Tutorial for quantum variable cell relaxation of LaH10

[anyi-mxh]

Dear Developer,

Recently, I followed the tutorial "Tutorial for quantum variable cell relaxation of LaH10" and performed variable cell relaxation calculations on the material I am studying (relax.vc_relax(target_press=0, ensemble_loc="ensembles")). I then used the final structure of SSCHA to run DFT self-consistent and electron-phonon coupling calculations. However, I found that the structure obtained by SSCHA free energy minimization shows significant stress (as shown in the figure). In contrast, the stress value of the DFT-optimized structure is close to zero. I now have the following questions: (1) Is it reasonable that the SSCHA-optimized structure shows large stress? (2) When combining SSCHA and EPW methods to calculate superconducting properties, should the SSCHA optimized structure be used for phonon spectrum calculation?

Hope to get your reply, thank you for your help!

[DjordjeDangic]

Hello,

1. Yes, it is expected that the DFT stress calculated for structure coming from SSCHA relaxation shows a non-zero stress. Also, DFT will show non-zero forces on atoms for the SSCHA structure. This is a consequence of the fact that the minimum of total free energy is not a minimum of the Born-Oppenheimer energy surface.
2. Yes, if you are going to use SSCHA phonons.

[anyi-mxh]

Thanks for your kindly reply. It is very nice for me to understand the non-zero stress.