Optimizing extraction of information from resonance lines in Fourier-transform infrared spectroscopy

The major analytic challenge in spectroscopy is to extract the maximum amount of information from an optical feature or features that are usually defined by a small number of data points.~ These in turn are often adversely influenced by both noise and the resolution function of the spectrometer. In dispersive-optical data the optimal approach for analyzing individual or closely spaced features is to Fourier-transform them and perform the analysis in reciprocal space, a procedure that requires the removal of endpoint-discontinuity artifacts.~ In Fourier-transform infrared spectroscopy (FTIR) the situation is more complicated in that the interferogram is measured and the spectrum calculated by Fourier analysis. We show that endpoint-discontinuity considerations provide insight into the effects of apodization on the Fourier transform of the interferogram.~ Once these effects are taken into account, individual structures can be analyzed as in the dispersive-optics case.