A solution to the azobenzene “entropy puzzle”
[J. Phys.: Condens. Matter201729314002] is provided. Previous computational studies of the thermal Z → E(back-)isomerization of azobenzene
could not describe
the experimentally observed large negative activation entropies. Here
it is shown that the experimental results are only compatible with
a more complicated multistate rotation mechanism that involves a triplet
excited state. Using nonadiabatic transition state theory, close to
perfect agreement is achieved between all calculated and experimental
Eyring parameters. We also provide new experiments that indicate the
presence of a noticeable external heavy-atom effect, which is a direct
result of spin–orbit coupling effects being important in the
proposed mechanism. These results suggest a reexamination of the mechanisms
of related thermal double bond isomerizations in other systems in
cases when an excited state of triplet (or other) multiplicity becomes
thermally accessible during a rotation process.