Application of Infrared Absorption Spectra at Low Temperatures to the Analysis of Hydrocarbon Mixtures

As is known, the difficulty of chemical analysis of hydrocarbon mixtures makes it technically advantageous to determine the chemical composition from infrared spectra. The simplicity, accuracy, and speed of analysis of multicomponent systems depend above all on the establishment of strong absorption bands, each of which would be characteristic of only one component and would lie in a region where absorption by the other components is absent.

The authors of the paper under review*) draw attention to the fact that a decrease in line width (and, upon a change in temperature to −195° C, the band width should decrease by half) increases the probability of finding non-overlapping absorption bands belonging to different components. In order to determine quantitatively to what extent the decrease in line width facilitates the problem of analysis, the authors proceed as follows. It is conventionally assumed that analysis in the infrared spectrum can be carried out in the case where at least one band of each of the components does not overlap with the bands of the other components. To make an approximate calculation, it is assumed that the bands of each of the components are distributed uniformly—

*) W. H. Avery and J. R. Morrison, J. Appl. Phys., 18, 961 (1947).

uniformly in a spectral region of some width (500–1300 cm\(^{-1}\)); that the widths of the bands and their intensities are all identical, and that each of the components has the same number of bands. Under these simplifying assumptions, the probability of the possibility of successfully carrying out the analysis in the sense indicated above can be calculated. The calculation gives an interesting result. It turns out that, when the band width is changed from 40 to 20 reciprocal centimeters for a system of 18 components, each of which contains 15 bands, the probability of being able to carry out the analysis in the infrared spectrum changes from 0.003 to 0.999. The authors of the article are fully aware of the crude simplification made in the calculation. They believe, however, that the order of these figures will remain correct for the real case. Hence the practical conclusion is drawn: in the analysis of complex mixtures, exceptionally great advantages are obtained by investigation at liquid-nitrogen temperature. Obviously, the possibilities of carrying out the analysis increase still more at liquid-helium temperature. If the law of reduction of band width (inversely proportional to the square of the temperature) remains unchanged, then the bands should narrow sevenfold. The authors believe that the advantages obtained may outweigh the difficulties of work even at such low temperatures.

The experimental part of the work consists in obtaining and comparing with one another the infrared spectra of four trimethylpentanes at temperatures of 0°C and \(-195^\circ\)C.

A film of hydrocarbon was condensed on the mirror bottom of a cylindrical vessel. This vessel was immersed with its lower end in crushed ice or liquid nitrogen. By means of a mirror, the radiation was directed to the bottom of the vessel, reflected from it, and, with the aid of the same mirror and other mirrors, directed into the spectrograph. The authors point out a whole series of advantages of this method of investigation, even if one is not seeking to obtain especially low temperatures. The spectra presented by the authors speak eloquently in favor of work at low temperatures.

A. K.