Straining the flatland: novel physics from strain engineering of atomically thin graphene and molybdenum disulfide
MetadataShow full item record
2D materials like graphene and MoS_2 are atomically thin, extremely strong and flexible, making them attractive for integration into strain engineered devices. Strain on these materials can change physical properties, as well as induce exotic physics, not typically seen in solid-state systems. Here, we probe the novel physics arising from distorted lattices of 2D materials, strained by nanopillars indentation and microelectromechanical systems (MEMS), using Raman and photoluminescence (PL) spectroscopy. From nanopillars strained multilayer MoS_2, we observe exciton and charge carrier funneling due to strain, inducing dissociation of excitons in to free electron-hole pairs in the indirect material. Using MEMS devices, we were able to dynamically strain monolayer and multilayer graphene. Multilayer graphene under MEMS strain showed signatures of loss in Bernal stacking due to shear of the individual layers, indicating that MEMS can be used to tune the layer commensuration with tensile strain. We further explore simulation of pseudo-magnetic fields (PMFs) generated in monolayer graphene strained by MEMS, using machine learning, to accelerate and optimize the strength and uniformity of the PMF in new graphene geometries. Nanopillars provide non-uniform, centrally biaxial strain to multilayer MoS_2 transferred on top. Raman E^1_2g and PL redshift across the pillar confirms 1-2% strain in the material. We also observe a softening in the A_1g Raman mode and an enhancement in the overall PL with an increase in radiative trions, under strain. The changes in these charge-dependent features indicates funneling of charge carriers and neutral excitons to the apex of the pillar, as strain locally deforms the band structure of the conduction and valence bands. DFT calculations of the band structure in bilayer MoS_2 under biaxial strain shows the conduction band is lowered, further increasing the indirectness of multilayer MoS_2. This should cause the PL intensity to decrease, whereas we observe an increase in MoS_2 PL intensity under strain. We theorize that this is due to a dissociation of excitons into free electron-hole pairs. The increase in charge carrier densities due to strain leads to a renormalization of the local band structure and increased dielectric screening, supporting free electron-hole recombination at the K-point without momentum restrictions. In turn, electron-hole recombination occurs around the K-point inducing a high intensity PL, which opens attractive opportunities for utilization in optoelectronic devices. MEMS chevron actuators can dynamically strain 2D materials, which we demonstrate through uniaxial strain in CVD and exfoliated graphene. We use a novel microstructure assisted transfer technique which can deterministically place materials on non-planar surfaces like MEMS devices. Building on previously reported 1.3% in monolayer MoS2 from our group, we report tunable 0.3% strain in CVD monolayer graphene and 1.2% strain in multilayer exfoliated graphene using MEMS chevron actuators, detected by Raman spectroscopy. The asymmetric-to-symmetric strain evolution of the 2D phonon line shape in multilayer graphene is evidence of changes in interlayer interactions, caused by shearing between layers. This demonstrates that MEMS can be used to tune the commensuration in few layer 2D materials, which is a promising avenue towards Moiré engineering. Using machine learning, we also simulate optimal monolayer graphene geometries for generating strong, uniform pseudo-magnetic fields by MEMS strain. The coupled use of finite-element methods, variational auto-encoder, and auxiliary neural network accelerates the search for PMFs in strained graphene, while optimizing the graphene shape for fabrication through electron-beam lithography. Our experimental and simulated work creates a road-map for rapid advancement in zero-field quantum Hall effect devices using graphene-integrated MEMS actuators.
RightsAttribution 4.0 International