aiidalab / aiidalab-qe

AiiDAlab App for Quantum ESPRESSO
https://aiidalab-qe.readthedocs.io/
MIT License
12 stars 15 forks source link

Handling Molecules #734

Closed AndresOrtegaGuerrero closed 1 month ago

AndresOrtegaGuerrero commented 6 months ago

The current app misses the possibility of treating properly Molecules and tricks have to be used in this case: the user has to override defaults and reduce k-point mesh accuracy to have only 1 k-point.

Moreover, Bandstructure is not deactivated.

AndresOrtegaGuerrero commented 6 months ago

@cpignedoli

cpignedoli commented 6 months ago

I would add a simple pre-check of the structure:

superstar54 commented 6 months ago

Thanks, @cpignedoli , for the suggestions.

what about other plugins?

This pre-check is done in step 1 when the user confirms the structure, so it applies to all plugins.

if the structure has no cell add a std cell (e.g. +5A in each direction) treat it as a molecule and inform the user of the action taken pointing out how this can be overridden

One suggestion of the cell size: since in PW, we usually use MT correction for an isolated system, we need to add a cell that guaranteed the cell is twice as large as charge density (~the molecule's size).

AndresOrtegaGuerrero commented 2 months ago

From #817

Instead, in the structure editor, under “Edit StructureData”, add also “no periodicity” as an option. And test for a molecule if this works. (E.g. all mesh of k-points become 1x1x1 = Gamma only). E.g. try PDOS of a molecule, we have ETFA molecule (add also a simple molecule like O2 or CO2 to the examples!).

In any case, ensure that the issue discussing what needs to be done for low-dimensional systems (e.g. what to relax, whether to add a Martyna-Tuckermann or similar, how many k-points, …) is in the milestone for the release in 6 months (~march 2025).ma