The Near Infra-Red Absorption Spectra of Some Organic Liquids

Near infra-red absorption spectra of thirty-one organic liquids to 2.5μ.— $A$ self-recording prism spectrograph, Littrow mounting, with dispersion equivalent to two 60° flint glass prisms—considerably greater than had hitherto been used in this region—was arranged so-that the spot of light from the mirror of the galvanometer in the thermopile circuit, was focused on a sheet of bromide paper on a drum rotated in gear with the prism table. Curves thus obtained are reproduced which show the transmission through 11 mm of certain organic liquids: ethyl, methyl, propyl, butyl, ethylene and methylene halides, butyl nitrite, benzene, toluene, pentane, hexane, heptane, octane, mesitylene, paraxylene, chloroform and bromoform. In a table are given the wave-lengths of the absorption maxima accurate to about.01μ, and, in many cases, the percentage absorption, accurate to five per cent. Curves obtained by visual observations, are also given for eight of the above compounds and for acetone and ethyl ether. Bands near 0.90, 1.02, 1.17, 1.38 and 1.70μ occur in most if not all of these spectra. Hence these bands are attributed to the $C-H$ bond, and appear to be, respectively the 7th, 6th, 5th, 4th, and 3rd harmonics of a fundamental frequency found by Coblentz at 6.9μ. There are also bands at 1.81, 1.91, 2.00 and 2.10μ which occur for several liquids. Shifts in the positions of the bands for different liquids, which may amount to ±.03μ, are not understood, but may be due to a change in the relation intensity of the two components of the band; for some of the bands are clearly double, like the infra-red bands due to gaseous compounds. The simple quantum theory of such gaseous spectra does not seem to hold directly for these liquid spectra, however, since the separation of the components of 1.38μ and also of 1.7μ is greater for the iodides than for the corresponding chlorides.