Starting with formaldehyde-C1" 1,2-dibromocyclopentane-4-C" ( X I ) was synthesized in eight steps. Conversion of X I to cyclopentadiene was accomplished in three different ways, namely: the dehydrobromination of X I with quinoline a t 195-198' C, the preparation of 1,2-bis-(trimethylammonium)-cyclopentane-4-14 hydroxide from X I followed by thermal decomposition, and the conversion of X I to 3,5-dibromocyclopentene followed by reaction of the latter with magnesium in d r y ether. Degradations of the samples of cyclopentadiene obtained by these three synthetic routes showed a completely randomized distribution of the Cld activity, with 20y0 of the total activity on each of the five carbon positions. These results led t o the conclusion that these preparations of cyclopentadieile very likely involved ionization to the cyclopentadienyl anion. T h e randomized distribution of C" in the resulting products thus furnishes a strong indication that all five carbon positions in the cyclopentadienyl anion are equivalent.
INTRODUCTIONT h e acidity of cyclopentadiene was first recognized i11 1900 by Thiele (I), and its potassium salt was prepared in 1901 (2). This salt is stable, though very reactive. The reason for the stability of salts containing the cyclopentadienyl anion has been given asearly as 1928, by Goss and Ingold (3), and is that this ion has six (n) electrons, distributed over the entire five-membered ring, thus constituting a stable aromatic system similar to benzene. T h e resonance energy of this ion has been calculated to be about 42 kcal per mole (4).Salts of cyclopentadiene with organic cations are also known. Pyridi~lium cyclopentadienylide has been described by Lloyd and Sneezum (5) and by Kosower and Ramsey (6). Triphenylphosphonium cyclopentadienylide has also been reported (7).From theoretical considerations and physical chemical measurements, a number of workers (4,(8)(9)(10)(11)(12) have predicted that the ~legative charge on the cyclopentadiellyl anion will not be localized on a partic~ilar carboll atom but will be distributed over theion as a whole, as depicted by I. In the present study, attempts were made to obtain, by the tracer technique, some direct experimental evidence ill support of these theoretical predictions.If it were possible to synthesize cyclopentadiene-5-C14 (11), conversion of this conlpound to its potassi~~m salt followed by regeneration should yield a cyclopentadiene in which the C14 distribution could provide definitive evidence for the struct~ire of the cyclopentadienyl anion. Should the anion have the symmetrical electronic structure I, on 'Manuscript