Monoaza-analogs† of trimethylenemethane. Isoelectronic similarities and differences

Organic molecules characterized by high-spin ground states are of interest for many reasons, including their potential use as spin reservoirs in ferromagnetic organic materials. For the design of high-spin organic molecules having an even number of electrons, one typically tries to take advantage of Hund’s rule 4 in constructing systems biased towards having triplet ground states. That is, one attempts to design a molecule where the frontier molecular orbitals (MOs) are degenerate (or nearly so), in which case the preferred occupation of those orbitals may be expected to generate, for the case of two electrons, the diradical triplet state. Trimethylenemethane (TMM) in D3h symmetry is an example of such an organic molecule. TMM is a non-Kekulé hydrocarbon (non-Kekulé implying that no resonance structure can be drawn placing its 2n pi electrons into n double bonds) characterized by two degenerate non-bonding π frontier MOs (Fig. 1). Coulson 5 credited Moffitt with first having appreciated this unusual (for an organic molecule) feature of TMM, and shortly thereafter Longuet–Higgins 6 provided a theoretical analysis of the orbital properties of non-Kekulé hydrocarbons in general. In spite of having been the subject of considerable theoretical study, however, TMM remained ‘undiscovered’ until 1966, when Dowd 7 successfully synthesized TMM and observed its EPR spectrum in a frozen matrix; later experiments confirmed the ground state of TMM to be the triplet. A recent photoelectron spectroscopy experiment quantitated the separation between the 3A92 ground state and the excited 1A1 state as 16.1 ± 0.1 kcal mol in the gas phase. High level theoretical calculations are in good agreement with this splitting. Although a small molecule, TMM proves a challenging test case for different theoretical models—achieving equivalent accuracies for different spin states in molecules having nearly degenerate frontier orbitals is not trivial. For best results, multireference wavefunctions corrected for dynamical correlation effects are required; a summary of prior theoretical work on TMM is available. Dowd 11 and Berson 12 were both particularly active in expanding knowledge of the chemistry of TMM and substituted derivatives, and subsequent workers have continued to find new uses for TMM derivatives, e.g., in the cleavage of double-stranded DNA. Larger non-Kekulé systems have also