Thanks for your advice.
Actually, I'm thinking the same way, just try to make sure.
after all, an external electric field is just a (constant)
force vector that scales with the magnitude of a particle's
charge. not much magic there and one of the things where
simulations have an advantage over experiments.
Since we are trying to study the dynamics of the system with electric
field, we have been worried about different boundary conditions may
it all depends on what kind of information
you want to extract and how sensitive that
information is to dielectric boundary conditions.
if you are after the response to a small varying
perturbation, yes, it is a concern.
give different results. but I think we'll just have to find a way
around this problem.
i would be *much* more concerned about the fact
that you have a non-equilibrium setup and how
you dissipate the injected energy as well as
how you avoid the flying ice cube syndrome. 
cheers,
axel.