Skip to content

Tuesday, 7. January 2020

To do

  • Look in to packmol/genmer(from molclus - also looks interesting)
  • Look into qcp
  • Get job submissions working
  • Submit all the optimisations
  • Read up on SAPT
  • Figure out why SSH boots me when I try to transfer files

What I did

Tried to find a nice way to convert between file formats. openbabel worked… eventually, but the version on monarch was a bit old and couldn’t seem to handle outputting to .xyz files. I found another utility called newzmat that comes with gaussian which can convert any gaussian file into a few different formats, including xyz, mol, pdb. I then wrote a script to iterate this through a string of arguments.

Yesterday I wrote a script to convert all my gjf scripts into slm scripts. it pulls the processor count and memory requirements out of the gjf and builds the slm script that way. Memory is mem+1 to account for conversion limitations and overhead costs.

I checked out genmer for generating initial structures and it seems to be interesting. Molclus also looks vaguely promising as a way to search for the best starting geometries, however I need to write a slurm script to be able to run it, since it calls gaussian/orca/mopac

10:30pm

I finally managed to get openbabel compiled and wrote a runscript for it. it requires zlib and boost to be loaded, so all commands I issue either load and unload these automatically or run through a script which does it for me.

I forgot to optimise the aromatic base molecules, so I just put them on to run tow.

Optimised structures

Alkyl:

  • Ethane
  • Ethene
  • Ethyne

Aromatic:

  • Benzene
  • Furan
  • Pyrrole

Cations:

  • Tetramethylammonium
  • Tetramethylphosphanium
  • 1,1-dimethylpyrrolidin-1-ium
  • 1,3-dimethyl-2H-imidazole

Anions:

  • \(\ce{BF4-}\)
  • \(\ce{BF2(CF3)2}\)
  • \(\ce{C(CN)3-}\)
  • \(\ce{N(CN)2-}\)
  • \(\ce{N(SOOCF3)2-}\)