Multiwfn forum

max

Member

Registered: 2020-03-30

Posts: 2

generate excited state .wfn in orca

Dear Multiwfn users.

I am still relatively new to orca and am struggling to figure out how to get orca to generate common output files. Currently, I am trying to figure out how to generate excited state .wfn file to use in Multiwfn charge-transfer analysis based on electron density differences (Multiwfn section 4.18.3). The orca_2aim program takes the electron density from the .gbw file and replots as a .wfn. This makes it trivial to generate ground state .wfn file.

My issue is in finding a way to generate the excited state .gbw fill which can be converted to a .wfn from an existing TDDFT calculation? This was simple in Gaussian09 using TD(read, root=#, NonEqSolv) with the Dencity=Current and out=wfn but i can not find the equivalent operation in the orca manual. Does anyone know if this is possible in orca?

Best
Max

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,468

Website

Re: generate excited state .wfn in orca

Unfortunately, AFAIK there is no keyword for directly generating .wfn/.wfx file containing wavefunction (i.e. natural orbitals) of excited state calculated by TDDFT method. You may ask this on ORCA forum, developers may have some workarounds.

However, via Multiwfn it is possible to generate (unrelaxed) excited state wavefunction file based on output file and .molden file of ORCA. Below is an example.

Run below task

! PBE0 def2-SV(P) tightSCF miniprint nopop
%maxcore  1000
%pal nprocs   4 end
%tddft
nroots 3
tprint 1E-8
end
* xyz   0   1
O     -0.00000000   -0.00000000    0.11930801
H      0.00000000    0.75895306   -0.47723204
H     -0.00000000   -0.75895306   -0.47723204
 *

Assume that output file is h2o.out, the gbw file is h2o.gbw. Then run

orca_2mkl h2o -molden

Now you have h2o.molden.input.

Then boot up Multiwfn and input
h2o.molden.input
18  // Electronic excitation analysis
13  // Generate natural orbitals of specific excited states
h2o.out
1-3  // Assume that you want to generate wavefunction file for all the three calculated excited states.

Now you will find NO_0001.molden, NO_0002.molden and NO_0003.molden in current folder. They respectively contain natural orbitals of the 1st, 2nd and 3rd excited states. Since in Multiwfn, .molden file provides richer information than .wfn file, you do not need to convert them to .wfn file (however, if you really need .wfn file, you can load one of them, then enter subfunction 2 of main function 100, you will find an option used to export current wavefunction file as .wfn file)

Above steps work for currently latest version, namely Multiwfn 3.7(dev).

max

Member

Registered: 2020-03-30

Posts: 2

Re: generate excited state .wfn in orca

My original input is for the orca job was:

! UKS BP SP PBE0 D3BJ ZORA ZORA-def2-SVP Sarc/J FinalGrid6 TightSCF RIJCOSX CPCM(Acetonitrile)
%maxcore 2000
%pal
nprocs 12
end
%tddft
    maxdim 10
      nroots 300
        DoNTO true
end
%scf
    MaxIter 500
end
%basis Newgto 58 "SARC-ZORA-TZVPP" end
       Newauxgto 58 "Sarc/J" end
       Newgto 16 "ZORA-def2-TZVPP" end
       Newauxgto 16 "Sarc/J" end
       Newgto 8 "ZORA-def2-TZVPP" end
       Newauxgto 8 "Sarc/J" end
end
%output
    Print[P_ReducedOrbPopMO_L] 1;
    Print[P_FragPopMO_L] 1;
end

do you recommend that i rerun using a tight print or is finalgrid6 enough?

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,468

Website

Re: generate excited state .wfn in orca

Please check beginning of Section 3.21 of Multiwfn manual on how to correctly prepare input file for most functions in electron excitation module.
The role "tprint" is very different to finalgrid6, it controls threshold of printing coefficients of configuration state functions in the TDDFT calculation. By default, the tprint is too large, making many coefficients ignored in the output file, and thus leading to inaccurate analysis result in Multiwfn.