[Prev][Next][Index][Thread]
Dielectric response - correlations: mini-bibliography (fwd)
Sender: bushb@merck.com
Subject: WSN: Dielectric response - correlations: mini-bibliography
Last week I posted a comment that linearized dielectric response of solvent
to a solute is a generalized susceptibility matrix, indexed by sites of
solute charges (dipoles...). Matrix elements could be simulated either
as relaxation free energies or as correlations of potential fluctuations.
Several labs have indeed implemented such calculations.
=== Tom Simonson (simonson@zinfandel.u-strasbg.fr) writes that
"this idea was the basis of two 1991 papers I wrote on the
microscopic dielectric properties of proteins: Biophysical Journal, 59,
670-690, and J. Mol. Biol., 218, 859-886"; further work will soon appear.
=== Ron Levy and co-workers (Mahfoud Belhadj and Doug Kitchen) at Rutgers
formulate the idea clearly in "Gaussian fluctuation formula for electrostatic
free-energy changes in solution", J.Chem.Phys. 95, 3627-3633 (1 Sept 1991),
and (with Karsten Krogh-Jespersen) apply it to "Solvent effects on the
adiabatic free energy difference between the ground and excited states
of methylindole in water", J. Phys. Chem. 95, 6756-6758 (1991).
They make the connection to integral theories (in which fluctuations are
implicit) and show how remarkably linear the response actually is; so much so
that precision is much improved by using second moments (correlations;
Gaussian exponents) to *extrapolate* the linear response, rather than e.g.
to simulate separately the relaxation response to initial and final solutes.
I'm sure that there is other work, past and present, based on the
correlation-susceptibility theorem ( Kubo, J.Phys.Soc. (Japan) 12, 570 (1957)).
There's a language barrier: most *physics* applications use plane waves
(indexed by wave-vector (k)) and frequencies -- an infinite basis
appropriate to spectroscopies. Applications to individual solute *molecules*
are better understood in terms of the charge-site-points on the solute.
Either approach has the merit of dealing with MEANINGFUL quantities (the
responses to defined perturbations). Though this seems hardly worth mention,
the biomolecular literature is unfortunately cluttered with arguments over
non-questions: "what is the dielectric constant of a protein (surface water,
water in clefts); is it the same for charge interactions or potentials, as
for dipole interactions or electric fields; if the dielectric constant is
small even around polar sidechains, how can a protein stabilize ions . ??? ..."
-- Bruce_Bush@merck.com