Probing the structure and phonon properties of solids with auxiliary-field quantum Monte Carlo
ORAL
Abstract
Determining the accurate structure of a material is a critical step in understanding its physics. Predictive computations in correlated materials remain a major challenge. We present a direct, ab initio computation of forces and stresses with auxiliary-field quantum Monte Carlo (AFQMC) using planewave basis and multiple projector pseudopotentials. Our method potentially allows determination of the potential energy surface at a much higher efficiency than an approach based on total energies alone. In addition, we propose a fast and robust structural optimization algorithm [1] for optimizations when the forces or gradients are statistically noisy. Applying this algorithm in combination with forces and stresses computed by AFQMC, we demonstrate efficient, accurate, and full degrees-of-freedom optimizations in solids. Furthermore, we show that the AFQMC forces can be used to obtain accurate phonon spectra in solids.
*Supported by the Simons Foundation. SC is also supported by DOE under Grant No. DE-SC0001303.
–
Publication: [1] S. Chen and S. Zhang, arXiv:2204.12074 (2022), in press at Nature Computational Science.
Presenters
-
Siyuan Chen
- William & Mary