Hard to assign differing charges to the same elements importing a CIF file

[bc118]

It appears for the general Zeolite case, we will have to find a creative way to assign the proper charges, as there are no bonds for reference.
For GOMC, the old way is to look "in order via atom names" and use .inp files. However, MoSDeF assigns per the bonds and atoms around it, and the standard method to do adsorption is not to use bonds. The only distinguishing feature in the CIF file between the same elements is the Atom name, which is unique for every atom.
This effects any case where the element does not have the same charge, but the same charge for any given element will not be an issue.

Has anyone else had this issue? If so, was it resolved and how?

Dr. Potoff suggested this as a possible method to calculate the charges for each system, so they can be done in a large-scale screening process. The paper below assigns charges to MOFs using a ML algorithm. This may be a great way to incorporate the large-scale screening of MOFs in MoSDeF. We could just use the LJ parameters from the FF XML and use these MOF charge assignments.

Fast and Accurate Machine Learning Strategy for Calculating Partial Atomic Charges in Metal–Organic Frameworks:
https://pubs.acs.org/doi/pdf/10.1021/acs.jctc.0c01229

• This paper notes that the charges are strongly dependant on the adsorbed molecules, especially if they have a dipole or quadrupole.

https://github.com/snurr-group/pacmof

There are 2 examples attached below:

• Zeolite with Water Example which works: This works as there is only 1 possible option for the charges, given the element type. However, this is likely a very small likelihood for the majority of zeolites

• Zeolite with Water Example which does not work: This fails as there are many possible options for the charges, given the same element type. This is most likely the case for most zeolites

Zeolite_MOSDEF_example_Rev_0.zip

Example of CIF File with charges:

cif_file_w_charges.zip