CALIBRATION OF AN FT-IR SPECTROMETER FOR AMBIENT AIR MONITORING USINGPLS

The objective of this work has been to develop a robust calibration me thod for simultaneous multigas detection with a Fourier transform infr ared (FT-IR) system. Calibration models for the identification and qua ntification of 23 gases in the presence of high concentrations of back ground gases such as water vapor, carbon dioxide, and methane have bee n obtained for an FT-IR instrument with 0.7-cm(-1) resolution. The cal ibration models have been tested on a breadboard instrument for trace gas measurement in manned space missions. The results show that FT-IR combined with multivariate methods such as partial least-squares (PLS) and proper pretreatment of the infrared spectra used in calibration i s well suited for this purpose. A procedure for baseline drift compens ation has been introduced to make the system insensitive to baseline d rift and variations in transmittance. This baseline drift compensation also reduces the need for background measurements. Further, a procedu re for incorporating a priori information about the instrument signal- to-noise ratio (SNR) and the absorption strength of interfering absorp tion lines has been developed. Indoor air monitoring and industrial pr ocess monitoring are other possible application areas for these techni ques. Parts of this work have been performed in a project for the Euro pean Space Agency (ESA) in cooperation with Kayser-Threde GmbH and Dai mler-Benz Aerospace, Dornier GmbH.