[lammps-users] Image charges

Hello everyone,

Does anyone here know of a way to include this in simulations of
metal/liquid interfacial phenomena? Has anyone successfully performed
a simulation including this using the pair_hybrid option?

Thanks!

Hello everyone,

Does anyone here know of a way to include this in simulations of
metal/liquid interfacial phenomena? Has anyone successfully performed
a simulation including this using the pair_hybrid option?

camilo,

i've been using an MD code with image charges for studying
water films on metal surfaces as an undergraduate student,
a very long time ago. that was a homegrown code, though.
it had a true 2d-ewald summation, which turned out to be
quite important. a paper containing all those details is:
e. spohr, j.chem.phys 107 (1997), pp6342-6348.

i've never seen it mentioned here, but the way i would try
to set this up with lammps is to set up a system in the normal
way. then you have to define your image plane, take all the
charged particles and replicate them with opposite charge
at the mirrored location. now i would use fix nvt or nve to
only propagate the real particles and write a fix to keep
the image pseudo particles at their corresponding image locations.

we were using a simple fixed corrugation potential to mimic
the roughness of the surface, i assume you want to use pair
hybrid to have the metal represented by eam or alike. in that
case you'd probably also have to parameterize the repulsion
between the metal and the adsorbed liquid (e.g. a repulsive
only morse or lj) for the interaction between the real metal
particles and the liquid atoms/molecules.

hope this helps.

cheers,
  axel.

Please explain how an image charge needs to
be treated.

Steve