DISSOCIATIVE RECOMBINATION OF POLYATOMIC IONS

The experimental data of Adams et al on products of the dissociative recombination of polyatomic ions are discussed and rationalized in terms of the valence bonds concerned. By combining the data with theoretical considerations branching ratios are derived. There are two main classes of saturated polyatomic ions: those for which the number of valences of the ionized atom is decreased by neutralization and those for which the number is increased. The hydride ions H3O+ and H4N+ are simple examples of the valence decrease class. It is reasoned that while H3O+ may shed two H atoms in dissociative recombination with thermal energy electrons H4N+ cannot do likewise because of a difference in the OH and NH2 relic radicals. The prediction is made that a bond between a pair of neutral atoms in an ion like CH3O+H-2 cannot be broken by dissociative recombination. Although the number of valences of an isolated C+ ion increases on neutralization the dissociative recombination of polyatomic ions containing C+ normally proceeds through a radiationless transition that causes a decrease. For example when CH3+ sheds an H atom by dissociative recombination the CH2 relic radical is mainly in the 1A1 metastable state with the C atom having two paired orbitals (a 'lone pair') so that it is effectively divalent. In contrast to CH3O+H-2 dissociative recombination, CH3C+H-2 dissociative recombination may break a bond between neutral atoms. In this event ethylene would be formed; but the branching ratio for the channel is not predicted mainly because of uncertainty in the Franck-Condon factor. The thesis is advanced that aside from the Franck-Condon factor (which depends on the position of the crossing and hence on the exothermicity of the channel) the rate of the radiationless transition involved in breaking a bond between an ionized atom and a neighbouring atom of a particular species is likely to be approximately independent of the other constituents of the polyatomic ion. It is pointed out that the distribution of the measured dissociative recombination coefficients of polyatomic ions has a cluster such as would be expected if they were constrained by having an upper limit (possibly having a value that depends on whether the ionized atom is tri- or tetravalent). This is rationalized by use of the maximum cross section theorem.