CCL: How to eliminate the spin contamination in the ORCA

 Sent to CCL by: =?utf-8?B?TWFyaXVzeiBSYWRvxYQ=?= [mariusz.radon:-:uj.edu.pl]
 > On 22 Apr 2022, at 11:38, Nikunj Kumar nkk749 a gmail.com
 <owner-chemistry : ccl.net> wrote:
 >
 >
 > Sent to CCL by: "Nikunj  Kumar" [nkk749..gmail.com]
 > Dear all,
 > I am working on an interesting catalyst system which has non-innocent
 ligand. I
 > have optimized the triplet structure using the ORCA 4.2.1 employing the
 TPPSh
 > functional. I was expecting the S^2 value to be around 2.0 but it came out
 to
 > be 2.71. Can we consider this as a triplet? Is this value reliable, if not
 how
 > can we get rid of the spin contamination.
 >
 > Thanks
 > Nikunj Kumar
 Dear Nikunj:
 Assuming that the contamination of your triplet originates from one single
 quintet state, the value of 2.71 would imply ~18% of the contaminant:
 <S2> = (1-x)*2 + x*6 => x=0.18
 To know how important is this effect for the energy, you need to also compute
 the energy of contaminating quintet state. As Marcel Swart already told you,
 there are relatively simple formulas to correct the energy for
 spin-contamination. You can find them in the Marcel’s paper or in many
 other papers, such as mine: https://dx.doi.org/10.1021/ic503109a.
 One important remark: you need to compute the the proper quintet state (i.e.,
 the one which contaminates your triplet state). To know how to obtain the proper
 quintet state, it can be useful to take a look at natural orbitals of your
 spin-contaminated triplet state. You will most likely see the following natural
 orbitals:
 - two orbitals with occupation numbers exactly 1.0 (phi_1, phi_2)
 - a pair of orbitals with fractional occupation numbers 2-v and v (phi_3,
 phi_4).
 For the proper quintet state, you need to put unpaired electrons into these four
 orbitals (phi_1, phi_2, phi_3, phi_4). You take a look into the above paper from
 Inorganic Chemistry (https://dx.doi.org/10.1021/ic503109a) for some examples.
 I don’t know if it is possible to get natural orbitals directly from
 Orca, but you can always compute them with my program Natorbs.
 https://tungsten.ch.uj.edu.pl/~mradon/natorbs/ or https://pypi.org/project/natorbs/
 Natorbs imports UHF-type molecular orbitals via either Molden format or cclib
 interface.
 Best wishes,
 Mariusz Radon
 --
 Mariusz Radon, Ph.D., D.Sc.
 Associate Professor
 Faculty of Chemistry, Jagiellonian University
 Address: Gronostajowa 2, 30-387 Krakow, Poland
 Room C1-06, Phone: 48-12-686-24-89
 E-mail: mradon : chemia.uj.edu.pl (mariusz.radon : uj.edu.pl)
 Web: https://tungsten.ch.uj.edu.pl/~mradon
 ORCID: https://orcid.org/0000-0002-1901-8521