search

bytedance / DeepSolid

A library combining solid quantum Monte Carlo and neural network.
Apache License 2.0
32 stars 7 forks source link

Question on ee part of ewaldsum #1

Closed mavenlin closed 1 year ago

mavenlin commented 1 year ago

Thank you for making this code open-source.

One question comes to me when I read the ewaldsum. When the electron distances are calculated the same way as the ion distances, does it imply that the electrons are sampled from one cell and mirrored periodically instead of sampled independently from each supercell? Would there be a problem doing so?

GiantElephant123 commented 1 year ago

Thanks for asking and sorry for the late response. You are right, electrons outside the chosen supercell are mirror placed. This is called supercell approximation in physics and it's employed because we can't treat all the electrons in solids (order ~ 10^23). And due to this approximation, our calculation will only make sense and reach the thermodynamic limit if we choose a large enough supercell (or a fine k-point mesh in SCF language).

mavenlin commented 1 year ago

It is a very prompt response! Thanks for clarifying.