AVS 69 Session AC+MI+TH-TuM: Magnetism, Electron Correlation, and Superconductivity in the Actinides/Rare Earths

As new classes of superconducting materials emerge, puzzles of high-temperature superconductivity continue to be one of the pressing issues in condensed matter physics and solid-state chemistry. In particular, iron pnictide superconductors still pose many open questions. Surprisingly, very few actinide-based analogues of iron pnictide high temperature superconductors have been reported so far, perhaps as a result of synthesis complications imposed by toxicity, reactivity, and high vapor pressure of constituent elements. In this work, we revisit the U-Fe-As ternary, in which only one compounds has been reported to exist so far – UFeAs₂ (P4/nmm space group) [1]. By implementing flux synthesis, we were able to grow large single crystals of UFe₅As₃ compound, which adopts UCr₅P₃ structure type [2] (P2₁/m, mP18) with lattice parameters a = 7.0501(17) Å, b = 3.8582(9) Å and c = 9.6342(13) Å, β = 100.2(8)° [3]. The refined composition agrees well with the elemental ratio established by the EDX analysis – U_9.9(2)Fe_58.7(2)As_31.4(2). The magnetic behavior of UFe₅As₃ was studied by magnetic susceptibility measurements in the temperature range 1.8 - 600 K and an antiferromagnetic ordering below T_N = 56 K was established. The Sommerfeld coefficient g = 138 mJ/mol K² indicates enhanced effective electron mass. More in-depth investigations of this and other U-Fe-As ternary compounds are currently underway.

References:

[1] D. Kaczorowski, J. Alloys. Compd. 186, 333-338 (1992).

[2] W. Jeitschko, R. Brink, P.G. Pollmeier, Z. Naturforsch. (1993) 48b, p. 52-57.

[3] N.Zaremba, M. Krnel, Yu. Prots, A. Leithe-Jasper, Yu. Grin, E. Svanidze, in preparation (2023)

Specific features of 5f magnetim in U-based systems, namely the strong spin-orbit coupling and very strong magnetic anistropy or giant magnetoresistance, are degraded by low temperatures of magnetic ordering. The variability of T_C or T_N in intermetallics has been explored and it turns out that involvement of magnetic d-metals such as Fe or Co does not help, unless very high fraction of the transition metals turns a compound in fact a regular d-magnet, in which the unique 5f features are suppressed. The reason is the 5f-d hybridization with a destabilizing action on both d and 5f moments.

Seeking additional degrees of freedom we turned towards compounds with more polar character of bonding, which drives U to the verge of metallicity. Interesting alternative to intermetallics is provided by Zintl phases, typically ternary compounds with one cation (e.g. an f-metal) and a complex covalently bond anion, stabilized by a charge transfer from the cation. An excellent insight is provided by such compounds with layered crystal structures, which leads to strongly anisotropic electronic properties. Our research focuses on Zintl phases with the trigonal CaAl₂Si₂ type (P-3m), formed by alternating cationic (Ca) and anionic (Al-Si) layers. Those without f-metals or with 4f metals tend to be narrow-gap semiconductors. U-based phases such as UCu₂P₂ are semi-metallic. The prominence of this compound dwells in ferromagnetism with high Curie temperature T_C = 216 K, which is further enhanced by pressure application to the vicinity of room temperature. Ab-initio calculations reveal it this material is actually a half-metal, with spin-split 5f band bringing spin-up only states to the Fermi level. The conduction electron states remain hybridized with the 5f states, which causes also the non-f electrons to be fully spin polarized. The magnetisms is of local moment type, with magnetic entropy exceeding R*ln2, but arising from 5fbands. UCu₂P₂ can be compared with EuZn₂P₂. Here the Eu²⁺ moments are organized in an AF structure. The field alignment of the moments impacts the width of the gap seen in transport data, but the material remains semiconducting.

This work was supported by the Grant Agency of the Czech Republic under the grant No. 21-09766S.

In the first part, we investigate the electronic structure of europium sulfide (EuS) with the aim to understand the valence-to-core resonant inelastic x-ray scattering (RIXS) spectra measured at the Eu L₃ edge. We show that the main part of the observed signal comes from the direct RIXS: an Eu 2p core electron is excited to an empty Eu 5d band above the Fermi level, and then another electron from an Eu 5d state hybridized with the S 3p bands (located below the Fermi level and hence occupied) fills back the core hole. Besides this straightforward channel, the measured RIXS spectra display two satellite features. After considering several candidates (such as excitations of the 4f shell excited by indirect RIXS processes, that is, by the interaction with the core hole in the intermediate state), we arrive at the conclusion that the satellites are excitons formed by a 4f hole and an 5d electron localized at the same Eu atom. Such excitons were suggested in the context of optical absorption a long time ago [1], but the concept is not widely accepted. It was argued that existence of these excitons is incompatible with photoconductivity measurements [2]. Our observations indicate that these arguments may need to be revisited.

In the second part, we simulate valence-to-core RIXS at the actinide M₅ edge in selected actinide oxides starting from the dynamical mean-field electronic structure obtained earlier [3] and using techniques similar to [4]. This investigation is motivated by the recent experimental study of UO₂ and UF₄ [5]. Our preliminary results suggest that the feature observed at an energy loss of roughly 10 eV above the white line is likely a charge-transfer excitation, the intensity of which is sensitive to the hybridization between the actinide 5f states and the ligand 2p states.

This work was supported by the Czech Science Foundation under the grant No. 21-09766S.

[1] T. Kasuya and A. Yanase, Rev. Mod. Phys. 40 (1968) 684, https://doi.org/10.1103/RevModPhys.40.684
[2] P. Wachter and B. Bucher, Physica B 408 (2013) 51, https://doi.org/10.1016/j.physb.2012.09.018
[3] J. Kolorenc, A. Shick, A. I. Lichtenstein, Phys. Rev. B 92 (2015) 085125, https://doi.org/10.1103/PhysRevB.92.085125
[4] J. Kolorenc, Physica B: Condensed Matter 536 (2018) 695, https://doi.org/10.1016/j.physb.2017.08.069
[5] J. G. Tobin et al., J. Phys.: Condens. Matter 34 (2022) 505601, https://doi.org/10.1088/1361-648X/ac9bbd

Historically, the bulk of actinide oxide research has focused on the properties of various uranium and plutonium compounds as they impact their use in the nuclear fuel industry. While the fundamental understanding of compounds such as UO₂ has improved, a fundamental grasp of the physical properties of other actinide oxides remains elusive. In actinide systems, the 5f electrons experience a uniquely delicate balance of effects and interactions having similar energy scales, which are often difficult to properly disentangle. The interplay of factors such as the dual character of 5f-states, competing interactions, and strong spin-orbit coupling results in magnetically unusual and intriguing behavior: multi-k antiferromagnetic ordering, multipolar ordering, mixed valence configurations, and more. In this presentation, I will provide an overview of the available synthesis techniques for selected binary actinide oxides. I will also show the current state of knowledge on their crystal structures and magnetic properties. Moreover, I will talk about the future opportunities that are vital for a fundamental understanding of these systems.

UCu₂P₂, a Zintl phase with the trigonal CaAl₂Si₂ structure type, is a 5f ferromagnet with a record-high Curie temperature among U compounds, T_C = 216 K [1]. While the size of the magnetic moment 2.0 μ_B/U is not surprising due to U-U spacing exceeding the Hill limit, the reasons for the high T_C are less understood. Ab-initio calculations reveal only very weak hybridization of the U-5f states with the 6d states as well as with electronic states of Cu and P. It seems that a transfer of U-6d states to the P-3p states is an important ingredient, which was highlighted by a rapid increase of T_C under hydrostatic pressure so that a room-temperature 5f ferromagnetism could be demonstrated. Besides magnetization, transport, and heat capacity studies on single crystals we performed also characterization of a polycrystalline material, which has T_C = 219 K. Thermal expansion study revealed a moderate increase of both lattice parameters just below T_C, so we can exclude that the pressure enhancement of T_C is driven simply by thermodynamics (via the Ehrenfest relation). Hence the reasons have to be attributed to the enhancement of specific U-U couplings upon compression. Indeed, ab initio calculations probing the energy enhancement upon moments reversal gave a semi-quantitative account of the observed tendency of T_C.

[1] D. Kaczorowski, R. Troc, Magnetic and transport properties of a strongly anisotropic ferromagnet UCu₂P₂, J. Phys. Condens. Matter, 1990, 2, 4185.

*This work was supported by the Czech Science Foundation under the grant No. 21-09766S.

The Unoccupied Density of States (UDOS) of ThO₂, UO₂ and PuO₂ have been investigated with a combined experimental and theoretical approach.^1,2Ligand field density functional theory calculations of the dioxides of thorium, uranium, and plutonium have been combined with high-energy-resolution fluorescence detection (HERFD) in x-ray absorption spectroscopy and inverse photoelectron spectroscopy (IPES) measurements to provide powerful insight into the underlying composition of the unoccupied 5f electronic structure in these 5f localized systems. Fine structure in the 5f_5/2 transitions in HERFD can be directly correlated with the fine structure in the leading edge of the IPES. The shapes, intensities, and systematics in HERFD and IPES are explained in a consistent and rigorous fashion in terms of the j-specific 5f electronic structure.Additionally, an actinide N_4,5 branching ratio analysis of PuO₂ and UO₂ has been performed, including measurements with a scanning transmission electron microscope at the Advanced Light Source and simulation with FEFF. It is shown that the 5f occupation of the Pu in plutonium dioxide is n = 5.

References

1. J. G. Tobin,H. Ramanantoanina, C. Daul, S.-W. Yu, P. Roussel, S. Nowak, R. Alonso-Mori, T. Kroll, D. Nordlund, T.-C. Weng, D. Sokaras, "The Unoccupied Electronic Structure of Actinide Dioxides," Phys. Rev. B 2022, 105, 125129.https://doi.org/10.1103/PhysRevB.105.125129
2. J. G. TobinandS.-W. Yu, "Pu5f Occupation in Plutonium Dioxide," Inorg. Chem. 2023, 62, 6, 2592–2598, https://doi.org/10.1021/acs.inorgchem.2c03202