Radiochim. Acta 88, 297Ϫ306 (2000) by Oldenbourg Wissenschaftsverlag, München
Thermodynamic model for the solubility of thorium dioxide + – – ° ° in the Na -Cl -OH -H2O system at 23 C and 90 C
By Dhanpat Rai1,*, Dean A. Moore1, Charles S. Oakes1 and Mikazu Yui2
1 Pacific Northwest National Laboratory, Richland, Washington 99352, USA 2 Japan Nuclear Cycle Development Institute, Tokai Works, Tokai, Japan
(Received August 30, 1999; accepted in revised form March 8, 2000)
Solubility / Thermodynamics / Thorium dioxide / ThO2(c) / at room temperature. Therefore, potential thorium concen-
ThO2(am) / Temperature effects on ThO2 crystallinity / trations in aqueous solutions contacting failed repositories Solubility product / Solubility products at different may be many orders of magnitude lower than indicated by temperatures measurements of ThO2(am) solubility at 25 °C. At 25 °C, the precipitation/dissolution kinetics of Summary. Data are extremely limited on the effects of tem- ThO2(c) are not known, which makes it difficult to accu- perature on crystallinity and the resulting changes in solubility rately ascertain the solubility product of ThO2(c) at this products of thermally transformed thorium oxide phases. Such temperature. Fortunately, in addition to being relevant to data are required to reliably predict thorium behavior in high- expected repository conditions, a moderate increase in tem- level waste repositories where higher than ambient tempera- perature accelerates the precipitation/dissolution kinetics tures are expected. Solubility studies were conducted as a func- such that the solubility of ThO2(c) can be approached from tion of pH and time and at 0.1 M NaCl for 1) ThO2(am) at both the oversaturation and undersaturation directions. ° 23 C, 2) ThO2(am c), a thermally transformed amorphous → ° ° Therefore, it is possible to study the effect of temperature thorium hydrous oxide at 90 C, and 3) ThO2(c) at 23 C and ° ° on the crystallinity of thorium hydrous oxides as well as 90 C. Results show that when ThO2(am) is heated to 90 C, it transforms to a relatively insoluble and crystalline solid the solubility products of these solid phases at elevated temperatures. While the solubility product of ThO2(c) can [ThO2(am c)]. At a fixed pH, the observed solubility of → ° be calculated (see Appendix A), experimental verification ThO2(am) at 23 C is more than 11 orders of magnitude greater ° 1 than those for ThO2(c) at 23 C or of ThO2(am c) and is required due to uncertainties in the thermodynamic ° → 0 4ϩ 0 ThO2(c) at 90 C. Solubility data were interpreted using the quantities for ThO2(c), Cp for Th , and the absence of Cp Pitzer ion-interaction model. The log of the solubility product values for aqueous Th4ϩ at temperatures Ͼ55 °C. Thus, the for the thorium dioxide dissolution reaction [ThO2(s) ϩ 2H2O primary objectives of this study were to demonstrate the 4ϩ Ϫ
Thorium hydrous oxide [ThO2(am)] is generally used to Stock solutions were prepared from reagent grade Th(NO3)4 estimate leachable thorium concentrations under waste re- ·4H2O(c), concentrated NaOH(aq) (Anachemia), concen- pository conditions (e.g., JNC [1]). The temperatures in trated reagent grade HCl(aq) (GFS), and reagent grade high-level nuclear waste repositories are expected to be NaCl(c) (Alfa Products). considerably higher than ambient and may reach 00°C. 1 A 100 g per liter solution of thorium was prepared in Under these conditions, predictions of leachable thorium 0.1 M HNO (aq) using the reagent grade Th(NO ) ·4HO. require a solubility model of the appropriate thorium oxide 3 3 4 2 Stock solutions of 5.0 M NaOH and 1.0 M NaCl were pre- phase as a function of temperature. Data are extremely lim- pared using the reagents mentioned above. Deionized dis- ited on the effects of temperature on crystallinity and re- tilled water was used in all solution preparations. sulting changes in the solubility products of thermally 1 0 4ϩ transformed phases. Prasad et al. [2] showed that the trans- For example, the reported Cp values for Th vary considerably. ° Values at 298 K are Ϫ60 J · molϪ1 ·KϪ1 based on Morss and McCue’s formation of ThO2(am) to crystalline ThO2 is slow at 25 C [5] data as recalculated by Hovey [6] and Ϫ224 J · molϪ1 ·KϪ1 report- (about 270 days) and much more rapid (about 12 days) Ϫ1 Ϫ1 ° ed by Hovey [6]; 111 and Ϫ228 J · mol ·K at 303 K reported by at 100 C. The solubility product (Ksp) of crystalline ThO2 Apelblat and Sahar [7] and Hovey [6], respectively. Hovey’s data are [ThO2(c)] has been shown to be approximately eight orders most recent and appear more reliable and thus were selected for calcu- of magnitude lower than that for the hydrous oxide [3, 4] lations reported here. However, Hovey’s heat capacity measurements extend to only 55°C; thus, extrapolation to higher temperatures is de- * Author for correspondence (E-mail: [email protected]). pendent upon the reliability of the extrapolation method. 298 D. Rai, D. A. Moore, C. S. Oakes et al.
Table 1. Comparison of filtered and unfiltered aqueous thorium con- Aliquots of suspensions equilibrated at 23 or 90°C were ° centrations from solutions maintained at 90 C. centrifuged at 5000 rpm for 10to15 minutes filtered
Ϫ1 a through Amicon Centricon-30 filters (30,000 MW cutoff), pH log (CTh/mol · kg ) acidified, and then analyzed for thorium. Thorium concen- filtered unfiltered trations in unfiltered but centrifuged and filtered samples were similar, indicating that a significant amount of col- 2.51 Ϫ5.52 Ϫ5.50 2.57 Ϫ5.25 Ϫ5.29 loidal material was absent and that the results from the un- 2.69 Ϫ5.46 Ϫ5.48 filtered samples are reliable (Table 1). This is particularly 2.78 Ϫ5.48 Ϫ5.48 significant to the experiments at 90 °C where available fil- ters may not be stable. a: All solutions were centrifuged at 5000 rpm for 10to15 minutes. For the 90°C experiments, all implements (e.g., pipette Amicon Centricon-30 filters (30,000 MW cutoff) were used. tips, sampling tubes, rinse water, pH buffers, centrifuge and centrifuge containers) that contacted or could change sample temperatures prior to pH measurement or sampling ThO (am) was precipitated from the nitrate solution by 2 were maintained at 90°C. For pH measurements, the elec- adding an aliquot to 10 mL of water and adjusting the pH trode was rinsed with 90 °C water before being placed in to about 10.5 using the NaOH(aq) stock solution. The re- samples. Maximum temperature decrease during pH sulting suspensions were centrifuged and the supernatant measurements was about 5 °C. Subsequently, suspensions discarded. Soluble nitrates, chlorides, and sodium were re- were centrifuged and small aliquots 1) from all suspensions moved by washing twice with 10-mL aliquots of water. were filtered through Amicon filters, equilibrated at 90 °C ThO (c) was prepared by initially dissolving a quantity 2 for 1 hour, and 2) a few selected samples were withdrawn of Th(NO ) ·4HO(c) in hot 0.3 M HNO (aq). Oxalic acid 3 4 2 3 for analysis. The filtered and unfiltered aliquots were then dihydrate was then vigorously stirred into the solution at analyzed by ICP-MS. The analytical data obtained in this 75 °C. The resulting precipitate and solution were then al- study are listed in Appendix Tables B.1 through B.4. lowed to rest for 30 minutes between 50 and 75 °C and then Three sets of experiments ranging in pH and amount of filtered through a coarse sintered glass filter. ThO (c) was 2 ThO (am) were conducted at 23 °C (Table 2). All three then produced by firing the oxalate in air for 2 hours at 2 sample sets were subsequently heated to 90 °C. Heating of 750 °C. X-ray diffraction (XRD) analysis confirmed the ThO2(am) suspensions was expected [2] to result in increas- solid to be crystalline ThO2. ing crystallinity and decreasing solubility [3] due to the
formation of crystalline ThO2. For example, heating 4ϩ Equilibration, sampling, and analysis ThO2(am) suspensions in equilibrium with soluble Th or
the samples in which the ThO2(am) dissolves completely, All experiments were conducted on a benchtop at room will result in the formation of crystalline ThO either temperature (23Ϯ 2°C) 2 or at 90°C. A heating block was 2 ° through solid-solid transformation (Eq. (1)) or by precipi- used for the experiments at 90 C. For each experiment, tation (Eq. (2)). The precipitation mechanism (Eq. (2)) will several milligrams of ThO2(am) or ThO2(c) were loaded also be accompanied by an increase in Hϩ concentration in into a number of 50-mL polypropylene centrifuge tubes proportion to the amount of precipitated thorium. containing approximately 38 mL of 0.1 M NaCl(aq). The pH of the suspensions was adjusted using HCl, and the ThO2(am) < ThO2(c) (1) tubes were then sealed with Nalgene caps. The amount of 4ϩ ϩ Th ϩ 2H2O Table 2. Listing of experiments conducted in this study. Set Solid pH range Equilibration Temperature numbera phase periods (days) (°C) Set I ThO2(am) 2.0 to 4.7 6, 21 23 b ThO2(am c) 1.5 to 2.2 76 90 → Set II ThO2(am) 4.2 to 5.1 5, 1323 b ThO2(am c) 2.0 to 2.1 52 90 → Set III ThO2(am) 1.9 to 4.2 7, 22 23 b ThO2(am c) 1.9 to 3.0 76 90 → c d Set IV ThO2(c) 1.3 to 3.6 11, 87, 687, 794 23 c Set V ThO2(c) 1.4 to 3.5 1290 a: The total approximate amounts of thorium solid phase used in each sample from different sets were 115 mg for Sets I and II, 5 mg in Set III, and 10 to 500 mg in Sets IV and V (the largest amounts of solid were used in low pH regions of Sets IV and V). ° b: This solid phase resulted from heating the ThO2(am) to 90 C. c: Produced by firing thorium oxalate at 750°C, see text for details. d: The 794-day analyses correspond to samples that were equilibrated at 23°C for 779 days and maintained at 90°C for 14 days and then readjusted to 23°C for one day. that steady-state concentrations are reached in Ͻ5 days. indicating that steady-state concentrations are reached by X-ray diffraction analyses indicate that the equilibrated sol- 687 days. At Ͻ687 days, aqueous thorium concentrations id phases are amorphous (Fig. 2). Previous studies [8Ϫ10] appear to be nearly independent of pH between pH values on the solubility of ThO2(am) have shown that equilibrium of 2 and 3. A continuous decrease in thorium concen- with these precipitates is reached within a few days. There- trations is expected under these acidic conditions with in- fore, the results obtained in this study, in conjunction with creasing pH. The reasons are not known for the observed data available in the literature, can be used to interpret these lack of dependence of thorium concentrations on pH in the data for equilibrium with ThO2(am). middle pH range; however, this behavior may be due to the ° The solubility of ThO2(c) as a function of pH at 23 C presence of a finite amount of a less crystalline material was found to be approximately 9 to 12 orders of magnitude (estimated from these experiments to be about one percent). lower than the solubility of ThO2(am) (Fig. 3). Aqueous thorium concentrations increased slightly with an increase in the equilibration period from 11 days to 687 days, but were similar to each other at longer equilibration periods, ° Fig. 1. Experimental ThO2(am) solubilities in 0.1 M NaCl at 23 C Fig. 2. X-ray diffraction patterns for the solid ThO2 phases examined plotted as the logarithm of total thorium versus measured pH. The in this study [(am) ϭ amorphous, (am c) ϭ transformation of → dotted line represents predicted solubility of ThO2(am) using the amorphous phase when heated to 90 °C, (c) ϭ fully crystalline phase modeling parameters reported in Tables 3 and 4. prepared by firing thorium oxalate at 750 °C]. 300 D. Rai, D. A. Moore, C. S. Oakes et al. Fig. 3. Solubility of ThO2(c) at 23°C and different equilibration peri- Fig. 4. Experimental ThO2(am c) and ThO2(c) solubilities in 0.1 M ° → ods. The 794 day data represent data from ThO2(c) suspensions that NaCl at 90 C plotted as the logarithm of total thorium versus mea- were equilibrated for 779 days at 23°C, maintained at 90°C for 14 sured pH. ThO2(am c) data are for sample Sets I to III that were days, and then readjusted to 23°C for one day before sampling. Lines originally equilibrated→ at 23°C and then equilibrated at 90 °C (for represent predicted solubilities of ThO2(am) and ThO2(c) using mode- details see Appendix B). The lines represent predicted solubilities of ling parameters reported in Tables 3 and 4. various thorium solid phases using the modeling parameters reported in Tables 3 and 4. Although not plotted, the ThO2(c) solubility line at ° ° 23 C falls between the ThO2(am(C) and ThO2(c) at 90 C. To confirm this hypothesis, we heated the ThO2(c) suspen- sions, which were previously equilibrated at 23°C for 779 siderably lower than those for identical samples equilibrat- ° days, at 90 C for 14 days; the suspensions were then main- ed at 23 °C (Fig. 4 and Appendix Tables B.1 through B.3). tained at 23 °C for one day and sampled (data represented Thus, at a given pH, the thorium concentration is approxi- by 794-day equilibration period in Fig. 3). This heating and mately 11 orders of magnitude lower at 90 °C than at 23 °C. cooling process did not result in significant changes in ob- However, it must be reiterated that upon heating each served thorium concentrations as a function of pH (Fig. 3; sample both pH and thorium concentration decreased. Table B.4) at pH values Ͻ2. However, the observed tho- These decreases in pH and thorium concentration must re- rium concentrations in the pH region 2 to 3 decreased by sult, as stated previously, from the transformation of up to several orders of magnitude and indicated a more ThO2(am) to a more crystalline form of ThO2. This con- continuous decrease in thorium concentrations with the in- clusion is consistent with XRD analyses of ThO2(am) crease in pH. These results indicate that heating the suspen- samples equilibrated at 23 °C and at 90 °C (Fig. 2). Peaks sions may have crystallized the somewhat less crystalline corresponding to crystalline ThO2 are absent from ThO fraction. Although the aqueous thorium concen- 2 ThO2(am) samples equilibrated at 23 °C, but all of the peaks trations reached steady state at pH values Ͻ 2 at longer ° from ThO2(am) samples heated to 90 C [hereafter referred equilibration periods, it is not known whether the samples to as ThO2(am c)] correspond to crystalline ThO2. reached equilibrium. We believe equilibrium may have At 90 °C and→ any pH, the observed solubilities of the been reached. This hypothesis is supported by the data of originally fully crystalline ThO2 are approximately two or- Prasad et al. [2], wherein ThO2(am) crystallized at Ͼ270 ders of magnitude lower than those for ThO2(am c) days, indicating that precipitation dissolution kinetics, al- → (Fig. 4). Comparison of XRD patterns for ThO2(am c) though not rapid, are certainly achievable at longer equili- → and ThO2(c) show that the ThO2(am c) peaks are broader bration periods. In addition, upon heating, a decrease in → and considerably lower in intensity than those for ThO2(c) thorium concentrations from the 687-day steady-state con- (Fig. 2). It is therefore not surprising to find that the solu- centrations between pH 2 and 3 suggests that for a given bility of ThO2(am c) is greater than the solubility of pH the ThO (c) solubility cannot be higher than that ob- → 2 ThO2(c). Although the solubilities of ThO2(am c) and served at 687 days. Thus, these data can be used to set → ThO2(c) differ somewhat, both phases are many orders of limits for the solubility of ThO (c) at 23°C. 2 magnitude (Ͼ11) less soluble than ThO2(am) at 23 °C. It also should be noted that there is very little scatter in the Thorium dioxide solubility at 90°C data at pH values between 1.0 and 2.2, and that the ob- served thorium concentrations decrease by about four or- At 90 °C, both the measured pH and thorium concentrations ders of magnitude with one unit increase in pH. Somewhat in the suspensions originally containing ThO2(am) are con- more scatter is observed in the data at pH values Ͼ 2.2. The Thermodynamic model for the solubility of thorium dioxide 301 Table 3. Dimensionless standard molar Gibbs energies of formation of Table 4. Pitzer ion-interaction parameters at 1 bar and at 23 and 90°C. different species used in calculations. Binary parametersa Refer- 0 Species ∆Gf /RT Reference ence β(0) β(1) β(2) 103 C For calculations at 25°C 4ϩ ° Th Ϫ284.227 [22] For calculations at 25 C ϩ Ϫ ThO2(am) Ϫ450.08 this study Na ,Cl 0.0754 0.2770 0.700 [24] ϩ Ϫ ThO2(c) ՆϪ477.72 this study Na ,OH 0.0807 0.2982 5.050 [25] ϩ Ϫ H2O Ϫ95.663 [23] H ,Cl 0.1769 0.2973 0.362 [26] OHϪ Ϫ63.435 [23] Th4ϩ,ClϪ 1.0920 13.7 Ϫ160 Ϫ112[18] For calculations at 90°C For calculations at 90°C Th4ϩ Ϫ228.37 see Appendix A Naϩ,ClϪ 0.0992 0.3253 Ϫ1.47 [24] ϩ Ϫ ThO2(am c) Ϫ377.629 this study Na ,OH 0.0868 0.3461 0.08 [25] → ϩ Ϫ ThO2(c) Ϫ382.738 this study (experimental) H ,Cl 0.1536 0.3345 Ϫ1.61 [26] 4ϩ Ϫ ThO2(c) Ϫ383.27 this study Th ,Cl 1.0920 13.7 Ϫ160 Ϫ112[18] (calculated; see Appendix A) b H2O Ϫ75.21 see Appendix A Ternary parameters Reference OHϪ Ϫ46.60 see Appendix A s θHϩ,Naϩ 0.036 [27] s θHϩ,Th4ϩ 0.600 [18] s θNaϩ,Th4ϩ 0.420 [8] s θ Ϫ Ϫ Ϫ0.050 [27] reason for this scatter is not known, but we surmise it may Cl ,OH ψ ϩ ϩ Ϫ Ϫ0.004 [27] be related to the crystallinity of the solid phase. H ,Na ,Cl ψNaϩ,OHϪ,ClϪ Ϫ0.006 [27] ψHϩ,Th4ϩ,ClϪ 0.080 [28] ψ ϩ 4ϩ Ϫ 0.2 0[8] Thermodynamic model Na ,Th ,Cl 1 1 ° ° Rai et al. [8] recently developed a model based on the a: These values are for 23 C and are assumed to apply to the 90 C data (see text for details). Pitzer ion-interaction formalism [11, 12] to reproduce b: These values are of 23°C and are assumed to apply to the 90°C ThO2(am) solubilities in dilute to concentrated NaCl and data. Because the experiments were conducted in relatively dilute ° MgCl2 solutions at 25 C. It is of interest to examine how NaCl solutions (0.1 M), the ternary parameters will not significantly closely that model [8] reproduces the results of this study. affect the calculations. The equilibrium of solid thorium dioxide [ThO2(s)] with the aqueous solution is the primary reaction (Eq. (3)) inves- tigated in this study. solubility in 0.1 M NaCl at 23 °C (Fig. 1). However, the observed and predicted dependence of aqueous thorium 4ϩ Ϫ ThO2(s) ϩ 2H2O ThO2(am), ThO2(am c), or ThO2(c)]. rium was achieved, these data certainly represent the lower → ° Predictions made using the Rai et al. [8] model are about solubility limit of ThO2(c) at 23 C. Thus, the log Ksp value ° 0.6 orders of magnitude lower than the observed ThO2(am) must be ՆϪ56.9 for ThO2(c) at 23 C. 302 D. Rai, D. A. Moore, C. S. Oakes et al. Table 5. Solubility products (Ksp) of different Solid Log K Ionic Reference thorium dioxide solids [ThO2(s) ϩ 2H2O < sp Th4ϩ ϩ 4OHϪ]. strength Values at 25 °C ThO2(am) Ϫ44.9Ϯ 0.5 0 this study, experimental Ϫ44.8 Ͻ0.02 [15] Ϫ45.5 0 [9] Ϫ44.7 a 0[16] Ϫ45.7 a Յ0.1 [17] Ϫ41.1 a 0.1 [29] Ϫ50.76 a 0.1 [30] b ThO2(act) Ϫ48.7 0 [20] ThO2(c) ՆϪ56.9 0 this study, experimental Ϫ54.2Ϯ 1.3 c 0 calculated from Rai et al. [3] Ϫ54.13 c 0 calculated from Wagman et al. [4] Ϫ54.2 c 0 calculatedd from Cox et al. [21], Wagman et al. [4], and CODATA [20] Values at 90 °C ThO (am c) e Ϫ49.2Ϯ 0.2 0 this study, experimental 2 → f ThO2(c) Ϫ51.4Ϯ 0.2 0 this study, experimental Ϫ51.6 0 calculatedd from Cox et al. [21], Wagman et al. [4], and CODATA [20] Ϫ50.8 0 Calculatedd from Cox et al. [21], Wagman et al. [4] and CODATA [20] 0 assuming that ∆Cp,rxn ϭ 0. a: The authors designated the solid phase as Th(OH)4(s), and is more likely an amorphous phase. b: act ϭ a designation given by the authors to represent a microcrystalline phase. c: These calculated Ksp values are similar because of similarities in the original source of the thermodynamic data. d: For details see Appendix A. e: This phase is expected to convert to fully crystalline ThO2(c) at longer equilibration periods and at higher temperatures. The standard deviation is for data Sets I and II. f: The standard deviation in the data is higher at higher pH values, possibly due to the presence of finite amounts of more soluble phase as discussed in text. The reported value is for data at pH Ͻ2.5. To interpret the 90°C data, values at 90 °C are required ported in this paper (Tables 3 and 4) will provide reliable 0 4ϩ Ϫ of ∆Gf /RT for Th ,OH ,H2O, and of ion-interaction pa- predictions for these systems. rameters for chloride and hydroxide with the dominant cat- The solubility products obtained in this study for 4ϩ ϩ 4ϩ ° ions (Th ,Na ) in the system. With the exception of Th - ThO2(am) and ThO2(c) at 25 C and those reported in the ClϪ, all of the appropriate ion-interaction parameters for literature are compared in Table 5. Although the primary this system are available or can be estimated (see Table 4 objective of this study was not to determine the solubility and Appendix A) for the temperature range between 23 product of ThO2(am), the value obtained in this study and 90 °C. Making a reasonable3 assumption that the ion- agrees closely with our earlier studies [9, 15] and is similar interaction parameters for Th4ϩ-ClϪ at 23 °C apply to the to the values quoted by Kovalenko and Bagdasarov [16] ° 0 90 C data, the ∆Gf /RT of the ThO2(am c) was deter- and Bilinski et al. [17] (Table 5). The large differences in → mined to be Ϫ377.629, which corresponds to the log Ksp some of the reported values appear to be due to differences of Ϫ49.2 for ThO2(am c). In a similar fashion, the log in experimental variables (e.g., the type and concentration → ° Ksp value for ThO2(c) at 90 C was determined to be Ϫ51.4 of electrolyte), and the choice of aqueous species or mode- based on the experimental data (Fig. 3). The solubility of ling parameters used in calculations. For example, Felmy ThO (am c) is many orders of magnitude lower than the 2 → et al. [19] pointed out that differences between values re- solubility for ThO2(am) and is only about 2.2 orders of ported in Östhols et al. [10] and those reported in Felmy et magnitude higher than the value for ThO2(c). The close al. [9] are due to assumptions about the aqueous phase mo- agreement between the observed and predicted solubilities dels rather than the absolute differences in solubilities. The (Figs. 1, 3, and 4) shows that the modeling parameters re- fact that ThO2(am) solubility in very dilute to concentrated chloride and cabonate solutions [8, 9, 19] is consistent with 3 Values of ion-interaction parameters for Th4ϩ-ClϪ are not avail- the value reported in Felmy et al. [9], which is similar to able at other than 25 °C. However, values of BThCl at 90°C, the only parameter important in these relatively dilute solutions, can be esti- the value obtained in this study, attests to the reliability mated by analogy with Th[ClO4]4 using Hovey’s [6] heat capacity data of the recently determined values. No other experimental (assuming (B /T) is similar to (B /T) and Roy et al. [18] ThCl P,I ThClO4 P, I ThO(c) solubility data at 25 or 90 °C are available for direct value for BThCl at 25 °C. On this basis and at an ionic strength of 0.1 M, comparisons. However, comparisons can be made with cal- BThCl changes from Ϫ9.6 at 25°CtoϪ9.7 at 90 °C, which is not signifi- cant. culated solubility products derived from calorimetric data. Thermodynamic model for the solubility of thorium dioxide 303 8. Rai, D., Felmy, A. R., Sterner, S. M., Moore, D. A., Mason, M. The log Ksp value of ՆϪ56.9 for ThO2(c) at 25 °C obtained in this study compares very favorably with previously cal- J., Novak, C. F.: The Solubility of Th(IV) and U(IV) Hydrous Oxides in Concentrated NaCl and MgCl2 Solutions. Radiochim. culated values (Ϫ54.2Ϯ1.3, Rai et al. [3]; Ϫ54.13, Wag- Acta 79, 239Ϫ247 (1997). man et al. [4]) and with the calculated value (Ϫ54.2) in 9. Felmy, A. R., Rai, D., Mason, M. J.: The Solubility of Hydrous this study using tabulated values from Wagman et al. [4], Thorium (IV) Oxide in Chloride Media: Development of an Aque- CODATA [20], and Cox et al. [21]. The experimental log ous Ion-Interaction Model. Radiochim. Acta 55, 177Ϫ185 (1991). K value of Ϫ51.4 for ThO (c) at 90 °C is almost identical 10. Östhols, E., Bruno, J., Grenthe, I.: On the Influence of Carbonate sp 2 on Mineral Dissolution: III. The Solubility of Microcrystalline to our calculated value (Ϫ51.6) derived from the selected ThO2 in CO2-H2O Media. Geochim. Cosmochim. Acta 58,613Ϫ 4ϩ data cited above and the heat capacity function for Th 623 (1994). given by Hovey [6]. The slightly larger log Ksp value of 11. Pitzer, K. S.: Thermodynamics of Electrolytes. I. Theoretical Basis ° and General Equations. J. Phys. Chem. 77, 268Ϫ277 (1973). Ϫ49.2 for ThO2(am c) at 90 C is consistent with the fact that this material is→ not fully crystalline. 12. Pitzer, K. S. : Ion Interaction Approach: Theory and Data Corre- lation: In Activity Coefficients in Electrolyte Solutions,2nd ed., CRC Press, Boca Raton, p. 75Ϫ153 (1991). 13. Sterner, S. M., Felmy, A. R., Rustad, J. R., Pitzer, K. S.: Thermo- Conclusions dynamic Analysis of Aqueous Solutions Using INSIGHT. PNWD- SA-4436, Pacific Northwest National Laboratory, Richland, The data presented in this paper show that ThO (am) con- Washington (1997). 2 14. Grenthe, I., Lagerman, B.: Studies of Metal Carbonate Equilibria. verts upon heating at 90 °C to a crystalline thorium dioxide 23. Complex Formation in the Th(IV)-H2O-CO2(g) System. Acta [ThO2(am c)] and that the Ksp for ThO2(am c) is many Chem. Scand. 45,231Ϫ238 (1991). → → ° orders of magnitude lower than that for ThO2(am) at 25 C. 15. Ryan, J. L., Rai, D.: Thorium(IV) Hydrous Oxide Solubility. In- ° org. Chem. 26,4140Ϫ4142 (1987). The measured ThO2(c) solubility product at 90 Cisin close agreement with the value calculated using selected 16. Kovalenko, P. N., Bagdasarov, K. N.: Use of the Photocolorimeter 4ϩ for Determination of the Solubility Product of Thorium Hydrox- calorimetric data and the Born model for Th given by ide. J. Applied Chem. USSR 34, 759Ϫ763 (1961). Hovey [6]. These thermodynamic data and models will pro- 17. Bilinski, H., Fu˝redi, H., Branica, M., Tezak, B.: Precipitation and vide reliable predictions for thorium behavior at tempera- Hydrolysis of Thorium(IV) in Aqueous Solution : Thorium Ni- tures reaching 90 °C. trate-Potassium Hydroxide. I. Determination of Solubility Con- stants of Th(OH)4. Croat. Chem. Acta 35, 19Ϫ30 (1963). 18. Roy, R. N., Vogel, K. M., Good, C. E., Davis, W. B., Roy, L. N., Acknowledgments. This research was conducted at the Pacific North- Johnson, D. A., Felmy, A. R., Pitzer, K. S.: Activity Coefficients west National Laboratory, operated by Battelle for the U.S. Department ° in Electrolyte Mixtures: HClϩThCl4ϩH2O for 5Ϫ55 C. J. Phys. of Energy (DOE) under Contract DE-AC06-76RLO 1830, in collabo- Chem. 96, 11065Ϫ11072 (1992). ration with Japan Nuclear Cycle Development Institute (JNC) under an 19. Felmy, A. R., Rai, D., Sterner, S. M., Mason, M. J., Hess, N. agreement with DOE for a project on “Development of Fundamental J., Conradson, S. D.: Thermodynamic Models of Highly Charged Thermodynamic and Adsorption Data”. We thank Drs. Sumio Masuda, Aqueous Species: Solubility of Th(IV) Hydrous Oxide in Concen- Hiroyuki Umeki, and Mr. Kaname Miyahara of JNC for their support trated NaHCO and Na CO Solutions. J. of Soln. Chem. 26, 233Ϫ and technical discussions throughout the study period. We also thank 3 2 3 248 (1997). Ms. Kay Hass for Text Processing and Ms. Kathi Hanson for editorial 20. CODATA 1978: Recommended Key Values for Thermodynamics. support of this document (1221B J ThO2(c) 2-11-00 final.doc). J. Chem. Thermodyn. 10, 903Ϫ906 (1978). 21. Cox, J. D., Wagman, D. D., Medvedev, V. A.: CODATA Key Val- ues for Thermodynamics. Hemisphere Pub. Corp., 271 pages References (1989). 22. Fuger, J., Oetting, F. L.: The Chemical Thermodynamics of Acti- 1. JNC, Japan Nuclear Cycle Development Institute. May 1999. H12 nide Elements and Compounds: Part 2, The Actinide Aqueous Project to Establish Technical Basis for HLW Disposal in Japan. Ions. IAEA, Vienna (1976). The draft second progress report on Research and Development 23. Harvie, C. E., Müller, N., Weare, J. H.: The Prediction of Mineral for the Geological Disposal of HLW in Japan. JNC TN1400 99- Solubilities in Natural Waters: The Na-K-Mg-Ca-H-Cl-SO -OH- 013. 4 HCO -CO -H O System to High Ionic Strengths at 25°C. Geo- 2. Prasad, R., Beasley, M. L., Milligan, W. O.: Aging of Hydrous 3 2 2 Thoria Gels. J. Electron Microscopy 16(2), 101Ϫ119(1967). chim. Cosmochim. Acta 48, 723 (1984). 3. Rai, D., Swanson, J. L., Ryan, J. L.: Solubility of NpO · 24. Pitzer, K. S., Peiper, J. C., Busey, R. H.: Thermodynamic Proper- 2 ties of aqueous Sodium Chloride Solutions. J. Phys. Chem. Ref. xH2O(am) in the Presence of Cu(I)/Cu(II) Redox Buffer. Radio- chim. Acta 42,35Ϫ41(1987). Data 13, 1Ϫ102 (1984). 4. Wagman, D. D., Evans, W. H., Parker, V. B., Schumm, R. H., 25. Holmes, H. F., Mesmer, R. E.: Isopiestic Molalities for Aqueous Halow, I., Bailey, S. M., Churney, K. L., Nuttal, R. L.: The NBS Solutions of the Alkali Metal Hydroxides at Elevated Tempera- Tables of Chemical Thermodynamic Properties, Selected Values tures. J. Chem. Thermodyn. 30,311Ϫ326 (1998). 26. Holmes, H. F., Busey, R. H., Simonson, J. M., Mesmer, R. E., for Inorganic and C2 and C2 Organic Substances in SI Units. J. Phys. Chem. Ref. Data Vol. 11, Suppl. 2 (1982). Archer, D. G., Wood, R. H.: The Enthalpy of Dilution of HCl(aq) 5. Morss, L. R., McCue, M. C.: Partial Molal Entropy and Heat Ca- to 648 K and 40 MPa. J. Chem. Thermodyn. 19, 863Ϫ890 (1987). pacity of the Aqueous Thorium(IV) Ion. Thermochemistry of Tho- 27. Pitzer, K. S., Kim, J.: Thermodynamics of Electrolytes. IV. Ac- rium Nitrate Pentahydrate. J. Chem. Eng. Data, 21, 337Ϫ341 tivity and Osmotic Coefficients for Mixed Electrolytes. J. Amer. (1976). Chem. Soc. 96, 5701Ϫ5707 (1974). 6. Hovey, J. K.: Thermodynamics of Hydration of a 4ϩ Aqueous 28. Felmy, A. R., Rai, D.: An Aqueous Thermodynamic Model for a 4ϩ 2Ϫ ϩ ϩ Ion: Partial Molar Heat Capacities and Volumes of Aqueous Tho- High-Valence 4:2 Electrolyte Th -SO4 in the System Na -K - ϩ ϩ 4ϩ 2Ϫ Ϫ rium(IV) from 10to55°C. J. Phys. Chem. B101, 4321Ϫ4334 Li -NH4 -Th -SO4 -HSO4 -H2O to High Concentration. J. Sol. (1997). Chem. 21, 407Ϫ423 (1992). 7. Apelblat, A., Sahar, A.: Properties of Aqueous Thorium Nitrate 29. Nabivanets, B.I., Kudritskaya, L. N.: Hydro Complexes of Th(IV). Solutions. Part 3. Partial Molar Heat Capacities and Heats of Di- Ukr. Khim. Zh. 30,891Ϫ895 (1964, in Russian). lution at 30°C. J. Chem. Soc. Faraday Trans. I, 71, 1667Ϫ1670 30. Moon, H.: Equilibrium Ultrafiltration of Hydrolyzed Thorium(IV) (1975). Solutions. Bull Korean Chem. Soc. 10, 270Ϫ272 (1989). 304 D. Rai, D. A. Moore, C. S. Oakes et al. 31. Haas, J. L. Jr., Fisher, J. R.: Simultaneous Evaluations and Corre- 33. Victor, A. C., Douglas, T. B.: Thermodynamic Properties of Tho- lation of Thermodynamic Data. Amer. J. Sci. 276, 525Ϫ545 rium Dioxide from 298 to 1200 K. Natl. Bur. Stds. J. Res. A. 65, (1966). 105Ϫ111 (1961). 32. Tanger, J. C., Helgeson, H. C.: Calculation of the Thermodynamic 34. Shock, E. L., Sassani, D. C., Willis, M., Sverjensky, D. A.: Inor- and Transport Properties of Aqueous Species at High Pressures ganic Species in Geologic Fluids: Correlations Among Standard and Temperatures: Revised Equations of State for the Standard Molal Thermodynamic Properties of Aqueous Ions and Hydroxide Partial Molal Properties of Ion and Electrolytes. Amer. J. Sci. 288, Complexes. Geochim. Cosmochim. Acta 61, 907Ϫ950 (1997). 19Ϫ98 (1988). Appendix A: of fits to the heat capacities for the individual species using Literature sources of thermodynamic quantities the function 0 Ϫ2 2 Ϫ1/2 used to calculate Ksp for ThO2(c) dissolution Cp ϭ q0 ϩ q1T ϩ q2T ϩ q3T ϩ q4T (A.3) 4+ − reaction [ThO2(c) + 2 H2O (Th +4OH] reported by Haas and Fisher [31], where the qi terms are fitting constants. The following sources were used for the The value of Ksp can be calculated from the thermodynamic relationships heat capacities of the species required for Ksp calculations: Hovey [6] for Th4ϩ(aq)1 ; Tanger and Helgeson [32] for 0 Ϫ ln Ksp ϭ ∆Gf, rxn/RT (A.1) OH (aq); Victor and Douglas [3] for ThO2(c); and Cox et 0 0 al. [21] for H2O(l). Values of ∆Hf (298 K) and ∆Sf (298K) and for the individual species were obtained from Shock et al. 0 0 0 [34], which in turn were taken from Wagman et al. [4] and ∆Gf, rxn ϭ ∆Hf,rxn Ϫ T∆Sf, rxn (A.2) CODATA [20] for Th4ϩ(aq), Tanger and Helgeson [32] for Ϫ where K is solubility product; ∆G0 , ∆H0 , ∆S0 are OH (aq), and from Cox et al. [21] for ThO2(c) and H2O(l). sp f, rxn f,rxn f,rxn Hovey’s experimental heat capacities were only measured standard Gibbs free energy, enthalpy, and entropy, of for- ° mation, respectively, of reaction (rxn); R is the gas con- to 55 C. Consequently, his Born model was used to pro- duce C0 (Th4ϩ) values to higher temperatures. Values of the stant; and T is the temperature in Kelvin. p fitting constants in Eq. (A.3) for each species are listed in At elevated temperatures, these calculations are most 0 0 0 Table A.1. Values of ∆Hf and ∆Sf for each species and accurate if heat capacity (Cp) data are available for the T calculated values of ln K at 298.15 K and 363.15 K are 0 temperature range of interest (e.g., and ͐ d∆Hf, rxn ϭ listed in Table A.2. Note that the 363.15 K ln Ksp ϭϪ117 T T T 298K 0 0 0 0 if the van’t Hoff relationship (∆Cp ϭ 0) is assumed versus ͐ ∆Cp,rxn dT and ͐ d∆Sf,rxn ϭ ͐ (∆Cp,rxn/T) dT). Valu- 298K 298K 298K ln Ksp ϭϪ119 for the case where heat capacities were 0 0 es of ∆Hf,rxn (T) and ∆Sf, rxn (T) were obtained by integration taken into account. Table A.1. Parameters for Eq. (A.3) and tem- 4ϩ Ϫ perature range (degrees Kelvin, K) over which Fitting constant Th OH ThO2(c) H2O heat capacities were fit. Ϫ3 10 q0 Ϫ22.204 Ϫ281.236 0.159053 Ϫ0.0187226 q1 12.90903 346.775 Ϫ0.051403 0.099873 Ϫ8 10 q2 Ϫ3.12312 Ϫ20.5431 q3 Ϫ0.217902 Ϫ3 10 q4 373.811 3801.96 Ϫ1.414711.11031 Temp. range (K) 273Ϫ373 273Ϫ423 263Ϫ373 293Ϫ373 Table A.2. Thermodynamic quantities of species taking part in Ksp calculation for ThO2(c) at 298 K and at 363 K. 4ϩ Ϫ a a Units Th OH ThO2(c) H2O Shock et al. [34] Tanger and Cox et al. [21] Cox et al. [21] Helgeson [32] Values at 298 K Ϫ1 0 J · mol ∆Hf Ϫ769 019 Ϫ230 024 Ϫ1 226 400 Ϫ285 830 Ϫ1 Ϫ1 0 J · mol ·K ∆Sf Ϫ215 Ϫ243.93 Ϫ191.61 Ϫ163.14 Ϫ1 0 J · mol ∆Gf Ϫ705 004 Ϫ157 297 Ϫ1169 271 Ϫ237 190 0 ∆Gf /RT Ϫ284.40 Ϫ63.45 Ϫ471.68 Ϫ95.68 Values at 363 K Ϫ1 0 J · mol ∆Hf Ϫ784 500 Ϫ237 100 Ϫ1 222 000 Ϫ280 900 Ϫ1 Ϫ1 0 J · mol ·K ∆Sf Ϫ261.4 Ϫ265.6 Ϫ179.0 Ϫ148.3 Ϫ1 0 J · mol ∆Gf Ϫ689 500 Ϫ140 700 Ϫ1157 000 Ϫ227 100 0 ∆Gf /RT Ϫ228.37 Ϫ46.60 Ϫ383.27 Ϫ75.21 0 0 0 0 0 a: Values of ∆Sf (298 K, ThO2) and ∆Sf (298 K, H2O) were calculated from the following values: S (Th) 51.8; S (O2, g) 205.043; S (ThO2, c), 0 0 65.23; S (H2, gas), 130.57; and S (H2O, l), 69.95 reported by Cox et al. [21]. Thermodynamic model for the solubility of thorium dioxide 305 Appendix B: Experimental data tables Table B.1. Measured ThO2 solubilities in Ϫ1 ° ThO2(am) 23Ϯ2 °C ThO2(am c) 23Ϯ 2°C 0.1 mol · kg NaCl(aq) at 23 and 90 C → (Set I). 6 days 21 days 76 days pH log pH log pH log Ϫ1 Ϫ1 Ϫ1 (CTh/mol · kg )(CTh/mol · kg )(CTh/mol · kg ) 2.030 Ϫ1.978 2.027 Ϫ1.985 1.466 Ϫ2.156 2.182 Ϫ1.995 2.191 Ϫ1.9811.469 Ϫ2.039 2.397 Ϫ1.999 2.419 Ϫ2.023 1.479 Ϫ2.262 2.564 Ϫ2.007 2.599 Ϫ2.003 1.480 Ϫ2.239 2.721 Ϫ2.007 2.764 Ϫ1.959 1.509 Ϫ1.986 2.915 Ϫ1.998 2.968 Ϫ2.010 1.620 Ϫ2.271 3.074 Ϫ2.036 3.130 Ϫ2.047 1.567 Ϫ2.380 3.170 Ϫ1.996 3.224 Ϫ1.977 1.615 Ϫ2.515 3.270 Ϫ2.011 3.332 Ϫ2.018 1.634 Ϫ2.551 3.546 Ϫ2.034 3.485 Ϫ2.035 1.722 Ϫ2.988 3.970 Ϫ2.134 3.748 Ϫ2.037 1.901 Ϫ3.240 4.157 Ϫ2.423 4.002 Ϫ2.234 2.000 Ϫ3.788 4.256 Ϫ2.650 4.105 Ϫ2.389 2.060 Ϫ3.916 4.506 Ϫ3.318 4.427 Ϫ3.039 2.192 Ϫ4.435 4.555 Ϫ3.469 4.517 Ϫ3.258 2.150 Ϫ4.441 4.718 Ϫ4.024 4.581 Ϫ3.436 2.206 Ϫ4.398 Table B.2. Measured ThO solubilities in 2 ° ° Ϫ1 ° ThO2(am) 23Ϯ2 C ThO2(am c) 90Ϯ 2 C 0.1 mol · kg NaCl(aq) at 23 and 90 C → (Set II). 5 days 13 days 52 days pH log pH log pH log Ϫ1 Ϫ1 Ϫ1 (CTh/mol · kg )(CTh/mol · kg )(CTh/mol · kg ) 4.207 Ϫ2.340 4.128 Ϫ2.417 1.970 Ϫ3.707 4.336 Ϫ2.681 4.277 Ϫ2.701 2.011 Ϫ3.939 4.377 Ϫ2.738 4.260 Ϫ2.804 1.983 Ϫ3.917 4.608 Ϫ3.138 4.527 Ϫ3.056 2.048 Ϫ4.044 4.823 Ϫ3.732 4.796 Ϫ3.520 2.081 Ϫ4.099 4.943 Ϫ3.918 4.782 Ϫ3.527 2.065 Ϫ4.267 4.964 Ϫ3.967 4.873 Ϫ3.625 2.076 Ϫ4.385 5.060 Ϫ4.252 4.982 Ϫ3.860 2.090 Ϫ4.320 Table B.3. Measured ThO2 solubilities in Ϫ1 ° ° ° ThO2(am) 23Ϯ2 C ThO2(am c) 90Ϯ 2 C 0.1 mol · kg NaCl(aq) at 23 and 90 C → (Set III). Note that the log thorium values for most of the 23°C samples are essentially con- 7 days 22 days 76 days stant due to complete solubilization of the solid phase. pH log pH log pH log Ϫ1 Ϫ1 Ϫ1 (CTh/mol · kg )(CTh/mol · kg )(CTh/mol · kg ) 1.809 Ϫ3.368 1.794 Ϫ3.406 ND a ND a 1.992 Ϫ3.347 1.984 Ϫ3.402 1.881 Ϫ3.179 2.124 Ϫ3.3192.109 Ϫ3.374 2.047 Ϫ3.210 2.281 Ϫ3.308 2.280 Ϫ3.374 2.055 Ϫ4.474 2.408 Ϫ3.341 2.415 Ϫ3.388 2.176 Ϫ4.775 2.587 Ϫ3.369 2.595 Ϫ3.421 2.202 Ϫ5.323 2.744 Ϫ3.360 2.759 Ϫ3.441 2.233 Ϫ6.278 2.832 Ϫ3.475 2.848 Ϫ3.530 2.320 Ϫ6.532 3.019 Ϫ3.344 3.047 Ϫ3.397 2.327 Ϫ5.697 3.174 Ϫ3.373 3.200 Ϫ3.426 2.381 Ϫ5.400 3.337 Ϫ3.375 3.360 Ϫ3.421 2.514 Ϫ5.521 3.594 Ϫ3.355 3.620 Ϫ3.402 2.568 Ϫ5.248 3.750 Ϫ3.371 3.791 Ϫ3.416 2.687 Ϫ5.458 3.876 Ϫ3.455 3.927 Ϫ3.490 2.778 Ϫ5.479 4.106 Ϫ3.777 4.133 Ϫ3.763 2.975 Ϫ5.894 4.218 Ϫ4.008 4.229 Ϫ4.004 NDa ND a a: ND ϭ no data. 306 D. Rai, D. A. Moore, C. S. Oakes et al. ° Table B.4. Measured ThO2(c) solubilities at 23 and 90 C. 23°C (Set IV)a 90 °C (set V) a 11 days 87 days 687 days 794 b days 12 days pH log pH log pH log pH log pH log Ϫ1 Ϫ1 Ϫ1 Ϫ1 Ϫ1 (CTh/mol · kg )(CTh/mol · kg )(CTh/mol · kg )(CTh/mol · kg )(CTh/mol · kg ) 1.308 Ϫ3.824 1.372 Ϫ3.616 1.391 Ϫ3.528 1.353 Ϫ3.187 1.367 Ϫ3.413 1.491 Ϫ4.059 1.518 Ϫ3.860 1.568 Ϫ3.743 1.466 Ϫ3.542 1.432 Ϫ3.548 1.688 Ϫ4.1011.719 Ϫ4.000 1.757 Ϫ3.892 1.682 Ϫ4.1311.753 Ϫ4.688 1.835 Ϫ5.098 1.876 Ϫ4.925 1.904 Ϫ4.870 1.816 Ϫ4.8011.886 Ϫ5.700 2.037 Ϫ4.957 2.074 ND c 2.103 Ϫ4.713 2.063 Ϫ6.350 2.058 Ϫ6.288 2.223 Ϫ4.954 2.267 Ϫ4.816 2.292 Ϫ4.694 2.225 Ϫ5.896 2.207 Ϫ6.798 2.431 Ϫ5.311 2.463 Ϫ5.160 2.463 Ϫ5.021 2.367 Ϫ6.366 2.381 Ϫ7.296 2.597 NDc 2.581 Ϫ4.988 2.623 Ϫ4.856 2.579 Ϫ7.063 2.525 Ϫ6.863 2.751 Ϫ5.191 2.786 Ϫ5.019 2.768 Ϫ4.991 2.718 Ϫ6.889 2.632 Ϫ6.926 2.865 Ϫ5.329 2.881 Ϫ5.235 2.878 Ϫ5.159 2.834 Ϫ8.142 2.758 Ϫ8.143 2.983 NDc 3.003 Ϫ5.312 2.981 Ϫ5.282 2.962 Ϫ7.261 2.830 Ϫ5.822 3.051 Ϫ5.609 3.045 Ϫ5.493 3.043 Ϫ5.498 2.991 Ϫ7.331 2.906 Ϫ8.172 3.185 Ϫ6.338 3.133 Ϫ6.311 3.168 Ϫ6.166 3.155 Ϫ9.369 3.232 Ϫ8.054 3.413 Ϫ6.894 3.358 Ϫ6.857 3.359 Ϫ7.127 3.357 Ϫ8.144 3.489 ND c 3.578 Ϫ7.747 3.591 Ϫ8.598 3.538 Ϫ9.367 3.551 Ϫ8.178 ND c ND c a: Different set of samples were used in Sets IV and V. ° ° ° b: ThO2(c) suspensions, previously equilibrated at 23 C for 779 days, were then equilibrated at 90 C for 14 days and then maintained at 23 C for 1 day before sampling. c: ND ϭ no data.