“…In addition, if these spectra were the result of ultrafast dynamics, some temperature dependence of the line shapes (more than would be caused by the higher rotational states) is anticipated. Considering Model 2 and assuming an energy difference of 1 kcal mol À1 between 12 a,b and 12 c, [34] for simple Boltzmann distributions at 503 and 563 K there would only be approximately 27 or 29 % respectively of 12 c. The IR spectra resulting from this distribution would be dominated by the spectra of 12 a,b, but would also contain, at lower intensity, the spectra of 12 c. [35] Given that DFT calculations tend to overestimate the stability of the delocalized forms of semibullvalenes and barbaralanes by up to 3 kcal mol À1 , [25] temperature alone cannot be responsible for the dramatic change in the IR spectra and structure of 12. Assuming an IR detection limit of 5 % of the major component under the vapor-phase conditions, for Model 3, 12 c must be at least 3.3 kcal mol À1 more stable than 12 a,b (a simple Boltzmann distribution calculated at 563 K gives ca.…”