Bitumen Characterization with Fourier Transform Infrared Spectroscopy and Multivariate Evaluation: Prediction of Various Physical and Chemical Parameters

摘要

As the aim of this work, the characterization of bitumen based on Fourier transform infrared spectroscopy combined with multivariate evaluation was investigated. For this, 90 bitumen samples were analyzed using the attenuated total reflection technology with multiple reflections. The gained spectra were divided into two partial spectra with the relevant peaks and pre-processed with a standard normal variate transformation and the first derivative. For a comprehensive evaluation, a multivariate analysis method in terms of the partial least square regression was used. On the basis of this procedure, models could be determined allowing for the description and prediction of different bitumen parameters, including the penetration, the complex shear modulus, the phase angle, and the asphaltene content. Further, the softening point can be estimated roughly. In contrast, the description of the flexural stiffness a well as the m value capturing the low-temperature behavior and the description of the content of the different maltene fractions are not possible. Concerning the structural relationships, increasing hardness, viscosity, stiffness, and elastic behavior are associated with an increase of the aromatic and oxygen-containing compounds and changes in the structure of the aromatics and alkenes.