It happens that the Aqvist parameters are somewhat arbitrary -
for each ion in water, there is a range of e/R* values that will
yield the same radial distribution and solvation free energies
in, say, TIP3 water. It is possible that the divergence of some
values from those yielded by the ff combining rules conversion
formula is not significant. Quite possibly the divergent values
are in error as you indicate, in any case.
Much more significant is the fact that the TIP3 and SPC water
parameters were derived for water-water interactions, not
interactions of waters with other molecules. The relevant
detail is that the H's have no vdw while the O has an expanded vdw
to contain them. Thus cation parameters developed for these water
models have smaller vdw in order to get the correct solvent
radial distribution around the ion. If you embed those ions in
molecules parameterized otherwise you must pray for some unknown
cancellation of errors. :-)
You can find a longer discussion in JACS 116, 6070 (1994), including
a simulation-based formula for determining cation (in water) R* given
e=.1 and a known first peak of radial distribution.
Finally, it is considered that multivalent ions are pushing the limits
of molecular mechanical methods, which may trump all other considerations.
Bill
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Fri Apr 22 2005 - 23:53:00 PDT
