On the Raman Spectrum of Plutonium dioxide: vibrational and crystal field modes M. Naji1, N. Magnani1, L. J. Bonales3, S. Mastromarino1, 2, J.-Y. Colle1, J. Cobos3, D. Manara1,* 1European Commission, Joint Research Centre, Institute for Transuranium Elements, P.O. Box 2340, 76125 Karlsruhe, Germany. 2Università degli Studi di Roma "La Sapienza", Piazzale Aldo Moro 5, 00185 Rome, Italy. 3Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas, CIEMAT Avenida Complutense, 40, 28040 Madrid, Spain. The Raman spectrum of plutonium dioxide is studied both experimentally and theoretically. Particular attention has been devoted to the identification of high-energy modes at 2120 cm-1 and 2625 cm-1, whose attribution has so far been controversial. The temperature dependence of both modes suggests an electronic origin. Crystal Field (CF) calculations reported in this work shows that these two modes can be respectively assigned to the Γ1 → Γ5 and Γ1 → Γ3 CF transitions 5 within the I4 manifold. These two modes, together with the only vibrational line foreseen by the group theory for the -1 Fm-3m PuO2 symmetry – the T2g Pu-O stretching mode observed at 478 cm – can thus be used as Raman fingerprint of fcc plutonium dioxide. PACS number(s): 78.30.Am, 07.35.+k, 62.50.+p I. INTRODUCTION temperature-independent magnetic susceptibility14 is compatible with its Γ1 singlet ground state (which is Plutonium dioxide spectroscopy can have important expected from the simple argument that the CF states applications in nuclear waste management, nuclear for Pu4+ ions in a given ligand geometry should roughly forensic and safeguard activities. 1, 2 In particular, the be the inverse as those for U4+).15 However, this single- Raman spectrum of plutonium dioxide (PuO2) has been ion picture turns out to be problematic upon further studied in a few research laboratories in the last couple inspection, in particular because the energy gap of decades. 3-6 Plutonium dioxide crystallises in the fcc between the two lowest CF levels detected by inelastic fluorite-like structure, belonging to the Fm-3m space neutron scattering (INS) is much smaller than the value group. The factor group theory at the Γ point for this which was estimated from magnetic measurements (in type of crystal symmetry was fully derived by order to maintain a constant paramagnetic susceptibility Shimanouchi et al. 7. It foresees only one triply up to 1000 K); this issue is still an open problem today. degenerate T2g (LO1) Raman active vibrational mode, It must be remarked that INS is only sensitive to the Γ1 and one infrared active T1u mode at a lower energy. → Γ4 transition, and a complete spectroscopic 5 Since the work by Shimanouchi et al., these modes have investigation of the lowest I4 manifold is still lacking been experimentally observed for various actinide and would be highly desirable in order to test it against dioxides and other fluorite-like materials,8-11 including current theories. plutonium dioxide. 4, 5, 12 However, the Raman spectrum Experimentally, in addition to INS, electronic Raman of plutonium dioxide was reported up to higher Raman scattering (ERS) played a crucial role to determine the -1 12 16 shift energies (beyond 1500 cm ) only recently. A few various split LSJ manifolds of the 4,5f ions. In UO2, crucial features of the high-energy part of the PuO2 ERS elucidate the manifestation of the strong electron– Raman spectrum remain to-date unclarified, which has phonon interactions in the resonant coupling between ′ ′ 4+ not permitted a satisfactory definition of a precise nLO (n = 1–4) phonons and 3퐻4 → 3퐹2 U Raman fingerprint for PuO2. Among those features, the intermultiplet CF excitations. Consequently, one would most remarkable are two strong peaks consistently expect that the Raman effect would be of some interest 12 -1 observed by Sarsfield et al. at 2120 ± 10 cm and in the study of CF transitions in PuO2. 2620 ± 10 cm-1. Despite some attempts, the nature of The experimental part of this study was carried out these two lines could not be clearly defined. Sarsfield with the help of an original encapsulation technique, et al. 12 analysed in depth various possible attributions described in a very recent publication.17 Among other for them, both in terms of higher-order phonon things, this technique permits, for the investigation of combinations and surface impurity modes, but could not such highly radioactive materials, the use of different find any satisfactory assignment. Finally, these authors excitation laser sources, and the employment of the suggested that these lines might have an electronic triple subtractive spectrometer configuration. The first origin. feature is important for the analysis of resonant and In this respect, there is a vast literature discussing the energy dispersive modes and for the detection of crystal field (CF) potential of UO2 and NpO2, fluorescence features in the Raman spectra; the second particularly in relation to their peculiar low-temperature is necessary for the study of low-energy Raman lines 13 magnetic properties. Less attention has been devoted and anti-Stokes peaks. In particular, the study of the T2g to PuO2, mainly because, at a first sight, its anti-Stokes band, carried out in this research for the 1 very first time on PuO2 in parallel to the Stokes in this research is a Jobin-Yvon T 64000 equipped with spectrum, allowed us to reasonably estimate a a 1800 grooves per mm grating, a low noise liquid temperature dependence of the observed Raman modes, nitrogen cooled symphony CCD detector, a subtractive especially the above-mentioned high-energy bands. pre-monochromator (in triple mode) which allows Moreover, CF calculations performed in the present access to anti-Stokes spectrum lines while blocking the research soundly support the attribution of these high- elastic Rayleigh line. Excitation sources used in this energy lines to PuO2 CF transitions. Such a conclusion, work are the 488 nm (2.53 eV) or 514.5 nm (2.41 eV) together with the already sound theoretical and lines of an Ar+ Coherent® Continuous Wave (CW) experimental description of the vibrational part of the laser or the 647 nm (1.91 eV) or 752 nm (1.65 eV) lines Raman spectrum, allows us to propose a consistent of a similar Kr+ laser. Both laser sources have a Raman fingerprint for plutonium dioxide. controllable nominal power up to 1.5 W. Typical laser power values at the sample surface ranged between 5 II. EXPERIMENTAL METHODS mW and 300 mW, whereby the highest power values were used in order to heat the sample in situ. The power A. Samples impinging the sample surface was measured by a Coherent® power-meter placed at a position The current Raman spectroscopy measurements were corresponding to the sample surface. It is lower by performed on a solid polycrystalline PuO2 disk, approximately a factor 5 than the actual power at the prepared following the procedure described by De exit of the laser cavity. Using the long focal x50 Bruycker et al. 18 The starting material was based on objective and the single spectrometer mode permits a beads obtained by gel-supported precipitation (SOL– good spectral resolution (± 1 cm-1) independently of the GEL). A disk-shaped sample of 8–9 mm in diameter surface shape, with a spatial resolution of 2 µm x 2µm and about 1 mm in thickness was obtained using a bi- on the sample surface. The spectrograph is calibrated directional press. The samples were then sintered in an with the T2g excitation of a silicon single crystal, set at -1 20 atmosphere of Ar + H2 with 1500 ppm of H2O to obtain 520.5 cm . The instrument is calibrated on a daily dense material. In order to obtain stoichiometric basis prior to measurements. material (O/M=2.00) and remove the defects produced during the fabrication, the PuO2 sample was subjected C. Temperature determination to two consecutive heat treatments under a flow of air at 1423 K for 8 hours. Already after the first heat a) Stokes and anti-Stokes Raman peak analysis treatment the measured weight gain of the sample corresponded to stoichiometric PuO2 separately For Raman laser heating experiments, the Raman measured by thermo-gravimetry (TG). Since, in laser was used as an excitation and heating source. The addition, no further change was noticed after annealing laser power was then increased stepwise and the the samples for a second time, it was considered that Stokes/anti-Stokes Raman spectra were collected at PuO2.00 stoichiometry had been reached. A lattice different powers. parameter of 0.5396(1) Å was measured by X-ray The absolute temperature (TSAS) was derived from the diffraction (XRD), corresponding to the literature value ratio (RSAS) between the experimental intensities of the 19 for PuO2. The average crystallite size measured by anti-Stokes and Stokes T2g peaks according to Bose- XRD and estimated with the help of scanning electron Einstein statistics. Prior to these calculations, Stokes microscopy (SEM) on a polished surface, was (5 ± 3) and anti-Stokes Raman spectra were corrected for the µm. instrumental response according to the procedure The isotopic composition of the plutonium employed described elsewhere.21, 22 Despite the instrumental was checked by High Resolution Gamma Spectroscopy correction, the value of TSAS can represent the real (HRGS) to be 93.4 wt.% 239Pu and 6.4 wt.% 240Pu. sample temperature in a rather inaccurate fashion, due Traces (<0.2 wt.%) of 241Pu and 241Am stemming from to uncontrollable sample-dependent factors affecting the β-decay of 241Pu were detected by Thermal Ionisation intensities of the Stokes and anti-Stokes lines, such as Mass Spectrometry (TIMS). impurity fluorescence, roughness scattering etc. Therefore, the value of TSAS should be considered as B.