Robust cepstral analysis at variable wavelength scan depth for narrowband tunable laser absorption spectroscopy

Abstract

This work focuses on optimizing the modified free induction decay (m-FID) method, or cepstral analysis, for accuracy, precision, and robustness in performing laser absorption spectroscopy at conditions of variable baseline distortion where laser tuning range is of a similar order of magnitude as the targeted spectral linewidth. The optimum selections of the initial and final time of the m-FID signal were assessed at variable scan indices (ratio of laser scan range to spectral linewidth), and guidance is offered in parameter selection based on scan index and noise characteristics of the optical system. The sensitivities of the m-FID method to multiple types of baseline distortion were also analyzed to reflect typical experimental factors such as beam attenuation, background radiation, and laser stability. Experimental demonstration of the technique was performed with an interband cascade laser targeting the CO2 transition (00010000) R(60) near 4.2 micron to measure gas concentration at variable scan indices in a static cell by using both the m-FID and direct absorption spectroscopy (DAS) method while varying injection current amplitude. An estimated baseline was determined by the measured laser output and superimposed with different types and levels of distortion. Under conditions of signal distortion, the m-FID method is shown to demonstrate superior accuracy and precision with smaller deviations in time than that rendered by typical DAS, especially at smaller scan depths, showing its potential for measurements at elevated pressures or at a high scan frequency where distributed feedback lasers typically yield narrower tuning range.