Temperature going to zero despite using Langevin Thermostat

Dear all

I am trying to simulate the behavior of mono disperse hard spheres in an insulating medium for my class project. I am using FLD. Inspite of using Langevin thermostat the temperature goes down to zero from a starting of 300 K. I am reproducing the script below.

Here is the script

Big colloid particles and small LJ particles

units lj
atom_style sphere
dimension 3

lattice fcc 0.005
region box block 0 4 0 4 -2 2
create_box 1 box
create_atoms 1 box

set group all type/fraction 1 1 23984
set type 1 diameter 20

velocity all create 300 4928459 dist gaussian

multi neighbor and comm for efficiency

neighbor 1 multi
neigh_modify delay 0
comm_modify vel yes

pair_style lubricateU 1.0 0 10 10.6 0.01
pair_coeff * *

timestep 0.00001

fix 1 all langevin 300 300 20000 48279 omega yes
fix 2 all nve

thermo_style custom step temp epair etotal press vol
thermo 1000

dump 1 all atom 500 dump.colloid

run 100000

You are using LJ units and setting the langevin thermostat
to 300 - does not make sense. You only have 1 type
of particle, so do not use atom_style sphere. Just
set sigma = 1, as there is only one length scale in
your model.