Layerwise stratification and band reordering in twisted multilayer MoTe₂.

Area of Science:

• Materials Science
• Condensed Matter Physics
• Computational Physics

Background:

• Multilayer moiré systems exhibit complex behaviors due to interlayer interactions.
• Accurate simulation of these systems is computationally challenging.
• Understanding structural and electronic properties is key to novel device applications.

Purpose of the Study:

• To develop an efficient, physics-informed strategy for generating training data for machine learning force fields in multilayer moiré systems.
• To enable machine learning force fields transferable across layer numbers and stacking configurations.
• To investigate the structural and electronic properties of multilayer twisted MoTe2 (tMoTe2).

Main Methods:

• Introduced a physics-informed training-data generation strategy.
• Developed a machine-learning force field transferable across layer numbers and stacking configurations.
• Applied the method to multilayer twisted MoTe2 (tMoTe2) to study interlayer interactions and stratification.

Main Results:

• Identified structural and electronic stratification in tMoTe2, with interface layers showing reconstruction and outer layers exhibiting attenuated distortions.
• Discovered that stratification is strongest at intermediate twist angles (2-5°), not ultrasmall angles.
• Observed suppression of interlayer hybridization across the moiré interface-bulk boundary, leading to electronic isolation.
• Found coexisting honeycomb and triangular lattice motifs in frontier valence bands of twisted double bilayer MoTe2.
• Demonstrated that twist angle and gating can induce Chern band reordering and nonlinear electric polarization.

Conclusions:

• The developed approach enables efficient simulation of multilayer moiré systems.
• Revealed novel structural-electronic separation phenomena absent in bilayer systems.
• The findings pave the way for designing materials with tailored electronic properties using moiré engineering.