The Occurrence of Harmonics in the Infra-Red Absorption Spectra of Diatomic Gases

Abstract

The study of the infra-red absorption bands of gases has within the last few years received a great impetus through Bjerrum'st explanation of their structure as the result of the superposition of the molecular rotations on the vibrational motion of the charged atoms, which is the primary cause of the absorption. Bjerrum's suggestion was quickly followed by the discovery of v. Bahr2 that these bands are, in reality, clusters of comparatively narrow lines, corresponding to the discrete angular velocities predicted for the molecular rotations by the older form of the quantum theory. In a recent number of the Physical Reviews Kemble has called attention to the fact that if the energy of the vibrational motion also takes on only the quantum values the corresponding infra-red absorption bands in the spectra of diatomic gases ought to be accompanied by faint harmonics. An examination of the literature showed only one case, viz., carbon monoxide, in which more than one band was known to exist and this was a faint one at very nearly half the wave length of the fundamental. The present article is a preliminary report on an experimental search for similar harmonics in the infra-red absorption spectra of other diatomic gases and an examination of their structure. First harmonics have been found in the spectra of hydrochloric acid and hydrobromic acid and their structure as well as that of the faint carbon monoxide band has been shown to be in general agreement with the theory. Search was also made for possible second and third harmonics in the spectra of carbon monoxide and hydrochloric acid, but without result. Apparatus and Method of Procedure.-The work was done on a prism spectrometer with a combination Wadsworth-Littrow mounting similar to that described by Gorton.4 This arrangement gives an unusually high resolving power by a double passage of the energy through the prism. The energy was measured with a thermopile. The prism was of quartz, the refractive indices given by Carvallo and Rubens5 being used in computing the wavelength scale. The absorption tube was 78 cm. long and had a minimum inside diameter of 3.5 cm. The ends were closed with windows of fused quartz. Each galvanometer deflection with the absorption tube in place was 'sandwiched' between two deflections with the tube drawn to one side.