 3.20 (2.02 ± 0.11) (1.60 ± 0.21) (0.45 ± 0.33) 4 4.37 (2.10 5 5.61 (2.20 ± ± 0.25) (0.58 ± 0.40) 0.14) (1.69 ° KCl at 25 C log I K K log I K log K log 1 10 K 2 10 10 3 m 10 molal M 0 0.000 (1.80 ± 0.30) (11.24 ± 0.03) (6.80 ± 0.02) (3.15 ± 0.02) 0.1 0.101 (0.98 ± ± 0.03) (6.21 ± 0.02) (2.72 ± 0.02) 0.31) (10.46 ± ± 0.03) (6.08 ± 0.02) (2.59 ± 0.02) 0.25 0.252 (0.76 0.34) (10.30 ± 0.44) (10.29 ± 0.04) (6.06 0.5 0.509 (0.66 0.03) (2.50 ± 0.03) ± 0.75 0.769 (0.65 ± 0.58) (10.38 ± 0.05) (6.12 ± 0.04) (2.46 ± 0.04) 1 1.03 (0.69 ± 0.73) (10.52 ± 0.06) (6.20 ± 0.05) (2.44 ± 0.05) 2 2.13 (1.0 1.4) (11.28 ± 0.11) (6.71 ± 0.09) (2.44 ± 0.09) ± 3 3.31 (1.5 ± 2.1) (12.21 ± 0.17) (7.34 ± 0.14) (2.50 ± 0.14) 0.19) (2.59 4 4.58 (2.1 2.9) (13.26 ± 0.24) (8.05 ± ± ± 0.19) (Continued on next page)

463 VIII.3 Acid-base equilibria of edta in aqueous solutions 421 Table VIII-8-a: (continued) KCl at 25 ° C I I K K log log K log 5 10 m 4 6 10 10 M molal ± ± 0.10) − (0.50 ± 0.20) 0 0.000 (2.23 0.05) (1.30 0.05) (1.31 ± 0.1 0.101 (2.04 − (0.27 ± 0.20) ± 0.10) ± ± 0.10) − (0.18 ± 0.20) 0.25 0.252 (2.01 0.06) (1.32 0.07) (1.34 (0.10 0.10) − ± ± 0.20) 0.5 0.509 (2.02 ± ± ± 0.10) − (0.03 ± 0.21) 0.75 0.769 (2.06 0.09) (1.36 ± 0.12) (1.38 ± 1 1.03 (2.11 ± 0.21) 0.11) (0.02 2 2.13 (2.37 0.22) (1.46 ± 0.13) (0.19 ± 0.24) ± ± 0.34) (1.54 0.17) (0.34 ± 0.28) 3 3.31 (2.67 ± 0.46) (1.63 0.21) (0.49 ± ± 0.34) 4 4.58 (3.01 ± at 25 ° C KNO 3 I K log K K K log I log log 1 10 10 2 K 10 10 m 3 M molal ± 0.30) (11.24 ± 0.03) (6.80 ± 0.02) (3.15 ± 0.02) 0 0.000 (1.80 ± 0.31) (10.45 ± 0.03) (6.20 ± 0.02) (2.72 ± 0.02) 0.1 0.101 (0.97 0.25 0.253 (0.74 ± 0.34) (10.29 ± 0.03) (6.07 ± 0.02) (2.58 ± 0.02) 0.5 0.512 (0.60 ± ± 0.04) (6.04 ± 0.03) (2.48 ± 0.03) 0.44) (10.27 ± 0.58) (10.35 0.05) (6.08 ± 0.04) (2.42 ± 0.04) 0.75 0.776 (0.57 ± ± ± 0.06) (6.16 ± 0.05) (2.39 ± 0.05) 1 1.05 (0.58 0.73) (10.48 ± 1.4) (11.22 ± 0.12) (6.64 ± 0.09) (2.34 ± 0.09) 2 2.19 (0.8 ± ± 5 3.44 (1.2 0.18) (7.25 ± 0.14) (2.35 ± 0.14) 2.2) (12.16 log K I I log K log K 4 5 m 10 6 10 10 molal M 0 0.000 (2.23 ± 0.05) (1.30 ± 0.10) − (0.50 ± 0.20) 0.1 0.101 (2.04 ± ± 0.10) − (0.27 ± 0.20) 0.05) (1.30 ± − ± 0.10) 0.25 0.253 (2.00 (0.19 ± 0.20) 0.06) (1.31 0.5 0.512 (2.00 0.07) (1.32 ± 0.10) − (0.12 ± 0.20) ± 0.75 0.776 (2.03 ± 0.09) (1.32 ± 0.10) − (0.07 ± 0.21) 1 1.05 (2.07 0.12) (1.33 ± 0.11) − (0.03 ± 0.21) ± 2 2.19 (2.28 ± 0.23) (1.36 ± 0.13) (0.09 ± 0.24) 3 3.44 (2.55 0.35) (1.39 ± 0.17) (0.20 ± 0.29) ± (Continued on next page)

464 VIII Discussion of data selection for edta ligand 422 Table VIII-8-a: (continued) at 20 ° C NaClO 4 log K K K log K log I log I Na 10 1 10 10 2 3 10 m molal M 0.20) (11.18 ± 0.03) (6.75 ± 0.02) (3.17 ± 0.02) 0 0.000 (2.79 ± 0.21) (10.36 ± 0.03) (6.14 ± 0.02) (2.75 ± 0.02) 0.1 0.101 (1.94 ± 0.23) (10.16 0.03) (5.99 ± 0.02) (2.62 ± 0.02) ± ± 0.25 0.253 (1.69 0.31) (10.07 ± 0.04) (5.92 ± 0.5 0.512 (1.52 ± 0.03) ± 0.03) (2.53 ± 0.42) (10.09 ± 0.04) (5.93 ± 0.04) (2.49 ± 0.04) 0.75 0.778 (1.45 ± 0.53) (10.14 ± 0.05) (5.97 ± 0.04) (2.47 ± 1 1.05 (1.43 0.04) 2 2.21 (1.5 1.1) (10.58 ± 0.10) (6.30 ± 0.08) (2.50 ± 0.08) ± ± 3 3.49 (1.7 ± 0.13) (2.59 ± 0.13) 1.7) (11.20 ± 0.15) (6.75 2.3) (11.94 0.21) (7.30 ± 0.19) (2.71 ± 0.19) ± 4 4.93 (2.0 ± 3.1) (12.81 ± 0.27) (7.94 ± 0.25) (2.87 ± 0.25) 5 6.55 (2.4 ± log log K I K K log I 4 5 m 10 6 10 10 molal M 0 0.000 (2.24 0.05) (1.30 ± 0.10) − (0.51 ± 0.20) ± ± 0.05) (1.31 ± 0.10) 0.1 0.101 (2.03 (0.27 ± 0.20) − 0.25 0.253 (1.97 ± 0.05) (1.33 ± 0.10) − (0.18 ± 0.20) 0.5 0.512 (1.95 ± 0.05) (1.35 ± 0.10) − (0.09 ± 0.20) 0.75 0.778 (1.94 ± ± 0.11) − (0.02 ± 0.21) 0.06) (1.38 ± 0.07) (1.40 0.12) (0.04 ± 0.21) 1 1.05 (1.95 ± 0.10) (1.52 0.16) (0.25 ± 0.25) ± ± 2 2.21 (2.03 0.15) (1.64 ± 0.21) (0.44 3 3.49 (2.15 0.32) ± ± ± 0.21) (1.78 ± 0.28) (0.65 ± 0.40) 4 4.93 (2.30 ± 0.27) (1.94 ± 0.37) (0.87 ± 0.50) 5 6.55 (2.47 C ° at 20 NaNO 3 K log I K log K log K log I 10 Na m 2 10 10 10 1 3 molal M 0 0.000 (2.79 ± 0.20) (11.18 ± ± 0.02) (3.17 ± 0.02) 0.03) (6.75 0.1 0.100 (1.94 ± 0.21) (10.36 ± 0.03) (6.13 ± 0.02) (2.74 ± 0.02) 0.25 0.252 (1.68 ± ± 0.03) (5.97 ± 0.02) (2.60 ± 0.02) 0.23) (10.14 ± ± 0.03) (5.88 ± 0.02) (2.49 ± 0.02) 0.5 0.508 (1.49 0.31) (10.03 ± 0.41) (10.03 ± 0.04) (5.87 0.75 0.768 (1.41 0.03) (2.43 ± 0.03) ± 1 1.03 (1.37 ± 0.53) (10.06 ± 0.04) (5.89 ± 0.03) (2.40 ± 0.03) 2 2.13 (1.4 ± 1.0) (10.39 ± 0.07) (6.11 ± 0.05) (2.33 ± 0.05) 3 3.32 (1.5 1.6) (10.86 ± 0.10) (6.44 ± 0.08) (2.32 ± 0.08) ± 4 4.60 (1.7 ± 2.2) (11.42 ± 0.13) (6.83 ± 0.10) (2.33 ± 0.10) 0.14) (2.36 5 6.00 (1.9 2.8) (12.06 ± 0.17) (7.27 ± ± ± 0.14) (Continued on next page)

465 VIII.3 Acid-base equilibria of edta in aqueous solutions 423 Table VIII-8-a: (continued) NaNO ° C at 20 3 I I K K log log K log 5 10 m 4 6 10 10 M molal ± ± 0.10) − (0.51 ± 0.20) 0 0.000 (2.24 0.05) (1.30 0.05) (1.30 ± 0.10) − (0.28 ± 0.20) 0.1 0.100 (2.02 ± 0.05) (1.31 0.25 0.252 (1.96 0.10) − (0.20 ± 0.20) ± ± 0.05) (1.31 0.10) − (0.13 ± 0.20) ± ± 0.5 0.508 (1.91 0.05) (1.32 ± 0.11) − 0.75 0.768 (1.88 ± 0.21) ± (0.08 ± 0.06) (1.32 ± 0.11) − (0.04 ± 0.21) 1 1.03 (1.87 ± 2 2.13 (1.86 ± 0.14) (0.07 ± 0.24) 0.08) (1.35 3 3.32 (1.88 0.11) (1.38 ± 0.18) (0.17 ± 0.29) ± ± 4 4.60 (1.92 0.23) (0.25 ± 0.35) 0.14) (1.41 ± 0.18) (1.44 0.29) (0.34 ± ± 0.43) ± 5 6.00 (1.96 C ° NaCl at 20 log I K K log K I log K log 10 1 Na 2 10 10 10 m 3 molal M 0.20) (11.18 ± 0.03) (6.75 ± 0.02) (3.17 ± 0.02) 0 0.000 (2.79 ± ± 0.21) (10.36 ± 0.1 0.100 (1.94 ± 0.02) (2.74 ± 0.02) 0.03) (6.14 0.25 0.251 (1.69 ± 0.23) (10.15 ± 0.03) (5.98 ± 0.02) (2.61 ± 0.02) 0.5 0.505 (1.53 ± 0.31) (10.06 ± 0.03) (5.91 ± 0.02) (2.52 ± 0.02) 0.75 0.761 (1.46 ± ± 0.04) (5.90 ± 0.03) (2.47 ± 0.03) 0.41) (10.06 ± 0.52) (10.10 0.04) (5.93 ± 0.03) (2.44 ± 0.03) 1 1.02 (1.44 ± 1.0) (10.47 0.05) (2.43 0.07) (6.19 ± ± ± 0.05) 2 2.08 (1.5 ± ± ± 0.10) (6.54 ± 0.07) (2.47 ± 0.07) 3 3.19 (1.7 1.5) (10.96 ± 2.1) (11.52 ± 0.13) (6.95 ± 0.10) (2.53 4 4.36 (1.9 0.10) ± 5 5.59 (2.2 2.6) (12.13 ± 0.16) (7.40 ± 0.13) (2.61 ± 0.13) ± log K I I log K log K 4 5 m 10 6 10 10 molal M 0 0.000 (2.24 ± 0.05) (1.30 ± 0.10) − (0.51 ± 0.20) 0.1 0.100 (2.03 ± ± 0.10) − (0.28 ± 0.20) 0.05) (1.31 ± ± 0.10) − (0.19 ± 0.20) 0.25 0.251 (1.97 0.05) (1.32 ± 0.05) (1.33 ± 0.10) 0.5 0.505 (1.93 (0.11 ± 0.20) − 0.75 0.761 (1.92 ± 0.05) (1.35 ± 0.11) − (0.05 ± 0.21) 1 1.02 (1.92 ± 0.06) (1.37 ± 0.11) (0.00 ± 0.21) 2 2.08 (1.96 ± ± 0.14) (0.16 ± 0.24) 0.08) (1.44 3 3.19 (2.02 0.10) (1.52 ± 0.17) (0.30 ± 0.28) ± 4 4.36 (2.11 ± 0.13) (1.60 ± 0.22) (0.44 ± 0.34) ± 5 5.59 (2.20 0.17) (1.68 ± 0.27) (0.57 ± 0.41) (Continued on next page)

466 VIII Discussion of data selection for edta ligand 424 Table VIII-8-a: (continued) C KCl at 20 ° log I I K K log K log K log K 2 1 10 10 m 3 10 10 M molal 0.30) (11.18 ± 0.03) (6.75 ± 0.02) (3.17 ± 0.02) 0 0.000 (1.79 ± ± 0.31) (10.40 0.03) (6.16 ± 0.02) (2.74 ± 0.02) 0.1 0.100 (0.97 ± 0.34) (10.24 0.03) (6.03 ± 0.02) (2.61 ± 0.02) ± ± 0.25 0.252 (0.75 0.44) (10.23 ± 0.04) (6.01 ± 0.5 0.508 (0.65 ± 0.03) ± 0.03) (2.52 ± 0.58) (10.32 ± 0.05) (6.07 ± 0.04) (2.48 ± 0.04) 0.75 0.768 (0.64 ± 1 1.03 (0.68 ± 0.06) (6.16 ± 0.05) (2.46 ± 0.05) 0.73) (10.46 2 2.13 (1.0 1.4) (11.22 ± 0.12) (6.66 ± 0.10) (2.46 ± 0.10) ± ± 3 3.30 (1.5 ± 0.15) (2.52 ± 0.15) 2.1) (12.14 ± 0.18) (7.29 2.9) (13.19 0.25) (8.00 ± 0.21) (2.61 ± 0.21) ± 4 4.57 (2.1 ± C KCl at 20 ° I K I log K log K log 10 5 4 10 10 m 6 M molal ± ± 0.10) − (0.51 ± 0.20) 0 0.000 (2.24 0.05) (1.30 0.1 0.100 (2.05 0.05) (1.31 ± 0.10) − (0.28 ± 0.20) ± 0.25 0.252 (2.02 ± 0.06) (1.32 ± 0.10) − (0.19 ± 0.20) 0.5 0.508 (2.03 ± ± 0.10) − (0.10 ± 0.20) 0.07) (1.34 ± 0.09) (1.36 0.11) − (0.04 ± 0.21) 0.75 0.768 (2.07 ± 0.12) (1.37 0.11) (0.01 ± 0.21) ± ± 1 1.03 (2.12 0.22) (1.45 ± 0.14) (0.18 2 2.13 (2.37 0.24) ± ± ± 0.34) (1.54 ± 0.18) (0.33 ± 0.29) 3 3.30 (2.67 ± 0.47) (1.63 ± 0.23) (0.48 4 4.57 (3.01 0.35) ± at 20 ° C KNO 3 K I K log K K I log log log 10 10 10 2 10 K 3 m 1 M molal 0 0.000 (1.79 ± 0.30) (11.18 ± 0.03) (6.75 ± 0.02) (3.17 ± 0.02) 0.1 0.101 (0.95 ± ± 0.03) (6.16 ± 0.02) (2.74 ± 0.02) 0.31) (10.39 ± ± ± 0.03) (6.02 0.25 0.253 (0.72 0.02) (2.60 ± 0.02) 0.34) (10.23 0.5 0.511 (0.59 0.44) (10.21 ± 0.04) (5.99 ± 0.03) (2.50 ± 0.03) ± 0.75 0.775 (0.56 ± 0.58) (10.29 ± 0.05) (6.04 ± 0.04) (2.44 ± 0.04) 1 1.04 (0.57 0.73) (10.42 ± 0.06) (6.11 ± 0.05) (2.41 ± 0.05) ± 2 2.18 (0.8 ± 1.4) (11.16 ± 0.12) (6.59 ± 0.10) (2.36 ± 0.10) 0.15) 5 3.43 (1.2 2.2) (12.09 ± 0.19) (7.20 ± 0.15) (2.37 ± ± (Continued on next page)

467 VIII.3 Acid-base equilibria of edta in aqueous solutions 425 Table VIII-8-a: (continued) KNO ° C at 20 3 K I I K K log log log 5 m 10 10 4 6 10 M molal 0.05) (1.30 ± 0 0.000 (2.24 − (0.51 ± 0.20) ± 0.10) ± ± 0.10) − (0.28 ± 0.20) 0.1 0.101 (2.04 0.05) (1.30 0.06) (1.31 ± 0.10) − (0.20 ± 0.20) 0.25 0.253 (2.00 ± ± 0.07) (1.31 ± 0.10) − 0.5 0.511 (2.01 ± 0.20) (0.13 0.75 0.775 (2.03 0.09) (1.32 ± 0.11) − (0.08 ± 0.21) ± ± 0.12) (1.33 0.11) − (0.04 ± 0.21) 1 1.04 (2.07 ± 0.23) (1.36 ± 0.14) (0.09 ± 0.24) 2 2.18 (2.29 ± ± 0.35) (1.39 ± 0.18) (0.19 ± 0.30) 3 3.43 (2.55 − 4 in Table VIII-8-b. Calculated values for the apparent first protonation constant of edta + + in some Na Molar units / K background electrolytes at 25 and 20°C. The SIT model for activity coefficients has been used with the ∆ε values obtained in this review. The † constant K includes the formation of alkali-metal complexes. For example, in sodium 1 electrolytes it corresponds to reaction: + − 3 − 3 + − 4 Na(edta) x ) edta + H Hedta + (1 x + x Na − U where x is an undetermined fraction of alkali metal complex. Care should be exercised in sodium media, because they are ghest ionic strengths when using values at the hi confirmed by few experimental data only, as indicated in the figures of Sections VIII.3.2 to VIII.3.6. Also note that there is no experimental support for the values in potassium media at ionic strengths above 1 molal. 25 ° C NaNO NaClO NaCl 3 4 † † † log log K I I I I K K log 10 10 m m 1 m 1 10 1 molal molal molal M 0 0.000 (11.24 ± 0.03) 0.000 (11.24 ± 0.03) 0.000 (11.24 ± 0.03) 0.1 0.101 (9.43 ± ± 0.19) 0.100 (9.42 ± 0.19) 0.19) 0.101 (9.42 ± ± 0.22) 0.252 (9.08 ± 0.22) 0.25 0.254 (9.09 0.22) 0.253 (9.08 ± 0.30) 0.509 (8.86 ± 0.5 0.513 (8.88 ± 0.30) 0.30) 0.506 (8.85 0.75 0.779 (8.79 ± 0.40) 0.769 (8.77 ± 0.40) 0.762 (8.75 ± 0.39) 1 1.05 (8.75 ± 0.52) 1.03 (8.72 ± 0.51) 1.02 (8.70 ± 0.50) 2 2.21 (8.8 ± ± 1.0) 2.09 (8.7 ± 1.0) 1.0) 2.14 (8.8 3 3.50 (9.0 1.7) 3.33 (8.9 ± 1.6) 3.20 (8.8 ± 1.5) ± 4 4.95 (9.4 ± 2.3) 4.61 (9.2 ± 2.2) 4.37 (9.0 ± 2.1) 5 6.58 (9.8 ± 3.1) 6.02 (9.5 ± 2.8) 5.61 (9.3 ± 2.6) (Continued on next page)

468 VIII Discussion of data selection for edta ligand 426 Table VIII-8 -b: (continued) C 25 ° KCl KNO 3 † † I K K log I I log m 1 10 m 10 1 M molal molal 0.03) 0.000 (11.24 ± 0 0.000 (11.24 ± 0.03) ± ± 0.15) 0.1 0.101 (10.17 0.15) 0.101 (10.17 0.21) 0.253 (9.91 0.20) ± ± 0.25 0.252 (9.91 0.31) 0.512 (9.79 ± 0.30) 0.5 0.509 (9.78 ± ± 0.45) 0.776 (9.77 ± 0.43) 0.75 0.769 (9.74 ± 0.61) 1.05 (9.80 ± 0.59) 1 1.03 (9.75 ± 2 2.13 (9.9 ± 1.3) 1.3) 2.19 (10.1 ± ± 3 3.31 (10.2 2.1) 3.44 (10.5 2.2) 3.0) ± 4 4.58 (10.6 20 ° C NaCl NaNO NaClO 3 4 † † † I K I I I K log K log log m 10 m m 10 1 1 10 1 M molal molal molal 0 0.000 (11.18 0.03) 0.000 (11.18 ± 0.03) 0.000 (11.18 ± 0.03) ± 0.1 0.101 (9.38 ± 0.19) 0.100 (9.37 ± 0.19) 0.100 (9.37 ± 0.19) 0.25 0.253 (9.04 ± 0.22) 0.251 (9.03 ± 0.22) 0.22) 0.252 (9.03 ± 0.30) 0.508 (8.81 0.30) 0.505 (8.81 ± 0.30) ± ± 0.5 0.512 (8.83 0.40) 0.768 (8.72 ± 0.40) 0.761 (8.70 0.75 0.778 (8.74 0.39) ± ± ± 0.52) 1.03 (8.67 ± 0.51) 1.02 (8.65 ± 0.50) 1 1.05 (8.70 ± 1.0) 2.13 (8.7 ± 1.0) 2.08 (8.65 ± 0.99) 2 2.21 (8.8 ± 3 3.49 (9.0 1.6) 3.19 (8.8 ± 1.5) 1.7) 3.32 (8.9 ± ± ± 2.2) 4.36 (9.0 ± 2.1) 4 4.93 (9.3 2.3) 4.60 (9.1 ± 3.1) 6.00 (9.4 ± 2.8) 5.59 (9.2 ± 2.6) 5 6.55 (9.7 KNO KCl 3 † † log I K log I K I 1 m 10 m 1 10 molal molal M ± 0.03) 0.000 (11.18 ± 0.03) 0 0.000 (11.18 0.1 0.100 (10.11 ± ± 0.15) 0.15) 0.101 (10.11 0.25 0.252 (9.86 0.20) 0.253 (9.86 ± 0.20) ± 0.5 0.508 (9.72 ± 0.31) 0.511 (9.74 ± 0.30) 0.75 0.768 (9.69 ± ± 0.43) 0.45) 0.775 (9.72 1 1.03 (9.70 0.60) 1.04 (9.75 ± 0.58) ± 2 2.13 (9.9 ± 1.3) 2.18 (10.0 ± 1.3) 3 3.30 (10.2 2.1) 3.43 (10.4 ± 2.2) ± ± 2.9) 4 4.57 (10.5

469 VIII.3 Acid-base equilibria of edta in aqueous solutions 427 Temperature effects VIII.3.8 4 − Enthalpy changes for the protonation of edta 4) − r ( ( r − 5) + + H (VIII.9) H edta edta H U r ( 1) r − have been determined mostly by calorimetric methods, although in a few references enthalpies were obtained from the temperature variation of protonation constants. A few of these studies have not been considered [77CHO/GOE] , [82TER/IVA] , [85MAR/EVA] [86MAR/EVA] , . These references are discussed in , [98KOC/VAS] [82TER/IVA] Appendix A. In the case of the reported enthalpy changes correspond to the differences in the literature protonation constants adopted by the authors at 20 and 25 ° C. As they were not the result of original measurements in [82TER/IVA] these data [98KOC/VAS] are not considered here. In the enthalpies are reported in a graph but partly corrected for ionic stre ngth effects with an unknown function, and the data cannot be included here. The (VIII.9) values reported in [66MOE/CHU] and [69SIL/SIM] ∆ H were rm ith a large uncertainty: they were obtained not considered because they are associated w from few values of log over a narrow temperature interval (20 to 40) ° C. K 10 + + + , K or R N Most of the remaining investigations have been performed in Na 4 (tetraalkylammonium) salts. Only few studies re port data in LiNO . Daniele et al. [85DAN/RIG] report data 3 for H ∆ ((VIII.9), r = 1), and only [73VAS/KOC] and [76VAS/KOC] give values of rm H ∆ = 3 and 4). For the second protonation step, there is data both in ((VIII.9), r rm [85DAN/RIG] , but the two data sets are strongly discrepant. For and in [74VAS/KOC] example, at 1 M LiNO − C and 25 H ∆ 0.2) r = 2), is reported to be ° (22.1 ± ((VIII.9), 3 rm − 1 kJ ⋅ [74VAS/KOC] , while the log mol K ( T , (VIII.9), r = 2) data in [85DAN/RIG] indi- 10 1 − cate instead ⋅ (9 4) kJ ± . As described in the discussions in Appendix A for these − mol two references, systematic errors in the LiNO data are suspected, especially in 3 [74VAS/KOC] . There are however no additional literature studies in lithium media to support the data in [85DAN/RIG] . Because the amount of e xperiments in sodium, po- + tassium and R N salts is sufficient to make a selection of values for H ∆ ((VIII.9), r = 4 rm − 4 1 and 2), enthalpy data for the first and second protonation of edta obtained in LiNO 3 were excluded in the evaluation described in this Section. + + + Literature data in Na , K N or R media are listed in Table VIII-9 and Table 4 VIII-10. Some of these studies are also discussed in Appendix A, as indicated in the corresponding table. The reported enthalpy changes for the first two individual protona- tion steps, corresponding to the protonation of the amine groups ( r = 1 and 2 in equation − 1 (VIII.9)), range between 12 kJ ⋅ mol 33 and − , depending on ionic media and tem- − perature. As expected, the reported enthalpy changes for the protonation of carboxylic groups (with r = 3 to 5) are small: H ∆ (VIII.9) values range between + 0.6 and rm − 1 + 8 kJ ⋅ mol , depending on ionic media and temperature.

470 VIII Discussion of data selection for edta ligand 428 Table VIII-9: Literature data on the enth alpy changes for ethy lenediaminetetraacetate step-wise protonation reactions in sodium, potassium or tetraalkylammonium media. 1 − 1 − ∆ H Units are: mol·L I mol ⋅ . ), °C for temperature and kJ for ionic strength ( for rm reactions are listed in Table VIII-10. Enthalpy changes for other protonation a ∆ t I H Electrolyte (VIII.9) References Method rm † r =2 r =3 r =4 r =5 =1 r b 1.6) (22.6 ± 1.6) − (15.2 ± − [53CAR/MAR] 0 KCl 0-30 * → 23.8 − 18.4 5.98 0.75 2.05 [58TIL/STA] cal 0.1 KCl 20 − 20 23.7 − 18.2 − cal 0.1 KNO , [63AND] 3 [64AND] e − * 0.1 KNO ± (20 − (33 ± 8) [66MOE/CHU] 25 2) 3 − − − (18.20 ± 0.21) 25 (15.10 ± 0.54) [69CHR/IZA] ClO cal 0.03 4 -0.17 e − 20.5 − [69SIL/SIM] 25 18.4 * 0.1 KNO 3 25 − 23.4 − 18.2 [70KUG/CAR] cal 0.1 KNO 3 c,d 25 (5.48 ± 0.04) (1.26 ± 0.04) [73VAS/KOC] cal 0.2 NaClO 4 0.5 (5.86 ± 0.08) (1.51 ± 0.04) ± ± 0.13) (1.63 0.75 0.04) (5.86 cal 0.2 NaNO (5.36 ± 0.13) (1.13 ± 0.08) 3 0.5 (5.23 ± 0.04) (0.96 ± 0.08) 0.75 ± 0.13) (0.92 ± 0.08) (4.81 (5.36 ± 0.04) cal 0.2 KNO 3 (5.40 ± 0.04) (1.21 ± 0.04) 0.5 (5.52 ± 0.04) (1.21 ± 0.10) 0.75 c,d 25 − (23.44 ± 0.21) − (19.87 ± cal 0.2 KNO [74VAS/KOC] 0.17) 3 0.5 − (24.58 ± 0.21) − (22.22 ± 0.13) ± (26.47 ± 0.21) − (24.10 1 0.13) − − cal 0.2 NaNO (19.25 ± 0.21) 3 − (21.21 ± 0.25) 0.5 1 − (23.97 ± 0.17) cal 0.2 NaCl (18.37 0.13) − ± − ± 0.13) 0.5 (20.63 (23.51 ± 0.21) 1 − 0.08) − (18.66 ± cal 0.2 NaClO 4 0.5 − (21.17 ± 0.25) − (24.14±0.17) 1 cal 1 Me NCl 20 − 27.95 − 21.55 [76AND] 4 0.1 − 24.48 − 15.94 1 KNO 20 − 28.20 − 23.47 3 17.57 − 25.10 − 0.1 (Continued on next page)

471 VIII.3 Acid-base equilibria of edta in aqueous solutions 429 Table VIII-9: (continued) ο a Electrolyte t Method H I ∆ (VIII.9) References rm † =3 r =2 r =1 r =4 r =5 r c,d [76VAS/KOC] (3.85 ± 0.21) 15 KNO cal 0.2 3 (3.68 ± 0.04) 0.5 0.08) ± 0.75 (3.85 (7.30 ± 0.27) KNO cal 0.2 35 3 (7.32 ± 0.13) 0.5 ± 0.06) 0.75 (7.09 (3.35 ± 0.08) cal 0.2 NaClO 15 4 (3.70 ± 0.08) 0.5 0.75 (4.10 ± 0.04) cal 0.2 NaClO (7.45 ± 0.04) 35 4 ± 0.21) 0.75 (7.95 c, e cal 0.5 KNO ± 0.2) [77CHO/GOE] 25 –(29.1 ± 0.2) –(22.7 3 c 15 − (20.17 ± 0.21) [78VAS/KOC] cal 0.2 KNO 3 0.3 − (25.52 ± 0.21) ± ± 0.17) − (22.13 (26.38 0.17) 0.5 − − (27.20 ± 0.21) 1 (24.31 ± 0.25) − cal 0.2 KNO 25 − (18.87 0.13) ± 3 0.3 − (24.87 ± 0.21) 0.5 (25.75 ± 0.17) − (20.84 ± 0.21) − − ± 0.21) − (22.93 ± 0.21) 1 (26.57 35 ± (16.95 cal 0.2 KNO 0.25) − 3 (24.00 ± 0.21 0.3 − − (24.64 ± 0.21) − (19.79 ± 0.21) 0.5 1 − (26.07 ± 0.25) − (22.01 ± 0.17) − (20.50 ± 0.17) cal 0.2 NaClO 15 4 0.17) (22.38 ± 0.5 − 1 − (25.52 ± 0.25) cal 0.2 NaClO 25 − (18.95 ± 0.21) 4 0.5 − (20.92 ± 0.21) (23.85 ± 0.25) 1 − 0.21) 35 − (17.74 ± cal 0.2 NaClO 4 0.5 − (19.66 ± 0.21) 1 − (23.05 ± 0.21) b,c,d ± − (17.6 ± 2.0) − (17.9 10-45 3.4) [85DAN/RIG] * 0.1 NaNO 3 0.3 − (16.8 ± 2.0) − (20.7 ± 3.4) ± 0.6 − (14.9 ± 2.0) − (23.1 3.4) (Continued on next page)

472 VIII Discussion of data selection for edta ligand 430 Table VIII-9: (continued) ο a Electrolyte t Method H ∆ (VIII.9) References I rm † =1 =3 r =2 r r r =4 r =5 − (12.5 ± 3.4) − (24.8 ± 4.9) 1 10-45 * 0.1 KNO (22.3 ± 2.0) − (17.0 ± 3.4) − 3 ± (22.3 ± 2.0) − (18.2 − 3.4) 0.3 − (23.9 ± 2.0) 0.6 (19.5 ± 3.4) − 1 − (22.3 ± 3.4) − (18.2 ± 4.9) NBr 10-45 (20.1 ± 2.0) * 0.1 Me − 4 − (22.2 ± 2.0) 0.3 − (23.4 ± 2.0) 0.6 1 − (22.0 ± 3.4) + ± N 10-45 − (20.1 * 0.1 Et 2.0) 4 0.3 − (22.2 ± 2.0) 0.6 (23.3 ± 2.0) − (22.3 ± 3.4) − 1 ± NBr 10-45 − (19.5 ± 2.0) − (17.0 ± 3.5) (6.9 ± 5.3) (2.5 * 0.1 Pr 8.0) 4 0.3 − (22.3 ± 1.5) − (18.0 ± 3.1) (5.8 ± 4.7) (1.1 ± 7.5) ± (23.9 ± 2.0) − (10.5 − 3.8) (5.5 ± 5.9) (0.8 ± 8.2) 0.6 1 − (22.3 ± 3.0) − (18.1 ± 5.2) (5.8 ± 7.1) (1.6 ± 6.9) cal 0.2 KCl 25 –22.3 –17.2 0.84 2.51 , [85MAR/EVA] c,e [86MAR/EVA] 0.1 NaCl 0-150 − (18 ± 2) * [95PAL/NGU2] †: The enthalpy change for the first protonation step of ethylen ediaminetetraacetate does not take into account the complex formation with alkali metals, and ther efore it refers to apparent equilibria such as: + − 3 + 4 − − 3 + Na(edta) + H ) edta U Hedta x . + x Na x (1 − a: Methods: cal = calorimetry; *: ∂ p K T = temperature dependence of protonation constants. / ∂ a review using the constan : Original log t heat capacity model. K ( T ) data recalculated in this b 10 c : See comments in Appendix A. d: Data also reported in LiNO ionic medium. 3 e: Values not considered in the review procedure.

473 VIII.3 Acid-base equilibria of edta in aqueous solutions 431 Table VIII-10: Literature data considered by this review on the enthalpy changes for ethylenediaminetetraacetate ise reactions. Units are: protonation, excluding step-w † − 1 − 1 H ∆ mol·L for ionic strength ( mol . Enthalpies for ), °C for temperature and kJ for ⋅ I rm step-wise protonation reactions are listed in Table VIII-9. a † t (ºC) Method H ∆ References I Electrolyte rm − + 4 − 2 H edta U edta + 2 H 2 b − (41.8 ± 0.8) [58YAT/KAR] cal 1 NaOH 25 − + 2+ 4 H edta edta U + 6 H 6 b 0 HClO ,HCl,HNO cal 25 − (32.0 ± 0.6) [77LYM/VAS] → 3 4 b,c HCl ± 0.4) [77VAS/LYM] 25 (58.0 cal 2 − (68.5 ± 0.5) 3 − − (79.7 ± 0.8) 4 2 HNO − 0.4) ± (60.5 3 − (72.3 ± 0.5) 3 − (83.8 ± 4 0.8) b,c 15 − (61.0 HCl 0.4) [77VAS/LYM2] cal 2 ± 3 − (71.1 ± 0.5) 4 − (81.6 ± 0.8) 2 (54.7 ± 0.4) 35 − ± (65.7 0.5) 3 − − (76.9 ± 0.8) 4 2+ + H edta(aq) + 2 H U edta H 6 4 b − (0.29 ± cal 1 (Li,Na,K)NO [79VAS/KOC] 25 0.21) 3 1.5 ± 0.33) (1.42 2 (2.64 ± 0.29) 3 (4.94 ± 0.29) 1 NaClO 0.25) 25 − (2.55 ± 4 − (5.15 ± 0.29) 2 (7.36 ± 0.50) 3 − 4 − do not take into account the complex forma- †: Enthalpy changes for prot onation reactions involving edta − 3 − 4 ) edta + x Na(edta) x − fer to apparent equilibria such as: (1 tion with alkali metals, and therefore they re 4) − + + n ( . edta H U H n + x Na + n a: Methods: cal = calorimetry. b: See comments in Appendix A. edta solutions with the acid i ndicated, and therefore, in th e presence of small amounts of c: From mixing K 4 + K ions.

474 VIII Discussion of data selection for edta ligand 432 rformed at temperatures ranging from Calorimetric measurements have been pe 15 to 35 C, cf. ° Table VIII-9 and Table VIII-10. In a few studies the enthalpies of proto- nation have been determined calorimetrically only at 20 [63AND] , C ° [58TIL/STA] , [64AND] ° C by taking into . These were extrapolated in this review to 25 [76AND] , account the temperature dependence of enthalpi es of reaction, which is given by the corresponding heat capacity changes: ⎛⎞ H ∂∆ rm . ∆= C ⎜⎟ r,m p T ∂ ⎝⎠ p Heat capacity changes have been determ ined both calorimetrically and from the temperature dependence of protonation constants. The calorimetric data by Hovey, , [88HOV/HEP2] : [86HOV/HEP] , [85HOV/TRE] Hepler and Tremaine are most precise 4 − 1 − 1 − ο − (310 ± 20) J ⋅ K (edta ⋅ mol C , 298.15 K) = p ,m 1 − 1 3 − − ο ⋅ − 2) J ⋅ K , 298.15 K) = (Hedta ± mol (82 C p ,m 1 − 1 ο − − 2 C edta 1) J ⋅ K (H ± ⋅ mol , 298.15 K) = (81 . 2 ,m p ο 1 − 1 − (VIII.9) = (228 ± 20) and (163 ± 2) J ⋅ K C for ⋅ mol ∆ These data result in p r,m r = 1 and 2, respectively, wh ich may be compared with: 1 − − 1 • ∆ ((VIII.9), = 1) = (98 ± 30) J ⋅ K in 0.1 M NaCl obtained from r ⋅ mol C p r,m [95PAL/NGU2] the variation of protonation constants with temperature , and T values obtained from the cal ∂∆ H Table / cf. orimetric measurements of ( ) , ∂ • p m r VIII-11. Another literature value of ∆ obtained from calorim etric measurements C p r,m ∂∆ of ( H T / ∂ : ) [77VAS/LYM2] is p r m 1 4 − − + 1 − ο 2+ K edta C H ∆ = (485 ± 60) J ⋅ . U ⋅ mol + 6 H edta 6 r,m p ° C for the first, second and third protonation enthalpies Literature values at 20 ° C by using appropriate values of were extrapolated to 25 (VIII.9), if available C ∆ r,m p − 1 1 − Table VIII-11), or by setting cf. ( (VIII.9) = 150 J ⋅ K otherwise. The val- ⋅ C mol ∆ p r,m ° [58TIL/STA] ues at 20 C for the fourth and fifth protonation enthalpies were not in- ± cluded in the review. The uncer tainty of such extrapolated values was increased by − 1 0.5 kJ mol ⋅ . Table VIII-9 contains some enthalpy changes derived from temperature varia- tions of protonation constants [53CAR/MAR] , [85DAN/RIG] , [95PAL/NGU2] . The primary log K ( T ) data was interpreted using the consta nt heat capacity model, either in 10 the original paper [95PAL/NGU2] or by this review [53CAR/MAR] , [85DAN/RIG] .

475 VIII.3 Acid-base equilibria of edta in aqueous solutions 433 ° C for the stepwise protonation reactions of Table VIII-11: Heat ca pacity changes at 25 − 4 edta . Values evaluated in this review from the temperature dependence of the corre- sponding enthalpy changes determined by calorimetric methods as listed in Table VIII-9. − 1 − 1 ∆ (VIII.9) (J ⋅ K I (M) ⋅ mol ) Medium C Reference r,m p r = 1 0.2 (110 20) [78VAS/KOC] KNO ± 3 2 0) 0.3 (76 ± 0.5 (82 20) ± ± 20) 1 (48 r = 2 0.2 (155 KNO ± 20) [78VAS/KOC] 3 0.5 (123 ± 20) 1 (121 20) ± 0.2 (113 ± 20) NaClO 4 20) ± 0.5 (138 1 (130 20) ± r = 3 KNO ± 70) [76VAS/KOC] 0.2 (173 3 ± 0.5 (182 70) 70) ± 0.75 (162 NaClO 0.2 (205 ± 70) 4 ± 70) 0.5 (212 ± 70) 0.75 (192 For calorimetric studies, the interpretation of the data depends on the assumed speciation of the initial and final solutions in the experiments. For the first and second 3 − 4 − III.9), the species edta protonation reactions, r = 1 or 2 in equation (V , Hedta and 2 − H have a relative large pH-range of predominance, but for the other protonated edta 2 tervals of pH at which they predominate. Therefore the forms of edta there are no in calorimetric determination of ∆ (VIII.9) is less prone to systematic errors for r = 1 H rm r ≥ 3. During the evaluation of the da ta, the 95% uncertainty for calorimet- or 2 than for − 1 − 1 ± 1 kJ ⋅ mol ric values was set to: = 1 or 2; ± 1.5 kJ ⋅ mol for r for r = 3; and − 1 ± 2 kJ mol ⋅ r = 4; unless the reported uncertainty in the original publication was for larger. The H ∆ (VIII.9) data in Tabl e VIII-9 were treated according to the SIT rm equations ( cf. Section V.3.6). Only data at 20 and 25 ° C were considered. Values at 20 ° C for the first, second and third protona ° C using tion enthalpies were extrapolated to 25 appropriate values of (VIII.9) as described above. ∆ C p r,m nd electrolyte were treated (VIII.9) data for each backgrou H ∆ Initially, the rm according to the SIT equations ( cf. Section V.3.6), using multi-linear weighted least-

476 VIII Discussion of data selection for edta ligand 434 ο and ∆ε ∆ r , MX), where MX is the background H squares fits to obtain values of ( L rm + r electrolyte, and is the number of H is preliminary regression in reaction (VIII.9). Th sults obtained for chloride, nitrate or perchlorate media analysis revealed that the re cannot be distinguished from a statistical point of view. Furthermore, for some protona- tion steps there was no difference between the results obtained in sodium and potassium media. Hence, all the data have been fitted subsequently with the minimum number of + + + + + + ∆ε ) parameters needed (where M , K (M or R is either Li N ). , Na L 4 The data treated according to the SIT equations, with the uncertainties assigned + ∆ε VIII-14. The enthalpy changes and in this review are shown in Figure r , M ) values ( L squares fits are listed in Table VIII-12. obtained from the multi-linear weighted least- + enthalpy changes, Table VIII-12. Selected (VIII.9) and corresponding ∆ε ) H ∆ (M L rm ° values at 25 C. † r r r = 3 r = 4 = 1 = 2 ο − 1 (1.9 mol H ⋅ ) − (19.8 ± 0.5) − (15.2 ± 0.4) (7.1 ± 0.4) ( kJ ± 1.5) ∆ rm 1 − 3 1 − − K (10 ⋅ mol ⋅ kg ) ∆ε L Cation + (3.4 3.0) Li ± + − ± 3) Na (24 + (3.4 ± 0.9) K + 1.0) (0.2 ± N (1.8 ± 1.2) R 4 + + Na ; K (6.5 ± 0.8) + + + ; K N ; Pr − (1.0 ± 2.3) Na 4 − (0.7 ± 3.4) all electrolytes 4 − do not take into account the complex forma- †: Enthalpy changes for prot onation reactions involving edta tion with alkali metals, and therefore they refer to apparent equilibria such as: 4) − n 4 − ( + + 3 − (1 x H x . H + edta + ) edta x − + Na(edta) Na n U n Because of alkali-metal complex formation, the data for the first protonation of − 4 edta refers to an apparent equilibrium, e.g., in the case of sodium background electro- lytes: + − 3 + 4 − − 3 ) edta x − (1 Na(edta) U Hedta + H x + x Na + † ” on equilibrium constants, enthalpy This is indicated with a superscript “ † ∆ H . Only two references, however, report a value of etc changes, r = 1) in ((VIII.9), rm + † Na [85DAN/RIG] , [95PAL/NGU2] . All other values of r = 1) H ∆ ((VIII.9), media rm + + were obtained in K N or tetraalkylammonium (R ) electrolytes, cf . Table VIII-9. The 4 † ο + + H ∆ differences between the values of r = 1) obtained in either K ((VIII.9), N or R 4 rm media are not large as compar ed with the values in NaNO cf . Table VIII-9 and Figure , 3 + x in the reaction above are small for K VIII-14. This could be expected if the values of + complex formation, or if the enthalpy of K coordination is small.

477 VIII.3 Acid-base equilibria of edta in aqueous solutions 435 ( r − 5) + ( r − 4) U H edta + H Figure VIII-14: edta Enthalpy changes for reactions: H 1) − r ( r first protonation reaction in thodology. The data for the plotted according to the SIT me alkali-metal media, where complexes with the medium cations may take place, + 4 − + − 3 − 3 − x to reaction: (1 ) edta corresponds instead Hedta + H Medta + U M x . + x + electrolytes; KCl: K [53CAR/MAR] 4– + 3– 4− 3− + Hedta + H U edta [58TIL/STA] z Hedta + H edta : KNO −5 3 [63AND], [64AND] −1 [70KUG/CAR] [74VAS/KOC] mol −10 ⋅ [76AND] [78VAS/KOC] [85DAN/RIG] / kJ L + −15 electrolytes: Na D ) [69CHR/IZA] (NaClO 4 [85DAN/RIG] (NaNO ) + 8 3 † 1 −20 [95PAL/NGU2] (NaCl) H r + R N electrolytes: ∆ 4 NCl) [76AND] (Me 4 −25 NBr) [85DAN/RIG] (Me 4 NBr) [85DAN/RIG] (Et 4 1.0 1.5 0.5 0.0 NBr) [85DAN/RIG] (Pr 4 I / molal m + K electrolytes; KCl: [53CAR/MAR] + 3− 2− 2– + 3– [58TIL/STA] edta + H H Hedta U H z edta + H Hedta 2 −10 2 KNO : 3 [63AND], [64AND] −1 [70KUG/CAR] mol [74VAS/KOC] ⋅ [76AND] −15 [78VAS/KOC] / kJ L [85DAN/RIG] + D Na electrolytes: [69CHR/IZA] (NaClO ) 4 −20 + 6 2 ) [74VAS/KOC] (NaNO 3 H r [74VAS/KOC] (NaCl) ∆ ) [74VAS/KOC] (NaClO 4 [78VAS/KOC] (NaClO ) 4 −25 ) [85DAN/RIG] (NaNO 3 + N electrolytes: R 4 0.0 0.2 0.4 0.6 0.8 1.0 1.2 NCl) [76AND] (Me 4 NBr) [85DAN/RIG] (Pr 4 I / molal m (Continued on next page)

478 VIII Discussion of data selection for edta ligand 436 Figure VIII-14: (continued) 16 + − 2− + 2– – H edta + H z H edta H edta + H U edta H [58TIL/STA] (KCl) 2 3 3 2 14 [73VAS/KOC], [76VAS/KOC]: −1 (KNO ) 3 12 (NaClO ) 4 mol ⋅ (NaNO ) 3 10 ) (LiNO 3 / kJ NBr) [85DAN/RIG] (Pr L 4 8 D 6 + 4 3 H r 4 ∆ 2 0 0.0 0.5 1.0 1.5 / molal I m 15 + – + − + H H edta U H edta(aq) 4 3 edta(aq) H edta + H z H [73VAS/KOC], [76VAS/KOC]: 3 4 (KNO ) 3 −1 10 ) (NaClO 4 mol ) (NaNO ⋅ 3 (LiNO ) 3 [85DAN/RIG] (Pr NBr) 5 4 / kJ L D + 2 4 0 H r ∆ −5 1.5 1.0 0.5 0.0 I / molal m [77VAS/LYM] , , The values in Table VIII-10 from [77LYM/VAS] [77VAS/LYM2] , [79VAS/KOC] correspond to the following reactions: 2+ 4 + − + 6 H U H (VIII.10) edta edta 6 + 2+ edta U H (VIII.11) edta(aq) + 2 H H 6 4 In all these calorimetric studies the authors assumed that their final solutions 2+ only contained H edta . However, calculations using the protonation constants selected 6 + by this review show that under the conditions studied the amount of H edta was not 5 ≈ negligible (from ≈ 30% in 2 M acid, to of this, the uncer- 15% in 4 M acid). Because − 1 ± 2 kJ ⋅ tainties in the experimental values were increased to mol . The data were then treated using the SIT methodology as discussed in Section V.3.6. The weighted multi- ο 1 − linear least-squares fits gave mol and ± (30.7 − (VIII.10) = ∆ ⋅ H 2.0) kJ rm − 1 ο ο (2.6 − (VIII.11) = ∆ ∆ H . The values of H (VIII.10) and mol ⋅ 2.3) kJ ± rm rm ο H ∆ (VIII.11) may be combined to obtain: rm

479 VIII.3 Acid-base equilibria of edta in aqueous solutions 437 − 4 1 − + ο ⋅ edta(aq) H + 4 H ∆ = − H ± 3.0) kJ edta mol (28.1 U 4 rm which agrees with the value calculated adding the four step-wise protonation enthalpies, 1 − 1.7) kJ mol ± (26.0 − namely: ⋅ . † ο ο ο = 6). H H ∆ ((VIII.9), r = 5) + r ((VIII.9), H ∆ (VIII.11) = ∆ In addition, rm rm rm ο Only one determination of H ((VIII.9), r = 5) has been reported, cf. Table VIII-9, ∆ rm + but this value is highly dependent on the calculated ratio between H and edta 5 H edta(aq), in the original publication. More data are needed to select individual values 4 ο for ∆ (VIII.9) with H = 5 and 6. r rm , listed in Table VIII-10, cannot be compared The data from [58YAT/KAR] + with other results, because in these experiments substantial Na complex formation took place. From the selected reaction data, the fo llowing enthalpies of formation are se- lected: 4– ο − 1 , 298.15 K) = − ∆ ± 3.8) kJ H mol (edta ⋅ (1704.8 fm ο 3– 1 − H ∆ (Hedta − (1724.6 ± 3.7) kJ ⋅ mol , 298.15 K) = fm 1 2– ο − edta ∆ , 298.15 K) = (H (1739.8 ± 3.7) kJ ⋅ mol H − 2 fm ο – − 1 H edta ∆ , 298.15 K) = . (1732.7 ± 3.7) kJ ⋅ mol (H − 3 fm Alkali metal edta compounds and complexes VIII.4 + + Complexes with Na VIII.4.1 and K 4 − It was found early that the protonation constants of edta depend on the nature of the background electrolyte. Especially, the first protonation constant was found to be sub- + + stantially lower in Na solutions as compared with K media. This was ascribed to the [47SCH/ACK] formation of sodium complexes, and Schwarzenbach and Ackermann + 4 − + determined the stability of the Na and Li in 0.1 M KCl media. complexes with edta Since then several studies have reported equilibrium constants for the formation of + + + ethylenediaminetetraacet ate complexes with Li , K , Na etc . Some of these references , are discussed in Appendix A [65PRI/SEB] , [83ARE/MUS] , [85DAN/RIG] , [87BER/VAS] . The references considered in this review are listed in Table VIII-13. The + coordination of Na in Na O(cr) has also been deter- edta ⋅ 5H 2H O(cr) and Na ⋅ H edta 2 2 2 4 2 mined [93FON/SOL] .

480 VIII Discussion of data selection for edta ligand 438 4 − + + and K Table VIII-13: Literature data on the formation of edta complexes with Na − 1 considered in this review. The equilibrium ⋅ L constants are expressed in units of mol , and correspond to reactions: ( m r − 4) ((4)) mqr +− q 4) − r ((4)) mqr + +− + ( M m ]/[M U L = [ M(HL) M(HL) + . q , β ] H [H ] L r , r , q r m mrq mrq b a + β t M Electrolyte log β Method log Reference I log β β log 1,1,1 10 10 1,2,1 1,0,2 10 1,0,1 10 + 1.66 ise-H 0.1 KCl [47SCH/ACK] 20 Na c + + + /Me 1.91 0.36 [63PAL] N ? Na ise 0.11 Na 4 d + 0.32 CsCl/NaCl 25 Na ∆ 1.79 2.47 0.49 [65BOT/CHA] p K a + − 0.31 CsCl/KCl K 0.96 + NCl/KNO 0.1 Me ∆ p 20 K K 0.8 [67AND] 3 4 a + K ∆ 1 Me NCl/NaClO [77AND] 20 Na p 1.27 [67AND] , 4 4 a e + [68WAT/SCH] ± gl 0.1 Me NCl 25 Na (1.82 0.01) 4 + K ± 0.06) 0.1 (0.55 + 25 K (0.69 ± 0.06) 1 e + K 0.1 Et ∆ NClO [71ROR/MAC] /NaClO 25 Na p 1.88 4 4 a 4 + Et /KNO K 0.70 NClO 4 3 4 e + ,sp 0.5 Me ∆ NCl/NaCl 25 Na p 1.45 [73CAR/SWA] K a 4 + 0.05) (1.69 ± NCl 20 Na [76AND] gl 1 Me 4 + K ± 0.1 NaCl Na p (1.85 ∆ 0.05) a e + 25 Na gl 0.1 Me (2.1 ± 0.15) (0.8 ± 0.2) [91SAL/BOO] NClO 4 4 + K K Me NClO 0.16) /KNO ± ∆ p (1.6 ± 0.3) (1.08 3 4 4 a + NCl 25 Na 0.01) (2.09 (2.35 ± gl 0.15 Me ± 0.02) (1.99 ± 0.02) [93CHE/REI] 4 + (1.60 ± 0.02) (1.57 ± 0.03) (1.51 ± 0.04) K ode; gl = potentiometry with pH-glass electrode; a: Methods: ise-H = potentiometry with hydrogen electr ise = ion selective electrode; sp = spectrophotometry; ∆ p K = the formation constants for alkali-metal a in protonation constants (alkali electrolytes complexation were obtained from differences tetraal- versus kylammonium salts). + + + b: Me N N onium, and tetrapropylammonium, , and Pr : tetramethylammonium, tetraethylamm , Et N 4 4 4 respectively c: A constant ionic medium was not used. The temperat ure is not given in the paper. The anionic compo- nents of the solutions are not given. − 2 d: Ionic strength not constant wh en determining the stability of Na edta . 2 e: Determined the stability of metal complexes from g lass electrode measurements in the activity scale (pH- buffer calibration). Several experimental methods have been used in these studies: emf measure- + ments with pH- and Na -electrodes, spectrop hotometry, calorimetry, and comparison of protonation constants obtained in different ionic media. The latter approach is based on − 4 the assumption that tetraalkylammonium ions do not form ion pairs with edta , and that any differences in the values for the protonation constants in different media with the same ionic strength are due exclusively to the formation of alkali-metal complexes. However, specific ion interactions (activity coefficient effects) also do induce differ- I > 0.1 M. This is probably the reason be- ences in protonation constants, especially at + + hind the postulated Me N and Et N complexes in [85DAN/RIG] . There is therefore a 4 4

481 VIII.4 Alkali metal edta compounds and complexes 439 larger uncertainty associated w ith all the references that used this methodology (labelled p ∆ K “ ” in Table VIII-13). a As equilibrium constants are reported both for 20 and 25°C, it is necessary to effects. Enthalpy changes for the formation consider first the available data on thermal 3 3 − − [54CHA2] , [65PRI/SEB] , and Kedta are reported in several studies of Na(edta) [76AND] , [77VAS/LYM] , [85DAN/RIG] , [87BER/VAS] , [76VAS/BEL] . The data [65PRI/SEB] from , are not included in this review, as [85DAN/RIG] , [87BER/VAS] discussed in Appendix A. Data from the remaining references are tabulated in Table VIII-14. Table VIII-14: Literature data on the enthalpy changes for the complex formation be- − − + 3 4 − 4 + tween edta + edta Na(edta) . The data correspond to reaction: Na and Na . U a b 1 − I t (°C) (M) ) References Method (kJ ⋅ mol Medium ∆ H rm c + − − cal 0.55 Me N 5.9 [54CHA2] 25 / Cl 4 NCl 20 − (10.8 ± 0.5) cal 1 Me [76AND] 4 0.1 NaCl (5.4 ± 0.5) − d cal 0.3 Me 25 NNO (9.71 ± 0.29) [76VAS/BEL] − 3 4 0.5 − (9.37 ± 0.33) 1 (8.54 ± − 0.13) 0.3 KNO 15 (10.04 ± 0.21) − 3 0.5 ± 0.13) − (8.58 1 ± 0.21) − (7.11 25 0.3 (9.46 ± 0.17) − 0.5 − ± 0.21) (8.58 1 (6.90 ± 0.25) − 0.3 35 (9.08 ± 0.29) − 0.5 (8.45 ± 0.17) − 1 − (6.49 ± 0.29) → 0 25 − (7.6 ± 1.2) [77VAS/LYM] cal a: cal = calorimetry. + b: Me N : tetramethylammonium. 4 3 − c: Titrated (Me edta with 1 M NaCl. Enthalpy change is based on log N) from (Na(edta) ) K 4 4 10 [47SCH/ACK] and protonation enthalpies from [53CAR/MAR] . 3 − d: Enthalpy change is based on log (Na(edta) ) K extrapolated to different from [47SCH/ACK] 10 temperatures and ionic strengths. The values have been treated with the SIT methodology ( cf. Section V.3.6), and are plotted in Figure VIII-15. The data in Figure VIII-15 show a large spread, which cannot be explained only by the differences in the ionic media, and there is no indication of an ionic strength dependence of these scattered data. Hence, this review selects a weighted average of the quantities, + 8 D in Figure VIII-15 to derive th e selected enthalpy value at zero H ∆ L rm ionic strength:

482 VIII Discussion of data selection for edta ligand 440 + − 4 − 3 Na(edta) Na (VIII.12) + edta U − 1 ο − (4 ± 3) kJ·mol (VIII.12)= ∆ H rm This selection yields: ο 3 − –1 . H , 298.15 K) = – (1949.1 ± ∆ (Na(edta) 4.8) kJ·mol fm 3 − plotted according to the Figure VIII-15: Enthalpy data for the formation of Na(edta) SIT model. The term 8 D is the Debye-Hückel term fo r ionic strength effects of en- L 4 − − + 3 thalpy data for the reaction Na + edta U ( cf . Section V.3.6) Na(edta) 2 −1 − 3 − 4 + 3– 4– + + edta Na Na(edta) U Na(edta) Na z + edta 0 mol ⋅ [54CHA2] [76AND] at 20°C −2 [76VAS/BEL] in KNO / kJ 3 L [76VAS/BEL] in Me NNO D −4 3 4 [77VAS/LYM] + 8 − −6 3 −8 Na(edta) H r ∆ −10 1.2 0.8 0.4 0.0 I / molal m y be calculated from results of the calo- The heat capacity change at 25°C ma [88HOV/HEP2] rimetric measurements in : ο 1 − ∆ C ⋅ mol 25) kJ (VIII.12) = (243 ± r,m p From these results it may be calculated that the equilibrium constant for the corresponding reaction will increase by (0.012 ± 0.009) log -units when the tempera- 10 ture is decreased to 20°C. This was taken into account when reviewing the literature on this equilibrium constant. 3 − − 3 The equilibrium constants fo r the formation of Na(edta) (Table and Kedta VIII-13) were converted to molal units and extrapolated to 25°C when necessary. The ± 0.15 log standard deviations for the individual data were increased to -units if the 10 reported uncertainty was missing or if it was below this value. When treated according cf. Appendix B), the data for each cation in the background electro- to the SIT model ( lyte should follow a different linear dependence on I . In this case most of the data has m been obtained in tetramethylammonium media (Table VIII-13). The plots in Figure VIII-16 show that the few values obtained in media not containing tetraalkylammonium + ions appear close to the data obtained in Me N solutions. 4

483 VIII.4 Alkali metal edta compounds and complexes 441 The weighted linear least-squares regressions gave the following equilibrium constants: ο 3 − ο log K ± , 298.15 K) = (2.8 (Na(edta) = 0.2), β log 10 Na 1 10 − 3 ο ο K = 0.3), ± , 298.15 K) = (1.8 log (Kedta β log 10 K 1 10 which are selected by this review. The slope of the SIT regressions corresponds to: − 3 + +− 4 − 1 (Me N , edta ) (Me N , Na(edta) ) ε ε − (0.24 ± 0.47) kg ⋅ mol – = , 4 4 3 +− 4 + − − 1 ε (Me N , edta ) (Me N , Kedta ) = – − (0.50 ± 0.64) kg ⋅ mol ε . 4 4 too large to allow any recommendations. The uncertainties in these values are If the literature values obtained from comparison of protonation constants in included in the regres- ” in Table VIII-13) are not K p ∆ different ionic media (labelled “ a 3 − ο sion, then the results are instead: (Na(edta) 0.3) and ) = (2.8 ± β log 1 10 3 − ο (Kedta ) = (2.0 ± 0.5), which agree with the selected values. log β 1 10 3 − are probably Additional complexes with stoichiometries different than Medta formed in solutions where the alkali cation concentration is perhaps 1000 times higher 2 − than that of edta. The formation of Na by com- edta was proposed in [65BOT/CHA] 2 paring titration curves obtained at diverse ionic strengths. Under such conditions the activity coefficients were not controlled. No value can be recommended for this com- plex at present. + + Equilibrium constants have been published for protonated Na and K com- 2 − − edta) and M(H , cf . Table VIII-13. The reported stability of these plexes: M(Hedta) 2 complexes is, with one exception, quite weak: 4) + − r (1+ ( r − 4) M 1 U edta) edta 1.0 + H K M(H r r The original studies were performed in such a way that it is not possible to ignore the possibility of variations in activity factors. In [93CHE/REI] the reported stability of these complexes is substantial for a wide pH range: the complexes account + + for etetraacetate in solutions ≈ 10 mM in Na 2 or K 20% of the ethylenediamin at pH ≥ 3. These results are in disagreement with previous studies that indicate that it is only the first protonation constant that is substantially affected by the medium cation. The reason for this discrepancy has not been elucidated. Due to these issues, equilibrium constants for these protonated complexes cannot be recommended in this review.

484 VIII Discussion of data selection for edta ligand 442 − 3 − 3 and Kedta Figure VIII-16: Equilibrium constants for the formation of Na(edta) plotted according to the SIT methodology. Symbols with black background correspond to tetramethylammonium media, those with grey background correspond to other tetra- alkylammonium media, and symbols with white background indicate other ionic media. Equilibrium constants have been converted to molal units, and extrapolated to 25°C when necessary. 4.0 [47SCH/ACK] m I [63PAL] (t = ?) ) − 3.5 [65BOT/CHA] X , [67AND, 77AND] + [68WAT/SCH] (Na 3.0 [71ROR/MAC] ε − [73CAR/SWA] D NCl [76AND] in Me 4 2.5 [76AND] in NaCl + 8 Na [91SAL/BOO] 3– + 4– K − + − 3 4 + edta Na Na(edta) U [93CHE/REI] 2.0 10 Na(edta) + edta z Na ° C at (20 – 25) log at (20 - 25)°C 1.5 0.0 0.2 0.4 0.6 0.8 1.0 1.2 I / molal m [65BOT/CHA] 3.0 [67AND] m I [68WAT/SCH] ) − [71ROR/MAC] 2.5 X , [91SAL/BOO] + [93CHE/REI] (K ε 2.0 − D 1.5 + 8 K − 4 − + 3 3– + 4– K U K Kedta + edta Kedta z K + edta 10 at (20 - 25)°C C ° at (20 – 25) 1.0 log 0.4 0.8 1.0 1.2 0.0 0.6 0.2 / molal I m

485 VIII.4 Alkali metal edta compounds and complexes 443 Magnesium and calcium edta VIII.5 compounds and complexes VIII.5.1 Magnesium and calcium edta compounds Many edta compounds containing magnesium and calcium have been reported in the of the 16 compounds listed in Table VIII-15 has been literature. The stoichiometry confirmed by elemental analysis. Solubility measurements in H O have been reported 2 for several of these compounds. Table VIII-15: Magnesium and calcium edta compounds. References reporting solubility data are marked with (sol.). O Reference Solubility in H Compound 2 3 − edta O 0.098 mol·dm Mg 9H [59VOR] , [74MYA/LOG] (sol.), [76MYA/LOG] (sol.) ⋅ 2 2 3 − Mg(H 6H O 0.0014 mol·dm edta) (sol.), [57BRI/PAR] (sol.), [59VOR] , [74DUD/SHT] ⋅ 2 2 1 − [74DUD/SHT2] (sol.), [76DUD] (sol.) 0.0010 mol·kg 3 − 2H O 0.04 mol·dm (sol.) edta [57BER/MUL] Ca ⋅ 2 2 3 − Ca edta (sol.) 7H [74MYA/LOG] O 0.036 mol·dm ⋅ 2 2 1 − [67BHA/KRI] edta) O 0.014 mol·kg 2H , [73RYK/SHT] (sol.) Ca(H ⋅ 2 2 8H O [49PFE/SCH] CaMg(edta) ⋅ 2 3 − CaMg(edta) 9H O 0.060 mol·dm [74MYA/LOG] (sol.) ⋅ 2 Mg(edta) [47SCH/ACK] Na 2 1 − 4H (sol.) O 2.3 mol·kg Mg(edta) [73DUD/SHT] , [58SAW/PAU] Na ⋅ 2 2 K Mg(edta) 5H O [42PFE/OFF] ⋅ 2 2 Ca(edta) [43PFE/SIM] , [47SCH/ACK] Na 2 1 − 2H O 2.0 mol·kg [70VOR/RYK] (sol.) Ca(edta) Na ⋅ 2 2 Na Ca(edta) 3.5H O [58SAW/PAU] ⋅ 2 2 [43PFE/SIM] Ca(edta) 6H O [49PFE/SCH] , Na ⋅ 2 2 [43PFE/SIM] Ca(edta) 4H [42PFE/OFF] O , K ⋅ 2 2 [Cu(NH ]Ca(edta) 8H O [49PFE/SCH] ) ⋅ 3 2 4 et al. [57BER/MUL] Bersin state that the solubility of Ca edta ⋅ 2H O is 1.6% 2 2 3 − 0.04 mol·dm ≈ ( et al. ) at room temperature leading to solutions of pH 4 to 5. Myachina − 3 [74MYA/LOG] give solubility values of 0.036 mol·dm edta for Ca 7H O, 0.060 ⋅ 2 2 − 3 − 3 mol·dm O. All solubilities O and 0.098 mol·dm for CaMg(edta) ⋅ for Mg 9H edta ⋅ 9H 2 2 2 were measured at (25.0 ± 0.1) ° C, but the pH of the resulting solutions is not given. Myachina [76MYA/LOG] et al. dissolved 3.3 weight-% Mg edta in water, which corre- 2 − 3 sponds to 0.098 mol·dm Mg edta ⋅ 9H O. 2 2 , measured a solubility of 0.6 mg [57BRI/PAR] Bricker and Parker, 3 − Mg(H ⋅ 6H ) at room temperature. The pH of O per ml of water ( ≈ 0.0014 mol·dm edta) 2 2 the resulting solution is not reported. Dudakov and Shternina [74DUD/SHT] ,

486 VIII Discussion of data selection for edta ligand 444 − 1 6H Mg(H [74DUD/SHT2] edta) ⋅ report 0.0010 mol·kg [76DUD] O at 25°C lead- and 2 2 ing to solutions of pH 4.2. Rykova and Shternina [73RYK/SHT] report a solubility of 1 − 0.014 mol·kg edta) ⋅ 2H Ca(H O at 25°C and pH 3.8. 2 2 − 1 at 25 measured a solubility of 2.0 mol·kg C ° [70VOR/RYK] et al. Vorob’ev Ca(edta) ⋅ 2H O. In close agreement with this value, Dudakov and and pH 8.2 for Na 2 2 − 1 [73DUD/SHT] Shternina for Na give a solubility of 2.3 mol·kg Mg(edta) ⋅ 4H O at 2 2 ° C and pH 8.0. 25 In qualitative terms, an overall consistent picture emerges from these solubility data. However, it is outside the scope of this review to develop a quantitative thermody- namic model for these rather soluble Mg and Ca edta compounds. Stability of magnesium and calcium edta complexes VIII.5.2 Complex formation in Mg and Ca edta systems have been studied by several investiga- tors. The equilibrium data found in the literature are summarised in Table VIII-16. From at the divalent alkaline earth cations form these experimental investigations we infer th − 2 − the following complexes with edta: M(Hedta) , Medta and M edta(aq), where M 2 stands for Mg or Ca. The equilibrium constants in Mg and Ca edta systems usually refer to these reactions: 2 − 4 − 2+ M U (VIII.13) + edta Medta 2 − − + U + H M(Hedta) (VIII.14) Medta + − 4 − 2+ + H M(Hedta) + edta (VIII.15) M U The coordination number, CN, of Mg and Ca can be larger than six, i.e. in − 2 2 − [Mg(H [77PAS/WHI] and in [Ca(H CN = O) edta] , CN = 7 has been found , O)edta] 2 2 2 8 has been reported [79BAR/UCH] . In addition, some edta donor atoms are not bound 2 − in the 1:1 complex. Hence, Medta can bind a second ligand X forming ternary com- plexes MedtaX, but their stability constants are small and difficult to be detected ex- 2+ actly. For Ni (see Table VIII-22 in Section VIII.7 .2) such ternary species have been studied, mainly using spectrophotometric measurements. This method does not work in the case of Mg and Ca as these cations give colorless solutions.

487 VIII.5 Magnesium and calcium edta compounds and complexes 445 Table VIII-16: Experimental equilibrium data for the Mg and Ca edta systems. The uncertainties are given as reported in the references. If the ionic medium is shown in parenthesis, the contribution of the reacting species to the total ionic strength has been considered. t C) log ( K Reference Method Ionic medium ° 10 2 − 2+ − 4 + edta Mg(edta) Mg U 20 8.69 ise-H 0.1 M KCl [47SCH/ACK] I ise-H 0 25 9.09 [54CAR/MAR] → ise-Hg 0.1 M NaClO 25 8.9 [56SCH/REI] 4 gl 0.1 M KNO 0.5 (8.49 0.02) [60BOH/MAR] ± 3 13.4 (8.57 0.02) ± 25.3 (8.64 0.02) ± 42.4 (8.73 0.01) ± ise-Hg 0.1 M (KNO ) 30 9.06 [64YEN/LIU] 3 0.1 M KNO dis 20 11 [65JOK/MAJ] 3 0.18 M (Me gl NCl) 10 8.86 [83ARE/MUS] 4 0.20 20 8.82 0.17 25 8.93 0.34 8.70 0.53 8.67 0.76 8.64 0.17 37 8.96 0.35 8.86 0.18 45 9.15 37 0.15 M NaCl (7.750 gl 0.006) [84DUF/MAY] ± pot 0.3 m NaCl 25 7.45 [2001CHO/BON] 1 m NaCl 25 (6.7 0.1) ± 2 25 (6.44 0.03) ± 3 25 6.338 25 (6.41 4 0.03) ± 5 25 6.52 2 + − − U Mg(edta) + H Mg(Hedta) ise-H 0.1 M KCl 20 3.85 [47SCH/ACK] gl 0.1 M (Me NCl) 25 (4.0 0.1) [83ARE/MUS] ± 4 2+ + 4 − − U Mg Mg(Hedta) + edta + H pot 0.3 m NaCl 25 11.59 [2001CHO/BON] 1 m NaCl 25 (11.0 0.3) ± 2 25 (10.0 0.5) ± 3 25 10.3 4 25 (10.7 0.2) ± 5 25 10.0 (Continued on next page)

488 VIII Discussion of data selection for edta ligand 446 Table VIII-16: (continued) t K Reference Method Ionic medium C) log ( ° 10 2 4 2+ − − U + edta Ca(edta) Ca 10.59 ise-H 0.1 M KCl [47SCH/ACK] 20 I 25 10.98 [54CAR/MAR] ise-H 0 → 0.1 M (KNO gl 0.05) [54SCH/GUT] ) 20 (10.70 ± 3 (10.96 pol 0.1 M (KCl) 0.40) ± 0.1 M NaClO ise-Hg 25 10.7 [56SCH/REI] 4 ise-Hg 0.1 M (NaNO [57SCH/AND] ) 21.7 (10.81 0.10) ± 3 0.1 M (KNO gl 0.05) ) 20 (10.69 ± 3 0.1 M KNO gl 0.03) [60BOH/MAR] 0.5 (10.94 ± 3 13.4 (10.62 0.03) ± 25.3 (10.42 0.02) ± 42.4 (10.11 0.01) ± 0.3 M NH cix ? ? 10.45 [60MAT/SAF] 4 gl 0.1 M KCl 30 (10.59 0.05) [63GRI/HUG] ± 0.1 M (KNO ise-Hg ) 30 10.92 [64YEN/LIU] 3 dis 0.1 M KNO 20 11 [65JOK/MAJ] 3 gl 1.6 M KNO 25 (9.4 [66KUL/REE] 0.1) ± 3 1.0 M (KNO gl 0.01) [68KUE/SCH] ) 25 (9.68 ± 3 a a ise-Ca 0.01 M Et 4.07 25 [68REC/LIN] Nac 4 0.1 M (NaClO gl ) 20 0.05) [70AND/MAL] (10.85 ± 4 0.1 M (KNO ise-Cu ) 25 (10.78 0.02) [73HAN/RUZ] ± 3 0.1 M (KNO gl ) 25 (10.73 0.04) [75AND/POD] ± 3 gl 0.5 M (Me NCl) 25 10.28 [75CAR/SWA] 4 cix ? [77MOY/FRI] rt 11.0 ise-Ca 25 10.93 [79CRA/MOO] 0.1 M (NaCl) gl 0.18 M (Me NCl) 10 11.03 [83ARE/MUS] 4 0.20 20 10.77 0.17 25 10.75 0.34 25 10.61 0.53 25 10.47 0.76 25 10.46 0.17 37 10.59 0.35 37 10.49 0.18 45 10.49 gl 0.15 M NaCl 37 (9.360 0.003) [84DUF/MAY] ± 0.1 M KCl 20 pot 22 (10.73 0.10) [89POC] ± − gl 0.1 M (NaClO 10.75 [92GLA/HUL] ) 25 4 cal ? 25 10.7 [99GRI] (Continued on next page)

489 VIII.5 Magnesium and calcium edta compounds and complexes 447 Table VIII-16: (continued) Method Ionic medium t C) log ( K Reference ° 10 2 + − − U + H Ca(edta) Ca(Hedta) 20 3.18 [47SCH/ACK] ise-H 0.1 M KCl 1.6 M KNO 0.2) [66KUL/REE] gl 25 (2.8 ± 3 0.1 M (KNO gl 0.04) [75AND/POD] ) 25 (2.94 ± 3 gl 0.5 M (Me NCl) 25 3.47 [75CAR/SWA] 4 0.1 M (Me gl (3.1 0.1) [83ARE/MUS] NCl) 25 ± 4 25 4.23 [85MAR/EVA] cal 0.2 M KCl [86MAR/EVA] 2 2+ − U Ca(edta) + Ca Ca edta(aq) 2 ise-H 20 < 0.7 [48SCH/ACK] 0.1 M KCl 0.3 M NH cix [60MAT/SAF] ? ? 2.07 4 nmr ? 29 (1.1 0.1) [68LEY/WHI] ± 2+ 2 − refers to [Ca(edta) ]/([Ca a: ]·[edta] The stability constant ) where [edta] is the total ligand species total total medium is tetraethylammonium acetate concentration at equilibrium. The ionic are of an order of mag- The stabilities of the Mg and Ca edta 1:1 complexes K 1 nitude that allows the direct investigation of their equilibria (VIII.13) in the pH range 3 to 10. A convenient method for determining the equilibrium constants of these magne- sium and calcium edta complexes is the alkalimetric titration of mixtures containing the neutral protonated ligand H edta and a salt of the investigated metal ion, preferably with 4 − − e.g., a non − complexing anion, or ClO NO . Both, the ligand and the metal salt should 3 4 with respect to the concentration of the be present at negligible total concentrations to keep the activity coefficien ts of all species constant background electrolyte in order during the experiment. Starting with H lt to realise, because in edta(cr), this is not difficu 4 water only millimolar solutions of this acid are attained due to its low solubility (see Section VIII.2.2). Fr om the graphic representation of the titration curves the composi- tion of the formed species can be inferred (Figure VIII-17) . Using solutions in the mil- limolar concentration range the metal ion complex formation occurs after the deprotona- − − tion of H edta edta and H and above pH 3 with simultaneous formation of M(Hedta) 4 3 − 2 − 2 Medta 8). The main features shown in from H (Figure VIII-17 and Figure VIII-1 edta 2 2 − Figure VIII-17 are caused by the formation of Medta . After addition of 4 equivalents of strong base per mol edta a large pH change is observed at a = 4, which increases with − 2 increasing stability of the formed Medta complex (Figure VIII-17). At a > 4 the titra- − tion curve represents the excess base in the solution. M(Hedta) always remains a minor species detectable in the titration curve only at a ≈ 2 at pH < 5 (Figure VIII-18). In order − to obtain larger conc entrations of M(Hedta) an excess of metal is needed, allowing its formation at low pH values as done in [47SCH/ACK] and [48SCH/ACK] , but also under such conditions less than 10% of the metal is bound to protonated complexes.

490 VIII Discussion of data selection for edta ligand 448 − 3 Figure VIII-17: Simulated titration curves of 1 molal H × edta(aq) in 1 molal NaCl 10 4 − 3 − 3 (upper curve) and 1 10 × × 10 m H in 1 molal NaCl (lower molal MgCl edta(aq) and 1 2 4 a denotes the moles of base added per moles of ligand present in curve). The symbol solution. Stability constants selected in this review have been used to calculate the titra- tion curves. 12 11 10 9 ] + 8 [H 10 7 log 6 5 4 3 2 012345 a Figure VIII-18: Distribution of complex specie s in the simulated titration of the Mg edta system (Figure VIII-17) in 1 m NaCl. 100 2+ [Mg ] 2 − 90 ] [Mg(edta) 80 2 − 70 [H edta ] 2 60 50 40 − % Concentration [H edta ] 3 30 3 20 − ] [Hedta 10 edta(aq)] − [H 4 ] [MgHedta 0 23456789 + [H ] log − 10

491 VIII.5 Magnesium and calcium edta compounds and complexes 449 The literature data collected in Table VIII-16 have been scrutinised in order to select reliable studies, summarised in Table VIII-17 on which the evaluation of recom- mended values is based. [57BJE/SCH] In agreement with the rules given in [64SIL/MAR] , , , the ionic medium in Table VIII-16 is given in parentheses, if the given [71SIL/MAR] effectively to the total numerical value corresponds -value, corrected for the sometimes I significant contribution of the reacting species, with concomitant reduction of the added amount of inert salt. Note that this rule was not respected by several authors and thus, a check of the original reference is always necessary. The largest part of the entries in Table VIII-16 refers to measurements of the concentration of hydrogen or a metal ion with a glass electrode (gl) or an ion specific electrode (ise). From these measurements th e concentrations of all species in solution forming during the alkalimetric titrations can be calculated. Note that by using an ion eded in order to obtain the concentration specific electrode also pH measurements are ne of the free ligand, although this is not exp column for the method licitly specified in the in Table VIII-16. Because of various shortcomings in the experimental procedures or the report- i.e. ing of the results ( electrodes used, pH calibration, ionic strength, composition of the solutions), the values of [54CAR/MAR] [60BOH/MAR] [60MAT/SAF] , , , [65JOK/MAJ] [66KUL/REE] , [68KUE/SCH] , [75CAR/SWA] , [77MOY/FRI] , , [85MAR/EVA] , , [89POC] , [92GLA/HUL] , [86MAR/EVA] have been dis- [99GRI] carded ( cf. Appendix A). The stability constants for Mg and Ca edta reported in [78NOV/LUC] most probably have been taken from [47SCH/ACK] and [54SCH/GUT] , although this is not explicitly stated in [78NOV/LUC] . However, no experimental work with edta is described by [78NOV/LUC] and hence, their Mg and Ca edta values are not included in Table VIII-16. In the case of [47SCH/ACK] only the equilibrium data for the protonated spe- cies have been considered in this review ( cf. Appendix A). [68REC/LIN] The conditional stability constant of is the only measurement in lues reported by Duffield et al. tetraethylammonium acetate. The va [84DUF/MAY] 2 2 − − represent the only determination of Ca(edta) stability at 37 and Mg(edta) C and 0.15 ° M NaCl. The data of both papers have not been included in the final data evaluation. The studies [56SCH/REI] , [57SCH/AND] , [70AND/MAL] in Na media need some comments. In the case of [56SCH/REI] , [57SCH/AND] the cation exchange equi- 2+ 2 − − 2+ 2 librium Hg(edta) has been studied by means of a Hg + Ca + Hg U Ca(edta) sensitive electrode. The Hg-edta complexation has been determined in the same me- dium (NaClO and NaNO , respectively), but the edta protonation constants used to 4 3 2 − calculate the stability constant of Hg(edta) refer to potassium media. [56SCH/REI] used the protonation constants of [47SCH/ACK] determined in 0.1 M KCl, and

492 VIII Discussion of data selection for edta ligand 450 determined the edta protonation constants in the same study in 0.1 M [57SCH/AND] KNO studied the edta complexation of Sb(III) and the cation ex- . [70AND/MAL] 3 2+ − − − 2 change equilibrium Sb(edta) Ca(edta) + Sb(OH) + Ca (aq) in 0.1 M + 3 OH U 3 NaClO [67AND] have . Again, edta protonation constants determined in 0.1 M KNO 4 3 pted values reported been used in the data evaluation. Hence, in all three cases the acce − 3 − 3 in Table VIII-17 have been corrected for Kedta for Na(edta) not complexation, and complexation, as one would infer by just looking at the column “ionic medium”. [54SCH/GUT] In the cases of , , [63GRI/HUG] , [57SCH/AND] [70AND/MAL] the uncertainties of Table VIII-16 have been multiplied [83ARE/MUS] , i.e. , by a factor (1.96) to obtain error limits closer to a 95% total uncertainty levels, including random and possible systematic errors. If this procedure resulted in uncer- ± 0.10 is used [73HAN/RUZ] tainty estimates < 0.1, a value of [75AND/POD] . For the , , [83ARE/MUS] , [2001CHO/BON] , the uncer- , [64YEN/LIU] [47SCH/ACK] review of tainty estimates are detailled in Appendix A. In the case of [56SCH/REI] an uncertainty ± 0.2 has been estimated considering the additional ambiguity arising from the use of of the protonation constants of [47SCH/ACK] . Table VIII-17: Accepted formation constants for the Mg and Ca edta systems to derive the selected values. Uncertainties have been estimated in this review. a Reference C) log Ionic medium t ( t C) log ( K log K K ° ° 10 10 10 2 2+ 4 − − U Mg(edta) Mg + edta (c) 0.1 M NaClO 25 (8.9 0.2) (9.2 [56SCH/REI] 0.2) ± ± 4 (b) (c) ) 30 (9.06 [64YEN/LIU] 0.30) 25 (9.0 0.3) 0.1 M (KNO (9.3 0.3) ± ± ± 3 0.17 M (Me (8.93 0.15) [83ARE/MUS] NCl) 25 ± 4 0.34 25 (8.70 0.15) ± 0.53 25 (8.67 0.15) ± 0.76 25 (8.64 0.15) ± d 25 (7.45 0.3 m NaCl 0.20) 0.10) (8.61 [2001CHO/BON] ± ± d 1 25 (6.70 0.20) 0.10) (8.16 ± ± d 2 25 (6.44 0.20) 0.10) (8.24 ± ± d 3 25 (6.34 0.10) (8.47 0.20) ± ± d 4 25 (6.41 0.10) (8.86 0.20) ± ± d 5 25 (6.52 0.20) 0.10) (9.28 ± ± 2 + − − U + H Mg(Hedta) Mg(edta) 20 0.1 M KCl 0.2) [47SCH/ACK] (3.9 ± 0.1 M (Me NCl) 25 (4.0 [83ARE/MUS] 0.2) ± 4 (4.1 25 0.3 m NaCl 0.5) [2001CHO/BON] ± 1 25 (4.3 0.5) ± 2 25 (3.5 0.5) ± 3 25 (4.0 0.5) ± 4 25 (4.3 0.5) ± 5 25 (3.5 0.5) ± (Continued on next page)

493 VIII.5 Magnesium and calcium edta compounds and complexes 451 Table VIII-17: (continued) a C) log Reference K t ( ( C) log t K log Ionic medium K ° ° 10 10 10 4 2+ 2 − − U + edta Ca(edta) Ca b c 0.10) 0.10) ) 20 (10.70 (10.90 [54SCH/GUT] 0.1 M (KNO 0.10) 25 (10.62 ± ± ± 3 c 0.2) (10.99 0.20) 25 (10.7 0.1 M NaClO [56SCH/REI] ± ± 4 b c 0.1 M (NaNO 0.20) 25 (10.76 ) 21.7 (10.81 (11.05 [57SCH/AND] 0.20) 0.20) ± ± ± 3 c b 0.1 M (KNO 0.10) 0.10) 25 (10.61 (10.89 ) 20 (10.69 0.10) ± ± ± 3 b c 0.1 KCl 30 (10.59 [63GRI/HUG] 0.10) 25 (10.67 0.10) 0.10) (10.96 ± ± ± b c 0.1 M (KNO 0.30) 25 (11.0 (11.3 [64YEN/LIU] 0.30) 0.30) ) 30 (10.92 ± ± ± 3 b c 0.1 M (NaClO 0.10) ) 20 (10.85 (11.05 0.10) 0.10) 25 (10.77 [70AND/MAL] ± ± ± 4 c 0.1 M (KNO 0.10) (11.07 [73HAN/RUZ] 0.10) ) 25 (10.78 ± ± 3 c 0.1 M (KNO 25 (10.73 [75AND/POD] 0.10) (11.02 0.10) ) ± ± 3 e 0.1 M (NaCl) (10.96 [79CRA/MOO] 0.10) ± (10.99 0.04) mean ± 2 4 2+ − − U Ca Ca(edta) + edta NCl) 25 (10.75 0.15) [83ARE/MUS] 0.17 M (Me ± 4 0.34 25 (10.61 0.15) ± 0.53 25 (10.47 0.15) ± 0.76 25 (10.46 0.15) ± + 2 − − U Ca(edta) Ca(Hedta) + H 0.1 M KCl (3.2 0.2) [47SCH/ACK] 20 ± ) 25 0.1 M (KNO 0.10) [75AND/POD] (2.94 ± 3 0.1 M (Me NCl) (3.1 0.2) [83ARE/MUS] 25 ± 4 final evaluation of selected values. a: Stability constant used in the ° C using enthalpy values ev aluated in this review (s ee Section VIII.5.3). At b: Temperature correction to 25 = 0.1 M the temperature correction for Ca from 20 to 25 ° C is ∆ log 0.001). I ± = − (0.081 K 1 10 3 − complexation using values evaluated in this review (see Table VIII-8-a and Table c: Corrected for Kedta VIII-8-b). At I = 0.1 M and 25 ° C the correction is ∆ log 0.015). K ± = (0.285 1 10 3 − d: Corrected for Na(edta) complexation using values evaluated in this review (see Table VIII-8-a). At . ° C the correction range is 1.1 < ∆ log I K < 2.8 depending on 25 1 10 cf. Appendix A). e: Evaluated in th is review (

494 VIII Discussion of data selection for edta ligand 452 2+ 4 2 − − U the data of [2001CHO/BON] , + edta Mg(edta) For the reaction Mg [64YEN/LIU] have been considered as one data set in the regression and [56SCH/REI] analysis. Together with the data of [83ARE/MUS] this resulted in a multi-linear least- squares regression analysis () . The selected values are: ο log K = (10.90 0.10), ± 10 1 1 − (Me (1.02 , 0.20) kg NCl) = mol ∆ε ± − ⋅ 4 1 − (NaCl)= (0.52 mol 0.04) kg . ± − ⋅ ∆ε 2– The SIT interaction coefficient (Mg(edta) (NaCl) allows an estimate of , ∆ε ε 4– + + 2+ – Na (Mg (edta (NaCl) + ) = , Cl ) + ) = – (0.52 0.04) + (0.32 0.14) + , Na ± ε ± ∆ε ε 1 − (0.19 0.02) = – (0.01 0.15) kg mol . ⋅ ± ± 2+ Multi-linear least-squares SIT regr ession plot for the reaction Mg + Figure VIII-19: − 4 2 − U edta . The data of Mg(edta) [56SCH/REI] , [64YEN/LIU] and [2001CHO/BON] the regression analysis. The results are have been considered as one data set in ο log K − 1 10 1 0.10), ∆ε (Me (NaCl) = NCl) = − (1.02 ± 0.20) kg ⋅ mol = (10.90 ± , ∆ε 4 − 1 mol − ⋅ (0.52 ± 0.04) kg . 14.0 13.5 13.0 12.5 D 16 12.0 + K 10 11.5 log 11.0 [83ARE/MUS] [2001CHO/BON] 10.5 [56SCH/REI] [64YEN/LIU] 10.0 0123456 / molal I m

495 VIII.5 Magnesium and calcium edta compounds and complexes 453 For the reaction: 4 2 2+ − − U Ca(edta) + edta Ca a weighted mean of log K = (10.99 = 0.04) was calculated from all accepted data at I ± 1 10 = 0.1 M (Table VIII-16). Extrapolation of this value to zero ionic strength assuming ∆ε 1 − 0.5) kg mol (0.5 (in analogy with the results obtained for Mg in NaCl but with an ⋅ ± − ο log K increased uncertainty) results in = (12.69 0.06). The constant obtained from ± 1 10 ο log K SIT extrapolation (Figure VIII-20) of data reported by [83ARE/MUS] , = 1 10 (12.70 0.22), fits well with the above results. The weighted mean of both values is ± selected in this review: ο log K = (12.69 0.06). ± 1 10 K for the formation of Figure VIII-20: Extrapolation to infinite dilution of log 10 1 2 − in tetramethylammonium chloride solution at 25 ° C [83ARE/MUS] . The Ca(edta) ο 1 − results are (1.06 = (12.70 ± 0.22) and ∆ε log . The NCl)= − K ± 0.31) kg ⋅ mol (Me 4 1 10 mean value derived from data at I = 0.1 M (Table VIII-16) is shown for comparison. 14.5 14.0 D 13.5 16 + K 10 13.0 log 12.5 83ARE/MUS 0.1 M mean value l 12.0 0.00.10.20.30.40.50.60.70.80.91.01.1 / molal I m

496 VIII Discussion of data selection for edta ligand 454 For the reaction, + 2 − − U + H Mg(Hedta) Mg(edta) [2001CHO/BON] the data of have been extrapolated to zero ionic strength. The results ο 1 − log K 0.49) and = (4.85 (NaCl) = (0.00 . The SIT interac- 0.13) kg mol are ⋅ ∆ε ± ± 10 – + (NaCl) allows an estimate of tion coefficient (Mg(Hedta) , Na ) = (NaCl) + ε ∆ε ∆ε + 2– + + (H , Na , Cl ) = (0.00 (Mg(edta) ) + 0.13) + (0.12 0.15) = 0.01) – (0.01 ε ± ± ± ε 1 − (0.11 mol 0.20) kg (Figure VIII-21). Extrapolation of the values of [47SCH/ACK] ⋅ ± [83ARE/MUS] and = (0.0 to zero ionic strength assuming 0.5) results in ∆ε ± ο log K 0.21) = (4.42 [83ARE/MUS] 0.21) [47SCH/ACK] and (4.50 . The weighted ± ± 10 mean of all three values is selected: ο log K = (4.5 0.2). ± 10 2 − K for the reaction Mg(edta) + Figure VIII-21: Extrapolation to infinite dilution of log 10 + − H U in NaCl at 25 ° C [2001CHO/BON] . The results are log = K º(VIII.14) MgHedta 10 − 1 (4.85 ± 0.49) and ∆ ± 0.13) kg ⋅ mol . The values of ε (NaCl)= (0.00 [47SCH/ACK] and [83ARE/MUS] are shown for comparison. 7.0 6.5 6.0 5.5 D 4 + 5.0 K 10 4.5 log 4.0 [2001CHO/BON] 3.5 [83ARE/MUS] [47SCH/ACK] l 3.0 0123456 / molal I m

497 VIII.5 Magnesium and calcium edta compounds and complexes 455 For the reaction: + − 2 − + H U Ca(Hedta) Ca(edta) , [83ARE/MUS] were extrapolated to zero , [75AND/POD] the data of [47SCH/ACK] – ± 0.5) (in analogy with Mg(Hedta) ionic strength assuming ∆ε = (0.0 ) resulting in ο ο = (3.75 ± log [47SCH/ACK] , K log K 0.21) = (3.45 ± 0.11) [75AND/POD] and 10 10 (3.66 ± 0.21) [83ARE/MUS] . The weighted mean of all three values is selected: ο K = (3.54 log ± 0.09). 10 2 2+ − + Ca U Ca edta(aq) no reliable data are available. For the reaction Ca(edta) 2 Enthalpy of complex formation VIII.5.3 rature are collected s extracted from the lite Enthalpy data for the Mg and Ca edta system in Table VIII-18. These data have been sc rutinised in order to obtain the accepted data (Table VIII-19) subsequently used for evaluating the values selected in this review. Table VIII-18: Experimental enthalpy data for the Mg and Ca edta systems. The uncertainities are given as reported in the references. If the ionic medium is shown in parenthesis, the contribution of the reacting species to the total ionic strength has been considered. − 1 ) Reference C) ( H ∆ (kJ ⋅ mol Method Ionic medium ° t rm 2+ 4 2 − − U Mg(edta) + edta Mg pK [54CAR/MAR] / 12 T I 25 0 0 − → ∂ − ∂ a c cal 25 13.0 [54CHA2] ? M NaNO 3 d 13.0 a 0.1 M KNO cal 20 (13.1 0.6) [56CAR/STA] ± 3 + (e) ? M Na [57JOR/ALL] /? 25 23.0 4) pK / T 0.1 M KNO [60BOH/MAR] 0 42 (8 − ± ∂ ∂ 3 a cal 0.1 M (KNO ) 20 14.60 [63AND] 3 (f) ? 20 25 20.1 [65PRI/SEB] − cal 0.3 M (KNO ) 15 (10.92 [76VAS/BEL2] 0.21) ± 3 (10.59 0.5 0.25) ± 1.0 (9.62 0.17) ± 25 (12.64 0.3 0.17) ± 0.5 (11.97 0.17) ± 1.0 (10.50 0.17) ± 0.3 35 (14.90 0.17) ± 0.5 (13.68 0.17) ± 1.0 (11.80 0.17) ± pK [83ARE/MUS] / 1.3) T 0.16 M (Me 45 (12.1 NCl) 10 − ± ∂ ∂ 4 a (Continued on next page)

498 VIII Discussion of data selection for edta ligand 456 Table VIII-18: (continued) − 1 Method Ionic medium H ∆ (kJ ⋅ mol ° C) ) Reference ( t rm 2 4 2+ − − U + edta Ca(edta) Ca / [54CAR/MAR] T pK 11 0 0 25 I − → ∂ ∂ − a c ? M NaNO 24.3 25 cal [54CHA2] − 3 d 27.2 − (b) [56CAR/STA] 20 0.6) (27.0 0.1 M KNO cal ± − 3 + 23.8 [57JOR/ALL] /? 25 ? M Na (e) − 0.54 – 0.86 M cal 25 (19.7 [58YAT/KAR] − ± 0.4) 1.58 25 24.02 [59YAT/KAR] cal 0.6 − − 23.10 − pK / [60BOH/MAR] T 0.1 M KNO 4) 0 (33 42 − ∂ ± ∂ − 3 a [63AND] ) 20 27.41 cal 0.1 M (KNO − 3 25 23.4 [65PRI/SEB] (f) ? 20 − − cal 0.3 M (KNO (30.46 0.21) [76VAS/BEL2] ) 15 − ± 3 0.33) (30.84 0.5 − ± 0.17) (30.92 1.0 ± − 0.3 25 (29.00 0.29) ± − (29.46 0.25) 0.5 − ± 0.17) (30.00 1.0 − ± 35 0.3 0.17) (27.41 − ± (28.12 0.17) 0.5 ± − 0.21) (28.53 1.0 − ± pK [83ARE/MUS] T 0.16 M (Me 1.3) NCl) 10 / 45 (25.1 − ± ∂ ∂ − 4 a 22.9 [99GRI] cal ? 25 − 2 + − − U + H Ca(Hedta) Ca(edta) 0.2 M KCl 25 8.24 [85MAR/EVA][86MAR/EVA] cal a: Actual ionic strength: 0.173 < I < 0.177 M. b: Actual ionic strength: = 0.165 M. I c: 1 equivalent edta in solution. d: 2 equivalents edta in solution. e: Thermometric titration. f: Continuous-flo w enthalpimetry. lved in Reaction (VIII.13), In order to obtain the heat invo 2+ 4 − 2 − U Medta + edta (VIII.13) , M it is very important to avoid the use of solutions in which other processes, accompanied by heat evolution, occur, especially if the corresponding evolved heat is not exactly known. This can be the case if the anion of the metal salt or the cation of the ligand edta

499 VIII.5 Magnesium and calcium edta compounds and complexes 457 are bound to the counterion in the separate solutions before their mixing. Then, in the subsequent complex formation, the heat consuming dissociation of the species present in the two initial solutions has to be taken into account. Such processes are investigated in solutions containing an inert salt, whose cation forms weak edta complexes. In gen- eral KNO was chosen as inert salt. In this cas e the best way to carry out calorimetric 3 measurements is using K edta and metal nitrates as reagents, and to maintain the ionic 4 strength near to that for which the equilibrium constant is known. This procedure was not always followed in calorimetric studies, causing widely varying results. For that reason the results of [54CHA2] [65WRI/HOL] , [57JOR/ALL] , are not accepted in this review. Enthalpy data obtained from the temperature variation of stability constants [54CAR/MAR] , [60BOH/MAR] , [83ARE/MUS] have also not been accepted in this review ( cf. Appendix A). [59YAT/KAR] , [65PRI/SEB] , , The rejection of the results of [58YAT/KAR] [85MAR/EVA] [86MAR/EVA] [99GRI] are discussed in Appendix A. , , have been included in the final [56CAR/STA] Note that only the Ca data of 2+ data evaluation because the authors used CaCl in the case of Ca , but MgSO in the 2 4 2+ case of Mg . The latter unfortunate choice causes additional ambiguities in the Mg [56CAR/STA] results and hence, the Mg data of have been discarded. have later been re- [75VAS/BEL] The enthalpy data for Mg edta reported in interpreted by [76VAS/BEL2] [76VAS/BEL2] . Only the corrected data set according to is included in Table VIII-18. Weighted least squares SIT-regression plots (Figure VIII-22 and Figure in this review (Table VIII -19) result in the following VIII-23) using the data accepted selected values: ο 1 − ∆ H 0.4) kJ mol ((VIII.13), M = Mg, 298.15 K) = (19.8 ⋅ ± rm 1 1 3 − − − mol 0.7) kg (0.5 K = 10 ⋅ ∆ε − ± ⋅ ⋅ L ο 1 − H ∆ (22.2 ((VIII.13), M = Ca, 298.15 K) = mol 0.4) kJ ⋅ − ± rm 1 3 1 − − − mol 0.6) = (2.4 K 10 kg ⋅ ⋅ ⋅ ± − ∆ε L , NaNO . Table and KCl ( cf For temperature corrections at = 0.1 M KNO I 3 3 VIII-17) this review calculates H ∆ ((VIII.13), 298.15K, I = 0.1 M) = 0.4) (27.2 ± − rm 2 − for Ca(edta) . This leads to = ∆ log K ∆ = − (0.081 ± 0.001) and K log 1 10 10 1 ± r a temperature correction from 20 to (0.079 0.001) using the van’t Hoff equation fo ° C and from 30 to 25 25 C, respectively. Likewise, fo r temperature corrections at I = 0.1 ° M KNO = 0.1 ( cf . Table VIII-17) this review calculates I H ∆ ((VIII.13), 298.15K, 3 rm − 2 1 − log ∆ K 0.4) kJ M) = (14.7 mol ± ⋅ . This leads to for Mg(edta) = − (0.042 ± 0.001) 10 1 using the van’t Hoff equation for a temperature correction from 30 to 25 ° C.

500 VIII Discussion of data selection for edta ligand 458 Table VIII-19: Accepted enthalpy data for the Mg and Ca edta systems to derive the [63AND] selected values. Uncertainties have been estimated in this review for or have [56CAR/STA] been multiplied by a factor (1.96) for , [76VAS/BEL2] to obtain error limits closer to a 95% total uncertainty level. -1 ° C) ) Reference H ∆ (kJ·mol Method Ionic medium t ( rm 2+ 4 − 2 − Mg U + edta Mg(edta) ) 20 (14.60 ± 0.40) [63AND] 0.1 M (KNO cal 3 0.3 M (KNO ) 25 (12.64 cal 0.33) [76VAS/BEL2] ± 3 (11.97 0.5 ± 0.33) ± 0.33) 1.0 (10.50 − 4 2+ − 2 Ca Ca(edta) U + edta cal 0.165 M − (27.0 ± 1.2) [56CAR/STA] 20 cal ) 20 − (27.41 0.1 M (KNO 0.40) [63AND] ± 3 cal 0.3 M (KNO ) 25 − (29.00 ± 0.57) [76VAS/BEL2] 3 − (29.46 ± 0.49) 0.5 (30.00 ± 0.33) − 1.0 Figure VIII-22: Weighted least squares SIT-regression plot of enthalpy data from 2 − ο . The results are for the formation of Mg(edta) , = H [76VAS/BEL2] ∆ [63AND] rm 1 − 1 − 3 − 1 − mol ⋅ 0.4) kJ ± (19.8 K 10 and ∆ε kg ⋅ − (0.5 ± ⋅ mol 0.7) ⋅ . = L 22 1 − 21 / kJ·mol L 20 D + 16 H r 19 ∆ [63AN D] [76VAS/BEL2] 18 1.0 0.8 0.6 1.2 0.2 0.0 0.4 I / molal m

501 VIII.5 Magnesium and calcium edta compounds and complexes 459 Figure VIII-23: Weighted least squares SIT-regression plot of enthalpy data from − 2 [56CAR/STA] [76VAS/BEL2] , [63AND] , . The results for the formation of Ca(edta) ο − − 3 − 1 1 1 − H ∆ are (2.4 − = mol ⋅ kg K ∆ε and ⋅ 10 mol ⋅ 0.4) kJ ± (22.2 − . = ⋅ 0.6) ± L rm -19 -1 -20 / kJ mol L -21 D + 16 H [63AND] r -22 ∆ [76VAS/BEL2] [56CAR/STA] Series11 Series12 Series13 -23 0.0 0.4 0.2 0.8 1.2 0.6 1.0 I / molal m Selenium edta compounds and complexes VIII.6 No information about thermodynamic properties of Se edta compounds and complexes could be found in the literature. VIII.7 Nickel edta compounds and complexes VIII.7.1 Nickel edta compounds A considerable number of nickel containing edta compounds have been reported in the t also amorphous edta complex compounds literature (Table VIII-20). Crystalline, bu ecause of their magnetic properties. have received particular attention b Solubility measurements in H O have been reported for a few of these com- 2 pounds. Astakhov and Verenikin [53AST/VER] O a solubility of edta ⋅ 5H report for Ni 2 2 1 − 33 g per 100 g H ≈ 0.7 mol·kg O ( ) and for Na Ni(edta) ⋅ 2H O 34.4 g per 100 g H O 2 2 2 2 1 − ( 0.8 mol·kg ≈ C, the pH of the resulting solutions ° ). Both values were measured at 17 et al. [74MYA/LOG] is not reported. Myachina give solubility values of 0.129 − 3 − 3 mol·dm edta ⋅ 6H O, and 0.130 mol·dm for Ni for MgNi(edta) ⋅ 6H O, both measured 2 2 2 at (25.0 ± 0.1) ° C. The pH of the resulting solutions is not given. It is outside the scope

502 VIII Discussion of data selection for edta ligand 460 of this review to develop a quantitative thermodynamic model for these highly soluble Ni edta compounds. − ckel edta compounds. References reporting solubility data are marked Table VIII-20: Ni crystal structures ar e marked with (str.), with (sol.). References reporting X-ray single ttering) analyses of amorphous solids; (en) (am.) refer to LAXS (Large-Angle X-ray Sca stands for ethylenediamine (C N H ) and (big) represents biguanide (C ). H N 7 2 8 5 2 2 Solubility in H Compound O Reference 2 –1 5H ⋅ O 0.7 mol·kg edta [53AST/VER] (sol.) Ni 2 2 –3 edta 6H , O 0.129 mol·dm Ni [74MYA/LOG] (sol.), [81BEL/DRI] , [82BEL/ESC] ⋅ 2 2 [84ESC/FUE] [85LEC/MOS] (am.), [84MOS/GAL] , (am.), , (str.), [88CAM/MUN] (am.), [86COR/DRI] [86ESC/FUE] [93ATZ/FIL] , [95BOR/COR] Ni(H edta) [42BRI/HES] , , [62AXT/HAN] [42KLE/RAD] 2 Ni(H edta) ⋅ H O [59SMI/HOA] (str.), [67BHA/KRI] 2 2 Ni(H edta) O [93SPI/RIB] ⋅ 2.5H 2 2 LiNi(Hedta) [86POL/FIL] (str.) –1 Na 2H (sol.) O 0.8 mol·kg Ni(edta) [53AST/VER] ⋅ 2 2 Na ⋅ O [59SAW/PAU] 4H Ni(edta) 2 2 Na ⋅ 6H Ni(edta) O [49PFE/SCH] 2 2 K Ni(edta) [90KOZ/VAS] 2 –3 MgNi(edta) [84ESC/FUE] ⋅ O 0.130 mol·dm (sol.), 6H , [86ESC/FUE] [74MYA/LOG] 2 CaNi(edta) 2H [49PFE/SCH] O ⋅ 2 CaNi(edta) 4H O [84NES/POR2] (str.) ⋅ 2 MnNi(edta) , ⋅ O [86DRI/COR] , [84ESC/FUE] , [86ESC/FUE] 6H 2 [88TOG/KOJ] [89COR/DRI] , Co [86GOM/JAM] ⋅ 2H Ni O edta 0.7 2 1.3 CoNi(edta) O ⋅ [86GOM/JAM] 2H 2 Co O ⋅ 2H edta [86GOM/JAM] Ni 0.5 1.5 2 CoNi(edta) 6H , O [81BEL/DRI] ⋅ [82BEL/ESC] , [84ESC/FUE] , 2 [84MOS/GAL] (am.), [86ESC/FUE] Co [Ni(edta)] 4H O [92SAP/COR] ⋅ 2 2 2 Ni[Cu(Hedta)] ⋅ O [49PFE/SCH] H 2 2 Ni(en)[Ni(edta)] 4H O [81SYS/AGR] (str.) ⋅ 2 Ni(en) (str.) [Ni(edta)] ⋅ 4H [86SYS/AGR] O 2 3 Ni(big) ⋅ 7H O [56DUT/RAY] [Ni(edta)] 2 2 Cu(en) [Ni(edta)] ⋅ 2H O [80AGR/SYS] (str.), [90KOZ/VAS] 2 2 Cu(en) [Ni(edta)] 4H O [85LAN/KRA] , ⋅ 2 2 Cu(big) [Ni(edta)] ⋅ 7H O [56DUT/RAY] 2 2 Zn[Ni(edta)] (str.) ⋅ O [84ESC/FUE] , [86ESC/FUE] , [87LEO/FRI] 6H 2 Pb[Ni(Hedta)(H O)]Cl [91LI/ZHA] 2 N H [94SAR/SIV] [Ni(Hedta)] ⋅ H O 5 2 2 Li Ni(edta) ⋅ 4Al(OH) ⋅ 4H O [95ISU/TAR] 3 2 2

503 VIII.7 Nickel edta compounds and complexes 461 Stability of nickel edta complexes VIII.7.2 Complex formation in Ni(II) edta systems has been studied by several investigators. The experimental equilibrium data are summarised in Table VIII-21 and Table VIII-22. In qualitative terms these data reveal that in aqueous solution edta may form the following − 2 − 3 − 2+ complexes with Ni : Ni(H , Ni(edta) and other edta)(aq), Ni(Hedta) , Ni(edta)OH 2 ternary complexes Ni(edta)X, where X is a second ligand. The equilibrium constants of these species usually refer to the following reac- tions: 2+ 4 − − 2 Ni Ni(edta) + edta (VIII.16) U 2 + − − Ni(Hedta) + H (VIII.17) U Ni(edta) − + + H U Ni(H edta)(aq) (VIII.18) Ni(Hedta) 2 (2+n) 2 n − − − Ni(edta)X (VIII.19) + X Ni(edta) U Table VIII-21: Experimental equilibrium data for the Ni edta system. The uncertainties are given as reported in the references. If the ionic medium is shown in parenthesis, the contribution of the reacting species to the total ionic strength has been considered. t ( ° C) log Method Ionic medium K Reference 10 4 − 2+ 2 − U Ni(edta) Ni + edta 63 Ni 20 19 [51COO/LON] 0.1 M (H,K)Cl 0.1 M (KCl) (18.56 20 gl ± 0.07) [51SCH/FRE] ≈ 0.1 M ? ? 17.6 [52MAR/PLU] sp 30 17.5 [53HUG/MAR] 0.1 M KNO sp 3 pol 0.1 M (KNO ) 20 (18.62 ± 0.06) [54SCH/GUT] 3 [63STA] 0.06) 20 (18.36 ± dis 0.1 M KClO 4 20 19 [65JOK/MAJ] dis 0.1 M KNO 3 pol 0.2 M KNO 25 18.12 [65OGI] 3 ± 1) 18.79 [74TER/NIK2] sol 0.10 - 0.12 M (22 ) 25 (18.52 ± 0.05) [75AND/POD] 0.1 M (KNO gl 3 sp ≈ 0.1 M (KCl) (20 ± 2) 17.83 [78KOR/VAL] (Continued on next page)

504 VIII Discussion of data selection for edta ligand 462 Table VIII-21: (continued) ( t ° K Reference Method Ionic medium C) log 10 − 2 + − + H U Ni(edta) Ni(Hedta) 25 pot [51COO/LON] 0.1 M KCl 3 0.1 M (KNO ) 20 (3.2 ± 0.1) [54SCH/GUT] pol 3 25 2.9 [58COO/LON] pot 1.25 M NaClO 4 [63BHA/KRI] 25 2.73 1 M NaClO sp 4 3.12 0.1 M (KNO gl ± 0.07) [69BRU/NAN] ) 25 (3.23 3 3.4 [74TER/NIK2] 0.01 - 0.15 M sol 22 1 M KNO gl 25 2.96 [79JAN/PFE] 3 − ? 3.27 [88EVS/SMI] 0.3 M gl 0.2 − + + H U Ni(H edta)(aq) Ni(Hedta) 2 sol 22 [74TER/NIK2] 0.02 - 0.06 M 1.4 [79JAN/PFE] 25 0.99 1 M KNO gl 3 − ? 1.46 [88EVS/SMI] 0.3 M gl 0.2 − 6 − 4 − 2 + edta Ni(edta) U Ni(edta) 2 − 0.3 M ? 1.95 [88EVS/SMI] gl 0.2 4 − 3 − − + Hedta Ni(Hedta) Ni(Hedta) U 2 gl 0.2 − ? 4.73 [88EVS/SMI] 0.3 M 2 − 2 − Ni(H Ni(H edta) edta)(aq) + H U edta 2 2 22 gl 0.2 − 0.3 M ? 6.90 [88EVS/SMI]

505 VIII.7 Nickel edta compounds and complexes 463 Table VIII-22: Experimental equilibrium data for the Ni edta X system where the ligand X forms a ternary complex Ni(edta)X. The un certainities are given as reported in the references. If the ionic medium is shown in parenthesis, the contribution of the reacting species to the total ionic strength has been considered. − 1 t K X Method Ionic medium ∆ ( H Reference kJ ⋅ mol ° C) log r m 10 n − − 2 (2+ n ) − + X Ni(edta) Ni(edta)X U − cal OH 0.1 M KNO 20 (1.8 ± 0.3) ≈ 13 [56CAR/STA] 3 sp 25 0.41 [63BHA/KRI] 1 M NaClO 4 0.55 cal 1.5 M (KNO ± 0.05) − (9.92 ) 25 (0.93 0.05) [85VAS/KOZ] ± 3 2 − sp [73BAR/FRI] NO 25 1.43 1.5 M NH ox 3 4 ) ? (0.48 ± 0.08) [82VAS/VAS] sp 1.0 M (KNO 3 ) (0.54 ± 0.08) 1.5 M (KNO 3 2.0 M (KNO (0.63 ) 0.13) ± 3 a 3 − 25 8.01 pot ? [91BAP] cit a 35 8.19 a 45 8.36 − CN kin 0.1 M NaClO 25 3.76 [70COO/MAR] 4 25 2.26 sp 1 M NaClO NH [63BHA/KRI] 4 3 2.56 [67JAC/MAR] 0.05) ± (1.35 25 sp 0.5 M KCl 25 (1.39 ± 0.01) 1 M NaClO 4 sp 1.5 M NH 25 1.30 NO [69FRI/DYA] 4 3 ± ) ? (1.24 0.07) [82VAS/VAS] 1.0 M (KNO sp 3 0.05) ± ) (1.21 1.5 M (KNO 3 2.0 M (KNO ) ± 0.02) (1.95 3 OH sp 1 M NaClO [65BHA/RAD] 25 (1.52 ± 0.06) NH 4 2 NH [65BHA/RAD] 0.06) 1 M NaClO 25 (1.65 ± sp NH 2 2 4 − 4 P [82VAS/VAS] 0.08) sp 1.25 M (KNO O ) ? (0.82 ± 3 7 2 ) (0.88 ± 0.06) 1.5 M (KNO 3 ) (0.95 ± 0.12) 2.0 M (KNO 3 pyridine 1 M NaClO sp ± 0.03) [65BHA/RAD] 25 (1.68 4 NO 25 0.33 [69FRI/DYA] sp 1.5 M NH 4 3 [65BHA/RAD] 0.05) ± 25 (2.25 1 M NaClO ethylene- sp 4 cal diamine 1.5 M (KNO ) 25 (2.36 ± 0.02) (46.0 ± 0.9) [81VAS/VAS] − 3 1.3) ± ± 0.04) − (46.0 35 (2.13 ≈ 0.05 M 25 (2.74 ± 0.35) [83KOR/PRO] sp propylenediamine sp 1 M NaClO 25 (2.3 ± 0.1) [65BHA/RAD] 4 (Continued on next page)

506 VIII Discussion of data selection for edta ligand 464 Table VIII-22: (continued) 1 − kJ K X Method Ionic medium ∆ Reference H ( C) log ⋅ mol t ° m r 10 − − n 2 ) n (2+ − Ni(edta)X + X Ni(edta) U 25 3.04 sp NO glycine 1.5 M NH [73BAR/FRI] 3 4 cal 1.5 M (KNO ) 25 (1.16 ± 0.05) − (21.8 ± 0.8) [82VAS/BEL] 3 1.5 M (KNO cal ) 25 (1.0 − (22.2 ± 1.3) [86VAS/KOZ] 0.2) ± 3 [87VAS/KOZ] 0.7) ± ) 25 (0.8 ± 0.1) − (4.0 cal ida 1.5 M (KNO 3 K [91BAP] is not clear, see Appendix A. a: The meaning of log in 10 in Table VIII-21 and Table VIII-22 that We can infer from the data collected the species Ni(H edta)(aq) predominates in solutions at pH < 2, and the species 2 − 2 − Ni(Hedta) has a very large pH range of pre- predominates from pH 2 to 4. Ni(edta) 3 − is already present and dominance from pH 4 to 13. At pH 13 the species Ni(edta)OH its concentration increases with increasing pH. The entries [51SCH/FRE] , , , [79JAN/PFE] in [69BRU/NAN] [75AND/POD] Table VIII-21 referring to the method “gl” (glass electrode) need some comments. In all 2+ these studies solutions with equimolar amounts of Ni edta have been used. In and H 4 − − 2 such solutions the 1:1 complexes Ni(H , Ni(edta) and edta)(aq), Ni(Hedta) 2 3 − Ni(edta)OH are present, depending on the pH of the solutions. Simple base titrations of such solutions yield log K (VIII.18) and log (VIII.17) values of the two protonated K 10 10 − species Ni(H , respectively, if the initial pH is very low edta)(aq) and Ni(Hedta) 2 − [79JAN/PFE] . If the initial pH is higher only log K (VIII.17) of Ni(Hedta) is obtained 10 − 2 [69BRU/NAN] . At pH > 4, Ni(edta) is the only Ni-edta complex present in equimolar solutions and simple base titrations yield no further information. However, if a suitable 3+ 2+ protonated polyamine (H tren , tren = tris(2-aminoethyl)amine) and Ca are present in 3 2 − 2+ such solutions at pH > 4, the Ni(edta) epwise Ni(tren) complex forms st and 2 − Ca(edta) by base titration. From the involved exchange equilibrium: 22+ 3+ 2+ 2 + − − (VIII.20) Ni(tren) + Ca(edta) + 3H Ni(edta) + Ca + H tren U 3 2 − is obtained if the protonation constants of the two the equilibrium constant of Ni(edta) − 2 2+ formation constant of Ni(tren) ligands edta and tren, and the are known and Ca(edta) [51SCH/FRE] , [75AND/POD] . Ni(II) has a coordination number of six, and edta is an ideal candidate for a 2+ perfect octahedral coordination with Ni in aqueous solution. This poses the question why any ternary complexes Ni(edta)X have been observed at all (Table VIII-22). The study of ternary Ni(edta) X species in 1 M NaClO ° C by Higginson and Samuel and 25 4 − 2 [70HIG/SAM] reveals that edta acts in Ni(edta) as hexadentate ligand only for 75% of the complex species, the remaining part being aquo-pentadentate. In the latter case, the exchange of the bound H O molecule with X could explain the formation of ternary 2

507 VIII.7 Nickel edta compounds and complexes 465 ds weak to moderately stable ternary species Ni(edta)X. Whereas for most ligan Ni(edta)X complexes have been reported ( log K (VIII.19) < 3), the most stable com- 10 − plexes seem to be formed with CN K (VIII.19) > 3). Also dinuclear ternary com- ( log 10 [72BAR/FRI] plexes could be formed [73BAR/FRI] . , , [73BAR/DYA] 2 − 2 − O) can be distinguished and Ni(edta)(H Although the two species Ni(edta) 2 by different colours in aqueous solution [70HIG/SAM] , no equilibrium data are avail- able for the reaction: 2 − 2– Ni(edta) U Ni(edta)(H O O) + H (VIII.21) 2 2 Usually equilibrium constants refer to equilibria with unambiguous stoichiome- try. This is not the case for equilibrium (V III.16) where the product actually represents 2 − 2 − and Ni(edta)(H . However, due to the O) the sum of the concentrations of Ni(edta) 2 n (VIII.21) this sum ca lack of an equilibrium constant for reactio nnot be de-convoluted. Table VIII-21 and Table VIII-22 have The experimental results collected in been scrutinised in order to select reliable studies, summarised in Table VIII-23, on which the evaluation of selected values is based. Table VIII-23: Accepted formation constants for the Ni edta systems to derive the selected values. Uncertainties have been estimated in this review. a log K K log K log Reference at 25 ° C ° at 20 Ionic medium C 10 10 10 2+ − 4 − 2 Ni(edta) Ni (VIII.16) + edta U b c (18.47 ± 0.1 M (KCl) (18.56 ± (18.75 ± 0.14) 0.14) [51SCH/FRE] 0.14) c b ) (18.62 ± (18.53 ± 0.20) 0.20) (18.81 ± 0.20) 0.1 M (KNO [54SCH/GUT] 3 c (18.52 ± 0.10) (18.81 ± 0.10) ) [75AND/POD] 0.1 M (KNO 3 ± 0.08) weighted mean (18.79 2 − − + Ni(Hedta) + H (VIII.17) U Ni(edta) d ) (3.2 ± 0.3) (3.18 ± 0.30) 0.1 M (KNO [54SCH/GUT] 3 (3.23 ± 0.14) [69BRU/NAN] 0.1 M (KNO ) 3 ± 0.13) weighted mean (3.22 a: Stability constant used in the final evaluation of selected values. (see Section VIII.7.3). ° C using enthalpy values evaluated in this review b: Temperature correction to 25 At = 0.1 M the temperature correction from 20 to 25 I − C is K ∆ = log (0.093 ± 0.001). ° 10 3 − complexation using values ev aluated in this review (see Table VIII-8-a). At I = c: Corrected for Kedta 0.1 M and 25 C the correction is K log ° ∆ = (0.285 ± 0.015). 10 = 0.1 ° C using enthalpy value of [69BRU/NAN] (see Table VIII-24). At I d: Temperature correction to 25 M the temperature correction from 20 to 25 ± C is log K ∆ ° 0.004). = – (0.022 10 The constants reported in [51COO/LON] would need corrections for the proto- − + nation of H edta to H and of Ni(Hedta) to Ni(H edta edta)(aq) because the pH range 2 4 5 in this study extended to very low values. However, the equilibrium data needed for such corrections at I = 0.1 M (H, K)Cl are not available and thus, the equilibrium con-

508 VIII Discussion of data selection for edta ligand 466 stants reported by [51COO/LON] are rejected in this review. log K [51SCH/FRE] The study of (VIII.16) value is reliable. However, the 10 1 originally reported in [51SCH/FRE] has been corrected in [54SCH/GUT] considering − 2 2+ . This corrected value is and Ca(edta) revised equilibrium constants for Cu(tren) included in Table VIII-21; its uncertainty has been increased by a factor of two in order to account for systematic errors (Table VIII-23). 2 − [53HUG/MAR] In the study of the dissociation of the complex Ni(edta) was + − investigated by increasing [H ] but without considering the formation of Ni(Hedta) in the data analysis ( Appendix A). Thus, their reported value is not considered in this cf. review. For the polarographic measurements of [54SCH/GUT] an acetic acid/sodium el had been added as NiSO acetate buffer was used and nick salt. The effects of sodium 4 acetate and sulphate on the nickel-edta equilib rium are difficult to assess. Nevertheless, the values reported by [54SCH/GUT] review but with increased are considered in this uncertainties (Table VIII-23). − 2 The high edta concentration (1 10 × [63STA] causes a significant M) used by I = 0.1 to 0.2 M during the experiments ( cf. Appen- increase of the ionic strength from dix A). This increase is outside the limits wh ere constant ionic activity coefficients can be assumed. Thus, the log K [63STA] is not considered in (VIII.16) value reported in 10 this review. The log (VIII.17) value of [69BRU/NAN] is reliable; its uncertainty has K 10 r to account for systematic errors (Table been increased by a factor of two in orde VIII-23). The log K (VIII.16) value of is reliable, its uncertainty has [75AND/POD] 10 r to account for systematic errors (Table been increased by a factor of two in orde VIII-23). The values reported in [79JAN/PFE] are rejected in this review because the e for their pH measurements. authors used the activity scal The results reported in [56MAR] , [63BHA/KRI] , [65JOK/MAJ] , [65OGI] , , and [88EVS/SMI] are not considered in this review ( cf. [78KOR/VAL] [74TER/NIK2] Appendix A). Ternary complexes (Table VIII-22) are evaluated in this review if the second ligand X refers to hydroxide, oxalate or citr ate. All other data in Table VIII-22 (below the dotted line) are given for illustrative purposes only. − 3 The species Ni(edta)OH is present in solutions of pH > 12 and only a few papers report quantitative data concerning its stability [56CAR/STA] , [63BHA/KRI] , [85VAS/KOZ] . However, as discussed in Appendix A, all these studies suffer from ο log K various shortcomings, and no value for (VIII.19) with X = OH can be recom- 10 mended.

509 VIII.7 Nickel edta compounds and complexes 467 4 − The stability of the species Ni(edta)(ox) has been investigated by two groups [73BAR/FRI] [82VAS/VAS] . The results obtained at similar ionic strength differ by , [82VAS/VAS] one order of magnitude (Table VIII-22). The results of have been re- jected by this review ( Appendix A), and based on a single determination at 1.5 M cf. NH [73BAR/FRI] NO , which is not unreasonable but needs confirmation, no value for 4 3 ο log K (VIII.19) with X = ox can be recommended. 10 . Because [91BAP] There is only one study of te rnary Ni-edta-citrate complexes of various shortcomings the paper of [91BAP] Appen- cf. is rejected by this review ( dix A). Summarising, for the evaluation of log K (VIII.16) remain only the values 10 , [54SCH/GUT] at 20 ° C (Table [75AND/POD] at 25 ° C and [51SCH/FRE] from VIII-23). The values at 20 ° C are corrected to 25 ° C by log K ∆ = − (0.093 ± 0.001), 10 1 − obtained with the van’t Hoff equation using = ± 0.4) kJ ⋅ mol ∆ H evaluated in − (31.0 rm this review (see Section VIII.7.3). In addition, all log values at 25 ° C are corrected K 10 + by ± = (0.285 log 0.015) in order to account for K K complexation with ∆ 10 3 − Kedta equilibrium constants recommended for 0.1 M KCl and 0.1 M KNO in this 3 VIII.3.7). A weighted mean of review (Table VIII-8-a, Section log K (VIII.16) = 10 ± 0.08) is obtained at I = 0.1 and 25 ° C. Because only data for I = 0.1 M are con- (18.79 ο ∆ε i.e. , , sidered the SIT extrapolation can be done neglecting log K (VIII.16) = 18.79 10 2 – ∆ z × D = 18.79 + 16 × 0.1095 = (20.54 ± 0.08). Assuming |∆ε| ≤ 1, extrapolating from I = 0.1 M to zero ionic strength results in an additional uncertainty of 0.1 at maxi- ± mum. Hence, the selected value is: ο K log (VIII.16)= (20.54 ± 0.13). 10 log K (VIII.17) remain only the values from For the evaluation of 10 [54SCH/GUT] C and [69BRU/NAN] at 25 at 20 C (Table VIII-23). The value at 20 ° C ° ° ° C by is corrected to 25 log ∆ = − (0.022 ± 0.004), obtained with the van’t Hoff K 10 − 1 2 − equation using mol ± 1.3) kJ ⋅ H ∆ = , the enthalpy of protonation of Ni(edta) − (7.5 rm measured at 0.1 M KNO (VIII.17) = [69BRU/NAN] . A weighted mean of log K 10 3 ± I = 0.1 and 25 ° C. Because only data for I = 0.1 M are con- (3.22 0.13) is obtained at ο , i.e sidered the SIT-extrapolation can be done neglecting ∆ε . log K (VIII.17) = 3.22 − 10 2 ∆ z × D = 3.22 + 4 × 0.1095 = (3.66 ± 0.13). Assuming |∆ε| ≤ 1, extrapolating from I = 0.1 M to zero ionic strength results in an additional uncertainty of ± 0.1 at maximum. Hence, the selected value is: ο K ± (VIII.17) = (3.66 log 0.16). 10 3 − edta)(aq), Ni(edta)OH , No equilibrium constants can be selected for Ni(H 2 4 − 5 − Ni(edta)(ox) , and Ni(edta)(cit) .

510 VIII Discussion of data selection for edta ligand 468 VIII.7.3 Enthalpy of nickel edta complex formation Enthalpy data for the Ni-edta systems extracted from the literature are collected in Table VIII-24. These data have been scrutinised in order to obtain the accepted data subse- quently used for evaluating the values selected in this review. Table VIII-24: Experimental enthalpy data for the Ni edta system. The uncertainities are given as reported in the references. If the ionic medium is shown in parenthesis, the contribution of the reacting species to the total ionic strength has been considered. − 1 C) ∆ Reference H Method Ionic medium kJ ⋅ mol t ( ° m r 2+ 4 − 2 − Ni(edta) Ni + edta U b 25 − 31.8 [54CHA2] ? M NaNO cal 3 c − 32.6 a 0.1 M KNO cal 20 − (34.94 ± 0.63) [56CAR/STA] 3 + (d) ? M Na /? 25 − 31.0 [57JOR/ALL] cal 0.22 1.58 M 25 − 28.79 − [59YAT/KAR] 31.30 − cal 0.1 M (KNO ) 20 − 31.59 [63AND] 3 ? 20 - 25 (e) − 30.1 [65PRI/SEB] cal 0.1 M KNO 25 − 35.6 [65WRI/HOL] 3 cal 0.3 M (KNO − ± 0.17) [76VAS/BEL3] ) 15 (33.76 3 0.5 ± 0.21) − (34.43 1.0 (35.56 0.21) − ± 0.3 25 − ± 0.17) (32.47 0.5 (33.10 ± 0.13) − 1.0 − 0.21) (34.64 ± 0.3 35 − ± 0.17) (31.63 0.5 (32.01 ± 0.08) − 1.0 − (33.35 ± 0.23) 2 − + − U Ni(Hedta) + H Ni(edta) 0.1 M (KNO ) 25 − (7.5 cal 1.3) [69BRU/NAN] ± 3 2 − − − 3 U + OH Ni(edta)OH Ni(edta) cal 20 ≈ 13 [56CAR/STA] 0.1 M KNO 3 cal 1.5 M (KNO ) 25 − (9.92 ± 0.05) [85VAS/KOZ] 3 a: Actual ionic strength: 0.173 < I < 0.177 M. d: Thermometric titration. b: 1 equivalent edta in solution. e: Continuous-flo w enthalpimetry. c: 2 equivalents edta in solution.

511 VIII.7 Nickel edta compounds and complexes 469 was chosen In order to obtain the heat involved in (VIII.16), in general KNO 3 carry out calorimetric measurements is using as inert salt. In this case the best way to K edta and metal nitrates as reagents, and to maintain the ionic strength near to that for 4 which the equilibrium constant is known. This procedure was not always followed in calorimetric studies, causing widely varying results. For example, [56CAR/STA] used NiSO , solutions. For that reason the results of [54CHA2] in KNO [56CAR/STA] , 3 4 [57JOR/ALL] cf. are not accepted in this review ( Appendix A). , [65PRI/SEB] , , [85VAS/KOZ] [65WRI/HOL] [59YAT/KAR] The results of are also rejected (see the discussions in Appendix A). For the final evaluation of enthalpy data for reaction (VIII.16) remain the val- 0.3 estimated in this review) and the data of (with an uncertainty of ± [63AND] ues of [76VAS/BEL3] C. In the latter case the uncertainties given in Table VIII-24 have at 25 ° error limits closer to a 95% total uncertainty been multiplied by a factor (1.96) to obtain ., including random and possible systematic errors. levels, i.e [63AND] A weighted linear regression using the enthalpy data of and the 25 ° C 1 − ο data of [76VAS/BEL3] gives ((VIII.16), 298.15K) = − (26.1 ± 0.4) kJ ⋅ mol H ∆ with rm − 3 − 1 − 1 ∆ε × 10 (Figure VIII-24). However, in evident contrast = kg ⋅ K − (2.1 ⋅ mol ± 0.7) L 2+ 2+ with the results obtained for Mg where the two data sets of and [63AND] and Ca [76VAS/BEL2] reveal a good overall linear correl ation (Figure VIII-22 and Figure 2+ in the case of Ni VIII-23), a strong deviation from the expe cted linearity is seen (Figure VIII-24). Figure VIII-24: Weighted least squares SIT-regression plot of enthalpy data from 2 − , for the formation of Ni(edta) . The results are [76VAS/BEL3] [63AND] ο − − 1 − 3 1 1 − ∆ H ± and ∆ε mol = − (2.1 ± 0.7) × 10 (26.1 − kg ⋅ K (VIII.16) = 0.4) kJ ⋅ mol ⋅ . L rm -23 -24 1 − -25 / kJ·mol L D -26 + 16 H r [63AND] ∆ -27 [76VAS/BEL3] -28 1.2 1.0 0.8 0.6 0.4 0.2 0.0 I / molal m

512 VIII Discussion of data selection for edta ligand 470 A possible explanation is that for the values at low ionic strength in Figure 2 − VIII-24 the complex Ni(edta)(H I O) values predominates, whereas at the larger 2 2 − Ni(edta) becomes the predominating complex (see discussion in Section VIII.7.2 of [70HIG/SAM] the results of Higginson and Samuel in terms of reaction (VIII.21)). The values of the enthalpy of complex formation H ∆ of the two complexes are expected rm (VIII.16). Enthalpy data to be different. The same statement in principle is true for K ed complex than the perhaps reveal much better a change in the structure of the form formation constants. This can be explained with the large compensation of the terms for H S ∆ ∆ and in reaction (VIII.21). The values of the conventional stability con- rm rm stants of similar complexes involving other ligands are very near [71AND/WEN] , but for edta such data have not been reported. Excluding the value for I = 1 M (KNO ) in the regression analysis, an ap- 3 − 2 2 − ο O) i.e can be derived, ∆ (Ni(edta)(H . H O) , proximate value for Ni(edta)(H 2 2 rm − 1 3 − − − 1 1 mol ⋅ 0.4) kJ (26.7 − 298.15 K) = ± 1.3) × 10 ∆ε − kg ⋅ K = and ⋅ mol (5.1 ± . How- L ever, this question can only be solved by more extended measurements, and this review selects the values derived from all experimental data in Figure VIII-24 , i.e ., ο − 1 ∆ (VIII.16) = − (26.1 0.4) kJ ⋅ mol H ± rm − 3 − 1 − 1 10 mol − kg ⋅ K (2.1 ± ⋅ = 0.7) × . ∆ε L This selection yields: ο –1 2– ∆ (Ni(edta) H , 298.15 K) = – (1785.9 ± 3.9) kJ·mol . fm and KCl ( cf. Table VIII-23) this I = 0.1 M KNO For temperature corrections at 3 − 1 review calculates ∆ ((VIII.16), 298.15 K, I = 0.1M) = − (31.0 ± 0.4) kJ ⋅ mol . This H rm − 0.001) using the van’t Hoff equation for a temperature K = ± (0.093 log ∆ leads to 10 ° C. correction from 20 to 25 Only one study concerns the heat involved in reaction (VIII.17) [69BRU/NAN] . The result seems to be reasonable and it has been used for temperature corrections in this review (Table VIII-23). However, based on a single determination ο only, which is not unreasonable but needs confirmation, no value for H ((VIII.17), ∆ rm 298.15 K) can be recommended. Technetium edta compounds and complexes VIII.8 VIII.8.1 Technetium edta compounds complex salt of edta, H (TcO) O, has been deter- (edta) ⋅ 5H The structure of a Tc(IV) 4 2 2 2 . However, no chemical thermodyna mic data are available for this mined [81BUE/AND] compound.

513 VIII.8 Technetium edta compounds and complexes 471 VIII.8.2 Technetium edta complexes Only three studies report data on the complex formation between Tc and edta. Gorski and Koch dissolved freshly prepared hydrous oxide of [69GOR/KOC] [70GOR/KOC] , Tc(IV) in HClO n containing an unknown trace con- and obtained an aqueous solutio 4 centration of technetium(IV). In the pH range 1 to 2.5 they investigated the ionic mobil- ity by electrophoresis, and they studied the formation of complexes in the presence of complexing ligands by cation exchange and by electrophoresis. The results were inter- 2+ + preted by assuming the presence of TcO at pH 1, reacting to TcO(OH) and 2 − TcO(OH) , (aq) at higher pH, and the concomitant formation of TcO(OH)nta 2 − 5 − 3 and TcO(OH)edta . The unexpectedly high value of TcO(OH)(nta) = 19.1 K log 1 10 2 for the reaction: + 4- 3- TcO(OH) + edta TcO(OH)edta (VIII.22) U + pO ( log K = N differs considerably from the value evaluated in this review for 10 1 2 (9.23 0.13) cf. Section VIII.11.2.3). ± Recent measurements show that Tc(IV) is only stable in reducing solutions. Consequently, experimental studies of Tc(IV) demand an exact control of the redox state and the composition of the species present in solution. However, Gorski and Koch [69GOR/KOC] [70GOR/KOC] do not mention any measures to control the redox state , of Tc in their experiments. The crystal structure determination of H (edta) (TcO) shows [81BUE/AND] 4 2 2 μ− O bridges with the two oxygen atoms, that the two Tc metal ions form two equal ganic ligand (see Figure VIII-4-a in Section leaving four Tc coordination sites for the or VIII.1). In accordance with these observations, the data obtained by Gorski and Koch 2+ [69GOR/KOC] [70GOR/KOC] for the diprotonic acid TcO strongly indicate the , presence of polymeric Tc species. Finally, reaching aqueous equilibrium in a system containing Tc(IV) needs normally days [83AND/GAS] but no hint about reaction times is given in [69GOR/KOC] , [70GOR/KOC] . , [69GOR/KOC] For these reasons all the equilibrium data reported in [70GOR/KOC] need confirmation, and no equilibrium constants have been selected in this review. 3+ From [Tc(thiourea) ] edta, a brown solid Tc(III)- [96REY/TER] and Na H 2 6 2 edta was obtained [98GON/KRE] . By a combination of different techniques (UV-Vis, 1 99 IR and Tc elemental analyses and cerimetric titrations, it was H-NMR spectroscopy, concluded that the 1:1 complex is formed (Na[Tc(edta)]·2H O). The complex is much 2 3+ more stable than [Tc(thiourea) ] . Solutions of Tc(III)-edta at pH values between 2 and 6 6 were not altered after storing for 5 days.

514 VIII Discussion of data selection for edta ligand 472 VIII.9 Zirconium edta compounds and complexes Zirconium edta compounds VIII.9.1 [57BOB/RAF] reporting zirconium edta compounds proposes a variety The first paper of “insoluble compounds” with zirconium to edta ratios 3:2, 2:1, 3:1, 4:1 and 6:1. These compounds have been inferred from titrations of zirconium nitrate with edta, but no quantitative data are given and the proposed compounds have not been confirmed by any other study. Zirconyl chloride reacts w ith edta or its disodium salt in an acidic medium to yield the [Zr(edta)]·4H O chelate [98MAR/SHL] . 2 Two papers reporting structures of Zr edta compounds were found: Zr(edta)(H O) [Zr(edta)CO [74POZ/POR] and (CN ] (see Figure VIII-3-a in Sec- H ) 6 2 2 3 2 3 [95MIS/SER] tion VIII.1) . These structure data reveal a coordination number of 8 for Zr 4+ complex salts, and we infer the same coordination number for Zr in aqueous solution. 4+ This means that two edta mo lecules can be bound by Zr , forming not only 1:1 but also 1:2 Zr edta complexes, as proposed by [77KOS/SHE] . No thermodynamic data are available for these Zr edta compounds. Stability of zirconium edta complexes VIII.9.2 4+ − 6 with dimerisa- Zr forms a very stable 1:1 edta complex, which hydrolyses at pH 5 [64INT/MAR] tion , . The equilibria in alkaline solution are complicated [67BOT/AND] by the precipitation of hydrolytic products. Under such conditions, also the formation of 4 − Zr(edta ) . was postulated [77KOS/SHE] 2 reports the formation of mixed Zr − edta − F [96YUC/HOK] A more recent study complexes. The maximum coordination number of 8 for Zr, as inferred from crystal 2 − Zr(edta )F structure analyses (see above) , is reached with the complex . 2 In Table VIII-25 experimental equilibri um data found in the literature are pre- i.e. sented. Data obtained from measurements under strong acidic conditions, measure- ments in 1 and 2 M HClO , are generally more important for deriving Zr edta stability 4 ch conditions the effects of zirconium hy- constants than other data, because under su difficult corrections have to be applied to drolysis are minimised. However, in all cases the experimental data in order to derive Zr edta stability constants. As discussed in detail in Appendix A, several papers [56MOR/JUS] , [63KYR/CAL] , [64INT/MAR] , [64PAN/VLA] , [66ERM/MAR] , [66LAP/PAN] , [67BUD/HAA] , , [68KOZ] , [75TER/SHE] [67TIK2] [77KOS/SHE] , [87JOA/BIG] are , not accepted in this review for evaluation of Zr edta stability constants. Recently, Vasil ′ ev et al . [99VAS/KAT] determined spectrophot ometrically K 1 K = by ligand competition. Their result, log at room temperature in 1 M HClO 10 1 4

515 VIII.9 Zirconium edta compounds and complexes 473 0.10), belongs to the more convincing ones because of the very low total metal (29.93 ± − 5 M) used in their experiments in order to avoid polymerization 10 × ion concentration (3 of zirconium. Considering the edta protonation constants they used in their data analy- has to be increased. Furthermore, took Zr [99VAS/KAT] K sis, the uncertainty of 1 , which are considered too high to et al. hydrolysis data from Nazarenko [78NAZ/ANT] . , with the evaluation of [76BAE/MES] e.g. be consistent Table VIII-25: Experimental equilibrium data for the Zr edta system. The uncertainties are given as reported in the references. + t Method Ionic medium [H ° C) log ] K Reference ( 10 4+ 4 − U + edta Zr(edta)(aq) Zr sp 0.1 M NaClO 0.01 M 25 19.4 [56MOR/JUS] 4 0.1 M 0.1 M 20 (29.5 ± 0.5) [63KYR/CAL] sorption (28.5 ± 0.3) 1 M 1 M HNO 3 5 M HNO 5 M (30.6 ± 0.2) 3 [64CAL/KYR] 1.0 M ? 29.70 sorption 3.0 M (H,Na)NO 3 1.2 M 30.10 1.4 M 29.65 1.6 M 30.42 1.8 M 30.53 2.0 M 30.61 2.2 M 30.73 2.4 M 30.86 2.6 M 30.86 2.8 M 30.77 1.0 M 28.46 1.0 M HNO 3 2.0 M HNO 30.58 2.0 M 3 3.0 M 31.11 3.0 M HNO 3 4.0 M 30.92 4.0 M HNO 3 5.0 M HNO 5.0 M 30.63 3 [64INT/MAR] 0.01 M 25 (29.0 ± 0.9) sp 0.1 M (KCl) ? 29.0 [66ERM/MAR] 0.23 M ix 0.23 M HClO 4 1 M HClO 1 M ? 28.0 4 1 M 20 27.7 [67BOT/AND] 1 M (H,Na)ClO red 4 ± 0.2) [67BUD/HAA] sp 25 (31.9 0.8 N HCl ? cix 1.2 M 20 (28.96 1.2 M HCl ± 0.04) [67TIK2] [69SOC/VOL] 28.4 1.78 M ? 3.6 M HNO pol 3 20 27.91 [70PRA/HAV] 2.0 M 2 M HClO ix 4 sp 1 M HClO 1 M r. t. (29.93 ± 0.10) [99VAS/KAT] 4 (Continued on next page)

516 VIII Discussion of data selection for edta ligand 474 Table VIII-25: (continued) + Reference C) log ( K ° t ] Method Ionic medium [H 10 + 2 4 − − U Zr(edta)(aq) + 2 OH + edta Zr(OH) 2 NO dis 1 M NH pH 9.9 ? (21.86 ± 0.05) [87JOA/BIG] 4 3 ± NO pH 9.9 (21.50 2.5 M NH 0.02) 3 4 a − 4 − 4 ZrL Zr(edta) + edta U II 2 pH > 6 ? (1.8 1 M 0.2) [77KOS/SHE] nmr ± 4 − − 4 − 4 Zr(edta) Zr + 2 edta U 2 0.01 M HClO 0.01 M ? (7.9 ± sp [66LAP/PAN] 0.1) 4 − + O U Zr(edta)(aq) + H + H Zr(edta)OH 2 pot 0.1 M (KCl) 0.01 M 25 − 6.2 [64INT/MAR] 1 M (H,Na)ClO 1 M 20 − 6.1 [67BOT/AND] red 4 2 − − 2 Zr(edta)OH (Zr(edta)OH) U 2 0.01 M 25 3.5 [64INT/MAR] pot 0.1 M (KCl) – – Zr(edta)F U Zr(edta)(aq) + F [96YUC/HOK] 25 4.62 ise-F 0.1 M KNO 3 − 2 – – U Zr(edta )F + F Zr(edta)F 2 0.1 M KNO 25 2.8 [96YUC/HOK] ise-F 3 a: See discussion of in Appendix A. [77KOS/SHE] r. t.: room temperature However, unlike the other metal ions considered in this review, no NEA se- lected values for Zr hydrolysis were availabl e at the time of the preparation of this chap- ter. Thus, a final consistent evaluation of Zr edta complexation has to be postponed until selected Zr hydrolysis data are available. For the time being our review only results in a list of promising papers for a [64CAL/KYR] later re-evaluation: , [67BOT/AND] , [69SOC/VOL] , [70PRA/HAV] , [96YUC/HOK] , [99VAS/KAT] . However, all data reported in these studies need diffi- cult corrections, and there is no guarantee that a re-evaluation using consistent auxiliary data will lead to selected values for Zr edta complexes. A final remark concerning Zr edta comp lexation might be in place: an investi- gation of Zr edta complexation avoiding hydrolytic equilibria was never realised, in contrast to several other metal ions. For inst ance, in the case of Pd(II), also exhibiting an intricate, difficult and not well known hydrolysis behavior, the complexation with − − − − Cl , I , Br and SCN is well known. This allowed not only the determination of the involved equilibrium constants, but subsequently also the derivation of formation con- stants with other polydentate ligands, for instance by use of potentiometric and spectro- photometric methods.

517 VIII.9 Zirconium edta compounds and complexes 475 Enthalpy of zirconium edta complex formation VIII.9.3 ium with edta have been determined in a The thermal effects of the reaction of zircon C in 2, 3, 3.7 and 4 M HClO othermal jacket at 25 calorimeter with an is ° 4 [78VAS/LYM] , and at 15 and 35 [78VAS/LYM2] C in 2, 3 and 3.7 M HClO . The ° 4 2+ solutions only contained H authors assumed that their edta . However, as discussed in 6 + Section VIII.3.8, the amount of H was not negligible under the experimental con- edta 5 [78VAS/LYM] ditions of (from 30% in 2 M acid, to ≈ 15% in 4 M , ≈ [78VAS/LYM2] acid). Because of this, the uncertainties in the experimental values should be increased − 1 to ± 2 kJ·mol in any further evaluation. However, a final consistent evaluation of the enthalpy of Zr edta complexation has to be postponed until selected Zr hydrolysis data are available. Table VIII-26: Experimental enthalpy data for the Zr edta system. The uncertainties are given as reported in the references. –1 t C) Method Ionic medium H ∆ (kJ mol ( ) Reference ° rm 4+ 2+ + + H Zr Zr(edta)(aq) + 6 H edta U 6 2 M HClO cal 25 (24.5 ± 0.6) [78VAS/LYM] 4 3 M HClO (23.9 ± 0.8) 4 (24.0 ± 0.8) 3 M HClO 4 0.7) ± (23.7 4 M HClO 4 2 M HClO (24.5 0.1) ± 4 (25.0 ± 0.3) 3 M HClO 4 0.2) (23.9 ± 3.7 M HClO 4 2 M HClO cal 15 (27.5 ± 0.3) [78VAS/LYM2] 4 (28.2 ± 0.3) 3 M HClO 4 0.3) ± (28.2 3.7 M HClO 4 2 M HClO 35 (23.0 ± 0.2) 4 0.2) (22.6 ± 3 M HClO 4 (22.5 ± 0.2) 3.7 M HClO 4 VIII.10 Uranium edta compounds and complexes Uranium edta compounds VIII.10.1 VIII.10.1.1 U(IV) edta compounds Some solid complex salts containing U(IV) have been isolated. The stoichiometry of the 14 compounds listed in Table VIII-27 has been confirmed by elemental analysis. [63ERM/KRO] report that the compound Uedta ⋅ n H can vary between 2.5 O, where n 2 − 3 and 3.5, “is stable in air for a long time and its solubility in water at 25 ° C is 6.45·10

518 VIII Discussion of data selection for edta ligand 476 M”. The same authors state that the K and NH compounds “have a low solubility in 4 s “are almost insoluble in water”, but no water”, whereas the Ca, Sr and Ba compound [63ERM/KRO] quantitative data are given by . Table VIII-27: U(IV) edta co mpounds. A reference reporting solubility data is marked + with (sol.). C(NH ) = guanidinium cation. 23 Compound Reference ⋅ 2H O [43BRI/THI] Uedta 2 2.5 − 3.5H [63ERM/KRO] O Uedta (sol.) ⋅ 2 ⋅ n H O [63ERM/KRO] edta(OH) U 2 4 2 9H H U (edta) [63ERM/KRO] ⋅ O K 2 2 3 2 2 K H U [63ERM/KRO] (edta) O ⋅ 18H 2 3 2 2 2 ⋅ ) H U [63ERM/KRO] (edta) O 8H (NH 2 2 3 4 2 2 ) [63ERM/KRO] H O U 16H (edta) ⋅ (NH 3 2 2 4 2 2 Ca U (edta) ⋅ 12H O [63ERM/KRO] 3 2 2 2 U (edta) ⋅ 18H O [63ERM/KRO] Sr 3 2 2 2 U [63ERM/KRO] (edta) O ⋅ 18H Ba 2 3 2 2 (C(NH ) ) [UedtaCl ] ⋅ 4H O [83SHC/MIK] 2 2 2 3 2 ) ) [UedtaF [83SHC/MIK] ] (C(NH 3 3 2 3 [83SHC/MIK] ) ] ) [Uedta(CO ) (C(NH 2 3 2 3 4 (C(NH 3H ) [Uedta(ox) ] ⋅ ) O [83SHC/MIK] 2 4 3 2 2 VIII.10.1.2 U(VI) edta compounds A number of solid complex salts containing U(VI) has been isolated and described in the literature. The stoichiometry of the 14 compounds listed in Table VIII-28 has been confirmed by elemental analysis. X ray crystal structure data are reported for only one − compound, (C(NH . The solubility of ) ) [85SHC/ORL] [(UO edta], in F ) 2 2 2 4 3 2 UO H edta) ⋅ (H O has been investigated by [59KLY/SMI3] . At a pH below 3 the solid 2 2 2 phase contains H ⋅ edta(cr) and UO (H O, and at a pH above 5 the solubility edta) H 2 2 2 4 increases sharply because of complex formation. In the pH range 3.0 – 4.5 4 − [59KLY/SMI3] ⋅ measured total uranium concentrations of about 2 M and total edta 10 concentrations of about 0.02 M and calculated a solubility product. However, the results of [59KLY/SMI3] are not accepted in this review ( cf. Appendix A).

519 VIII.10 Uranium edta compounds and complexes 477 mpounds. A reference reporting solubility data is marked Table VIII-28: U(VI) edta co ray single crystal structure is marked with with (sol.), and a reference reporting an X − + C(NH ) (str.), see also Figure VI II-4b in Section VIII.1. = guanidinium cation and ac 23 = acetate anion. Compound Reference , [64BHA/KRI] (sol.), [59KLY/SMI3] , [42BRI/HES] UO H (H ⋅ edta) O 2 2 2 [85SHC/ORL] ) edta (UO ⋅ [85SHC/ORL] O H 2 2 2 (UO edta ) O) (H [85SHC/ORL] 2 2 2 2 ) [85SHC/ORL] ⋅ 4H , O [64BHA/KRI] , [83SHC/MIK] edta (UO 2 2 2 [85SHC/ORL] ) , ) [83SHC/MIK] [(UO (str.) F edta] ) (C(NH 2 4 3 2 2 2 (C(NH (H ) ) [(UO ) [85SHC/ORL] (NCS) , [83SHC/MIK] edta] O) 2 2 3 2 2 2 2 2 [85SHC/ORL] ) ) ) (NCS) [(UO edta] (C(NH 4 4 3 2 2 2 [83SHC/MIK] [(CH ) SO] edta , [85SHC/ORL] ) (UO 2 2 2 4 3 (UO [(CH ) ) SO] edta [85SHC/ORL] 2 2 2 3 2 [85SHC/ORL] ) , ) [83SHC/MIK] [(UO O ) 3H (ac) ⋅ edta] (C(NH 2 2 2 3 2 2 2 3H [(UO ) (ac) edta] ⋅ ) O [83SHC/MIK] , [85SHC/ORL] ) (C(NH 2 3 2 3 3 2 2 (C(NH edta] ) ) (CO ) ) [(UO ⋅ 2H O [83SHC/MIK] , [85SHC/ORL] 4 2 2 3 2 2 2 3 (CO ) ) [85SHC/ORL] [(UO , ) [83SHC/MIK] edta] ) (C(NH 3 3 2 2 2 6 3 [85SHC/ORL] ) , ) [83SHC/MIK] [(UO O ) 2H (ox) ⋅ edta] (C(NH 3 2 6 3 2 2 2 VIII.10.2 Uranium edta complexes VIII.10.2.1 U(V) edta complexes No thermodynamic data are available for U(V) edta complexes. U(III) edta complexes VIII.10.2.2 Uranium(III) is not stable in aqueous solution because it is rapidly oxidised (see Section V.4.2.1.2.d, p.200, in [92GRE/FUG] ). The stability constant for U(III) edta complexa- ο K log tion, = 20.3, found in Table 1 of [69MOS] , has been “determined by extrapola- 10 1 3+ tion, also using the literature data on the stability of acido-complexes of Am and 3+ Cm ”. This estimated value for an unstable species, which is not based on an actual experimental study, is not credited in this review. Table VIII-29: Equilibrium data for the U(III) edta system. t ( ° C) log Method Ionic medium K Reference 10 3+ 4 − − U U Uedta + edta rev I → 0 ? 20.3 [69MOS]

520 VIII Discussion of data selection for edta ligand 478 VIII.10.2.3 U(IV) edta complexes Although already in one of the first papers on aqueous uranium(IV) edta chemistry [61PAL/HSU] the formation of 1:1 and 1:2 U(IV) − edta chelates has been postulated, based on spectrophotometric investigations, we have scarce quantitative information on the composition and on the stability of the formed complexes from only five papers [59KLY/SMI4] [83PER/MIS] , [68CAR/MAR] , [62KRO/ERM] . The , [63ERM/KRO] , papers are summarised in Table VIII-30. equilibrium constants reported in these data for the U(IV) edta system. The Table VIII-30: Experimental equilibrium uncertainties are given as reported in the references. Method Ionic medium ( ° K Reference C) pH range log t 10 4+ 4 − U Uedta(aq) U + edta ∼ 0.025 M ? 1.7 (25.6 ± 0.4) [59KLY/SMI4] titr sp 20.0 1.4 − 1.5 (25.8 ± 0.2) [68CAR/MAR] 0.10 M KCl 4+ + + + H Uedta(aq) + 5 H edta U U 5 1.0 M (H,Na)ClO 25 − 0.2 − 0.4 (2.28 sp 0.04) [62KRO/ERM] ± 4 (1.98 1.5 M (H,Na)ClO − 0.2 − ± 0.03) 0.3 4 2.0 M (H,Na)ClO − 0.2 − 0.3 (1.90 ± 0.03) 4 (1.65 − 0.2 − 0.3 ± 0.04) 2.5 M (H,Na)ClO 4 0.0 − 0.3 (1.61 ± 0.03) 3.0 M (H,Na)ClO 4 4 − 4 − 2 Uedta(aq) + edta U U(edta) 23 pot 25 3.5 − 5 11.9 [63ERM/KRO] 0.1 M KCl ? 4.5 − 6 (9.25 ± 0.03) [83PER/MIS] sp 0.2 M − 2 2 − U edta UH(edta) 2 Uedta(aq) + H 2 3 22 0.04 M KCl 25 1 − 5 (3.64 ± 0.06) [63ERM/KRO] sp 0.09) 0.1 M KCl (3.57 ± ± 0.11) (2.85 0.5 M KCl ± 0.16) (3.30 2.0 M KCl (Continued on next page)

521 VIII.10 Uranium edta compounds and complexes 479 Table VIII-30: (continued) Method Ionic medium C) pH range log t K Reference ( ° 10 3 − − 2 + UH(edta) 2 Uedta(aq) + H edta + H U 2 23 25 1 − 5 (0.46 ± 0.07) [63ERM/KRO] sp 0.04 M KCl (0.79 ± 0.03) 0.1 M KCl 0.06) ± (0.82 0.5 M KCl 0.04) ± (1.07 2.0 M KCl + + Uedta(aq) + H U U(Hedta) sp 1.0 3.0 M 25 ≤ 1.5 [62KRO/ERM] − − − UedtaOH Uedta(aq) + OH U 25 0.01 M KCl − 6 9.00 [63ERM/KRO] pot 4 0.1 M KCl 9.07 0.25 M KCl 9.08 0.5 M KCl 9.17 1.0 M KCl 9.13 − + Uedta(aq) + H O U UedtaOH + H 2 pot 0.10 M KCl 25.3 3 − 5 − ( 4.72 ± 0.01) [68CAR/MAR] 1.0 M KCl − ( 4.58 ± 0.01) 2 − − U (UedtaOH) 2 UedtaOH 2 0.01 M KCl 25 4 − 6 2.84 [63ERM/KRO] pot 0.1 M KCl 2.75 0.25 M KCl 2.79 0.5 M KCl 2.48 1.0 M KCl 2.86 [68CAR/MAR] 0.05) ± − 5 (2.91 3 25.3 pot 0.10 M KCl ± 0.05) 1.0 M KCl (2.48 − 2 − − UedtaOH U + OH Uedta(OH) 2 pot 0.01 M KCl 25 7 − 9 5.91 [63ERM/KRO] 0.1 M KCl 6.29 6.41 0.25 M KCl 0.5 M KCl 6.49 6.87 1.0 M KCl

522 VIII Discussion of data selection for edta ligand 480 In Table VIII-31, information is summarised about the formation of ternary edta complexes, showing the tendency of Uedta(aq) to bind further complexing agents. Table VIII-31: Experimental equilibrium data for the U(IV) edta X system where the ligand X forms a ternary complex UedtaX. The uncertainties are given as reported in the references. ° C) log X Method Ionic medium K t Reference ( 10 − 4 − 3+ U UedtaX + edta UX − pol 0.1 M NaClO F 20 17.50 [59SMI] 4 − n − n Uedta(aq) + X U UedtaX H 0.03) X sp ? ? (3.00 ± Oxalic acid [83PER/MIS] 2 Iminodiacetic acid X pot 0.1 M KCl 25.3 (8.2 ± 0.1) [67CAR/MAR] H 2 0.1) X (4.2 ± H Phtalic acid 2 a 8 − Hydroxyquinoline − H X (9.72 ± 0.04) 2 5 − sulfonic acid − − Dihydroxybenzene 1,2 0.05) ± (15.61 X H 4 disulfonic acid − 3,5 5 − Sulfosalicylic acid X (11.08 ± H 0.05) 2 1,8 − Dihydroxynaphthalen − H ± X 0.01) (16.22 4 − 3,6 disulfonic acid Catechol H X (14.16 ± 0.5) 2 [80PER/POL] Benzoylacetone HX sp ? ? 3.22 Thenoyltrifluoroacet 2.80 one HX 2.50 Dibenzoylmethane HX 2 − 3 − + K + H − ( a: In addition, log = O U UedtaXOH 7.14 ± + H 0.01) is reported for UedtaX . 10 2 , the solution of the 1:1 com- [68CAR/MAR] and [63ERM/KRO] As found by in acidic medium at pH > 3, and polynuclear hydroxo com- plex is hydrolysed already plexes are formed. At pH > 7 a pH drifting without precipitation is observed in titration experiments [68CAR/MAR] . This phenomenon is an indication of even further hydroly- sis and polymerization of the U(IV) chelate. Probably ve ry complex polynuclear che- lates containing hydroxo bridges are pr oduced. When the pH exceeds 11, U(OH) (s) is 4 precipitated [63ERM/KRO] . Apparently the six edta donor atoms are not sufficient for replacing all water molecules coordinated with the U(IV) ion, and more complex polydentate ligands would be needed to avoid these complications. The octadentate dtpa (diethylenetrinitrilopenta-acetate) is better in this respect, but only the decadentate ttha (trietylenetetranitrilohexa-acetate) forms a 1:1 complex without hydrolytic products [68CAR/MAR] . In this context, probably not only the number of the donor atoms of the

523 VIII.10 Uranium edta compounds and complexes 481 steric strains of the formed complexes are ligand determines its behavior, but also the important. In the first quantitative study of the U(IV) edta system [59KLY/SMI4] differ- ent quantities of U(IV) sulfate were titrated with edta solution and arsenazo indicator until the color of the solutions changed from blue to pink. However, no background c strength during the titration experiments. electrolyte was used, causing a varying ioni cf. For this and other reasons ( Appendix A) the results of [59KLY/SMI4] are not con- sidered in this review. [62KRO/ERM] The spectrophotometric study of the U(IV) edta system by in acidic (H,Na)ClO media is considered as reliable by this review. The results have been 4 interpreted in terms of the reaction: 4+ + + U U Uedta(aq) + 5 H edta + H (VIII.23) 5 + 2+ edta and H edta determined by the authors involving equilibrium constants for H 5 6 3+ [62KRO/ERM] [50KRA/NEL] . The study of U(IV) hydroly- taken from and for UOH 3+ [50KRA/NEL] sis by is one of the best source of data for UOH and has been accepted, among other studies, for the NEA selection in [92GRE/FUG] . The edta protonation constants reported by [62KRO/ERM] have been re-analysed in this review ( Appen- cf. dix A) and differences up to 0.14 log units have been found. Considering these differ- 10 0.2 has been assigned to the experimental data of ± ences, an overall uncertainty of [62KRO/ERM] shown in Table VIII-30. A weighted SIT least squares regression analy- ο log K sis gives ± 0.19) and ∆ε (VIII.23) = − ( 0.02 ± 0.09). (VIII.23) = (4.80 10 ο log K ± (VIII.23) = (4.80 0.19) with the edta protonation con- Combining 10 ο log β = (24.72 ± 0.12) evaluated in this review (see Section VIII.3.7) results in stant 10 5 ο log K (VIII.24) = (29.52 ± 0.22) for the equilibrium: 1 10 4+ − 4 + edta U Uedta(aq). (VIII.24) U The SIT interaction coefficient ∆ε (VIII.23) = − ( 0.02 ± 0.09) allows an esti- − − 4+ + ClO ClO ε mate of (Uedta(aq), NaClO , (H ) = ) + ε (U ) + , ∆ε (VIII.23) – 5 ε 4 4 4 − 4+ − + + − ε (H ) = (0.76 ε (H , , ClO ClO edta ) = (0.14 ± 0.02) and ε (U 0.06) , ± ClO ) using 5 4 4 4 + − [92GRE/FUG] from ClO edta and , ε (H ± ) = − ( 0.23 0.15) evaluated in this review 5 4 (see Section VIII.3 .7). The result, (Uedta(aq), NaClO ε ) = 0.19 ± 0.19), is consistent − ( 4 with ε (H edta(aq), NaClO ± 0.14) (see Section VIII.3.7). However, the de- − ( 0.29 ) = 4 4 viation of ε (Uedta(aq), NaClO ) from zero, as generally assumed for neutral species in 4 this review, is at the verge of statistical significance. Considering the more recently − 4+ discussed value of (U ε ClO , ) = (0.84 ± 0.06), see footnote I on p.818 in 4 [2001LEM/FUG] , results in ε (Uedta(aq), NaClO 0.19). Hence, a statisti- ) = – (0.11 ± 4 cally significant deviation of ε (Uedta(aq), NaClO ) from zero cannot be established 4 from the data of [62KRO/ERM] .

524 VIII Discussion of data selection for edta ligand 482 [68CAR/MAR] Another reliable study of the U(IV) edta system is reported by . The equilibrium constant K (VIII.24) is obtained from log a spectrophotometric study 10 1 of a cation exchange equilibrium in the system U(IV) – edta –