AVS2015 Session AC+AS+MI-WeM: Magnetism, Complexity and Superconductivity in the Actinides and Rare Earths

Plutonium (Pu) is arguably the most complex elemental metal known because its 5f electrons are tenuously poised at the edge between localized and itinerant configurations. This complex electronic structure leads to emergent behavior—all a direct consequence of its 5f electrons—including six allotropic phases, large volumetric changes associated with these transitions of up to 25%, and mechanical properties ranging from brittle α-Pu to ductile δ-Pu. Pu also exhibits a Pauli-like magnetic susceptibility, electrical resistivity and a Sommerfeld coefficient of the specific heat that are an order of magnitude larger than in any other elemental metal. Finally, while experiments find no sign for static magnetism in Pu, most theories that use the correct volume predict a magnetically ordered state. This discrepancy might be reconciled by recent Dynamical Mean Field Theory (DMFT) calculations that suggest that the electronic ground state of δ-Pu is a quantum-mechanical admixture of localized and itinerant valence configurations. The question whether the ground state of δ-Pu is indeed a true quantum-mechanical superposition may only be answered via observation of the associated virtual valence (charge) fluctuations among the distinct 5f⁴, 5f⁵, and 5f⁶ configurations. The characteristic energy scale for the associated spin fluctuations is expected to T_K= 800 K (E_sf ≈ 70 meV) that will result in a dynamical spectral response centered at this energy for T < T_K. We have performed high-energy inelastic neutron spectroscopy at room temperature using a large polycrystalline sample of δ-²⁴²Pu with a total mass of m ≈ 21 g at the Lujan Center and at the Spallation Neutron Source. Our measurements demonstrate the existence of high energy magnetic fluctuations centered at E_sf = 84 meV, in good agreement with the DMFT calculations. In addition, they allow us to extract the magnetic form factor of δ-Pu, yielding critical information about its valence state. These unprecedented results place show that the magnetism in Pu is not "missing" but dynamic, but dynamic, and is driven by virtual valence fluctuations. Our measurements provide a straightforward interpretation of the microscopic origin of the large, Pauli-like magnetic susceptibility of δ-Pu and associated Sommerfeld coefficient. Furthermore, because the various valence configurations imply distinct sizes of the Pu ion, the valence-fluctuating ground state of Pu also provides a natural explanation for its complex structural properties and in particular the large sensitivity of its volume to small changes in temperature, pressure or doping.

Interfacial exchange interaction in bilayers consisting of two dissimilarly ordered magnetic materials (e.g. an antiferromagnet (AF) and a ferro- or ferrimagnet (F)) may give rise to a phenomenon called the magnetic exchange bias (EB) effect [1]. The EB manifests itself as a shift of a magnetic hysteresis loop along the field direction when the bilayer is field-cooled below the Néel temperature (T_N) of the AF. This property is of great value for magnetic recording applications. Despite the conceptual simplicity, a generally accepted theory that predicts the EB behavior for an apt pair of materials is still missing. The reason for that might be in poorly defined interface structure in both magnetic and crystallographic aspects.

Critical dependence of EB on magnetic anisotropy brings us the possibility to use actinides with strong spin-orbit interaction as the key ingredient. Here we report exchange bias studies in magnetic bilayers consisting of a stoichiometric UO₂ film grown epitaxially on different substrates and covered with polycrystalline metallic (Ni₈₀Fe₂₀ and Fe) and highly textured oxide (Fe₃O₄) layers of variable thickness. Large longitudinal exchange bias ~2.6 kOe is found in UO₂/Fe₃O₄ bilayers [3] while UO₂ combined with metallic ferromagnets displays perpendicular exchange coupling with an order of magnitude smaller EB. Interestingly, unusual effects in UO₂/Fe₃O₄ were observed, namely, exchange bias did not vanish at T_N of UO₂. Apart from the fact that single layers of magnetite were showing some EB (not more than 25 % of the total effect in UO₂/Fe₃O₄), the EB in UO₂/Fe₃O₄ bilayers exceeded notably that of the single Fe₃O₄'s to approx. 70 K that was attributed to possible proximity effects of Fe₃O₄ on T_N of UO₂ and/or to the magnetic anisotropy of UO₂ preserved locally above T_N. The effects were observed for the samples of quality controlled by different methods such as X-ray Photoelectron Spectroscopy, conventional X-ray Diffraction, Transmission Electron Microscopy and Rutherford Backscattering Spectroscopy. The work has been supported by the Czech Science Foundation, grant No. 13-25866P.

[1] W. H. Meiklejohn and C. P. Bean, “New magnetic anisotropy”, Phys. Rev. B 102, 1413 (1956).

[2] V. Sechovsky, L. Havela, in: Magnetic Materials, K.H.J. Buschow (Ed.), Elsevier, Amsterdam, 1998, Vol. 11, p. 1.

[3] E. A. Tereshina et al., Appl. Phys. Lett. 105, 122405 (2014).

The interplay between different electronic ground states, especially magnetism and superconductivity, has evolved in a climate of discovery in which many of the fundamental rules of condensed matter physics are questioned by materials with unexpected properties. These “emergent properties,” such as complex magnetism, heavy-fermion superconductivity, the coexistence of magnetism and superconductivity, and/or Kondo physics emerge from complex materials in which quasiparticles develop different states of organization and correlation. The majorities of these bizarre electronic ground states are encountered in f-electron systems and are linked to the hybridization between the f-states and ligand electrons. How these properties evolve with the progressive filling of the f-shells remains an open question, but it is a key ingredient for their understanding. The 4f and 5f strongly correlated electron systems at the border of magnetism are of active current interest, particularly because the accompanying quantum criticality provides a route towards both strange-metal, non-Fermi-liquid behavior, and unconventional superconductivity. In spite of large theoretical and experimental efforts the nature of the electronic behaviors is still unclear. One way to address the electronic properties of these fascinating materials is it to perform extensive transport studies such as Hall, Nernst, or Seebeck effects. In particular, the latter one has gained importance in recent years in thermoelectric materials as potential solutions for applications, such as spot cooling of electronic components, waste heat recovery system and/or remote power generation in space stations and satellites. In addition, the Seebeck coefficient is a sensitive probe of energy relative to the Fermi level, it can therefore be used as a tool to characterize the electronic structure of materials, especially in the vicinity of the narrow gap or pseudo-gap. During the talk I will provide a general introduction to the magnetic and transport characteristics of 4f and 5f electron systems. Then, I will present thermoelectric properties of selected 4f and 5f materials and give an overview on how the thermoelectric power studies can be used to probe electronic properties in this class of materials. I will discuss implications of the results, and their limitations.

We have been studying some members of large family of A₂T₂X (A = Rare-Earth (RE) or actinide, T = transition metal, X = p-metal) compounds crystallizing in the Mo₂FeB₂ structure type (space group P4/mbm). U₂T₂X interact with H₂ only at high pressure (≈ 100 bar) reaching 2 H/f.u. The H absorption produces a lattice expansion, while the tetragonal structure is preserved. The H atoms presumably enter the 8k position inside the U₃T tetrahedra occupied randomly up to 50%. On the other hand, it was found that some RE₂T₂X compounds can absorb more hydrogen compared to their U-counterparts in much lower H pressures. The amorphization of the structure upon hydrogenation (4 H/f.u.) was found for RE₂Pd₂In(Sn) with light RE (La, Nd). RE₂Pd₂In(Sn) with heavy RE behave similar way as U₂T₂X compounds (2 H/f.u., crystal structure type not changed). For the isostructural indide Nd₂Ni₂In the hydrogen absorption of 7 H/f.u. leads to the orthorhombic distortion of the crystal structure.

Recently, we have been studying several isostructural RE₂T₂Mg (T = Ni, Pd). The H absorption in these compounds depends on the type of transition metal. The compounds with Ni tend to absorb up to 8 H/f.u. The crystal structure is changed to monoclinic (space group P21/c). If the transition metal is Pd, the hydrogen absorption is lower, reaching approx. 6 H/f.u. and the crystal structure is modified in different way. The Tb₂Pd₂Mg-hydride crystallizes in a new ternary structure type (Fmmm). In the case of Mg-compounds, the hydrogen absorption in not reversible. Upon desorption of H the initial crystal structure is not restored.

We have found that in U-compounds the ordering temperatures increase upon hydrogenation contrary to the RE-compounds where the ordering temperatures are dramatically reduced. Magnetic properties of U-compounds strongly depend on the inter-U distances. Hydrogen intrusion modifies the lattice by expanding it without changing the crystal-structure type leading to a band narrowing. As a consequence doping of U intermetallics by interstitial hydrogen leads to stronger magnetic properties. On the other hand, the hydrogen absorption has opposite effect on magnetic properties of RE₂T₂X compounds. For RE compounds, hydrogenation affects mainly the 4f – magnetic moments and their ordering. The exchange coupling is reduced presumably by reducing the concentration of conduction electrons.

A previously unknown neptunium-transition-metal binary compound Np₂Co₁₇ has been synthesized and

characterized by means of powder x-ray diffraction, ²³⁷Np Mössbauer spectroscopy, superconducting-quantum-interference-device magnetometry, and x-ray magnetic circular dichroism. The compound crystallizes

in a Th₂Ni₁₇-type hexagonal structure with room-temperature lattice parameters a= 8.3107Å and c=

8.1058Å. Magnetization curves indicate the occurrence of ferromagnetic order below T_C>350 K. Mössbauer

spectra suggest a Np³⁺ oxidation state and give an ordered moment of μ_Np=1.57μ_Band μNp=1.63μ_B

for the Np atoms located, respectively, at the 2b and 2d crystallographic positions of the P6₃/mmc space group.

Combining these values with a sum-rule analysis of the XMCD spectra measured at the neptunium M4,5 absorption

edges, one obtains the spin and orbital contributions to the site-averaged Np moment [μ_S=−1.88μ_B,

μ _L =3.48μ_B]. The ratio between the expectation value of the magnetic-dipole moment and the spin magnetic

moment (m_md/μ_S=1.36) is positive as predicted for localized 5f electrons and lies between the values

calculated in intermediate-coupling (IC) and jj approximations. The expectation value of the angular part of

the spin-orbit-interaction operator is in excellent agreement with the IC estimate. The ordered moment averaged

over the four inequivalent Co sites, as obtained from the saturation value of the magnetization, is μ_Co~1.6_μB.

The experimental results are discussed against the predictions of first-principles electronic-structure calculations

based on the spin-polarized local-spin-density approximation plus the Hubbard interaction. The structural behavior of Np₂Co₁₇ is investigated by means of high pressure diamond-anvil compression measurements and is compared with that of the isostructural compounds Lu₂Co₁₇ and Lu₂Ni₁₇. The Th₂Ni₁₇-type hexagonal crystal structure is preserved with no measurable discontinuous volume collapses up to the highest achieved pressure, p=43GPa. For Np₂Co₁₇, fits to the Birch-Murnaghan and Vinet equations of state give values of the isothermal bulk modulus and

its pressure derivative of B₀=286GPa and B₀=3, revealing that this Np compound is a highly incompressible solid with stiffness comparable to that of superhard covalently bonded materials. The isothermal equation of state for the studied compounds are in excellent agreement with the results of ab initio fully-relativistic, full potential local spin-density functional calculations. Theoretical estimates of the bulk modulus are given also for Np₂Ni₁₇, for which B₀ is predicted to assume values intermediate between those measured for Lu₂Ni₁₇ and Np₂Co₁₇.

Several routes of preparation of alloyed U trihydrides, UH₃, were discovered. Starting from the U₆T compounds, hydrogenation leads to T atoms embedded in the β-UH₃ structure, with transition-metal atoms T occupying one of the U sites. We have been hydrogenating γ-U alloys, using various transition metals helping (together with ultrafast cooling) to retain the bcc U structure down to low temperatures. As such alloys are much more resistant to hydrogen attack, high pressures of H₂ gas had to be applied. The H absorption corresponds to approx. 3H/atoms per 1 U atom. In none of the cases the alloying metals segregate and two different structures were obtained. The hydrides (UH₃)_1-xZr_x form the α-UH₃ structure, i.e. the bcc structure expands and fills by H. Hence basic electronic properties of α-UH₃ could be established. Starting from U_1-xMo_x, we obtained (UH₃)_1-xMo_x, which tends to be β-UH₃ like, but has the grain size is 1 nm only, i.e. practically amorphous. This amorphous phase also easily accepts additional dopants, as Zr, Fe, Ti, V...however magnetic properties remain only weakly affected. It is quite remarkable that all such materials are ferromagnets with the Curie temperature in the range 160-205 K, even if the active U sublattice is diluted by more than 30% of other metals. In this respect the hydrides are different than conventional band ferromagnets, sensitive to inter-atomic spacings and alloying. Albeit all are metallic, the U-H interaction, which can have a partly ionic character, plays clearly important role. Electronic structure calculations (performed for the α-UH₃ structure and in ferromagnetic or Disordered Local Moment state, with possible random Zr occupancy) suggest a transfer of U-6d and 7s electrons into H-1s states, reducing the hybridization of 5f and non-f states, supporting thus magnetism even if the U-U spacing is below the Hill limit [1].

[1] I. Tkach et al. Phys.Rev. B 91, 115116 (2015).