CaZrO3/KTaO3 (111) Interface: 2DEG & Rashba Effects

• The paper reveals that the CaZrO3/KTaO3 (111) interface supports a confined 2DEG exhibiting 2D superconductivity (Tc ≈ 0.43 K) and pronounced, anisotropic Rashba spin–orbit effects.
• Epitaxial growth via pulsed laser deposition and advanced structural/transport measurements confirm high-quality, coherent interfaces with minimal interdiffusion and controlled carrier accumulation.
• Optical tuning dramatically enhances nonreciprocal and nonlinear Hall effects by modulating Rashba coupling and carrier density, indicating promising routes for device applications.

Searching arXiv for the specified interface and closely related KTaO$_3$ (111) papers to ground the article in current literature. arXiv search query: "CaZrO3 KTaO3 (111) interface KTaO3 (111) 2DEG Rashba superconductivity" CaZrO$_3$/KTaO$_3$ (111) is an epitaxial oxide heterointerface in which a CaZrO$_3$ film is grown on a (111)-oriented KTaO$_3$ substrate and a two-dimensional electron gas (2DEG) forms on the KTaO$_3$ side. Reported CaZrO$_3$/KTaO$_3$ (111) heterostructures combine a confined interfacial metallic state, two-dimensional superconductivity, strong inversion asymmetry, substantial Rashba spin–orbit coupling, and pronounced nonlinear transport responses, including giant light-tunable nonreciprocal transport and a sign-reversible nonlinear Hall effect (Zhang et al., 2024, Zhang et al., 19 Jul 2025). Because the interface is buried, its low-energy electronic structure is usually interpreted by reference to angle-resolved photoemission and tight-binding studies of KTaO$_3$ (111) 2DEGs, which establish the Ta $5d$ $_3$0-derived subband structure, anisotropic Fermi surfaces, and nontrivial spin textures expected on the KTaO$_3$1 side (Mallik et al., 2023, Bruno et al., 2019, Al-Tawhid et al., 20 Feb 2025).

1. Heterostructure, crystallography, and interfacial charge accumulation

Reported CaZrO$_3$2/KTaO$_3$3 (111) devices are epitaxial heterostructures grown by pulsed laser deposition on single-crystal KTaO$_3$4 (111) substrates. One study used 50 nm CaZrO$_3$5 films grown at $_3$6 and $_3$7 mbar oxygen pressure with a 248 nm KrF excimer laser at a fluence of $_3$8 and 1 Hz (Zhang et al., 2024). Another used 10 nm CaZrO$_3$9 films on $_3$0 KTaO$_3$1 (111), grown at $_3$2 and $_3$3 Pa with a 248 nm KrF excimer laser at 2 Hz and a fluence of $_3$4 (Zhang et al., 19 Jul 2025). The reported lattice mismatch is small, $_3$5 or $_3$6, enabling coherent or high-quality epitaxy (Zhang et al., 2024, Zhang et al., 19 Jul 2025).

Structural characterization identifies atomically sharp and chemically sharp interfaces. In one report, $_3$7–2$_3$8 x-ray diffraction and STEM-HAADF established coherent epitaxy and a flat interface, while STEM-EDX found no discernible interdiffusion (Zhang et al., 2024). In another, AFM showed an atomically flat surface with RMS roughness $_3$9, consistent with layer-by-layer growth (Zhang et al., 19 Jul 2025).

The KTaO$_3$0 (111) side sets the interfacial symmetry. KTaO$_3$1 is cubic perovskite, but along (111) the stacking is polar and the surface or interface belongs to $_3$2. The (111) projection of the B-site network yields a triangular lattice of Ta atoms, while a KTaO$_3$3 (111) slab can also be described as a buckled honeycomb arrangement of Ta$_3$4 ions, with three successive Ta (111) planes forming a hexagonal arrangement with three inequivalent Ta sites (Mallik et al., 2023, Al-Tawhid et al., 20 Feb 2025, Zhang et al., 19 Jul 2025). This symmetry is fundamental for the trigonal Fermiology, anisotropic spin texture, and tensor form of nonlinear and spin-galvanic responses.

The microscopic origin of the 2DEG is attributed primarily to oxygen vacancies at or near the KTaO$_3$5 surface or interface, together with interfacial polarity and band alignment. One study states that the interfacial conduction is attributed primarily to oxygen vacancies at the KTaO$_3$6 surface, which provide in-gap donor states with two main distributions around 2.3 eV and 3.1 eV below the conduction band and electron-dope the Ta $_3$7 $_3$8 conduction band (Zhang et al., 2024). Another states that, in combination with polar discontinuity and oxygen-vacancy-related in-gap states, electrons are transferred or accumulated on the KTaO$_3$9 side because the Ta $_3$0 conduction-band minimum lies below the relevant higher-lying CaZrO$_3$1 states (Zhang et al., 19 Jul 2025). A common misconception is that the conducting channel is a CaZrO$_3$2 metallic layer; the reported interpretation instead places the low-energy carriers in a KTaO$_3$3-side interfacial quantum well.

2. KTaO$_3$4 (111) electronic structure as the reference normal state

Direct band mapping of buried CaZrO$_3$5/KTaO$_3$6 (111) interfaces is not available from standard VUV ARPES, so the reference normal-state electronic structure is taken from KTaO$_3$7 (111) 2DEGs studied at surfaces and ultrashallow interfaces (Mallik et al., 2023, Bruno et al., 2019). These studies establish that the relevant conduction states are Ta $_3$8 $_3$9-derived and, at experimentally relevant energies, are entirely derived from the bulk $_3$0 manifold, with the $_3$1 band lying about 400 meV higher in bulk KTaO$_3$2 (Bruno et al., 2019).

Quantum confinement along [111] produces multiple subbands. ARPES on superconducting KTaO$_3$3 (111) 2DEGs reveals multiple dispersive conduction bands crossing the Fermi level and a subband structure fitted by four spin-split band pairs within the measurement window, including three Ta $_3$4 $_3$5-derived pairs and an additional confined subband represented by a second copy of the $_3$6 manifold (Mallik et al., 2023). Earlier ARPES on surface-stabilized KTaO$_3$7 (111) 2DEGs resolved two main electron-like subbands with a bandwidth of $_3$8 meV and a third low-density band near $_3$9 (Bruno et al., 2019).

The Fermi surface is strongly noncircular. ARPES studies report a star-shaped outer contour centered at $_3$0 together with inner sheets that are hexagonal or more isotropic (Mallik et al., 2023, Bruno et al., 2019). In one superconducting KTaO$_3$1 (111) 2DEG, the outer star-shaped Fermi surface has a major diameter of $_3$2 along $_3$3-M and a minor diameter of $_3$4 along $_3$5-K; inner contours are described as star-like, nearly hexagonal, and nearly circular (Mallik et al., 2023). In an earlier surface 2DEG, the outer star extends to $_3$6 along $_3$7-M and the inner hexagon to $_3$8 (Bruno et al., 2019). Reported ARPES-derived 2D carrier densities for these KTaO$_3$9 (111) 2DEGs are $_3$0 and $_3$1 (Bruno et al., 2019, Mallik et al., 2023).

Orbital character is strongly hybridized. Unlike KTaO$_3$2 (001), where $_3$3 and $_3$4 can often be separated, the KTaO$_3$5 (111) 2DEG does not exhibit a simple pure orbital character for individual bands; instead, the $_3$6 orbitals are strongly mixed, although large-$_3$7 segments of the star-like contour can become locally dominated by $_3$8, $_3$9, or $_3$0 character (Mallik et al., 2023, Bruno et al., 2019). This suggests that CaZrO$_3$1/KTaO$_3$2 (111) should be treated as a multi-orbital, symmetry-constrained quantum well rather than as a single-band Rashba electron gas.

3. Interfacial transport and two-dimensional superconductivity

CaZrO$_3$3/KTaO$_3$4 (111) interfaces are metallic at low temperature and support a confined conduction channel. In a 50 nm-film study, the sheet carrier density is $_3$5 at 300 K and decreases to $_3$6 at low temperature because of carrier freeze-out from in-gap states; the mobility is $_3$7 at 2 K and $_3$8 at 250 K (Zhang et al., 2024). In a 10 nm-film study, the interface remains metallic from 2 to 300 K with $_3$9, $5d$0, and $5d$1 at 2 K (Zhang et al., 19 Jul 2025).

The conduction is two-dimensional. Angular magnetoresistance was used to estimate a conduction thickness of $5d$2 nm via

$5d$3

and the resulting metallic channel was described as confined to the interface region (Zhang et al., 2024). In the dark, Hall transport is linear and negative, consistent with a dominant electron-like band (Zhang et al., 2024).

A superconducting transition is observed at low temperature. For the 50 nm CaZrO$5d$4/KTaO$5d$5 (111) interface, the sheet resistance drops to zero with a critical temperature $5d$6 K, defined at 50% of the normal-state resistance (Zhang et al., 2024). The superconducting state is described as two-dimensional and localized at the interface, comparable to epitaxial LaVO$5d$7/KTaO$5d$8 (111) and lower than some non-epitaxial KTaO$5d$9 interfaces (Zhang et al., 2024). Within the broader KTaO$_3$00 literature, the (111) orientation is singled out as hosting the highest $_3$01 among KTaO$_3$02 orientations (Zhang et al., 2024).

Reference superconducting KTaO$_3$03 (111) 2DEGs also display strong in-plane anisotropy. In Eu-oxidized KTaO$_3$04 (111), $_3$05 K for current along $_3$06 and $_3$07 K for current along $_3$08, while the normal-state mobilities are $_3$09 and $_3$10 (Mallik et al., 2023). This suggests that the anisotropic Fermi surface and spin texture established for KTaO$_3$11 (111) are relevant to superconductivity at CaZrO$_3$12/KTaO$_3$13 (111) as well.

4. Rashba spin–orbit coupling, spin texture, and spin-to-charge conversion

Broken inversion symmetry along the interface normal, combined with strong Ta $_3$14 spin–orbit coupling, gives rise to Rashba physics. A standard phenomenological form used for KTaO$_3$15 (111) interfaces is

$_3$16

or equivalently

$_3$17

but band-structure studies show that the actual splitting is multi-orbital, momentum-dependent, and strongly anisotropic (Al-Tawhid et al., 20 Feb 2025, Zhang et al., 2024, Mallik et al., 2023).

Tight-binding fits to superconducting KTaO$_3$18 (111) 2DEGs use a Ta $_3$19 basis on a 2D triangular lattice with $_3$20, $_3$21 eV, $_3$22 eV, onsite energies $_3$23 and $_3$24 eV for two effective subband copies, atomic spin–orbit coupling $_3$25 eV, and inversion-breaking orbital mixing $_3$26 and $_3$27 meV (Mallik et al., 2023). In these fits, the Rashba-like splitting emerges from the combination of $_3$28, on-site $_3$29, and inversion-breaking $_3$30, rather than from a single explicit Rashba term (Mallik et al., 2023).

Band-resolved Rashba parameters near $_3$31 are reported as $_3$32 meV$_3$33\AA\ for the outer pink band pair and $_3$34 meV$_3$35\AA\ for the orange pair, while the green and cyan pairs are anisotropic with $_3$36 along $_3$37-M and $_3$38 and $_3$39 meV$_3$40\AA\ along $_3$41-K, respectively (Mallik et al., 2023). Earlier ARPES-based analysis defined an effective $_3$42 and reported $_3$43 and $_3$44, with corresponding momentum splittings of $_3$45 and $_3$46 (Bruno et al., 2019). These values show that no single isotropic $_3$47 characterizes KTaO$_3$48 (111).

The spin texture is equally nontrivial. Calculations for KTaO$_3$49 (111) find that spin-momentum locking holds only on high-symmetry directions, while a strong out-of-plane spin component renders the spin texture threefold symmetric (Bruno et al., 2019). In the superconducting KTaO$_3$50 (111) 2DEG, the outer star-shaped contour exhibits helical-like spin winding with strong warping; spins are purely in-plane at the star tips along $_3$51-M, but acquire large out-of-plane components between tips, and inner sheets can show both orthoradial and radial orientations with local helicity reversals near band intersections (Mallik et al., 2023).

Spin-to-charge conversion has been measured directly in AlO$_3$52/KTaO$_3$53 (111) 2DESs by spin pumping. There, the Hall sheet density is $_3$54 at 300 K and $_3$55 at 3–4 K, the mobility at 3 K is $_3$56, and the mean free time is $_3$57 ps (Al-Tawhid et al., 20 Feb 2025). The inverse Rashba–Edelstein length is defined by

$_3$58

and reaches $_3$59 nm at 10 K. Using

$_3$60

an effective $_3$61 was inferred (Al-Tawhid et al., 20 Feb 2025). The angle dependence deviates from simple $_3$62 behavior and is fitted by

$_3$63

with $_3$64 at 10 K and $_3$65 at 70 K (Al-Tawhid et al., 20 Feb 2025). An important clarification is that stronger atomic SOC does not automatically imply stronger spin-to-charge conversion: the KTaO$_3$66 analysis attributes the modest IREE partly to multi-orbital compensation and to $_3$67 states near $_3$68 with nearly vanishing magnetic moment (Al-Tawhid et al., 20 Feb 2025).

5. Light-tunable nonreciprocal transport

A defining result for CaZrO$_3$69/KTaO$_3$70 (111) is the giant optical enhancement of nonreciprocal transport in the normal state (Zhang et al., 2024). Nonreciprocity is described phenomenologically by

$_3$71

with the experimentally extracted coefficient

$_3$72

in the linear regime $_3$73 T (Zhang et al., 2024).

In the dark at 5 K and $_3$74A, $_3$75, already comparable with LaAlO$_3$76/SrTiO$_3$77 and larger than many Rashba semiconductors (Zhang et al., 2024). Under 330 nm ultraviolet illumination, $_3$78 reaches $_3$79 at several tesla and $_3$80, representing a three-orders-of-magnitude enhancement (Zhang et al., 2024). The signal decreases with increasing temperature and disappears around 40 K (Zhang et al., 2024).

The optical response is wavelength dependent. Around 450–360 nm, photons excite electrons from oxygen-vacancy in-gap states into the conduction band; at 330 nm, slightly above the KTaO$_3$81 band gap of $_3$82 eV, photoconductance is maximal because abundant electrons are pumped from the valence-band maximum into the Ta $_3$83 conduction band and into the interfacial well (Zhang et al., 2024). At 300 nm, the resistance increases again because photoexcited electrons occupy higher subbands, electron–phonon scattering and recombination increase, and weak localization competes with weak antilocalization (Zhang et al., 2024).

Hall transport under illumination becomes nonlinear and is analyzed with a two-band model. For $_3$84 nm, only intrinsic carriers are reported, with $_3$85 and $_3$86. At $_3$87 nm, a second photocarrier channel appears and a Lifshitz transition occurs at $_3$88 nm (Zhang et al., 2024). At 330 nm, $_3$89 jumps to $_3$90, $_3$91, and $_3$92 (Zhang et al., 2024).

Weak antilocalization analysis by modified Maekawa–Fukuyama theory yields a peak spin–orbit field

$_3$93

at 330 nm, together with a maximum Rashba coefficient

$_3$94

and a spin splitting energy

$_3$95

using $_3$96 (Zhang et al., 2024). The reported microscopic scaling is

$_3$97

and the interpretation is that UV illumination increases both $_3$98 and $_3$99 by about an order of magnitude, so the product $_3$00 dominates over the increase in $_3$01 and drives the giant enhancement of $_3$02 (Zhang et al., 2024).

6. Nonlinear Hall effect, skew scattering, and Berry-curvature engineering

The CaZrO$_3$03/KTaO$_3$04 (111) interface also exhibits a giant, light-tunable nonlinear Hall effect (NLHE) at zero magnetic field (Zhang et al., 19 Jul 2025). In a six-terminal Hall-bar geometry, an AC drive current $_3$05 generates a second-harmonic transverse voltage $_3$06, from which the second-order transverse conductivity is extracted as

$_3$07

For current mainly along $_3$08, the dark signal at 10 K is positive, with $_3$09 at $_3$10; for current along $_3$11, the NLHE is much weaker (Zhang et al., 19 Jul 2025).

The key symmetry point is that the standard Berry-curvature-dipole mechanism is forbidden. Because the nonmagnetic KTaO$_3$12 (111) interface retains in-plane threefold rotation $_3$13, the Berry-curvature dipole vanishes by symmetry, even though inversion symmetry is broken and Berry curvature itself is nonzero (Zhang et al., 19 Jul 2025). The measured NLHE is instead assigned to extrinsic skew scattering, with side jump providing a subdominant contribution (Zhang et al., 19 Jul 2025).

Optical excitation acts as an efficient gate. At $_3$14 nm and $_3$15 mW, the sign of $_3$16 reverses; at $_3$17 mW, $_3$18 at $_3$19; at $_3$20 mW and $_3$21, $_3$22, compared with $_3$23 in the dark (Zhang et al., 19 Jul 2025). The corresponding $_3$24 increases from $_3$25 in the dark to $_3$26 under strong illumination, nearly five orders of magnitude larger (Zhang et al., 19 Jul 2025).

The transport scaling analysis uses

$_3$27

Here the $_3$28 term is attributed to skew scattering and the $_3$29 term to side jump, since the Berry-curvature-dipole contribution is symmetry-forbidden (Zhang et al., 19 Jul 2025). In the low-$_3$30 regime, $_3$31 and $_3$32; in the high-$_3$33 regime, $_3$34 and $_3$35 (Zhang et al., 19 Jul 2025). The sign reversal is therefore driven by the skew-scattering term.

First-principles calculations connect this sign change to the Berry curvature triple

$_3$36

the symmetry-allowed third angular moment of Berry curvature on the Fermi contour (Zhang et al., 19 Jul 2025). For a 6-unit-cell KTaO$_3$37 (111) slab, $_3$38 is positive for $_3$39 eV, crosses zero at $_3$40 eV, and becomes negative above that value (Zhang et al., 19 Jul 2025). The experimental interpretation is that light raises the Fermi level toward a band crossing near M, changes the sign of the Berry curvature triple, and simultaneously introduces a high-mobility carrier channel with $_3$41 increasing from 2403 to 9060 $_3$42 as power rises from 0.1 to 15 mW (Zhang et al., 19 Jul 2025).

7. Modeling strategies, experimental constraints, and broader significance

Two complementary modeling frameworks are used for KTaO$_3$43 (111) and provide the natural starting point for CaZrO$_3$44/KTaO$_3$45 (111). One is a self-consistent tight-binding supercell built from relativistic DFT and Ta $_3$46 Wannier functions, with a 30-unit-cell slab along [111], an on-site confinement potential $_3$47, and self-consistent Poisson–Schrödinger solution with a field-dependent dielectric constant (Bruno et al., 2019). The other is a reduced 2D triangular-lattice tight-binding model with explicit $_3$48, $_3$49, and inversion-breaking $_3$50 terms and parameters fitted to ARPES (Mallik et al., 2023). Both encode the essential point that Rashba-like splitting and spin texture emerge from confinement, inversion breaking, and multi-orbital Ta $_3$51 physics rather than from a minimal single-band model.

For CaZrO$_3$52/KTaO$_3$53 (111), CaZrO$_3$54 can be treated as a wide-gap insulating overlayer that sets electrostatic boundary conditions, potential drop, strain, and orbital mixing on the KTaO$_3$55 side (Mallik et al., 2023). This suggests that interface-specific modeling should modify the confinement profile, subband occupancy, and inversion-breaking parameter rather than replace the KTaO$_3$56-side $_3$57 framework. Strain may be incorporated as anisotropic changes in hopping or as crystal-field terms that distinguish the three $_3$58 orbitals (Mallik et al., 2023).

A major experimental limitation is spectroscopic accessibility. Standard VUV ARPES probes only the top $_3$59–2 nm and is therefore suitable for surface 2DEGs or ultrashallow buried interfaces, but not for a buried CaZrO$_3$60/KTaO$_3$61 channel. Soft x-ray ARPES can probe deeper, though with limited energy and momentum resolution near $_3$62 (Mallik et al., 2023). As a result, buried CaZrO$_3$63/KTaO$_3$64 interfaces are presently constrained primarily by transport, weak-antilocalization analysis, nonlinear response measurements, and theory guided by KTaO$_3$65 (111) electronic-structure benchmarks (Mallik et al., 2023, Zhang et al., 2024, Zhang et al., 19 Jul 2025).

Several conceptual clarifications follow from the current literature. First, the KTaO$_3$66 (111) interface is not well described by an isotropic circular Rashba 2DEG; the actual Fermi surface is warped and multi-sheeted, and the spin texture includes substantial out-of-plane canting (Bruno et al., 2019, Mallik et al., 2023). Second, strong atomic SOC does not guarantee maximal spin-galvanic efficiency because spin and orbital channels can partially cancel (Al-Tawhid et al., 20 Feb 2025). Third, in the NLHE of CaZrO$_3$67/KTaO$_3$68 (111), the canonical Berry-curvature-dipole mechanism is symmetry-forbidden, and the experimentally relevant quantity is the Berry curvature triple controlling skew scattering (Zhang et al., 19 Jul 2025).

Taken together, the available work identifies CaZrO$_3$69/KTaO$_3$70 (111) as a noncentrosymmetric, multi-orbital, Ta $_3$71-electron interface in which the polar (111) geometry, strong spin–orbit coupling, and optically tunable band filling produce an unusual coexistence of two-dimensional superconductivity, anisotropic Rashba physics, giant nonreciprocal transport, and a sign-reversible nonlinear Hall effect (Zhang et al., 2024, Zhang et al., 19 Jul 2025).