Hello everyone,
I'm studding thermal properties of amorphous SiO2(a-SiO2) via molecular dynamics(LAMMPS) using Tersoff potential(Article Interatomic potential for Si–O systems using Tersoff parameterization
). But I'm experiencing a continuous temperature increasing issue in nve ensemble with a time step of 0.5fs. Problem can be solved by reducing the time step to 0.1fs. But my calculations are kinda time consuming ones so I really need to have the energy conservation with the 0.5fs time step. Anyone who experienced a similar issue with Tersoff potential (for a-SiO2)? Hope someone could give me an advice on my issue.Stay safe.
Thank you
Chamara Somarathna
Chamara Somarathna I have worked with Tersoff potential and SiO2 in NVE ensemble with even larger timestep (1 fs) with no problems. How did you prepare your system? Did you proceed via melt & quench route? Usually if the potential is correctly chosen and defined in the input script, the drift is a result of poorly equilibrated structure.
Hope this helps.
Dear Anže Hubman
First, thank you very much for your kind reply.
Back to my problem, I tried to create amorphous silica by melting and quenching method. Initially I tried to equilibrate Beta-Cristobalite at 1800K in NPT and NVE ensembles respectively. However, in the NVE ensemble, system temperature begins to increases as shown in the Fig. 1.
Then I skipped the NVE equilibration and used NPT for the melting and quenching process following http://dx.doi.org/10.1063/1.4983753. However the resultant amorphous silica also not able to handle the NVE ensemble keeping a stable temperature. My timestep size is 0.5fs and used the SiO.tersoff potential file which comes with LAMMPS.
I have attached my input script (SiO2.in) and the SiO.tersoff files herewith. I'm guessing that my initial atomic positions are the issue here since I have borrowed it from the LAMMPS mailing list. I have also attached the initial structure (SiO2.data lammps data file format) of my system herewith.
Can you please tell me about your procedure of creating amorphous silica, such as the initial crystal structure used, timestep size, whether is it ok to use the SiO.tersoff from LAMMPS or did you used a different potential file. I would be really grateful if you could help me out.
- Badges
- Science topic
- Similar topics
- Chemistry
- Theoretical Chemistry
- Computational Chemistry
- Molecular Dynamics