[gmx-users] Virtual sites and diffusion coefficients.

[sujithkakkat .]

Hello Mark,

     I have tested both models in simulating a variety of properties. The
original parametrization results involved predicting the critical point of
CO2 and also density at various P and T. I did several NVT simulations to
predict density-vs-pressure data. In this case the model with virtual
interaction sites gave much better results.

     In addition to this, simulations performed were performed on pure CO2
at supercritical conditions and mean square displacement was calculated for
the CO2 molecule. When compared to the reported ab initio dynamics results
the one obtained  using virtual sites gave much better agreement.

     The density of CO2 dissolved in H2O at very dilute concentrations was
also studied using both models of CO2 and the results were in good
agreement with experimental data for the model with virtual sites. In this
case the one without virtual sites also did reasonably well.

      However, the disparity appears when considering the Measn Square
displacement of dilute CO2 solution in water. Here the model without
virtual sites is closer to the experimental data.

    This has become very confusing now. The original parametrization of he
model did not involve predicting the transport properties. Also in the
published results using the EPM2 model  I don't find any mention of the
virtual interaction sites.

Regards,

Sujith.


On Tue, Feb 3, 2015 at 1:40 PM, Mark Abraham <mark.j.abraham at gmail.com>
wrote:

> Hi,
>
> With which set of parameters are you able to reproduce the original
> parametrization results?
>
> Mark
>
> On Tue, Feb 3, 2015 at 8:04 AM, sujithkakkat . <sujithks58 at gmail.com>
> wrote:
>
> > Dear all,
> >
> >      I tried simulating the CO2-water mixture with CO2 at a low
> > concentration (0.003 mole fraction) at various temperatures at 1 atmos.
> >
> > I used the EPM2 parameters for the CO2 molecules which includes a
> flexible
> > OCO angle. I did a separate simulation on the similar system with CO2
> model
> > constructed using virtual sites (following Justin Lemkul's tutorial) ,
> with
> > the interaction parameters same as the EPM2 model.
> >
> > In both cases the water model TIP4P was used.
> >
> > I ran 5ns simulation after equilibrium was attained for the T and P
> values.
> >
> >   When plotting the mean square displacement and fitting the graph to get
> > the diffusion coefficients, what I find is that the diffusion coefficient
> > obtained in the case of the CO2 model using virtual sites is much higher
> > (approximately two times) of that obtained for the EPM2 model without the
> > virtual sites.
> >
> >   Let me know if any one came across this before. I would like to hear
> from
> > the more experienced users, what they think about this.
> >
> > I will provide further details on the topology and MD parameters if
> > necessary.
> >
> > Regards,
> >
> > Sujith.
> > --
> > Gromacs Users mailing list
> >
> > * Please search the archive at
> > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > posting!
> >
> > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >
> > * For (un)subscribe requests visit
> > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> > send a mail to gmx-users-request at gromacs.org.
> >
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>