Can't get rid of imaginary frequency

[mstcomp]

Hello everyone. I couldn’t get rid of an imaginary frequency of a molecule (b36). Psi4 gave an imaginary frequencies consistently but Gaussian09 did not give any imaginary frequencies at this level of theory with same structure. What could be the reason?

I couldn’t get the intensities of the vibrations. Is there a way to get them with Psi4?

b36_in.txt (2.8 KB)
b36_out.txt (174.4 KB) (only last part of displacement because it’s a big file)

[jmisiewicz]

There are two unrelated questions here.

1. From the fragment of your output file you’ve provided, something wrong is happening during post-projection. Please provide your full output file if at all possible.
2. Psi4 1.5 does not compute intensities for vibrational frequencies computed by finite difference. We hope to add this feature in the Nov./Dec. 2022 release, Psi4 1.7.

[mstcomp]

Thank you for your effort. I uploaded it to google drive. Actually it’s not a big file. Forum allows max 4 mb, and I’m waiting impatiently for new abilities of psi4!

https://drive.google.com/file/d/1s-Q6m98DryKM0RX2rJNONbZ4eilJ43zj/view?usp=sharing

[hokru]

The gaussian default grid is much tighter than our default.
I recommend these settings, besides different pruning it’s the same as UltraFine, the new-ish gaussian default.

set dft_radial_points 99
set dft_spherical_points 590
set dft_pruning_scheme robust