[lammps-users] A problem for the 'pair_style hybrid/overlay'

Dear all:
  I am trying to simulate a system with one layer graphite on the silicon. The potential for the silicon and graphite is Tersoff potential, but I want to add LJ potential between the silicon and the graphite. So I have to use the 'pair_style hybrid/overlay' option. The input file is as follow:

### 3d Carbon nanotube thermal conductivity simulation using fix heat command.

units metal
boundary s s s
### read data
atom_style atomic
read_data silicon_graphite

pair_style hybrid/overlay tersoff lj/cut 10.0
pair_coeff 1 2 lj/cut 0.0468 3.377
pair_coeff * * SiC.tersoff Si C

mass 1 28.09
mass 2 12.011

dump mydump all xyz 200 silicon_graphite.xyz
thermo_style custom step temp pe ke etotal
thermo 200
minimize 1e-9 1e-9 50000 200000

But when I run the lammps, it tell me 'ERROR: Pair coeff for hybrid has invalid style'. So if that's to say I cann't use Tersoff and LJ with the 'pair_style hybrid/overlay' option. How can I simulate my system?
Another problem is for the Tersoff 'pair_coeff' option: when I use 'pair_coeff 1 1 Si.tersoff Si', it tell me 'ERROR: Incorrect args for pair coefficients'. Why I cann't use this method to define the pair_coeff for the Tersoff potential?
Any help would be appreciated.

pair_coeff 1 2 lj/cut 0.0468 3.377
pair_coeff * * SiC.tersoff Si C

The 2nd command needs * * tersoff SiC.tersoff ...

Your model seems odd. You will get Tersoff
interactions between Si and graphite, and then
add LJ to that. Probably not what you want. Unless
you set the Si-C interactions to 0 in your Tersoff potential
file.

Steve