Invited article: Linearization and signal recovery in photoacoustic infrared spectroscopy.

Photoacoustic (PA) infrared spectroscopy enables the characterization of a wide variety of materials, affording the spectroscopist several advantages over more traditional infrared methods. While PA spectra are readily acquired using commercial instrumentation, the quality of the data can be improved substantially through the use of specialized numerical and experimental procedures. Two of these methods are the subject of this review. Specifically, this article describes (a) linearization of PA infrared spectra, a calculation that incorporates phase and amplitude information to extend the range of linearity for strongly absorbing samples, and (b) lock-in and digital signal-recovery procedures in step-scan phase-modulation PA infrared spectroscopy. Linearization yields significant improvement in band definition, especially in the low-wavenumber region. This numerical method succeeds in situations where the PA phase of the sample is less than that of the reference (carbon black). When this criterion is not met initially, the sample or reference interferograms can be manipulated prior to the calculation. The steps involved in linearization are illustrated in detail and approximations are discussed. Lock-in demodulation of the step-scan phase-modulation signal is compared to digital (software) demodulation in this study; the lock-in technique is found to be superior in several cases. The imaginary interferograms in these experiments sometimes lack a strong central feature, a situation that necessitates the application of less commonly used methods for phase correction and spectrum calculation. These methods, which are available in commercial software, include two-quadrant and stored-phase corrections. The PA phase spectrum resembles amplitude and absorption spectra when real and imaginary PA spectra are correctly calculated.