Dedicated to Professor Rolf W. Saalfrank on the occasion of his 70 th birthdayAb initio and density-functional theory (DFT) calculations have been used to investigate the model rearrangements of quadricyclane to norbornadiene catalysed by single CuSO 4 and SnCl 2 molecules. The isolated reactions with the two molecular catalysts proceed via electron-transfer catalysis in which the hydrocarbon is oxidised, in contrast to systems investigated previously in which the substrate was reduced. The even-electron SnCl 2 -catalysed reaction shows singlet-triplet two-state reactivity. Solvation by a single methanol molecule changes the mechanism of the rearrangement to a classical Lewis acid-base process.
IntroductionThe quadricyclane (1) to norbornadiene (2) rearrangement [1] has long been of interest as a prominent example of a thermally forbidden, photochemically allowed exothermic process that can be catalysed by many very different species.Our interest in this rearrangement is strengthened because it represents a very early example of a holecatalysed [2] reaction. Haselbach et al. [3] first observed that the radical cation generated by ionising quadricyclane in a Freon matrix at 77 K, and later Kelsall and Andrews at 4 K [4], rearranged spontaneously under their reaction conditions to the norbornadiene radical cation. Later, Turro and Roth [5] were able to show that the quadricyclane radical cation does have a significant lifetime using CIDNP measurements and later time-resolved ESR spectroscopy.The rearrangement of 1 +• to 2 +• is also of considerable theoretical interest as it served as the ma-0932-0776 / 10 / 0300-0347 $ 06.00 c 2010 Verlag der Zeitschrift für Naturforschung, Tübingen · http://znaturforsch.com jor example that led to our current understanding of hole-catalysed reactions in particular and openshell pericyclic transformations in general. Soon after Woodward and Hoffmann's original publication of the "HOMO rule" that governs the stereochemistry of electrocyclic reactions [6], Longuet-Higgins pointed out [7] that all radical electrocyclic reactions are Woodward-Hoffmann forbidden. However, his conclusion that these reactions should therefore be slow was revised in Haselbach's seminal paper [3] in which he pointed out that the rearrangement of 1 +• to 2 +• can take place via a very low-lying state crossing, rather than through a continuous change characteristic of an allowed reaction. However, Haselbach's treatment of the symmetrical reaction path, in which the two cyclopropane bonds open simultaneously, neglected the Jahn-Teller effect in the transition state. Bischof [8] pointed out that a classical Jahn-Teller situation exists at the crossing point (because at this point the two "crossing" orbitals are degenerate by definition), so that the C 2v structure suggested by Haselbach is a hilltop, rather than a transition state. Later calculations [9] showed the hilltop to have C 2 symmetry. This structure undergoes Jahn-Teller distortion orthogonal to the reaction path to yield two enantiomeric C 1 trans...