On Mon, Sep 05, 2016, BLEY Michael wrote:
>
> Currently, I am working on aqueous electrolyte solutions involving
> chloride at different concentrations using the POL3 water model with
> AMBER 14.
> However, the output file shows me rather high temperatures (dependent
> on the concentration: more than 10 K) in the last line about the Dipole
> convergence and temperature like in the following example:
>
> Dipole convergence: rms = 0.795E-01 temperature = 4.95
It's been a long time since I've run a polarizable simulation in Amber
using the extended Lagrangian idea to propagate the dipoles. The
effective temperatures you will get depend (I think a lot) on the
thermostating parameters that are being used. Having dipole temperatures
of a few degrees was generally considered OK in the old days, but you are
certainly some way away from the Born-Oppenheimer limit where the dipoles
are converged at every step. I suspect that you would have to compare
properties for your system with and without treating the dipoles as extra
variables, in order to get a good idea of the practical limits of this
scheme.
[Note: as polarizable potential development in Amber has languished, so
have code developments to accelerate the polarization SCF calculation. So,
running a Born-Oppenheimer calculation with sander may be pretty slow.]
> For understanding this problem, I would like to know how this output
> value is calculated within the AMBER/SANDER package.
Look for the "diptemp" variable in ew_force.F90. The dipoles are assigned
a mass of 0.33 AMU.
> As far as I understand this temperature is proportional to the chosen
> atomic/ionic polarizability for the chloride,
I don't think this is true: the "temperature" of the dipoles is a complex
function of the polarizabilities of all polarizable atoms (most of which are
water) the the thermostating scheme you are using.
....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Sep 06 2016 - 07:00:04 PDT
