Role of disorder in magnetic and conducting properties of U–Mo and U–Mo–H thin films

https://doi.org/10.1016/j.matchemphys.2020.124069Get rights and content

Highlights

  • Preparation of U–Mo and UH3–Mo thin films by sputter deposition.

  • Superconductivity in U0.79Mo0.21 below Tc = 0.55 K.

  • UH3–Mo hydride films are ferromagnets; the Curie temperature of (UH3)0.74Mo0.26 is 165 K.

  • Negative temperature coefficient of resistivity, TCR = dρ/dT in all films attributed to the randomness on atomic scale.

Abstract

Crystal structure, magnetic and electrical properties of thin U–Mo films and their hydrides prepared by reactive sputter deposition from metallic targets were studied. The films are metallic but with high values of electrical resistivity. Both bcc U0.79Mo0.21 thin film and the hydride (UH3)0.74Mo0.26 (β-UH3 structure type) have a very strong texture. Superconductivity is identified below Tc = 0.55 K for U0.79Mo0.21. The critical field of 1.0 T is exceeding that of α-U (0.3 T). The hydride film (UH3)0.74Mo0.26 is a ferromagnet with the Curie temperature TC = 165 K, higher concentration of Mo somewhat reduces the TC value. Deposition on a cooled substrate reduces the grain size of the hydride, approaching the limit of amorphous state. The U–Mo films and hydrides exhibit the net negative temperature coefficient of resistivity, TCR = dρ/dT, attributed to the randomness on atomic scale, yielding very strong scattering of electrons and weak localization.

Introduction

In quest for low-enriched fuels, the required combination of high uranium density and mechanical integrity was found in U–Mo alloys. These are formed with the bcc structure. Typically 18 at.% Mo gives a single-phase bcc material, corresponding to the γ-phase of U metal [1], stable above 1049 K [2]. At lower Mo concentrations, the γ-phase is metastable and alloys contain a significant fraction of the orthorhombic α-U-like phase [1]. Both α- and γ-U are superconducting. In the former case, however, the superconductivity is not a bulk property and it is eliminated when the charge density wave state is realized at low temperatures [3]. The bulk superconducting ground state is apparently one of general characteristics of the bcc U-M alloys, formed with a number of transition metals M. The superconducting transition temperature Tc of bcc U–Mo alloys can reach ≈2 K [4,5]. Recently, the low-temperature conductivity of the U–Mo alloys was studied on samples obtained by ultrafast cooling (splat-cooling) [6,7]. This technique allowed stabilization of the pure, but weakly distorted γ0 phase down to room temperature at Mo concentrations as low as 11–12 at.% while the bcc γ-U phase was observed for 13–15 at.% of Mo. The splat-cooled U–Mo alloys demonstrated the increase of Tc from 1.24 K to 2.1 K as the Mo content increased from 0 to 15 at.%.

Superconductivity of the γ-phase is very robust, with the upper critical field μ0Hc2 > 5 T, which is much higher than μ0Hc2 = 0.3 T found for α-U superconductivity [6.7]. Another general feature of γ-U-Mo is an enhancement (in comparison to α-U with 10 mJ/mol K2) of the Sommerfeld coefficient of electronic heat capacity [7]. In the normal state, a striking enhancement of electrical resistivity and gradually developing negative slope (TCR) dρ/dT < 0 was observed and attributed to strong impurity scattering due to random distribution of alloying atoms, giving the potential strongly varying between the U and Mo ions [8]. As the α-U phase can evidently incorporate only very low concentrations of the alloying elements, the γ-phase purity can be actually proved by an absence of any low-temperature ρ(T) decrease.

Although the 5f states of U could in principle give rise to the Kondo effect, at which a strong exchange coupling of conduction electrons to local magnetic moments can provide dρ/dT < 0 due to conduction electrons condensing around local magnetic moments, such phenomena are naturally impossible if the broad 5f band in weak Pauli paramagnetic γ-U phases does not yield local U moments. It is, however, less obvious whether the Kondo effect can be in principle operable in materials where U magnetic moments are formed and ordered at low temperatures. The possibility to prepare hydrides of the U–Mo alloys, which are ferromagnetic due to a pronounced lattice expansion (ΔV/V0 > 60%), and which preserve the randomness of the Mo distribution, led to the recognition that magnetic fluctuations in the paramagnetic state with the chemical randomness both contribute to the weak localization, which enhances resistivity with decreasing temperature. The weak localization can be disrupted both by electron-phonon scattering at elevated temperatures and/or by external magnetic field reducing the magnetic disorder in the nanocrystalline ferromagnet.

The sputter deposition technique can yield systems far from thermodynamic equilibrium, being even more diverse than in the ultrafast-cooled materials. Moreover, it is in principle possible, by means of reactive sputtering, to prepare hydrides of the U alloys, having the same concentrations of metal atoms. Therefore it is tempting to explore the possibilities of the sputtering technique for the U-M alloys and related hydrides for a broad concentration range of M. Here we present results of the study of crystal structure and physical properties of selected films, discussed from the point of view of occurrence of superconductivity and magnetism. The possibility to probe metastable structures was demonstrated in the successful synthesis of UH2 with the CaF2 structure type [9], known so far for most of 4f and 5f elements except uranium, for which UH3 represents the first stable hydride. Recently there has been also an effort to synthesize epitaxially grown single crystalline γ-U-Mo films [10]. However, no data on the conducting properties of the materials was presented. The structure, electrical, and magnetic properties of U–Mo and U–Mo–H films are here compared with available bulk data. We demonstrate how the structure and properties are affected by the substrate temperature. While the hydride films are ferromagnets, the superconductivity was observed for the U–Mo films.

Section snippets

Experimental details

Thin film samples of pure and hydrogenated U–Mo (and pure U for reference) were prepared by reactive sputter deposition realized in a home-built setup, using a miniature U target in the form of a chip (natural uranium, 99.9 wt% purity) and an electron emitting thoriated tungsten filament stabilizing the plasma [11]. Mo wire was used as the second target to produce U–Mo alloys. Typical deposition rates were 0.1 nm/s. After testing runs using a Si wafer substrates, used for the XPS

XPS study of U–Mo and UH3–Mo thin films

Variations of the 5f states in the U–Mo alloys can be generally attributed either to the difference of the α-U and γ-U phases or to the random Mo distribution. As shown in Ref. [7], fundamental parameters such as Sommerfeld coefficient of electronic specific heat γ is only modestly enhanced from γ ≈ 10 mJ/mol K2 to 16 mJ/mol K2 for the U0.85Mo0.15 alloy, being of the single phase bcc structure. Also the Pauli paramagnetic susceptibility remains practically the same, on the 5 × 10−9 m3/mol

Concluding remarks

The films studied in this work are metallic with high values of electrical resistivity. In the thin films we observed properties analogous to bulk materials while having more degrees of freedom to vary at their synthesis. Such parameters as strain, grains size and texture can be tuned by synthesis conditions (use of various substrates, substrate temperatures, deposition rates etc.). Importantly, the (UH3)–Mo thin film hydrides are stable in contrast to bulk UH3, which self-ignites in air when

CRediT author statement

Evgenia A. Tereshina-Chitrova: Experimental investigation, Data curation, Original draft preparation, Ladislav Havela: Conceptualization, Supervision, Writing – Reviewing and Editing, Mykhaylo Paukov: Samples preparation, Milan Dopita: XRD Measurements, Lukáš Horák: XRD measurements, Oleksandra Koloskova: Samples preparation, Zbyněk Šobáň: Preparation of samples for resistivity measurements, Thomas Gouder: Methodology, Samples preparation, Frank Huber: Methodology, Samples preparation, Alice

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The work was supported by the project “Nanomaterials centre for advanced applications”, project no. CZ.02.1.01/0.0/0.0/15_003/0000485, financed by ERDF. We also acknowledge the support of Czech Science Foundation under the grant no. 18-02344S. The work at JRC Karlsruhe was supported by the Actuslab under contract with the European Commission. Part of the authors are grateful to the INTER-COST project (No. LTC18024) for support. Z.S. thanks the Ministry of Education of the Czech Republic, Grants

References (28)

  • V.P. Sinha et al.

    J. Alloys Compd.

    (2010)
  • B.S. Chandrasekhar et al.

    J. Phys. Chem. Solid.

    (1958)
  • T.G. Berlincourt

    J. Phys. Chem. Solid.

    (1959)
  • I. Tkach et al.

    Physica C

    (2014)
  • A.M. Adamska et al.

    Thin Solid Films

    (2014)
  • I. Tkach et al.

    J. Alloys Compd.

    (2012)
  • M. Paukov et al.

    Appl. Surf. Sci.

    (2018)
  • L. Havela et al.

    J. Electron. Spectrosc. Relat. Phenom.

    (2020)
  • R. Troc et al.

    J. Alloys Compd.

    (1995)
  • L. Havela et al.

    Physica B

    (2018)
  • L. Havela et al.

    J. Alloys Compd.

    (2006)
  • L. Havela et al.

    J. Magn. Magn Mater.

    (2016)
  • I. Grenthe et al.

    The chemistry of the actinide and transactinide elements

  • G.H. Lander et al.

    Adv. Phys.

    (1994)
  • Cited by (5)

    • Multi-phase interaction accompanied with the phase transformation process in U-21 at%Nb alloy quantified by in-situ neutron diffraction during heating

      2022, Journal of Nuclear Materials
      Citation Excerpt :

      Uranium and U-based materials have therefore attracted much attention for both the intense scientific interest and significant technological importance [3–6]. Due to the fundamental role on the properties (e.g., mechanical strength and thermal conductivity) of fuel materials [6–9], phase structure and transformation for U-M systems have been experimentally and theoretically investigated for decades [10–19]. For instance, pronounced effort has been made to establish the phase diagrams for those binary systems [10–16].

    View full text