Multiwfn official website: http://sobereva.com/multiwfn. Multiwfn forum in Chinese: http://bbs.keinsci.com/wfn
You are not logged in.
When I try to export a fchk file to .47, multiwfn complains about the fchk file containing spherical gaussians. So, I set the variable to force conversion of spherical to cartesian gaussians in the ini file. Reboot multiwfn and the export succeeds.
Then, when I use the .47 file as input to NBO7, I get this message:
Cartesian contaminants detected in the AO basis set; 428 functions discarded
So, why is it necessary for multiwfn to convert spherical to cartesian when NBO just deletes the cartesian contaminants? How does this conversion and then deletion affect the NBO results?
I'm using multiwfn 3.8 dev no gui version running on RHEL 8.
Last edited by jimkress (2022-10-04 22:40:35)
Offline
NBO doesn't use normal definition of spherical basis functions, which is also not supported by Multiwfn. Therefore, Multiwfn can only export .47 file based on Cartesian basis functions.
The prompt "Cartesian contaminants detected in the AO basis set" generally is not harmful, which should (I think) indicate that some redundant and useless basis functions corresponding to linear combination of Cartesian basis functions are removed for better numerical stability and reducing computational cost.
Offline
Thanks for the reply.
Interesting to note that I see this in the Q-Chem calculations I use for these fchk files:
Smallest overlap matrix eigenvalue = 3.19E-04
It would appear that there is no elimination of basis functions due to redundancy issues. This is not surprising since I use def2-SVP basis that doesn't have really diffuse functions.
Anyway, the NBO results look OK so I'll just continue to ignore the message.
Offline