MOs for a doublet system using Restricted open-shell

How To
#1

Dear All,
How can I have one set of orbitals for a doublet system using DFT method?
I used “UNRESTRICTED = false” key, but I got alpha and beta sets of orbitals with different energies.
And there is a strange result I got for the order of alpha MOs’ energies [a.u.], -0.4639 (HOMO) is more stable than -0.3556 (HOMO-1). How is possible?

End of Output:

                Orbital Energies (a.u.)

Alpha MOs
– Occupied –
-0.3556 -0.4639
– Virtual –
-0.2592 -0.2578

Beta MOs
– Occupied –
-0.3931 -0.3556
– Virtual –
-0.2839 -0.2585

Input:
$molecule
2 2
Cu 4.805355 1.736522 4.318146
O 4.715676 4.133872 10.308799
N 6.511375 1.215674 3.420703
C
H
…
$end

$rem
jobtype = SP
method = B3LYP
basis = def2-SVP
thresh = 14
scf_convergence = 6
max_scf_cycles = 1000
gui = 2
sym_ignore = true
print_orbitals = false
UNRESTRICTED = false
$end

#2

ROHF orbitals can be misleading and they are not uniqely defined so the results between different packages can differ. You can read here about Koopmans theorem for open-shell species:
https://en.wikipedia.org/wiki/Koopmans%27_theorem
ROHF can also cause convergence problems, so these calculations need to be very carefully tested.
I recommend to look at UHF orbitals instead. Even better – look at the quantities such as spin densities to get an idea of where unpaired electrons are.

1 Like