Some Comments Concerning Quantitative Infrared Spectroscopy

It is obvious to all consumers of infrared spectral data that the direct transference of quantitative intensities from one instrument to another would be a tremendous boon. It is equally clear that this transference is not possible at the present time [1]. For many reasons it is convenient to use the extinction coefficients at band peaks in quantitative analyses, but these coefficients are strongly affected by experimental conditions—notably the finite slit width. Unfortunately, the method of calculating the “effective” slit width from the mechanical slit width has not been established [2,3]. Recently, Russell and Thompson [4] have concluded that the effective slit width employed must be less than one fifth of the band width at half-peak optical density if the band-peak extinction coefficients are to be obtained. This means for spectra in solution a slit width of 1 cm⁻¹ which is considerably less than that obtainable with standard prism spectrometers. It is not surprising, in view of the above comments, that spectra taken of the same substance with instruments of different optical design are not superposable. However, Copelin's results [1] clearly indicate that instruments of the same make operating presumably under identical conditions yield extinction coefficients that are alarmingly dissimilar. It has not yet been demonstrated whether this variation is due to inconsistencies in the slit width or to other variations in the instrument. The development of a dependable “standard sample” is a critical step in unraveling the instrument problem.