For the time being, I plan to use crystalline silica surface for protein detection method in molecular dynamics simulation study, but I still couldn't find any suitable method to build it.
You can just take a crystal unit cell and replicate it by writing a script, or even simply in vmd.
Attached are a file with a unit cell an a small replication of it. Note that these files only contain types and positions, meaning that you will have to add bond information if you want to use a flexible silica force field (as opposed to freezing the atom positions).