On Mon, Sep 24, 2012 at 7:20 AM, Sanjib Paul <sanjib88paul.gmail.com> wrote:
> Dear Amber Users,
> For a test I calculated vibrational modes
> of a single water molecule by MMPBSA.py. There is a WAT unit in xleap
> and the inpcrd and prmtop file is generated form that. Then I run
>
By default you are using TIP3P waters, unless you've changed the water
model. See Chapter 2 section 12 for choosing water models.
> MMPBSA.py using the following input file-
> normal mode calculation
> &general
> startframe = 1,
> endframe = 1,
> /
> &nmode
> nmode_igb = 0
> /
>
> The vibrational normal modes obtained from here are
> bending = 1996.938 cm^-1
> symmetric stretch= 3735.841 cm^-1
> antisymmetric stretch = 5952.802 cm^-1
>
> But in literature I have seen that the last two frequencies in gas phase
> are
> symmetric stretch = 3657 cm^-1
> antisymmetric stretch = 3756cm^-1
>
> So, here the difference is significant.
>
This is not at all surprising, especially for TIP3P. TIP3P is a rigid
water model, and as such must be used with SHAKE (or the analytical SETTLE
variant). Therefore, there is little incentive for carefully parametrizing
the bonds to recover the experimental vibrational frequencies, since the
H-O bonds are constrained, anyway.
You could try a flexible water model and see if that gives you better
answers, but I would not expect it.
I tried it in quantum mechanical software TURBOMOLE. Here also I used
> same initial coordinates. The result is good here.
> bending = 1642.33 cm ^-1
> symmetric stretch = 3651.30 cm^-1
> antisymmetric stretch = 3754.15cm^-1
>
This would also be QM-method dependent, but I would expect much better
answers using ab-initio methods for this problem.
HTH,
Jason
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Sep 24 2012 - 05:00:03 PDT
