Computations to Evaluate Dewar's Split P -Orbital (SPO) Method

The split p-orbital (SPO) wavefunction derived from the single determinant | 2pz2p̄z | corresponding to the singlet state in which a single 2pz orbital is doubly occupied, correlates the electronic motion so that the electrons always occupy opposite lobes of the 2pz orbital. This SPO singlet wavefunction may be written as 2—½{— | 2pz2p̄z | + | | 2pz | | 2p̄z | |} and is the lowest-energy linear combination of this pair of two-electron determinants. In this two-electron context, the triplet function | 2pz | 2pz | | is degenerate with the SPO singlet. In order to test the persistence of this qualitative picture when occupied sigma core orbitals are explicitly included, we have performed analogous rigorous computations on the 1s22s22pz2 nonstationary state of carbon, using corresponding determinants | 1s1s̄2s*2s̄*2pz2p̄z | and | 1s1s̄2s*2s̄* | 2pz | | 2p̄z | |, which include occupied 1s and 2s core orbitals. The lowest-energy singlet linear combination is found to be almost entirely composed of the first determinant and corresponds to a conventional, rather than an SPO wave-function. The triplet determinant | 1s1s̄2s*2s̄*2pz | 2pz | | is of 2.5 a.u. higher energy than the lowest singlet. While these calculations do not support the SPO method as stated for this one-center problem, it is possible that computations with other basis functions will suggest vertical correlation of pi-electronic motion.