Generating framework.def and labeling cif by atom type

[shu815]

Hello!

I have a question about utilizing flexible frameworks in a GCMC simulation in Raspa:

Is there an easy way to generate a cif which is labeled by atom types such that each atom type can be uniquely defined in bonds/angles/dihedrals/impropers a framework.def file? I hope to write my own framework.def file for UFF atom types, I see that there are some example framework.def files in the examples directory in the RASPA install, but they don't seem to have all the types in UFF e.g. C atoms in rings C_R.

One approach I was thinking of using was taking a cif -> lammps datafile generated from the lammps_interface from Boyd et al. (https://github.com/peteboyd/lammps_interface) since it determines the hybridizations and identifies atom types corresponding to UFF, and then using that to re-generate a cif file which has unique atom names (e.g. C_3, C_R, C_R, C_R...  from the original cif of just C, C, C, C...), from which I can then write the framework.def file of every unique grouping for bonds/angles/dihedrals/impropers.

Is there a better or more conventional way to do this than going through the hassle of using lammps_interface to do this?

Thanks in advance!

[David Dubbeldam]

It is very customary to write scripts precisely for the reason you mention. Different software use different input, and you often need to convert.
It pays off to write scripts to take care of that.