I would like to utilize the COMPASS and PCFF force fields for my system. When using the COMPASS force field, I encountered a warning regarding the system having a charge. This warning is resolved when I set “system charge false,” allowing the generation of output files. However, I am concerned about potential implications of this assumption. Additionally, for the PCFF force field, I am experiencing issues with missing torsions for my system. These may be due to parameters not yet incorporated into the force field or because my SMILES description is not fully understandable for EMC. Could you please review my input file to provide reassurance?
A small remark about the issue with the partial charges: in page 32 of the manual it is written that system_charge checks if the total system charge equals to zero or not. The default value is true and I guess the reason is to guarantee that the resulting configuration is neutral. Setting the system_charge to false, does not perform this check and the user will not get any message (warning or error) if the resulting configuration is not neutral. To my understanding, the implication is: if you create an initial configuration and it is not neutral, then the system_charge option will warn you. if you keep the same settings and assign false to system_charge then the initial configuration will still not be neutral but this time you will not get any warning or error message. It is more of masking a problem than resolving. You can also try to simulate the system with LAMMPS and most likely you will also get a warning that the system is not neutral. Unless it is a special case, the system is expected to be always neutral.
Evangelos is correct with respect to setting system_charge to false. It suppresses the check, but does not neutralize the system. It allows you to at least check the resulting build to see what corrective action you would have to take. In this case, your carbon in the *OC(=O)N(H)* moiety is erroneously typed as a carbon from a deprotonated acid group, which is the result of how the typing rules are defined (as is stated in ‘Trying to get polyurethane initial structure’).
It appears to me – by manually interpreting typing rules in ./field/pcff/pcff_templates.dat, that PCFF typing rules do not include the carbon in the *OC(=O)N(H)* moiety, thus not supporting typing urethane groups. This is also the case for the compass force field. You will have to add the necessary rules and parameters, if you want to use either of these class 2 force fields. Parameters are derived by using proper quantum chemical procedures.
You can find out how rules are applied or fail by adding field_debug reduced or field_debug full to the ITEM OPTIONS section of your EMC Setup .esh.
