On Wed, Apr 27, 2016, Lorenzo Gontrani wrote:
>
> The system is ICH2CN, and I calculated the electrostatic potential ab
> initio as usual, obtaining these point charges:
>
> ESP charges:
> 1
> 1 C -0.898387
> 2 C 0.631150
> 3 I 0.032933
> 4 H 0.353024
> 5 H 0.353045
> 6 N -0.471765
>
> When I use the ordinary RESP method, I get the following result:
> 4 I1 i M 3 2 1 1.540 111.208 -180.000 -0.113877
> 5 C1 c3 M 4 3 2 2.208 151.520 -180.000 -0.101590
> 6 H1 h1 E 5 4 3 1.090 105.533 -58.378 0.140478
> 7 H2 h1 E 5 4 3 1.090 105.521 58.407 0.140478
> 8 C2 c1 M 5 4 3 1.447 112.551 -180.000 0.346442
> 9 N1 n1 M 8 5 4 1.162 179.005 180.000 -0.411931
>
> the charge on nitrile carbon is less than half, and the iodine changes
> sign..
This is what I would expect: the point of the "R" in "RESP" is to avoid large
charges on atoms that are partially or nearly buried. This is what you are
seeing here, where C1 goes for -.89 (ESP) to -.10 (RESP), and C2 goes
from 0.63 (ESP) to 0.35 (RESP). A consequence is that I goes from being very
slightly positive to slightly negative.
Be sure to look at the literature on modeling halogen bonds with force
fields. I don't know the details, but most studies find that off-center
charges or other strategems (beyond simple atom-based fixed charges) are
needed to get reasonable results.
....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Apr 27 2016 - 05:30:04 PDT
