Metal Organic Framework (MOF)

LAMMPS LAMMPS Beginners
,
1

I am trying to carry out pervaporation MD simulations for MOF on LAMMPS (or gromacs) (Zr, O, C, and H). I am facing a lot of issues:

  1. I have the forcefield parameters: (bonds, angles, dihedrals, impropers: coefficients of all these)
  2. I have the .cif file of the MOF, which has labels for each atoms (distinguished based on hybridization: C1, C2, C3, O1 and O2). When i convert this to an xyz file or pdb file or any format to that matter, it will lose this info and just retain the C, H, O, Zr, making it difficult to map which carbon is what hybrid.
  3. Also, I tried bond, angle and dihedral guessing using VMD topotools, but it fails to guess bonds between metal and other atoms, which i cannot do manually because the MOF is about 1700 atoms. Other tools like openbabel, etc also fail to do so.
  4. I tried with starting the MOF with an asymmetric unit and assign these parameters manually (this asymmetric unit has 9 atoms and i hoped that i can apply symmetry operations using a python code). But after I apply this symmetry operations, i fail to get the desired MOF.
  5. I tried fftool by taking the xyz file and ff file that i organised myself to make the system. But the issue is that the Zr is not recognised by fftool. I renamed it as Zn to see if it can at least generate a data.lammps file, where I can manually rename Zn as Zr and change the atomic mass etc. But that also failed and i got this error:
    (base) user@user-Vostro-3268:~/fftool$ python3 fftool -b 41.4,41.4,20.7 -m a -l -a -p xy 1728 supercell.xyz
    density 80.878 mol/L volume 35479.0 A^3
    molecule_file species nmol force_field natom nbond source charge
    warning: no force field for /home/user/fftool/ZnMOF.ff
    Because, the atom naming and multiple types is not being compatible…

idk what to do now please advise

2

From what I can see, this is not a LAMMPS issue, it’s an input preparation issue, and this forum focuses on LAMMPS itself rather than how to generate force field files for arbitrary systems. Preparing correct input files is part of our job as LAMMPS users; writing your own script to parse the CIF and assign atom types accordingly is sometimes the best solution.

Simon

2 Likes