Moderators: Global Moderator, Moderator
Hi
I'd like to run a long NVT MD trajectory with a MLFF, using a Nose-Hoover thermostat. I am looking at Li diffusivity in LiCoO2. I would like to couple only the Co and oxygen atoms to the thermostat, but am not clear on how to accomplish this. Any guidance you have would be greatly appreciated.
Thanks
Kirk
If you don't mind using the Andersen thermostat, then you can follow the instructions for "multiple Andersen thermostats". With that method you can define up to three groups of atoms and you can have a separate Andersen thermostat act on each of them. Unfortunately multiple thermostats are currently only implemented for the Andersen thermostat.
You can also try and MD with selective dynamics with any thermostat. Please read the wiki for the POSCAR for a how to. Within selective dynamics you can freeze the motion of atoms in given directions completely.