CCL: Spin contamination increased by annihilation of the first-order contaminant

 Sent to CCL by: Mariusz Radon [mariusz.radon]_[gmail.com]
 On 05/01/2015 05:00 PM, Igors Mihailovs igors.mihailovs0::gmail.com wrote:
 > Dear all,
 >
 > the spin
 > contamination was not descreased by removal of the first-order contaminant:
 >  (...)
 >  Annihilation of the first spin contaminant:
 >  S**2 before annihilation     2.1080,   after     8.9661
 >
 >
 > Is there some way to interpret this result?
 >
 Dear Igors:
 Yes, this could happen if your wave function is contaminated by higher
 spin states (not only S+1, but also S+2, S+3, etc; where S = M_S is the
 desired spin state; zero in your case).
 Take a look at the following paper: Davidson, E. R. & Clark, A. E.
 "Spin
 Polarization and Annihilation for Radicals and Diradicals"
 Int. J. Quantum Chem., 2005, 103, 1-9. The most relevant conclusion:
 "The annihilation procedure greatly increases the weights on the higher
 spin states and can easily lead to an increase, rather than a decrease,
 in <S^2>."
 You have probably a biradical or multiradical system. In such case,
 analysis of natural orbitals (eigenvectors of the density matrix) or
 natural spin density orbitals (eigenvectors of the spin density matrix),
 or Amos-Hall corresponding orbitals can be useful to identify the pairs
 of weakly coupled electrons in your system.
 Best regards,
 Mariusz
 --
 Dr Mariusz Radon, Ph.D.
 Coordination Chemistry Group
 Faculty of Chemistry
 Jagiellonian University
 ul. Ingardena 3, 30-060 Krakow, Poland
 http://www2.chemia.uj.edu.pl/~mradon