Investigations into the Polymorphs and Hydration Products of UO3 Lucas E. Sweet, Edgar C. Buck, Charles H. Henager Jr., Shenyang Hu, David E. Meier, Shane M. Peper, Jon M. Schwantes, Yin-Fong Su, Robert L. Sams, Thomas A. Blake and Timothy J. Johnson* Pacific Northwest National Laboratory, P.O. Box 999, Richland, WA 99352
Thomas J. Kulp, Ricky L. Sommers, Joshua D. Sugar, and Jeffrey D. Chames Sandia National Laboratory - Livermore, Livermore, CA 9455 ABSTRACT This work focuses on progress in gaining a better understanding of the polymorphic nature of the UO3 and UO3-water system; one of several important materials associated with the nuclear fuel cycle. The UO3-water system is complex and has not been fully characterized, even though these species are common throughout the fuel cycle. For example, most production schemes for UO3 result in a mixture of up to six different polymorphic phases, and small differences in these conditions will affect phase genesis that ultimately results in measureable changes to the end product. Here we summarize our efforts to better characterize the UO3-water system with optical techniques for the purpose of developing some predictive capability of estimating process history and utility, e.g. for polymorphic phases of unknown origin. Specifically, we have investigated three industrially relevant production pathways of UO3 and discovered a previously unknown low temperature route to β-UO3. Powder x-ray diffraction and optical spectroscopies were utilized in our characterization of the UO3-water system. Pure phases of UO3, its hydrolysis products and starting materials were used to establish optical spectroscopic signatures for these compounds. Preliminary aging studies were conducted on the αand γ- phases of UO3. Keywords: UO3, uranium trioxide, polymorphs, Raman spectroscopy. 1. INTRODUCTION There are several uranium chemical species that serve as key components in the nuclear fuel cycle whose identities and properties are well known. These include the triuranium octaoxide U3O8 (the major component of “yellowcake”), the oxide products UO3 and UO2, as well as the fluoride products UF4 and UF6 resulting from conversion facilities. One of the key components is clearly UO3; it is found in the earlier stages of the nuclear fuel cycle in the milling, refinement and conversion processes that precede isotope enrichment. In addition, spent nuclear fuel usually undergoes a refinement process that produces the intermediate species UO3 from the denitrification of uranyl nitrate hexahydrate (UNH) after purification by solvent extraction. The resulting uranium trioxide, UO3, is reduced to UO2 before the fluoride conversion processes. Of these key species, some are relatively well understood in terms of their phases and uranium coordination environments, however the polymorphic nature of UO3 is not as well understood. The lack of knowledge as to specifics of these polymorphs, as well as how the phases are produced from one another, and their relationship to the hydrolysis products which some phases (readily) form has proven to be something of a limitation as well as a continuing source of discussion in the literature. This work has focused on a more thorough investigation of the pure phases of UO3 and related compounds as well as various analytical methods that can be used to characterize the phases. It turns out that the structural phase that UO3 adopts is a direct result of the production route used to produce the material, and is thus a characteristic of the process. With an understanding of the polymorphic nature of UO3 and with *
; phone 509-372-6058
Chemical, Biological, Radiological, Nuclear, and Explosives (CBRNE) Sensing XIII, edited by Augustus Way Fountain III, Proc. of SPIE Vol. 8358, 83581R · © 2012 SPIE · CCC code: 0277-786X/12/$18 · doi: 10.1117/12.919706
the structural forms of the end products, one could potentially use such information to discern the production method. In order for this to happen, however, it is necessary to be able to reliably prepare and characterize the different phases, as well as the sundry hydrolysis and other transformation products of the UO3 system. To that end, we have investigated three industrially relevant production pathways of UO3 and discovered a previously unknown low temperature route to the phase known as β-UO3. The three routes of UO3 production we have chosen to investigate include synthesizing UO3 from any of three different starting products including uranyl nitrate hexahydrate (UNH) UO2(NO3)2•6H2O, ammonium uranyl carbonate (NH4)4UO2(CO3)3, or meta-studtite, UO4•2H2O. Producing UO3 from UO2(NO3)2•6H2O is a common route to UO3 when reprocessing spent fuel. Another route for producing UO3 during ore processing1 is via the precipitation and conversion from ammonium uranyl carbonate, (NH4)4UO2(CO3)3. The third route that we investigated produces UO3 from UO4•2H2O (meta-studtite) or UO4•4H2O (studtite): This method is typically used when obtaining UO3 when processing and refining uranium from ores. As mentioned above, each of the three methods (starting materials) results in the formation of a distinct polymorphic phase of the trioxide species UO3, vide infra. Because many of the UO3 phases are susceptible to hydrolysis or hydration under ambient conditions (humid air at room temperature), we also examined the formation of certain UO3 hydrolysis products by intentionally synthesizing and characterizing some of the hydrolyzed species. The objective of this research is to provide methods and information so as to identify and discriminate amongst these tightly inter-related species. 2. MATERIAL PREPARATION AND CHARACTERIZATION METHODS In order to investigate the properties of different UO3 polymorphs, the samples were derived, primarily by heating, from different starting reagents. However, most of the reagents all had to be prepared in our laboratory, and their identities also had to be confirmed in all cases. In addition to preparing the different phases of UO3, preliminary aging studies were also conducted on the pure α- and γ- phases of UO3 to understand the times scales associated with metamorphosis / decomposition of the pure phases. 2.1 Material Preparation In order to prepare the polymorphs of UO3, the separate samples of UO2(NO3)2•6H2O, UO4•2H2O and (NH4)4UO2(CO3)3 were individually heated to a variety of temperatures ranging from 350°C to 500°C for 60 hrs. However, these samples had to be prepared individually. The first UO3 synthesis involved the heating of UNH, UO2(NO3)2•6H2O, and this sample was simply used as supplied by the chemical vendor International Bioanalytical Laboratories, Incorporated (IBI). The reagent was verified as pure via XRD as described below. The product had the rich orange color characteristic of the γ-phase of UO3 solid. The second route to preparing UO3 we investigated involves the conversion of UO4•2H2O (meta-studtite), so the metastudtite was prepared separately in our laboratory. This was accomplished by quantitatively adding 3 mL of 30% aqueous H2O2 (from Sigma-Aldrich) to 5 mL of a 1 M aqueous solution of UO2(NO3)2 (obtained as a solid from IBI chemicals, 99.9% pure). A light yellow precipitate formed upon addition of the hydrogen peroxide. The solution was heated to 80°C for 24 hours to dry the sample. Powder XRD of the light yellow reaction product was used to confirm the yellow powder was in fact UO4•2H2O. In the third synthetic route to UO3, the ammonium uranyl carbonate (NH4)4UO2(CO3)3 used to make UO3 had to first be synthesized by adding 1.1965g of (NH4)2CO3 (Sigma-Aldrich 99.9%) to 2.17 mL of a 0.96 M aqueous solution of UO2(NO3)2 (IBI 99.9%). The yellow precipitate was allowed to settle and the water was decanted off. The wet solid was baked in a furnace at 80°C for 3 hours to remove the remaining water. By comparing the XRD powder pattern collected on the product to the powder pattern calculated from the crystal structure of (NH4)4UO2(CO3)3, the product was confirmed to be pure. In terms of the hydrated species, two of the most common hydrolysis products of UO3 are α-UO2(OH)2 (uranyl hydroxide) and (UO2)4O(OH)6•5H2O (meta-schoepite). Therefore, pure α-UO2(OH)2 was prepared by stirring a sample
of γ-UO3 in water (forms a slurry) and then heating to 80°C for 24 hours to drive off the excess (unreacted) water. (UO2)4O(OH)6•5H2O, a further hydrated form of UO3, was prepared by stirring a sample of γ-UO3 in water and allowing the water to evaporate at room temperature. Each of these samples was analyzed using powder XRD, Raman microscopy and fluorescence microscopy. Both of the hydrated forms of UO3 were coarse (large grain) powders that were bright yellow in color in contrast to the very finely divided (small grain size) dark orange powder that is typical for the dehydrated UO3. 2.2 Analytical Methods The purity of the phases of the starting materials, the UO3 polymorphs, as well as the UO3/UO3-water hydrolysis products were established via x-ray diffraction necessary so as to serve as “ground truth” before recording and establishing the optical spectroscopic signatures for these key fuel cycle compounds. Powder XRD patterns were collected on all starting materials as well as the different phases of UO3 and the hydrolysis products using a Rigaku Ultima IV powder diffractometer. The Ultima IV uses a Cu Kα graphite monochromated source and is outfitted with a D/Tex silicon strip detector. Typically, the collection of the XRD patterns took 30 minutes per material. Fits were determined by Rietveld refinements using the TOPAS software package. An example of an experimental XRD pattern and the reference diffraction pattern to which it was matched is seen in Figure 1 for the hydrolysis product α-UO2(OH)2. The diffraction pattern was matched via Rietveld refinement to the reference pattern of α-UO2(OH)2 from the literature data of Weller et al.]
Figure 1. X-ray Diffraction (XRD) pattern plot for the hydrolysis product α-UO2(OH)2 (uranyl hydroxide) synthesized by adding H2O to γ-UO3 and heating to 80°C for 24 hours (black). The blue pattern was calculated from the published crystal structure of α-UO2(OH)2.
The uranium samples were also analyzed using a DeltaNu ExamineR Raman spectrometer, which uses a 785 nm laser for its excitation source. The ExamineR module was attached to an Olympus BX51 compound microscope equipped with a 10x objective. Raman spectra were collected in a dispersive configuration with a spectral range of 200-2000 cm-1 Raman shift and a resolution of 5 cm-1. For the Raman spectra, a typical collection time per spectrum was approximately one minute. 3. GENERATION AND CHARACTERIZATON OF DISCRETE UO3 PHASES 3.1 Uranyl Nitrate Hexahydrate UO2(NO3)2•6H2O Used to Form γ-UO3 The first UO3 synthetic route that we explored was the conversion of UO2(NO3)2•6H2O to form γ-UO3. After heating 1g of UO2(NO3)2•6H2O at 350°C for 60 hours, the XRD powder pattern of the orange powder sample matched that of the crystal structure reported for pure γ-UO3  and the sample had the rich orange color characteristic of the γ-phase of this material. The sample showed no impurities via XRD analysis. The sample was then further heated to 400°C for 60 hrs.
The XRD powder pattern of the sample after heating at 400°C still appeared to be that of the γ-UO3 phase. This behavior is consistent with what has been previously reported. The resulting Raman spectrum of the pure γ-UO3 sample is shown in lower trace of Fig. 2; we believe these represent some of the few spectra of pure γ-UO3, as this species is quite susceptible to hydrolysis.
Figure 2. Raman spectra of (UO2)4O(OH)6y5H2O (top), α-UO2(OH)2 (middle) and γ-UO3 (bottom)
However, because the γ-UO3 isomorph is known to readily hydrolyze under ambient conditions (humid air, room temperature),[12-16] the Raman spectra for the related materials α-UO2(OH)2 (middle) and (UO2)4O(OH)6•5H2O (metaschoepite, top trace) are also plotted in Figure 2. We note that both hydrolysis products were confirmed as pure by XRD and that both were prepared starting from the same γ-UO3 material described above. While the spectroscopic distinction between the spectra of γ-UO3 and the hydrolysis products is clear, the spectroscopic distinctions between the two hydrolysis products are more subtle. The Raman bands at 840 cm-1 for α-UO2(OH)2 and 841 cm-1 for (UO2)4O(OH)6•5H2O represent the well-known ν1 symmetric stretching frequency of the uranyl (O=U=O)2+ cation, and are fairly typical UO22+ frequencies for such species, as both have discrete uranyl ions in the crystal.[17-19] That is to say, the UO22+ cation has only relatively weakly interactions with other atoms / ions in the unit cell and the vibrational frequencies approximate those of the vapor-phase species. This is because for both the uranyl ν1 symmetric stretch and the ν3 anti-symmetric stretch it has long been known that the peaks can shift strongly depending on the local environment of the UO22+ cation.[17, 20] The peaks at 840 and 841 cm-1 of the two hydrolysis products are typical of the uranyl symmetric frequency in many uranyl-bearing compounds and minerals such as the 815 cm-1 mode reported for kamotoite, the ~840 cm-1 frequency of euxenite, and the 865 cm-1 value reported for UO2F2.[11, 23] On the other hand, the γ-UO3 starting reagent does not have clearly isolated uranyl groups as there are different uranium site symmetries with different degrees of coordination to the neighboring oxygen atoms. Its vibrational spectrum is thus not as well understood, but has as its strongest band the peaks at 767, 484 and 339 cm-1.[11, 24] We can thus deduce that as γ-UO3 hydrolyzes, it is marked by the disappearance of these vibrational bands and the emergence of a Raman band associated with the symmetric stretching of a discrete UO22+ cation in the typical 830 to 870 cm-1 range as seen in Figure 1. 3.2 Meta-studtite UO4•2H2O Used to Form α-UO3 The meta-studtite (UO4•2H2O) sample was heated to prepare a sample of nearly pure α-UO3 as confirmed by powder xray diffraction. Unfortunately, we could not obtain a reasonable Raman spectrum of α-UO3 because of the fluorescence resulting from the 785 nm laser used for Raman excitation. It is hoped that switching to 1064 nm excitations will abate or eliminate this problem.[25, 26] Interestingly, during some ancillary florescence studies, the bulk of the α-UO3 sample did not fluoresce in the visible/near IR region when exposed to excitation light in the 375-560 nm spectral band. Because
the sample is light yellow in color, this indicates that it absorbs light in the violet and blue, possibly into green wavelengths as well. It is anticipated that it will have minimal absorbance at the 1064 nm wavelength. As an ancillary study, the sample of α-UO3 was also allowed to age in a capped vial for 45 days. From Rietveld analysis of the resulting powder XRD pattern it was determined that the sample consisted of 90% α-UO2(OH)2, 5% α-UO3 and 5% (UO2)4O(OH)6•5H2O. Efforts are currently underway to establish the rate of hydrolysis under a given set of conditions. 3.3 Ammonium Uranyl Carbonate (NH4)4UO2(CO3)3 Used to Form Amorphous UO3, β-UO3, and α,β-UO3. A series of preparation temperatures was tried for the formation of UO3 from the pure (NH4)4UO2(CO3)3 starting material. Figure 3 shows the Raman spectra that resulted from the sequential heating of the ammonium uranyl carbonate sample to 350 °C and 450 °C, with the spectra recorded after having cooled back to room temperature. It has previously been reported that (NH4)4UO2(CO3)3 forms amorphous UO3 at 400°C. From the XRD powder patterns (not shown here), however, the products that formed after heating (NH4)4UO2(CO3)3 to 350°C are largely amorphous. The XRD signal-to-noise ratio on the sample prepared at 350°C was very low indicating a large degree of amorphous product. However, the discernible peaks for the sample prepared at 350°C match the crystal structure of β-UO3 quite well. This is a previously unknown route to relatively pure β-UO3 production. After further heating the sample to 450°C, the resulting product was relatively pure with 18% α-UO3 and 82% β-UO3 as determined by Rietveld refinement using TOPAS software package. We note that for the top Raman trace in Figure 3 is likely due primarily to the β-UO3 since αUO3 fluoresces and decomposes under the 785 nm laser light.
Figure 3. Raman spectra of starting reagent and the products formed after heating (NH4)4UO2(CO3)3 to 350 degrees for 60 h (then cooling back to room temperature). For the 350 oC sample, XRD indicated β-UO3 / largely amorphous β-UO3. The top trace is the Raman spectrum of product formed after heating to 450 degrees C for 60 h and then cooling: 82% β-UO3 and 18% α-UO3.
4. SUMMARY AND CONCLUSIONS
This study has generated three different crystalline phase of the key fuel cycle compound uranium trioxide, UO3. X-ray diffraction patterns have shown pure compounds of α- UO3, β-UO3, and γ-UO3 with a new synthetic route to β-UO3 being reported for the first time. We have also synthesized and purified two of the more common hydrolysis products of UO3, namely (UO2)4O(OH)6•5H2O (meta-schoepite) and α-UO2(OH)2 (uranyl hydroxide). These were both synthesized and characterized as pure phases by XRD and Raman spectroscopy. We have recorded Raman spectra of some of these pure phases for the first time. It is hoped that by employing a 1064 nm Raman system that we will be able record the Raman spectra of all known phases of UO3. We hope to carry on this research to elucidate signatures of the different production process and storage conditions of UO3.
5. ACKNOWLEDGMENTS This research was supported in part by the National Technical Nuclear Forensics Center (NFNFC, a department of the U.S. Department of Homeland Security), and was co-funded by NA-22 in the National Nuclear Security Administration / Office of Nonproliferation & Verification Research and Development. The work was conducted at the Pacific Northwest National Laboratory, a multiprogram national laboratory operated by Battelle for the U.S. Department of Energy under Contract DE-AC05-76RL01830. We thank our sponsors for their support.
REFERENCES               
M. Benedict, T. H. Pigford, and H. W. Levi, Nuclear Chemical Engineering (2nd Edition) McGraw-Hill, (1981). I. Grenthe, J. Drożdżyński, T. Fujino et al., Uranium*: The Chemistry of the Actinide and Transactinide Elements, Springer Netherlands, (2011). H. R. Hoekstra, and S. Siegel, “The Uranium-Oxygen System - U3O8-UO3,” Journal of Inorganic & Nuclear Chemistry, 18, 154-165 (1961). E. H. P. Cordfunke, and A. A. Van Der Giessen, “Pseudomorphic Decomposition of Uranium Peroxide into UO3,” Journal of Inorganic and Nuclear Chemistry, 25(5), 553-555 (1963). L. Eary, and L. Cathles, “A Kinetic Model of UO2 Dissolution in Acid, H2O2 Solutions That Includes Uranium Peroxide Hydrate Precipitation,” Metallurgical and Materials Transactions B, 14(3), 325-334 (1983). R. Graziani, G. Bombieri, and E. Forsellini, “Crystal Structure of Tetra-ammonium Uranyl Tricarbonate,” Journal of the Chemical Society, Dalton Transactions(19), 2059-2061 (1972). TOPAS v 4.2 Bruker AXS, (2009). J. C. Taylor, and H. J. Hurst, “Hydrogen Atom Locations in Alpha-Form and Beta-Form of Uranyl Hydroxide,” Acta Crystallographica Section B-Structural Crystallography and Crystal Chemistry, B 27(OCT15), 2018-& (1971). R. Engmann, and P. M. D. Wolff, “Crystal Structure of Gamma-U3,” Acta Crystallographica, 16(10), 993-& (1963). H. R. Hoekstra, and S. Siegel, “Chemistry and Crystallography of Uranium Oxide Systems,” Journal of Inorganic & Nuclear Chemistry, 7(1-2), 174-175 (1958). D. P. Armstrong, R. J. Jarabek, and W. H. Fletcher, “Micro-Raman Spectroscopy of Selected Solid UxOyFz Compounds,” Applied Spectroscopy, 43(3), 461-468 (1989). V. J. Wheeler, R. M. Dell, and E. Wait, “Uranium Trioxide and the UO3 Hydrates,” Journal of Inorganic & Nuclear Chemistry, 26(11), 1829-1845 (1964). P. Taylor, D. D. Wood, A. M. Duclos et al., “Formation of Uranium Trioxide Hydrates on UO2 Fuel in Air Steam Mixtures Near 200-Degrees-C,” Journal of Nuclear Materials, 168(1-2), 70-75 (1989). M. T. Weller, M. E. Light, and T. Gelbrich, “Structure of Uranium(VI) Oxide Dihydrate, UO3.2H2O; Synthetic Meta-schoepite (UO2)4O(OH)6.5H2O,” Acta Crystallographica Section B, 56(4), 577-583 (2000). R. J. Finch, F. C. Hawthorne, and R. C. Ewing, “Structural Relations Among Schoepite, Metaschoepite and "Dehydrated Schoepite",” Canadian Mineralogist, 36, 831-845 (1998).
            
H. R. Hoekstra, and S. Siegel, “Uranium Trioxide-Water System,” Journal of Inorganic & Nuclear Chemistry, 35(3), 761-779 (1973). M. Gál, P. L. Goggin, and J. Mink, “Vibrational Spectroscopic Studies of Uranyl Complexes in Aqueous and Non-aqueous Solutions,” Spectrochimica Acta Part A: Molecular Spectroscopy, 48(1), 121-132 (1992). S. P. McGlynn, and J. K. Smith, “The Electronic Structure, Spectra, and Magnetic Properties of Actinyl Ions: Part I. The Uranyl Ion,” Journal of Molecular Spectroscopy, 6(0), 164-187 (1961). R. L. Frost, J. Cejka, and M. L. Weier, “Raman Spectroscopic Study of the Uranyl Oxyhydroxide Hydrates: Becquerelite, Billietite, Curite, Schoepite and Vandendriesscheite,” Journal of Raman Spectroscopy, 38(4), 460-466 (2007). H. R. Hoekstra, “Uranium-Oxygen Bond Lengths in Uranyl Salts: Uranyl Fluoride and Uranyl Carbonate,” Inorganic Chemistry, 2(3), 492-495 (1963). R. L. Frost, M. L. Weier, J. ô. ƒåejka et al., “Raman Spectroscopy of Uranyl Rare Earth Carbonate Kamotoite(Y),” Spectrochimica Acta, Part A, 65(3‚Äì4), 529-534 (2006). R. L. Frost, S. J. Palmer, and B. J. Reddy, “Raman Spectroscopic Study of the Uranyl Titanate Mineral Euxenite (Y,Ca,U,Ce,Th) (Nb,Ta,Ti)2O6,” Journal of Raman Spectroscopy, 42(5), 1160-1162 (2011). A. J. P. R. Kips, M. R. Houlton, A. Leenaers, J. D. Mace, O. Marie, F. Pointurier, E. A. Stefaniak, P. D. P. Taylor, S. Van den Berghe, P. Van Espen, R. Van Grieken, R. Wellum, “Determination of Fluorine in Uranium Oxyfluoride Particles as an Indicator of Particle Age,” Spectrochimica Acta, Part B 64, 199-207 (2009). S. D. Gabelnick, G. T. Reedy, and M. G. Chasanov, “Infrared Spectra of Matrix-isolated Uranium Oxide Species. II. Spectral Interpretation and Structure of UO3,” The Journal of Chemical Physics, 59(12), 6397-6404 (1973). J. F. Kelly, T. A. Blake, B. A. Bernacki et al., “Towards a Practical Raman Line Imager for Trace Analyses of Explosive Residues,” International Journal of Spectroscopy, submitted, (2012). Y.-F. Su, T.J. Johnson, K.H. Jarman, B.M. Kunkel, J.C. Birnbaum, A.G. Joly, E.G. Stephan, R.G. Tonkyn, R.G. Ewing, and G.C. Dunham, , “Demonstrated Wavelength Portability of Raman Reference Data for Explosives and Chemical Detection,” International Journal of Spectroscopy, submitted, (2012). C. Greaves, and B. E. F. Fender, “Structure of Alpha-UO3 By Neurton and Electron-Diffraction,” Acta Crystallographica Section B-Structural Science, 28(DEC15), 3609-3614 (1972). P. C. Debets, “Structure of Beta-UO3,” Acta Crystallographica, 21, 589-& (1966).