A question about CG-MD of graphene by lammps

hello everyone

while i want to follow the coarse-grained modeling of graphene, i knew Ruiz had developed one way to model the CG of graphene, but the modeling is hard to develop. So one Chinese professor developed another way which used the tersoff potential with the Lj potential. (https://www.researchgate.net/publication/312061386_Coarse-grained_molecular_dynamics_model_and_mechanical_properties_of_multi-layer_graphene)

from this paper, it said the bonded interaction is used tersoff potential, and non bonded interactions is used LJ potential. but i cannot minimize the model correctly, so i want to ask some for help?

here is my part script and tersoff potential:

units metal
atom_style full
boundary p p s

neighbor 0.3 bin
neigh_modify delay 0 every 1 check yes

read restart file

read_data 1.pov
#set pair potentials

pair_style hybrid tersoff lj/cut 12
pair_coeff * * tersoff C.tersoff C
pair_coeff * * lj/cut 0.02292 3.469

mass * 48.042800


min_style cg

minimize 1.0e-4 1.0e-6 1000 10000

velocity all create 300 13450 dist gaussian

create equlibrium process

fix 1 all nve

fix 2 all momentum 1 linear 1 1 1

# output of thermal dynamics

thermo 1000
thermo_style custom step pe ke etotal press temp vol

dump 1 all atom 1000 dump.tersoff_cg.lammpstrj
log log.tersoff_cg


timestep 0.001
run 10000