Collisional deactivation, inefficient

It is apparent that two collisional deactivation processes must be working in competition—one a very efficient removal of some intermediate which yields nonstereospecific products, and the other a relatively inefficient production of some other intermediate which yields nonstereospecific products. The very efficient process, which produces a sharp drop in the amount of tripletlike products at low pressures of inert gas, undoubtedly involves collisional stabilization of the hot cyclopropane formed by singlet addition, while the less efficient process would seem to involve intersystem crossing to triplet methylene, just as Anet and Frey proposed. 

Previously, various aspects of the specific dissociation probabilities, kt, were described for simplified conditions. Since, in practice, rnono-energetic systems are rarely achieved, a direct measurement of kt is usually not made, and it is necessary to relate the measured quantities to the theoretical quantities. In nearly all cases (mass spectra are an exception), experimental results depend to some extent upon the competition between collisional deactivation and reaction. Some of the conclusions derived from the formulation depend upon what allowance is made for weak or inefficient collisions. All available evidence indicates that, provided the deactivating molecule is not small, most collisions lead to deactivation. The strong collision formulation will be used the behavior expected if this restriction is relaxed is described in Sec. II-E,3. 

Several anomalies appear to exist in these data. First, the temperature dependence of rate seems excessive for the last group. Second, the diethyl ketone rates are much higher than those in the ethylene system and led Heller and Gordon to attribute greatly reduced collisional deactivation efficiency to the ketone relative to ethylene. If anything, however, we believe that the efficiency inequality should be reversed. Finally the experimental k values were calculated on the assumption that hydrogen or deuterium gas present in the mixture was completely inefficient as a collisional deactivator. We believe that this assumption is too extreme and that a reasonable lower estimate17 of their collisional efficiency would be 0.20. On this basis, all rate constants would be doubled or tripled. 

In the oxygen system at approximately 50 mm. pressure (collision frequency — 109 sec. 1) half of the 02 " ions are stabilized before emission can take place (13). In the condensed phase, therefore, deactivation should compete to the exclusion of electron emission. The much higher probability of collisional deactivation in liquids may explain why compounds such as C02 and CH3C1, for which attachment is very inefficient in the gas phase, are often effective electron scavengers in liquid systems. One must be wary, therefore, of using even relative gas-phase electron attachment coefficients in liquid-phase studies. For molecules with very small electron affinities (< 0.1 e.v.) the reversibility of Reaction 3 may have to be considered even after the excitation energy of the negative ion has been removed by collision. 

---