Help needed with Energy minimization using Monte Carlo for Quartz using Buck Potential

I am currently using the following Buck potential to minimize the energy of Quartz:

Buck potentials

pair_style hybrid buck(slash)coul(slash)long 11.0 lj(slash)cut 12.0
kspace_style pppm 1.0e-4
kspace_modify mesh 24 24 30 order 6
pair_coeff 1 1 buck(slash)coul(slash)long 32007.28 0.3623 4035.5
pair_coeff 1 2 buck(slash)coul(slash)long 415172.24 0.2235 3078.51
pair_coeff 2 2 lj_cut 0.0 0.0

neighbor 2.0 bin
neigh_modify every 1 delay 0 check yes

velocity all create 300.0 9999999 rot yes dist gaussian

fix 1 all nvt temp 298.15 298.15 20 drag 0.02
restart 50000 restart.*.zuizhong
timestep 0.001

thermo_style custom step temp press etotal ke pe
thermo 10
dump 1 all custom 10 dump.sio2 type x y z
run 100000

The Input file used:

LAMMPS data file

72 atoms
2 atom types

0.000000 9.832000 xlo xhi
0.000000 8.514762 ylo yhi
0.000000 10.81080 zlo zhi

Masses

1 28.085501 # Si
2 15.999400 # O

Atoms

1 1 1 2.4 2.60700 0.00000 3.60400
2 1 1 2.4 2.60700 0.00000 2.00900
3 1 1 2.4 0.14900 4.25700 3.60400
4 1 1 2.4 0.14900 4.25700 9.00900
5 1 1 2.4 7.52300 0.00000 3.60400
6 1 1 2.4 7.52300 0.00000 9.00900
7 1 1 2.4 5.06500 4.25700 3.60400
8 1 1 2.4 5.06500 4.25700 9.00900
9 1 1 2.4 -1.30300 2.25800 1.80200
10 1 1 2.4 -1.30300 2.25800 7.20700
11 1 1 2.4 -3.76100 6.51500 1.80200
12 1 1 2.4 -3.76100 6.51500 7.20700
13 1 1 2.4 3.61300 2.25800 1.80200
14 1 1 2.4 3.61300 2.25800 7.20700
15 1 1 2.4 1.15500 6.51500 1.80200…

I have modified the existing monte carlo script from the examples in LAAMPS and made it 3D from 2D. The program runs and minimizes the structure energy but when I use an already stabilized Quartz structure, I do not expect much motion to happen. Also, the result of the code brings the Si molecules really close to each other which is most likely undesirable. Any input on the Buck potential or the stabilized input to be used?

Thanks,
Sanchari

data.72siO2 (3.07 KB)

in.quartzmc (3.74 KB)

this is not really an answer to your inquiry but i noticed your input only does MC moves in x and y direction, not z direction. so you cannot really claim to have changed the script from 2d to 3d.

axel.

Thank you Axel for your response. At an earlier version I had done the MC conversion to 3D. Must have switched back accidentally.

I am attaching the new code and 72 atom data file (as the 67 atom one loses atoms). Please let me know if you have any further corrections to make.

When I moved a single Si from the data file I expect to see the same convergent result from the moved and unmoved versions, unfortunately that is not the case. I want to make sure it isnt my input file, the Buck potential or the code itself that is defective.

Thanks,
Sanchari

data.72siO2 (3.07 KB)

in.quartzmc_d (4.03 KB)

Thank you Axel for your response. At an earlier version I had done the MC conversion to 3D. Must have switched back accidentally.

I am attaching the new code and 72 atom data file (as the 67 atom one loses atoms). Please let me know if you have any further corrections to make.

When I moved a single Si from the data file I expect to see the same convergent result from the moved and unmoved versions, unfortunately that is not the case. I want to make sure it isnt my input file, the Buck potential or the code itself that is defective.

why should the system go back to the unperturbed state?
i don’t see anything in your input that should do that.
you are doing single particle MC moves alternating with a single MD step for all atoms at a given temperature. of course your system will move and there is nothing restraining atoms back to their original positions. in fact, the MC moves will disrupt the MD time integration and thus will cause a significant increase in kinetic energy. if you change your thermo output to also output the temperature, you can easily see that. the nose-hoover thermostat will try to counteract the effect of the MC moves, but this reaches a dynamic equilibrium at around 1370K. clearly, your system will not remain in a 0K configuration under such conditions.

axel.