Broken Simmetry NEL

[lorenzo briccolani]

Dear Users, 

I am reaching out as I am uncertain about the correct assignment of NEL values in the BS routine. The system under investigation consists of a molecule with a central Cu2+ atom coordinating two ligands in a square planar geometry. In a standard spin density calculation (UPBE/DZVP-MOLOPT) with a multiplicity of two, the spin densities are well reproduced, with the unpaired electron localized precisely on the copper atom. The Hirshfeld output reports the following values:

Element  Kind  Ref Charge     Population       Spin moment  Net charge
           Cu     4      11.000        6.182   5.514            0.668     -0.695

My objective is to replicate this result using the BS approach. For a Cu2+ ion, I assume the configuration should be as follows:
&KIND Cu
     &BS ON
      &ALPHA
       L = 0 2
       N = 4 3
      NEL = -2 ?
      &END
      &BETA
       L = 0 2
       N = 4 3
      NEL = -2 ?
      &END
     &END

I have conducted several tests; however, I am unable to obtain the correct number of valence electrons and the expected multiplicity in the output. I have attached the input file for reference (opt.inp).

I would be sincerely grateful for any suggestions regarding the appropriate assignment of the NEL values.

Best regards,

Lorenzo Briccolani

[Krack Matthias]

Dear Lorenzo

 

The valence electron configuration of the Cu-q11 pseudopotential employed by you is 3d10 4s1. The following &BS section should create an initial 3d9 doublet configuration for a Cu(2+) atom:

 

 UKS

  MULTIPLICITY 2

 

  &BS

    &ALPHA

      N    3  4

      L    2  0

      NEL +1 -1

    &END ALPHA

    &BETA

     N    3  4

      L    2  0

      NEL -1 -1

    &END BETA

  &END BS

Note that these BS setting affects only the electron configuration of the initial atomic guess and the converged result might differ.

 

HTH

 

Matthias

 

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cp2k/7b7ad053-f176-49f2-a9ea-0580b9d0d584n%40googlegroups.com.

[lorenzo briccolani]

Dear Matthias,

Thank you very much for your help, as always—it is greatly appreciated.

Just to ensure I fully understand, the first column of values defines the target orbital, while the second column corresponds to the initial orbital. Therefore, in the case of the alpha electron, I add one electron with spin-up (NEL +1) in the 3d orbital by removing it from the 4s orbital. For the beta electron, I do not add any electron (NEL -1), transferring it from the 4s to the 3d orbital.

I apologize if I seem insistent, but I would like to clarify this point, as the manual refers to a single real value.

Following your recommendations, I assigned an initial multiplicity of two to the Cu atom. However, I am unsure why the calculation still assigns 10 valence electrons overall, while for the beta electrons, it assigns 4.5 instead of 4. Because the values enclosed in square brackets, I believe, refer to the number of alpha or beta electrons contained in each orbital. This is why, in the case of beta electrons, I expected D 4.00 instead of 4.50. Below, I have included the relevant lines from the output:
Guess for atomic kind: Cu

 Electronic structure
    Total number of core electrons                                         18.00
    Total number of valence electrons                                      10.00
    Total number of electrons                                              28.00
    Multiplicity                                                         doublet
    Alpha Electrons
    S   [  1.00  1.00  1.00]
    P   [  3.00  3.00]
    D      5.00
    Beta Electrons
    S   [  1.00  1.00  1.00]
    P   [  3.00  3.00]
    D      4.50

Best Regards 

Lorenzo Briccolani

To view this discussion visit https://groups.google.com/d/msgid/cp2k/ZRAP278MB08276547DC4F2701A2644E5EF4CB2%40ZRAP278MB0827.CHEP278.PROD.OUTLOOK.COM.

[Krack Matthias]

Dear Lorenzo

 

Sorry, my suggestion for the &BS section was wrong, try this instead

 

  &BS

    &ALPHA

      N    3  4

      L    2  0

      NEL  0 -1

    &END ALPHA

    &BETA

     N     3  4

     L     2  0

     NEL  -2 -1

    &END BETA

  &END BS

 

Best

 

Matthias

 

To view this discussion visit https://groups.google.com/d/msgid/cp2k/CAJPdbE-4Apgw%3DoyYG-LT%2BmkkwVrMJmfEw29VQ9VVqohHoCF85A%40mail.gmail.com.

[lorenzo briccolani]

Great, it's correct, thank you so much

Guess for atomic kind: Cu

 Electronic structure
    Total number of core electrons                                         18.00

    Total number of valence electrons                                       9.00
    Total number of electrons                                              27.00

    Multiplicity                                                         doublet
    Alpha Electrons
    S   [  1.00  1.00  1.00]
    P   [  3.00  3.00]
    D      5.00
    Beta Electrons
    S   [  1.00  1.00  1.00]
    P   [  3.00  3.00]

    D      4.00
 
Based on what criteria did you assign NEL 0 -1 to alpha and NEL -2 -1 to beta?

Thank you again, you have been very helpful.

Lorenzo