Hi all,
I'm trying to run a single point calculation with swapped MO ( I used guess=Alter to swapped them.)
I checked that the MO order is swapped at the beginning of the SCF but it goes back to original ordering at the end of calculation. I want to just restrict the order not freeze MO.
Does anyone know the keyword to do this?
Thank you.
Dear W. Park, I explain what I used to do:
(1) Perform a DFT calculation to obtain a ground-state relaxed wavefunction.
(2) Use the Gaussian standard keyword Guess=(Read,Alter) to obtain the target state: Guess=Read recovers the ground-state wavefunction, and
Guess=Alter specifies orbitals to be alternated from the ground-state wavefunction.
(3) Perform a UDFT calculation keeping the modified wavefunction so as to reach a self-consistent field solution for the target state.
Tip: Use SCF=QC
In other words, read the wavefunction of the ground state, then from the checkpoint modify the orbitals you want.
Best, Pablo
Dear Pablo Mtz,
Thank you for your explanation.
I can see that it reorder the MO when I used Guess=Alter. Do you mean that it restrict the MO order during SCF when I run UDFT?
If you don't mind could you please share the input for UDFT calculation? I'm new to this Gaussian, and I'm not sure if I'm doing correctly.
Best, Woojin
Dear Woojin Park
"I can see that it reorder the MO when I used Guess=Alter."
Yes, it does, but the SCF will try to find the ground state anyway, going back to original ordering at the end of calculation.
Here an input example (after you obtained the ground state):
%oldchk=your_ground_state.chk
%chk=your_new_state.chk
#P UCAM-B3LYP/6-31G** sp Geom=CheckPoint
# Guess=(Read,Alter) SCF=QC
your calculation title
0 1
xyz
432 433
Here orbital alpha-HOMO 432 is swapped with alpha-LUMO 433. Do not forget blank lines as epxlained in the 'examples' section:
https://gaussian.com/guess/
It is important to use SCF=QC.
Let me know if you get your new wavefucntion.
Best, Pablo
Dear Pablo Mtz
Thank you and sorry to bother you again.
I'm getting strange error when I used Geom=Checkpoint.
"Atom specifications unexpectedly found in input stream."
I couldn't find the solution.
Could you please check my attached files?
q : DFT calculation
q-read : file which met error
Btw, I want to optimize triplets and switch MO1 and HOMO.
Best, Woojin
Dear Pablo Mtz
Both way gives me same error "Unexpected end-of-file while reading orbital switches."
Please see the attached files.
Best, Woojin
Dear Woojin Park according to the description of Guess=alter in the Gaussian website:
"... The alteration sections consist of a set of transpositions indicating that one of these occupied orbitals is to be replaced by one of the other (virtual) orbitals ..."
It seems only LUMOs can be swapped. Please try one LUMO to confirm this.
See the 'examples' section again, it seem Guess=(read,permute) is another option in which you order the orbitasl as you want. I've never tried this way, but it seems that i may work.
Best, Pablo
Dear Pablo Mtz
I can swap the MOs using Guess=(read,permute). It prints the symmtry of initial orbital "(B1) (A1) (B2) (A1) (A1) (A1)".
(I swapped the MO1 and MO5)
However, I met the original problem. It relaxes to the ground state wavefunction. After the SCF, it print the symmetry of orbitals "(A1) (A1) (B2) (A1) (B1) (A1)" and the eigenvalues have ascending order, meaning it didn't restrict the ordering.
Please see the attached file if it is needed.
Best, Woojin
Dear Woojin Park perhaps this configuration is not possible, the first MO is for core electrons, you are trying to switch the 1s orbital of oxygen with a 2p orbital, which is challenging for a SCF calculation.
Triplet electrons are localized in p orbitals, If you are trying to analyze different triplet states, I'd suggest to swap between p orbitals instead of 1s.
Best, Pablo
https://commons.wikimedia.org/wiki/File:Molecular_Orbitals_for_Water.png
Dear Pablo Mtz
I need to swap 1s orbital but this may not be a good idea.
Thank you so much for your help.
Wish you all the best
Woojin
- Badges
- Science topic
- Similar topics
- Human-Computer Interaction
- Collaboration