Hello,
I am considering the problem of liquid crystals confined by metal surfaces. May I ask if there is any method to include the polarization of the metal induced by the liquid crystals in MD simulation ? Can this be done in LAMMPS? Does anyone have experiences with this issue ?
Thank you very much,
Wei
Hello,
I am considering the problem of liquid crystals confined by metal surfaces. May I ask if there is any method to include the polarization of the metal induced by the liquid crystals in MD simulation ? Can this be done in LAMMPS? Does anyone have experiences with this issue ?
i am not aware of anything. there are some potential types
implemented in LAMMPS that support charge equilibration,
but i am not sure how well this translates for your needs.
quite a while back, we had a discussion on the list here
about how to implement a simple image charge model,
that would be the simplest way to incorporate what you
are looking for. however, this never led somewhere, since
nobody seemed to have the time to implement it, or didn't
get it done, or did get it done and didn't talk about it.
axel.
Wei,
I will dare to drop a comment here although I am far from knowing much
about the topic.
First, you should be browsing the lit first to see if a similar study
has been attempted using classical physics.
Polarization effects due to charge transfer are incorporated in force
fields such as Reaxff via the QEq method.
Reaxff is implemented in lammps but there might not be any
parametrization available for the system you intend to model. Even it
you were to find one, my guess is that it will not do very well for
this your particular problem. One of the reasons is that QEq cannot
describe the dissociation limit of molecules and fails to obtain the
correct size dependence of the molecular polarizability. See: T.
Verstraelen, V. Van Speybroeck, M. Waroquier , Journal of Chemical
Physics , 131 (4), 044127 , 2009 , 3.093
The best candidate to QEq that exist so far is the Split charge method
(SQE) but it is not implemented in Lammps and when done, it will
require to be parametrized for your system of course.
There are ways to model polarization effects that are not related to
charge transfer but to the deformation of the electronic cloud, i.e.
the shell model. I don't think this one is implemented in lammps yet
but it exists in other MD codes. Yet, for the metal problem you have
charge transfer might be the dominant mechanism.
Carlos