Article pubs.acs.org/est
Synthesis of Coffinite, USiO4, and Structural Investigations of UxTh(1−x)SiO4 Solid Solutions
Sabrina Labs,*,† Christoph Hennig,‡ Stephan Weiss,‡ Hilde Curtius,† Harald Zan̈ ker,‡ and Dirk Bosbach† †
Insitute of Energy and Climate Research (IEK-6), Nuclear Waste Management, Forschungszentrum Jülich GmbH, Wilhelm-Johnen-Straße, 54245 Jülich, Germany ‡ Institute of Resource Ecology, Helmholtz-Zentrum Dresden-Rossendorf, Bautzner Landstrasse 400, 01328 Dresden, Germany S Supporting Information *
ABSTRACT: The miscibility behavior of the USiO4−ThSiO4 system was investigated. The end members and 10 solid solutions UxTh(1−x)SiO4 with x = 0.12−0.92 were successfully synthesized, without formation of other secondary uranium or thorium phases. Lattice parameters of the solid solutions evidently follow Vegard’s Law. Investigation of the local structure with EXAFS reveals small differences between the U and Th environment attributed to different atomic radii of the metal atoms but no implications for a miscibility gap. The data provided confirm complete miscibility for the system USiO4−ThSiO4. The structure of the end members was studied in detail with XRD and discussed with special regard to the oxygen positions and the often neglected Si−O bond length. USiO4 could be obtained without UO2 impurities and the lattice parameters derived from Rietveld refinement as c = 6.2606(3) Å and a = 6.9841(3) Å. The Si−O distance in USiO4 appears to be 1.64 Å, which is more reasonable than earlier reported values.
■
INTRODUCTION In many countries with a nuclear energy program, spent nuclear fuel and/or vitrified high level radioactive waste will be disposed in a deep geological repository. Demonstrating the long-term safety (usually 106−109 yrs) of such a repository system is a major associated scientific challenge. The potential release of radionuclides into the environment strongly depends on the availability of water and the subsequent corrosion behavior of the waste form as well as the formation of secondary phases, which control the radionuclide solubility. The U solubility is typically derived from the solubility product of UO2 for reducing conditions in the repository system. Some U(IV) minerals, which could form as secondary phases, may have a lower solubility. This is indicated for coffinite,1 USiO4, from thermodynamic data.2 While investigations on this matter are scarce, the most recent study3 indeed questions the lower solubility. It is possible to demonstrate the long-term safety using only UO2 solubility data. For certain scenarios knowledge of validated thermodynamic data of coffinite would be very valuable. In natural uranium deposits such as Oklo, Gabon, and Cigar Lake, Canada, coffinite can be abundantly found as the alteration product of uraninite,4−6 which is considered a natural analogue of UO2.7 The precipitation of USiO4 as a secondary phase should be favored in contact with silica-rich groundwater (silica concentration >10−4 mol/L, Langmuir’s criterion).8 Natural samples are often metamict due to the long exposure to α-radiation and are associated with other minerals and organic © 2013 American Chemical Society
matter. Hence the determination of accurate thermodynamic values from these samples is not feasible. The structure of USiO4 was determined in analogy to thorite (ThSiO4) to be tetragonal with space group symmetry I41/amd and Z = 4. It is an orthosilicate isostructural to zircon (ZrSiO4), hafnon (HfSiO4), and thorite (ThSiO4). While ThSiO4 can easily be synthesized in the thorite as well as in the huttonite9 modification, synthesis of pure USiO4 proves to remain rather challenging. By applying the original hydrothermal procedure developed by Hoekstra and Fuchs,10 samples often turn out to contain the oxide, UO2, and are of rather poor crystallinity.11,12 For the last decades attempts to obtain synthetic USiO4 by other methods13,14 have failed. The possibility of a solid solution between USiO4 and ThSiO4 was already predicted by Goldschmidt,15 and first attempts were performed by Fuchs and Gebert.16 From this limited data it seems likely that solid solutions exist and follow Vegard’s Law. Förster17 has investigated natural uranothorite and found a maximum solubility of 30 mol % Th in coffinite and maximum 36 mol % U in thorite. Calculations by Ferriss et al.18 suggest that the maximum amount of Th in USiO4 or U in ThSiO4 should not exceed 12 mol % at 500 K with complete immiscibility at 1000 K. Costin et al.19 assigned the problems Received: Revised: Accepted: Published: 854
September 9, 2013 November 26, 2013 December 2, 2013 December 2, 2013 dx.doi.org/10.1021/es403995b | Environ. Sci. Technol. 2014, 48, 854−860
Environmental Science & Technology
Article
XAFS Data Acquisition and Analysis. XAS data were collected at the ROBL beamline (BM20) at the European Synchrotron Radiation Facility (ESRF). A Si(111) doublecrystal monochromator and a pair of collimating and focusing rhodium-coated mirrors was used. Data were collected in transmission mode at Th L3 (Ek=0 = 16310 eV) and U L3 (Ek=0 = 17185 eV) edges. Energy was calibrated using the K edge (E = 17038 eV) of an yttrium metal foil. For each X-ray absorption fine structure (EXAFS) measurement, the spectra of the reference foil was systematically collected at the same time. Several sweeps were performed during the measurement on the same sample to improve the signal-to-noise ratio. The measurements were performed at room temperature, the ones of the end member samples at 15 K employing a closed cycle He cryostat. The EXAFS oscillations were extracted from the raw absorption spectra by using WinXAS 3.1,26 and the edge step extracted with ATHENA from the demeter package.27 Curve fitting was done using the EXAFSPAK software.28 The interatomic scattering paths were calculated using the ab initio code FEFF8.29 Since the theoretical backscattering phases of U and Th are almost identical, no effort was made to generate mixed scattering phases. The related error in the interatomic distance R is less than the standard deviation. The data analysis was conducted by fixing the coordination numbers, N, to the crystallographic values and refining the distances, R, and Debye−Waller factors, σ2, simultaneously. The amplitude reduction factor (S02) was set at 0.90, and the shift in the threshold energy (E0) was linked during the fit for all paths of a data set.
associated to the synthesis of pure USiO4 to kinetics and studied the synthesis behavior of the solid solutions. While the authors were able to synthesize UxTh(1−x)SiO4 solutions up to x = 0.8, the products in the U-rich members were associated with an oxide phase, UyTh(1−y)O2, which dominated the final synthesis product. This study aims to analyze the structural properties of UxTh(1−x)SiO4 solid solutions throughout the full range of x = 0−1. A simple synthesis route to avoid the formation of UyTh(1−y)O2 through using an excess of SiO2 is presented. The study focuses on the characterization of the USiO4−ThSiO4 system and to indicate whether a miscibility gap exists. Lattice parameters over the complete range of the UxTh(1−x)SiO4 solutions are investigated by X-ray diffraction. EXAFS is further applied to characterize the local structure.
■
MATERIALS AND METHODS All reagents are were analytical grade from common suppliers. Uranium metal turnings were a gift of Prof. Florian Kraus from TU Munich. Experimental steps are performed in an anoxic glovebox under N2 atmosphere (O2 below 10 ppm). An UCl4 solution was prepared by dissolution of uranium metal turnings in 6 M HCl. Before dissolution the uranium turnings were washed in EtOH abs. to remove remaining oil from the cutting surface. The oxide layer was removed by washing the turnings with 0.1 M HNO3. As soon as the hydrogen evaporation ceased the obtained solution was filtered to remove any possibly remaining solids.20 The uranium concentration was then determined by ICP-MS, and the oxidation state of uranium(IV) was confirmed through UV−vis. ThCl4 solution was prepared by weighing the desired amount of Th(NO3)4·5H2O and dissolving it in 6 M HCl. Afterward several steps of dissolution and evaporation were performed.21 Ultimately, the concentration of thorium in solution was measured by ICP-MS. A 0.1 M silica solution was prepared by dissolving Na2SiO3·9H2O in carbonate-free Millipore water. Synthetic coffinite, USiO4, was prepared by adding UCl4 solution to a 6-fold amount of silica solution. The pH value was then adjusted to ∼8, and the greenish precipitate was centrifuged. The clear supernatant was pipetted off, and the remaining green slurry was mixed with NaHCO3 buffer. This mixture was transferred into a glass ampule and sealed. The sealed ampules were hydrothermally treated at 250 °C for 7 d and cooled to room temperature at a rate of 20 °C/d. After cooling the product was washed several times with water and dried. The solid solutions were prepared employing the same procedure, but using a mixture of the UCl4 and ThCl4 stock solutions in the desired stoichiometry. Synthetic thorite, ThSiO4, was prepared according to the protocol of Frondel and Collette.9 Initially supernatants were analyzed with ICP-MS to determine the precipitation yield of uranium and thorium, which proved to be quantitative. XRD Data Acquisition and Analysis. X-ray powder diffraction measurements were carried out at room temperature using a Bruker D8 Advance diffractometer with Bragg− Brentano geometry (Cu Kα1,2 radiation, 30 kV, and 45 mA) equipped with a VÅnTech linear position-sensitive detector. The powder patterns were recorded using a step size of 0.028 with an exposure time of 10 s across the angular range 2θ = 15−130°. Lattice parameters were refined by the Le Bail method22 using the JANA2006 program.23 Further Rietveld analysis of the pure sample was conducted using GSAS program.24,25
■
RESULTS AND DISCUSSION Calculation of the Mole Fractions. In the present study 10 UxTh(1−x)SiO4 solid solutions with x = 0.12−0.92 were successfully synthesized without the presence of UyTh(1−y)O2 through the use of excess silica. From the absorption edge of the XANES spectra the relative amounts of uranium and thorium in the sample were calculated. From these values the mole fractions were calculated, which were confirmed by EDS analysis (see Supporting Information for detailed description). The uranium amount in the product phases was always slightly higher than the initial ratio. In the study of Costin et al. always less uranium was detected in the desired silicate phase, UxTh(1−x)SiO4, than expected. At x = 0.24 an accompanying dioxide phase, UyTh(1−y)O4, appeared to which Costin et al. could attribute the missing uranium. Results from XRD Data. Coffinite. The X-ray diffraction pattern of the pure USiO4 shows the expected zircon-type pattern for spacegroup I41/amd. By qualitative investigation of the spectra the presence of other uranium phases could be excluded. The Rietveld refinement,30 therefore, was carried out taking into account only the silicate phase (cf. Figure 1a). The region of 2θ = 20.5−22.5°, which is dominated by a broad peak resulting from glassy silica, was excluded from the refinement to achieve a better agreement between experimental and calculated pattern. Lattice constants were determined as c = 6.2606(3) Å and a = 6.9841(3) Å from the Rietveld refinement with reliability factors of Rp = 7.89% and wRp = 8.28%. A comparison between the lattice constants from our refinement and those reported in literature is given in Table 1. The formation of USiO4 can be achieved within reasonable reaction times and without the presence of the UO2 phase through the modification of the synthesis, i.e., the use of excess SiO2. 855
dx.doi.org/10.1021/es403995b | Environ. Sci. Technol. 2014, 48, 854−860
Environmental Science & Technology
Article
oxygen bond length is Si−O = 1.64 Å. The earlier reported value of Fuchs and Gebert suggests a Si−O bond length of 1.58 Å.16 This is extremely short for a silicate, especially when taking into account that U−O bonds ought to establish significant covalent bonding behavior. Typical actinide and rare earth silicates are in the range of 1.61−1.80 Å.31 The same difficulties were experienced with ThSiO4, where oxygen positions were uncertain until Taylor and Ewing successfully determined them from single crystal measurements.32 The Si−O distance then was found to be 1.63 Å. Values for the synthesized ThSiO4 were determined from Rietveld refinement (see Figure 1b) and were in fair agreement with the single crystal data. Solid Solutions. By qualitative analysis of the X-ray diffraction pattern, the synthesized solid solutions of UxTh(1−x)SiO4, regardless of composition, exhibit the expected pattern of spacegroup I41/amd. The presence of the dioxide, UyTh(1−y)O2, can be excluded. The excess of SiO2 in the synthesis results in amorphous silica, and in some samples even in crystalline α-SiO2. Samples with a uranium mole fraction below x < 0.4 show preferred presence of crystalline α-SiO2, and samples with x > 0.4 show amorphous silica. The lattice parameters of the UxTh(1−x)SiO4 samples are listed in Table 3. The SiO2 phases were considered in the refinement process and amorphous silica was treated as a virtual phase with symmetry P m-3m. An overview of the collected X-ray pattern is shown in Figure 2. Reflections are shifted to larger 2θ values with increasing x, as the unit cell of ThSiO4, is larger than that of USiO4. In none of the pattern reflections split up. It is conclusive that no segregation into two separate UxTh(1−x)SiO4 phases takes place. Vegard like behavior has been observed for these phases by Fuchs and Gebert in 1959, yet only few compositions were investigated. While the most recent work3 reported data on the Th-rich members of the system, up to x = 0.46, Costin et al.19 synthesized UxTh(1−x)SiO4 up to x = 0.61. Above x = 0.17 the pattern of the silicate phase were accompanied with an additional dioxide phase, which dominated the product toward the U-rich members. The formation of UyTh(1−y)O2 was suppressed in these syntheses through excess amount of SiO2. Both lattice parameters as well as the cell volume decrease linearly with increasing uranium content (cf. Figure 3a and b). The gathered data as listed in Table 3 show that the complete
Figure 1. Rietveld refinement of USiO4 (a) and ThSiO4 (b). Calculated pattern (black), observed pattern (red), and difference curves (purple) are shown. (a) Reflection positions of the USiO4 (blue) and α-SiO2 phase (green) are marked; Rp = 7.89%, wRp = 8.28%. (b) Reflection positions of the ThSiO4 (blue) are marked; Rp = 6.73%, wRp = 8.74%.
Table 1. Comparison between the Lattice Constants [Å] of USiO4 Obtained in This Study and Reported in the Literature parameter
Pointeau et al.11
Hoekstra and Fuchs10
Fuchs and Gebert16
this study
a=b c
7.0135(4) 6.2669(6)
6.981(4) 6.250(5)
6.995(5) 6.263(4)
6.9842(2) 6.2606(2)
Atomic Oxygen Positions and Bond Lengths. The oxygen positions for USiO4 from measurement are y = 0.0691(2) and z = 0.2087(1), comparable to those reported by Fuchs and Gebert,16 y = 0.070(1) and z = 0.222(1) within the typical error limits. Data on the atomic positions of oxygen are rather scarce; an overview for USiO4 and ThSiO4 is given in Table 2. The resulting uranium−oxygen bond distances from this refinement are U−O1 = 2.30 Å and U−O2 = 2.44 Å, whereas the silicon−
Table 2. Overview of the Crystallographic Data of USiO4 and ThSiO4 Obtained in This Study and Reported in the Literature USiO4 atomic position Fuchs and Gebert (1959)16 Bose et al. (2009)33 [calcd] this study
bond distance [Å]
y
z
U−O1
U−O2
Si−O
0.070(1) 0.076 0.0691(2)
0.222(1) 0.212 0.2087(1) ThSiO4
2.32(8) 2.27 2.298(1)
2.52(9) 2.41 2.439(1)
1.58(9) 1.627a 1.64(1)
y
z
Th−O1
Th−O2
Si−O
0.084(1) 0.081 0.0732(1) 0.079 0.0706(8)
0.222(1) 0.232 0.2104(2) 0.215 0.2092(8)
2.46(8) 2.45 2.37(1) 2.35 2.363(5)
2.50(9) 2.51 2.47(1) 2.45 2.467(6)
1.55(9) 1.56 1.63(1) 1.627a 1.658(9)
atomic position 16
Fuchs and Gebert (1959) Sinha and Prasad (1973)34 Taylor and Ewing (1978)32 Bose et al. (2009)33 [calcd] this study a
bond distance [Å]
Fixed. 856
dx.doi.org/10.1021/es403995b | Environ. Sci. Technol. 2014, 48, 854−860
Environmental Science & Technology
Article
Table 3. Unit Cell Parameters, Cell Volume and Reliability Factors from Lebail Refinement of the UxTh(1−x)SiO4 Uranothorite Solid Solutions (Data of the End Members from Rietveld Refinement) x (U)XAS
a [Å]
c [Å]
V [Å3]
wRp [%]
Rp [%]
0 0.12(1) 0.26(1) 0.36(1) 0.42(1) 0.55(1) 0.56(1) 0.71(1) 0.75(1) 0.84(1) 0.92(1) 1
7.1816(1) 7.1129(8) 7.0949(8) 7.0775(6) 7.0697(7) 7.0450(4) 7.0410(5) 7.0180(3) 7.0105(4) 7.0017(2) 6.9952(4) 6.9842(2)
6.2946(1) 6.3301(8) 6.3194(8) 6.3103(7) 6.3106(7) 6.2904(6) 6.2844(6) 6.2802(4) 6.2680(6) 6.2636(3) 6.2727(5) 6.2606(2)
324.66(1) 321.39(5) 318.10(5) 316.09(3) 315.41(5) 312.20(4) 311.81(4) 309.32(2) 308.06(2) 307.07(1) 306.94(3) 305.38(3)
8.74 8.09 8.10 8.34 7.74 10.42 9.70 5.73 10.49 10.74 12.28 8.28
6.73 5.67 5.82 6.05 5.67 7.81 7.33 4.50 8.12 8.42 8.38 7.89
series of UxTh(1−x)SiO4 solid solutions systematically follow Vegard’s law. With increasing Th-content reflections not only shift to lower 2θ values but also experience a significant broadening (see Figure 2b). As all samples were processed under the same conditions, an influence of the reaction time or temperature as reason for this effect can therefore be ruled out. The breadth increase of the peak width can result from crystallite size (i.e., the length of the coherent scattering domains). It is reasonable to assume that disordering effects, stacking faults, etc. have a significant influence. From the peak width of the reflections crystallite sizes were determined using the Williamson−Hall method.35 These decrease from 30 nm with decreasing uranium content to ∼9 nm below x = 0.4. Costin et al.19 and also Szenknect et al.3 found that crystallite size was rather constant between 10 and 30 nm in their samples. Both observed that grain sizes, i.e., the size of the particles, were usually an order of magnitude larger than the crystallite size. Costin et al. also reported that the particle size increases with x. In the present study, the opposite can be observed. The particles of the Th-rich phases reach a
Figure 3. (a) Linear decrease of the lattice parameters a = b and c with mole fraction x. (b) The cell volume V linearly decreases with increasing mole fraction x, while the c/a ratio also increases.
much bigger size than those in the U-rich as can be observed by the SEM investigation (see Supporting Information). To provide a possible explanation for this behavior it is useful to take a look at the local environment. The USiO4 unit cell is smaller than the ThSiO4 unit cell, and for high U content the substituted Th atoms are isolated and due to their slightly larger radius fixed in smaller U position. With increasing Th content the lattice becomes wider, so for the case of x → 0 U atoms are confined within a rather large Th gap, allowing the U atom to shift out of the equilibrium position. This promotes disordering
Figure 2. X-ray diffraction patterns of UxTh(1−x)SiO4 solid solutions with increasing x (top to bottom), 15−90° 2θ range shown (a) and close up of the 2θ = 30−35° region (b). The shift of the 211 and 112 reflection is observable, and the broadening in the Th-rich samples becomes evident. Intensities from the amorphous silica and α-SiO2 are indicated with an arrow. 857
dx.doi.org/10.1021/es403995b | Environ. Sci. Technol. 2014, 48, 854−860
Environmental Science & Technology
Article
RTh−Th. For x → 1 only a small number of Th atoms appear in a dominating environment of U neighbors, and thus a → 0 and the experimentally determined value approaches the distance RTh−U. The intermediates follow eq 2. The distance distribution for x → 0 approaches accordingly the limiting distance RU−Th. Figure 5 shows an approximation with Gaussian normal distribution functions. The dashed line adheres to the Vegard line connecting RTh−Th and RU−U. The dotted line represents RTh−U.
and allows for stacking faults resulting in more crystallites and smaller crystallite size. Local Structure. The analysis of the local structure of Th1‑xUxSiO4 provides information about if the solid solution can be considered as a true mixture or if it comprises potentially a miscibility gap. EXAFS spectroscopy is well suited for such analysis because of its element selectivity.36,37 The direct coordination of U and Th in Th1−xUxSiO4 do not provide enough information, because the M−O distances (M = metal, either Th or U) in the [UO8] polyhedron and the Th−O distances in the [ThO8] polyhedron depend mainly on the central atom and do not show a significant change with varying x (see Supporting Information). In contrast, the shortest M−M distance provides more information on the local structure. Half of the [ThO8] and [UO8] are directly linked via M−O−M bonds, and half of the polyhedra are isolated from each other through [SiO4] polyhedra (Figure 1). The M−M distances indicated in Figure 4 are well suited for an analysis of the local structure in UxTh(1−x)SiO4.
Figure 5. Distances (blue and red) of UxTh(1−x)SiO4 determined with EXAFS and the weighted averages (purple). The measurements at the U L3 and Th L3 absorption edges were performed successively with the same sample.
The background of the EXAFS at Th L3 absorption edge at Ek=0 = 16310 eV is not disturbed, whereas the EXAFS at U L3 absorption edge at higher energy, Ek=0 = 17185 eV, is to a certain extent influenced by the underlying EXAFS of the Th L3 absorption edge. Thus, the experimental values of Th L3 EXAFS are closer to the expected curve than the values of the U L3 EXAFS. The weighed mean cation−cation distances vary linearly with composition and are close to the Vegard line, implying a complete solid solution without notable clustering. The data are suitable to discuss whether there potentially exists a miscibility gap and to what extent. The values are rough indicators for the size of a miscibility gap. The hypothetical case that the system is completely separated in two phases, ThSiO4 and USiO4, could be indicated by EXAFS, because it would remain equal to RTh−Th for any x and reversed, would remain equal to RU−U. The distances are expected to follow a linear trend combined with a discontinuity35 if there is a miscibility gap. The true limiting distances are related with the size of the miscibility gap. Within the experimental error limits, there is no indication of a miscibility gap in the system. In conclusion, USiO 4 and the complete range of UxTh(1−x)SiO4 solid solutions were synthesized via a hydrothermal route. To suppress the formation of the oxide phase, UyTh(1−y)O2, an excess of silica was used, employing Le Chatelier’s principle. By this means it was possible to synthesize USiO4 in good crystallinity, with no additional uranium phase present and within short reaction times. Significant amounts of α-SiO2 and amorphous silica in the product, however, have to be taken into account. Quartz reflections are easily distinguished from those of USiO4 in the XRD, and furthermore silica does not disturb the EXAFS spectra; therefore this additional phase does not cause problems in the analysis of the structure of USiO4.
Figure 4. Schematic drawing of the metal (M) coordination in UxTh(1−x)SiO4. RM−M represents the three possible distances RTh−Th and RTh−U, and RU−U under discussion.
The solid solution comprises three types of these M−M distances, RTh−Th, RTh−U ,and RU−U. (It should be metioned that RTh−U and RU−Th are equivalent.) The experimentally determined average distance is given by R̅ =
aRTh − Th + bRTh − U + cRU − U a+b+c
(1)
The diffraction measurements reveal the average distance given by eq 1, whereas EXAFS reveals only two types of distances: at the Th L3 absorption edge the distances RTh−Th and RTh−U and at the U L3 absorption edge the distances RU−U and RU−Th. RTh−Th is given by the limiting species ThSiO4, RU−U by USiO4. Taking into account only the interaction with the next neighboring metal atoms, it is reasonable to assume that RTh−U ≈ (RTh−Th + RU−U)/2. The experimental determined distances at the Th L3 absorption edge are R̅ ThL3 =
aRTh − Th + bRTh − U a+b
(2)
In other words, at the Th L3 absorption edge there is no information on RU−U. In the structure of UxTh(1−x)SiO4 with x = 0, i.e., ThSiO4, there exist exclusively distances of RTh−Th, and thus the experimentally determined value is identical with 858
dx.doi.org/10.1021/es403995b | Environ. Sci. Technol. 2014, 48, 854−860
Environmental Science & Technology
Article
(5) Janeczek, J.; Ewing, R. C. Dissolution and alteration of uraninite under reducing conditions. J. Nucl. Mater. 1992, 190, 157−173. (6) Min, M.; Fanga, C.; Fayek, M. Petrography and genetic history of coffinite and uraninite from the Liueryiqi granite-hosted uranium deposit, SE China. Ore Geol. Rev. 2005, 26, 187−197. (7) Janeczek, J.; Ewing, R. C.; Oversby, V. M.; Werme, L. O. Uraninite and UO2 in spent nuclear fuel: a comparison. J. Nucl. Mater. 1996, 238 (1), 121−130. (8) Langmuir, D. Uranium solution-mineral equilibria at low temperatures with applications to sedimentary ore deposits. Geochim. Cosmochim. Acta 1978, 42, 547−569. (9) Frondel, C.; Collette, R. L. Hydrothermal synthesis of zircon, thorite and huttonite. Am. Mineral. 1957, 42, 759−765. (10) Hoekstra, H. R.; Fuchs, L. H. Synthesis of coffinite - USiO4. Science 1956, 123, 105. (11) Pointeau, V.; Deditius, A. P.; Miserque, F.; Renock, D.; Becker, U.; Zhang, J.; Clavier, N.; Dacheux, N.; Poinssot, C.; Ewing, R. C. Synthesis and characterization of coffinite. J. Nucl. Mater. 2009, 393 (3), 449−458. (12) Mumpton, F. A.; Roy, R. Hydrothermal stability studies of the zircon-thorite group. Geochim. Cosmochim. Acta 1961, 21 (3−4), 217− 238. (13) Lungu, S. Some theoretical aspects in the spinodal decomposition and crystallization of silicate melts. experimental results for the UO2-SiO2 system. J. Nucl. Mater. 1973, 48, 165−171. (14) Robit, V.; Poinssot, C.; Spahiu, K.; Cui, H. Mater. Res. Soc. Symp. Proc. 2009, 932, 489. (15) Goldschmidt, V. M. Geochemistry; Oxford University Press: Oxford, U.K., 1954. (16) Fuchs, L. H.; Gebert, E. X-ray studies of synthetic coffinite, thorite and uranothorites. Am. Mineral. 1958, 43, 243−248. (17) Förster, H.-J. Composition and origin of intermediate solid solutions in the system thorite-xenotime-zircon-coffinite. Lithos 2006, 88 (1−4), 35−55. (18) Ferriss, E. D. A.; Ewing, R. C.; Becker, U. Simulation of thermodynamic mixing properties of actinide-containing zircon solid solutions. Am. Mineral. 2010, 95 (2−3), 229−241. (19) Costin, D. T.; Mesbah, A.; Clavier, N.; Dacheux, N.; Szenknect, S.; Raveaux, J. How to explain the difficulties in the coffinite synthesis from the study of uranothorite? Inorg. Chem. 2011, 50, 11117−11126. (20) Dacheux, N.; Brandel, V.; Genet, M. Synthesis and characterization of mixed-calence uranium orthophosphate−U(UO2)(PO4)2. New J. Chem. 1996, 19, 15−25. (21) Dacheux, N.; Brandel, V.; Genet, M.; Bak, K.; Berthier, C. Solid solutions of uranium and thorium phosphates: synthesis, characterization and x-ray photoelectron spectroscopy. New J. Chem. 1996, 20, 301−310. (22) LeBail, A.; Duroy, H.; Fourquet, J.-H. Ab-initio structure determination of LiSbWO6 by X-ray powder diffraction. Mater. Res. Bull. 1988, 23, 447−452. (23) Petricek, V.; Dusek, M. and Palatinus, L. Jana2006 the crystallographic computing system. Technical report; Institute of Physics: Praha, Czech Republic, 2006. (24) Larson, A. C. and von Dreele, R. B. General structure analysis system (gsas). Technical Report LAUR 86-748; Los Alamos National Laboratory: Los Alamos, NM, 2000. (25) Toby., B. H. Expgui, a graphical user interface for gsas. J. Appl. Crystallogr. 2001, 34, 219−221. (26) Ressler, T. WinXAS: A XAS data analysis program under MS windows. J. Synchrotron Radiat. 1998, 5, 118. (27) Ravel, B.; Newville, M. ATHENA, ARTEMIS, HEPHAESTUS: data analysis for X-ray absorption spectroscopy using IFEFFIT. J. Synchrotron Radiat. 2005, 12, 537−541. (28) George, G. N.; Pickering, I. J EXAFSPAK, a suite of computer programs for analysis of X-ray absorption spectra; Stanford Synchrotron Radiation Laboratory: Stanford, 2000. (29) Ankudinov, A. L.; Ravel, B.; Rehr, J. J.; Conradson, S. D. Realspace multiple-scattering calculation and interpretation of x-rayabsorption near-edge structure. Phys. Rev. B. 1998, 58, 7565−7576.
According to X-ray diffraction data UxTh(1−x)SiO4 forms a complete solid solution under hydrothermal conditions. Refined lattice parameters of the UxTh(1−x)SiO4 phases as well as cell volume follow Vegard’s Law and regularly decrease with increasing uranium content. This observation is well supported by the small difference in ionic radii of U4+ (1.00 Å) compared to that of Th4+ (1.05 Å).38 Results from EXAFS measurements of Th L3 and U L3 edges feature a continuous correlation between increasing uranium content and decreasing metal−metal bond lengths. This confirms the presence of a complete solid solution in the system USiO4−ThSiO4. We are now able to complete the data for this system and moreover offer a congruent set of measurements as a useful base for further investigation into the involved kinetics of formation.
■
ASSOCIATED CONTENT
S Supporting Information *
The comparison of initial U/Th ratio and the U/Th ratio measured in the products (S1); a further discussion on the calculation of the mole fractions from XANES edge and the comparison of these to the mole fractions determined by SEMEDS (S2), as well as SEM micrographs of the prepared samples (S3 and S4); comparison of the bond lengths of USiO4 and ThSiO4 obtained from XRD and EXAFS (S5); comparison of lattice parameters a = b and c of the UxTh1−xSiO4 solid solutions obtained in this study and reported by Costin et al. and Szenknecht et al. (S6, S7); EXAFS spectra and FT magnitudes (S8 and S10); tables containing the data obtained from EXAFS (S9 and S11); and an illustration of U−O1/2 and Th−O1/2 distances with varying x (S12). This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS Beam time at ROBL beamline (ESRF) was granted through ACTINET-i3 (proposal No. AC4-JRP04). Financial support was granted by German Federal Ministry of Education and Research (BMBF), ImmoRad Project (Grant No. 02NUK019C). The authors would like to thank Prof. Florian Kraus from TU Munich for providing the uranium metal turnings and Aline Ritter for the ICP-MS measurements.
■
REFERENCES
(1) Stieff, L. R.; Stern, T. W.; Sherwood, A. M. Coffinite, a uranous silicate with hydroxyl substitutions: A new mineral. Am. Mineral. 1956, 41 (9 and 10), 675−689. (2) Hemingway, B. S. Thermodynamic properties of selected uranium compounds and aqueous species at 298.15 K and 1 bar and at higher temperatures. Preliminary models for the origine of coffinite deposits; U.S. Geological Survey: Reston, VA, 1982; Open-File Report 82-619. (3) Szenknect, S.; Costin, D. T.; Clavier, N.; Mesbah, A.; Poinssot, C.; Vitorge, P.; Dacheux, N. From uranothorites to coffinite: A solid solution route to the thermodynamic properties of USiO4. Inorg. Chem. 2013, 52 (12), 6957−6968. (4) Arcos, D.; Prez del Villar, L.; Bruno, J.; Domnech, C. Geochemical modeling of the weathering zone of the “Mina Fe” U deposit (Spain): A natural analogue for nuclear spent fuel alteration and stability processes in radwaste disposal. Appl. Geochem. 2008, 23 (4), 807−821. 859
dx.doi.org/10.1021/es403995b | Environ. Sci. Technol. 2014, 48, 854−860
Environmental Science & Technology
Article
(30) Giacovazzo, C. Fundamentals in Crystallography; Oxford University Press: Oxford, U.K., 1992. (31) (a) Brown, G. E.; Gibbs, G. V.; Ribbe, P. H. The nature and the variation in length of the Si-O and Al-O bonds in framework silicates. Am. Mineral. 1969, 54, 1044−1061. (b) Brown, G. E.; Gibbs, G. V. Oxygen Coordination and the Si-O Bond. Am. Mineral. 1969, 54, 1528−1539. (32) Taylor, M.; Ewing, R. C. The crystal structures of the ThSiO4 polymorphs: Huttonite and Thorite. Acta Crystallogr. 1978, 34, 1074− 1079. (33) Bose, P. R.; Mittal, R.; Chaplot, S. L. Lattice dynamics and high pressure phase stability of zircon structured natural silicates. Phys. Rev. B 2009, 79 (17), 174301. (34) Sinah, D. P.; Prasad, R. On the synthetic preparation and lattice structure of thorite. J. Inorg. Nucl. Chem. 1973, 35, 2614−2615. (35) Williamson, G. K.; Hall, W. H. X-ray line broadening from filed aluminium and wolfram. Acta Metall. 1953, 1 (1), 22−31. (36) Scheinost, A.; Stanjek, H.; Schulz, D. G.; Gasser, U.; Sparks, D. Structural environment and oxidation state of Mn in goethite-groutite solid-solutions. Am. Mineral. 2001, 86, 139−146. (37) Hubert, S.; Purans, J.; Heisbourg, G.; Moisy, P.; Dacheux, N. Local Structure of Actinide Dioxide Solid Solutions Th1‑xUxO2 and Th1‑xPuxO2. Inorg. Chem. 2006, 45, 3887−3894. (38) Shannon, R. D. Ion radii. Acta Crystallogr. 1976, A32, 751−767.
860
dx.doi.org/10.1021/es403995b | Environ. Sci. Technol. 2014, 48, 854−860
