WorksPhysics and Astronomy
Scalable and deterministic construction of moiré superlattice in 2D materials using stressor films
Abstract
Moiré superlattice in two-dimensional (2D) materials provides a powerful platform to engineer emergent electronic states, yet the construction of moiré superlattices remains lab-scale, involving much trial and error and with little control. Here, we demonstrate the construction of a heterostrain-induced moiré superlattice in transition metal dichalcogenides using a scalable process that deterministically induces strain to 2D materials. By applying patterned thin-film stressors and probing the resulting structures with scanning transmission electron microscopy, we directly resolve the induced heterostrain, lattice deformations, and stacking variations that produce the moiré superlattice. We find that uniaxial and biaxial heterostrain give rise to distinct moiré patterns, including stripes and distorted hexagonal patterns. With this approach, we create in-plane polar distortions and thus in-plane polarization at the domain boundaries of the moiré superlattice in MoS. The deterministic and scalable construction of moiré patterns using a well-established scalable process opens opportunities to design new moiré geometries in 2D materials.
Corresponding Author: [email protected]
Moiré superlattices form when layered materials are stacked with a small relative twist or lattice mismatch, producing periodic modulations on a length scale much larger than the atomic lattice [1, 2, 3]. They have led to many ground-breaking phenomena including superconductivity [4, 5], topological edge states [6, 7], exciton trapping [8], correlated insulating phases [9, 10], and spontaneous polarization [11, 12, 13]. At the microscopic level, local stacking orders in the moiré superlattice control the interlayer coupling strength, intralayer strain and lattice reconstruction, and form moiré domains whose boundaries play a central role in determining the electronic ground states [14, 15, 16]. Thus, controlling the layer stacking and strain has become a key strategy to tailor the properties of moiré superlattices.
To date, most studies of moiré systems have focused on twist-controlled bilayers such as twisted bilayer graphene [4, 5, 9, 17] and twisted transition metal dichalcogenides (TMDs) such as bilayer [18, 19], [20, 21], [22, 23] and [24, 25, 26]. Strain-induced moiré patterns have been realized in vertically stacked heterobilayers with intrinsic lattice mismatch, such as [27], [28, 29], and graphene/hexagonal boron nitride [30, 31], where the difference in lattice constants produces moiré modulations. More recently, attention has turned to heterostrain in homostacks [32], a state in which adjacent layers of the same material experience different lattice deformations, offering an additional means to further tune moiré geometry and electronic structure. While its potential has been explored in theory [33, 34, 35, 36, 37], experimental progress remains limited. Most observations arise from unintentional strain, and only a few device demonstrations have achieved controlled heterostrain [38, 39, 40]. Most importantly, both twist-controlled and strain-controlled moiré constructions remain lab-scale, without obvious pathways for scalability and industry adoption.
Strain engineering via stressor film deposition is a standard industry process in complementary metal-oxide-semiconductor (CMOS) integrated circuit fabrication, where controlled stress is engineered to enhance carrier mobility [41, 42]. Adapted to 2D materials, stressor thin films can transfer strain to the underlying lattice, providing a scalable and controllable route to achieve heterostrain [43, 44]. By further patterning the stressor films with lithography, both biaxial and uniaxial strains can be locally introduced in a site-specific manner [45, 46]. The strain orientation can be aligned with specific crystallographic directions, while its spatial distribution and magnitude can be tuned by adjusting the geometry and thickness of the stressor films. This tunability enables precise control of strain fields in the materials.
Results
Patterned stressor films and heterostrain
Experiments were conducted on flakes exfoliated from single crystals of 2H-. We first transferred flakes onto Si substrates using Scotch tape exfoliation and a polypropylene carbonate (PPC) stamp. The flakes were then transferred onto thin membrane TEM grids for high-resolution STEM imaging. Next, a patterned stressor film was deposited onto exfoliated flakes via e-beam lithography and e-beam evaporation (see Methods and Supplementary Fig. 1 for details). The compressive stressor film consists of a 50 nm thick layer sandwiched between two 10 nm thick layers and was patterned into 10 m-wide stripes spaced 10 m apart. The resulting sample geometry is flakes capped with a patterned stressor film, as illustrated in Figure 1a.

Figure 1. Stressor film-induced heterostrain in MoS₂. a, Schematic of patterned Al₂O₃/SiO₂/Al₂O₃ stressor film transferring heterostrain onto MoS₂ locally. Red arrows indicate the strain direction. b, Optical image of a MoS₂ flake under the patterned film on the Si substrate. Scale bar, 5 µm. The Raman spectra shown in c,d come from regions indicated by the arrow. c,d, Raman spectra of and modes of MoS₂ across the uncapped to capped regions, respectively. The spectra are plotted with a vertical offset for visual clarity, with the spectra from the uncapped region at the top and from the capped region at the bottom. e, Corresponding and peak position profiles obtained by Lorentzian fitting. f, Cross-sectional atomic-resolution HAADF-STEM images of MoS₂ under the region close to the stressor edge. Scale bar, 5 nm. The panels on the left and right show magnified views of the regions highlighted by the green and yellow boxes. Scale bars, 1 nm. g, Layer-dependent Mo-Mo interatomic distances from experiment (blue) and simulation (red). Experimental values were obtained by quantitative atom tracking from f. The error bars are the standard deviations of experimental Mo-Mo distances. Simulated values were calculated from molecular dynamics for each layer. The dashed lines are polynomial fits to the data, drawn as a visual guide to the overall trend. h, Simulated atomic configuration model of MoS₂ multilayer after application of 1%
To confirm the presence of strain in flakes due to the stressor film, we first performed Raman spectroscopy since the mode is highly sensitive to in-plane strain as it reflects in-plane intralayer vibrations [47, 48]. Figure 1b shows an optical image of a 20 nm-thick flake covered with the patterned stressor film on a Si substrate for Raman measurements. We focus on the mode at 384 across regions from the uncapped to the stressor-capped areas of the flake (Figure 1c). The mode first shifts to higher frequencies near the stressor edge, indicative of an in-plane compressive strain. Within the capped region under the stressor, the mode redshifts, indicating tensile strain transferred by the stressed film. The change of the peak position of the mode is shown in Figure 1e. The out-of-plane vibrational mode at 409 cm also shows a downshift in the capped region (Figure 1d,Figure 1e). The peak is sensitive to carrier concentration [49], and the unchanged full width at half maximum (FWHM) of the peak indicates that no additional defects were introduced during film evaporation (Supplementary Fig. 2). This suggests that the peak shift is from an accompanying out-of-plane strain in the capped region [48]. The observed Raman spectral shifts due to stressor films are consistent with previous reports [45, 50], confirming that the patterned stressor film effectively induces strain in MoS (Supplementary Fig. 3). Microscopically, the stressor film edge exerts a lateral force that pushes the underlying flake sideways, inducing a localized compressive strain near the edge and tensile strain under the stressor film (Figure 1a schematic).
Next, we investigate how deeply the applied strain penetrates into the MoS. Cross-sectional STEM imaging directly resolves in-plane lattice distortions layer-by-layer with atomic resolution. Figure 1f shows the atomic resolution high-angle annular dark field (HAADF)-STEM image of the top several MoS layers viewed along the [] direction near the edge of the stressor film (see Supplementary Fig. 4). In the 2H stacking polytype (space group P6/mmc), Mo atoms in every second layer are vertically aligned. However, a closer look at the topmost layer reveals a gradual shift of Mo positions: the left side retains the 2H registry, while on the right, the Mo atoms in the first and third layer are no longer aligned. This indicates a local stacking rearrangement driven by the stressor-induced strain in the topmost MoS layer (Supplementary Fig. 4). Figure 1g shows the averaged in-plane Mo-Mo distance for each layer measured by quantitative atom tracking, which shows that the topmost layer is under compressive strain of %, and the second layer is under % strain, while the rest of the layers remain unstrained.
To confirm these observations, Molecular Dynamics (MD) simulations were performed on a five-layer MoS model using the LAMMPS package [51] with the REBOMoS potential [52] for intralayer interactions and a Lennard-Jones potential for interlayer coupling (see Methods for details) [37]. After energy minimization, a total of 1% compressive strain was applied to the top layer through incremental ramping while the bottom layer was fixed to mimic the substrate constraint. The simulated Mo-Mo distances (Figure 1g) agree well with the experimentally measured values. The simulated atomic configuration and magnified stacking views (Figure 1h) also match the stacking rearrangement observed in STEM. In aggregate, the experimental and simulated results show that the strain imparted by the patterned stressor film is mainly concentrated in the top two layers of .
Moiré pattern formation by heterostrain
Having identified the heterostrain in the top two layers of , we now focus on plan-view imaging to visualize the structural arrangements at the nanoscale. Figure 2a shows an optical image of a flake capped with a -wide stressor film on a holey membrane. To measure the local strain, we use a nanobeam 4D-STEM approach [53] combined with exit-wave power cepstrum (EWPC) analysis [54], which enables high precision strain mapping across large areas ranging from a few to several hundred nanometers while being robust against sample tilts and noise in electron diffraction patterns. Figure 2b shows a reconstructed, virtual annular dark-field STEM image from the 4D-STEM data, focusing on the region near the stressor film edge where the strain variations are most pronounced. A moiré superlattice with varying domain size and patterns is clearly visible. The corresponding strain maps (Figure 2c-Figure 2e) reveal a largely biaxial tensile strain of in along with shear components under the stressor and near the stressor edge. We note that the measured strain values represent the projected strain averaged through the entire nm flake thickness, while in reality the top two layers experience the strain of and , as shown in Figure 1f,Figure 1g.

Figure 2. Evolution of strain distribution in MoS₂ under the stressor film. a, Optical microscopy image of MoS₂ under the patterned stressor film. Scale bar, 5 m. b, Reconstructed virtual annular dark-field image from the 4D-STEM dataset, obtained by integrating the signal in the diffraction pattern over an angular range of 34-107 mrad at each scan position. The image corresponds to a zoomed-in region near the stressor film edge marked in a. Scale bar, 500 nm. c-e, Strain maps for , , components, obtained from the same region in b through exit wave power cepstrum analysis. The overlaid dashed lines depict the stressor film edge. Scale bars, 500 nm.
Regions away from the stressor film exhibit a predominantly uniaxial compressive strain. This transition from a biaxial strain near the stressor to uniaxial strain away from the stressor produces a distinct change in the resulting moiré patterns. Directly under the stressor film (top part of Figure 2b), we observe distorted hexagonal (almost diamond-like) domains that are elongated perpendicular to the stressor edge, with domain boundaries appearing as bright lines. The larger strain in this region results in denser moiré domains with moiré periodicity of 30-40 nm. Moving farther from the stressor, the domains become progressively larger and distorted, eventually evolving into a stripe pattern under uniaxial compressive strain away from the stressor film. These moiré patterns persist over several micrometers. We note that defects introduced during the exfoliation locally modulate the strain field and introduce discontinuities in the moiré periodicity, particularly the diagonal straight fold in Figure 2b. Nevertheless, on the right side of the fold the moiré pattern evolves from hexagonal to stripe, and merges into the stripe domains at the bottom of the flake. Strain maps further show that the fold perturbs the lattice only along the narrow line, and the dominant variations align parallel to the stressor edge across both sides of the fold. The overall resulting moiré orientation remains perpendicular to the stressor edge. As a control, pristine flakes without stressor films show no moiré patterns or evident bending or defects (Supplementary Fig. 5), confirming that the observed moiré structures are induced by the stressor. Stressor-induced moiré patterns are also observed in rhombohedral-stacked - flakes (Supplementary Fig. 6), showing that the effect is not limited to the 2H phase.
Atomic structure of reconstructed lattice
The stressor film-induced strains thus alter the layer stacking in the top few layers of and create moiré patterns of different geometries. To gain deeper insights into the atomic-scale mechanism governing the moiré structure, we acquire atomic-scale plan-view HAADF-STEM images of the sample. The reconstructed, virtual dark-field STEM images clearly show the stripe and distorted hexagon moiré domains with distinct periodicities (Figure 3a,Figure 3d). The symmetry and magnitude of the applied strain field vary with the spatial position relative to the stressor film edge and determine the moiré pattern geometry: Stripe patterns due to uniaxial strain away from the stressor edge (Figure 3b) and distorted hexagonal moiré networks due to biaxial strain close to the stressor edge (Figure 3e).

Figure 3. Atomic structure of moiré domains in MoS₂. a, Plan-view low-magnification HAADF-STEM image of stripe-type moiré with domain boundaries highlighted by white dashed lines. Scale bar, 30 nm. b, Schematic of rigid moiré pattern generated by uniaxial and shear heterostrain. c, Zoomed-in filtered HAADF-STEM images of selected areas marked by colored dots across a domain boundary in a from 2H,
Atomically resolved STEM images were acquired across the domain boundary for both moiré patterns (Figure 3c,Figure 3f and Supplementary Fig. 7), which reveal changing stacking configurations. Schematics of two layers experiencing heterostrain and an interlayer shift, are drawn below the STEM images. These stacking arrangements reproduce the observed STEM images (see Supplementary Fig. 8 for simulations). Thus, a continuous change in atomic registry occurs over the span of -nm-wide domain boundary that separates neighboring 2H-stacking domains. From left to right, the stacking sequence goes from 2H to the bridge (Br) region, then to saddle point (SP) stacking, back to Br, and finally to 2H stacking. Importantly, in the stripe moiré region (Figure 3c), in-plane lattice displacements are along the zigzag direction across the domain boundary, whereas in the hexagonal moiré region (Figure 3f) the lattice shifts are along the armchair direction. Previous studies have shown that interlayer sliding along the zigzag direction avoids unfavorable metal-metal and chalcogen-chalcogen overlap, which stabilizes stripe domains, while sliding along the armchair direction allows the lattice to relax into hexagonal networks [39]. Our observations are consistent with this microscopic understanding. The uniaxial strain drives interlayer displacements along the zigzag direction, leading to stripe moirés away from the stressor, while the biaxial strain couples to both crystallographic axes and leads to lattice shifts along the armchair direction, producing a distorted hexagonal moiré structure near the stressor edge.
Polarization textures in moiré patterns
The continuous stacking changes over the domain boundaries can induce polarization locally due to shifts in the atom registry. The interlayer shifts modulate charge redistribution and produce electric polarization with a spatially varying texture [11]. We employ nanobeam 4D-STEM combined with a cepstral-based algorithm of the exit wave imaginary cepstrum (EWIC) [55] to directly measure the polarization at domain boundaries. In EWIC, the logarithm of nanobeam electron diffraction (NBED) patterns at each scan position are Fourier transformed, and the imaginary component of the transformed data is analyzed. This suppresses intensity variations of NBED coming from extrinsic factors such as sample thickness, tilt, and noise, isolating asymmetries that correspond to intrinsic local polar displacements. Unlike Bragg-disk intensity methods, EWIC compiles information from the entire diffraction pattern, making it dose-efficient and resolving displacement vectors with picometer precision, enabling robust polarization mapping in beam-sensitive 2D materials. Figure 4a,Figure 4e show the reconstructed ADF STEM images of the stripe and hexagonal moirés, respectively. Representative NBED patterns at the domain boundary are shown in Figure 4b,Figure 4f. The corresponding EWIC transforms (Figure 4c,Figure 4g) reveal asymmetric intensity features that can be interpreted as dipoles [56], highlighted by arrows. Each dipole encodes both the direction and magnitude of the local interlayer displacement vector, with the dipole length corresponding to the displacement magnitude. The measured dipole lengths are in the range of 1.5–2.0 Å near the domain boundaries for the stripe and hexagonal patterns, in the same range of the unit cell shifts of Å along the armchair direction and of Å along the zigzag direction, where is the lattice constant, 3.16 Å. By scanning over the entire area, these local dipoles are mapped to yield the full polar displacement field of the superlattice. The resulting polar displacement maps for the stripe and hexagonal moiré patterns are shown in Figure 4d,Figure 4h. Each pixel color encodes the direction of the local displacement vector. Within the centrosymmetric 2H domains, no polar distortion is observed. Strong in-plane polar displacements are observed in the vicinity of the domain boundaries, where the polarization vectors on opposite sides of the boundary are oriented in opposite directions. These relative displacements originate from the interlayer stacking transitions going from Br→SP→Br between adjacent 2H domains. Owing to the broken mirror symmetry, the domain boundaries develop in-plane polar textures. Hence, in-plane polarization is induced in non-polar 2H MoS flakes due to controlled construction of moiré patterns using stressor films.

Figure 4. In-plane polar textures in heterostrained MoS2. a,e, Reconstructed virtual annular dark field images of stripe and hexagonal moiré patterns, respectively. b,f, Representative nano-beam electron diffraction (NBED) patterns with logarithmic intensity scale from the regions marked by colored dots in a,e. c,g, Corresponding exit wave imaginary cepstrum (EWIC) patterns calculated from b,f. The arrows indicate the directions of polar displacements. d,h, In-plane polarization maps of the stripe and hexagonal
The orientation of these polarization vectors is directly coupled to the strain state and the stacking configuration. In the stripe moiré, where uniaxial strain is present, the relative sliding of adjacent layers compresses the lattice perpendicular to the boundary. This produces polarization vectors pointing toward the domain boundary under compressive strain. In the hexagonal moiré, biaxial tensile strain couples to both crystallographic axes, while an additional shear component distorts the ideal biaxial strain and skews the hexagonal domains. As a result, the polarization vectors at the boundaries are not strictly perpendicular to the domain boundaries but instead tilt, as well as outward-oriented along the hexagonal domain boundaries. These observations show that the symmetry of the applied heterostrain not only reflects the moiré geometry but also governs the orientation of the polarization vectors at the domain boundaries.
The lattice reconstruction in moiré superlattices naturally generates intralayer strain. The spatially inhomogeneous strain fields are particularly important, as they create asymmetric distributions of electron orbitals at the domain boundaries, hence giving rise to a polarization [57]. We examine the local strain distribution in the stripe and hexagonal moiré patterns and their coupling to polarization textures. Figure 5a,Figure 5b show the strain maps for and components in the stripe moiré region, obtained from the cepstrum analysis of 4D STEM data. Pronounced spatial variations are observed at the moiré domain boundaries, where opposite strain values appear on either side of the boundary. This alternating strain results from lattice relaxation and interlayer coupling under uniaxial deformation. The resulting strain gradients are concentrated at the boundaries, and through flexoelectric coupling they should induce an electric polarization according to:

Figure 5. Strain-gradient induced polarization at domain boundaries. a,b, Strain maps for and components of the moiré stripe. The overlaid dashed lines depict the moiré domain boundaries. c, Strain gradient map calculated from the derivative of the strain along and axis in a,b. d,e, Strain maps for and components of the moiré hexagon. f, Strain gradient map calculated from the derivative of the strain along and axis in d,e. Color denotes direction of polarization vectors and strength gives magnitude. All scale bars, 20 nm.
(See Supplementary Information Section 9 for details). Consistent with Figure 4d, the corresponding strain gradient map (Figure 5c) reveals polarization vectors pointing oppositely relative to the stripe boundaries. In the hexagonal moiré region, the strain maps (Figure 5d,Figure 5e) also show strong variations at the domain boundaries. The corresponding strain gradient map (Figure 5f) shows a close correspondence between the strain field and polarization texture (Figure 4h). The strain gradient orientation aligns with the polar displacement vectors and follows the distorted hexagonal boundaries. This correlation suggests that the polar distortion originates not only from relative lateral translations between layers (Figure 4), but also from accompanying lattice distortions localized at the domain boundaries (Figure 5).
Conclusion
In this work, we demonstrate controlled heterostrain and scalable and deterministic construction of moiré superlattice in using patterned stressor thin films and present structural information from both cross-sectional and plan-view perspectives. The applied strain is primarily transferred to the top two layers and concentrated laterally near the stressor edge. The strain field is biaxial in regions near the stressor edge and evolves into uniaxial farther away, giving rise to distinct moiré geometries: distorted hexagonal networks under combined biaxial and shear strain and stripe domains under uniaxial strain. At the atomic scale, these moiré patterns originate from in-plane lattice displacements along different crystallographic axes that relax into different types of domain networks. Stacking changes at the domain boundaries induce in-plane polar distortions, and the resulting polarization textures mirror the underlying strain state. These findings establish patterned stressor films as a scalable and controllable means to engineer designer moiré patterns, going beyond what is possible with simple twist or intrinsic lattice mismatch between layers, to induce novel properties, such as in-plane polarization in non-polar .
Methods
Sample preparation
- bulk crystals (HQ Graphene) were mechanically exfoliated using Scotch tape onto pre-cleaned substrates. Target flakes were identified with an optical microscope and transferred to membrane TEM grids and Si substrates using a polypropylene carbonate stamp. The TEM grids were mounted onto a custom holder and spin-coated with poly(methyl methacrylate) at 2000 rpm for 2 min, and baked at 170°C for 2 min. Electron-beam lithography (Nabity NPGS, Zeiss Supra SEM) was used to write stripes spaced apart. After pattern development, a stressor film of (10 nm)/ (50 nm)/ (10 nm) was deposited using e-beam evaporation at base pressure of 5e-6 torr and at a rate of Å/s. The nominal film thickness was measured using quartz crystal monitor during deposition. The film exhibits compressive stress of -105.7 MPa, Supplementary Figure 1 shows the surface profilometer data.
Raman spectroscopy
MoS samples for Raman spectroscopy were exfoliated on Si substrates, followed by the same fabrication conditions for TEM samples. Raman spectra were taken using a WITec Alpha300R Confocal Raman Microscope at an excitation wavelength of 532 nm with 1800 gr/mm diffraction gratings. The laser spot size is estimated to be 0.7 m. Raman line scans were performed along the direction perpendicular to the stressor’s edge and were collected at 0.15 m steps. The laser power was kept below 1 mW to prevent laser induced damage to the flakes. The and Raman modes were analyzed by fitting with Lorentzian functions to extract the frequency and full widths at half maximum.
Scanning transmission electron microscopy characterization
Cross-section STEM specimens were prepared using a Thermo Fisher Scientific Helios G4 UX focused ion beam (FIB) using standard lift out and thinning methods from the samples on Si substrates, with a final milling at 2 kV. Plan-view samples were prepared by transferring exfoliated MoS flakes onto SiN TEM grids, followed by the stressor film fabrication process described above.
Electron microscopy data were acquired using a Cs-corrected Thermo Fisher Scientific Spectra 300 X-CFEG operating at 120 kV equipped with an Electron Microscope Pixel Array Detector (EMPAD) detector for 4D-STEM. A convergence angle was 30 mrad, and inner and outer collection angles were approximately 60 and 200 mrad, respectively, for HAADF-STEM imaging. To obtain high signal-to-noise-ratio images, multiple fast acquisition images were acquired and subsequently aligned and summed using a rigid registration process optimized for noisy images. EELS was used to estimate the sample thickness of 20 nm. The measured thickness was 0.36 mean free path , we used an inelastic mean free path of 55 nm for MoS at 120 kV.
The 4D-STEM datasets were acquired at 120 kV with camera length of 295 mm. A 0.5-mrad convergence angle was used for the low-magnification strain maps (Figure 2), leading to a 3.42 nm probe size (defined by Full Width Half Maximum (FWHM) probe diameter) and 13.4 m depth of focus. A 3-mrad convergence angle was used for the polar distortion maps (Figure 4,Figure 5), leading to a nm probe size and m depth of focus. For all datasets, 2 ms dwell time and real-space pixels were used to acquire the EMPAD 4D datasets. The data in Fig. 2 was taken at 53k magnification with a 1.88 m field of view, and the data in Fig. 4,5 were taken at 850k magnification with a 117 nm field of view. All reconstructed virtual annular dark-field images from the 4D-STEM datasets were obtained by integrating the signal in the diffraction pattern over an angular range of 34-107 mrad at each scan position.
The EWPC method was used to measure the strain fields of MoS. The raw diffraction pattern was first transformed to logarithmic scale. The power cepstrum image was acquired through fast-Fourier transform of the diffraction. The EWPC spots corresponding to the () and () spacings were measured through the center of mass method to determine the two lattice vectors along x and y directions. The strain maps for , , and components can be calculated through polar decomposition method. For EWIC analysis, the imaginary component of the cepstral transform of the logarithm of the diffraction patterns was extracted. Dipole features, appearing as positive-negative lobe pairs in the EWIC transform, were fit with a 2D Gaussian function to extract peak positions. The separation of the lobes corresponds to the magnitude of polar displacement, and their orientation defines the displacement direction. The resulting polar vector fields were mapped across the scanned region to reconstruct polarization distributions.
Theoretical simulations
To construct an atomistic model for MoS multilayer, we build a five-layer MoS structure with dimensions of approximately 158.02 164.22 Å. MD simulation is performed using the LAMMPS package [51]. The reactive empirical bond-order potential for MoS (REBOMoS) [52] is employed to describe the intralayer interactions, while van der Waals forces described by the Lennard-Jones potential account for the interlayer interactions. An initial minimization step using the conjugate gradient method with convergence force criteria of was used, which resulted in an equilibrium interlayer distance of 6.13 Å, while the intralayer Mo–Mo lattice length is found as 3.16 Å. Ramped strain is applied in the MD simulations by introducing a 0.02% compressive strain to all atoms of the top layer, such that the leftmost atoms remain fixed while the rightmost atoms are displaced by 0.02% of the top-layer length, imposing a half-symmetry boundary condition. During this process, the bottom layer and the left edge are fixed to prevent relaxation of the imposed strain. This procedure is repeated 50 times, resulting in a total compressive strain of 1% applied to the top layer. The strain transfer length scale and the Mo-Mo length is found by calculating the average strain in each layer. The average results are found by randomly selecting five sets of Mo-Mo pairs within each layer, excluding atoms at the edges to avoid boundary effects. For each pair, the interatomic distance is measured, and the corresponding strain is calculated by dividing the displacement by the original length. The results are then averaged as shown in Figure 1g.
ACKNOWLEDGEMENTS
The work was supported by the DOE ECMP Award number DE-SC0023905. Sample fabrication was performed in part at the Cornell NanoScale Facility, a member of the National Nanotechnology Coordinated Infrastructure (NNCI), which is supported by the National Science Foundation (Grant NNCI-2025233). Electron microscopy was supported by the National Science Foundation under Cooperative Agreement No. DMR-2039380, and additional microscopy characterization made use of the Cornell Center for Materials Research Shared Facilities, with the Thermo Fisher Helios G4 X FIB supported by NSF (DMR-1539918).
AUTHOR CONTRIBUTIONS
Y.-M.W. and J.J.C. conceived and designed the experiments. Y.-M.W., S.D.F., and S.S. performed the STEM measurements and analysis. Y.-M.W., S.L., and N.L.W. fabricated the samples. Y.-M.W., S.L. and N.L.W. performed and analyzed Raman measurements. Y.X. and H.A. performed the theoretical simulations. Y.-M.W., S.L. and J.J.C. wrote the manuscript, with input from all authors.
COMPETING FINANCIAL INTERESTS
The authors declare no competing financial interests.
compressive strain to the top layer. All atoms are rendered using gray color to enhance the visualization of a few select atoms of Mo (cyan) and S (yellow). Magnified views of the green and yellow boxed regions are shown for comparison with experiment.
to the bridge region (Br), then to saddle point (SP), and back to Br, and 2H stackings with corresponding atomic schematics. The arrow indicates the lattice shift along the zigzag direction. Scale bars, 1 nm. d, Low-magnification HAADF-STEM image of distorted hexagonal moiré with domain boundaries highlighted by white dashed lines. Scale bar, 50 nm. e, Schematic of rigid moiré pattern generated by biaxial and shear heterostrain. f, Zoomed-in filtered HAADF-STEM images of selected areas across a domain boundary in d with corresponding atomic schematics. The arrow indicates the lattice shift along the armchair direction. Scale bars, 1 nm.
moiré, respectively. Color denotes direction of polarization vectors, as specified by the color wheel, and strength gives magnitude. Insets show zoomed-in views of the maps from the dashed boxes. All scale bars, 20 nm.
References
- [1]Balents, L., Dean, C. R., Efetov, D. K. & Young, A. F. Superconductivity and strong correlations in moiré flat bands. Nature Physics 16, 725–733 (2020).DOI
- [2]Andrei, E. Y. et al. The marvels of moiré materials. Nature Reviews Materials 6, 201–206 (2021).DOI
- [3]Mak, K. F. & Shan, J. Semiconductor moiré materials. Nature Nanotechnology 17, 686–695 (2022).DOI
- [4]Cao, Y. et al. Unconventional superconductivity in magic-angle graphene superlattices. Nature 556, 43–50 (2018).DOI
- [5]Yankowitz, M. et al. Tuning superconductivity in twisted bilayer graphene. Science 363, 1059–1064 (2019).DOI
- [6]Sunku, S. et al. Photonic crystals for nano-light in moiré graphene superlattices. Science 362, 1153–1156 (2018).
- [7]Huang, S. et al. Topologically protected helical states in minimally twisted bilayer graphene. Physical Review Letters 121, 037702 (2018).DOI
- [8]Seyler, K. L. et al. Signatures of moiré-trapped valley excitons in MoSe₂/WSe₂ heterobilayers. Nature 567, 66–70 (2019).DOI
- [9]Cao, Y. et al. Correlated insulator behavior at half-filling in magic-angle graphene superlattices. Nature 556, 80–84 (2018).
- [10]Zhou, Y. et al. Bilayer wigner crystals in a transition metal dichalcogenide heterostructure. Nature 595, 48–52 (2021).DOI
- [11]Ko, K. et al. Operando electron microscopy investigation of polar domain dynamics in twisted van der Waals homobilayers. Nature Materials 22, 992-998 (2023).DOI
- [12]Yasuda, K., Wang, X., Watanabe, K., Taniguchi, T. & Jarillo-Herrero, P. Stacking-engineered ferroelectricity in bilayer boron nitride. Science 372, 1458–1462 (2021).
- [13]Rogée, L. et al. Ferroelectricity in untwisted heterobilayers of transition metal dichalcogenides. Science 376, 973–978 (2022).DOI
- [14]Woods, C. et al. Commensurate–incommensurate transition in graphene on hexagonal boron nitride. Nature Physics 10, 451–456 (2014).DOI
- [15]Yoo, H. et al. Atomic and electronic reconstruction at the van der Waals interface in twisted bilayer graphene. Nature Materials 18, 448–453 (2019).DOI
- [16]Weston, A. et al. Atomic reconstruction in twisted bilayers of transition metal dichalcogenides. Nature Nanotechnology 15, 592–597 (2020).DOI
- [17]Lu, X. et al. Superconductors, orbital magnets and correlated states in magic-angle bilayer graphene. Nature 574, 653–657 (2019).DOI
- [18]Wang, L. et al. Correlated electronic phases in twisted bilayer transition metal dichalcogenides. Nature Materials 19, 861–866 (2020).DOI
- [19]Ghiotto, A. et al. Quantum criticality in twisted transition metal dichalcogenides. Nature 597, 345–349 (2021).DOI
- [20]Li, H. et al. Mapping charge excitations in generalized Wigner crystals. Nature Nanotechnology 19, 618–623 (2024).DOI
- [21]Li, H. et al. Imaging moiré excited states with photocurrent tunnelling microscopy. Nature Materials 23, 633–638 (2024).DOI
- [22]Wang, P. et al. One-dimensional Luttinger liquids in a two-dimensional moiré lattice. Nature 605, 57–62 (2022).DOI
- [23]Yu, G. et al. Evidence for two dimensional anisotropic Luttinger liquids at millikelvin temperatures. Nature Communications 14, 7025 (2023).DOI
- [24]Anderson, E. et al. Programming correlated magnetic states with gate-controlled moiré geometry. Science 381, 325–330 (2023).
- [25]Cai, J. et al. Signatures of fractional quantum anomalous hall states in twisted MoTe₂. Nature 622, 63–68 (2023).DOI
- [26]Zeng, Y. et al. Thermodynamic evidence of fractional Chern insulator in moiré MoTe₂. Nature 622, 69–73 (2023).DOI
- [27]Regan, E. C. et al. Mott and generalized Wigner crystal states in WSe₂/WS₂ moiré superlattices. Nature 579, 359–363 (2020).DOI
- [28]Li, T. et al. Quantum anomalous Hall effect from intertwined moiré bands. Nature 600, 641–646 (2021).DOI
- [29]Zhao, W. et al. Realization of the haldane Chern insulator in a moiré lattice. Nature Physics 20, 275–280 (2024).arxiv.org/abs/2207.02312
- [30]Yankowitz, M. et al. Emergence of superlattice Dirac points in graphene on hexagonal boron nitride. Nature Physics 8, 382–386 (2012).DOI
- [31]Yankowitz, M., Ma, Q., Jarillo-Herrero, P. & LeRoy, B. J. van der Waals heterostructures combining graphene and hexagonal boron nitride. Nature Reviews Physics 1, 112–125 (2019).DOI
- [32]Huder, L. et al. Electronic spectrum of twisted graphene layers under heterostrain. Physical Review Letters 120, 156405 (2018).DOI
- [33]Bi, Z., Yuan, N. F. & Fu, L. Designing flat bands by strain. Physical Review B 100, 035448 (2019).DOI
- [34]Mannaï, M. & Haddad, S. Twistronics versus straintronics in twisted bilayers of graphene and transition metal dichalcogenides. Physical Review B 103, L201112 (2021).DOI
- [35]Kögl, M. et al. Moiré straintronics: a universal platform for reconfigurable quantum materials. npj 2D Materials and Applications 7, 32 (2023).arxiv.org/abs/2207.12115
- [36]Escudero, F., Sinner, A., Zhan, Z., Pantaleón, P. A. & Guinea, F. Designing moiré patterns by strain. Physical Review Research 6, 023203 (2024).
- [37]Dey, A., Hasan, N., Wu, S. M. & Askari, H. Memorization of strain-induced moiré patterns in vertical van der waals materials. ACS Applied Materials & Interfaces 17, 16223–16233 (2025).DOI
- [38]Mesple, F. et al. Heterostrain determines flat bands in magic-angle twisted graphene layers. Physical Review Letters 127, 126405 (2021).DOI
- [39]Edelberg, D., Kumar, H., Shenoy, V., Ochoa, H. & Pasupathy, A. N. Tunable strain soliton networks confine electrons in van der Waals materials. Nature Physics 16, 1097–1102 (2020).DOI
- [40]Halbertal, D. et al. Moiré metrology of energy landscapes in van der Waals heterostructures. Nature Communications 12, 242 (2021).
- [41]Orain, S., Fiori, V., Villanueva, D., Dray, A. & Ortolland, C. Method for managing the stress due to the strained nitride capping layer in MoS₂ transistors. IEEE transactions on electron devices 54, 814–821 (2007).
- [42]Thompson, S. E., Sun, G., Choi, Y. S. & Nishida, T. Uniaxial-process-induced strained-Si: Extending the CMOS roadmap. IEEE Transactions on electron Devices 53, 1010–1020 (2006).DOI
- [43]Hou, W. et al. Strain engineering of vertical molybdenum ditelluride phase-change memristors. Nature Electronics 7, 8–16 (2024).DOI
- [44]Zhang, Y. et al. Patternable process-induced strain in 2D monolayers and heterobilayers. ACS nano 18, 4205–4215 (2024).DOI
- [45]Peña, T. et al. Strain engineering 2D MoS₂ with thin film stress capping layers. 2D Materials 8, 045001 (2021).DOI
- [46]Azizimanesha, A., Peña, T., Sewaket, A., Hou, W. & Wu, S. M. Uniaxial and biaxial strain engineering in 2D MoS₂ with lithographically patterned thin film stressors. Applied Physics Letters 118 (2021).
- [47]Lee, J.-U. et al. Strain-shear coupling in bilayer MoS₂. Nature Communications 8, 1370 (2017).
- [48]Li, H. et al. From bulk to monolayer MoS₂: evolution of Raman scattering. Advanced Functional Materials 22, 1385–1390 (2012).
- [49]Debnath, R. et al. Evolution of high-frequency Raman modes and their doping dependence in twisted bilayer MoS₂. Nanoscale 12, 17272 (2020).
- [50]Liu, Q. et al. Programmable strainscapes in a two-dimensional (2D) material monolayer. ACS Nano 19, 30125-30136 (2025).
- [51]Plimpton, S. Fast parallel algorithms for short-range molecular dynamics. Journal of Computational Physics 117, 1–19 (1995).
- [52]J Stewart and Douglas Spearot. Atomistic simulations of nanoindentation on the basal plane of crystalline molybdenum disulfide. Modelling and Simulation in Materials Science and Engineering 21, 045003 (2013).
- [53]Ophus, C. Four-dimensional scanning transmission electron microscopy (4D-STEM): From scanning nano diffraction to ptychography and beyond. Microscopy and Microanalysis 25, 563–582 (2019).
- [54]Padgett, E. et al. The exit-wave power-cepstrum transform for scanning nanobeam electron diffraction: robust strain mapping at subnanometer resolution and subpicometer precision. Ultramicroscopy 214, 112994 (2020).
- [55]Holtz, M. E. et al. Mapping polar distortions using nanobeam electron diffraction and a cepstral approach. Microscopy and Microanalysis 29, 1422–1435 (2023).
- [56]Deb, P. et al Imaging polarity in two dimensional materials by breaking Friedel’s law. Ultramicroscopy 215, 113019 (2020).
- [57]McGilly, L. J. et al. Visualization of moiré superlattices. Nature Nanotechnology 15, 580–584 (2020).DOI