Crystal structures and photoelectron spectra of some trimethanoanthracenes, tetramethanonaphthacenes, and pentamethanopentacenes. Experimental evidence for laticyclic hyperconjugation

Abstract

Photoelectron (p.e.) spectra of the series of dienes (4), (10), (11a)–(13a), and crystal structures for the dodecachlorodienes(11b)–(13b) are reported. The spectra revealed large π-splitting energies of 0.32 and 0.52 eV for (4) and (11a) respectively. The value of (4) is attributed to the presence of orbital interactions through six σ bonds (OIT-6-B), and this, taken with the p.e. data for the dienes (1)–(3), provides convincing evidence for the long-range nature of OIT-n-B. The larger π-splitting energy of 0.52 eV observed for (11a), compared with (4), has been explained in terms of the presence of two reinforcing orbital interactions: OIT-6-B and a new variant of hyperconjugation, called laticyclic hyperconjugation, in which the π MOs of the double bonds mix with the _ψπ MO of the central CH₂ bridge of (11a). The surprisingly large π-splitting energy of 0.29 eV observed for (12a) is also due largely to laticyclic hyperconjugation involving both pairs of symmetry adapted _ψπ MOs of the CH₂ bridges. No π-splitting energy could be detected in the p.e. spectrum of (13a). From the crystal structures of (11b)–(13b), it was found that the distance between neighbouring CH₂ groups, and between a double bond and its closest CH₂ neighbour is about 3 Å in all three compounds. HF/STO-3G calculations on model ethene ⋯(CH₄)_n⋯ ethene complexes suggest that laticyclic interactions, like OIT-n-B, are very long-range processes, the predicted π-splitting energies amounting to ca. 10^–3 eV for two double bonds separated by ca. 33 Å. The relevance of these orbital interactions to biological electron-transfer processes is briefly discussed.