Re: Why the Lennard−Jones parameters E for anions are usually larger than cations?

From: Axel Kohlmeyer (
Date: Mon Jun 15 2015 - 05:36:16 CDT

On Mon, Jun 15, 2015 at 5:02 AM, <> wrote:

> Hi everybody :
> I am simulating the solution systems comprising water and ions. And I note
> that in most popular classical forcefield, the Lennard−Jones parameters E
> (or \eps) for anions are usually larger than cations. And the adsorption
> energy (on metals or graphene) is also higher for anions than cations.
> Is there some different treatment between cations and anions when optimize
> the parameters? or some physical nature lies on the results?
​please note that you cannot view things in this simplified fashion. there
are multiple issue to keep in mind:
- you cannot compare the LJ well depth epsilon without also considering the
diameter sigma and the coulomb interaction. anions and cation have
different effective diameters and different polarizabilities. both need to
be represented in an average fashion by the LJ parameters.
- parameters in classical force fields are optimized for specific
environments. so ions cannot be looked at "naked", but in combination with
their solvation shell(s) of water molecules.
- ​when you talk about adsorption energies, you have to be more specific as
what it is exactly what you are computing, since people rarely look at
isolated species in vacuum with conventional classical force field. once
you have additional atom present, the adsorption process is rather complex
with multiple contributions.

​in general, i would recommend to look up the original publications
describing the parameterization of those entities and study the
justifications and references given there.


> Thank you very much.
> Table. L−J parameters E in (kcal/mol)
> Na K Cl charmm -0.0469 -0.087 -0.15 amber -0.00277 -0.000328 -0.1
> opls_aa -0.0028 -0.0003 -0.118
> Best regards,
> J. Liu

Dr. Axel Kohlmeyer
College of Science & Technology, Temple University, Philadelphia PA, USA
International Centre for Theoretical Physics, Trieste. Italy.

This archive was generated by hypermail 2.1.6 : Tue Dec 27 2016 - 23:21:10 CST