In both reactor systems, the temperature was controlled with an external, electrically

one might expect, each reaction exhibits a wavelength dependence, but the mechanisms foreach are not intuitively obvious. In all three examples, photochemical induction followedby thermal propagation results in higher conversions with fewer by- products when comparedto strictly thermal synthetic routes.IntroductionThere are numerous chemical reactions which can be initiated by laser irradiation. Formany of these examples, the laser is a convenient laboratory source for power and wave-length control, while the ultimate light source for an industrial process would be a metalarc lamp. The present work entails the optimization of three photochemical processes whichcould be scaled up to pilot plant scale reactions. In two of the examples, the commercialsource would be a controlled Hg arc lamp, while one example would probably employ a laserin a larger scale reaction.The examples which will be considered from the present work are all long chain free -radical reactions. Photon costs can be significant for photochemical processes, but chainreactions have an economic advantage in that each absorbed photon can result in the forma-tion of many molecules of product. Two of these systems involve chlorine atom initiatedchlorination or dehydrochlorination reactions producing l,l,l- trichloroethane and vinylchloride respectively. The third example is a photoinitiated oxidation of cumene producingcumene hydroperoxide. All three products are important commercial materials produced on alarge scale.ExperimentalThe gas phase photochemical reactor employed for the chlorine atom initiated reactionsis depicted in Figure 1. In both cases, the reactants were preheated and passed continu-ously through a reactor with quartz windows where they were irradiated with various lasersources. The reactor effluent was trapped and the products were analyzed by gas chromato-graphy /mass spectrometry and capillary gas chromatography with internal standards.