[gmx-users] bond stretching force constant from QM data

[Valentin Gogonea]

Dear CCLers and GMX users,

I am trying to estimate the force constant for Fe-S bond stretching 
from QM calculations and use it with Gromos87 force field (gromacs). I 
carried out UB3LYP/6-31+G** calculations for Fe-SCH3 
(charge=+1,multip=5) and Fe-Imidazole (+2,5) with G98 and I got the 
following force constants (I used a conversion factor from mDyne/A to 
KJ/(mol nm2) of 60230):
		mDyne/A		kJ/(mol nm2)		Gromos87
Fe-S		1.839		110763
Fe-N		0.757		45594			418400


1) Is it acceptable to scale properly (using the QM and Gromos Fe-N 
force constants) the Fe-S force constant  to be in the same range as 
the Fe-N force constant in the Gromos87 force field. Did anybody use 
this kind of approach or there is a better way to estimate and adjust 
the QM determined force constant to a force field force constant?

2) For an angle bending the force constant required in the forcefield 
has units of kJ/(mol rad2), while the one obtained from QM calculations 
has units of kJ/(mol nm2). I believe that the former can be obtained 
from the latter if one assumes that the oscillator length for this 
particular normal mode is the distance between the endpoints of the 
angle. This should be valid for a torsion angle too. Is this correct?

I would appreciate any insight that I can get from the list and I thank 
you for your help.

Valentin Gogonea
Department of Chemistry
Cleveland State University