Silicon Oxide Structure

Dear LAMMPS users,

I am trying to create Silicon Oxide (Si02) structure in LAMMPS but I am encountering some difficulties. Could anyone please help on how to accomplish that using “lattice” command or any other appropriate command in LAMMPS? I can also be directed to anywhere where similar problem has been addressed. I have navigated through this website but I can’t seem to find anything related.


This is more of a question for the Science Talk section of the website. I can share some general advice, as I have used LAMMPS to model crystalline and amorphous SiO2, check DOI: 10.1021/la400857s.

  • If you want to model a crystal structure, you should start by converting the CIF file to a P1 symmetry group (e.g. no symmetry operations in your unit cell and all atoms explicitly defined), then create a valid DATA file that can be replicated to the desired size.
  • For an amorphous structure, you can melt the crystal to high temperature, equilibrate, and then quench the liquid to a glassy state. Details on the protocol in the paper I have mentioned.
  • You can use only a pairwise potential without bonded interactions in both cases. If you use Buckingham + Coulomb, the simulation may experience lost atoms during melting. This can happen as there is a chance to overcome the energy barrier of the Buckingham potential. In this case, either use conservative settings (small timestep, neighbour list updated at every step, etc.) or replace the Buckingham potential with a Lennard-Jones one, which has a hard barrier at low distances.

EDIT: when I wrote that article, I wasn’t so proficient with LAMMPS. You can also create a decent amorphous sample from an initial random structure with the right stoichiometry, followed by minimisation and annealing at high temperature.

What difficulties?

You first need to tell us what you have tried and how it failed and what exact kind of structure you want to build.

Bottom line, the more you tell us, the better we can help. So far, you don’t tell us much, so there is only a limited amount of help you can get. If you look up the crystal structure, it should be straightforward to set this up. The documentation of the lattice command discusses the setup of a Wurtzite crystal structure. That should give you a blueprint. Also, you could consider using external builder tools (e.g. Atomsk) to create a data file and import that.

I seriously doubt that. I recall quite a few discussions on that subject. Make sure you also search for “silica” instead of SiO2 (and not Si02) or Silicon Oxide.