Keeping carbonate planar without improper dihedrals

GROMACS version:2020.5
GROMACS modification: No

Hi All,

I would like the carbonates in my simulation to remain planar. I do not have parameters for improper dihedrals. So, I was wondering, in order to achieve the planarity, would it be sufficient for me to apply constraints to the three C…O bond lengths, and the three O…O bond lengths within a carbonate unit?

As I was scanning through the old forum posts, I found some recommendations on using virtual sites to maintain planarity in nitrate ions. In order to maintain planarity in the specific case of carbonate ion, would I need to employ virtual sites in addition to the constraints I mentioned in the above paragraph? If I do need to use virtual sites, would it be sufficient to define carbon as a mass-less virtual-site whose coordinates are defined from the positions of the three oxygen atoms? Would the moment of inertia be accurate with just this definition, or would some additional steps need to be taken to ensure we obtain the correct moment of inertia?

I would really appreciate help from the community!


Tej S. Choksi

Assistant Professor
School of Chemical and Biomedical Engineering
Nanyang Technological University, Singapore
Block N1.2, B1-18, 62 Nanyang Drive
Singapore 637459, Office: +65 6316 8940

Constraints will not work.

I suggest to use an improper dihedral. The parameters are not critical, as there should be no motion. Just use values from a similar improper dihedral in another force field.