FTIR spectroscopy using a thermal light source has been the dominant method for obtaining infrared spectra since the 1950’s. Unfortunately the limited surface brightness and low spatial coherence of black-body radiators limits the spectral SNR in microspectroscopy and stand-off detection. Two recent innovations are addressing this problem a) FTIR instruments illuminated by high-spatial coherence broad-band supercontinuum sources and b) high spatial coherence narrow-band EC-QCL’s.
Here we ask whether these two approaches offer equivalent sensitivity. By noting an analogy with near-infrared optical coherence tomography we rigorously show that the high temporal coherence of the EC-QCL brings an additional, very large SNR advantage over an FTIR instrument illuminated by a supercontinuum source under otherwise matched conditions. Specifically if a spectrum containing N points is recorded by both instruments using the same illumination intensity and the same detector noise level, then the EC-QCL can deliver a given spectral SNR in a time xN shorter than the FTIR instrument. This factor can reach x100, potentially even x1000, in realistic applications.
We exploit the analogy with OCT further by developing a mid-infrared “swept laser”, using commercially available components, in which the tuning rate is much higher than in commercial EC-QCL devices. We use this swept laser to demonstrate the SNR advantage experimentally, using a custom-made EC-QCL spectrometer and PDMS polymer samples. We explore the potential upper limits on spectral acquisition rates, both from the fundamental kinetics of gain build-up in the external cavity and from likely mechanical limits on cavity tuning rates.
|