Modern Data Processing Techniques for Fourier Transform Infrared Spectroscopy

Abstract

Quantitative analysis of mixtures by infrared spectroscopy can be considerably improved using modern interferometric techniques compared to classical dispersion methods. The improvement arises from two different benefits. First, the signal-to-noise ratios for the Fourier transform spectra are considerably enhanced and secondly, the digitized nature of the spectra allows the introduction of computerized data processing techniques. This paper will discuss the nature of these new data processing methods as utilized for the quantitative analysis of the mixtures. The logic pattern to be followed in developing the spectral methods is shown in Fig. 1. The FT-IR spectrum of the mixture is obtained and recorded as a digitized absorbance spectrum. The first question to be answered concerns the number of spectroscopically distinguishable components in the mixture. Two methods are available for making this spectral evaluation without chemical or physical separation of the components. If only two components are present, absorbance subtraction is simplest while the absorbance ratio method is more suitable for more complex mixtures [2, 3].