Abstract:The integrated experimental‐machine learning framework that extracts crucial hidden phonon thermal transport properties from ultrafast diffraction. The phonon transport results in local atomic vibrations that further links to the time‐dependent diffraction intensities. By using adjoint‐state scientific machine learning approach to solve the inverse thermal transport containing thousands of coupled differential equations, the hidden phonon properties can be reliably uncovered. One central challenge in understanding phonon thermal transport is a lack of experimental tools to investigate frequency‐resolved phonon transport. Although recent advances in computation lead to frequency‐resolved information, it is hindered by unknown defects in bulk regions and at interfaces. Here, a framework that can uncover microscopic phonon transport information in heterostructures is presented, integrating state‐of‐the‐art ultrafast electron diffraction (UED) with advanced scientific machine learning (SciML). Taking advantage of the dual temporal and reciprocal‐space resolution in UED, and the ability of SciML to solve inverse problems involving O(103) coupled Boltzmann transport equations, the frequency‐dependent interfacial transmittance and frequency‐dependent relaxation times of the heterostructure from the diffraction patterns are reliably recovered. The framework is applied to experimental Au/Si UED data, and a transport pattern beyond the diffuse mismatch model is revealed, which further enables a direct reconstruction of real‐space, real‐time, frequency‐resolved phonon dynamics across the interface. The work provides a new pathway to probe interfacial phonon transport mechanisms with unprecedented details.

Probing precise interatomic potentials by nonadiabatic nonlinear phonons

Probing the interatomic potential of solids with strong-field nonlinear phononics

High-harmonic spectroscopy probes lattice dynamics

High Harmonic Spectroscopy Probes Lattice Dynamics

Probing anharmonic phonons in WS2 van der Waals crystal by Raman spectroscopy and machine learning

Probing Far-Infrared Surface Phonon Polaritons in Semiconductor Nanostructures at Nanoscale

Panoramic Mapping of Phonon Transport from Ultrafast Electron Diffraction and Scientific Machine Learning

Probing Phonon dynamics and Electron-Phonon Coupling by High Harmonic Generation in Solids

Exploring Phononic Properties of Two-Dimensional Materials using Machine Learning Interatomic Potentials

Probing anharmonic phonons by quantum correlators: A path integral approach

Nonequilibrium Phonon Tuning And Mapping In Few-Layer Graphene With Infrared Nanoscopy

Nanoscale imaging of phonon dynamics by electron microscopy

Experimental Realization of Nonreciprocal Adiabatic Transfer of Phonons in a Dynamically Modulated Nanomechanical Topological Insulator.

Engineering Acoustic Phonons and Electron-Phonon Coupling by the Nanoscale Interface

Modulation of probe signal in coherent phonon detection revisited: Analytical and first-principles computational analyses

Accelerating High-Throughput Phonon Calculations via Machine Learning Universal Potentials

First-principles theory of nonlinear long-range electron-phonon interaction

Phonon non-equilibrium effects on interface thermal resistance between graphene and substrates

Excitation protocols for nonlinear phononics in bismuth and antimony

Developing machine learning potential for classical molecular dynamics simulation with superior phonon properties

Phonon-Coupled High-Harmonic Generation for Exploring Nonadiabatic Electron-Phonon Interactions