Twistronic control of shift current in multilayer moiré system

This paper demonstrates that twisting multilayer H-MoS2 moiré systems breaks crystal symmetry to activate new tunable in-plane shift current components, utilizing a validated Slater-Koster tight-binding model to establish twist angle as an effective parameter for engineering nonlinear optical responses in large-scale moiré materials.

The Gist

Imagine you have a solar panel. Usually, to turn sunlight into electricity, you need a specific internal "traffic cop" (called a p-n junction) to push electrons in one direction. But there's a different, more exotic way to do this called the Bulk Photovoltaic Effect. Think of it like a crowd of people (electrons) who, when hit by a flash of light, instinctively jump a few steps to the left or right all at once, creating a flow without needing a traffic cop. This specific "jump" is called the Shift Current.

This paper is about learning how to control that jump in a special family of materials called MoS2 (Molybdenum Disulfide), which are made of atom-thin sheets.

Here is the breakdown of their discovery, using simple analogies:

1. The Problem: The "Pixel" Limit

To study how these materials behave, scientists usually use super-powerful computer simulations (like a high-resolution camera) called Density Functional Theory (DFT). However, when you twist two sheets of MoS2 on top of each other, they create a giant, complex pattern called a Moiré pattern (like when you hold two window screens slightly out of alignment).

If you try to simulate this twisted pattern with the "high-resolution camera," the computer crashes because there are too many atoms to count. It's like trying to count every single grain of sand on a beach with a magnifying glass; it takes too long.

2. The Solution: The "Sketch" Model

The authors built a smarter, faster tool. Instead of counting every grain of sand, they created a tight-binding model (think of it as a detailed sketch or a map). They first used the high-resolution camera to learn the rules of the game for a single sheet, and then they taught their sketch how to mimic those rules.

They tested this sketch against the real data for 1, 2, and 3 layers of MoS2, and it worked perfectly. It was fast, accurate, and could handle the giant twisted patterns that the heavy computer simulations couldn't.

3. The Discovery: The "Twist" Switch

The most exciting part of the paper is what happens when you twist the layers.

• Untwisted (Stacked neatly): Imagine a stack of pancakes perfectly aligned. The electrons can only jump in one specific direction (let's say, North). The "South" direction is locked off.
• Twisted (Rotated slightly): Now, imagine rotating the top pancake slightly. This breaks the perfect symmetry. Suddenly, the "South" direction unlocks! The electrons can now jump in new directions that were previously forbidden.

The authors found that by changing the twist angle (how much you rotate the layers), you can tune the electricity.

• At a small twist, the new direction is weak.
• At a larger twist, the new direction becomes just as strong as the original one.

It's like having a dimmer switch for the direction of electricity. You can rotate the layers to decide exactly where the current flows.

4. The "Jump" Distance

The paper also looked at how far the electrons jump. They found that twisting the layers changes the "dance steps" of the electrons. In the twisted versions, the electrons physically shift their position in space by a larger amount (about 1 Angstrom, which is tiny but significant at the atomic scale) compared to the untwisted versions. This bigger jump is what creates the stronger, more controllable electricity.

5. The Multi-Layer Experiment

They also tested what happens with three layers:

• Twisting just the middle layer: This creates a strong current in two different directions.
• Twisting the top and bottom layers in opposite directions: This acts like a cancel-out mechanism. The middle layer stays straight, and the "twist" effects cancel each other out, keeping the current flowing in just the original direction.

The Bottom Line

This paper proves that you don't need a traffic cop to generate solar power in these materials. Instead, you can use twistronics (twisting the layers) as a dial to control how much electricity is made and which way it flows.

They showed that their "sketch" method is a reliable way to design these materials without needing supercomputers that would take years to run. This opens the door to building future solar devices where you can simply twist the layers to get the exact performance you need.

Technical Summary

Technical Summary: Twistronic Control of Shift Current in Multilayer Moiré Systems

Problem Statement
The conversion of light into electrical current in non-centrosymmetric materials via the bulk photovoltaic effect (BPVE) offers an alternative to traditional p-n junctions, potentially overcoming the Shockley–Queisser efficiency limit. Among BPVE mechanisms, the shift current (SC) is particularly significant due to its dependence on the geometric properties of electronic wavefunctions. While two-dimensional (2D) materials like transition-metal dichalcogenides (TMDs) are promising candidates for such applications, accurately modeling their nonlinear optical responses in twisted multilayer configurations presents a computational challenge. First-principles density functional theory (DFT) calculations become prohibitively expensive for large moiré supercells, which can contain thousands of atoms even at relatively small twist angles. Consequently, there is a need for efficient computational frameworks capable of capturing the essential electronic and geometric features of these large-scale systems to explore twist-induced control over shift currents.

Methodology
The authors developed and validated a Slater–Koster (SK) tight-binding (TB) model parametrized from first-principles DFT calculations to investigate H-MoS₂ systems.

• DFT Benchmarking: DFT calculations were performed for mono-, bi-, and trilayer H-MoS₂ using the Quantum ESPRESSO package. Maximally localized Wannier functions were constructed using Mo 4d and S 3p orbitals to facilitate the fitting of TB parameters.
• Tight-Binding Model: A TB Hamiltonian was constructed using the TBPLaS package. The model includes intralayer hopping (Mo–Mo, Mo–S, S–S) and interlayer interactions. For twisted systems, the interlayer SK parameters were scaled according to the exponential distance dependence relative to the AA-stacked configuration, utilizing a smooth cutoff function to account for the varying interatomic distances in the moiré lattice.
• Shift Current Calculation: The shift current conductivity tensors were calculated using the WannierBerri package for the TB model and postW90 for DFT data. The calculations utilized a 150 × 150 × 1 k-point grid. Spin-orbit coupling was neglected, as previous studies indicated its negligible effect on the MoS₂ shift current profile.
• System Scope: The study covered non-twisted mono-, bi-, and trilayers, as well as commensurate twisted bilayers and trilayers with specific twist angles (e.g., 21.787°, 13.174°, 9.430°, 7.341°) derived from integer pairs defining the moiré lattice vectors.

Key Contributions and Results

1. Model Validation: The SK-TB model was shown to accurately reproduce the electronic band structure near the band edges and the main spectral features of the shift current conductivity for mono-, bi-, and trilayer H-MoS₂ when compared against DFT and Wannier-interpolated results. The model successfully captured the magnitude and spectral position of the main peaks in the solar energy spectrum range without requiring additional fitting for multilayer systems.
2. Layer-Dependent Enhancement: In untwisted multilayer systems, the magnitude of the shift current was found to be approximately proportional to the number of layers. This suggests that interlayer coupling modifies the positions of charge centers, leading to an overall enhancement of the nonlinear response.
3. Symmetry Breaking and Tensor Activation: The introduction of a twist angle breaks the crystal symmetry of the AA-stacked configuration. In untwisted bilayers, specific conductivity tensor components (e.g., $\sigma_{yyy}^{sc}$) are zero due to symmetry. The study demonstrated that twisting activates these previously forbidden components. For a twist angle of 21.787°, the $\sigma_{yyy}^{sc}$ component becomes non-zero and increases in magnitude with the twist angle, eventually reaching the same order of magnitude as the dominant $\sigma_{xxx}^{sc}$ component.
4. Twist Angle Tunability: The analysis of the shift distance (a real-space measure of charge center displacement) revealed a direct connection between the twist-induced modification of electronic wavefunctions and the increase in nonlinear response. The shift distance along the $y$-direction, which is zero in the untwisted case, increases with the twist angle, reaching approximately 1 Å for the largest angle considered.
5. Trilayer Configurations: The study extended the analysis to twisted trilayers. A configuration with only the middle layer twisted showed enhanced anisotropy with increased contributions to both $x$ and $y$ components. Conversely, a configuration with top and bottom layers rotated by equal and opposite angles ($\pm\theta$) preserved the in-plane symmetry of the untwisted middle layer, resulting in only the original non-zero tensor components.

Significance and Claims
The paper establishes the twist angle as an effective parameter for engineering shift current generation in multilayer TMD-based systems. The primary significance lies in demonstrating that tight-binding approaches, when properly parametrized, provide a practical and computationally efficient route for exploring nonlinear optical phenomena in large-scale moiré materials, bypassing the limitations of conventional first-principles calculations.

The authors claim that their results establish moiré systems as a promising strategy for controlling nonlinear photovoltaic phenomena in 2D materials. Specifically, the ability to activate and tune in-plane photocurrent components through relative layer rotation offers a new degree of freedom for designing tunable nonlinear responses. The work calls for dedicated experimental verification of these predictions and suggests that the introduced methodology can be applied to larger moiré superlattices of twisted van der Waals materials for bulk-photovoltaic applications. Future work is noted to potentially include heterobilayers and correlated moiré systems.