JOURNAL OF PHYSICAL CHEMISTRY A, v.125, no.21, pp.4548 - 4557
Abstract
The Stone-Wales bond rotation isomerization of nonicosahedral C-60 (C-2v-C-60) into isolated-pentagon rule following icosahedral C-60 (I-h-C-60 or IPR-C-60) is a limiting step in the synthesis of I-h-C-60. However, extensive previous studies indicate that the potential energy barrier of the Stone-Wales bond rotation is between 6 and 8 eV, extremely high to allow for bond rotation at the temperatures used to produce fullerenes conventionally. This is also despite data indicating a possible fullerene road mechanism that necessitates low-temperature annealing. However, these previous investigations often have limiting factors, such as using the harmonic approximation to determine free energies at high temperatures or considering only the reverse I-h-C-60 to C-2v-C-60 transition as a basis. Indeed, when the difference in energy between I-h-C-60 and C-2v-C-60 is accounted for, this barrier is generally reduced by similar to 1.5 eV. Thus, utilizing the recently developed density functional tight binding metadynamics (DFTB-MTD) interface, the effects of temperature on the bond rotation in the conversion of C-2v-C-60 to I-h-C-60 have been investigated. We found that Stone-Wales bond rotations are complex processes with both in-plane and out-of-plane transition states, and which transition path dominates depends on temperature. Our results clearly show that at temperatures of 2000 K, the free energy for a C-2v-C-60 to I-h-C-60 transition is only similar to 4.21 eV and further reduces to similar to 3.77 eV at 3000 K. This translates to transition times of similar to 971 mu s at 2000 K and similar to 34 ns at 3000 K, indicating that defect healing is a fast process at temperatures typical of arc jet or laser ablation experiments. Conversely, below similar to 2000 K, bond rotation becomes prohibitively slow, putting a lower threshold limit on the temperature of fullerene formation and subsequent annealing.