[gmx-users] CHelpG charges on Fe-PPIX

[Valentin Gogonea]

Dear CCLers and GMX users,

A couple of months ago I asked a question about what van der Waals 
radius should be used for Fe(2+) to obtain CHelpG charges (to be used 
with OPLS parameters). I got few replies and consequently, following 
the suggestions from the list I tried few exploratory calculations on 
Fe(2+)-porphyrin IX (without axial ligands). I am listing here the 
results obtained for different van der Waals radii in the range of 
1.00-2.00A:

CHelpG charges for Fe(2+)-PPIX
Total Charge=0, Multip=5

Radius-Fe(A)     Charge(au)
1.0		      4.226756
1.2		      3.069658
1.5		      0.943646
1.7		    -0.342193
2.0		    -1.644652

Mulliken charge is  0.318069

Radius for C,N,O is 1.50A and for H 1.45A (Breneman model as 
implemented in Gaussian98)

There is a large variability in the charges and I am wondering how to 
get a value which is acceptable. Further MD simulations will clearly 
depend on the charge values. I used UHF/LanL2DZ basis set on Fe and 
STO-3G for the atoms in the porphyrin ring.

I would really appreciate any suggestions on how to proceed in order to 
get acceptable charges for Fe.

Thank you,

Valentin Gogonea
Cleveland State University
Cleveland, Ohio, USA