In the past years, infrared spectroscopy has proven to be our workhorse for studying the chemistry of oil paint. While taking a simple IR spectrum is routine for many chemists, we have made some progress in pushing the possibilities of IR spectroscopy to maximise the amount of information we can get about our samples.
With a heterogeneous material like oil paint, we are nearly always confronted with the problem of band overlap. Obtaining quantitative information on the concentration or evolution of chemical species in large datasets is only possible when you can accurately measure the intensity of each component band. Therefore, we are continuously developing tailored algorithmic methods that can automatically correct baselines, normalise spectra, subtract spectral features or fit the components of an IR band envelope (1).
Time-dependent IR spectroscopy
For the study of oil paint degradation, we do not just want to know which chemical species are in a paint sample, but also how fast the composition of that sample changes under various conditions. To make these time-dependent measurements possible, we built a custom sample cell in which sample films can be exposed to solvents, solutions or various atmospheres while being continuously monitored with ATR-FTIR spectroscopy (1,2). The resulting datasets can then be analysed, and used as input for mathematical models that describe reaction and diffusion processes in oil paint systems.
While many interesting questions about oil paint ageing can be answered with conventional IR spectroscopy, we are also exploring the application of 2-dimensional IR spectroscopy to the study of polymer structure. With pump-probe 2D-IR spectroscopy, the light absorption of a sample is compared before and after irradiation with a bright and short (<1 ps) IR laser pulse. The resulting data can give detailed information on the coupling of vibrational modes, the distance and angles between chemical bonds, and help with the identification of chemical species in strongly cluttered spectra (3).