PCFF vs PCFF_ore files

Hi all -

This is probably not a super-important question, but it’s a distinction I have been confused about as of late.

I’ve been using EMC to initialize structures for LAMMPS, where I’m doing all-atom simulations of gas transport in polymer membranes (things like H2, N2, O2, CO2, etc.). I’ve found building polymer systems with the PCFF force field to be pretty smooth, all things considered.

My trouble here started with assigning atom types for CO2 - based on the rules established for PCFF, it was unable to do so, and it seems like the atomtypes didn’t explicitly exist for it. I’ve tried using values from the LAMMPS source files for CFF91 (results were totally wrong) and COMPASS (much closer to agreement with literature).

I just noticed that, in the field folder in EMC, there is a second force field called pcff_ore. After going through the .frc file for it, it seems like this force field has the necessary atomtypes for CO2 that I’d want even though the rules to assign them are not in place. The bond length (1.16A for CO) and angle (180deg for OCO) are defined such that it seems like this is what I should use, and it’s closer to COMPASS.

My question is, what’s the difference between pcff and pcff_ore in EMC, and where exactly does the latter file come from?



Hi Sam,

The pcff files are based off the original pcff files, with some minor alterations as are documented at the top of the pcff_templates.dat file. pcff_ore does not differ much in rules from pcff. The difference lays in the addition of a set of parameters for ores to the pcff_ore.frc parameter file. These parameters stem from the work by Heinz and coworkers (e.g. https://doi.org/10.1021/la3038846). Unfortunately, these parameters are not directly available due to the absence of rules regarding these parameters in the rules file pcff_ore_templates.dat. This means, that the parameters are inaccessible through the regular typing schemes as implemented in EMC. This is the case for types as depicted in the mass section of pcff_ore.frc from line 186 and onward. I would like to point out, that metal oxide surface parameters are accessible by combining the pcff force field with metal oxide force fields such as cff/sio2 from the cff collection, which are related to the Heinz work. Normally, one sets the force field upon import of the crystal surface. This implies, that EMC applies rules from such a force field to the imported surface only. One can use pcff for the bulk.

1 Like

Hi Pieter,

Thank you for the clarification! This is very good to know.