On Path Length, Beam Divergence, and Retroreflector Size in Open-Path FTIR Spectroscopy

Abstract:

Open-Path Fourier Transform InfraRed (OP-FTIR) spectroscopy is an established technique used to measure boundary layer trace gas concentrations, e.g., in fenceline monitoring of industrial emissions. One limitation of OP-FTIR, especially in ambient environments, is the relatively high detection limit, i.e., ~4–7 ppb for benzene, toluene, and xylene (BTX) species. This detection limit is specified as the target gas spectral absorption feature exceeding the noise level by a factor of 3, and assuming a 450 m optical path difference (225 m separation between the retroreflecting array and the spectrometer). Common ambient volatile organic compounds (VOCs, e.g., isoprene, formaldehyde, monoterpenes) have lower detection limits of ~1–2 ppb, however, their ambient concentrations are also usually lower, making their detection difficult, except in conditions of enhancement. The detection limit is directly proportional to optical path, which controls target gas spectral absorption feature depth; however, depending on the specifics of the spectrometer and telescope optics, beam divergence can begin overfilling the distant retroreflector array for paths greater than ~300 m, leading to a loss of what would be long-path returning radiation at the detector. In this case, the absorption signature of the target gas increases, but the signal to noise ratio of the recorded spectrum does not, making detection difficult. We describe the results of an experiment where retroreflector array size was increased to collect a larger fraction of returning radiation at a one-way separations ranging from ~100 – 500 m (optical paths of ~200 – 1000 m). We present an analysis of both underlying spectra (e.g., signal to noise ratio, target gas absorption features) and also of quality indicators for retrieved concentrations (e.g., root mean square residual of spectral fit, retrieved concentration error) for a selection of VOC and other trace gas species. Finally, we will present quantitative information about the effectiveness of cleaning retroreflector arrays after extended field use, including spectral intensity changes and surface characterization.