ABSTRACT
States........................................................................... 359 6.3 The Photochemical [1,7] Hydrogen Shift in 1,3,5-
Cycloheptatrien .......................................................... 363 6.4 Zilberg and Haas’s Method for Locating Conical
Intersections .............................................................. 367
6.5 Semiempirical Structure Determinations of Conical Intersections.................................................. 374
6.6 Ab Initio Verification of Semiempirical Results....... 379 6.7 Triplet Photoreactions................................................ 384 6.8 Conclusions ................................................................. 388 Appendix .............................................................................. 389 References............................................................................ 389
6.1 INTRODUCTION
Semiempirical methods are generally quite useful in exploratory work, especially on larger systems.1 This is true not only for the ground-state chemistry, but also for excited-state phenomena such as photochemical reaction mechanisms, provided the semiempirical method used is flexible enough for a description of excited states. Excited-state potential-energy surfaces (PESs) can be determined with little computational effort, and stationary points such as minima, barriers, and, in particular, conical intersections2 can be located quite reliably by semiempirical MR-CI (multireference configuration interaction) calculations even for larger systems.3 Thus, such calculations may be quite useful in investigating organic photoreactions, as will be shown using the photochemistry of cycloheptatriene (CHT), cyclooctatetraene (COT), and 2H-azirine as examples to demonstrate various aspects of the procedure.
