For several years we have been interested by the stabilization of highly reactive species, such as silyl cations, [1] carbocations, [2a] carbenes, [2b] diradicals, [2c] and antiaromatic heterocycles; [2d] thus, cyclopentadienyl cations are of special interest for us. The parent compound (C 5 H 5 ) is supposed to have antiaromatic character, [3] and the triplet ground state predicted from simple H¸ckel theory has been confirmed by ESR spectra [4] and by the latest ab initio calculations. [5] Like the C 5 H 5 ion the pentamethylcyclopentadienyl cation (C 5 Me 5 ) is predicted to have a triplet ground state, with the singlet state being 4.2 kcal mol À1 higher in energy. [5b] Jutzi and Mix [6] reported that the reaction of bromopentamethylcyclopentadiene with silver tetrafluoroborate or hexafluoroantimonate below À 30 8C, led to a deep violet solution. In the 13 C NMR spectrum, only a broad signal attributed to the solvent was observable, which indicated the triplet nature of the cation. Polymers were formed upon warming this solution, but the pentamethylcyclopentadienyl cation has been trapped by various nucleophiles.Based on this prior literature, the recent report on the isolation of the pentamethylcyclopentadienyl cation 1 by one of us [7] was therefore really fascinating and unexpected. The tetrakis(pentafluorophenyl)borate salt of 1 was described as a crystalline material, stable for weeks at room temperature, and that can be left open to the atmosphere without serious decomposition.There were two major differences between the calculations and observations: the observed C4ÀC5 bond length (1.51 ä) was 11 % longer than the calculated one (1.36 ä), and in contrast to the calculated geometry, which predicted no pyramidalization of C4 and C5, their attached methyl groups protruded appreciably from the plane (CH 3 -C4-C5-CH 3 dihedral angle 106.98; Figure 1). These differences were explained by crystal packing between the anion and the cation, which would pyramidalize C4 and C5, a distortion permitted by the weak p bonding. These would have been noncovalent nonbonded interactions. The resulting deformations would have been a trade off between coulombic attractions and nonbonded repulsions. Figure 1. Comparison of 13 C NMR chemical shifts and geometric parameters for 1 and 2 (1: solid-state NMR spectroscopy and X-ray diffraction study; 2: solution NMR and calculations).Although reasonable these explanations were not totally convincing. Moreover, the observed dihedral angle between the methyl groups bonded to C4 and C5 and the C4ÀC5 bond length were in perfect agreement with a saturated fragment (C4 À C5 single bond and the methyl groups in trans position). In other words, hydrogen atoms (difficult to observe by X-ray diffraction studies) might have been present on C4 and C5 and therefore the actual structure would be the already known [8] pentamethylcyclopentenyl cation 2. This assumption was reinforced by the 13 C NMR chemical shift for the C4 and C5 carbons of 1 (d 60), which are in the expected range for...
