Is there a transition state for the unimolecular dissociation of cyclotetraoxygen (O₄)?

The cyclo‐O₄ molecule is isoelectronic with cyclobutane and has been mentioned as a potential new high energy density material (HEDM). The important unresolved question has been whether or not cyclo‐O₄ is a genuine minimum on the O₄ potential energy hypersurface. Here the transition state for cyclo‐O₄ dissociation to two O₂ molecules has been located at a number of levels of theory using a double zeta plus polarization (DZP) basis set. The theoretical methods with which the transition state geometry was optimized include two‐configuration self‐consistent‐field (TCSCF), configuration interaction including all single and double excitations (CISD), coupled cluster including all single and double excitations (CCSD), and the latter with a noniterative correction for connected triple excitations, CCSD(T). The equilibrium geometry of O₄ has D_2d symmetry, while a stationary point of D₂ symmetry is of Hessian index two (i.e., two imaginary vibrational frequencies) at the highest level of theory, DZP CCSD(T). The true transition state lies only slightly lower energetically (classical barrier 7.9 kcal/mole) and is of C₂ symmetry with the DZP CCSD(T) method. The activation energy was predicted in a completely consistent manner to be 6.2 kcal/mole. The possibility that this barrier could disappear entirely at higher levels of theory is discussed. It is concluded that the O₄ barrier is too low for O₄ to be useful as a HEDM.