When I run LAMMPS with ReaxFF force field, I wanted to divide the model into several groups and set each group to charge-neutrality.
But I found that when solving qeq equations, LAMMPS can only apply charge-neutrality constrain to the whole model, rather than to several groups in the model.
So, I would like to ask our developer if it is possible to add additional constraints to allow to constrain each group to charge-neutrality? （Maybe in the further version of LAMMPS)
At the moment I am not aware of anybody that is planning to implement such a feature. Please note that most of LAMMPS was not written by the core LAMMPS developers but contributed by people like you that needed a specific feature and implemented it and then contributed their implementation for inclusion into the distribution.
Something like this may be possible in a few months’ time as a side effect of ongoing development of the ELECTRODE package. The short range shielding used in constant potential molecular dynamics (Gaussian charges) is quite different from ReaxFF’s QEq (which I don’t follow).
But it is not a main priority for us as it doesn’t have any obvious physical motivation (see below), and so I cannot promise any timeline.
In any case, I am not sure this is a physically appropriate model, since the underlying electrostatic meaning of this is that each group in your simulation is attached to a separate charge reservoir which keeps it electroneutral and the potential differences between those reservoirs adjust instantaneously as needed to obtain that effect.