<<

AECL

SOLUBILITY BEHAVIOUR OF (III) AND ANTIMONY(V) IN BASIC AQUEOUS SOLUTIONS TO 300°C

bY

Robert J. Lemire’, Nancy B. Tosello’ and James D. HaWay

‘Reactor Chemistry Branch 2Analytical Chemistry Branch Chalk River Laboratories Chalk River, Ontario, KOJ 1JO

1999 December

AFCL-12064 EACL

COMPORTEMENT DE LA SOLUBILII% DES MAT&ES SOLIDES D’ANTIMOINE(III) ET D’ANTIMOINE(V) DANS LES SOLUTIONS AQUEUSES BASIQUES JUSQU’A 3oO°C

Par Robert J. Lemire’, Nancy B. Tosello’ et James D. Halliday’

Resume

Le role et l’importance des isotors 122Sb et ‘%Sb dans le transport d’activite a l’int&ieur du circuit primaire d’un reacteur CANDU D ont Cte associes a l’entree d’oxygene lors de l’arr& du reacteur. Dans le cadre d’un programme visant a reduire au minimum la liberation et la redeposition de ces , on a mesure la solubilite des sels et des oxydes d’antimoine(IIl) et (V) dans des solutions basiques a des temperatures allant de 25 B 3OOOC. Les result&s fournissent des renseignements sur la charge et la stabi1it.e en fonction de la temperature des esp&ces d’antimoine en solution et servent de guide dans la d&ermination des variations de la solubilite des mat&es solides d’antomoine en fonction de la temperature.

Dans les solutions dans lesquelles l’oxydation de l’antimoine (Ill) en antimoine (V) est reduite au minimum, la solubilid du Sb203 augmente d’environ deux ordres de grandeur entre 25 et 200°C, puis se stabilise ou decroit legerement. A 250°C, dans les solutions oxydantes, on a trouve que le SbzOs.xHzO et l’antimoniate de simple Btaient instables dans les solutions d’hydroxyde de sodium en ce qui concerne la mat&e solide, Na&H(H20)]2_&b206, qui presente une structure de pyrochlore. La solubilite de cet antimoniate de sodium partiellement protone croit de 25 B 2OOOC et decroit aux temperatures sup&ieures a 250°C. Ces variations de solubilite des mat&es solides d’antimoine(V) correspondent aux variations de la stabilite des esp&ces anioniques de la solution d’antimoine (SbOj ou Sb(OH);), m&me si la composition des mat&es solides contenant de l’antimoine dans les solutions oxydantes basiques depend fortement des cations et de leur concentration en aqueuse.

On pourrait s’attendre a ce que tous les solides utilises dans ces experiences produisent des concentrations totales d’antimoine en solution 2 0,00005 moldm” dans n’importe quelle solution aqueuse neutre ou basique (en supposant qu’aucun se1 de sodium ne soit ajot@. Par consequent, dans les conditions du circuit primaire, la precipitation d’oxydes d’antimoine ou d’oxydes mixtes est peu probable. On ne peut pas Bcarter l’hypothese que le Sbz05 hydrate (en particulier la forme de pyrochlore) puisse Ctre moins soluble dans des solutions presque neutres, de faible force ionique.

’ Chimie des reacteurs 2 Chimie analytique Laboratoires de Chalk River Chalk River (Ontario) KOJ 1JO Dt?cembre 1999

ABCL-12064 ABCL

SOLUBILITY BEHAVIOUR OF ANTIMONY(II1) AND ANTIMONY(V) SOLIDS IN BASIC AQUEOUS SOLUTIONS TO 300°C

Robert J. Len-rim’, Nancy B. Tosello’ and James D. Halliday2

Abstract

The major contributions of the isotopes 122Sb and ‘“Sb to activity transport in a CANDU@ reactor primary heat transport system (HTS), have been associated with ingress during reactor shutdown. As part of a program to minimize the release and redeposition of these isotopes, the of antimony(lll) and (V) and salts have been measured in basic solutions at temperatures from 25 to 300°C. The results provide information on the charge and the stability as a function of temperature of antimony solution species and, hence, a guide to the trends in the temperature dependence of the solubilities of antimony solids. ln solutions in which oxidation of antimony(lll) to antimony(V) is minimized, the solubility of Sb203 increases by about two orders of magnitude between 25 and 2OO”C, and then levels out or decreases slightly. At 25O”C, in oxidizing solutions, Sb2Os.xHzO and simple sodium antimonate(V) were found to be unstable in solutions with respect to the , Na2,[H(H20)]2_2,Sb206, which has a pyrochlore structure. The solubility of this partially protonated sodium antimonate increases from 25 to 200°C and decreases at temperatures above 250°C. These solubility changes for the antimony (V) solids reflect changes in the stability of the anionic antimony solution species (SbOj or Sb(OH)$, even though the compositions of antimony-containing solids in basic oxidizing solutions are strongly dependent on the cations and their aqueous phase concentrations.

All solids used in the present experiments would be expected to generate total solution antimony concentrations 2 0.00005 moldrn-3 in any neutral or basic aqueous solutions (assuming no added sodium salts). Therefore, under HTS conditions, precipitation of any antimony oxides or mixed oxides is unlikely. It cannot be ruled out that hydrated Sb205 (especially the pyrochlore form) might be less soluble in near-neutral, low-ionic-strength solutions.

‘Reactor Chemistry Branch 2Analytical Chemistry Branch Chalk River Laboratories Chalk River, Ontario, KOJ 1JO

1999 December

ABCL- 12064 i

CONTENTS

1. INTRODUCTION ...... 1

2. THB CI-IFMISTRY OF ANTIMONY(III) AND ANTIMONY(V) ...... 2 2.1 The Aqueous Species ...... 2 2.2 The Solids ...... 2 2.3 Previous Solubility Measurements ...... 3 2.3.1 Previous Solubility Measurements for Sb203 ...... 3 2.3.2 Previous Solubility Measurements for Sb205 ...... 5 2.3.3 Previous Solubility Measurements for Sodium Antimonate(V) ...... 6

3. EXPERIMENTAL SOLUBILITY MEASUREMENTS ...... 7 3.1 General Procedures for Measurements for T 2 200°C ...... 7 3.2 Antimony(III) ...... 7 3.3 Antimony(V) ...... 7 3.3.1 Preparation and Characterization of the Solid Phases ...... 7 3.3.2 Solubility Experiments for Temperatures Below 100°C...... 12 3.3.2.1 Preliminary Results ...... 12 3.3.2.2 The Solubilities of Solids B and C at 25 and 75°C ...... 12 3.3.2.3 Other Experiments ...... 13 3.3.3 Solubilities for Temperatures from 200 to 300°C ...... 15 3.3.3.1 Preliminary Results ...... 15 3.3.3.2 The Solubility of Solid B as a Function of Temperature and Hydroxide Concentration...... 17

4. DISCUSSION ...... 17 4.1 Antimony(IlI) ...... 19 4.2 Antimony(V) ...... 23 4.2.1 Rationale for the Measurements Using “NaSb(OH)h(s)” and Other Sodium Antimonates ...... 23 4.2.2 Solubility of NaSb03.3H20(s) (NaSb(OH)h) in Basic Solutions ...... 24 4.2.3 Solubility of Na2,[H(H20)]2_2$b206.H20, a = 0.75 in Basic Solutions 27 4.2.3.1 Comparison of the Solubility with Other Solids at 25 and 75°C ...... 27 4.2.3.2 Solubility of Na2,[H(H20)]2_2aSb206.H20 from 25 to 300°C ...... 28

5. CONCLUSIONS ...... 30

6. ACKNOWLEDGMENTS ...... 31

7. RBFBRBNCES ...... 31 ii

Appendix A: Literature Thermodynamic Data for Aqueous Antimony Species and Selected Solids...... A.1 Simple Aqueous and Hydrolysis Species of Antimony...... 37 A.2 Antimony(III) and Antimony(V) Oxide Solids ...... 37 A.3 Chemical Thermodynamic Measurements for Mixed Oxides Containing Antimony ...... 40

LIST OF TABLES

Table 2- 1: Reported Solubilities of Sb203 ...... 4 Table 2-2: Reported Solubilities of SbzOS ...... 6 Table 2-3: Reported Solubilities of Sodium Antimonate in Water ...... 6 Table 3-l: Experimental Solubility Measurements for SbzOs/Sb Mixtures ...... 8 Table 3-2: Results of neutron activation analyses of solid B ...... 11 Table 3-3: per Mole Sb of Various Antimony(V) Compounds Containing Oxygen, Sodium or Hydroxide Ions or Water ...... 11 Table 3-4: Experimental Solubilities of NaSb(OH)e (initially Solid A) from the Present Study ...... 13 Table 3-5: Total Antimony Concentrations for Solids B and C as Measured for Basic Oxidizing Solutions at 25 and 75°C ...... 14 Table 3-6: Results of Equilibration of Mixed Antimony Solids with Water at 75°C (unless otherwise noted) ...... 15 Table 3-7: Total Antimony Concentrations as Measured over Solid B (initially solid A, but converted to solid B during the experiment) for Basic Oxidizing Solutions at 250°C ...... 16 Table 3-8: Total Antimony Concentrations as Measured over solid B (initially solid A, but converted to solid B during the experiment) for Basic Oxidizing Solutions after Heating to 250°C and Cooling to Room Temperature...... 16 Table 3-9: Total Antimony Concentrations for Solid B as Measured for Basic Oxidizing Solutions at 200 to 300°C or after Cooling to Room Temperature. 18 Table 4- 1: Calculated Thermodynamic Quantities for the Dissolution of Sb203...... 22 Table 4-2: Values of the Solubility Product for Solids Nominally Hydrated NaSb03 or NaSb(OH)h ...... 26 Table 4-3: Activity Products for Na2u[H(H20)]2_2,Sb206.H20 (logioK,) with Values of a Calculated from the Experimental Results for Each Temperature and from Values of logloK,(25”C) and a Calculated from the Results at all Temperatures ...... 29 Table Al: Literature Tabulations of Chemical Thermodynamic Values for Antimony Aqueous Species ...... 37 Table A2: Literature Tabulations of Gibbs Energy of Formation Values for Antimony(III) and Antimony(V) Oxides at 25°C ...... 37 Table A3: Literature Tabulations of Enthalpy of Formation Values for Antimony(III) and Antimony(V) Oxides at 25°C ...... 38 Table A4: Literature Tabulations of Entropy Values for Antimony Oxides at 25°C ...... 38 . . . 111

Table A5: Literature Tabulations of Heat Capacity Values for Antimony Oxides at 25°C ...... 38 Table A6: Temperature-Dependent Heat-Capacity Values for Antimony Oxides ...... 39 Table A7: Temperature-Dependent Heat-Capacity Values for Alkali /Antimony(V) Mixed Oxides ...... 40

LIST OF FIGURES

Figure 3-l: X-Ray Diffraction Pattern for NaSb(OH)e - Solid A ...... 9 Figure 3-2: X-Ray Diffraction Pattern for (a) the Pyrochlore Structure Sodium Antimonate - Solid B; (b) Partially Dehydrated Sb205 [ 18]...... 10 Figure 3-3: X-Ray Diffraction Pattern for NaSb(OH)e - Solid C ...... 10 Figure 4- 1: Total concentrations of antimony(III) in aqueous solution in equilibrium with Sb203 from 15 to 50°C (for details, see text) ...... 19 Figure 4-2: Solubility of Sb203 at temperatures from 90 to 300°C ...... 22 Figure 4-3: Calculated total solution concentrations of Sb(III) as a function of temperature and hydroxide concentration ...... 23 Figure 4-4: Sodium ion concentration (M) as a function of total Sb(V) concentration for solubility measurements of sodium antimonates in basic solutions at 25 and 75°C ...... 25 Figure 4-5: Solubility measurements for solid B, a mixed oxide of antimony(V) (hydrated pyrochlore-structure sodium salt, Na~.&IO.~Sb&l&O) ...... 29 1. INTRODUCTION

Among the major contributors to activity transport in CANDU* primary heat transport systems (HTS) are ‘22Sb and ‘%b. These antimony isotopes are activation products from irradiation of the naturally occurring isotopes of antimony, 121Sb and 123Sb, an element found as a minor (but unmeasured and unspecified) component of some reactor materials [ 11. Antimony is also present in some pump seals and bearings at Gentilly-2. Oxygen excursions during shutdown at Gentilly-2 have resulted in large increases in out-of-core radiation fields, adversely affecting scheduled maintenance. This has led to the routine use of an oxidizing antimony removal process at the start of each annual maintenance shutdown [ 11. Although it is not certain whether the antimony is initially mobilized by a physical or chemical mechanism, it is clear that antimony can be readily released and transported in solutions under oxidizing conditions. The primary impetus for the work described in this report was to enhance our understanding of the high- temperature (50 to 300°C) solution properties of antimony, for both reducing and oxidizing conditions, and hence to help determine conditions that minimize antimony release, transport and deposition.

The nature of antimony species in aqueous solutions, and the solids, stable and metastable, that can exist in contact with such solutions, is very complex, especially for oxidizing conditions. There does not appear to be any single paper or document that describes antimony solution species and solids in a coherent, comprehensive manner. This is particularly true if the behaviour as a function of temperature is of interest. Several studies of the solubility of antimony(lII) oxide, Sb203, in neutral to basic solutions have been reported [2-71. These span a fairly wide range of temperature (15 to 200°C). Belinskaya and Militsina [8] thoroughly reviewed the qualitative features of the antimony(V) oxide solids related to the preparation of inorganic ion- exchangers, and that review provided many useful insights about the behaviour of antimony(V) in neutral and basic solutions. For both antimony solids and aqueous species (see Appendix A), values in several standard tables of chemical thermodynamic data differ substantially. Solubility studies are one means of obtaining information about changes in aqueous species as a function of pH and temperature.

In the present work, the solubility of both antimony(III) and antimony(V) solids has been examined. A very limited study was carried out using the antimony(III) oxide, to ensure that the solubility of Sb203 did not change greatly between 200 and 3OO”C, and to confirm the literature results at 200°C [7]. A more detailed study was made for antimony(V) solids. Different sodium antimonate(V) solids were prepared, and characterization attempted. Solubilities of these antimony(V) solids in basic solutions were measured at temperatures from 25 to 300°C. The results from these studies were used to draw conclusions about the relative stabilities of the antimony solids, the nature of the aqueous antimony species and, for oxidizing conditions, constraints on the probable total concentrations of antimony species in solution as a function of temperature.

* CANDU: Deuterium ; registered trademark. 2

2. THE CHEMISTRY OF ANTIMONY@) AND ANTIMONY(V)

2.1 The Aaueous Snecies

At equilibrium in aqueous solution, antimony generally forms antimony(III) species under reducing conditions, and antimony(V) species under oxidizing conditions. Antimony (III) chemistry is moderately well understood, at least near room temperature. This is not true for antimony(V). In some general references [9, lo], it is implied that HSb(OH)6 is a moderately strong acid, ionizing at pH values between 1 and 2 at room temperature. However, at macro- concentrations, Sb(V) forms polymeric species. As for many hydrolytic , once formed these are slow to depolymerize, even though simpler species may be thermodynamically stable in a particular solution. Therefore, even though there have been several studies of the species formed in weakly acidic and neutral solutions (e.g., [ 1 l-14]), it is not clear what species will form in very dilute solutions of Sb(V). Lefebvre and Maria [ 141 proposed a series of anionic dodecamers H12_,(SbO& . Recently, Nakano et al. [ 151 reported the isolation and of a salt containing the [SbsOi2(OH)& anion. Jander and Ostmann [13] reported finding evidence for formation only for acidic solutions with total antimony concentrations greater than 10” mol.dm-3. The same authors, using absorption spectroscopy, found that for basic Sb(V) solutions (pH > 12) there is an equilibrium between two monomolecular anionic species, with the most probable reaction being: Sb(OH); + OH- * SbO(OH):- + H20 (2.1) Although, the existence of SbO$ -(aq) in solution does not yet appear to be proven, Prasad [ 161, based on potentiometric and conductometric experiments, proposed that addition of acid to a solution of “K3Sb04” resulted in the sequential conversion:

SbOi (aq) + 0.5 Sb20’: (aq) + SbOj (aq) + 0.5Sb205(aq) (2.2) This does not appear to be compatible with the spectroscopic study of Jander and Ostmann [ 131.

2.2 The Solids

Antimony(III) oxide exists in two forms. The commercially available orthorhombic form of the oxide () is easier to prepare, and occurs more commonly in nature. However, the cubic form (senarmontite) is reportedly more stable near room temperature [3].

In 1970, Stewart and Knop [ 171 noted that samples of commercially available “anhydrous Sb205)’ were actually found to be either an amorphous, partially reduced solid (of approximate composition Sb204.4) or Sb204, and this was also reported by other groups [8]. We found similar problems with currently available “Sb205)‘, and consider any studies based on commercially available Sb205 (or even on material prepared in research laboratories) to be suspect unless proper characterization is provided for the solid. Antimonic acid (hydrated Sb205) cannot be dehydrated to Sb205 by heating in air at 1 bar [8]. Heating antimonic acid (i.e., hydrated Sb205) to between 650 and 850°C instead to a partially reduced solid, Sb204.35, that has a defect pyrochlore structure [ 181. Further heating to 935°C yields P-Sb204 [ 181. Rumpel et al. [ 191 have 3 reported the preparation of Sb204.4(OH)r.2, Sb204.s(OH)0.4 and two forms of anhydrous Sb205 by heating (at 300 to 750°C) various antimony oxides in the presence of small amounts of water with oxygen gas at pressures from 80 to 900 bar. The Sb204.s(OH)O.4 is closely related to (or possibly identical to) the compound Sb50i20H.H20, for which the crystal structure was reported by Jansen [20].

A wide variety of solids containing antimony(V) and alkali have been reported. The different salts with antimony to ratios of 1: 1 apparently either do not have the same structures, or, at least in some cases, are not simple hexahydroxy salts. Although crystals of NaSb(OH)G are tetragonal[21-231, crystals of LiSb(OH)h are reported to be hexagonal or trigonal [21]. The and sodium salts, as recovered from concentrated aqueous solutions, have the apparent stoichiometries MSbO3.xI-IzO (x = 3.5-3.6) [24], whereas the salt is recovered as KSb03.2.6H20 [25]. The latter formula is not compatible with a MSb(OH)e structure. Lisichkin et al. [26] suggested K[HSb03(0H)].H20, and other possibilities were discussed by Balicheva and Roi [27]. Similarly, Stewart and Knop [ 171 reported that analyses of commercially available “KSb(OH)e” corresponded to the compound being KSb03.2.3H20, and suggested that “NaSb(O&” is probably a mixture of NaSb(OH)h and an amorphous solid with a formula NaSbOs(3-x)H20. A full crystal determination has been done for NaSb(OH)e as prepared by treatment of a dilute solution of “KSb(OH)6” with NaCl(aq) [23]. The structure has Sb(OH)i octahedra with distortion to accommodate the Na+ ions. Beintema [22] found, in agreement with Knorre and Olschewsky [28], that the hexahydroxy-compound ((‘Na2H2Sb207 + 5 H20”) precipitated from hot solution, but that “Na2H2Sb207 + 6HzO” (possibly the same as the NaSb03.3.5HzO of Dravotsky and Karlicek [24]) was recovered from cooler solutions. The difference has been attributed [22] to inclusions of some of the mother liquor in the crystals prepared at lower temperatures. It is therefore unfortunate that Asai [23] gave no indication of any analysis of his material for hyperstoichiometric water.

Alkali metal antimonates having a pyrochlore structure have been prepared by wet and dry methods. Montmory et al. [29] prepared a pyrochlore-type solid “Sb2Na205(OH)2” by dehydration of “SbOsNa” trihydrate (NaSb(OH)h) at 180 to 320°C. Baetsle and Huys [30] and Abe [3 l] demonstrated ion-exchange properties for hydrated Sb205 in acidic media, while Bauer et al. [32] used different concentrations of alkali metal hydroxide solutions at room temperature to control the extent of substitution of the metal cations into the oxide pyrochlore structure. Gol’dshtein et al. [33] reported a mixed-oxidation-state pyrochlore-structure solid, NaSbmSbT07. The nature of the solids is discussed more thoroughly by Belinskaya and Militsina [S].

2.3 Previous Solubilitv Measurements 2.3.1 Previous Solubility Measurements for Sb203 Solubility values for Sb203 in water and basic aqueous solutions have been reported previously by a large number of authors [2-71 (Table 2-l; also c$ Figure 4-l). Except for the work of Popova et al. [7] and a single experiment by Schulze [2], all these measurements were done using solutions at temperatures between 15 and 50°C. In general, agreement between the results of the different studies is excellent-much better than most studies of oxide solubilities. 4

Table 2-l: Reported Solubilities of Sb203

T/“C Form of Reference SbzO& 15 6.71x10-’ 5.50x10” r? PI 15 0.3375” 1.3ox1o-3 r? WI 15 0.6749* 2.57~10” r? WI 15 1.0122* 4.12~10” r? 161 15 1.3493* 5.78~10” r? WI 15 1.6862” 7.63~10-~ r? WI 15 2.0232” 9.43x10” r? WI 15 2.3597” 0.0117 r? [61 25 1.oox1o-7 9.20~10-~ r [31 25 1.oox1o-7 4.oOx1o-5 C [31 25 1.oox1o-7 5.20~10-~ r c51 25 0.00505 9.80~10-~ r [51 25 0.0101 1.48~10~ r 151 25 0.0202 2.38~10-~ r 151 25 0.0404 4.28~10~ r [51 25 0.04 3.78~10~ r 151 25 0.0749 7.58~10~ r [51 25 0.0998 9.98~10~ r 151 25 0.3372” 1.63~10-~ r? [61 25 0.6742* 3.26~10” r? [61 25 1.0112* 5.06~10-~ r? WI 25 1.3482* 6.86x1o-3 r? 161 25 1.684” 9.78~10” r? WI 25 2.0209” 0.0117 r? U-d 25 2.3572* 0.0142 r? WI 35 1.45x1o-7 9.1ox1o-5 r? [41 35 1.58~10-~ 6.1~10~ r? [41 35 6.~10-~ 4.5ox1o-5 r? [41 35 3.49x10a 4.80~10‘~ r? [41 35 6.~10~ 6.8ox1o-5 r? r41 35 6.x10” 9.7ox1o-5 r? [41 35 0.0103 1.3OxlO~ r? [41 35 0.0424 2.56~10~ r? 141 35 0.0915 4.87~10~ r? 141 35 0.458 2.52~10-~ r? [41 35 0.702 4.32~10-~ r? [41 35 1.99 0.0129 r? [41 35 0.6732* 4.29~10-~ r? [61 ...... 35...... -1.0097* ...... 6.60x10”...... _ ...... r?...... _ ...... [61 Cont’d.. . 5

Table 2- 1 (Concluded) T/“C t[QH’y [Sh]T/ Form of Reference mol*ti3 mol*ti3 SbzO& 35 1.3464” 8.66x10” r? [61 35 1.6827* 0.0111 r? WI 35 2.0185” 0.0141 r? WI 35 2.3553” 0.0161 r? WI 50 0.6715* 5.95x1o-3 r? WI 50 1.0077” 8.65~10‘~ r? WI 50 1.3432” 0.0119 r? [61 50 1.6789* 0.0149 r? [61 50 2.0134* 0.0192 r? [61 90 5.95x1o-7 3.1OxlO~ C [71 90 5.95x1o-7 4.oOxlO~ C [71 90 5.95x1o-7 3.20~10~ C [71 90 5.95x1o-7 2.70~10” C [71 90 5.95x1o-7 3.3ox1o-4 C [71 90 5.95x1o-7 290x10~ C II71 100 7.28~10-~ 3.40x10” r? [21 200 2.23~10-~ 4.44x1o-3 C [71 200 2.23~10‘~ 4.52~10-~ C [71 200 2.23~10-~ 4.36~10” C [71 200 2.23~10-~ 4.53x10” C [71 200 2.23x 1O-6 4.90x10” C [71 200 2.23~10-~ 4.4ox1o-3 C [71 200 2.23~10-~ 3.90x10” C [71 200 2.23~10-~ 4.35x10” C [71 200 0.01 5.87~10” C [71 200 0.01 5.92~10” C 171 200 0.03 8.04x 10” C [71 200 0.03 8.85~10” C [71 200 0.1 0.0145 C [71 200 0.1 0.0153 C 171 t Initial hydroxide ion concentration, except when marked by an asterisk (*). In those cases, the hydroxide concentration was also determined at the end of the experiment, and is the value reported here. $ r:orthorhombic form (valentinite); c: cubic form (senarmontite)

2.3.2 Previous Solubility Measurements for Sb205

The solubility of (hydrated) Sb205 at 35°C in acid solutions was reported by Tourky and Mousa [4]. Their solid was prepared by hydrolysis of the Sb(V) hydrochloride salt and dried by heating to 90°C. Although the reported total concentration of aqueous antimony species in water in contact with this solid was quite low (Table 2-2), Baes and Mesmer [lo] concluded that it is unlikely a pure solid phase was present in these experiments, because of the ease with which gels 6 are prepared from such solutions. Glixelli and Przyszczypkowski [34] derived a “maximum” solubility value from a 15-minute equilibration at 100°C.

Table 2-2: Reported Solubilities of Sb205

T/“C [Sb]T/mO&i3 Reference Medium 35 0.000487 [41 4.600 mol.dm” HCl 35 0.000372 [41 4.092 moldm” HCl 35 0.000287 141 3.748 moldm” HCl 35 0.000125 [41 2.900 moldm” HCl 35 0.000059 141 2.458 moldm” HCl 35 0.000035 [41 1.981 moldm” HCl 35 0.000043 [41 1.064 mol.dm-3 HCl 35 0.000057 141 0.5 16 mol.dm-3 HCl 35 0.00007 1 141 0.100 mol.dm-3 HCl 35 0.ooo101 [41 0.050 mol.dm-3 HCl 35 0.00027 1 141 water 100 0.0212 [341 water

2.3.3 Previous Solubility Measurements for Sodium Antimonate(V)

There have been several reports of solubility measurements for sodium antimonate in water [35-391 (Table 2-3).

One of the problems with the reported solubilities is that the pH or ion concentration was not reported for any of the experiments. Consequently, the nature of the anionic antimony

Table 2-3: Reported Solubilities of Sodium Antimonate in Water

T/“C [Sb],/mol*ti3 Reference Notes 15 (?) 0.00538 [371 18 0.00229 1361 25 0.00299 1361 25 0.0053 [381 25 0.0033 1391 25 0.012 ~251 avg. Sb, Na analyses 33.5 0.00412 1361 35 0.0044 c391 50 0.015 [381 50 0.0060 [391 70 0.0084 [391 75 0.030 [381 80 0.0093 r391 100 0.03 [371 7 species is not known. The values reported by Tomula [36] and Blandamer et al. [39] are similar, while those reported by Urazov et al. [38] are markedly higher. The value based on the % composition analyses of Dravotsky and Karl&k [25], where both the sodium and antimony concentrations were measured, is very high, and indeed greater than the solubility of the lithium salt as reported in the same paper.

3. EXPERIMENTAL SOLUBILITY MEASUREMENTS

3.1 General Procedures for Measurements for T 2 200°C

The solubility measurements at 200 to 300°C were carried out in aqueous sodium hydroxide solutions in a 300 mL Autoclave Engineering autoclave equipped with a stirrer. Solutions were sampled at temperature by preheating the stainless-steel filter holder and sampling line to the temperature of the autoclave. The sampling-line connection to the autoclave was then opened, and liquid driven by the hydrostatic pressure in the autoclave was forced through a 0.45 pm filter. Immediately after passing through the hot filter, the sample was condensed, weighed and acidified. Samples of the final solid(s) were recovered at the end of the experiment after the autoclave had been cooled to room temperature. Powder X-ray diffraction (XRD) patterns were obtained for the antimony solids using a Siemens Diffractometer with Cu Ko,i radiation. Antimony concentrations were determined by neutron activation and inductively coupled plasma-atomic emission spectroscopy (ICP-AES).

3.2 Antimonv(III)

Sb203 (Aldrich, 99.999%) was used in these experiments without further purification. The XRD pattern for the oxide showed lines only for the orthorhombic form (valentinite). Because it was uncertain whether traces of oxygen would generate Sb(V) in our solutions at elevated temperatures, the Sb203 was mixed with metallic antimony (Alfa, 99.9999%), and this mechanical mixture was used as the charge in the static autoclave. Excess oxygen would then be taken up by the metal, the overall reaction being:

Sb(c) + 1.502(g) + Sb203(c) (3.1)

The XRD pattern of the oxide in the mixture recovered from the autoclave after the 300°C solubility experiments was essentially identical to the pattern for the initial valentinite. The measured solubilities are listed in Table 3- 1.

3.3 Antimonv(V)

3.3.1 Preparation and Characterization of the Solid Phases

Three different solids were used in the present solubility studies, and at least one of these materials was probably a mixture or solid solution. 8

Table 3-l: Experimental Solubility Measurements for Sb2O$Sb Mixtures

-3 Run T/“C Contact [Sbhd PH’liniti~ [OH’]e,,d mol*dm Duration/d mol*ti3 mol*ti3 14-1 200 6 0.0054 0.0104 0.0067 17-1 200 3 0.0023 0.0028 1 0.00042 14-2 250 1 0.0032 0.0104 0.0075 15-2 250 1 0.0016 0.0104 0.0062 16A-2 250 1 0.002 1 0.0028 1 0.0022 17-2 250 2 0.0016 0.0028 1 0.0012 15-1 300 2 0.0017 0.0104 0.0049 16A-1 300 6 0.0023 0.0028 1 0.0016 14-3 25 1 0.00064 0.0104 not measured 16A-3 25 5 0.00048 0.0028 1 not measured 17-3 25 2 0.00034 0.0028 1 0.0014

Sodium antimonate, nominally NaSb(OH)b, (solid A) was prepared as described in the literature [38,39]. Potassium antimonate (Aldrich Chemicals, 5 g) was dissolved in cooled boiled-out, distilled, deionized water (100 cm3) at 75”C, and the solution temperature was then adjusted to 50°C. Precipitation of sodium antimonate was initiated by the addition of 1 g NaCl dissolved in 10 cm3 of distilled water (at 50°C) and then the solution was immediately cooled, first to 25°C (for several hours), then in an ice-water bath. The crystals were filtered and washed with 250 cm3 of ice-cold distilled water and 250 cm3 of cold ethanol. The solid was dried in an oven at 105°C.

Solid B was found to form on heating sodium antimonate (solid A) in contact with 0.05 mol dm” NaOH(aq) solution in a titanium autoclave at 250°C for one week. The residual solid was washed with a very dilute solution ( lo4 moldm”) of aqueous sodium hydroxide, either at room temperature, or by treatment for 1 d at 25O”C, and was filtered and oven-dried at 105°C.

Solid C was found to form on heating sodium antimonate (solid A) in contact with aqueous sodium hydroxide solution at 75°C for times ranging from several days to several weeks. The solid was recovered from the solution by filtration and oven-dried at 105°C.

Powder XRD patterns for the three solids were obtained using a Siemens Diffractometer with Cu &i radiation (Figures 3-1,3-2(a) and 3-3). The powder XRD patterns found for some samples of solid A were consistent with that reported for NaSb(GH)h [21,40], although there was considerable variation from sample to sample. In some spectra, several of the peaks in the diffraction pattern did not appear (the pattern was then consistent with a face-centred cubic structure, a = 0.80 nm), while many of the remaining peaks were very strong and sharper than found for other samples. It may be that certain samples as prepared for XRD analysis were layered, causing some peaks to be weak. However, the “cubic” pattern could not be specifically 9 identified with any found in the literature. Initially, it seemed that the “cubic” solid could be identified with solid C discussed below; however, additional experiments showed the system to be more complicated.

Solid B provided a pattern that was identifiable with the pyrochlore structure, similar to the pattern of partially hydrated antimony(V) oxide, Sb205.xH20. Stewart et al. [18] reported a series of XRD patterns showing the transformation on stepwise heating of “antimonic acid” in air from room temperature to 735°C (at this temperature, reduction to Sb204.35 is complete). The reported changes in the peak positions during the heating are small; however, the intensities change markedly. The pattern of the material recovered from our experiments (presumably in continual contact with water during the experiments) very closely resembled the pattern reported by Stewart et al. [18] for samples fired in air at 220°C for 50 h (Figure 3.2(b)). However, elemental analyses of our pyrochlore-type solid B (see below) indicate that it is not a simple hydrated oxide, as it contains substantial amounts of sodium. For samples of solid C, the XRD patterns were related to those of the initial “NaSb(OH)i’ solid. ln some cases, the patterns of the solid recovered after long equilibration periods at 75°C resembled the initial solid; in other cases, certain peaks were markedly weaker.

Neutron activation analysis was carried out on samples of solids B and C. Solid C was found to be 8.7 wt.% Na, 46 wt.% Sb, i.e., a Na:Sb atomic ratio of (1.00 f 0.07), and an apparent molar mass of (265 * 13) g per mole antimony (assuming uncertainties of 5% in the neutron-activation analysis values). Analyses for three separately prepared samples of solid B are shown in Table 3-2. All three samples were prepared by heating solid A in 0.05 mol.dm-3 NaOH(aq) at 250°C for seven days. Sample B-6 was further washed with 0.0001 mol.dm‘3 NaOH(aq) at room temperature, and conditioned for five days at 250°C in 0.01 mol+dm-3 NaOH(aq). Sample B-9A was washed with 0.0001 moldm‘3 NaOH(aq) at room temperature, and conditioned for a further day at 250°C in that medium.

0 4 Xl 60 50 40 30 20 10

Figure 3-l: X-Ray Diffraction Pattcm for NaSb(OH)h - Solid A 10

100 mw (3) I (b) I I 80

g- i, % Km- Iii. i 8 am- I40 1 z IWO- al

O- 70 0 60 60 40 30 2-l 10 70 00 50 40 30 20 10 28 28

Figure 3-2: X-Ray Diffraction Pattern for (a) the Pyrochlore Structure Sodium Antimonate - Solid B; (b) Partially Dehydrated Sb205 [ 181

Figure 3-3: X-Ray Diffraction Pattern for NaSb(GH)h - Solid C 11

Table 3-2: Results of neutron activation analyses of solid B

Apparent Sample wt.% Na wt.% Sb Na:Sb Molar Mass per Sb 6 8.5 57 0.79 z!z 0.06 214 7 8.0 57 0.74 + 0.05 214 9A 7.8 58 0.71 f 0.05 210 avg. 8.1 57.3 0.75 + 0.03 212

From the average, assuming 5% uncertainties in the analyses, solid B was found to be 8.1 wt.% Na and 57 wt.% Sb, i.e., a Na:Sb atomic ratio of (0.75 + 0.03), and an apparent molar mass of (212 f 4) g per mole antimony. Table 3-3: Molar Mass per Mole Sb of Various Antimony(V) Compounds Containing Oxygen, Sodium or Hydroxide Ions or Water

Formula App. Molar Mass per Sb solid B 212+4 solid C 265 f 13 Sb2G5 161.8 Sb205.H20 170.8 NaSb03 192.7 NaSbOs.H20 210.7 NasHSb40i2.6H20 214.3 NaSb(GI& 246.7 Na2H2Sb207.5H20 246.7 Na2H2Sb207’6H20 255.8 NaSb03.3.5HzO 255.8 264.8

Comparison of the apparent molar masses (Table 3-3) with those for various possible solids suggests that solid C is probably a hydrated sodium antimonate or a hydrated “dihydropyro- antimonate” (but see Ref. 8). The estimated water-to-antimony ratio for solid C is slightly greater than might be expected from earlier work [28]. Solid B has a fairly high sodium content, considering the similarity of the XRD pattern to that of the pure hydrated oxide. On the basis of the diffraction pattern, it was concluded that solid A decomposes in water at 250°C, and forms a pyrochlore solid akin to the hydrated oxide, but incorporating at least some sodium in the solid. However, the material recovered from the autoclave is sufficiently insoluble, under most conditions, that quantitative elemental analysis by standard methods is difficult. The peaks in the diffraction pattern are quite sharp, whereas the presence of a large percentage of amorphous impurity in the solid might have been expected to cause substantial broadening or an erratic baseline. Various compounds based on polymeric hydrated antimony(V) oxide and having the 12 pyrochlore-type structure generate almost identical XRD patterns [8]. For example, Bauer et al. [32] reported patterns for a series of partially substituted solids Na2,(H30)2-22,Sb2G6’H2G, and Gol’dshtein et al. [33] reported a very similar pattern for a solid characterized as NaSbmSbT07. Stewart and Knop [ 171 concluded that the pyrochlore structure would he found only for sodium antimonate compounds having a Na:Sb ratio of 5 0.67. The analyses of our solid B are consistent with a gross formula with a somewhat greater ratio, (Sb205)(Na20)0.75.3.04H20, but this is within the continuous range of solid solutions with the pyrochlore structure (0.20 5 a I 0.87) reported by Bauer et al. [32]. The Na:Sb ratio in the three analyzed samples decreased in the order B-6 > B-7 > B-9A, whereas the concentration of Na+(aq) in the last solutions contacting the solids was B-7 > B-6 > B-9A, and the differences in the ratios are within the uncertainty limits of the analyses. Thus, any correlation between the value of a and the concentration of the last NaOH(aq) to contact the solution is weak. Of course, the extent of hydration of the solid may well have changed during cooling of the autoclave to room temperature. It is also possible that solid B as synthesized in the present work is a mixture of a mixed oxide and the hydrated oxide and/or an amorphous sodium-containing solid. However, except for the extent of hydration, the formulation is consistent with that proposed by Bauer et al. [32] (Na2a[H(H20)]2_~Sb206.H20, with a = 0.75).

3.3.2 Solubility Experiments for Temperatures Below 100°C

3.3.2.1 Preliminary Results

The solubility measurements at lower temperatures were carried out in Nalgene high- polyethylene bottles held in a thermostated bath. In a preliminary study of the solubility of sodium antimonate (solid A), the solutions were initially equilibrated at (25.0 + 0.5)“C for 17 days, sampled, equilibrated at 75°C for four days, sampled and finally re-equilibrated at 25°C for ten days before final sampling and examination of the solids. The results are listed in Table 3-4. Unfortunately, insufficient solid was used in the experiments with 0.001 moldrn-3 NaOH(aq), and all solutions were undersaturated both at 25 and 75”C, as was the 0.01 mol.dm-3 NaOH(aq) solution at 75°C. The solubilities were considerably greater at 75°C than at 25°C for both the 0.01 and 0.1 moldrn‘3 NaOH(aq) solutions.

The apparent increase in the solubility of the solid on cooling to 25°C might suggest that the solid that formed at 75°C is less stable than solid A at 25°C. However, the kinetics for transformation of solid C to the original solid A must then be quite slow at 25”C, as the solutions were left to re-equilibrate at the lower temperature for ten days before the solubilities were measured. It is also possible that solids A and C are essentially identical, and that the experimental equilibration periods used were too short.

3.3.2.2 The Solubilities of Solids B and C at 25 and 75°C

Experiments were also done (Table 3-5) to establish the solubilities of solids B and C in 0.003 and 0.04 moldm-3 NaOH(aq) solutions at 25 and 75°C. These results are discussed in Sections 4.2.2 and 4.2.3. 13

Table 3-4: Experimental Solubilities of NaSb(GH)h (initially Solid A) from the Present Study

Initial [NaOHJ/moldni3 T/“C [Sb]T/mol*dm” 0.001 25.4 2 (0.0016 f 0.0001) 0.001 25.4 2 (0.0016 * O.OOOl)* 0.01 25.4 (0.00060 * 0.00004) 0.01 25.4 (0.00065 f 0.00004) 0.01 25.4 (0.00132 + 0.00008)” 0.10 25.4 (0.000100 + O.OOOOO9) 0.10 25.4 (0.000100 f O.OOOOO9) 0.10 25.4 (0.00015 f o.OOOOl)* 0.10 25.4 (0.00014 f 0.00001)”

0.001 75.4 2 (0.0016 i 0.0001) 0.01 75.4 2 (0.0029 f 0.0002) 0.1 75.4 (0.0023 f 0.0001) 0.1 75.4 (0.0023 f 0.0001) * Values were obtained from re-equilibration of the solutions held four days at 75.4”C and are probably solubilities for the metastable solid C (see text).

3.3.2.3 Other Experiments

A further series of experiments was carried out at 75°C (and one additional experiment at 25”C), in an attempt to determine the relative stabilities of solids B and C. Samples of the solids (approximately 0.2 g of each) were contacted with 10 cm3 of deionized, distilled water in Nalgene high-density polyethylene bottles held in a thermostated bath. Each bottle was contained in a closed outer bottle to minimize the possible ingress of COz(g). Bottles were sampled after 2 1 to 155 days and solutions were submitted for Sb and Na analysis by ICP-AES. The hydroxide ion concentration of each final solution at room temperature was determined by measuring (Accumet 25 pH meter) the potential of an Accumet 13-620-295 electrode (for high pH with low sodium ion error) against an Accumet 13-620-5 1 calomel reference electrode. The potentials for several standard NaOH(aq) solutions of different known concentrations were measured with the same electrodes on the same day, and the unknown hydroxide concentration was determined by comparison with these standards. Checks were done to ensure that the response of the electrodes was approximately Nemstian over a wide range of pH (by also using measurements with standard acid solutions), and that sodium errors were either negligible or could be corrected by comparison with the NaOH(aq) standards. The residual solids were recovered for XRD analysis. The analysis results are listed in Table 3-6.

Preliminary nuclear magnetic resonance (NMR) results from 121Sb line-width measurements indicate that the Sb(V) in aqueous basic solutions is not in a totally symmetrical environment (or that more than one species is present). This result may be related to whatever phenomenon was responsible for the absorption spectroscopic results of Jander and Ostmann [ 131. 14

Table 3-5: Total Antimony Concentrations for Solids B and C as Measured for Basic Oxidizing Solutions at 25 and 75”C*

Initial Final [Sb]&nohlm-3 wnxpt Solid Duration/d [NaOHj/mol-ti3 [Na]T/mol-ti3 /mol-fi3 25°C 0.0029 (0.0034 * Oooo2) (Oooo34 f 0.oooo2) 0.0018 B-9A 46 0.0029 (0.0034 f oooo2 j (oooo35 f ooooo2 j 0.0016 B-9A 46 0.0029 (0.0027 f O.OOOl) (O.OOO18 f O.OOOOl) 0.0010 B-7B 83 0.0029 (0.0025 f 0.ooo1) (0.ooo18 f 0.oooo1) 0.0011 B-7B 83 0.0438 (0.044 f 0.003) (0.ooo17 z!z 0.oooo1) 0.040 B-7B 43 0.0438 (0.044 + 0.003) (0.ooo17 + 0.oooo1) 0.040 B-7B 43 0.0438 (0.043 +- 0.ooo) (0.ooo17 f 0.oooo1) 0.040 B-7B 83 0.0438 (0.044 f ooooj (O.OOol8 f oooool j 0.049 B-7B 83 0.0029 (O.oo51~ 0.003) (0.0018 f O.OOOl) ooo19 c2 43 0.0029 (Ooo50 St Oooo3) (0.0018 f ooool j 0.0018 C2 43 0.0029 (Ooo50 + Oooo3) (0.0018 + O.OOOl) ooo15 Cl 83 0.0029 (O.oo5 1 f 0.ooo1) (0.0018 f O.OOOl) ooo15 Cl 83 0.0438 (0.048 f 0.003) (O.OOO27 f O.OOOO2) 0.043 Cl 43 0.0438 f f 0.045 Cl 43 0.0438 (0.049 f 0.003) (Oooo30 f 0.oooo2) 0.049 Cl 83 0.0438 (0.044 f Oooo) (O.OOO25(O.OOO26 + O.OOOO2) 0.039 Cl 83 0.0029 (U.oo33 f O.oooo)

Table 3-6: Results of Equilibration of Mixed Antimony Solids with Water at 75°C (unless otherwise noted)

g g J32O Test l@?JW~ lO?Wr [OH’1 solid B NaSb:OHjd Duration moldm” mol*dmJ moldm” * Days 0.1922 0.1932 c 9.955 21 9.35 + 0.65 7.76 f 0.45 < 10-5, 2 1o-9 0.2627 0.1989 A 9.941 25 5.83 * 0.44 2.71 f 0.16 < 10-5, 2 1o-9 0.2637 0.2003 A 9.950 25 5.83 + 0.44 2.92 f 0.18 5 10-5, 2 1o-9 0.2637 0.2017 A 9.936 42 15.1 + 0.9 10.8 f 0.7 5 lo-5, ;r 1o-9 0.2597 0.2024 A 9.961 42 14.4 f 0.9 10.5 f 0.7 < 10-5, 2 1o-9 0.187 0.169 C = 10 83 3.31 f 0.17 2.63 Z!I 0.16 5 10-5, 2 1o-9 0.2628 0.2004 A 9.953 213 27.8 f 1.7 16.4 + 1.6 6 x 1O-5 0.1965 0.1835 c 9.925 155 3.72 f 0.24 2.63 f 0.16 I 105, 2 1o-9 0.2019 0.1992 c 9.965 1 43 (25°C) 9.79 f 0.65 7.72 f 0.45 < 10-5, 2 1o-9 * Initially solid A or C, but probably all solid A was converted to C by the end of an experiment.

3.3.3 Solubilities for Temperatures from 200 to 300°C

The solubility measurements at 200 to 300°C were carried out in aqueous sodium hydroxide solutions in a titanium autoclave equipped with a stirrer, as described in Section 3.1.

3.3.3.1 Preliminary Results

Two preliminary experiments were carried out to determine the approximate solubility of antimony(V) in basic solutions at 250°C. In both experiments, the initial solid was solid A, which was found to convert to solid B during the experiment. The solubilities measured from these samples taken at 250°C are listed in Table 3-7. The solubilities reported for the solutions with the same initial NaOH(aq) concentration are for samples taken from the same experimental run at different times.

After the autoclave had been cooled to room temperature, but prior to removal of the autoclave lid, gas was used to slightly pressurize the autoclave to force liquid samples through the sampling line. The solution concentrations of antimony and sodium species after cooling the autoclave to room temperature are listed in Table 3-8. This table includes the results from later experiments (runs done to synthesize further samples of solid B), in which the solutions were not sampled at 250°C.

As discussed previously in Section 3.3.1, the solids recovered from both preliminary experiments had powder X-ray diffraction patterns that bore no resemblance to that of the original solid A. Instead, these hard, compacted solids had essentially the same XRD (pyrochlore structure) pattern as a slightly hydrated oxide [ 181 and various mixed antimony oxide solids. 16

Table 3-7: Total Antimony Concentrations as Measured over Solid B (initially solid A, but converted to solid B during the experiment) for Basic Oxidizing Solutions at 250°C

Initial [Sb]T/mol*dm” Time/d [NaOH)/mol*dm.3 0.01 (0.0027 f 0.0002) 3 0.01 (0.0027 f 0.0002) 6 0.05 (0.0034 f 0.0002) 4 0.05 (0.0027 f 0.0002) 6 0.05 (0.0029 it 0.0002) 7

Table 3-8: Total Antimony Concentrations as Measured over solid B (initially solid A, but converted to solid B during the experiment) for Basic Oxidizing Solutions after Heating to 250°C and Cooling to Room Temperature

Initial Final [Sb]&uol*dm” Sampled [NaOHlhuokdm” [Na]TlmoldmJ O.oool (0.0058 it 0.0003) (0.0021* 0.0001) after 250°C expt. 0.01 (0.03 1 f 0.003) (0.00044 f 0.OOoo3) after 250°C expt. 0.05 (0.066 f 0.005) (0.00023 f 0.00002) after 250°C expt. 0.05 (0.073 + 0.005) (0.00015 f 0.OoOO2) after 250°C expt. (0.077 + 0.005) (0.00017 f o.ml) re-equilib. 8 d at 25°C (0.08 1 + 0.005) (0.00018 + 0.00002) re-equilib. 25 d at 25°C (0.085 * 0.005) (0.00019 f 0.OOoO2) re-equilib. 41 d at 25°C (0.082 f 0.005) (0.00017 f 0.OoOO2) re-equilib. 132 d at 25°C 0.05 (0.077 It 0.005) (0.00017 f 0.OOoO2) after 250°C expt. 0.05 (0.075 f 0.004) (0.00026 f 0.00003) after 250°C expt.

The hydroxide ion concentrations of the final solutions were expected to be dependent almost entirely on the quantity of sodium hydroxide formed from decomposition of the salt.

2NaSb(OI-&(s) + Sb205.xH20(s) + 2NaOH(aq) + (5 - x)HzO(l) (3.2)

Indeed, the final solution from the second experiment (on cooling to room temperature) was found to be strongly basic (> 0.2 moldm‘3 hydroxide). The analytical results for the final sodium ion concentrations (Table 3-8) are markedly greater than the initial concentrations, but lower than 0.1 mol.dm-3. The similar antimony concentrations found for the solutions at 250°C in the two experiments (Table 3-7) suggested that the final hydroxide concentrations in the two experiments were probably similar, or that the solubility of solid B at 250°C was almost independent of the hydroxide concentration in the moderately strong basic solutions. The final total concentrations of antimony in solution at room temperature (after cooling from 250°C) appeared to vary with the measured sodium ion concentrations, but it was unclear whether the differences might also have resulted from variable equilibration times at the lower temperature. Therefore, samples of 17 solid and solution from a third experiment were re-equilibrated with 10 cm3 aliquots of the final solution at 25°C for 8 to 132 d (using procedures similar to those described in Section 3.3.2.1). The total solution concentrations of antimony and sodium species from this third experiment are also listed in Table 3-8.

3.3.3.2 The Solubility of Solid B as a Function of Temperature and Hydroxide Ion Concentration

Samples of solid B were washed with water and equilibrated with aqueous NaOH (0.001,0.005 or 0.01 mol.dms3) at 250°C for 48 h, sampled, left for an additional 24 h, sampled again, cooled to 25°C overnight, and sampled again. Each solution sample was immediately mixed with an excess of 2 moldrn-3 HCl(aq). The total antimony concentrations and the solution sodium ion concentrations in the samples were determined (Table 3-9). After the autoclave had been cooled to room temperature and the contents sampled, the solid was recovered, washed with very dilute sodium hydroxide solution (usually 0.0001 moldm~3), and re-equilibrated with a fresh sample of the aqueous sodium hydroxide. The hydroxide concentrations of the sampled solutions were measured by comparison with suitable concentration standards (as discussed in Section 3.3.2.3), although in some cases the results showed considerable scatter. The procedure was carried out for duplicate experiments using three different concentrations of sodium hydroxide. Experiments were then carried out using 0.01 moldm-3 NaOH(aq) at 200 and 300°C. For the experiments at 2OO”C, slightly longer equilibration periods were used.

Over the course of several equilibrations, small amounts of a very fine black solid were formed in mixture with the white solid B. An XRD analysis of the (mechanically separated) black solid showed peaks at 0.372 and 0.413 nm that do not occur in the pattern for hydrated Sb205. A sample was dissolved using a lithium tet.ra/metaborate fusion method, and analyzed by inductively coupled plasma- (ICP-MS). This indicated that the major identifiable component in the black solid was antimony, with a similar (atom fraction) amount of sodium and small amounts of and (normal impurities in antimony compounds). Thus, the black solid is probably another unidentified sodium antimonate and not a product of a reaction involving the titanium of the autoclave.

4. DISCUSSION

From a literature survey and our own preliminary results for both antimony(III) and antimony(V), it appears that the behaviour of antimony solids and solutions is best understood for reducing conditions near room temperature. For temperatures above 50°C, the only previous useful results for aqueous solution species appear to be for antimony(III) in neutral and basic reducing conditions to 200°C [7].

For oxidizing conditions, with the exception of acidic concentrated antimony solutions at room temperature, it is not certain what solution species are formed nor what solids are thermo- dynamically stable. There are good indications that one or more anionic species are formed in 18

Table 3-9: Total Antimony Concentrations for Solid B as Measured for Basic Oxidizing Solutions at 200 to 3OO”C, or after Cooling to Room Temperature

Initial Final [Sb]&uol*dm” [OH7 sampled** [NaOH]/mo&dm” [Na]+uol*dm” 200°C 0.01 (0.0141 f 0.0009) (0.0045 f 0.0003) 0.0075 at 2OO”C(a) 0.01 (0.0141 f 0.0009) (0.0044 f 0.0003) 0.0098 at 2OO”C(b) 0.01 (0.0109 f 0.0007) (0.0021 f 0.0001) 0.0068 at 25”C(c) 0.01 (0.0133 f 0.0008) (0.0040 f 0.0002) 0.0078 at 2OO”C(a) 0.01 (0.0148 + 0.0009) (0.0044 +: 0.0003) 0.0077 at 2OO”C(b) 0.01 (0.0108 A 0.0006) (0.0012 zk 0.0001) 0.0065 at 25”C(c) 250°C 0.001 (0.0075 f 0.0004) (0.0042 f 0.0002) 0.0022 at 25O”C(a) 0.001 (0.0072 f 0.0014) (0.0042 f 0.0002) 0.0023 at 25O”C(b) 0.001 (0.0070 f 0.0004) (0.0046 zt 0.0003) 0.0009 at 25”C(c) 0.001 (0.0061 f 0.0003) (0.0040 f 0.0002) 0.0005 at 25O”C(a) 0.001 (0.0064 f 0.0003) (0.0041 f 0.0002) 0.0015 at 250”C(b) ------__-___0.001 ‘_(o~oi._~o.._)__@0052 f 0.0003) ~0.0037(~~~s70~ti~~-- f 0.0002) .------o.ooo2 .------at 25”C(c) 0.005 0.0072 at 250°C (a) 0.005 (0.0135 f 0.0008) (0.0037 f 0.0002) 0.0073 at 250”C(b) 0.005 (0.0106 It 0.0006) (0.0017 f 0.0001) 0.0059 at 25”C(c) 0.005 (0.0104 f 0.0006) (0.0038 f 0.0002) 0.0056 at 250”C(a) 0.005 (0.0110 f 0.0007) (0.0040 f 0.0002) 0.0039 at 250”C(b) ------0.005 ------(0.0098 f OXKIO6) (_0.0034--______f 0.0002~ __ .------0.0038 at 25”C(c) 0.01 (0.0148 + 0.0009) (0.0039 f 0.0002) 0.0083 at 25O”C(a) 0.01 (0.0157 f 0.0010) (0.0037 f 0.0002) 0.0112 at 250”C(b) 0.01 (0.0117 f 0.0007) (0.0015 f 0.0001) 0.0067 at 25”C(c) 0.01 (0.0145 + 0.0009) (0.0036 L+Z 0.0002) 0.0099 at 250”C(a) 0.01 (0.0150 f 0.ooo9) (0.0036 zt 0.0002) 0.0089 at 250”C(b) 0.01 (0.0125 i 0.0008) (0.0025 + 0.0001) 0.0082 at 25”C(c) 300°C 0.01 (0.0171 f 0.0010) (0.0024 + 0.0001) 0.0092 at 3OO”C(a) 0.01 (0.0160 zt 0.0010) (0.0020 f 0.0001) 0.0128 at 300”C(b) 0.01 (0.0158 f 0.0010) (0.0009 f 0.ooo1) 0.007 1 at 25”C(c) 0.01 (0.0124 * 0.0008) (0.0018 f 0.0001) 0.0070 at 3OO”C(a) 0.01 (0.0118 f 0.0007) (0.0015 + 0.0001) 0.0074 at 300”C(b) 0.01 (0.0147 + 0.0009) (0.0021 f 0.0001) 0.0073 at 25”C(c) 0.01 (0.0114 f 0.0007) (0.0012 + 0.0001) 0.0065 at 3OO”C(a) 0.01 (0.0137 f 0.0009) (0.0018 i 0.0001) 0.0098 at 300”C(b) 0.01 (0.0152 + 0.0010) (0.0012 + 0.0001) 0.0056 at 25”C(c) 0.01* (0.0035 f 0.0002) (0.0010 It 0.0001) 0.0018 at 3OO”C(a) 0.01* (11.1 f 1.1)10-5 (0.7 z!z o.2)10-5 < lo-’ at 300”C(b) 0.01* (20.0 zt 1.6) 1O-5 (1.4 f o.2)10-5 < lo-’ at 3OO”C(b) 0.01” (0.0478 f 0.0026) (0.0010 f 0.0001) _*** at 25”C(c) * Experimental problems (loss of water from the autoclave during the run, possible sampling line problems). ** Samples marked (a), (b) and (c) are from the same autoclave run; (a), (b )and (b’) are successive samplings on different days at the same temperature; (c) is the sample taken after the autoclave was brought back to room temperature (usually on the same day). *** Not measured. 19 basic solutions at room temperature, but prior to the present study there has been essentially no information as to whether this remains so at higher temperatures. The previous solubility measurements are suspect, because interconversion of several solids appears to be possible with changes in temperature. Any calculations of the speciation of antimony under oxidizing conditions based on the earlier solubility studies, particularly for temperatures greater than 25”C, should be regarded as, at best, highly speculative.

Antim0nvUI.I)

The following is an analysis of the solubility data obtained for SbzOs, listed in Table 3-l. The results are fairly scattered, and there are no evident trends of solubility with temperature or base concentration. The concentrations of base as measured in the equilibrated samples are probably less reliable than the total solution concentrations of antimony, and the “final” measured hydroxide ion concentrations were not used in the data analysis.

Solubility values for Sb203 in water and basic aqueous solutions have been reported previously by a large number of authors [2-71 (cf. Table 2-l). Except for the work of Popova et al. [7] and a single experiment by Schulze [2], all these measurements were done using solutions at temperatures between 15 and 50°C (Figure 4-l).

lop

15% W [1883SCH] 0 [73VAS/SHO] 25°C [39SLO] lo-*- [52GAY/GAR] $1 [73VAS/SHO] m+m _ this work au - 5 35°C X [48TOU/MOU] m + [73VASISHO] E 0 10”: 50°C % [73VAS/SHO] ti X ?a X X lOA- X X X n X x X

I I I II III 1o‘7 10” 1o-5 lo4 lo* 10” 10” loo [OH-] mol-dms3

Figure 4-l: Total concentrations of antimony(II1) in aqueous solution in equilibrium with Sb203 from 15 to 50°C (for details, see text).

In general, agreement between the results of the different studies is excellent-much better than most studies of oxide solubilities. Almost all the measurements, again except for those of Popova et al. [7], were done using the orthorhombic (valentinite) form of the solid rather than the cubic (senarmontite) form that is reportedly stable at these temperatures, or else the Sb203 was 20 synthesized under conditions known to yield the orthorhombic form. Bloom [3] carried out measurements over many months to compare the solubilities of the two forms at room temperature, and found a difference of only 0.36 in the logarithm (base 10) of the values; i.e., approximately 2 kJ.mol-’ at 25°C. This is a smaller difference in stability than the 7-9 kJ.mol-’ that has been suggested in several standard compilations [41-43], or the 5.7 k.Lmol-’ difference found in a reanalysis of available heat capacity, enthalpy and transition thermodynamic data, summarized in Appendix A.

Calculations based on the same assessment suggest that the total antimony(III) concentrations of Popova et al. [7] at 200°C over senarmontite should be systematically lower than our measured values over valentinite by a factor of 2.65. However, Popova et al’s values are, if anything, greater than those found in the present study. This result could indicate a small amount of valentinite impurity in the senarmontite sample of Popova et al., overestimation of the calculated difference in the Gibbs energies of formation of the two polymorphs, or systematic errors in either study. Regardless of the cause, most of the following data analysis has been done without differentiating between the solid actually used in each study. Based on the diffkulties we have encountered in attempting to synthesize senarmontite from aqueous solutions, it seems that in the unlikely circumstance that SbzOs were to precipitate from solution in a reactor system, the solid would probably be the less-stable valentinite.

The solubility results for Sb203 in neutral and basic solutions have generally been interpreted in terms of two equilibria:

Sb203(c) + 3H20 -T 2Sb(OH)3(aq) (4.1)

SbzOs(c) + 3H20 + 20H-(aq) * 2Sb(OH)i(aq) (4.2)

Many of the reported measurements list only the initial base concentration, not the final concentration. Where the only major aqueous antimony(III) species is a neutral species, or if the initial base concentration is much greater than the measured total concentration of antimony in solution, the data analysis is not affected. However, if the initial base concentration and the final antimony concentration are of the same order of magnitude, the final base concentration may be reduced from its initial value from the stoichiometry of reaction 4.2. The reported base concentrations of Gayer and Garrett [S] were stated to be initial concentrations, and it was assumed that this was also the case for the experiments of Tourky and Mousa [4], and Popova et al. [7]. It is the initial base concentrations that are shown in Figure 4- 1.

The equilibrium constant for reaction 4.1 depends on the activities of HzO(aq) and of Sb(OH)s(aq). In the present analysis, we have assumed that the activity for HzO(aq) is equal to 1.0, as is the activity coefficient of the neutral aqueous species, Sb(OH)s(aq). The equilibrium constant for reaction 4.2 can be written as Kz= a2(Sb(OH),) / (a2(OH-) a3(H20(aq))) where terms ¬ed [A] are molar concentrations, “a” is an activity and y is a molar activity coefficient. The ratio of the activity coefficients, Q,~(~B)&B-, is equal to 1 .O, provided the ionic medium in the solution is sufficiently low that the Debye-Htickel equation [44] (or extended versions of that equation, such as the Davies equation [45], that do not contain terms specific to the ion) is adequate.

Therefore, to a fust approximation, the total concentration of antimony(III) in solution can be expressed as:

[Sb]r = KWOHhl + PWHM

= KT + K; [OH-] (4.3)

Some of the studies [4,6] used hydroxide solutions with concentrations greater than those consistent with the Davies equation (i.e., IO.2 to 0.3 M). Unfortunately, such solutions also have provided the best evidence of formation of the Sb(OH)i anion at low temperatures.

Values at 25°C from the present study are systematically greater than those reported previously. Our measurements represent solutions that were saturated at 200 to 3OO”C, and these solutions apparently remained supersaturated when they were cooled to a temperature where the solubility is less. They had not returned to equilibrium in the 1-to-5-day before sampling. No other experiments seem to have been reported for the SbzOs-Hz0 system in which equilibrium was reached starting from supersaturated solutions.

A comparison of our solubility measurements with other measurements reported for temperatures between 90 and 200°C is shown in Figure 4-2. In general, our measured solubility values from 200 to 300°C are less than those measured by Popova et al. [7] at 200°C; this is true even if their values in water are compared with our values in basic solutions. This could also indicate that our solutions were undersaturated, but we cannot be certain without further experimental work.

Attempts were made to use the Sb203 solubility data to generate thermodynamic quantities ArGo, AS” at 25°C and an average value of A& for 15 to 300°C, A$ravg, for reactions 4.1 and 4.2.

Equation 4.3 was combined with the equation

K = -( l/RT) exp { AGO25 + (A&, avg - AS”&(T-298.15) -TA,C, avg ln(T/298.15)} (4.4)

where R is the gas constant (8.31451 J-K-‘.mol’) and T is the temperature in kelvin. Parameter values (Table 4- 1) of A,G”, A,S” and A& avs (i.e., A&, 15_3~oc, the average heat capacity of reaction between 15 and 300°C) were calculated using data from all the studies discussed above; 22

90°C 0 [75POPlKHO] 100°C 0 [1883SCH] 200°C [75POPKHO] 2w”c thiswork 250% A ltlkwork 0 thiswork : ..----arbitraryline : .’ slope=1 ,’ : ,’

,’

3ooc

[OH-] moledmd

Figure 4-2: Solubility of Sb203 at temperatures from 90 to 300°C. The “best-fit” line for 25°C based on the low-temperature data is shown for comparison, as is an arbitrary line with a slope of 1 .O. reported “initial” hydroxide ion concentrations were adjusted to “final” values using the “fitted” value of K2. The calculations were repeated, but in an attempt to minimize activity coefficient effects on the derived thermodynamic quantities, only measurements for solutions with total ionic strengths less than or equal to (a) 1.0 M and (b) 0.3 M were used.

Table 4- 1: Calculated Thermodynamic Quantities for the Dissolution of Sb203

4G” 4-s” 4c, 15-300°C Mrnol-’ /JK’~mol~l /PK-‘*mol-’ all data reaction (4.1) 24.9 f 3.5 47 Zk 66 -155 f41 reaction (4.2) 12.9 f 2.0 35 Ik 120 -23 f 76 all data [OH-] I 1 .O M reaction (4.1) 24.9 + 3.5 46f64 -151 AI 40 reaction (4.2) 12.9 f 3.0 56+204 -130 f 128 all data [OH-] IO.3 M reaction (4.1) 24.9 f 3.3 42f54 -134 f 33 reaction (4.2) 12.1 k4.6 -4 f 505 124f315 23

It appears that the thermodynamic quantities related to the formation of Sb(OH)i are still ill- defined, especially at higher temperatures. The calculated values of A~4.&, avg and Ar(4.2$‘25 are highly correlated. Indeed, it appears that Sb(OH)i is probably less important for temperatures 2 200°C than for those near room temperature. The temperature dependence of dissolution to form Sb(OH)s(aq) is much better defined, and it is apparent that although the solubility of Sb203 in neutral to basic solutions increases by more than an order of magnitude between room temperature and 2OO”C, it then remains constant, or begins to decrease only slightly between 200” and 300°C. Even at 300°C the total equilibrium concentration of aqueous antimony species in equilibrium with Sb203 is near 10s2 M for all basic solutions with [OH-] IO.3 M, more than an order of magnitude greater than at room temperature.

Based on the parameters for hydroxide ion concentrations I 0.3 M, the calculated total concentrations of antimony species at various temperatures, as shown in Figure 4-3, were estimated using the constants from an equal weight “least squares” treatment of all available Sb2Os solubility measurements (without regard to the particular polymorph) for hydroxide ion concentrations < 0.3 M.

L

I I I I I I I lo-' 10" 10" lo-' lo9 lo4 10"

[OH-] moLdma Figure 4-3: Calculated total solution concentrations of Sb(III) as a function of temperature and hydroxide concentration.

4.2 Antimonv(V)

4.2.1 Rationale for the Measurements Using “NaSb(OH)&)” and Other Sodium Antimonates On the basis of literature values for acidic solutions, (hydrous) SbzOs “antimonic acid” would be expected to have a high solubility in neutral and basic solutions, with the formation of anions 24 containing Sb(V). The easily purified monosodium salt, NaSb(OH)h(s) (probably better written as NaSb03*3H20), is reported to be stable and is only sparingly soluble. Therefore, measurements of the solubility of this solid as a function of aqueous sodium hydroxide concentration were expected to provide a method for investigating the behaviour of aqueous antimony(V) species in neutral and basic solutions. For example, the dissolution reaction NaSb03.3H20(s) * Na+(aq) + SbOj (aq) + 3H20(1) (4.5) has an equilibrium constant such that the concentration of antimonate in solution should vary inversely as the concentration of aqueous sodium hydroxide. If the reaction was 20H-(aq) + NaSbOs.3H20(s) * Na+(aq) + SbOi(aq) + 4H20(1) (4.6) the concentration of antimonate in solution would vary directly as the concentration of aqueous sodium hydroxide. If the reaction was

NaSb03.3H20(s) * Na+(aq) + OSSb205(aq) + OH(aq) + 2.5H20(1) (4.7) the concentration of antimonate in solution would vary as the inverse square of the concentration of aqueous sodium hydroxide. In each case, it should be possible to relate measurements at different temperatures to changes in stabilities of the aqueous antimony(V) species as a function of temperature.

However, the present work has established that in dilute sodium hydroxide solutions the simple sodium antimonate (solid A) is probably converted to solid C at some temperature between 25°C and 75”C, and then to solid B at higher temperatures. Thus, neither of the (presumably) simple compounds A or C is suitable for the study of solubility as a function of temperature. Nevertheless, solubility measurements using the simple sodium antimonates (A and C) provide a satisfactory starting point for a study of Sb(V) in basic solutions at low temperatures.

The use of the more complex, probably non-stoichiometric, solid B is more problematic. It seems to be reasonably easy to synthesize, and analyses suggest that variation in the stoichiometry is not extensive in basic solutions, and is not affected substantially by alteration in the washing or ripening procedures. Also, as discussed in a later section, solid B is stable with respect to solids A and C even at 25”C, except at high solution concentrations of Na+ (and/or OH-). However, solid B is still not well-characterized, although apparently similar material has been synthesized by other methods [29,32]. Study of the solubility of (hydrated) Sb205 might also have provided useful information, but that solid is not particularly simple to synthesize in a pure form [8, 171, and in contact with sodium hydroxide solutions the simple oxide would be converted, at least in part, to a sodium salt (or salts) [8,32]. Further, the extent of its hydration changes markedly with temperature [ 181. Thus, the results would be no easier to interpret than those for the sodium salts used in the present study.

4.2.2 Solubility of NaSb03*3H20(s) (NaSb(OH)6) in Basic Solutions

As Figure 4-4 shows, at 25°C solid C appears to be more soluble (i.e., less stable) than solid A. At 75°C solid A is unstable with respect to solid C, and the latter is much more soluble than at 25

25°C. At both 25°C and 75°C the antimony concentration of solutions in contact with solids A and C decreases with increasing Na+ (NaOH(aq)) concentration. This decrease is approximately linear, and the slope of the line in Figure 4-4 is close to - 1 (closer than if it is assumed that A and C are the same solid). We are unable to decide, based on present evidence, whether solids A and C are essentially the same or if they are two distinct phases.

From the experimental data listed in Table 3-5, the solubility (activity) products of solid C at 25 and 75°C have been calculated based on reaction 4.5

Ks, = aNa+ a SbO;* (4.8)

The activity coefficients were estimated for these relatively dilute solutions using an extended Debye-Htickel equation:

logloe = -A I?( 1 + 1.5 Ic.5) (4.9) with A=0.509 at 25°C and A = 0.564 at 75°C [46]. The total ionic strength was calculated from the experimental Na+(aq), SbOj (aq) (i.e., total Sb) and, where reasonable, the experimental OH(aq) concentrations.

v A25”C n B25’C B 75% 0 C 25% ‘\ A C 75°C

‘\ ‘\ -\ ‘\

10” [Sb], moLdme

Figure 4-4: Sodium ion concentration (M) as a function of total Sb(V) concentration for solubility measurements of sodium antimonates in basic solutions at 25 and 75°C. Lines from the least-squares fits to each set of values are also shown.

However, as can be seen from Table 3-5, the experimentally measured final hydroxide ion concentrations were substantially lower than the initial concentrations of the aqueous sodium 26 hydroxide solvent, especially when the initial concentration of NaOH(aq) was low. The reason for this is not clear, although the antimony solids themselves are unlikely to have been the primary cause (sample blanks with no solid showed the same tendency (Table 3-5)). The samples may have been contaminated with atmospheric CO2 despite the precautions (N2 gas flushing prior to closing the vials, and double containment) taken against this happening, or small amounts of acid may have been leached from the plastic bottles despite preconditioning by soaking in basic solutions and water. When the measured hydroxide ion concentration was substantially lower than that of the initial NaOH(aq), the ionic strength was assumed, for the purpose of the activity coefficient calculation, to be equal to the final concentration of Na+(aq). The extra uncertainties introduced by this procedure are less than 0.01 in log&,.

The solubility measurements listed in Table 3-4 were also used to determine values of the solubility products for solids A (at 25°C) and C (25 and 75°C). In these experiments, the final Na+(aq) and OH-(aq) concentrations were not determined. However, the initial concentrations of NaOH(aq) were sufficiently great that the final concentrations of these species could be assumed to be equal to the initial concentrations without introducing undue errors.

The average values of the solubility products (based on the measurements reported in both Tables 3-4 and 3-5), and comparison values from the literature, are listed in Table 4-2.

Table 4-2: Values of the Solubility Product for Solids Nominally Hydrated NaSbOs or NaSb(OH)e

Solid T/“C MediUm NUlIlber Avg. Sigma 95% Confid. Reference

NaSb(OH)e 18 Hz0 -5.325 1361 A 25 NaOH(aq) 4 -5.256 fo.046 fo.074 present work C 25 NaOH(aq) 11 -5.072 fo.050 io.034 present work NaSb(OH)e 25 H20 -5.100 [361 NaSb(OH)e 25 H20 -4.55 1 [381 NaSb(OH)e 25 H20 -5.017 [391 NaSb(OH)d 33.5 H20 -4.83 1 [361 NaSb(OH)b 35 H20 -4.776 [391 NaSb(OH)6 50 Hz0 -3.648 [381 NaSb(OH)h 50 Hz0 -4.518 1391 NaSb(OH)6 70 H2O -4.241 1391 C 75 NaOH(aq) 10 -4.094 fo.146 fo.105 present work NaSb(OH)a 75 H20 -3.046 [381 NaSb(O& 80 Hz0 -4.159 [39]

The log&, values of Urazov et al. [38] are markedly less negative than those from other sources, and it must be presumed that their analyses were affected by the presence of colloidal antimony solids. For 25°C our values for solid A are in fair agreement with, but slightly lower than, those of Tomula [36] and Blandamer et al. [39] for NaSb(OH)b. Our values for solid C at 25OC are actually more concordant with the literature values than those for solid A. This might 27 be explained if the methods used to synthesize or treat the solids [36,39] involved heating to slightly higher temperatures. The values of Blandamer et al. for 70 and 80°C are similar to ours (for solid C) within the uncertainties.

The similar results for the solubility products determined in water [36,39] and in 0.003 to 0.1 moldm~3 NaOH(aq) are consistent with the assumptions that these solids dissolve to form Na+(aq) and a singly charged anionic monoantimonate species (i.e., SbOj(aq) or Sb(OH)&aq)), and that the speciation is not dependent on the hydroxyl ion concentration in these solutions.

4.2.3 Solubility of Na2,[H(H20)]2_&b206.H20, a = 0.75 in Basic Solutions

4.2.3.1 Comparison of the Solubility with Other Solids at 25 and 75°C

The apparent equilibrium concentration of antimony in aqueous solution over Na2,[H(H20)]2_2,Sb206.H20 (a = 0.75) at 25 and 75°C in 0.003 to 0.04 mol.dm-3 NaOH(aq) (Table 3-5) is less than that of solid C (Figure 4-4), and this suggests that solid B is more stable than solid C under these conditions.

One of the simplest descriptions of the dissolution equilibrium is:

Na2a[H(H20)]2_2aSb206’H20(S) + (2-2a)OH(aq) * 2aNa+(aq) + 2SbOj(aq) + (5-4a)H20(1) (4.10)

and the equilibrium constant K, is defined by

(4.11)

For a = 0.75, this reduces to

0.75 ,-_p.25 = L4,75 aSbO; aNa+ (4.12)

Thus, the antimony concentration over hydrated Nai.sH.&b206(s) would be expected to decrease as the square root of increasing sodium hydroxide concentration (i.e., more slowly than in the case of the simple sodium antimonate). If this stoichiometry is accepted for solid B (cf. Section 4.2.3.2), assuming no changes in the predominant antimony species in solution with increasing hydroxide ion concentration (probably an oversimplification [ 13]), and further assuming that the antimony solution species over all the solids is the same, solid B would be expected to become unstable with respect to solid C only at NaOH(aq) concentrations of approximately 0.2 moldm-3 at 25°C and 1 moldm-3 at 75°C. Solid B would become unstable with respect to formation of solid A for NaOH(aq) concentrations of approximately 0.08 mol.dm-3 at 25°C. The solubility of Sb205 in water at 35°C (0.00027 mol.dm-3) reported by Tourky and Mousa [4] is very similar to the solubility of solid B in dilute basic solutions. 28

It was hoped that long-term equilibration of mixtures of two solids (B and C at 75”C, B and A at 25°C) in contact with water would generate a solution in equilibrium with both. The total antimony, sodium and hydroxide concentrations would then be fixed (for a specific value of a). These concentrations could then be compared with the values calculated from the two solubility products. Alternatively, if the solubility differences were large, one of the solids could be completely converted to the other. However, contacting water with mixtures of solid B and sodium antimonate (solids A or C) at 75°C even for more than 200 days apparently did not result in establishment of equilibrium between the solids, probably in part because the samples were not agitated continuously. The XRD patterns of the residual material show that both initial solids were still present after all of the experiments, although qualitatively the ratio of solid C to solid B appears to have decreased after extended equilibration times. The final solution concentrations of Na+(aq) were consistently greater than the total solution concentrations of antimony, and the final measured hydroxide concentrations were low (Table 3-6). This is consistent with simple partial dissolution of the solid C (or A) from the mixture, with some portion of the sample of solid C never coming into contact with the bulk of the solution. At present this is the best explanation we have found for the results listed in Table 3-6. However, there then is no apparent reason why the duplicate experiments over 25 days should have given essentially identical results, and the same applies to the pair of 42-day experiments. The same difficulty would arise even if we were to assume that equilibrium cannot be attained with one of the pure solids within these periods of time.

4.2.3.2 Solubility of Na20r[H(H20)]2_2,Sb206.H20 from 25 to 300°C

As for the results at lower temperatures, the solubility of solid B at 200 to 300°C varied only slightly with changes in Na+(aq) and OH- concentrations. If anything, the variation was less, especially near 250°C. The results from samples taken on successive days at 250°C (i.e., samples from the same autoclave run) are fairly consistent; the values from successive runs under the same conditions are somewhat less so. The total solution concentrations of antimony on cooling the autoclave to 25°C are lower than the high-temperature values, but generally greater than from solutions equilibrated for longer periods of time (Table 3-5). If all the results are considered together, regardless of the actual base concentration, the solubility of solid B in basic solutions increases from 25 to 2OO”C, but decreases slightly at 25O”C, and even more so at 300°C (Figure 4-5).

The calculated value of the activity product (KJ is strongly dependent on the value of a. If a is not independent of temperature, a set of values for the solubility product at different temperatures will not yield useful information concerning changes in stability of the aqueous antimony species as a function of temperature. Therefore, values of the activity product and a were first calculated from the results at each temperature for which measurements were done (Table 4-3). 29

o.ooo1 0 50 100 150 200 250 300 Temperature (“C) Figure 4-5: Solubility measurements for solid B, a mixed oxide of antimony(V) (hydrated pyrochlore-structure sodium salt, Nat.sH&b&I&o). Hydroxide ion concentration (mol.dm-3): 25°C O.OOlO-O.O49,75”C 0.0001 l-O.O42,2OO”C 0.0075-O.O098,25O”C 0.0005-0.0112, and 300°C 0.0065-0.0128.

Except for the value calculated from the 300°C solubility measurements, a does not differ greatly from the Na:Sb ratio of (0.75 It 0.03) found by neutron activation analysis of three separately prepared samples of solid B.

Table 4-3: Activity Products for Nah[H(H20)]2_2aSb206.H20 (logroK,) with Values of a Calculated from the Experimental Results for Each Temperature and from Values of logloKX(25”C) and a Calculated from the Results at all Temperatures*

T/“C Number of hh& &de ~ogloKx@[email protected]), Measurements 25 8 -(4.10 f 0.13) (0.62 f 0.03) -(4.52 z!z 0.11) 75 8 -(4.05 f 0.10) (0.76 z!z 0.02) -3.71 200 4 -(3.42 f 1.81) (0.77 f 0.44) -3.16 250 12 -(3.45 f 0.08) (0.76 f 0.02) -3.22 300 6 -(1.26 zk 1.01) (0.14 f 024) -3.37 all 36 - (0.7 1 & 0.02) - * The other fitted parameters are A&S = (38.4 -I 9.5) J.K-‘.mol-’ and A& = -(218 If: 27) J.K“.mol-‘.

Using the results for all five temperatures, a single (T independent) value for a, a value for logtoK, for 25”C, and values of AS and A&, for reaction 4.10 (assumed to be independent of temperature) were determined using a non-linear least-squares fit (Table 4-3). Using the fitted 30 value a = 0.71, none of the calculated values for the antimony concentrations differ from the corresponding experimental values by more than a factor of 2.1. Considering the possible sampling problems at high temperatures, the apparently slow approach to equilibrium at low temperatures, and the fact that the calculations were done using the simplification of assuming that the heat capacity of reaction is independent of temperature, this agreement is reasonably good.

5. CONCLUSIONS

Based on the work described above, we draw the following conclusions: l the solubility of Sb203 in basic solutions increases from 25 to 200°C and probably decreases slightly between 200 and 300°C; l it seems that in the unlikely circumstance that Sb203 were to precipitate from solution in a reactor system (unlikely because the concentrations of antimony solution species are too low), the solid would probably be the less-stable valentinite; l all of the antimony(V) solubility measurements are consistent with formation of a monoanionic antimony species in oxidizing basic solutions for temperatures from 25 to 25O”C, and probably to 300°C; l comparison of the solubility product of sodium antimonate(V) as determined in basic solutions with values reported in the literature suggests that the same antimony solution species is predominant in oxidizing solutions at 25 to 75°C from neutral solutions to solutions containing between 0.01 and 0.1 mol.dm-3 NaOH(aq); l simple sodium antimonate is converted to a hydrated pyrochlore-structure sodium salt, Na~[H(H20)]2_2,Sb206.H20 (a = 0.75 f 0.03) in basic aqueous solutions at 250°C; l the solubility of this pyrochlore-structure sodium antimonate increases from 25 to 2OO”C, and decreases at temperatures above 25O”C, probably primarily reflecting changes in the stability of the anionic antimony solution species (SbOj or Sb(OH);); l the temperature-dependence of the solubility of other antimony(V) solids in basic oxidizing solutions would be expected to change similarly (again based on the reasonable assumption that the solubility changes are primarily controlled by changes in the stabilities of the antimony solution species); l metastable antimony(V) solids can persist for extended periods of time in contact with oxidizing basic solutions at temperatures at or below 75°C; l although Sb(V) solids are generally not shown as having a region of predominance in potential-pH diagrams [9,47], it is probable that alkali metal antimonates are stable relative to Sb203 in basic oxidizing solutions; l the nature of insoluble antimony solids in basic oxidizing solutions is probably strongly dependent on the nature of the solutes, especially simple cations; l for modeling total antimony concentrations in oxidizing solutions, the use of antimony(V) species as discussed in Baes and Mesmer [lo] would overestimate the solubility of antimony at pH 12 even at 25°C; 31

all solids used in the present experiments would be expected to generate total antimony concentrations of 2 0.00005 moldm-3 in any neutral or basic solutions, assuming that no sodium salts have been added; this work demonstrates that under III’S conditions, precipitation of any antimony oxides or mixed oxides is unlikely; since the total concentrations of antimony species in the CANDU HTS are low (- 10s6 mol.dm-3 during antimony removal, and lo-” moldm-3 during operations), these results strongly suggest that precipitation of antimony oxides is not an important process in antimony activity transport; it cannot be ruled out that hydrated SbzOs (especially the pyrochlore form) might be less soluble in near-neutral, low-ionic-strength solutions; the formation of monomeric antimony species is consistent with the removal of antimony from basic solutions onto anionic ion-exchange resins.

ACKNOWLEDGMENTS

The ICP-AES analyses for antimony were carried out by S.L. Mitchell, C.J. Everall and R. Ryan, and the XRD patterns for the solids were obtained by J. Winegar. M. Totland did the ICP-MS analyses and, with P. Robinson, the neutron activation analyses. D. Guzonas and C. Stuart provided useful comments on a draft of the report. We also wish to thank D. Guzonas for many useful discussions.

7. REFERENCES

[l] Dundar, Y., Odar, S., Streit, K., Allsop, H., Guzonas, D., “Application of KWU Antimony Removal Process at Gentilly-2”, Proc. 7th International Conference on Water Chemistry of Nuclear Reactor Systems, BNES, Vol., 1 pp. 266-268, 1996.

[2] Schulze, H., “Antimontrisulfid in wasseriger Losung”, J. Prakt. Chem. 27 320-332 (1883).

[3] Bloom, M.C., “The Mechanism of the Genesis of Polymorphous Forms”, Amer. Mineralog. 24,281-292 (1939).

[4] Tourky, A.R., Mousa, A.A., “Studies on Some Metal Electrodes. Part V. The Amphoteric Properties of Antimony Tri- and Pent-oxide”, J. Chem. Sot. 759-763 (1948).

[S] Gayer, K.H., Garrett, A.B., “Equilibria of Antimonous Oxide (Rhombic) in Dilute Solutions of and Sodium Hydroxide at 25”“, J. Amer. Chem. Sot. 74, 2353-2354 (1952).

[6] Vasil’ev, V.P, Shorokhova, V.I., “Determination of the Thermodynamic Characteristics of Antimony(III) in Alkaline Solutions by a Solubility Method”, Russ. J. Inorg. Chem., l8, 161-164 (1973). (English translation from Zh. Neorg. Khim. 18 305-310 (1973)). 32

171 Popova, M.Ya., Khodakovsky, I.L., Ozerova, N.A., “Experimental Determination of the Thermodynamical Properties of Hydroxo- and Hydroxofluoride Complexes of Antimony at Temperatures up to 2OO”C”, Geokhimiya (6) 835-843 (1975).

PI Belinskaya, F.A., Militsina, E.A. “Inorganic Ion-exchange Materials Based on Insoluble Antimony(V) Compounds”, Russ. Chem. Rev. 49,933-952 (1980); (Eng. transl. from Usp. Khim. 49,1904-1936 (1980)).

[91 Pitman, A.L., Pourbaix, M., de Zoubov, N., “Potential-pH Diagram of the Antimony-Water System”, J. Electrochem. Sot. @4,594-600 (1957).

[lo] Baes Jr., C.F., Mesmer, R.E., “The Hydrolysis of Cations”, Wiley-Interscience, New York, 1976.

[ 1 l] Souchay, P., Peschanski, D., “Contribution ?t l’etude des phenomenes de condensation en chimie minerale. III. Emploi des mesures d’absorption lumineuse”, Bull. Sot. Chim. France 439-446 ( 1948).

[ 121 Ricca, B., D’ Amore, G., Bellomo, A., “Ricerche sugli antimoniati acidi”, Ann. Chim. (Rome) 491-507 (1956).

[13] Jander, G., Ostmann, H.-J., “Messungen der Lichtabsorption und des Diffusionsvermiigens von Antimonat(V)-liisungen bei verschiedenen Wasserstoffionen-Konzentrationen”, Z. anorg. allg. Chem. m,241-249, ( 1962).

[ 141 Lefebvre, J., Maria, H., “Etude des equilibres dans les solutions recente de polyantimoniates”, Comptes Rendu a,3 12 l-3 124 ( 1963).

[ 151 Nakano, H., Ozawa, Y., Yagasaki, A., “Synthesis and Characterization of the First Polyantimonate, [SbsOi2(OH)2&“, J. Am. Chem. Sot. m,12007-12008 (1995).

[ 161 Prasad, S., “Electrometric Investigations on the Formation of Different Antimonate Anions as a Function of pH”, Bull. Electrochem. 6,163-165 (1990).

[ 171 Stewart, D.J., Knop, O., “Pyrochlores. VI. Preparative Chemistry of Sodium and Silver Antimonates and Related Compounds”, Can. J. Chem. 48, 1323-1332 (1970).

[ 181 Stewart, D.J., Knop, O., Ayasse, C., Woodhams, F.W.D. “Pyrochlores. VII. The Oxides of Antimony: an X-Ray and Miissbauer Study”, Can. J. Chem. 50,690-700 (1972).

[19] Rumpel, H., Brendt, W., Adam, K., Schwarzmann, E., “Neue Oxid- und Hydroxid-Phasen des ftinfwertigen Antimons”, Z. Naturforsch. m, 39-42 (1978).

[20] Jansen, M., “Darstellung und Kristallstruktur von Sb205.3/5H20 = SbsO&OH).H&“, Rev. Chim. min. lS,242-247 (1978). 33

[2 l] Schrewelius, N., “Riintgentersuchung der Verbindung NaSb(OH)e, NaSbF6, NaSbOs und gleichartiger Stoffe”, Z. anorg. allg. Chem. a,241-254 (1938).

[22] Beintema, J., “Die Struktur einiger krystallisierten Antimonate”, Recu. Trav. Chim. Pays- Bas 56,93 l-967 (1939).

[23] Asai, T., “Refinement of the Crystal Structure of Sodium Hexahydroxoantimonate(V), NaSb(OH)<’ Bull. Chem. Sot. Japan, 48,2677-2679 (1975).

[24] Dravotslj, M., Karlizek, J., “Lithium and Sodium Antimonates”, Chem. Zvesti 35, 629-640 (1980).

[25] Dravotsky, M., Karl&k, J., “Conditions for the Existence of Potassium Antimonates with a M(I):Sb Ratio Greater Than One”, Coll. Czech. Chem. Corm-nun. 46,3074-3082 (198 1).

[26] Lisichkin, I.N., Kerimbekov, A.V., Kerimbekova, N.A., Manashirov, O.Ya., “Physicochemical Properties of Potassium and Sodium Antimonates”, Russ. J. Inorg. Chem. 16 1138-1140 (1971).

[27] Balicheva, T.G., Roi, N.I., “Determination of the Composition and Structure of Potassium Monoantimonate and its Condensation Products by Spectroscopy and Thermal Analysis”, Russ. J. Inorg. Chem. 17 1556-1559 (1972).

[28] v. Knorre, G., Olschewsky, P., “Betrage zur Kenntniss der Verbindung der AntimonsWre mit Kalium und Natrium”, Chem. Ber. l&2353-2362 (1885).

[29] Montmory, M.C., Durif-Varambon, Pare, X., “Transformation <

[30] Baetsle, L.H. and Huys, D., “Structure and Ion-exchange Characteristics of Polyantimonic Acid”, J. Inorg. Nucl. Chem. 30,639-649 (1968).

[3 l] Abe, M., “Synthetic Inorganic Ion-exchange Materials - XVIII. Ion-exchange Equilibria of Crystalline Antimonic(V) Acid with Alkali Metal Ions”, J. Inorg. Nucl. Chem. &I, 85-89 (1979).

[32] Bauer, A.A., Kleshchev, D.G., Burmistrov, V.A., Konev, V.N., Kleshchev, G.V., “X-ray Diffraction Investigations of M, H(HzO)-forms of Hydrated (M = Li, Na, K)“, Russ. J. Inorg. Chem. 29, 1412-1415 (1984).

[33] Gol’dshtein, L.M., Kalinichenko, I.I., Sheimkman, AI., Trofimov, V.G., Structure of Solid Solutions in the System Ti02-Sb203, and the Influence of Sodium, present as an Impurity, on its Phase Formation, Inorg. Mater. l0, 150-152 (1974) (Eng. transl. from Izv. Akad. Nauk, SSSR, Neorg. Mater. l0, 176-178). 34

[34] Glixelli, S. and Przyszczypkowski, A., “The Solubility of Antimony Pentoxide in Water”, Rocz. Chem. l4,474-485 (1934). CA 2424 (1935).

[35] Beilstein, F., v. Blase, O., “Untersuchungen tiber die Basixitit der Antimons&re”, Chem. Central. 60,803 (1889).

[36] Tomula, ES., “iiber die Antimons&rren und die analytische Anwendung des Natriumantimonats”, Z. anorg. allg. Chem. m,81-92 (1921).

[37] Vasil’eva, L.A. Tr. Kazan. Khim. Tekh. Inst. 2,41-49 (1934).

[38] Urazov, G.G., Lipshits, B.M., Lovchikov, V.S., “Solubility Isotherms of the System Na20-H20-Sb205 at 25” and 75”C”, Russ. J. Inorg. Chem. &198-201 (1959).

[39] Blandamer, M.J., Burgess, J., Peacock, R.D., “Solubility of Sodium Hexahydroxoantimonate in Water and in Mixed Aqueous Solvents”, J. Chem. Sot. Dalton, 1084- 1086 (1974).

[40] Williams, S., “Mopungite - a new Form”, Mineral. Rec. 16,73 (1985).

[41] Naumov, G.B., Ryzhenko, B.N., Khodakovsky, I.L., “Handbook of Thermodynamic Data”, 1971, translated from the Russian by G.J. Soleimani, edited by I. Barnes and V. Speltz, Geological Survey, report USGS-WRD-74-001 (1974). Available from National Technical Information Service, United States Department of Commerce, Springfield VA 2215 1, as document PB-226722.

[42] Wagman, D.D., Evans, W.H., Parker, V.B., Schumm, R.H., Halow, I., Bailey, S.M., Churney, K.L., Nuttall, R.L., “The NBS Tables of Chemical Thermodynamic Properties. Selected Values for Inorganic and Ci and C2 Organic Substances in SI Units”, J. Phys. Chem. Ref. Data 11, supplement No.2 (1982).

[43] Barin, I., “Thermochemical Data of Pure Substances”, 3rd edn., VCH, Weinheim, 1995.

[44] Robinson, R.A., Stokes, R.H., “Electrolyte Solutions”, 2nd edn. 5th impress. rev., Butterworths, London, 1970.

[45] Davies, C.W., “Ion Association”, Butterworths, London, 1962.

[46] Grenthe, I., Fuger, J., Konings, R.J.M., Lemire, R.J., Muller, A.B., Nguyen-Trung, C., Wanner, H., “Chemical Thermodynamics of Uranium”, North Holland, Amsterdam, 1992.

[47] Vi& B.W., “Stability Relations of Antimony and Compounds in the light of Revised and Extended Eh-pH Diagrams”, Chem. Geol. m,21-30 (1996).

[48] Thomsen, J., “Thermochemische Untersuchungen”, Barth, Leipzig (1882-1886). 35

[49] Past, V., “Antimony” in “Standard Potentials in Aqueous Solution”, Ed: Bard, A.J., Parsons, R, Jordan, J., Marcel Dekker Inc. New York, 1985, pp. 172-179.

[50] Vasil’ev, V.P, Shorokhova, V.I., “Determination of the Standard Thermodynamic Characteristics of the Antimony1 ion SbO+ and Antimony Oxide by a Potentiometric Method”, Sov. Electrochem. 8,178- 183 (1972) (Eng. transl. from Electrokhim. 8, 185190).

[5 l] Vasil’ev, V.P, Shorokhova, V.I., Kovanova, S.V., “Potentiometric Investigation of Alkaline Solutions of Antimony(III)“, Sov. Electrochem. 2,953-957 (1973) (Eng. transl. from Electrokhim. 9, 1006-1011).

[52] Pankratz, L.B., “Thermodynamic Properties of Elements and Oxides”, Bulletin 672, Bureau of Mines, United States Government Office, Washington D.C. (1982).

[53] Kubaschewski, O., Evans, E.L., Alcock, C.B., “Metallurgical Thermochemistry”, Pergamon Press, London, 1967.

[54] Glushko, V.P., Medvedev, V.A. et al., ed., “Termicheskie Konstanty Veshchestv”. Moscow: Nauka, 1968 (Part 3), as cited by Barin et al. [58].

[55] Knacke, O., Kubaschewski, O., Hesselmann, K. “Thermochemical Properties of Inorganic Substances”, Vol. I, II, Springer-Verlag, Berlin, 199 1.

[56] Kubaschewski, O., Alcock, C.B., Spencer, P.J., “Materials Thermochemistry”, Pergamon Press, Oxford, 6th edn., revised, 1993.

[57] Bar-in, I., Knacke, O., “Thermochemical Properties of Inorganic Substances”, Springer, Berlin, 1973.

[58] Barin, I., Knacke, O., Kubaschewski, O., “Thermochemical Properties of Inorganic Substances. Supplement”, Springer-Verlag, Berlin (1977).

[59] Gorgoraki, E.A., Tarasov, V.V., “Low Temperature Heat Capacity and Certain Thermochemical Data on the of As and Sb”, Tr. Mosk. Khim.-Tekhnol. Inst. No. 49, 1 l-15(1965).

[60] Anderson, C.T., “The Heat Capacities at Low Temperatures of Antimony, , and Antimony Pentoxide”, J. Am. Chem. Sot. 52, 2712-2720 (1930).

[61] White, W.R., Dachille, F., Roy, R., “High-pressure polymorphism of A~203 and Sb203”, Z. Kristallog. m,450-458 (1967). 36

[62] Chang, S.S., Bestul, A.B., “Heat Capacities of Cubic, Monoclinic, and Vitreous Arsenious Oxide from 5 to 350 K”, J. Chem. Phys. 55,933-946 (197 1).

[63] Behrens, R.G., Rosenblatt, G.M., “Vapor Pressure and Thermodynamics of Orthorhombic Antimony Trioxide (valentinite)“, J. Chem. Thermodynam. &173-188 (1973).

[64] Cox, I.D., Wagman, D.D., Medvedev, V.A., “CODATA Key Values for Thermodynamics”, Hemisphere Publishing Corp., New York., 1988.

[65] Mixter, W.G., “The heat of formation of trisodium orthophosphate, trisodium orthoarsenate, the oxides of antimony, trioxide; and fourth paper on the heat of combination of acidic oxides with ”, Am. J. Sci. 28, 103-l 11 (1909).

[66] Rossini, F.D., Wagman, D.D., Evans, W.H., Levine, S., Jaffe, I., “Selected Values of Chemical Thermodynamic Properties”, U.S. National Bureau of Standards, Circular 500, Washington, D.C. (1950).

[67] Wagman, D.D., Evans, W.H., Parker, V.B., Schumm, R.H., Nuttall, R.L., “Selected Values of Chemical Thermodynamic Properties. Compounds of Uranium, , , , and the Alkali Metals”, U.S. National Bureau of Standards, Technical Note 270-8 (1981).

[68] Kasenova, Sh.B., Kasenov, B.K., Mustafin, ES., Aldabergenov, M.K., “Thermodynamic Properties of Sodium Orthoantimonate NasSbO4 and the SbOi- Ion in Standard Aqueous Solution”, Russ. J. Inorg. Chem. 40, 1614-1616 (1995).

[69] Kasenova, Sh.B., Kasenov, B.K., Mustafin, E.S., “Heat Capacity and Thermodynamic Functions of MSbOs (M-Na, K, Cs) in the Temperature Range 298.15-673 K”, High Temperature 34,481-483 (1996) (Eng. transl. of Teplo. Vys. Temp. %,485-486 (1996)).

[70] Kasenov, B.K., Zhakibaev, B.K., Kasenova, Sh. B., “Evaluation of the Thermodynamic Properties of Alkali Metal Antimonates and of Gas-Phase Ions SbO; , SbzO; -, SbsO& , and SbO$“, Russ. J. Phys. Chem. 67,2230-223 1 (1993).

[7 l] Glushko, V.P., ed., “Termicheskie Konstanty Veshchestv”. Moscow: Nauka, 198 1. (Vol. 10, Part l), as cited by Kasenov et al. [70]. 37

Appendix A: Literature Thermodvnamic Data for Aaueous Antimonv Snecies and Selected Oxide Solids

A.1 Simnle Aaueous Ions and Hvdrolvsis Snecies of Antimony

Tabulated values from the literature for simple aqueous ions and hydrolysis species of antimony are listed in Table Al. A value for one additional species should be mentioned: AfH’(HSb(OH)b) = -1478.6 kJmo1“ [41,42] is traceable to Thomsen [48]. In all cases, proposed species differing by integral multiples of H20 have been treated as identical, and reported chemical thermodynamic values have been adjusted using the values for H20(1). Except for SbO+, the values for the solution species from Popova et al. [7] were based on their own high- temperature solubility results. Values for SbO+ and SbO as assessed by Past [49] were based on the electrochemical studies of Vasil’ev et al. [50,51].

Table A 1: Literature Tabulations of Chemical Thermodynamic Values for Antimony Aqueous Species

A&?VlcJ-mo~’ at 25°C So/J-C’moP at 25°C Species /ref. 191 1421* 1411 171 [491 [421* 1411 171 r491 SbO’ -175.7 -177.11 -179.6 -175.8 -175.64 -7.1 22.33 Sb(OH)3. -645.1 -644.7 -647.3 -644.7 116.3 125.5 192.9 SbOi -345.2 -340.19 -342.9 -339.5 -339.74 56.8 55.2

!%O; -274.1 * The year of compilation was 1964.

A.2 Antimonv(III) and Antimony(V) Oxide Solids

Tabulated values from the literature for simple oxides of antimony (III) and antimony(V) are listed in Tables A2-A6.

Table A2: Literature Tabulations of Gibbs Energy of Formation Values for Antimony(III) and Antimony(V) Oxides at 25°C

&“/kJ-mol’l Compound / ref. r91 1421* [411 [71 WI r491 WI Sb203, cubic -623.4 -634.4 -641.0 -626.8 -632.2 -634.3 Sb203, ortho. -615.0 -626.5 -631.8 -624.7 -626.6 -624.7 -626.3 sb20, -838.9 -829.2 -864.7 -829.2 -829.3 -829.1 * The year of compilation was 1964. 38

Table A3: Literature Tabulations of Enthalpy of Formation Values for Antitnony(III) and Antimony(V) Oxides at 25°C

A&P/kJmol-’ Compound / ref. [421* [531? Kw 1411 r71 1521 [491 WI WI 1431 Sbz03, cubic -720.3 -708.85 -715.5 -720.4 -709.4 -711.6 -716.1 -720.3 Sbz03, ortho. -708.S5 -708.8 -706.9 -708.8 -701.6 -708.8 -708.6 sbzo5 -971.9 -971.9 -1007.5 -1007.5 -971.9 -971.9 -993.7 -1007.5 -971.9 * The year of compilation was 1964. t As reported by Barin and Knacke [57]. $ As reported by Barin et al. [58]. 0 Probably the value for the orthorhombic form.

Table A4: Literature Tabulations of Entropy Values for Antimony Oxides at 25°C

S”/JdlllOl-l compolmd I ref. [42]* [541T [41] VI [521 [49] [SSI [Sal 1431 Sb&, cubic llO.& 132.6 132.4 122.2 132.6 132.7 f 4.2 110.45 SbzO3, ortho. 123.0 141.0 123.0 123.0 141.0 123.0 134.6 Sb2@ 125.1 125.1 125.1 124.9 125.1 124.9 125.1 i 8.4 * The year of compilation was 1964. t As reported by Barin et al. [58].

Table A5: Literature Tabulations of Heat Capacity Values for Antimony Oxides at 25°C

Cg/JKbd-’ Compound I ref. [421* [541? [41] rlw [521 [55.l [561 1431 Sb203, cubic lll.S$ 104.6 x 104.4 111.85 Sb&. ortho. 101.38 111.9 x+7.3 101.38 101.38 111.8 101.4

The year of compilation was 1964. As reported by Barin et al. [SS]. Probably the value for the orthorhombic form from Gorgoraki and Tarasov [59]. The problem may only be in the tables of Barin [43] and Barin et al. [58], and not in the tables of Glushko et al. [54]. The latter set of tables was not available to the authors of the present report. Based on the authors’ equation for the temperature dependence of the heat capacity of transformation [7]. A value for Ci(Sb203, orthorhombic) is not given explicitly. 39

Table A6: Temperature-Dependent Heat-Capacity Values for Antimony Oxides

Compound @JK’mor’ = a +lO%(T/K) + lO%/(T/I# Reference b C SbzO3. cubic 7:9 71.5 1411 t 75.31 97.49 1561 92.048 66.107 [431 Sb& ortho. 114.01 8.318 -13.435 [43,521 t Sbzos 141.33 -3.732 -20.112 1521 t From the same data [60]; probably the sample used for the measurement was primarily the orthorhombic form.

For Sb203, two solids exist-the orthorhombic valentinite and the cubic senarmontite [61]. Although the cubic form is more stable at low temperatures, the orthorhombic form is metastable over a wide temperature range, and most of the earlier chemical thermodynamic measurements reported in the literature for Sb203(s) have been carried out using the orthorhombic form of the solid. Anderson [60] measured the heat capacity of the orthorhombic form from -2 13.4 to 21.2”C. Some compilers [41,49,54,56] have preferred the heat capacity values for both forms of Sb203(s) as measured between -208 and 27°C by Gorgon&i and Tarasov [59]. This work showed that the values of C”,(Sb203, orthorhombic) are (7&2)% greater than C;(Sb203, cubic) throughout the temperature range of the measurements. The values of the latter authors for the orthorhombic solid are 58% greater than those of Anderson. However, based on results for arsenic oxides, Chang and Best& [62] have suggested that there may have been a systematic problem in results of Gorgon&i and Tarasov. If the heat capacities of Gorgoraki and Tarasov [59] are incorrect [63], the entropy values derived from the measurements are also likely to be incorrect.

Assuming that the compounds were correctly prepared and characterized, Gorgoraki and Tarasov [59] showed that Ci(Sb203, cubic) is systematically less than Ci(Sb203, orthorhombic) by 7.2 J K-‘.mol-’ near 25°C; also, S”zoc is 8.6 J.KW1.mol-’ less for the cubic form than for the orthorhombic form Although there are doubts as to the absolute values for the heat capacities and derived entropy values, the differences should be approximately correct. Therefore, the values of C;(T) and S”(25”C) for the orthorhombic form, based on the measurements of Anderson [60], are accepted, and the values Ci(Sb203, cubic) = (94 f 10) J-K-‘.mol-’ and S’(Sb203, cubic) = (114 & 10) J.R’.mol-’ can be selected (the uncertainties are estimates). From these values, and the assumption that the enthalpy of transition at 606°C is 4.184 kJ.mol-’ (i.e., 1.0 kcal.mol-’ [61]), the consistent function for C,(Sb203, cubic, T) is:

Cr(Sb203, cubic, T/K)/J.K-‘.mol“ = 66.296 + 0.092918T

In the absence of a recent, detailed analysis of literature values for A$I“(Sb203(s), 25”C), or recent experimental values for either the orthorhombic or cubic form, the value AfH’(Sb203, orthorhombic, 25°C)) = -708.55 kJ.mol-‘, selected by the U.S. National Bureau of Standards [42], is used. From this, the transition enthalpy and the selected expressions for Ci(Sb203(s), T), S’(Sb203, cubic, 25°C) and CODATA consistent auxiliary data [64], 40

AG’(Sb203, cubic, 25°C) = -632.08 kJ.mol-‘. The value of AG”(Sb20s, orthorhombic, 25°C) is -626.39 kJ.mol-* and the difference in the Gibbs energies of formation between the two forms is 5.7 kJ.mof’.

It is probable that the synthesis method used by Mixter [65] to prepare anhydrous Sb205 yielded a partially reduced solid [ 181, and the products of the combustion reactions were not well characterized. Anderson [60] did his low-temperature heat-capacity measurements using samples of hydrated Sb205 mixed with a lower oxide (presumed to be Sb204). After allowing for the lower oxide, Anderson reported Ci(Sb205.0.3 17H20) = 3 1.10 cal.K“.mol-’ (130.1 J~R’~rnol’ ) and (Sb20ss2.224H20) = 52.03 cal.K-l.mol‘l (217.7 J.K’.mol-‘) for 17°C. All tabulated values near 118 J.K-‘.mol-’ for Ci(Sb205) are derived from these measurements. The values for the heat capacity of Sb205 in at least two compilations [54,56] (original source not known to the present reviewers) are greater than that measured by Anderson [60] for a slightly hydrated oxide. Most compilers have proposed approximately the same value for S”(Sb205,25’C) based on Anderson’s work [60], even if the compiled heat capacity is quite different. All chemical thermodynamic values for Sb205 for temperatures greater than 25’C are extrapolated. There are substantial questions with respect to all of the primary experimental chemical thermodynamic data for Sb205.

A.3 Chemical Thermodynamic Measurements for Mixed Oxides Containing Antimony

Calorimetric data are extremely limited for the Sb(V) salts and/or mixed oxides. Mixter [65] reported a value for the enthalpy of formation of NasSb04 based on a determination of the heat of reaction of “pulverized antimony and sodium ”. A value based on this heat of reaction, AfH”(NasSb04,25”C) = -352 kcalmol’ (-1473 kJ.mol-‘), was recalculated for the U.S. National Bureau of Standards Circular 500 [66], but was dropped from subsequent editions of these tables [42,67]. This presumably was done because the reviewers decided that the value was unreliable. Certainly, Mixter’s reaction product was not well characterized. Kasenova et al. [68,69] determined the heat capacity (298.15-4OO”C) of the same compound and of a series of MSb03 solids. Their values are summarized in Table A7.

Table A7: Temperature-Dependent Heat-Capacity Values for Alkali Metal/Antimony(V) Mixed Oxides

Compound C~JK’*moK’ = A +lO%(T/K) + 10%!/(T/K)2 Reference A B C NaSbOJ (65 f 4) (156rt9) -(4.8 f 0.4) if391 Na$bO, (209 f 17) (9.9 f 0.8) -(73.1 f 5.8) I681 (132 f 8) (7.0 f 0.4) (35.6 f 2.2) [691 (183 f 13) (20.8 f 1.5) (76.8 & 5.4) 1691

The enthalpy of solution of Na3Sb04 in water was also reported [68] as being (18.2 f 0.4) kJ.mol-‘. However, the nature of the aqueous antimony species so formed (assumed by the authors to be SbO$(aq)) was not established. The entropy of NajSbO4, and chemical thermodynamic values for a wide variety of other anhydrous alkali metal antimonates, 41 were recently estimated by Kasenov et al. [70]. The enthalpy values are relative to Afw(NasSb04) = -1485.3 kJmol-’ from a set of standard tables [71] unavailable to the present reviewers. However, it is probable that this key value was also based on a (different) recalculation of Mixter’s measurements [65] discussed above. The entropies were estimated from values for arsenates and phosphates. Thus, there appear to be no reliable directly determined enthalpy of formation or entropy values for any of these solids, nor for the Sb(V) aqueous species. AEXL-12064

ISSN 0067-0367 To identify individual documents in the series, we have assigned an AECL-number to each. Please refer to the AECL-number when requesting additional copies of this document from: Information Centre, Inforrnation Management Branch AECL Chalk River, Ontario Canada KOJ 1 JO Fax: (613) 584-8144 Tel.: (613) 584-3311 ext. 4623

Pour identifier les rapports individuels faisant partie de cette series, nous avons affect& un nun&o AECL-a chacun d’eux. Veuillex indiquer le nun&o AECL-lorsque vous demandez d’autres exemplaires de ce rapport au: Service de Distribuition des Documents Officiels EACL Chalk River (Ontario) Canada KOJ 1 JO Fax: (613) 584-8144 Tel.: (613) 584-3311 poste 4623

Copyright 0 Atomic Energy of Canada Limited, 2000