Abstract:Spin relaxation and decoherence is at the heart of spintronics and spin-based quantum information science. Currently, theoretical approaches that can accurately predict spin relaxation of general solids including necessary scattering pathways and are capable of nanosecond to millisecond simulation time are urgently needed. We present a first-principles real-time density-matrix approach based on Lindblad dynamics to simulate ultrafast spin dynamics for general solid-state systems. Through the complete first-principles descriptions of pump, probe, and scattering processes including electron-phonon, electron-impurity, and electron-electron scatterings with self-consistent electronic spin-orbit couplings, our method can directly simulate the ultrafast pump-probe measurements for coupled spin and electron dynamics over nanoseconds at any temperatures and doping levels. We first apply this method to a prototypical system GaAs and obtain excellent agreement with experiments. We found that the relative contributions of different scattering mechanisms and phonon modes differ considerably between spin and carrier relaxation processes. In sharp contrast to previous work based on model Hamiltonians, we point out that the electron-electron scattering is negligible at room temperature but becomes dominant at low temperatures for spin relaxation in n -type GaAs. We further examine ultrafast dynamics in novel spin-valleytronic materials: monolayer and bilayer WSe2 with realistic defects. We find that spin relaxation is highly sensitive to local symmetry and chemical bonds around defects. For the bilayer WSe2 , we identify the scattering pathways in ultrafast dynamics and determine relevant dynamical properties, essential to its utilization of unique spin-valley-layer locking effects. Our work provides a predictive computational platform for spin dynamics in solids, which has potential for designing new materials ideal for spintronics and quantum information technology. https://doi.org/10.1103/PhysRevB.104.184418 ©2021 American Physical Society

Data-driven compression of electron-phonon interactions

Ab Initio Electron-Phonon Interactions in Correlated Electron Systems

Accelerating the calculation of electron-phonon coupling strength with machine learning

Accurate Electron-phonon Interactions from Advanced Density Functional Theory

Adaptively compressed polarizability operator for accelerating large scale \textit{ab initio} phonon calculations

Long-range quadrupole electron-phonon interaction from first principles

Efficient Calculation of the Lattice Thermal Conductivity by Atomistic Simulations with Ab Initio Accuracy

The Electron-Phonon Renormalization in the Electronic Structure Calculation: Fundamentals, Current Status, and Challenges

Ab-initio predictions of spin relaxation, dephasing and diffusion in solids

Dynamic mode decomposition of nonequilibrium electron-phonon dynamics: accelerating the first-principles real-time Boltzmann equation

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

Exciton Dispersion and Exciton-Phonon Interaction in Solids by Time-Dependent Density Functional Theory

Directly characterizing the relative strength and momentum dependence of electron-phonon coupling using resonant inelastic x-ray scattering

A machine-learning framework for accelerating spin-lattice relaxation simulations

Long-range Electron-Phonon Correlation and Peierls Dimerization in the One-Dimensional Molecular-Crystal Model

The Role of Electron Correlation Beyond the Active Space in Achieving Quantitative Predictions of Spin-Phonon Relaxation

First-principles interatomic potentials for ten elemental metals via compressed sensing

Interpolating many-body wave functions for accelerated molecular dynamics on the near-exact electronic surface

Ab initio ultrafast spin dynamics in solids

Accelerating High-Throughput Phonon Calculations via Machine Learning Universal Potentials