Evidence of Trivalent Am Substitution into U3O8 - Inorganic Chemistry

Article pubs.acs.org/IC

Evidence of Trivalent Am Substitution into U3O8 Marie Caisso,†,‡,∇ Pascal Roussel,§ Christophe Den Auwer,∥ Sébastien Picart,‡ Christoph Hennig,⊥ Andreas C. Scheinost,⊥,# Thibaud Delahaye,*,‡ and André Ayral∇ †

CEA, DEN, DTEC/SECA/LFC, F-30207 Bagnols-sur-Cèze Cedex, France CEA, DEN, DRCP/SERA/LCAR, F-30207 Bagnols-sur-Cèze Cedex, France § Unité de Catalyse et Chimie du Solide, UMR 8012 CNRS, 59652 Villeneuve d’Ascq Cedex, France ∥ Institut de Chimie de Nice, Université Nice Sophia Antipolis, UMR 7272, F-06108 Nice cedex 2, France ⊥ HZDR, Institute of Resource Ecology, 01314 Dresden, Germany # Rossendorf Beamline, ESRF, 38043 Grenoble, France ∇ Institut Européen des Membranes, UMR 5635 CNRS-ENSCM-UM2, CC047, Université Montpellier 2, F-34095 Montpellier Cedex 5, France ‡

ABSTRACT: U3O8 is considered to be the most stable phase for uranium oxide. Its structural properties must be accurately understood to foresee and manage aspects such as its leaching behavior when spent nuclear fuel is stored in an oxidative environment. Moreover, as fuel irradiation causes the formation of fission products and activation products such as plutonium and minor actinides, it is probable that U3O8 will be mixed with other chemical elements under real conditions of oxidation. The storage issue can be extended to americium transmutation, where the irradiated compounds are mixed oxides composed of uranium and americium. This study thus focused on determining the structural properties of a solid solution containing uranium and trivalent americium (U/Am ratio = 90/10) and synthesized so as to obtain conventional U3O8 oxide. This paper presents the possibility of combining trivalent americium with uranium in a U3O8 mixed oxide for the first time, despite the high valence and atomic ratio differences, and proposes novel structural arrangements. X-ray diffraction measurements reveal americium substitution in U3O8 uranium cationic sites, leading to phase transformation into a U3O8 high-temperature structure and general lattice swelling. X-ray absorption near-edge spectroscopy and extended X-ray absorption fine structure experiments highlight an excess of U+VI organized in uranyl units as the main consequence of accommodation.

■

INTRODUCTION The oxidation of uranium dioxide (UO2) into U3O8 has been studied in the field of nuclear fuel fabrication for many years. It was first investigated in the context of nuclear fuel reprocessing and is now considered in the waste management domain and for spent fuel behavior in dry storage conditions.1,2 Over 473 K, UO2 slowly oxidizes into U3O8.3 Focusing on spent fuels, even if most remains composed of UO2, the presence of additional highly radioactive radionuclides such as fission products or minor actinides (MA) means a sufficient heat load can be generated for UO2 to be oxidized into U3O8. This accelerates intermediate phase transformations, such as U4O9 and U3O7.4 The issue also concerns compounds fabricated in the field of MA transmutation,5,6 where the latter can be integrated in uranium oxides to form, in the case of americium (Am), dense U1−xAmxO2±δ pellets. The storage and post-irradiation behaviors of these compounds, under accidental oxidative conditions where heat load could not be controlled, remain unknown, but they have been identified as crucial factors for future experimental irradiations. In parallel, the U 3 O 8 © XXXX American Chemical Society

compound is also used as mineral pore-former in nuclear fuel fabrication.7,6 Impurity solubility in this oxide thus needs to be well-known to anticipate any chemical reaction between U3O8 as pore-former and other actinides found in nuclear fuels, such as MOX fuel8,9 or compounds used for transmutation. Examining first the literature on pure U3O8 at room temperature (RT), its crystalline structure has been determined to be orthorhombic (space group C2mm) and is called αU3O8.10−13 When heated to 483 K, the α form is converted into α′ with transition to the hexagonal space group P6̅2m. This transition and the consequent atomic position modifications were studied by Loopstra et al.14 and Ackermann et al.,15 who defined the current space groups and positions for U and O in the α-U3O8 and α′-U3O8 forms. While many characterizations have been performed on U3O8 to explain crystallographic structures10,12,14,16,17 few papers exist on U3O8 with another substituted chemical element. Tetravalent Pu and Np were Received: July 13, 2016

A

DOI: 10.1021/acs.inorgchem.6b01672 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

microspheres, and at the U LIII edge for pure U3O8 microspheres. The use of the U LII edge was necessary because of the presence of neptunium impurities in Am, with its LIII absorption edge interfering with the U LIII EXAFS region. The fluorescence signals were measured with a 13-element Ge solid-state detector using a digital spectrometer (XIA-XMap). Metallic foils with K edges close to the edges of interest, that is, Y (17 038 eV), Zr (17 998 eV), and Mo (20 000 eV), were used as references for energy calibration. EXAFS spectra were recorded at Am LIII, U LII, and U LIII edges, up to k = 18, 13.5, and 16 Å−1, respectively. During all the measurements, a He cryostat was used to maintain the sample temperature around 15 K. The thermal contribution to the Debye−Waller factors was thus greatly reduced, and only the static structural disorder contribution remained. Data refinements were performed using the IFEFFIT27,28 software and FEFF929 for ab initio calculations of EXAFS spectra. XANES spectra were normalized using a linear function for pre- and postedge approximation. E0 was arbitrarily assigned to the edge inflection point, determined as the knot of the first derivative. The U LIII edge spectra were compared to that of a UO3 compound also recorded at ROBL under the same conditions, while the Am LIII spectra were compared to those of AmO2 and pure Am oxalate,30 as Am+IV and Am+III references, respectively. For the pure U3O8 reference, a Fourier transform of the EXAFS spectrum (in k2) was performed with a Hanning window between 2.2 and 16 Å−1, while for the Amsubstituted U3O8, a Fourier transform of the EXAFS spectra (in k2) was performed with a Hanning window between 3 and 13 Å−1 and 2.8 and 11.8 Å−1 for the U LII and Am LIII edges, respectively. A 7.5 Å cluster of U3O8 was used for data fitting, using the lattice parameters determined through XRD measurements. Model U3O8 was fitted with the C2mm space group. In this crystallographic structure, U is coordinated by two axial O, five equatorial O, two axial U, and four equatorial U as nearest neighbors. Multiple scattering along axial directions was taken into account through the addition of triple scattering (U1−O2−OX−U1) and quadruple scattering (U1−O2−UX− O2−U1). The structural parameters of the multiple scattering paths were linked to those of corresponding single scattering paths to reduce the number of free parameters. Two space groups were tested for the Am-substituted U3O8, namely, P3̅m1 and P6̅2m, but only the latter allowed the fit to converge. For P6̅2m, the configuration of single scatterings around U was the same as for C2mm with a few changes in distances. Consequently similar triple and quadruple scattering paths were also added and improved the fit quality with structural parameters linked to the single paths. Debye−Weller factors were found equal to 0.005(1) and 0.0080 Å2, and 0.0200 and 0.0150 Å2 for ULII and AmLIII edges, respectively.

integrated into U3O8 to form a solid solution. Benedict et al. thus report some space group change with Pu incorporation,18 as do Finch et al. for Np.19 However, no structural models have been presented to explain how these elements, present as impurities, are accommodated in the structure. More recently, Remy synthesized a similar kind of mixed oxide containing tetravalent Ce.20 In the study described here, Am was chosen as the substituted element, meeting the transmutation target research need for post-irradiation characterizations. Mixed oxides composed of U and Am were synthesized under oxidative atmosphere, in the same conditions as those required for U3O8 formation.21−23 The crystalline structure obtained was fully characterized and compared to pure U3O8. First, X-ray diffraction (XRD) results presenting the new crystalline structure and its difference compared to the pure compound are exposed and analyzed. These results were then completed by X-ray absorption near-edge spectroscopy (XANES) measurements assessing the formal oxidation states of U and Am in the mixed oxide. Finally, X-ray absorption fine structure (EXAFS) data fitting was used to reconstruct the close environment of the U and Am cations and to establish the metal−O and metal−metal first distances, highlighting how Am can be substituted into the U3O8 structure.

■

EXPERIMENTAL SECTION

Sample Preparation. A wet chemical route based on the use on ion-exchange resin microspheres was selected for this work to ensure a homogeneous atomic scale distribution of the actinide element in the final oxide. This process of (U,Am)-based samples was the same as that described in previous works for actinide-based mixed-oxide synthesis.22,23 Stoichiometric amounts of uranyl and americium cations dissolved in aqueous solution with a 90/10 atomic U/Am ratio were fixed on the exchanger sites of weak acid resin (WAR) microspheres, releasing proton ions. After complete exchange (stable pH of the solution), loaded resin was extracted and dried at 373 K. Microspheres were then calcined for 4 h at 973 K under air, following a heat-up rate of 5 K·min−1. This calcination was performed in a tubular furnace (Lenton LTF), in which the microspheres were placed as a single layer in a quartz crucible. The same protocol was followed for pure U3O8 reference microspheres. At the end of the calcination, oxide microspheres were obtained and used for following characterizations. X-ray Diffraction. RT XRD diagrams were recorded using a Bruker D8 Advance apparatus equipped with a Cu Kα1,2 radiation and a linear Lynx-Eye detector, in a θ−θ Bragg−Brentano configuration. The step was fixed at ∼0.02°, with a time per step of 0.3 s, for an angular domain of 15 to 110° 2θ. Bruker EVA DIFFRACplus software was used for phase identification. The lattice parameters of the synthesized compounds were determined using Jana Suite software,24,25 with a Le Bail method for refinements.26 A pseudo-Voigt profile function was used for lattice parameter refinement. Oxide microspheres were crushed to enable XRD anlysis on oxide powder. A gold reference was added to the powdered sample for position calibrations. The Au phase profile was also fitted, but the lattice parameters were kept fixed, as it is used as a reference. X-ray Absorption Spectroscopy Experiments. X-ray absorption spectroscopy (XAS) experiments were performed at the European Synchrotron Radiation Facility (ESRF, Grenoble, France) on the Rossendorf Beamline (ROBL), with the storage ring operating at 6.0 GeV and 170−200 mA. The synchrotron beam was collimated with a 1.3 m long, Rh-coated, meridionally bent mirror, monochromatized with a double-crystal monochromator with a water-cooled Si (111) crystal pair, and focalized with a 1.2 cm long, Rh-coated toroidal mirror. XANES spectra were collected in transmission mode at the Am LIII (18 510 eV) and the U LIII (17 166 eV) edges with a step size of 0.7 eV, using ionization chambers. EXAFS measurements were performed in both transmission and fluorescence modes, at the Am LIII (18510 eV) and U LII (20948 eV) edges for Am-substituted U3O8

■

RESULTS AND DISCUSSION The following results present the features of the crystallographic structure obtained after (U,Am) loaded resin thermal conversion under air, performed at 973 K. These previously unknown features are compared to pure U3O8 microspheres. X-ray Diffraction Results. XRD diagrams recorded for pure U3O8 and Am-based U3O8 microspheres are shown in Figures 1 and 2, respectively. For the pure U3O8 X-ray phase diagram, phase identification matches the orthorhombic C2mm crystallographic structure. In the general context of uranium oxides, it corresponds to the α-U3O8 phase, a structure stabilized at RT. Le Bail refinement performed on this diagram gave the following lattice parameters: a = 6.720(2) Å; b = 11.952(2) Å; c = 4.146(2) Å, for a lattice volume of 333.02(1) Å3. The refinement parameters are summarized in Table 1, and the corresponding plots are shown in Figure 1. Structural values compared to the literature12,14,15 are typical of a standard αU3O8 phase and validate the synthesis of the U3O8 compound through thermal conversion of WAR microspheres under air at 973 K. B


Article

Inorganic Chemistry

hexagonal P6̅2m space group, stabilized at high temperature (over 500 K) by U3O8.15 Both have a similar cationic sublattice with single cationic position, where differences occur in the vicinity of the O atoms. In the case of a trigonal structure there are only two O positions, for five positions in a hexagonal one, consistent with greater anionic organization in the case of the hexagonal structure. The adjustment of the crystalline structure when Am is present, through this change of space group, is typical of Am substitution inside U3O8 phase. Le Bail refinements were performed for both possibilities. The lattice parameters determined for the P6̅2m space group, whose refinement is visible in Figure 2, are a = 6.838(2) Å and c = 4.171(2) Å, for a total lattice volume of 169.05(1) Å3. When refined with the P3̅m1 space group, the following lattice parameters were obtained: a = 3.948(2) Å and c = 4.167(2) Å, for a total lattice volume of 56.33(1) Å3, where a factor of the trigonal structure is found equal to that of the hexagonal one divided by √3. If comparing the now normalized volume of each crystalline structure (Table 2) with pure U3O8, a swelling

Figure 1. XRD diagram and refinement results for powdered pure U3O8 oxide microspheres converted for 4 h at 973 K under air. Peak matches with an * correspond to the gold phase. Experimental data are in black, Le Bail calculated diagram is in red, the difference between them is in blue, and Bragg positions are in green.

Table 2. Summary of the Lattice Volume Calculated for Pure U3O8 and Am-Substituted U3O8 pure U3O8

Am-substituted U3O8

space group

C2mm

P6̅2m

P3̅m1

volume (Å3) normalized volume (Å3)

333.02(2) 333.02(2)

169.05(2) 338.10(2)

56.33(2) 337.98(2)

of the structure clearly appears in the sample containing Am. The global increase is close to 5 Å3, which is a significant value for lattice modification and favors the hypothesis of Am substitution on a U site in the U3O8 oxide. X-ray Absorption Spectroscopy Results. X-ray Absorption Near-Edge Spectroscopy. The spectra recorded at the Am LIII edge for Am-substituted U3O8 and model compounds are shown in Figure 3. The comparison between the sample spectrum and the reference for Am+IV and Am+III clearly shows that Am is totally trivalent in the Am-substituted U3O8 structure. Concerning the corresponding U LIII edge, spectra (and corresponding derivatives) recorded for pure α-U3O8, UO3, and Am-substituted U3O8 are presented in Figure 4. UO3

Figure 2. XRD diagram and refinement results for (U,Am)-based oxide microspheres converted for 4 h at 973 K under air. Peak matches with an * correspond to the gold phase. Experimental data are in black, Le Bail calculated diagram is in red for P6̅2m reference space group, the difference between them is in blue, and Bragg positions are in green.

Table 1. Fit Parameters Obtained after Le Bail Refinement of the Two Samples Studied phase

pure U3O8

Am-substituted U3O8

space group

C2mm

P6̅2m

P3̅m1

GOF Rp wRp

1.44 2.54 3.48

1.30 2.79 3.60

1.50 2.82 3.74

Considering the X-ray diagram (Figure 2) collected after thermal conversion of WAR microspheres loaded with U and Am, a diagram showing a monophasic compound was also obtained after conversion, but if it seems similar to that of αU3O8 phase, a distinct difference can be noticed on the two peaks around 26.5° (enclosed zoom) and 34°. While the pure U3O8 phase showed two significant peaks, these merged into a single one in the new phase including Am. In phase identification, two crystallographic structures match the diagram collected. The first crystallizes in the trigonal P3̅m1 space group taken by α-UO3 phase,31 and the second in the

Figure 3. Normalized XANES spectra collected at the Am LIII edge and 20 K. AmO2 is in dark gray, Am-substituted U3O8 is in blue, and the Am+III reference is in red. C


Article

Inorganic Chemistry

increase in lattice volume for Am-substituted U3O8 visible in Table 2. An EXAFS study was also performed to better understand how the crystalline structure accommodates the presence of trivalent Am, what is in the close environment of U and Am, and what the differences are compared to pure U3O8. Extended X-ray Absorption Fine Structure. The k2weighted EXAFS spectrum and corresponding Fourier transform of pure U3O8 are presented in Figures 5 and 6. The best-

Figure 4. Normalized XANES spectra collected at the U LIII edge and associated derivatives; 20 K. U3O8 is in dark gray, Am-substituted U3O8 is in blue, and UO3 is in red.

corresponds to pure U+VI cation, illustrated by two distinct bands after the absorption edge in Figure 4, typical of two relatively wide U−O distances present: 2.1 Å for apical distance and ∼2.38 Å for equatorial distances.32,33 For the α-U3O8 reference, the spectrum shows a single significant band after the absorption edge, characteristic of the C2mm crystallographic structure with similar U−O. The valence of U in UO3 is +VI, while U3O8 presents two different U oxidation states, assessed to be +V and +VI by Kvashnina et al., thus ensuring electroneutrality inside the oxide.34 Considering U3O8 as a stoichiometric oxide at RT, the U element is 33% U+VI and 66% U+V, for an average valence of 5.33. When looking at the Am-substituted U3O8 spectrum, it appears to be intermediary between the two pure uranium oxides. Although not as pronounced as for UO3, two bands are present on the absorption derivative after the absorption edge, illustrated by a small shoulder around 17 180 eV, which is completely absent from the U3O8 spectrum. This observation would thus be consistent with the presence of two different U− O distances in the (U,Am)-based oxide, approaching the UO3 configuration. Figure 4 suggests that U lies in a mixed-valence state with unknown +V/+VI. Only high-resolution XANES would allow a precise confirmation of the mixed-valence and quantitative proportions of +V and +VI. Here, the resolution of the core hole at the An LII,III edge (∼7 eV) is far too small to enable such determination with accuracy. It can be concluded that an Am-substituted U3O8 XANES derivative (post edge shape) is thus qualitatively closer to that of UO3 than U3O8. If considering 10 atom % of trivalent Am on +V/VI U sites, the average U charge for stoichiometric oxide will shift from 5.33 (for pure U3O8) to 5.59. This would correspond to the presence of 60% of U+VI in the mixed oxide instead of 33% in pure U3O8, to compensate for trivalent Am presence. Considering the XRD results, XANES validates the hypothesis of substitution due to the change observed in the crystallographic structures from pure U3O8 to Am-substituted U3O8. Knowing the oxidation state of Am in the sample (+III), a comparison can be made between the sizes of the different cations present. Considering that the ionic radius of trivalent Am in a hexahedral environment equals 1.115 Å, while it is 0.87 Å for U+VI and 0.9 Å for U+V,35 helps to explain the global

Figure 5. k2-weighted EXAFS spectra of the pure U3O8 and Amsubstituted U3O8. Experimental spectra are plotted in dots (dark gray: pure U3O8 at U LIII edge; blue and light blue: Am-substituted U3O8 at the ULII and AmLIII edges, respectively, adjustments in black lines).

Figure 6. Fourier transform of the EXAFS spectra for pure U3O8 and Am-substituted U3O8. Experimental spectra are plotted in dots (dark gray: pure U3O8 at U LIII edge; blue and light blue: Am-substituted U3O8 at the ULII and AmLIII edges, respectively, adjustments in black lines).

fit structural parameters are summarized in Table 3. For this C2mm model structure, the fitted distances for single paths were close to the crystallographic data. The first U−O axial distance has a high uncertainty, and the corresponding Debye− Weller factor also exhibits quite a high value (0.03 Å2). This is due to complicated signal extraction at low k range for this kind of oxide, resulting in large Debye−Waller factors and high uncertainties for the first shell. Nevertheless, the other distances confirm the correspondence with the C2mm space group defined for pure α-U3O8, particularly through sub-cationic D


Article

Inorganic Chemistry Table 3. EXAFS Data on Pure U3O8 and Am-Substituted U3O8, and Comparison with Theorya pure U3O8 reference C2mm U LIII edge Rfactor = 3.1% Δχ2ν = 0.57 S02 = 0.8 Δe0 = 5.85 ε = 0.004 Am-substituted U3O8

U LII edge Rfactor = 3.6% Δχ2ν = 27.48 S02 = 0.9 Δe0 = −4.77 ε = 0.003 Am LIII edge Rfactor = 7.7% Δχ2ν = 4.56 S02 = 1.0 Δe0 = 6.02 ε = 0.012

bond type

distance (Å)

crystallography ref (Å)

σ2 (Å2)

two axial U−O four equatorial U−O one equatorial U−O two equatorial U···U two equatorial U···U two equatorial U···U two axial U···U eight diagonal U···U (50%) two axial U−O (50%) two axial UO′ four equatorial U−O one equatorial U−O six equatorial U−U (50%) two axial U−U two axial Am−O four equatorial Am−O one equatorial Am−O six equatorial Am−U two axial Am−U

1.99(5) 2.21(1) 2.59(1) 3.70(1) 3.84(1) 4.14(1) 4.16(1) 5.62(1) 2.03(1) 1.78(1) 2.16(1) 2.62(1) 4.01(2) 4.17(1) 2.26(1) 2.46(2) 2.93(2) 3.69(2) 4.27(2)

2.07 2.16 2.54 3.74 3.88 4.18 4.15 5.59 2.07 2.07 2.21 2.67 3.83 4.14 2.07 2.21 2.68 3.83 4.15

0.0301 0.0077 0.0237 0.0084 0.0138 0.0065 0.0065 0.0070 0.0165 0.0036 0.0085 0.0159 0.0137 0.0015 0.0033 0.0032 0.0022 0.0178 0.0028

So2 is the EXAFS global amplitude factor, Δe0 is the energy threshold, ε is the average noise, and Rfactor is the agreement factor of the fit. Numbers in round brackets are uncertainties. a

lattice and U···U interatomic distances that were fitted until 5.6 Å. Considering Am-substituted U3O8, both P3̅m1 and P6̅2m space groups were first tested as model references for fitting at the U edge. The trigonal structure did not show good correlation between experimental plot and fit, tending to suggest that metal−O and metal···U distances in the sample were closer to those identified in the hexagonal reference configuration. Therefore, as a starting point for fitting, the path lengths in the P6̅2m structure were considered as models for both Am and U edges. Focusing on the Am environment, Table 3 presents the bestfit parameters. XRD previously showed the substitution of Am on U site. Good results obtained for EXAFS spectrum fitting tend to confirm these results, as Am−O and Am···U distances are in good agreement with those corresponding to the P6̅2m structure. The increase in interatomic Am−O distances is in agreement with the decrease in formal oxidation state from +5.33 for U in U3O8 C2mm to +3 for Am in the structure studied here, although this is only indicative, as the two cations do not have the same geometry. Moreover, as Am+III has a larger crystal radius (1.115 Å) than U in the same hexahedral configuration (0.87 Å for U+VI and 0.9 Å for U+V),35 it should also contribute to the expansion of the O environment. The fit of the Am···U distances was only possible when considering a considerable shortening of the four equatorial Am−U paths, whereas axial distances increase. This significant variation, opposing the general swelling tendency, can be understood if anionic and cationic paths are de-correlated. This can be assumed to be the case, as anionic interatomic distances undergo variations that are different from cationic sublattice, and cations and anions are not aligned together in the same direction. It thus renders Am···U equatorial distance shrinkage possible, despite swelling of the anionic sublattice. On the contrary, for the axial direction where Am and O are aligned, both Am−O and Am···Am distances increase.

Fitting the U LII EXAFS spectrum for Am-substituted U3O8 evidences several structural modifications in the U closest environment compared to the P6̅2m structure, summarized in Table 3. Concerning the U−O single scattering, optimization of the fit led to a split of axial U−O distances into two different contributions, with an almost 50/50 distribution. The first equaled 2.03(1) Å and was very similar to the crystallographic value of 2.07 Å in P6̅2m. The second was found to be significantly shorter than the conventional distance in U3O8, at 1.78(1) Å. Attempts to remove the contribution from this short distance or to compensate for it with a higher Debye−Waller value for the first shell led to a very significant deterioration of the quality factor (Δχ2ν = 27 to 265). In addition, this result is correlated to the above discussion on XANES spectra and suggests that 50% of the U−O bond can be assimilated to “uranyl”, as its short distance is very close to the theoretical uranyl distance (1.76 Å36). As a result, it may be assumed that trivalent Am substitution in U3O8 induces structural and charge modifications with an increase in U+VI quantity, as expected with previous calculations of charge compensation leading to 60% of U+VI required, instead of 33%. An excess of the +VI cations thus needs to be organized within the structure through the formation of uranyl-like units suggested by XANES and fitted by EXAFS. Concerning the equatorial U−O distances, the values determined for the five paths are in agreement with the P6̅2m model: 2.16(1) Å compared to 2.21 Å for the four symmetric surrounding O, and 2.62(1) Å compared to 2.67 Å for the single O. When focusing on the cationic sublattice, the U···U interatomic distance also reveals important modifications with a considerable increase in one distance (+0.16 Å), linked to the four equatorial U···U paths, whereas simple axial interatomic distances remain close to the crystallography measurements. The addition of triple and quadruple scattering paths linked to 2.07 Å U−O axial distances considerably improved the quality of the fit. This result strengthens the assumption of the coexistence of uranyl-like bonds together with normal U−O axial bonds linked to the sub-cationic lattice. E


Article

Inorganic Chemistry

increase of interatomic distances around it. U undergoes more structural rearrangements with the formation of yl-like bonds, resulting in structural reorganization of U+VI in uranyl units, as U+VI amount increases to 60% instead of 33%, to compensate for trivalent Am presence. The planar geometry of U3O8-type oxides suggests alternating planes of U+VI with uranyl units and U+V in a conventional P6̅2m configuration. These results are the first new insights into the possibility for U3O8 to accommodate when submitted to high stress, such as the substitution of a trivalent actinide. They offer first promising approaches in the field of structural arrangements.

The structural arrangements around U and Am in pure U3O8 and in Am-substituted U3O8 are thus very different. According to the above data, the Am cation substitutes U and provokes a general swelling of its close environment, except for the Am···U equatorial distances that decrease. The U and O atoms of the Am equatorial plane are not aligned and make this Am···U distance modification possible. This is also evidenced by the U···U distortion caused by the presence of Am, with different structural constraints on cationic and anionic sublattices. The major change to the P6̅2m structure comes from the necessity to add shorter free U−O bonds to the system, which would be related to axial uranyl-like units as illustrated in Figure 7. The

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: 0033466796542. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors would like to thank F. Boucher for his precious advice on oxide structures. They also thank M. Bataille and P. Coste for sample preparation, as well as A. Gauthé, I. Jobelin, P. Grangaud, and J.-M. Pomarède for sample synthesis. We acknowledge the ESRF Synchrotron Light Source team’s help in providing beamtime at the ROBL beamline, and P. Colomp for sample transportation. M. Caisso is grateful for her Ph.D. fellowship funding by the CEA PACFA program.

Figure 7. (left) First anionic and cationic coordination shells for the P6̅2m space group used for EXAFS fitting. (right) First anionic and cationic coordination shells for U+VI, involving yle configuration. Green atoms are metals (U, Am), light green atoms are U, and red are O. The black atom is considered as an absorbing atom. Orange bonds are single, while blue bonds are double.

■

figure thus shows the two kinds of cationic and anionic environments around U and Am. The left part shows the P6̅2m configuration taken by Am and 50% of U, mostly U+V, in spite of anionic distance differences. The green atoms correspond to U or Am cations around the absorbing cation, while the red atoms are O. The configuration on the right illustrates the environment determined with “yle” units around U, attributed to U+VI, identified by the two blue double UO bonds. To conclude, from the U environment, the anionic sublattice undergoes most of its structural changes through uranyl double bond formation induced by a higher amount of U+VI in the mixed oxide, to compensate for the presence of Am+III. As U3O8 oxides have a planar layout, this specific configuration means that it probably favors alternating planes composed either of U+VI surrounded by uranyl units or U+V in conventional P6̅2m environment.

REFERENCES

(1) McEachern, R. J.; Taylor, P. A Review of the Oxidation of Uranium Dioxide at Temperatures below 400°C. J. Nucl. Mater. 1998, 254, 87−121. (2) Gras, J.-M.; Quang, R. D.; Masson, H.; Lieven, T.; Ferry, C.; Poinssot, C.; Debes, M.; Delbecq, J.-M. Perspectives on the Closed Fuel Cycle − Implications for High-Level Waste Matrices. J. Nucl. Mater. 2007, 362, 383−394. (3) Ferry, C.; Poinssot, C.; Cappelaere, C.; Desgranges, L.; Jegou, C.; Miserque, F.; Piron, J. P.; Roudil, D.; Gras, J. M. Specific Outcomes of the Research on the Spent Fuel Long-Term Evolution in Interim Dry Storage and Deep Geological Disposal. J. Nucl. Mater. 2006, 352, 246− 253. (4) Thomas, L. E.; Einziger, R. E.; Buchanan, H. C. Effect of Fission Products on Air-Oxidation of LWR Spent Fuel. J. Nucl. Mater. 1993, 201, 310−319. (5) Warin, D. Status of the French Research Program on Partitioning and Transmutation. J. Nucl. Sci. Technol. 2007, 44, 410−414. (6) Delahaye, T.; Lebreton, F.; Horlait, D.; Herlet, N.; Dehaudt, P. Application of the UMACS Process to Highly Dense U1‑xAmxO2±δ MABB Fuel Fabrication for the DIAMINO Irradiation. J. Nucl. Mater. 2013, 432, 305−312. (7) Bejaoui, S. Americium-Bearing Blanket Separate-Effect Experiments: MARIOS and DIAMINO Irradiations. In Proceedings of GLOBAL 2011; Makuhari, Japan, 2011. (8) Haas, D. MOX fuel fabrication experience at Belgonucleaire; Technical committee meeting on recycling of plutonium and uranium in water reactor fuel; International Atomic Energy Agency: Vienna (Austria), 1997. (9) Baron, D. D.; Hallstadius, L. 2.19 - Fuel Performance of Light Water Reactors (Uranium Oxide and MOX). In Comprehensive Nuclear Materials; Konings, R. J. M., Ed.; Elsevier: Oxford, 2012; pp 481−514. (10) Andresen, A. F. The Structure of U 3 O 8 Determined by Neutron Diffraction. Acta Crystallogr. 1958, 11, 612−614. (11) Ball, R. G. J.; Dickens, P. G. Calculation of Structural and Defect Properties of α-U3O8. J. Mater. Chem. 1991, 1, 105−112.

■

CONCLUSION These experiments investigated the possibility of trivalent Am being substituted in an α-U3O8 structure for the first time. The combination of XRD and XAS experiments led to the conclusion that substitution is possible thanks to a structure accommodation, through a transformation from the initial orthorhombic to a hexagonal structure for an Am-substituted oxide. The latter, corresponding to the P6̅2m space group, has a single cationic site, which increases the global symmetry of the structure, but also attests to a higher metallic disorder. The global cationic charge order of the orthorhombic structure (C2mm space group), where U+V and U+VI obtain their own metallic sites, is broken by the presence of trivalent Am. This leads to a space group change. Focusing on the close environments of U and Am in the mixed oxide as determined through EXAFS analysis, Am is well-fitted with distances characteristic of a hexagonal structure but causes a general F


Article

Inorganic Chemistry (12) Loopstra, B. O. Neutron Diffraction Investigation of U3O8. Acta Crystallogr. 1964, 17, 651−654. (13) Wen, X.-D.; Martin, R. L.; Scuseria, G. E.; Rudin, S. P.; Batista, E. R.; Burrell, A. K. Screened Hybrid and DFT + U Studies of the Structural, Electronic, and Optical Properties of U3O8. J. Phys.: Condens. Matter 2013, 25, 25501. (14) Loopstra, B. O. The Phase Transition in α-U3O8 at 210°C. J. Appl. Crystallogr. 1970, 3, 94−96. (15) Ackermann, R. J.; Chang, A. T.; Sorrell, C. A. Thermal Expansion and Phase Transformations of the U3O8‑Z Phase in Air. J. Inorg. Nucl. Chem. 1977, 39, 75−85. (16) Loopstra, B. O. The Structure of β-U3O8. Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem. 1970, 26, 656−657. (17) Desgranges, L.; Palancher, H.; Gamaléri, M.; Micha, J. S.; Optasanu, V.; Raceanu, L.; Montesin, T.; Creton, N. Influence of the U3O7 Domain Structure on Cracking during the Oxidation of UO2. J. Nucl. Mater. 2010, 402, 167−172. (18) Benedict, U. X-Ray Study of (U, Pu)3O8. J. Nucl. Mater. 1970, 35, 356−361. (19) Finch, R. J.; Jeremy Kropf, A. Neptunium Substitution into the Structure of Alpha-U3O8. MRS Online Proc. Libr. 2002, 757, nullnull.10.1557/PROC-757-II3.3 (20) Remy, E. Étude de La Synthèse de Sphères D’oxyde D’actinides Et/ ou de Lanthanides et de Leur Aptitude À La Céramisation; Université de Montpellier 2, 2013. (21) Remy, E.; Picart, S.; Delahaye, T.; Jobelin, I.; Dugne, O.; Bisel, I.; Blanchart, P.; Ayral, A. Fabrication of Uranium Dioxide Ceramic Pellets with Controlled Porosity from Oxide Microspheres. J. Nucl. Mater. 2014, 448, 80−86. (22) Remy, E.; Picart, S.; Delahaye, T.; Jobelin, I.; Lebreton, F.; Horlait, D.; Bisel, I.; Blanchart, P.; Ayral, A. Fabrication of Uranium− americium Mixed Oxide Pellet from Microsphere Precursors: Application of CRMP Process. J. Nucl. Mater. 2014, 453, 214−219. (23) Caisso, M.; Picart, S.; Belin, R. C.; Lebreton, F.; Martin, P. M.; Dardenne, K.; Rothe, J.; Neuville, D. R.; Delahaye, T.; Ayral, A. In Situ Characterization of Uranium and Americium Oxide Solid Solution Formation for CRMP Process: First Combination of in Situ XRD and XANES Measurements. Dalton Trans. 2015, 44, 6391−6399. (24) Carjaval, J. R. FULLPROF: A Program for Rietveld Refinement and Pattern Matching Analysis. Satell. Meet. Powder Diffr. XV IUCr Congr. 1990. (25) Rodríguez-Carvajal, J. Recent Advances in Magnetic Structure Determination by Neutron Powder Diffraction. Phys. B 1993, 192, 55−69. (26) Le Bail, A. Whole Powder Pattern Decomposition Methods and Applications: A Retrospection. Powder Diffr. 2005, 20, 316−326. (27) Newville, M. IFEFFIT: Interactive XAFS Analysis and FEFF Fitting. J. Synchrotron Radiat. 2001, 8, 322−324. (28) Ravel, B.; Newville, M. ATHENA, ARTEMIS, HEPHAESTUS: Data Analysis for X-Ray Absorption Spectroscopy Using IFEFFIT. J. Synchrotron Radiat. 2005, 12, 537−541. (29) Rehr, J. J.; Kas, J. J.; Vila, F. D.; Prange, M. P.; Jorissen, K. Parameter-Free Calculations of X-Ray Spectra with FEFF9. Phys. Chem. Chem. Phys. 2010, 12, 5503−5513. (30) Arab-Chapelet, B.; Martin, P.; Costenoble, S.; Delahaye, T.; Scheinost, A. C.; Grandjean, S.; Abraham, F. Multiscale Structural Characterizations of Mixed U(IV)-An(III) Oxalates (An(III) = Pu or Am) Combining XAS and XRD Measurements. Dalton Trans. 2016, 45, 6909. (31) Hoekstra, H. R.; Siegel, S.; Fuchs, L. H.; Katz, J. J. The Uranium−Oxygen System: UO2.5 to U3O8. J. Phys. Chem. 1955, 59, 136−138. (32) Siegel, S.; Hoekstra, H. R. An Examination of the Symmetry of Alpha Uranium Trioxide. Inorg. Nucl. Chem. Lett. 1971, 7, 497−504. (33) Greaves, C.; Fender, B. E. F. The Structure of α-UO3 by Neutron and Electron Diffraction. Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem. 1972, 28, 3609−3614.

(34) Kvashnina, K. O.; Butorin, S. M.; Martin, P.; Glatzel, P. Chemical State of Complex Uranium Oxides. Phys. Rev. Lett. 2013, 111, 253002. (35) Shannon, R. D. Revised Effective Ionic Radii and Systematic Studies of Interatomic Distances in Halides and Chalcogenides. Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr. 1976, 32, 751−767. (36) Jones, L. H. Determination of U-O Bond Distance in Uranyl Complexes from Their Infrared Spectra. Spectrochim. Acta 1959, 15, 409−411.

G


Evidence of Trivalent Am Substitution into U3O8 - Inorganic Chemistry

Recommend Documents