I want to create the molecular [Th-9H2O]4+, but it is
impossible for me to bond Th4+ ion with 9 water O ions because of the
hybridization problem.
Does Th really have 9 covalent bonds requiring angle and torsion terms?
E.g. in nature, is it possible that the waters hardly ever exchange?
If it is normal ion-water coordination, perhaps not even restraints
should be necessary.
Note that molecular mechanics are considered unable to model atomic
charges greater than +1 (less than -1) accurately. The most work on
this did involve bond-like potentials I believe - see the Vedani
reference in the vdw section of Deriving New Parameters on the
amber web. To implement Vedani's potentials, significant programming
would be required in leap and sander/gibbs, modifying the prmtop
format. Depending on the goal, I would first try modeling Th4+ in
solution and see what coordination results, then try adding restraints
if necessary. If detailed accuracy is required, I suggest quantum
methods; perhaps Roar could be used to handle this system locally
in a larger context.
Bill Ross
Received on Thu Mar 22 2001 - 09:44:16 PST
