NAMBE 2023 Session NM-TuA2: Novel Oxides and Superconductors

For the realization of the next generation of fast, energy-efficient nanoelectronics, there is a great need for new materials whose electrical and optical conductivities can be sensitively tuned between high (on) and low (off) states by altering a thermodynamic control parameter, e.g., strain or temperature. Unfortunately, most materials are either metallic or insulating and their conductivities cannot be changed substantially. Materials exhibiting a metal-insulator transition (MIT) at or above room temperature are quite rare, limiting their applicability in devices. The archetypical compound displaying an MIT is VO₂ with a transition temperature of 65 °C. Shortly after the discovery of the MIT in 3d¹ VO₂[1], a similar effect was discovered in 4d¹ NbO₂ albeit at a much higher temperature of 807 °C[2]. Thus far, the 5d¹ analog TaO₂ has remained elusive.

Here, we show the growth of epitaxial thin films of phase-pure, rutile-like 5d¹ TaO₂ using suboxide MBE. This recently developed flavor of MBE makes use of the TaO₂ molecular vapor emanating from a Ta₂O₅ charge in an effusion cell heated to temperatures around 1700 °C[3]. This approach avoids the notoriously unstable electron-beam evaporation of Ta metal and need for subsequent oxidation using a background gas or plasma. The latter is particularly challenging to control in the quest for TaO₂ as the stable bulk phase of tantalum oxide is the 5d⁰ compound Ta₂O₅, a band insulator without an MIT, similar to the case of 3d⁰ V₂O₅ and 4d⁰ Nb₂O₅. In the suboxide MBE approach, the Ta⁴⁺ is delivered to the substrate from a pre-oxidized molecular beam of TaO₂.

Unlike VO₂ that can be formed at back-end-of-line-compatible temperatures below 400 °C[4], we find that exceptionally high substrate temperatures above 1000 °C are needed to crystallize TaO₂ by suboxide MBE. Such high temperatures exceed the range of typical MBE systems but are attainable at the PARADIM Thin Film Facility, an NSF-supported national user facility[5], thanks to a recently installed CO₂ laser-based substrate heater.

Ongoing efforts entail the detailed investigation of the effect of epitaxial strain on the structural and spectroscopic properties of TaO₂ thin films on various substrates. To this end, we are employing an ensemble of X-ray, optical, and electrical transport techniques, searching for signs of a structural and electronic phase transition in this candidate 5d¹ MIT compound.

[1] Morin, F.J., Phys. Rev. Lett.3, 34 (1959)
[2] Janninck, R.; Whitmore, D., J. Phys. Chem. Solids27 (1966)
[3] Schwaigert, T. et al, J. Vac. Sci. Technol. A41, 022703 (2023)
[4] Paik, H. et al., Appl. Phys. Lett.107, 163101 (2015)
[5] www.PARADIM.org

Strain-engineering is a powerful means to tune the polar, structural, and electronic instabilities of ferroelectrics. KTaO₃ is an incipient ferroelectric, with a very large spin-orbit coupling, in which highly anisotropic superconductivity emerges near a polar instability in electron doped samples[1, 2]. Growth of high-quality epitaxial films provides an opportunity to use epitaxial strain to finely tune electronic and polar instabilities in KTaO₃. KNbO₃ is a well-known ferroelectric with multiple structural transitions[3]. Using a molecular beam of the suboxides TaO₂ and NbO₂ emanating from effusion cells containing Ta₂O₅ or Nb₂O₅ in combination with a molecular beam of potassium emanating from an indium-potassium intermetallic in an oxidant (~10% O₃ + 90% O₂) background pressure of 1x10^–6 Torr, KTaO₃, KNbO₃ and KNb_xTa_1-xO₃ films are grown under conditions of excess potassium in an absorption-controlled regime. Biaxial strains ranging from -0.1 % to -2.1 % are imposed on the commensurately strained KTaO₃ films by growing them upon SmScO₃, GdScO₃, TbScO₃, DyScO₃ and SrTiO₃ substrates, all with the perovskite structure. Reciprocal space mapping shows the epitaxial thin films are coherently strained to the underlying perovskite substrates provided the films are sufficiently thin. Cross-sectional scanning transmission electron microscopy does not show any extended defects and confirms that the films have an atomically abrupt interface with the substrate. X-ray diffraction rocking curves (full width at half maximum < 30 arc sec on all of the above substrates) are the narrowest reported to date for KTaO₃, KNbO₃ and KNb_xTa_1-xO₃

films grown by any technique. Laue fringes confirm that the films are smooth with a well-defined thickness. Atomic force microscopy reveals atomic steps at the surface of the grown films. SIMS measurements confirm that the films are free of indium contamination.

References

[1] Ueno K et al. Discovery of superconductivity in KTaO₃ by electrostatic carrier doping. Nat. Nanotechnol. 2011; 6:408

[2] Bruno FY et al. Band structure and Spin-orbital Texture of the (111)-KTaO₃ 2D Electron Gas. Adv. Electron. Mater. 2019; 5:1800860

[3] Hewat, A. W. Cubic-tetragonal-orthorhombic-rhombohedral ferroelectric transitions in perovskite potassium niobate: neutron powder profile refinement of the structures. J. Phys. C: Solid State Phys. 1973; 6.16:2559.

SrCoO_x (SCO) exhibits contrasting crystalline, electronic, and magnetic states with varying oxygen content. While in brownmillerite (BM) phase, SrCoO_2.5 is anantiferromagnetic insulator with ordered oxygen vacanices, SrCoO₃ is a ferromagnetic metal with the cubic perovskite (P) structure. Although SCO has been a relatively well-studied system, its growth using molecular beam epitaxy (MBE) has remained very limited. P-phase SCO films show high sensitivity to cation stoichiometry and oxygen chemical potential, with secondary phases present depending on growth conditions. In this work, we grew SrCoO_x using MBE- via both shuttered and co-deposition methods- in oxygen plasma, with 2.5 ≤ x ≤ 2.87 as suggested by X-ray diffraction (XRD). X-ray absorption study (XAS) of Co L and O K edges confirms Co³⁺ or higher cobalt oxidization states. In-situ reflection high-energy electron diffraction (RHEED) indicates excellent quality and highly epitaxial growth of our films. Stoichiometry ratios between Co and Sr were determined using in-vacuo X-ray photoelectron spectroscopy (XPS), followed by Rutherford backscattering (RBS). We grew our films on three different substrates: LaAlO₃ (LAO), La_0.3Sr_0.7Al_0.65Ta_0.35O₃ (LSAT), and SrTiO₃ (STO) which have lattice mismatches of - 1.0%, 1.0% and 2.0% with P-SCO respectively. Therefore, our study also provides a scope to explore strain-induced oxygen vacancies in the SCO films and their impacts on phase stability. Measurements of these phases were performed using temperature dependent transport property measurements and Scanning transmission electron microscopy (STEM). An evolution of electronic structure of 3d SCO films when coupled with 5d Ir/ Ta based systems in superlattice/double-perovskite structures can provide a further scope to study charge transfer in metastable oxide perovskites.

While paraelectric (Ba,Sr)TiO₃ films were once used as tunable dielectrics in radio frequency (RF) circuits, dielectric loss above 10 GHz renders (Ba,Sr)TiO₃ incompatible with the high frequency future of RF electronics.¹The related Ruddlesden-Popper titanates—(ATiO₃)_nAO with A = (Ba,Sr)—have demonstrated low loss up to 100 GHz, but these experiments have used interdigitated capacitors compatible with the in-plane dielectric tunability of these phases,^2,3 rather than commercially preferable¹ metal-insulator-metal (MIM) capacitors requiring out-of-plane dielectric tunability. To achieve out-of-plane tunability in a Ruddlesden-Popper film, first-principles calculations indicate the concentration of barium and the series member, n, should both be maximized,^3,4 but synthesizing such films is extremely challenging. Here, we refine existing synthesis techniques to grow a high-n Ruddlesden-Popper (n = 20), containing the highest concentration of barium ever accomplished in a Ruddlesden-Popper (A = Ba_0.6Sr_0.4) as shown in Fig. 1(b).⁵ With a firm grasp on the synthesis, we have demonstrated epitaxially strained heterostructures of (ATiO₃)_nAO dielectric layers with metallic SrRuO₃ electrodes (Fig. 1(b)). Measurements confirm that such Ruddlesden-Popper films are, in fact, ferroelectric (Fig. 1(c)) and that the dielectric constant is highly tunable at room temperature (Fig. 1(d)). To assess their relevance to the future of tunable dielectrics for GHz electronics, it remains to evaluate the dielectric loss of these new phases at frequencies greater than 10 GHz.

References

¹ G. Subramanyam, M.W. Cole, N.X. Sun, T.S. Kalkur, N.M. Sbrockey, G.S. Tompa, X. Guo, C. Chen, S.P. Alpay, G.A. Rossetti, K. Dayal, L.Q. Chen, and D.G. Schlom, J. Appl. Phys. 114, 191301 (2013).

² C.-H. Lee, N.D. Orloff, T. Birol, Y. Zhu, V. Goian, E. Rocas, R. Haislmaier, E. Vlahos, J.A. Mundy, L.F. Kourkoutis, Y. Nie, M.D. Biegalski, J. Zhang, M. Bernhagen, N.A. Benedek, Y. Kim, J.D. Brock, R. Uecker, X.X. Xi, V. Gopalan, D. Nuzhnyy, S. Kamba, D.A. Muller, I. Takeuchi, J.C. Booth, C.J. Fennie, and D.G. Schlom, Nature 502, 532 (2013).

³ N.M. Dawley, E.J. Marksz, A.M. Hagerstrom, G.H. Olsen, M.E. Holtz, V. Goian, C. Kadlec, J. Zhang, X. Lu, J.A. Drisko, R. Uecker, S. Ganschow, C.J. Long, J.C. Booth, S. Kamba, C.J. Fennie, D.A. Muller, N.D. Orloff, and D.G. Schlom, Nat. Mater. 19, 176 (2020).

⁴ K. Lee, W. Lee, M. Jeong, Y. Kim, E. Kim, H. Mun, J. Lee, C. Lee, K. Cho, D.G. Schlom, C.J. Fennie, N.M. Dawley, G.H. Olsen, and Z. Wang, U.S. Patent No. 11,133,179 B2 (2021).

⁵ M.R. Barone, M. Jeong, N. Parker, J. Sun, D.A. Tenne, K. Lee, and D.G. Schlom, APL Mater. 10, 91106 (2022).

Complex nickelates (R/A)NiO₃ (where R denotes lanthanide and A denotes alkaline earth metals) are of great interests owing to their diverse structures and functionalities. Dynamic tunability of Ni valence states in nickelate thin films (from 1+ to 4+) has offered emergent properties such as superconductivity, enhanced electrocatalytic activity, quantum confinement effect, and metal insulator transitions. Despite many efforts, stabilization of the quadrivalent (Ni⁴⁺)state through Sr doping in rare earth nickelates has been proven difficult. In this talk, I will present our effort in growing perovskite La_1-xSr_xNiO₃ in thin film and superlattice forms by oxygen-plasma assisted molecular beam epitaxy. We show that it is difficult to achieve high-quality single-phase epitaxy at higher Sr concentration (x>0.5) as bulk SrNiO₃ has a hexagonal phase. Phase segregation (SrNiO_x to SrNi₂O₃ + Sr₂NiO₃) and phase transition (perovskite to hexagonal) are revealed for the end member at different growth and strain conditions by XPS, XRD, and STEM.

In bulk, the Bi₂Sr₂Ca_n–1Cu_nO_2n+4 (i.e., Bi-cuprate) family provides the highest superconducting transition temperature—up to 105 K [1]—without highly toxic constituents like thallium or mercury. When prepared as epitaxial films, however, the superconducting transitions of this family are broad and depressed for as-grown epitaxial films, with the highest zero-resistance T_c reported to date of 97 K [2]. In this work, we use ozone-assisted MBE to grow Bi₂Sr₂Ca_n-1Cu_nO_2n+4 (n=1–3) single-layers and bi-layers on (100) SrTiO₃ substrates, where weuse a non-superconducting Bi₂Sr₂CuO₆ (Bi-2201, n=1) buffer layer to reduce disorder caused by the underlying substrate. We demonstrate that combining adsorption-controlled co-shuttered growth and the initial precise flux calibration provides the best structural and superconducting properties for Bi-cuprate films formed at growth rates of 0.1 µm/hr or higher.

In-situ reflection high-energy electron diffraction exhibits incommensurate structural modulation and also proves the absence of any secondary phases. The structural quality of the samples is confirmed using a combination of four-circle x-ray diffraction, atomic force microscopy, and transmission electron microscopy. The superconducting properties are studied by resistivity vs temperature measurements. Bi-2201/Bi-2212 bilayers exhibit the smoothest reported surfaces with subnanometer root-mean-square (rms) roughness of ~0.4 nm and the sharpest superconducting transition width (ΔT_c) ~10 K, similar to Bi-2212 single crystals. While the T_c(R=0) for the as-grown bilayer is low, i.e., ~50 K, this is a matter of oxygen content and can be modified via a post-growth process. This presentation focuses on achieving the high-structural and surficial quality and the sharp ΔT_c of the as-grown samples. We conclude that combining the co-shuttered growth and the initial precise flux calibration provides the best structural and superconducting properties [3].

FeSe, a bulk superconductor with a T_C of 9K has attracted a high level of attention since a skyrocketing boost in T_C was reported for a single unit cell (UC) layer of FeSe grown on SrTiO₃(001) by molecular beam epitaxy (MBE) to as high as 100K. FeSe-SrTiO₃ heterostructures have since been fabricated by many groups but the record T_C proved difficult to reproduce and thus the mechanism behind it remains concealed. After extensive work in the past, the field appears to agree on certain key “ingredients” in the sample preparation that are believed essential for the boost in T_C. Those are; 1. an ultra-clean substrate surface of a double TiO₂ termination realized by a chemical and thermal ex-situ and/or thermal in-situ substrate preparation; 2. ultra-thin – one UC thickness – limit of FeSe; 3. a high number of Se vacancies in the FeSe film ensured through post-growth annealing steps in ultra-high vacuum (UHV) for several hours; 4. followed by a capping layer growth protecting FeSe against oxidation during ex-situ characterization. We present our findings on FeSe thin film growth by MBE and present a roadmap for high-T_C - 222% higher than the reported bulk value in ex-situ transport measurements - circumventing above mentioned steps 1, 2, and 3 by simple in-situ Se/Fe flux ratio and temperature control during FeSe growth. FeSe films of 20-UC-thickness grown at varying temperatures and Se/Fe flux ratios and the structural and morphological properties of the obtained uncapped FeSe films were analyzed. The morphology of the films showed a sensitive dependence on the growth temperature and flux ratio spanning from perfectly smooth and continuous films with atomic terraces at 450 °C growth temperature and a low flux ratio of 2.5 to exclusively disconnected island growth of large height but smooth top surfaces at lower temperatures and/or higher flux ratios. Surprisingly, the tetragonal P4/nmm crystal structure of FeSe was maintained for all investigated films and the in-situ observed diffraction pattern in reflection high energy diffraction also maintained the streaky pattern characteristic for smooth FeSe films even for the samples with the most pronounced island growth resulting in a root mean square atomic force microscopy roughness of more than 18nm. Smaller flux ratios than 2.5 resulted in mixed - FeSe/elemental Fe - phase samples. FeSe films grown under optimized conditions at 450°C and a flux ratio of 2.5 (but without any post-growth UHV anneal) and capped with the commonly used FeTe (300°C) and elemental Te (room temperature) layers yielded superconducting onset temperatures of about 30K and a T_C of 20K.