Environ. Sci. Technol. 2008, 42, 9357–9362
Stabilization/Solidification of Radioactive Salt Waste by Using xSiO2-yAl2O3-zP2O5 (SAP) Material at Molten Salt State HWAN-SEO PARK,* IN-TAE KIM, YONG-ZUN CHO, HEE-CHUL EUN, AND HAN-SOO LEE Advanced Fuel Cycle Development, Korea Atomic Energy Research Institute, 150 Deokjin-dong, Yuseong-gu, Daejeon 305-353, Korea
Received July 20, 2008. Revised manuscript received September 24, 2008. Accepted October 10, 2008.
The molten salt waste from the pyroprocess is one of the problematic wastes to directly apply a conventional process such as vitrification or ceramization. This study suggested a novel method using a reactive material for metal chlorides at a molten temperature of salt waste, and then converting them into manageable product at a high temperature. The inorganic composite, SAP (SiO2-Al2O3-P2O5), synthesized by a conventional sol-gel process has three or four distinctive domains that are bonded sequentially, Si-O-Si-O-A-O-P-O-P. The P-rich phase in the SAP composite is unstable for producing a series of reactive sites when in contact with a molten LiCl salt. After the reaction, metal aluminosilicate, metal aluminophosphate, metal phosphates and gaseous chlorines are generated. From this process, the volatile salt waste is stabilized and it is possible to apply a high temperature process. The reaction products were fabricated successfully by using a borosilicate glass with an arbitrary composition as a chemical binder. There was a low possibility for the valorization of radionuclides up to 1200 °C, based on the result of the thermo gravimetric analysis. The Cs and Sr leach rates by the PCT-A method were about 1 × 10-3 g/(m2 day). For the final disposal of the problematic salt waste, this approach suggested the design concept of an effective stabilizer for metal chlorides and revealed the chemical route to the fabrication of monolithic wasteform by using a composite as an example. Using this method, we could obtain a higher disposal efficiency and lower waste volume, compared with the present immobilization methods.
2. Experimental Section
1. Introduction In the management of radioactive wastes, the molten salt waste from a pyroprocess to recover uranium and transuranic elements, which mainly consists of metal chlorides, is one of the problematic wastes not directly applicable to a conventional solidification process because of its physicochemical properties such as its volatility and low comparability with silicate glass (1). At present, there are two approaches to the immobilization of these radioactive wastes, Cl-containing material as a host matrix and non-Cl-containing matrix by dechlorination. * Corresponding author e-mail:
[email protected]. 10.1021/es802012x CCC: $40.75
Published on Web 11/08/2008
There would be a few minerals to be considered as a host matrix for a Cl-based waste, where the matrix meets the condition that the waste loading is reasonable for the final waste volume. Among many Cl-containing minerals, sodalite (Na8Al6Si6O24Cl2) is a practical example to treat a metal chloride waste, which was suggested by Argonne National Laboratory (ANL) (2-6). Immobilization of metal chloride waste by using phosphate-based glass has been studied (7-9). Different from these direct immobilization methods, some dechlorinations to remove the intrinsic limitation due to the Cl-induced disadvantages have been investigated by other countries. Leturcq et al. suggested a wet conversion by using a KOH solution, where the Cl was recovered as AlCl3 in an aqueous media to immobilize it into wadalite (Ca6Al5Si2O16Cl3); fission products were converted into hydroxides, and then vitrified by a conventional process (10). Donze and his co-workers also studied the thermal conversion of metal chlorides into a phosphate glass by using NH4H2PO4 as a phosphate source, where NH4Cl or HCl were generated by two reaction routes (11). Ikeda et al. have studied the thermal conversion reactions of metal chlorides into oxides with boric acid to develop a method for the vitrification of radioactive molten salt wastes; except for CsCl, the metal chlorides reacted with boric acid to produce borate glass and generate HCl (12). Our research group studied the chemical conversion of metal chlorides into manageable products at a high temperature using a sol-gel process based at an aqueous solution system (Na2SiO3-H3PO4-Al(NO3)3-H2O system) and investigated the chemical durability and physical properties of its wasteform; the products were successfully solidified as a monolithic wasteform at a high temperature, without vaporizing radionuclides or other elements in the waste and showed reasonable properties (13, 14). This approach indicated that the phosphate-related products are compatible with borosilicate glass at some mixing ratios. However, the reaction system based at an aqueous state has intrinsic problems due to its highly acidic condition, though avoiding the limitations on a host matrix and processing temperature for an immobilization of metal chloride wastes. The previous study, named GRSS (gel-route stabilization/ solidification), was to establish a reliable reaction system and to obtain information on the compatibility between reaction products and borosilicate glass. On the basis of the results in the previous study, this paper proposed a new approach to treat metal chloride wastes by using “a stabilizer, SAP,” that consisted of SiO2, Al2O3, and P2O5 and was prepared by a conventional sol-gel process. The design of the SAP and its reactivity for a metal chloride waste were investigated and the results could provide important information on the immobilization of metal chlorides.
2008 American Chemical Society
Materials. LiCl, CsCl, and SrCl2 (99% purity, Merck) with a composition of 90, 6.8, and 3.2 wt %, respectively, were used to simulate the waste salt. The composite, SiO2-Al2O3-P2O5 (SAP), was prepared by a sol-gel process. Tetraethyl orthosilicate (TEOS, Aldrich, 98%), aluminum chlorides (AlCl3 · 6H2O, Junsei, 98%) and phosphoric acid (H3PO4, Junsei, 85%) were used as sources of Si, Al, and P, respectively. The molar ratio of Si/Al/P was adjusted to 3/2/2 (SAP 067), 1/1/1 (SAP 100), 1/1/1.25 (SAP 125), and 1/1/1.5 (SAP 150). All reagents were dissolved in EtOH/H2O and the mixture was placed in an electric oven at 55-70 °C after being tightly sealed. After a gelling/aging for 3 days, the transparent hydrogels were dried at 110 °C for 2 days and then thermally treated at 600 VOL. 42, NO. 24, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
9357
°C for 2 h. A series of final products (SAPs) after pulverizing to about 100 µm were used as a stabilizer for treating the salt wastes. Method. The prepared SAPs were reacted with metal chlorides at 650∼850 °C after mixing it with different metal chloride/SAP mixing ratios. Also, the simulated salt waste was treated with the same procedures. The reaction products were mixed with glass powder with a mixing ratio of 75(gel product):25(glass) by weight. This powder mixture was heattreated at 1100 °C for 4 h without pressing. Characterization. The leach-resistance of solidified products was evaluated by PCT-A leaching method (ASTM Standard C 1285-02, ref 15). The conversion rate of the metal chlorides was calculated from the Cl concentrations in the leachant under the PCT-A leaching test condition. The concentration in leachate was measured by IC, AAS and ICPAES. The reaction products were characterized by X-ray powder diffraction (XRD, Rikaku, Cu KR radiation) and TG/ DTA (10 °C/min, 200 mL air/min, SEIKO 6300, Japan). 29Si, 27Al, and 31P MAS NMR spectra for SAPs were recorded at ambient temperature with a Bruker DSX 400 multinuclear spectrometer. Spin speeds were 5.5 kHz for 29Si and 12 kHz for 27Al and 31P. The resonance frequencies observed were 79.5 MHz for 29Si, 104.2 kHz for 27Al, and 161.9 MHz for 31P. Chemical shifts were recorded with respect to TMS for 29Si, [Al(H2O)6]3+ for 27Al, and H3PO4 for 31P.
FIGURE 1. Schematic chemical structure of reactive material containing Si, Al, and P; Al is an intermediate element to connect the Si-O-Si and P-O-P chemical bond. P-based domain is unstable when containing molten metal chlorides. This activates the Si-based side chain for metal chlorides. These phenomena produce a series of reactive site to metal chlorides.
3. Results and Discussion Synthesis of a Reactive Material to Metal Chloride. It is well-known that alkali metal chlorides can dissolve ionic compounds at a molten state due to their high electronegativity difference. This causes some materials to be unstable in a molten salt and to decompose into alkali metal compounds. Under this situation, if a component in an unstable material is properly chosen, stable alkali compounds could be obtained, namely, the removal of the limitations on host matrix due to the volatility of metal chlorides or the compatibility with conventional silicate glass. In Figure 1S of the Supporting Information, the XRD patterns of zeolite4A after a contact with some molten salts are presented. Different from the molten NaCl-KCl, LiCl-containing molten salt decomposed the zeolite-4A structure into Li-aluminosilicate. This is caused by the LiCl chemical reactivity, not the reaction temperature. When the zeolite-4A is applied to the LiCl waste, the possible route to a sodalite at about 900 °C might be as follows. 4Na12Al12Si12O48 + 12LiCl f Li12Al12Si12O48 + 12NaCl
(1-A)
4Na12Al12Si12O48 + 12LiCl f Li12Al12Si12O48+ 6Na8Al6Si6O24Cl2 (1-B) Na12Al12Si12O48 + 4LiCl f 2Na6Li2Al6Si6O24Cl2
(1-C)
1-B and 1-C routes to a sodalite are the extreme cases where LiCl decomposes zeolite-4A or not. In a real situation, the property of the LiCl salt is gradually changed by the NaCl generated from the reaction 1-A, and then a metal chloridesloaded zeolite-4A is proceeded to the 1-C route. At this point, it is noted that the molten LiCl does change zeolite-4A into LiAlSiO4, instead of being decomposed into SiO2 or Al2O3 as a main network of zeolite-4A; Li-metal aluminosilicate is wellknown as a stable material among the Li-related compounds. From this observation, it could be concluded that the Si-O-Al network is reactive to molten LiCl. Phosphorus-based compounds are well-known as a reactive or unstable material for metal chlorides. For example, a precipitation by metal phosphate in molten LiCl-KCl or NaCl-KCl (16). This research indicated that the P-O-P or P-O-M (M ) Li or Na) bond is very reactive to molten metal chlorides. 9358
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 24, 2008
FIGURE 2. MAS NMR spectra of synthesized SAPs with different P mol fractions after heat-treatment at 600 °C. From these considerations, the chemical bond, Si-O-Al or P-O-P or P-O-M, could be a desirable component in reactive materials for molten metal chlorides. They are also very useful elements in a ceramized or vitrified wasteform. The material containing Si, Al, and P can be designed as shown in Figure 1. When this material contacts with a molten salt, the P-based domain is unstable, so that it breaks the link between Si-O-Al and P-O-Al, generating a series of reactive sites for metal chlorides. This conceptual approach can be realized by a conventional sol-gel process. Figure 2 show the MAS NMR spectra for a series of SAPs; Figures 2S and 3S in the Supporting Information show the XRD patterns of a series of SAP and the photographs of realized SAP powder or disk-typed. As shown in Figure 2S in the Supporting Information, the synthesized SAPs are roughly amorphous, where a SiO2 crystalline phase was observed and no crystalline phases were detected. For the 29Si NMR spectra, resonances around -90 to -110 ppm are ascribed to tetrahedral silicon atoms bound via oxygen to aluminum atoms with 0∼4, Si(4Al), Si(3Al), Si(2Al), Si(1Al) and Si(0Al) (17-19). As shown in Figure 3, the distribution of the chemical shift was not greatly changed with the P mol fraction. From the chemical shifts, the fraction of the tetrahedral silicon surrounded by the tetrahedral silicon (Si-O-Si) is larger than the fraction of the tetrahedral silicon connected with the tetrahedral aluminum (Si-O-Al). This would mean that the silanol group (Si-OH) on the silicate
FIGURE 3. Possible chemical structure of SAPs from the MAS NMR anaysis. A, silicate domain (Si-O-Si); B, aluminosilicate domain(Si-O-Al); C, aluminophosphate domain (Al-O-P); and D, random domain (P-O-P or P-O-Al). chain formed at a highly acidic condition is subsequently connected with Al; Si-O-Si is the main chain and Si-O-Al exists as a side chain. The 27Al NMR spectra show a large resonance centered at 36 ppm, where the chemical shifts, 36 ppm, is ascribed to the tetrahedrally coordinated aluminum (20, 21). Low chemical shifts, -10 to -16 ppm, are attributed to the pentahedrally and octahedrally coordinated aluminum. The distribution of the chemical shift is sharpened with the increase of the P mol fraction. This trend is reverse to that of the 29Si MAS NMR spectra. At 36 ppm, Al connects with tetrahedral Si or P (Al(4Si) or Al(4P)). From the XRD analysis, the sharpness in the distribution of chemical shift in Al spectra is attributed to the increase of Al(4P), whereas the change in Si spectra is caused by the increase of the Si-O-P bond. This can be proved by the change of the P spectra. For the 31P NMR spectra of SAP 150, a sharp peak at -29 ppm and a shoulder peak at -36 ppm is assigned to the Q4 tetrahedral P site with neighboring AlO4, P(4Al), and Q3 P sites, respectively (22-25). Similar to Al, the main resonance for the P MAS NMR spectra was sharpened with the P mol fraction. It is noted that the chemical shift, -36 ppm, appeared at SAP 125. The shoulder peaks at SAP 125 and 150 is assigned to the Q3 P connected with -O-P, -O-Al, or -O-Si. The increase of the main resonance at -29 ppm means an increase of the fraction of Al(4P) or P(4Al), and an excess of P over the amount required to connect with Al would be connected with Si or P itself. This is the reason that the chemical shift, -36 ppm, appeared at SAP 125 and SAP 150.
From the MAS NMR analysis, the chemical structure is roughly defined as indicated in Figure 3. There are four distinctive domains related to silicate, aluminosilicate, aluminophosphate and a random chemical structure. It is noted that Al situated at the random domain decreases with the P mol fraction while Q3 P in the random domain increases. The MAS NMR analysis proves the schematic model of SAPs, which is valid under the given synthesis conditions for a series of SAPs. Reaction of SAPs with Metal Chlorides. Figure 4 indicates the XRD patterns of the reaction products for each metal chloride, LiCl, CsCl, SrCl2 and CeCl3. The reaction ratio was set to LiCl/MCl ) 4 on a mol base and salt/SAP ) 0.7 on a weight base. The reaction products are metal-aluminosilicate, metal-aluminophosphate, and metal phosphate as expected in Figure 1. These reactions are simply expressed as follows. LiCl + SAP f Li3PO4+LixAlxSi1-xO2-x+Cl2
(2)
LiCl + CsCl + SAP f Li3PO4+Cs2AlP3O10+ (Li, Cs)-alumniosilicate + Cl2 (3) LiCl + SrCl2 + SAP f Li3PO4+LixAlxSi1-xO2-x+Sr5(PO4)3Cl + Cl2 (4) LiCl + CeCl3 + SAP f Li3PO4+LixAlxSi1-xO2-x+CePO4+ Cl2 under the oxidative condition (5) As expected in the previous description on the SAP design, the moletn LiCl attacks the unstable P-rich region to break the linkage of “Si-O-Al-O-P-O-P”, generating a series of reactive sites, Si-O-Al, Al-O-P, P-O-P. From this route, metal chlorides react with each site to produce a series of products. All these products are well-known as thermally stable compounds. It means “not vaporized” at a high temperature. It is noted that the reaction product for LiCl as a main component in the salt waste did not changed with the reaction temperature from 650 to 850 °C, as indicated in Figure 4S in the Supporting Iinformation; the reaction route is determined at the molten temperature. Figure 5 presents the XRD patterns of the reaction products of SAP 125 with different reaction ratios of simulated salt waste. As shown in figure 5, the main crystalline phases were Li3PO4, Li-aluminosilicate and AlPO4, where the peak of AlPO4 gradually increased with the reaction ratio. Metal chlorides react with the reactive site in the formed SAP by contacting with the molten salt. At this moment, in the case that the metal chloride is enough to break the linkage and react with a sereis of reactive sites, AlPO4 would not exist in the reaction products. If not, the residual sites in the borken
FIGURE 4. XRD patterns of reaction products of SAP 100 for each metal chloride (at 650 °C for 4 h). VOL. 42, NO. 24, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
9359
FIGURE 5. XRD patterns of reaction products of SAP 125 with different reaction ratio in weight basis. (f, Li-aluminosilicate; [, AlPO4(Berlinite); g, Li3PO4; at 650 °C for 16 h).
FIGURE 6. Residual Cl % in reaction products at different reaction conditions. (This was calculated from the initial Cl amount in the mixture and the Cl weight released from the reaction products using the PCT-A method). linkages would be connected with each other and developed to Si-, Al-, or P-related compounds. On the basis of the results in Figure 5 and the Supporting Information, Figure 5S, AlPO4 can be considered as the reaction indicator for the validity of reaction ratio. Namely, the existence of AlPO4 means that the SAP still has an ability to react with metal chloride at a given reaction condition. From this result, the increase of the P mol fraction in SAP (from SAP 067 to SAP 150) enhanced the capability to react with metal chlorides; this can be proved by analyzing Cl in reaction products. Figure 6 presented the residual Cl in the reaction product for each SAP with different reaction ratios; the residual Cl% is the ratio of the leached amount of Cl of the reaction product by PCT-A test to the inital amount of Cl of the mixture before the reaction. The dechlorination rate related to the reaction has the same trend as expected from the XRD analysis and it was about 98-99% at some reaction conditions. Solidification of the Reaction Products. Use of SAP to treat a salt waste is to remove the Cl-induced disadvantages such as the limitations on host matrix or processing temperature. As described in the previous sections, the dechlorination of the metal chlorides could be accomplished to produce nonchloride compound by using SAP. This can provide high freedom on how to fabricate a monolithic wasteform. At this point, it is important whether the radionuclides are vaporized or not during a high temperature process. Figure 7 indicates the results of the thermo gravi9360
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 24, 2008
FIGURE 7. Thermogravimetric analysis of reaction products for each SAP at the reaction ratio, SAP/salt ) 2 (200 mL air/min, heating rate ) 10 °C/min). metric analysis for the reaction products. If metal chlorides are not interacted with SAPs, the LiCl, CsCl and SrCl2 start to vaporize at about 600, 800, and 900 °C under a given TGA condition, respectively. Depending on the kind of SAP, there were one or two weight loss steps between about 400 and 1000 °C; the unreacted LiCl and CsCl might be vaporized at about 600 and 800 °C. For SAP 125, the weight loss at this temperature range was about 0.71 wt %. The unreacted fraction of the metal chlorides could be calculated roughly and confirmed by using TGA, residual Cl% and the amount change of product. Under the assumption, a perfect conversion reaction, the metal oxide from the metal chlorides occupied 16.49 wt% of the reaction products (83.51 wt % SAP). If the total loss at the temperatures range is attributed to the unreacted metal chlorides, the conversion rate based on the TGA result is about 98.3% (see the calulation in the Supporting Information).The residual Cl% in Figure 6 conrresponds well with the results in Figure 7. On the basis of these data, it could be concluded that the dechlorination rate for SAP 125 and SAP 150 is about 98-99% under a given condition in this study. However, as shown in figure 5, the XRD analysis with different reaction ratios could indicate that the reaction ratio, SAP/salt ) 2, is sufficient to convert the given amount of metal chlorides into desirable compounds in that AlPO4 as a reaction indicator exist at SAP/salt ) 2 but there no existence of AlPO4 at SAP/salt ) 1. Therefore, the unreacted amount, 1 or 2% of the given metal chlorides might be attributed to the immperfect mixing between SAP and salt at a solid state.
TABLE 1. Leach Rate of Solidified Products by PCT-A Method (units g/(m2 day))a SAP/salt ) 2
Li Na Cs Ca Sr Si Al B P a
SAP/salt ) 3
SAP 067
SAP 100
SAP 125
SAP 150
SAP 067
SAP 100
SAP 125
SAP 150
0.0992 0.0027 0.0014 0.0045 0.0034 0.0143 0.0122 0.1327 0.0640
0.1083 0.0027 0.0012 0.0070 0.0015 0.0204 0.0114 0.0809 0.0714
0.1198 0.0162 0.0014 0.0058 0.0017 0.0145 0.0084 0.1422 0.0802
0.1216 0.2000 0.0013 0.0038 0.0017 0.0108 0.0075 0.4563 0.0804
0.2902 0.0213 0.0020 0.0040 0.0014 0.0126 0.0239 0.4738 0.1555
0.1762 0.1300 0.0057 0.0051 0.0034 0.0086 0.0153 0.3730 0.1324
0.1139 0.0753 0.0042 0.0054 0.0012 0.0131 0.0086 0.1893 0.0689
0.1895 0.0097 0.0016 0.0067 0.0007 0.0189 0.0214 0.2821 0.0828
Overall leach-rate: 0.02-0.04 g/m2day. Duplicated experiment (mean deviation: about 15∼35%).
FIGURE 8. SEM and elemental mapping image for a solidified product by using SAP 125, salt, and a borosilicate glass as a binder; solid line, magnified region; dashed line, a region for elemental mapping analysis. With this thermal stability, we fabricated a monolithic wasteform by using a borosilicate glass as a binder; this concept is based on a glass-bonded sodalite wasteform suggested by ANL. In the previous study using an aqueous system, the GRSS method, a borosilicate glass functioned as a chemical binder for aluminosilicate and a physical binder for P-containing products. The reaction products in this study are very similar to those in the previous study, where the products were Li3PO4 and amorphous aluminosilicate or aluminophosphates. As an arbitarary choice, a borosilicate glass (7.74 wt % Na2O, 1.60 wt % CaO, 22.84 wt % B2O3, 8.84 wt % Al2O3 and 58.98 wt % SiO2) used in the previous study was mixed with the reaction products for a sereis of SAPs. We investigatged the compatibility between the borosilicate and reaction products. Figure 6S in the Supporting Information indicated the photographs of the solidified samples. There were no bulk phase separation and it looked like a highly uniform feature at SAP 125. From the TGA analysis, at the reaction ratio, SAP/salt ) 3, a higher amount of unreacted SAP would exist, and its solidified products would be less densified. The uniformity and densification decreases with the P content in the SAP. The given glass as a binder can interact with reaction products by two routes, chemically or physically binded. As shown in Figure 6S in the Supporting Information, the borosilicate does interact physically with the unreacted SAP and is chemically bound with the reaction products. When using the given glass as a binder, a highly monolithic wasteform can be obtained at the reaction condition, SAP/salt ) 2 for SAP 125; at this condition the dechlorination rate was about 98% and there is below a 1% weight loss up to 1200 °C from the TGA anaysis. Table 1
shows the result of the PCT-A leaching test mehod for Cs and Sr. Regardless of the SAP compositions, the leach rates for Cs and Sr were about 1 × 10-3 g/(m2 day), respectively (overall leach rate ) 1 × 10-2g/(m2 day)). This leach-resitance of these radionuclides is reasoanble, compared with the result in other researches or wasteform. The SEM image of the solidified products by using SAP 125 are shown in Figure 8. The grains with a 1-2 µm size were uniformly distributed in the matrix and they would be a relativley Si-rich phase. The P-rich phase was situated between the grains, whereas Al was distributed unformly. Considering the powder size, 100-200mesh (75-150 µm), used in this study, the weak separation between two phases on a 1 µm scale indicated that the given glass as a binder has good compatibility with the reaction products; For the GRSS method in the previous study there was a distinctive separation between a P-rich phase and a Si-rich phase. In summary, for the final disposal of a salt waste, it requires about 2 g of SAP and 1 g of borosilicate for the treatment of 1 g of a salt waste. During the reaction, metal chlorides are converted into thermally stable compounds up to 1200 °C and the products are immobilized into a monolithic wasteform by using a borosilicate as a chemical binder. By using this method, we can reduce the final waste volume for long-term storage, when compared with the present immobilization methods.
Acknowledgments This project has been carried out under the Nuclear R&D program by the Ministry of Science and Technology in Korea. VOL. 42, NO. 24, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
9361
Supporting Information Available XRD analysis and photograph of a series of solidified products (PDF).This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Metcalfe, B. L.; Donald, I. W. Candidate wasteforms for the immobilization of chloride-containing radioactive waste. J. NonCryst. Solids. 2004, 348, 225–229. (2) Lexa, D.; Leibowitz, L.; Kropf, J. On the reactive occlusion of the (uranium trichloride + lithium chloride + potassium chloride) eutectic salt in zeolite 4A. J. Nucl. Mater. 2000, 279, 57–64. (3) Lambregts, M.; Frank, S. M. Characterization of cesium containing glass-bonded ceramic waste forms. Microporous Mesoporous Mater. 2003, 64, 1–9. (4) Ebert, W. L.; Lewis, M. A.; Johnson, S. G. The precision of product consistency tests conducted with a glass-bonded ceramic waste form. J. Nucl. Mater. 2002, 305, 37–51. (5) Morss, L. R.; Lewis, M. A.; Lichmann, M. K.; Lexa, D. Cerium, uranium, and plutonium behavior in glass-bonded sodalite, a ceramic nuclear waste form. J. Alloys Compd. 2000, 303-304, 42–48. (6) Lichmann, M. K.; Reed, D. T.; Kropf, A. J.; Aase, S. B.; Lewis, M. A. EXAFS/XANES studies of plutonium-loaded sodalite/glass waste forms. J. Nucl. Mater. 2001, 297, 303–312. (7) Mesko, M. G.; Day, D. E. Immobilization of spent nuclear fuel in iron phosphate glass. J. Nucl. Mater. 1999, 273, 27–36. (8) Mesko, M. G.; Day, D. E.; Bunker, B. C. Immobilization of CsCl and SrF2 in iron phosphate glass. Waste. Manage. 2000, 20, 271–278. (9) Day, D. E.; Wu, Z.; Ray, C. S.; Hrma, P. Chemically durable iron phosphate glass wasteforms. J. Non-Cryst. Solids 1998, 241, 1– 12. (10) Leturcq, G.; Grandjean, A.; Rigaud, D.; Perouty, P.; Charlot, M. Immobilization of fission products arising from pyrometallurgical reprocessing in chloride media. J. Nucl. Mater. 2005, 347, 1–11. (11) Donze, S.; Montagne, L.; Palavit, G. Thermal conversion of heavy metal chlorides (PbCl2, CdCl2) and Alkaline chlorides (NaCl, KCl) into phosphate glasses. Chem. Mater. 2000, 12, 1921–1925. (12) Ikeda, Y.; Takesima, Y. Conversion reaction of metal chlorides into oxides with boric acid. J. Nucl. Sci. Technol. 1995, 32, 1138– 1145. (13) Park, H. S.; Kim, I. T.; Kim, H. Y.; Ryu, S. K.; Kim, J. H. Stabilization/solidification of radioactive molten salt waste via gel-route pretreatment. Environ. Sci. Technol. 2007, 41, 1345– 1351.
9362
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 24, 2008
(14) Park, H. S.; Kim, I. T.; Cho, Y. J.; Eun, H. C.; Kim, J. H. Characteristics of solidified products containing radioactive molten salt waste. Environ. Sci. Technol. 2007, 41, 7536–7542. (15) ISO-6961: Long-Term Leach Testing of Solidified Radioactive Wasteforms; International Standard Organization: Geneva, Switzerland, 1982. (16) Volkovich, V. A.; Griffiths, T. T.; Thied, R. C. Treatment of molten salt wastes by phosphate precipitation: removal of fission product elements after pyrochemical reprocessing of spent nuclear fuels in chloride melts. J. Nucl. Mater. 2003, 323, 46–56. (17) Buchholz, A.; Wang, W.; Arnold, A.; Xu, M.; Hunger, M. Thermal stability and dehydroxylation of Brønsted acid sites in silicoaluminophosphates H-SAPO-11, H-SAPO-18, H-SAPO-31, and H-SAPO-34 investigated by multi-nuclear solid-state NMR spectroscopy. Microporous Mesoporous Mater. 2002, 56, 267– 278. (18) Buchholz, A.; Wang, W.; Arnold, A.; Xu, M.; Hunger, M. Successive steps of hydration and dehydration of silicoaluminophosphates H-SAPO-34 and H-SAPO-37 investigated by in situ CF MAS NMR spectroscopy. Microporous Mesoporous Mater. 2003, 57, 157– 168. (19) Yamashita, H.; Yosino, H; Nagata, K.; Inoue, H.; Nakajin, T.; Maekawa, T. Nuclear magnetic resonace studyies of alkaline earth phosphosilicate and aluminoborosilicate glasses. J. NonCryst. Solids 2000, 270, 48–59. (20) Belkeir, A.; Rocha, J.; Esculcas, A. P.; Berthet, P.; Gilbert, B.; Gabelica, Z.; Llabres, G.; Wijzen, F.; Rulmont, A. Structural characterisation of glassy phases in the system Na2O-Al2O3P2O5 by MAS and solution NMR, EXAFS and vibrational spectroscopy. Spectrochim. Acta A 1999, 55, 1323–1336. (21) Hartmann, P.; Vogel, J.; Friedrich, U.; Jager, C. Nuclear magnetic resonance investigations of aluminum containing phosphate glass-ceramics. J. Non-Cryst. Solids 2000, 263, 94–100. (22) Pires, R.; Abrahams, I.; Nunes, T. G.; Hawkes, G. E. Non-random cation distribution in sodium-strontium-phosphate glasses. J. Non-Cryst. Solids 2004, 337, 1–8. (23) Metwalli, E.; Brow, R. K. Modifier effects on the properties and structures of aluminophosphate glasses. J. Non-Cryst. Solids 2001, 289, 113–122. (24) Egan, J. M.; Wenslow, R. M.; Mueller, K. T. Mapping aluminum/ phosphorus connectivities in aluminophosphate glasses. J. NonCryst. Solids 2000, 261, 115–126. (25) Cody, G. D.; Mysen, B.; Saghi-Szabo, G.; Tossell, J. A. silicatephosphate interaction in silicate glass and melt: I. multinuclear (27Al, 29Si, 31P)MAS NMR and ab initio chemical shielding (31P) study of phosphorous speciation in silicate glasses. Geochim. Cosmochim. Acta 2001, 65, 2395–2411.
ES802012X
