[Octopus-users] How to set the electronic occupations

[Yasumitsu Suzuki]

Dear Octopus developpers and users,

Hi, everyone.
This time I want to ask you about how to set the initial electronic
occupations of the states properly.
I tried to perform the TDDFT-Ehrenfest MD simulation for "Combustion of
acetylene (acetylene molecule + oxygen molecule)" which is introduced in
the octopus web page.
There I wondered how I can set the electronic occupations so that oxygen
molecule has triplet multiplicity.
When I set inp file like:

CalculationMode=gs
Units=eV_Angstrom
Dimension=3
PeriodicDimension=0
Boxshape=sphere
radius=10
spacing=0.2
%Coordinates
"O" | 5.25 | 0 | 0
"O" | 4.00 | 0 | 0
"H" | -1.00 | 0 | 0
"C" | -2.06 | 0 | 0
"C" | -3.27 | 0 | 0
"H" | -4.33 | 0 | 0
%
ExtraStates=1
%Occupations
1 | 1
%
SpinComponents = unpolarized

I observed that the KS orbital obtained correctly reproduces the triplet
multiplicity of O2 molecule. (Two degenerated HOMO has one electron
respectively, and they are anti-bonding orbital of O2 which was
confirmed by XCrysDen visulalization.)
But, I think that it resulted in success by chance, because I didn't
know whether anti-bonding orbital of O2 will be HOMO or not when I set
%Occupations initially.
So, when we set the electronic occupations of such systems, must we
investigate which occupations works correctly by trial and error?
Is there any proper way for that?
Also I want to know wheter this initially set occupations will be fixed
in afer TD simulation or not.
I mean, for example in this case, if initial HOMO become not highest and
another orbital become HOMO during TD simulation, this HOMO still has
two electrons and the orbiatl that was HOMO initially still has one
electron?
Any advice or hint about this matter will be greatly appreciated.
Please tell me some comments.

Sincerely,

Yasumitsu Suzuki
Department of Chemical System Engineering
Graduate School of Engineering
The University of Tokyo