[17a-C302-1~12] 13.8 Optical properties and light-emitting devices

11:00 AM - 11:15 AM JST(2:00 AM - 2:15 AM UTC)

[17a-C302-9] Non-adiabatic Molecular Dynamics Calculations Combined with Time Series Machine Learning Methods for Analysis of Intermediate Level Carrier Dynamics in Er-doped GaAs

〇Yuya Makino¹, Yusuke Oteki², Yoshitaka Okada², Tomah Sogabe¹ (1.Univ. of Electro-Comm, 2.RCAST, Univ. of Tokyo)

Keywords:

semiconductor,Neural Network,Nonadiabatic Molecular Dynamics

In this research, we propose an approach that integrates non-adiabatic molecular dynamics calculations with machine learning. Excited energies and non-adiabatic couplings are generated using first-principles calculations, and the obtained data is used to predict the remaining data, reducing computational costs. Non-adiabatic molecular dynamics simulations are performed using the complemented data to efficiently analyze the carrier dynamics in Er-doped GaAs, aiming to optimize the Er doping concentration for longer carrier lifetimes at intermediate levels.