A specified study was made of the phenomenon of infrared-active nitrogen adsorption on nickel, the existence of which was first shown by Eischens and Jacknow. The study combines an infrared spectroscopic investigation of the adsorbed nitrogen with an electron-microscopic determination of the metal particle size distribution for the various nickel-on-silica samples. It is shown that this form of nitrogen adsorption is not limited to nickel but also exists on platinum and palladium (and, doubtless, also on other metals). However, this type of adsorption is measurable only if the diameter, d, of the metal crystallites is within the range ~ 15 Å < d < ~ 70 A. This is because crystals within this range possess a large number of special surface sites which are absent, or much scarcer, on larger or smaller crystals. This has appeared from a study of their surface structure in which use was made of marble models and of mathematical computations. The infrared activity of the nitrogen adsorbed on these special sites is due to the induction of a transition-moment by the strong polarizing field of the special sites. The observed absorption band is, in fact, the nitrogen Raman band (2331 cm −1) which has shifted under the influence of the Stark effect; in nickel, palladium and platinum it is observed at 2202, 2260 and 2230 cm −1 resp. Measurement of the heat and entropy of adsorption revealed that the initial heat of adsorption is about 12 kcal/mole and that the adsorption of the molecules is immobile at low coverage. Although the high heat of adsorption might suggest chemisorption, it is made plausible, i.a. in view of the non-specificity of the phenomenon to a particular metal, that this type of nitrogen adsorption is, in fact, physisorption.
Read full abstract