Do potential energy functions in MD forcefields have a sense of mathematical identity with their force constants between two atoms A and B?
I am using an Amber Force Field (so that is ....k(r-r_eq)^2 as the potential energy contribution from bond stretch) and I am trying to derive force constants from a QM hessian matrix (the second order of the energy landscape relative to coordinate displacement of each at).
If I calculate the bond stretch force constant for A-B, should it be the same as the force constant for B-A? I don't think it should be the same.
In A-B, atom A feels the force from the displacement of atom B. In B-A, it is the other way round.
I have tested my code out, and using HF/6-31G* on O=O I get 1914.71 kcal/mol.ang^2 for A-B and for B-A I get the same.
However, when I test something like NCH, when hydrogen is seen as the displacing atom, the value is 1026.32 kcal(mol.ang^2), but the other way around (C is seen is being displaced) then the value is 406.35. This makes perfect sense to me, but my question is a simple one:
When adding the force constants into the amber frcmod file, which value should I use? A-B or B-A? I've established that they are not the same.
Thanks
Anthony
- Badges
- Science topic
- Similar topics
- Mathematics