Inhomogeneous superconductivity in (001), (110) and (111) KTaO$_3$ two-dimensional electronic gas: $T_c$ driven from electronic confinement

Using a self-consistent tight-binding slab model, this study reveals that the orientation-dependent superconducting critical temperature in KTaO$_3$-based two-dimensional electron gases is primarily driven by variations in the spatial confinement of the electron gas and the resulting redistribution of the density of states, rather than changes in the pairing interaction.

The Gist

Imagine you have a giant, perfect cube of a special crystal called KTaO3 (Potassium Tantalate). This crystal is like a high-tech playground for electrons. When you slice this cube open and create a surface, electrons get trapped right at the edge, forming a thin, two-dimensional "sea" of electricity. This is called a 2D Electron Gas (2DEG).

Now, here's the magic trick: under the right conditions, these electrons can start dancing in perfect unison, a phenomenon called superconductivity (where electricity flows with zero resistance).

But here's the puzzle: Scientists found that if you slice the crystal in different directions, the "dance" changes completely.

• Slice it one way (001): The electrons barely dance at all.
• Slice it another way (110): They dance a little better.
• Slice it a third way (111): They throw the wildest, most energetic party, becoming superconducting at much higher temperatures.

The big question was: Why? Is it because the electrons are "friendlier" to each other in the (111) direction? Or is it just about how the room is arranged?

This paper answers that question with a simple, elegant explanation: It's all about the room layout, not the dancers.

The Analogy: The Dance Floor and the Crowd

Think of the electrons as dancers and the crystal surface as a dance floor.

1. The "Room" (The Crystal Orientation)
When you cut the crystal, you change the shape of the dance floor.

• The (001) Floor: This is like a small, cramped closet. The dancers are squeezed tightly against the wall. They can't spread out.
• The (111) Floor: This is like a massive, open ballroom with high ceilings. The dancers have plenty of room to spread out and move deeper into the room.

2. The "Music" (The Pairing Interaction)
The "music" that makes the electrons pair up and dance together is the same in all three rooms. The paper assumes the "attraction" between electrons is identical regardless of the direction. It's the same song playing in the closet, the hallway, and the ballroom.

3. The "Crowd Density" (The Density of States)
This is the key. In the cramped closet (001), the dancers are so squished that they can't find enough partners to pair up efficiently. In the open ballroom (111), the dancers are spread out over a larger area. This allows for a much better "redistribution" of the crowd.

What the Scientists Did

The authors built a virtual simulation (a digital twin) of these three different crystal slices. They didn't just guess; they calculated exactly how the electrons behave in each "room."

They used a model that accounts for:

• The Walls: How the electric field pushes the electrons against the surface.
• The Spin: How the electrons' internal "spins" interact with their movement (a quantum effect called Rashba coupling).
• The Dance: How they pair up to become superconductors.

The Big Discovery

They found that the superconducting temperature ($T_c$)—the point where the electricity stops having resistance—depends entirely on how far the electrons can spread out.

• In the (111) direction: The electrons spread out over about 10 layers of the crystal. Because they have more space, they can access more "energy states" (more spots on the dance floor) right at the edge where the magic happens. This makes it much easier for them to pair up, leading to a higher critical temperature.
• In the (001) direction: The electrons are stuck in just 2 layers. They are too cramped to form a strong superconducting state, so the temperature required to make them superconduct is very low (or they don't do it at all under normal conditions).

The "Aha!" Moment:
The paper proves that you don't need to change the "music" (the pairing force) to explain why the (111) crystal is a better superconductor. You just need to realize that the (111) crystal offers a larger, more spacious dance floor. The electrons are the same, the rules are the same, but the geometry of the space allows them to succeed where the others fail.

Why This Matters

This is a huge deal for future technology.

• Designing Better Chips: If we want to build superconducting computers or quantum devices, we shouldn't just look for new materials. We should look at how we cut the materials we already have.
• Geometry is King: It turns out that the shape of the interface (the cut) is just as important as the material itself. By choosing the right angle (like the (111) cut), we can unlock superconductivity without needing extreme cold or exotic chemicals.

In a nutshell: The paper tells us that for these special crystals, superconductivity isn't about who the electrons are, but where they are allowed to stand. The (111) direction gives them the best standing room, making them the best dancers in the quantum ballroom.

Technical Summary

1. Problem Statement

Two-dimensional electron gases (2DEGs) formed at KTaO$_3$ (KTO) interfaces exhibit superconductivity with a critical temperature ($T_c$) that depends strongly on the crystallographic orientation of the interface ((001), (110), and (111)). Experimental data shows a hierarchy where $T_c^{(111)} > T_c^{(110)} > T_c^{(001)}$, with the (001) orientation often failing to show superconductivity under similar conditions.

The central theoretical challenge addressed in this work is determining the microscopic origin of this anisotropy. Specifically, the authors investigate whether the hierarchy in $T_c$ arises from:

1. Orientation-dependent pairing interactions (e.g., different phonon spectra or coupling strengths for different surfaces).
2. Geometric confinement and electronic structure (e.g., differences in the spatial extent of the 2DEG, orbital composition, and density of states) while keeping the microscopic pairing interaction constant.

2. Methodology

The authors employ a unified microscopic framework based on a self-consistent tight-binding slab model to isolate the effects of geometry and electronic structure.

• Model Hamiltonian:
  • Bulk Description: A tight-binding model describing the $t_{2g}$ orbitals ($d_{yz}, d_{zx}, d_{xy}$) of Ta 5d electrons. It includes nearest-neighbor ($t_D$) and next-to-nearest-neighbor ($t_I$) hopping amplitudes.
  • Spin-Orbit Coupling (SOC): Atomic SOC ($\lambda \approx 0.27$ eV) is included.
  • Confinement Potential: A self-consistent electrostatic potential $\phi(z)$ is calculated to simulate the interface. This breaks inversion symmetry, inducing orbital Rashba couplings.
  • Crystal Field Distortions: Specific distortions are applied for (001)/(110) (tetragonal) and (111) (trigonal) symmetries to lift orbital degeneracies.
• Self-Consistency: The model iteratively solves the Schrödinger equation for the electronic wavefunctions and the Poisson equation for the electrostatic potential. A non-linear dielectric function $\epsilon(F)$, fitted to experimental ARPES data, is used to screen the electric field.
• Superconductivity Treatment:
  • A local spin-singlet s-wave pairing interaction ($U$) is introduced within a mean-field approximation.
  • Crucially, the interaction strength $U$ is kept identical for all three orientations to test if geometry alone drives the $T_c$ differences.
  • Inhomogeneous Order Parameter: The superconducting gap $\Delta_z$ is allowed to vary spatially across the slab layers (transverse direction $z$) using a variational ansatz $\Delta_z = \Delta e^{-z/z_0}$, where $z_0$ is the penetration depth.
• Pairing Scenarios: Two limiting cases are analyzed to provide bounds for $T_c$:
  1. Phonon-mediated (Retarded): Pairing occurs only within a narrow energy shell ($\hbar\omega_i$) around the Fermi level.
  2. Instantaneous (Non-retarded): Pairing occurs over the entire energy bandwidth (upper bound).

3. Key Results

A. Electrostatic and Electronic Structure Differences

• Electric Field: The (111) interface exhibits the strongest interfacial electric field (up to 4x higher than (001) at high carrier densities), despite having a more spatially extended 2DEG. This is due to dominant out-of-plane hopping in (111) versus in-plane dispersion in (001)/(110).
• Spatial Confinement:
  • (001): The 2DEG is highly confined, spanning only $\sim 2$ layers.
  • (110): Intermediate confinement ($\sim 5$ layers).
  • (111): The 2DEG is the most spatially extended ($\sim 10$ layers), with charge density peaking away from the first layer.
• Density of States (DOS): The Local Density of States (LDOS) at the Fermi level reflects these confinement regimes. The (111) orientation shows a broader distribution of spectral weight deeper into the slab, while (001) is sharply localized.

B. Superconducting Instability and $T_c$ Hierarchy

• Consistent Hierarchy: In both the phonon-mediated and instantaneous pairing scenarios, the calculated critical temperatures follow the experimental hierarchy:
  $$T_c^{(111)} > T_c^{(110)} > T_c^{(001)}$$
• Role of Spatial Extension: The (111) orientation yields the highest $T_c$ because the spatially extended 2DEG allows for a larger weighted density of states entering the pairing instability condition. The inhomogeneous order parameter ($z_0$) optimizes to match the spatial extent of the electronic states.
• Interaction Strength:
  • For the (001) orientation, a threshold-like behavior is observed in the instantaneous model, requiring a significantly higher interaction strength $U$ to induce superconductivity compared to (111).
  • The (001) interface often fails to show superconductivity within experimentally accessible $T_c$ ranges unless $U$ is unrealistically high.
• Penetration Depth ($z_0$): The optimal penetration depth of the superconducting order parameter correlates with the spatial extent of the 2DEG. For (111), $z_0$ is large, allowing the condensate to utilize bulk-like states, whereas for (001), $z_0$ is small and constrained by the tight confinement.

4. Key Contributions

1. Microscopic Explanation of Anisotropy: The paper demonstrates that the hierarchy of $T_c$ in KTO interfaces does not require orientation-dependent pairing interactions. Instead, it is driven purely by geometric confinement and the resulting redistribution of the density of states.
2. Inhomogeneous Mean-Field Theory: The authors rigorously treat the superconducting order parameter as a function of the transverse coordinate ($z$). They show that assuming a homogeneous order parameter (constant across the slab) fails to reproduce the experimental hierarchy, highlighting the necessity of accounting for spatial inhomogeneity in oxide 2DEGs.
3. Unified Framework: By using identical bulk parameters and dielectric functions for all orientations, the study isolates the geometric and symmetry-breaking effects (Rashba coupling, crystal field) as the primary drivers of the observed physics.
4. Orbital Rashba Couplings: The work provides a detailed derivation of orbital Rashba terms for (001), (110), and (111) orientations, including next-to-nearest-neighbor (NNN) processes, which are crucial for accurate modeling of the multi-layer slab.

5. Significance

• Resolving Experimental Discrepancies: The findings provide a coherent theoretical explanation for why (111) KTO interfaces are "superconducting hotspots" compared to (001) and (110), resolving previous ambiguities about whether the effect was due to material quality or intrinsic electronic structure.
• Design Principles for Oxide Electronics: The study establishes that spatial extension and orbital engineering are critical levers for enhancing superconductivity in oxide heterostructures. It suggests that maximizing the spatial extent of the 2DEG (e.g., via specific crystal orientations) can boost $T_c$ without needing to increase the pairing interaction strength.
• General Applicability: The framework developed here is not limited to KTO; it offers a general approach for understanding inhomogeneous superconductivity in other oxide interfaces where confinement and multiorbital physics play a dominant role.

In conclusion, the paper argues that the "geometric confinement" of the 2DEG dictates the spatial profile of the superconducting condensate and the effective density of states, thereby driving the orientation-dependent critical temperatures observed in KTaO$_3$ interfaces.