Molden inp files not showing symmetry

Find answers or ask questions regarding Dalton calculations.
Please upload an output file showing the problem, if applicable.
(It is not necessary to upload input files, they can be found in the output file.)

Post Reply
rlt92
Posts: 15
Joined: 20 Nov 2018, 12:29
First name(s): Richard
Last name(s): Thurston
Affiliation: UC Berkeley
Country: United States

Molden inp files not showing symmetry

Post by rlt92 » 03 Oct 2020, 07:03

Hi, I've been working on a wave function calculation on nitrobenzene with explicit C2v symmetry and to help ensure I'm setting up the calculation correctly I've been visualizing the orbitals using the output molden inp file. In comparison with literature orbitals I was expecting to see orbitals with C2v symmetry however the orbitals I computed appear to not have any symmetry. Is there an issue with how I'm setting up my calculation that's producing these oddly shaped orbitals, or could it be that there is an issue with how I'm visualizing the orbitals, or maybe I have a misunderstanding of how dalton is computing these orbitals and this is the expected output? Any advice would be appreciated.

I tried this calculation with a couple of different AO bases and with different set ups of the calculation. Attached are the output files and the associated molden inp files from a couple of my trials.
Attachments
state-1_sym-1_freqd-0.0_freqp-0.0_spin-1_CN_disp-0.0_ONO_rot-0.0_NBopt_dunningZ-2_molden.inp
(943.92 KiB) Downloaded 15 times
state-1_sym-1_freqd-0.0_freqp-0.0_spin-1_CN_disp-0.0_ONO_rot-0.0_NBopt_dunningZ-2.out
(144.5 KiB) Downloaded 15 times
nb_ground_dipole_hf_pol_dunning_molden.inp
(418.13 KiB) Downloaded 14 times
nb_ground_dipole_hf_pol_dunning.out
(86.66 KiB) Downloaded 12 times

taylor
Posts: 599
Joined: 15 Oct 2013, 05:37
First name(s): Peter
Middle name(s): Robert
Last name(s): Taylor
Affiliation: Tianjin University
Country: China

Re: Molden inp files not showing symmetry

Post by taylor » 03 Oct 2020, 18:24

I downloaded your molden.inp file for the ground dipole moment and as far as I can see the MOs display perfect C2v symmetry (as I would expect because I don't recall we have ever had a problem with this). I looked at the HOMO, which is labelled as A2 and indeed has a2 symmetry. The HOMO-1 is labelled B2 (this is not the Mulliken convention and most people would label it b1) and again has the correct symmetry. It is unfortunate that we have limits on what sort of file can be uploaded because I would really like to see a screen scrape or an exported graphic of the Molden visualization that shows an example of an orbital that does not have the correct symmetry. Perhaps you can PM me attaching such a picture?

Best regards
Pete

hjaaj
Posts: 394
Joined: 27 Jun 2013, 18:44
First name(s): Hans Jørgen
Middle name(s): Aagaard
Last name(s): Jensen
Affiliation: Universith of Southern Denmark
Country: Denmark

Re: Molden inp files not showing symmetry

Post by hjaaj » 04 Oct 2020, 18:22

A comment on the MCSCF calculation: the active space in symmetries 3 and 4 are fine, but in symmetries 1 and 2 you have only active orbital in each, which means it is a waist of computer resources. This is because there are no correlating orbitals for them, as you can see in the converged MCSCF for which the natural orbital occupation numbers of these two orbitals are very, very close to 2.0.

Hans Jørgen.

rlt92
Posts: 15
Joined: 20 Nov 2018, 12:29
First name(s): Richard
Last name(s): Thurston
Affiliation: UC Berkeley
Country: United States

Re: Molden inp files not showing symmetry

Post by rlt92 » 04 Oct 2020, 23:03

Hi Prof. Jensen,

Thanks for your reply. for this particular calculation i was replicating a literature MCSCF calculation of nitrobenzene. From what I gather they wanted to include the effects of the non-bonding orbitals around the oxygens which is why they only included 1 orbital for b1 and 1 orbital for a1 symmetries (however they also did their calculation in C1 symmetry so maybe they had other reasons why they didn't include correlating orbitals for a1 and b1 symmetries?). I'm curious to know if this reasoning is sound or if this is not a correct treatment of non-bonding orbitals. Any thoughts you have on the matter would be appreciated.

best,
Richard

rlt92
Posts: 15
Joined: 20 Nov 2018, 12:29
First name(s): Richard
Last name(s): Thurston
Affiliation: UC Berkeley
Country: United States

Re: Molden inp files not showing symmetry

Post by rlt92 » 05 Oct 2020, 00:48

As a follow up to the original issue, it looks like the software I was using to visualize the molecular orbitals (avagadro) wasn't correctly reading the files Jmol was able to correctly read the files. I'm still not sure why avagadro was having a hard time with these files but at least the issue wasn't with the calculation. Thanks for the help!

hjaaj
Posts: 394
Joined: 27 Jun 2013, 18:44
First name(s): Hans Jørgen
Middle name(s): Aagaard
Last name(s): Jensen
Affiliation: Universith of Southern Denmark
Country: Denmark

Re: Molden inp files not showing symmetry

Post by hjaaj » 06 Oct 2020, 14:30

Many have used ad-hoc guesses for active spaces in the past instead of using e.g. MP2 predicted occupation numbers, as you can do in Dalton.

There are NO reasons for including strongly occupied orbitals without correlating orbitals. As the occupation becomes very close to 2.0, any properties will be very close to the values with these two orbitals as inactive orbitals. The only effect is that the MCSCF costs more computer time ...

rlt92
Posts: 15
Joined: 20 Nov 2018, 12:29
First name(s): Richard
Last name(s): Thurston
Affiliation: UC Berkeley
Country: United States

Re: Molden inp files not showing symmetry

Post by rlt92 » 10 Oct 2020, 02:47

Prof. Jørgen,

Thanks for all the advice! You've been very helpful. I'll likely simplify the active space for future calculations.

Best,
Richard

Post Reply

Who is online

Users browsing this forum: No registered users and 6 guests