Research

The innovation in materials synthesis drives future technologies and discovery. We work at the interface of materials science and chemistry to alleviate the challenges of incorporating “stubborn” metals that are difficult to evaporate and difficult to oxidize. Through experimental studies within our group and collaborative computational studies, we seek to identify the factors at the atomic and electronic levels that determine how crystals of “stubborn” metals and their oxides grow in ultra-high vacuum. Our objectives are threefold: 

• Develop low-energy, radical-based MBE, and chemical beam epitaxy (CBE) for single-crystalline films of stubborn metal oxides (rutile, perovskite, Ruddlesden-Popper phases)

• Study the growth kinetics of stubborn-metal-containing oxides which are thermodynamically unstable in their bulk form

• Investigate the relationship between growth kinetics, defect structure and electronic properties

Figure 1: Different methods of supplying metals and oxygen in an oxide molecular beam epitaxy system. Adapted from W. Nunn et al., J. Mater. Res. 36, 4846 (2021).

Selected Reading:

Perovskite-based alkaline-earth stannates possess wide-to-ultra-wide band gaps that make them highly suitable candidates for transparent conductors, power electronic devices, and high electron mobility transistors. The perovskite structure is uniquely suited to these applications because (1) they are structurally compatible with and can be integrated on silicon and (2) have exceptionally versatile properties that can be modified by tuning their composition and strain. 

While the high room-temperature mobility in bulk single crystals of alkaline-earth stannates is exciting, replicating this property in thin films has been a long standing challenge. We grow thin films and heterostructures of alkaline-earth stannates (Ba,Sr,Ca)SnO₃ using radical-based MBE which achieves the atomic level control required to study the mobility-limiting mechanisms in these thin films. Through a combination of experimental and computational techniques, we investigate the factors that influence electronic properties such as defect formation, structural phase transformation, and strain, and develop synthesis strategies to overcome these bottlenecks. 

Figure 2: Tuning the magnitude and the type of bandgap by alloying in alkaline-earth stannates and germanates. (a) Band gap as a function of pseudocubic lattice parameters of alkaline-earth stannates and germanates. The top axis shows the pseudocubic lattice parameters of commercially available single crystalline substrates. Crystal structures of alkaline-earth (b) stannates and (c) germanates. ^† denotes metastable crystal structures quenched to ambient conditions. * represents a rhombohedral distortion at room temperature. Adapted from F. Liu et al., Commun. Mater. 3, 69 (2022).

Selected Reading:

Complex oxides with perovskite structure (ABO₃) are impressively multi-functional, exhibiting exotic properties such as colossal magnetoresistance, multiferroicity, and strongly-correlated electron behavior. Heterostructures of these materials also display interface-stabilized ground states such as two-dimensional electron gas (2DEG), 2D superconductivity, and novel magnetism, which do not exist in their bulk single crystal counterparts. These phenomena are extremely sensitive to composition and can be gainfully controlled by designing novel devices such as Mott field-effect transistors. 

We investigate the role of disorder and doping on superconductivity in (rare-earth) titanate thin films and heterostructures, their unusual magnetic ground states and strongly-correlated Mott-Hubbard-type insulator characteristics. With the excellent control over stoichiometry, dimensionality and strain offered by hybrid MBE, we seek to understand their effect on properties such as exotic magnetism and superconductivity using "built-to-order" quantum structures.

Figure 3 (left): Analyzing the broken-gap of SrTiO₃/NdTiO₃ heterostructures. (a-b) density of states for [STO]₉/[NTO]₁ and [STO]₆/[NTO]₄heterostructures. (c) Average electrostatic potential (red line) and Fermi energy E_F (dashed line) calculated for the [STO]₆/[NTO]₆ heterostructure strained coherently on LSAT substrates. Adapted from P. Xu et al., Adv. Mater. Interfaces 3, 1500432 (2016).

Figure 3 (right): Explaining the measured dielectric constant using band bending and extracting its intrinsic value in homoepitaxial SrTiO₃ films. (A) Schematic energy diagram for our homoepitaxial capacitor structure showing band bending due to Fermi level equilibration (left) and the resultant equivalent capacitance (right). (B) A/C_eff as a function of the thickness of the undoped SrTiO₃ layer at different temperatures. (inset) shows A/C_eff vs. t at 2 K. Dashed lines are linear fits. (C) Extracted intrinsic dielectric constant along the [001] crystallographic direction as a function of temperature and their comparison with the corresponding values in bulk SrTiO₃ single-crystals along different crystallographic directions (dashed lines). Adapted from Z. Yang et al., Proc. Natl. Acad. Sci. U.S.A. 119, e2202189119 (2022).

Selected Reading:

4d and 5d transition metal oxides have attracted considerable interest due to their strong spin-orbit coupling, leading to a delicate interplay between the on-site Coulomb interaction and crystal field splitting energy. Consequently, a wide range of exotic phases can emerge, including spin-orbit-assisted Mott insulators, novel magnetic phases such as altermagnets, quantum spin liquids, multipolar ordered materials, unconventional superconductors, and topological semimetals. However, many of these 4d and 5d “stubborn” metals pose serious challenges in conventional MBE due to their low vapor pressures and low oxidation potentials. 

We have developed novel synthesis approaches to enable atomically precise synthesis of 4d/5d metal oxide thin films within the framework of MBE. This approach leverages the properties of solid metal-organic sources of the “stubborn” metals to eliminate the challenges with pure metal sources. We are actively studying ruthenates and iridates, notably RuO₂, SrRuO₃, Sr₂RuO₄, IrO₂, and SrIrO₃, which hold great potential in electrocatalysis and next-generation of electronic and spintronic devices. The high-quality synthesis has deepened our understanding of epitaxial synthesis and enabled the discovery of new quantum phenomena in these material systems.

Figure 4: Effect of different epitaxial strains and IrO₂ film thickness on the structural properties. (a) Schematics showing the epitaxial strain states with the strain type and magnitude along the orthogonal in-plane directions for IrO₂ films grown on (top to bottom) (001), (101), and (110) oriented TiO₂ substrates. (b) X-ray diffraction 2θ–ω coupled scans for thin IrO₂ films grown on these different orientations of TiO₂ substrates. (c–e) Corresponding atomic force micrographs showing atomically smooth surfaces for the (c) (001), (d) (101), and (e) (110) oriented IrO₂ films. (f) X-ray diffraction 2θ–ω coupled scans for thicker IrO₂ films grown on different orientations of TiO₂ substrates. (g–i) Corresponding AFM images showing well-defined cracks along family of planes for IrO₂ (001) (g), rod-shaped Ir metal features on the surface of IrO₂ (101) (h) and a smooth surface for IrO₂ (110) (i) with spot-like surface Ir metal features observed in optical microscopy (inset). Adapted from S. Nair et al., Nat. Nanotechnol. 18, 1005 (2023).

Figure 5: Structural and electrical characterization of a superconducting Sr₂RuO₄ film grown by hybrid MBE. (a) High-resolution X-ray diffraction coupled scan of a 100-nm-thick Sr₂RuO₄ film grown on a LSAT (001) substrate. (b) Temperature-dependent resistivity of the film from 300 to 0.5 K. The inset shows an enlarged view of the superconducting transition. (c) Low temperature resistivity showing suppression of the superconducting transition on increasing the applied magnetic field from 0 Oe to 700 Oe. Adapted from R. Choudhary et al., APL Mater. 11, 061124 (2023).

Selected Reading:

Single crystalline membranes can move past the discrete strain states, reduced flexibility, and availability of lattice-matched substrates that limit the understanding of the intrinsic properties of epitaxially grown and clamped thin films. We study the growth and exfoliation of single crystalline membranes of functional oxides including high-k dielectrics on (1) van der Waals materials such as graphene and (2) in-situ grown, water-soluble sacrificial layers such as alkaline earth oxides. 

In the first approach, we investigate the interactions and interfaces between the hybrid MBE-grown high-k dielectrics, the van der Waals (vdW) material, and the substrate. We focus on understanding the mechanisms for thin film growth on different van der Waals materials to enable the growth of large-area membranes. In the second focus area, we explore thin-film growth on and exfoliation from water-soluble sacrificial layers with a simpler chemical composition and improved exfoliation than conventionally used sacrificial layers. We perform detailed structural and electronic transport studies on membranes grown by both techniques to control specific defects, investigate the local structure, and measure the carrier densities and mobilities to develop high-quality, field-effect devices for nanoelectronics.

Figure 6: Structural properties of SrTiO₃ films grown by hybrid MBE on bi-layer (BL graphene covered SrTiO₃ (001) substrates after growth, exfoliation and transfer onto a host r-plane Al₂O₃ substrate. Confocal Raman spectroscopy and microscopy of (A) BL-Gr/SrTiO₃ (001) before growth, (B) the resulting SrTiO₃/BL-Gr/SrTiO₃ (001) after growth, and (C) the restored BL-Gr/SrTiO₃ (001) after exfoliating the grown film. Each Raman micrograph shows the integrated intensity from one graphene peak scanned over the surface of the sample. (D) HRXRD 2θ-ω–coupled scans of the sample before growth and after growth, exfoliation, and then transfer to an r-plane Al₂O₃ substrate. Adapted from H. Yoon et al., Sci. Adv. 8, eadd5328 (2022).

Figure 7: The growth and exfoliation of SrTiO₃ (STO) and SrTiO₃/CaSnO₃ (STO/CSO) membranes grown on different binary alkaline-earth oxide sacrificial layers. (a) Schematic of the exfoliation and transfer method for an STO membrane. (b) Wide-angle X-ray diffraction 2θ–ω coupled scan of 422 nm, 100 nm, and 11 nm thick STO membranes exfoliated from a SrO sacrificial layer and transferred on an Au-coated Si substrate. The inset shows a representative optical image of an exfoliated PMMA/11 nm STO membrane transferred on an Au/Ti/Si substrate. (c) Fine-coupled X-ray diffraction 2θ–ω coupled scan around the STO (002) peak after the transfer of an STO/CSO heterostructure membrane grown on different Ba_1–xCa_xO sacrificial layers. (d) Comparison of the dissolution time of binary oxide sacrificial layers with complex sacrificial layers demonstrating faster membrane synthesis using binary oxide sacrificial layers. The inset shows a STO/CSO membrane on Si. Adapted from S. Varshney et al., ACS Nano 18, 8, 6348 (2024).

Selected Reading:

The discovery of new methods of generating energy without adversely affecting the environment is a compelling scientific problem of our time. In collaboration with Prof. Richard James in the Department of Aerospace Engineering and Mechanics at the University of Minnesota, we develop devices for the “direct” conversion of heat to electricity [here, “direct” means that electricity is generated by the material itself, without the need for a separate electrical generator]. Using phase-transforming materials with an abrupt change in the polarization during the transformation, we hope to better our understanding of the first-order phase transformation and the origin of thermal hysteresis in hard materials. Guided by the mathematical design of materials (led by the James group), we employ hybrid MBE to synthesize the thin film heterostructures and control their properties using strain engineering and alloying.

Figure 8: Demonstration of direct heat to electricity conversion using a bulk single crystal of BaTiO₃ (001) on cycling temperature across the tetragonal-to-cubic phase transformation (T_c ~ 120 °C) under an applied electric field. An output current of ~ 1 µA is measured across a load resistor across this phase transformation. Figure inspired by and adapted from A. Bucsek et al., Phys. Rev. Appl. 12, 034043 (2019).

Selected Reading:

The ability to control materials’ properties via external stimuli is a powerful approach to investigate materials. We use a field-effect transistor (FET) configuration to modulate the density of charge carriers in oxide semiconductors and semimetals and consequently, to manipulate the material properties. In the past decade, the use of electrolyte-based dielectric materials, such as ionic liquids and ion gels, instead of solid dielectrics has emerged as an alternative to control the electronic and magnetic properties of oxide thin films within a single thin film. By using high-dielectric-constant materials such as SrTiO₃ and electrolytes such as ion gels, we seek to modulate functional oxide thin films and decouple the effects of disorder and electron density on the modulated properties.

Figure 9: Modulation of electrical properties of La-doped BaSnO₃-based MOSFETs. (a) Device schematic, (b) optical micrograph of the fabricated device, (c, d) I_DS vs V_DS characteristics of the device as a function of V_GS measured at 300 and 77 K respectively, (e, f) transfer curve (I_DS vs V_GS) (left axis) and field effect mobility (μ_FE) (right axis) at 300 and 77 K respectively, (g, h) I_DS^1/2 vs V_GS (left axis) and g_m vs V_GS (right axis) at 300 and 77 K, respectively. Adapted from J. Yue et al., ACS Appl. Mater. Interfaces 10, 21061 (2018).

Selected Reading:

Figure 10: Electrostatic gating-based control of the metal-to-insulator transition in La-doped SrSnO₃ films. (a) Schematic of an ion gel-gated Hall bar structure illustrating the formation of EDL at the SSO film/ion gel interface, (b) optical micrograph of the patterned Hall bar, (c) logarithmic sheet resistance versus T^–1/4 plot for the pristine 12 nm La-doped SSO/12 nm undoped SSO/GSO (110) film, (d) temperature-dependent R_s of the same sample as a function of applied gate bias. Inset shows the R_s vs ln T plots for Vg ≥ 2 V. Dotted lines are a fit to the data. Arrows indicate the temperature at which the slope of the linear fits appears to deviate. Adapted from L. R. Thoutam et al., ACS Appl. Mater. Interfaces 11 (8), 7666 (2019).

Selected Reading:

The active sites of materials used in catalysis are sensitive to the density of charges on their surface. A material with a high dielectric constant (high-k) can accumulate a large surface charge density by increasing its ability to store charge (“capacitance”) or increasing the voltage threshold at which charge leakage is significant. Coupling materials for catalysis and high-k dielectrics in a heterostructure can enable the electric-field modulation of the charge density, the active sites, and the surface chemistry leading to higher reaction rates. 

The device structures have three key features: (1) high capacitance density (> 1.5 μF/cm²), (2) low leakage current density (< 10^-6 A/cm²) and (3) high charge density (> 2 × 10¹³ cm^-2).  We investigate thin films of perovskite oxide materials with high dielectric constants such as SrTiO₃ (k ~ 300) and BaTiO₃ (k ~ 600)  are ideal candidates for this application. Freestanding thin-film membranes based on these materials are also being explored for creating artificially-assembled heterostructures on desired substrates. We study the interplay of and trade-offs between different synthesis parameters on the capacitance, leakage current, and breakdown voltage to reach the threshold values for these parameters. 

Figure 11: Device heterostructure illustrating the concept of programmable catalysis using an oscillating voltage signal V_CAT. Figure inspired by and adapted from Center for Programmable Energy Catalysis, University of Minnesota. 

Selected Reading: