Author Topic: NTO Generation of X2C Calculations  (Read 1569 times)

hherbol

  • Newbie
  • *
  • Posts: 2
  • Karma: +0/-0
NTO Generation of X2C Calculations
« on: October 28, 2024, 09:19:21 PM »
I'm trying to make Natural Transition Orbitals (NTOs) of some molecules in Turbomole; however, it appears that a lot of the resulting files are undocumented in the Turbomole manual (same with the workflow).

Currently I'm doing it as follows:

1. Run an RIDFT calculation using X2C on
2. Run an ESCF calculation with X2C on
3. Use the "proper" program to generate the NTOs (proper -> mos -> dftnto -> choice of excitation -> choice of NTO definition)
4. Adding "$pointval nto 1-6 fmt=cub" to the control file and running "ridft -proper"

Now, three issues I have is that first off, the resulting NTO files are labelled in an interesting fashion that I can't understand.  For example, the hole for the first NTO (nto_occ*) output the following files:

nto_occ_1_x_d.cub
nto_occ_1_x_4.cub
nto_occ_1_x_3.cub
nto_occ_1_x_2.cub
nto_occ_1_x_1.cub

I believe from the manual where they outline how X2C mos are output these are:
  *_1 = Real part of the alpha spinor
  *_2 = Imaginary part of the alpha spinor
  *_3 = Real part of the beta spinor
  *_4 = Imaginary part of the beta spinor
  *_d = Total Density

Secondly, the NTO definitions aren't elaborated on in the documentation, so I don't really understand the intricacies of them:

    renormalized excitation   part (XX)       (1)
    excitation + deexcitation part (X+Y)(X+Y) (2)
    renormalized deexcitation part (YY)       (3)
    excitation + deexcitation part (XX)+(YY)  (4)

Thirdly, I don't understand why options (3) and (4) of the NTO definitions only exists when I do X2C calculations (when I did a closed shell calculation without X2C I only see options (1) and (2)).
« Last Edit: October 28, 2024, 09:31:14 PM by hherbol »