CASSCF and CASPT2 study on O- and Cl-loss predissociation mechanisms of OClO (A(2)A2)

For studying O- and Cl-loss predissociation mechanisms of OClO (A(2)A(2)), we calculated O- and Cl-loss dissociation potential energy curves (adiabatic minimum-energy dissociation paths) of several low-lying doublet and quartet states at the CASPT2 level and located the MECPs (minimum energy crossing points) for many pairs of the potential energy surfaces (PESs) at the CASPT2 and CASSCF levels. On the basis of our calculation results (including the spin-orbit couplings at the MECPs), we predict three processes for O-loss predissociation of A(2)A(2) and four processes for Cl-loss predissociation of A(2)A(2). The most favorable process for O-loss predissociation is OClO (A(2)A(2)) → A(2)A(2)/1 (2)B(2) MECP → 1 (2)B(2) (1 (2)A') → O ((3)P(g)) + ClO (X(2)Π) (the first O-loss limit), and the needed energy for this process from X(2)B(1) is 2.92 eV. The most favorable process for Cl-loss predissociation is OClO (A(2)A(2)) → A(2)A(2)/1 (2)B(2) MECP → TS1 (1 (2)B(2)) → 1 (2)B(2)/1 (2)A(1) MECP → Cl ((2)P(u)) + O(2) (X(3)Σ(g)(-)) (the first limit), and the needed energy is 3.08 eV. In the previously suggested mechanisms (processes), the A(2)A(2) state was considered to go to the important 1 (2)B(2) state via 1 (2)A(1) (A(2)A(2) → 1 (2)A(1) → 1 (2)B(2)). In the present study we have found that the A(2)A(2) state can directly go to 1 (2)B(2) (at the located A(2)A(2)/1 (2)B(2) MECP the CASPT2 energy (relative to X(2)B(1)) and CASSCF spin-orbit coupling are 2.92 eV and 61.3 cm(-1), respectively). We have compared our processes (A(2)A(2) → 1 (2)B(2) → ...) with the processes (A(2)A(2) → 1 (2)A(1) → 1 (2)B(2) → ...) suggested in the previous MRCI studies and rewritten by us using our calculation results. Energetically the MRCI process for O-loss predissociation (to the first limit) is only slightly (0.13 eV) more favorable than our process, and the MRCI processes for Cl-loss predissociation (to the first and second limits) need the same energies as our processes. By considering the probabilities of radiationless transitions, the MRCI processes are less favorable than our processes since the MRCI processes proceed via more PES/PES crossings (more MECPs). The experimental facts concerning the photodissociation are explained.