Design structure of FCC lattice in VESTA (Example Gallium arsenide)
Convert file to pdf and then use Atomsk software to convert that in lammps file
The problem is I create the bonds in VESTA but they disappear when I convert to pdf file.
When I try to create the bond within lammps itself using create_bonds command, I get the following error: ERROR: Cannot use create_bonds with non-molecular system (src/create_bonds.cpp:50)
That is very unlikely. PDF is not a file format for molecular simulations.
The why use a file format that does not allow storing the bond topology? That renders using the VESTA software are rather useless for your purpose.
The Atomsk documentation clearly states that it only supports atomic system, so that cannot be a good choice for your purpose.
You cannot create bonds, if your atom style does not support storing bonds.
You need to use a different atom style (e.g. bond or molecular or full) and a different software to create the atoms (e.g. LAMMPS plus create_bonds to create atoms and bonds) or a different protocol, e.g. if VESTA cannot create LAMMPS data files directly, export your system in a file format that does support including the bond information (you may need to export 2 files like for NAMD or CHARMM where you have to have a .psf file for the topology and a .coord or .pdb file for the coordinates) and then use a tool that can read and write both all required file types for convert to the LAMMPS data file.
Of course, then you have the challenge of finding suitable force field settings and parameters. In the LAMMPS distribution, for example, is a potential for GaAs, but that will give wrong results with explicit bonds, since it is a manybody potential where the bonds (or rather the bond order) are computed internally as part of the pair style.