[Prev][Next][Index][Thread]
Re: WSN: Difference in Dielectric property of water (fwd)
Sender: Michael A. Lee <lee@nematic.kent.edu>
Subject: Re: WSN: Difference in Dielectric property of water
Hello water list readers,
I am going to append a recent question and comments to this mail, so it
will be a bit long. You may want to read the question/comments
before you read what I have to say. (Of course, deleting all is an option.)
This mail was written before I saw the mail from Dongqing, Wei (attached too),
so it is in many respects supportive of that point, but that
point was made with useful looking references attached.
In the comment/question below, the individuals do not specifically
say what they mean by dielectric constant, but I assume that
are talking about a response function in some statistical average.
Perhaps they will be willing to expand further and put us the correct
track. In any case, I assume we are interested in either
a) approximating water near a large molecule in a molecular
dynamics type calculations or
b) predicting light scattering or other spatially averaged quantities
that one might get our of some ensemble average by some Monte
Carlo technique (or other theory).
The crux of my comment is that we all know the dielectric properties
are a function of omega, but we are so used to bulk
properties (as the answerer below comments) that we forget that
it is also a function of k (2 pi/lambda). The response of a finite
system (averaged over time) will depend on k and omega. So,
one might ask for the dielectric response of a large molecule in
water (as measure by light scattering, so averaged over many
configurations and orientations of the surrounding water).
In that case, the appropriate way of modeling it is to
made a position dependent dielectric function or more
appropriately, a wave vector dependent dielectric function.
If one is trying to do MD and not account for the specific water
molecules, then you are approximating a system of many degrees of
freedom by fewer degrees (i.e. sans water). Since Mother Nature
really has all these degrees of freedom, there is no
rigorous way to do this, so one searches for an adequate
approximation. (You can't fool Mother Nature.) The appropriate
approximation must depend on the question being asked
and the measurement being "simulated". If it is a protein
conformation one wants, then some approximation to the
time averaged water molecule response will likely be adequate
since the time scales for water are much faster than proteins.
This is again basically the spatial average, zero frequency,
response function.
Forgive me for adding the obvious, but the question of why
water has a different response to an electric field near a large
molecule than it does in the bulk, is because it is acted
on by forces that are different from those
of a bunch of surrounding water molecules. Proteins or other
large molecules 1) occupy space that that would otherwise
be filled with water 2) have their own charge distributions
and hence electric fields that affect nearby water orientation and
let water present a different molecular response 3) waters in
a modified configuration around a molecule affect each other too.
My comment may be general, but I hope that it eeks towards
a more specific question. There is for certain, (as the commentor
below suggests) no universal fudge factor that always give accurate
account of degrees of freedom that are omitted from a theory/simulation.
I would hope to hear from others on this matter, as I am likely being
too general to address specific research questions of specific research problems.
I could learn something about the research questions by hearing about
the specific problems where these questions are raised.
Regards,
Mike Lee,
Physics
Kent State
________________Attachment______________________________
From: bear@ellington.pharm.Arizona.EDU (Soaring Bear)
To: Multiple recipients of list <water@gibbs.oit.unc.edu>
Subject: Re: WSN: Difference in Dielectric property of water
>Dr. Mrigank asks:
>I want to get information/ideas on the the difference in the dielectric
>properties of water in different place in in biological cell.
>Like Buried in protein, surrounding protein, hydration spine of DNA, and
>what are the reasons for this difference.
You're hardly the first to wonder. I've been interested
in this problem in trying to model oligonucleotides with the
correct dielectric in the computer runs.
The fundamental problem is that dielectric is a bulk
property and trying to think local about bulk doesn't work
very well. Some theoreticians have tried different approaches
but it is an unsolved problem (perhaps unsolvable?).
Good Luck,
Bear
_______________________________Mail from Dongqing, Wei
Date: Sat, 19 Feb 94 15:59:58 -0500
From: dongqing@CERCA.UMontreal.CA
Subject: WSN: dielectric function
Hello there:
There have been someworks on the k-dependent or spatial dependent dielectric
functions in the context of solvation dynamics of ions and dipoles in polar
solvents. Foundamentally one is interested in the response of the solvent
polarization to a infinitely small by spatially varying field. This is
different from the concepts of dielectric constants where a uniform field is
applied.
Here is some references if anyone is interested.
D.Q. Wei and G.N. Patey, "Dynamics of molecular liquids: a comparison of
different theories with application to wave vector dependent dielectric
relaxation and ion solvation", J. Chem. Phys. 93, 1399 (1990).
P. Attard, D.Q. Wei and G.N. Patey, "Critical comments on the nonlocal
dielectric function employed in recent theories of the hydration force",
Chem. Phys. Lett., 172, 69 (1990).
A. Chandra, D.Q. Wei and G.N. Patey, "Microscopic theory of solvation dynamic
in polar liquids", J. Chem. Phys., Oct., 1st issue, 1993.
Hope this is helpful.
Regards. Dongqing