IS N=60 A MAGIC NUMBER FOR C-N(+) CLUSTERS OR PART OF A MAGIC SHELL

Metastable fractions were calculated for a series of consecutive C-2 eliminations from fullerene cations: C-n(+) -> C-n-2(+) + C-2, n = 64-54. The results were compared with experimental data obtained on fullerene samples using a VG-ZAB/2F instrument. Calculations were performed on two models: (1) a 'magic shell' model, in which C-60(+) is considerably more stable than C-62(+) but only slightly more stable than C-58(+); and (2) a 'magic number' model, in which C-60(+) is considerably more stable than both C-62(+) and C-58(+). RRKM/QET calculations of rate-energy dependences were performed. These led to parent ion and metastable ion breakdown curves. A non-uniform energy deposition function derived from electron ionization experiments on C-60 (H. Foltin et al., J. Chem. Phys., 98 (1993) 9624) was applied. The experimental results can be fit by both models, but are in better agreement with the magic shell model, provided the activation entropies do not vary strongly from one fullerene ion reaction to the next. Taking a binding energy of 7.1 eV for C-60(+) due to Foltin et al., the following values are obtained for the other binding energies: C-64(+), 5.9 +/- 0.3 eV; C-62(+), 5.1 +/- 0.7 eV; C-58(+) , 6.7 +/- 0.7 eV. The results substantiate earlier expectations by Klots (C.E. Klots, Z. Phys. D., 21 (1991) 335) based on the evaporative ensemble model.