The temperature of adsorbate in fix gcmc full to zero

Hi all,
I am simulating the adsorption of methane in silica nanopore (300K,10atm)using fix gcmc. The diameter of pore is about 1nm, and the surface of the pore are hydroxyl fully. When I perform the adsorption, I consider the ineraction of Si-CH4 and O-CH4 only. But, in the result, I found the temperature of the methane group is about 300K initially, but it fall to zero suddenly during the simulation. I don’t know whether this result is normal. If I ignore the temperature fluctuation, can I get the correct adsorption amount of methane in pore? How can I settle this problem? I have attempted many ways, modifying the tfac_insert keyword, using “fix nve/limit + fix langevin”, etc. But this problem still exist. So, could you give me some suggestions? I attach my input script and part of my outfile (I killed the simulation when I found the temperature fall to zero.).
Looking forward to your response. Thanks to your help.


in.pore1nm2sorp (2.59 KB)

pore1nm2sorp.out (6.71 KB)

You are running a version of LAMMPS from 16 March 2018, so I assume you have not incorporated the changes to the examples in examples/gcmc that we made in April 2018. The Ndof for temperature computes are now updated dynamically. You should do the same in your script.