27. January 2020¶

To do¶

Read deeper into SAPT
Check on Ni/Pd UHF opt
- Re-opt in CPCM?
metal-Ni-UHF-PCM opt
~~wxmacmolplt as X forwarded XYZ viewer?~~ crashes :(
Check SAPT job completeness against SAPT processing notebook
Check Ni opt jobs
Check ethyne-Pd test job

What I did¶

I started CPCM optimisations for the successful UHF and RHF gas phase optimisations

I noticed that the Psi4 jobs weren’t running in their scratch directories, so I set the slurm script to cd into it before it ran, to ensure that even if they’re outputting to ./. they won’t fill my home directory and error out. (especially with Psi4 installed locally… )

I explored displaying geometry in jupyter notebooks, though this seems difficult to say the least. There are a couple of packages, though they seem to be really finicky about things like slashes in the geometry title in XYZ files. one of them that I tried seemed to imply some really unnatural geometry of BF4, so I’m not entirely sure that I trust it.

OPTIMISE WITH RFO … okay, this still doesn’t work in gaussian

I tried to optimise the solvated molecules in Psi4 but the process is incredibly slow without a lot of memory and for some reason kept crashing for me.

In ORCA, the system implodes much like in gaussian.

I fixed up some issues in the psi4 script. It now seems to be using scratch properly. I also figure out that sbatch can take memories in GB rather than MB, so more exact totals can be given more easily (using 1GB = 1024MB, rather than 1GB = 1000MB)