Mechanistic Studies of the Autoxidative Curing of an Oil-Based Paint Model SystemPublic Deposited
Microkinetic modeling is a powerful tool for creating dynamic and quantitative descriptions of complex systems. These detailed mechanistic models compliment experimental techniques and provide an ability to achieve deeper insights into chemical processes where numerous intermediates are highly reactive and difficult to quantify in the laboratory. This thesis discusses the development of a microkinetic model for the autoxidative curing of an oil-based paint model system relevant to the cultural heritage science community. Grasping the mechanisms of the formation and resulting composition of crosslinked and non-crosslinked species from autoxidative processes is critical to understanding and predicting the long-term chemical and physical stability of painted objects. The fatty acid ester studied in this work, ethyl linoleate (EL), is frequently used as a model system for oil paint binders. Consequently, quantitative experimental data was available for model validation and comparison to theoretical predictions for a few key metrics such as the peroxide content and oxygen absorption of the system, as well as the evolution of small volatile molecules such as hexanal and pentanal. The mechanism and parameters for autoxidation were assembled from a variety of literature sources. This preliminary microkinetic model of EL curing catalyzed by cobalt-2-ethyl hexanoate (Co-EH) up to the formation of single crosslinked species revealed that the mechanisms governing the formation of the volatile species hexanal were still not well understood and that a much larger reaction network would be required to thoroughly describe the cured system. An in depth study of mechanistic postulates for the formation of hexanal was conducted and β-scission of higher rank oligomeric products was demonstrated to be an effective route to form these volatile products at the mild reaction conditions of interest. Gas Chromatography/Mass Spectrometry (GC/MS) headspace analysis was also conducted to provide quantitative targets for hexanal and pentanal production in the early cure regime (< 24 hours) which were previously unavailable. Quantum chemical calculations were performed to derive a structure-reactivity relationship for the formation of experimentally observed epoxide species. A wide range of chemical space was explored for the first time, including observations for allylic and benzylic radical reactants. This work revealed that the reactant radical type can have a strong effect on the calculated rate parameters and that separate kinetic correlations are required to treat saturated and unsaturated species accurately. A second generation model for EL autoxidation was constructed incorporating these new mechanistic insights and used computational automation to generate the reaction network. This larger microkinetic model represented a significant improvement in capturing experimentally observed products at longer curing times (up to 100 hours). The larger model was also able to capture accelerated rates of curing in response to increased temperature and Co-EH concentration. Building a detailed microkinetic model for an oil-based model system expanded understanding of the underlying chemical mechanisms involved in the formation of stable paint films.