Hello community,
I brought my attention to ASE since I am trying to simulate Raman spectroscopy. I’m very new to this, and It’s hard to judge the limitation of the Resonant Raman calculator. I just have the impression that it is suitable for single molecules and/or very small systems.
My goal is to calculate it for a simulation cell (that I got from a MD) that contains 300000 atoms. Does this seem reasonable to experienced people or computationally impossible?
If impossible, would you have any recommendation on how to compute my Raman spectra for this kind of simulation cell?
Thanks a lot
2 Likes
Have you found a solution yet?
This seems a challenging thing to do, with or without ASE. Certainly people simulate Raman for solid state unit cells. But in the harmonic lattice dynamics formalism, each of the 3N modes has a Raman frequency to be determined by diagonalisation of 3Nx3N force constants, and an intensity depending on the response to electric field and structure perturbabations along that mode. It’s not really an approach that should scale to such a cell size.
There are other methods based on molecular dynamics (not implemented in ASE) but again they require some information about electronic structure and it is not generally feasibly to perform electronic structure calculations on such a system, let along along the course of an MD trajectory. There will need to be some kind of inexpensive surrogate model involved.