Development of an automated FTIR analysis framework for asphalt binders using asymmetric least squares smoothing and deconvolution

Fourier transform infrared (FTIR) spectroscopy is widely used to assess asphalt binder chemical composition and aging. This study evaluates commonly used FTIR analysis methods, including peak recognition with baseline, adjusted baseline, tangential area calculations, and fixed-range methods. A simplified curve-fitting deconvolution method was also developed to provide fast, user-independent FTIR analysis, which requires preprocessed spectra (baseline correction and normalisation). A total of 288 asphalt binder samples, subjected to various field and laboratory aging conditions, were analysed to calculate carbonyl (ICO) and sulfoxide indices (ISO). Results showed that the ICO and ISO effectively track short- and long-term aging. ICO was further used to evaluate the effectiveness of recycling agents in accommodating reclaimed asphalt pavement. While the peak recognition method offered the highest accuracy, the deconvolution method provided a fast, automated alternative. To facilitate FTIR analysis, AutoFTIR was developed as an open-source platform for efficient analysis and database management of FTIR spectra.