[gmx-users] Simulation of a charged system

[Mark Abraham]

On 7/12/2011 11:31 AM, Surya Prakash Tiwari wrote:
> Dear Gromacs users,
>
> I am simulating a charged system with periodic boundary conditions. My
> system has 506 water molecules and one ion.
> I am trying to calculate the free energy of an ion.
> I do not want to use any counter-ions to neutralize the system,
> because I don't have force-field between my ion and the counterion.
>
> In particular, I want to reproduce the following paper: 
> Horinek, D., Mamatkulov, S. I.,&  Netz, R. R. (2009). Rational design
> of ion force fields based on thermodynamic solvation properties. The
> Journal of chemical physics, 130(12), 124507. doi:10.1063/1.3081142
>
> Their system is also charged, has one ion and 506 water molecules.
> They are using AMBER software. On page number 7, they have discussed
> the correction terms due to ewald summation (equation 6) in a charaged
> system.
> On the same page, they further say that AMBER has implemented first
> term in eqn. 6 to account for ion’s Coulomb interaction with its
> periodic images:
> "In the PME implementation in AMBER, a self-energy correction of
> N*e^2 *xi /(8*pi*epsilon) is already accounted for."
>
> I just want to know whether, Gromacs has the same implementation in
> their PME to account for ion’s Coulomb interaction with its periodic
> images.

Sounds strange to me. I'd check the AMBER manual for what this is and 
how it works (don't think you'll find anything) and then ask the authors 
of that paper what they really mean. As Darden et al note (JCP 109 
10921) at the end of section IIB, the so-called "self term" and the 
"self potential" are different things, and this is possibly a problem.

Mark