Thomas3R
commented
1 week ago Hi,
The GFN-FF for Ln/An is currently only available in the bleeding edge version. You can get it here: https://github.com/grimme-lab/xtb/releases/tag/bleed
The GFN-FF model is not designed for ions in an aqueous solution and I believe there are more accurate special-purpose force fields for this matter. The best you can do is probably this DFTB method: https://doi.org/10.1021/acs.jctc.0c00089 However, I do not know if these force fields or the DFTB method are freely available and easily applicable.
If you want to try the GFN-FF model for this purpose, make sure that the neighbor lists of your ions are as intended (they are not bound to other atoms). You can check this in the xtb output or - if you add the --wrtopo nb flag to your xtb call - in the gfnff_lists.json file. You can find a description of this feature here: https://xtb-docs.readthedocs.io/en/latest/gfnff.html#topology-file
If they are bound to other atoms you can adjust the neighbor list using the detailed input file to adjust the neighbors. See https://xtb-docs.readthedocs.io/en/latest/gfnff.html#manual-neighbor-list-setup
This feature is only applicable for molecular calculations, but it still works for PBC if the bond is in the unit cell (both atoms are in the unit cell and neither of them is an image in a neighboring cell).
Yes, the calculations can be performed with periodic boundary conditions.
What do you want to investigate with these simulations?
Best regards, Thomas
Dear Developers,
My question is:
How to use GFN-FF_Ln_An model in xtb-6.7.1?
How reliable is the simulation of uranyl ions in aqueous solution using the GFN-FF model in xtb-6.7.1? Can calculations be performed in periodic boundary conditions?
I appreciate your expertise and look forward to your guidance on this matter.
Thank you for your time and attention.
Best regards!