Conical intersections and intersystem crossings explain product formation in photochemical reactions of aryl azides

Photochemistry of 3- and 4-methoxyphenyl azide at 266 nm has been studied by means of the complete active space self-consistent field (CASSCF) and multi-configurational second-order perturbation (MS-CASPT2) methods. Minima and interstate crossing points have been optimized using the CASSCF method. The calculations predict that the key step of the photolysis of both azides is a non-radiative process. However, an important difference is found when we compare the reactivity of both isomers of azide, deactivation of 3-methoxyphenyl azide (1) can occur via two reaction channels (internal conversion or intersystem crossing), which lead to formation of the dimer of 2H-azepine derivative (2a) and 3,3′-dimethoxyazobenzene (2b). In contrast, deactivation of 4-methoxyphenyl azide (3) takes place via a singlet to triplet intersystem crossing, which leads to formation of 4,4′-dimethoxyazobenzene (4). After initial deactivation, both isomers follow a cascade of surface crossings until they reach the final nitrenes, respectively. The reference active space for the two azides is 14 electrons in 13 orbitals and comprises the six π-type orbitals of the aromatic ring plus four σ-(N-N2) and five π-type orbitals of the –N3 moiety.