Carbon-13 Hyperfine Splitting in the Electron Spin Resonance Spectra of Semiquinones and Cyclooctatetraene

Abstract

Measurements have been performed on the carbon-13 hyperfine splitting in the ESR spectra of a series of methyl- and chloro-substituted semiquinones and of the cyclooctatetraene anion radical. The splittings were observed from carbon-13 nuclei present in natural abundance (approx. 1%), and the carbon-13 satellites were distinguished from spurious low-intensity lines arising from side-reaction produced radical impurities by means of careful intensity measurements as well as other techniques. The largest observed carbon-13 splitting had a magnitude of 1.7 gauss, and in most spectra smaller splittings were found. The carbon-13 satellites were overlapped by the major lines in the spectra (arising from proton splitting), but after correcting the measured intensities for this overlap, good agreement was found between the predicted and measured intensities in those spectra in which a detailed analysis was possible. In the p-benzosemiquinone ion, a previously observed line of low intensity has been identified as a splitting arising from carbon-13 nuclei, and has been assigned, on the basis of intensity measurements, to the four CH carbon atoms. No splitting was found from the carbon atoms in the CO bonds of this compound, and it was concluded that the splitting is less than 0.6 gauss. Splitting constants tentatively identified with the methyl carbon atoms of the methyl-substituted compounds were in the neighborhood of 1.5 gauss. The splitting in the cyclooctatetraene anion indicates that the three sigma bonds attached to a carbon atom lie in a single plane. The experimental results are in agreement with the theory of carbon-13 hyperfine splittings recently developed by Karplus and Fraenkel. Comparison of theory and experiment indicates that these results provide a very sensitive measure of the pi-electron spin densities in the radicals. The valence-bond theory of the spin-densities in the methyl substituents is shown to be inadequate in its present form, but the molecular-orbital theory appears to give good agreement. This latter agreement is based on certain as yet unjustified assumptions, and therefore both theories require re-examination. The spectrum of the previously observed semiquinone ion formed from 2,5-dihydroxyquinone was interpreted in order to obtain an estimate, in conjunction with other data, of the hyperfine splitting at the carbon-13 nucleus of the carbon atom in a CO bond. The analysis also shows that the two CH carbon atoms in this compound have negative pi-electron spin densities.