Resolving Nonequilibrium Phonon Dynamics in Moiré Heterostructures with Machine Learned Interatomic Potentials for Enhancing Thermal Transport

ORAL

Abstract

Heterostructures involving van der Waals (vdW) materials host moiré patterns that strongly modulate their thermal properties, including interfacial vibrational coupling and thermal boundary conductance (TBC). We have previously shown that, in some transition metal dichalcogenide (TMD) bilayers, nonequilibrium phonon populations created by optical excitation can increase the effective TBC due to specific anharmonic phonon coupling involving the scattering of a K phonon in one layer and the emission of a low-energy, near-Γ phonon in the other [1]. However, a complete mode-resolved picture of nonequilibrium phonon dynamics in moiré systems has been inaccessible: ab initio calculations of 3-phonon scattering processes are intractable at moiré scales, and traditional force fields and machine-learned interatomic potentials (MLIPs) often struggle to capture both intra- and interlayer interactions to the accuracy required to describe 3-phonon scattering mechanisms.

Here, we present a multiphonon-scattering approach integrating split-MLIPs, recently developed for layered materials [2], to recover momentum-resolved phonon populations and scattering pathways in moiré heterostructures from ultrafast electron diffraction analysis from our collaborators. Applied to bilayer MoSe₂/WSe₂, we find long-lived M-, K-, and Q-valley phonons with lifetimes varying systematically with twist and material pairing, identifying optimal twists and pairings that maximize interlayer phonon coupling.

*This work was supported by the Office of Naval Research through the Multi-University Research Initiative (MURI) on Twist-Optics (Grant # N00014-23-1-2567) and by the National Science Foundation CAREER award through grant no. DMR-2238328. JDG acknowledges the partial support of the Natural Sciences and Engineering Research Council of Canada (NSERC), PGS D-568202-2022.

Publication: [1] Johnson, A.C. et al. (2024) 'Hidden phonon highways promote photoinduced interlayer energy transfer in twisted transition metal dichalcogenide heterostructures', Science Advances, 10(4).
[2] Georgaras, J.D. et al. (2025) Accurate, transferable, and verifiable machine-learned interatomic potentials for layered materials, arXiv: .

Presenters

  • Johnathan Dimitrios Georgaras

    • Stanford University

Authors

  • Johnathan Dimitrios Georgaras

    • Stanford University
  • Felipe H da Jornada

    • Stanford University
  • Fang Liu

    • Stanford University
  • Archana Raja

    • Lawrence Berkeley National Laboratory
  • Akash Ramdas

    • Stanford University