droplet accelerating via external electric field in gas atomsphere and gas's temperatue tends to zero

Hi all,
I want to simulate such problem: a droplet is accelerated in gas atomsphere via external e-field, and i get some trouble now.
I found lammps has got the command that “fix efield”, so after i equilibrized the system using command “fix all nvt”, i added an external electric field reads " fix kick all efield 0.01 0.0 0.0 " along x axis, and then ran for another period.

The input script is like this:

The result showed in VMD indicated that the droplet can be accelerated to very high speed, but the gas molecules moves slowly and slowly
as time elapsed and finally became static.
I guess that as i use the “fix all nvt” command the system will be fixed to the target temperature even when the external e-field accelerates the droplet and its temperature rises, and the fraction of gas-phase-temperature will be diminished. This conflict finally results in gas molecules’ static state. But i don’t konw how to modify the script to fix the problem. Can you give me some advices to improve it? Thanks a lot!

Don't include the droplet atoms in your thermostat.
As it goes to high velocity the other atoms have to
be cooler to achieve an overall low temperature,
if they are all in the same thermostat. You can use
multiple thermostats on different groups of atoms
if you like.

Steve