White, Friend or Foe?: Understanding and predicting photocatalytic degradation of modern oil paintings

This dissertation presents a study into the ultraviolet irradiation-initiated degradation phenomena occurring in titanium white containing oil paints, commonly referred to as photocatalytic degradation. The topic of this thesis can be summarized as: the (photocatalytic) properties of titanium white pigments in oil paints and their consequences for collections containing modern art. The thesis consists of three parts: 1) characterization of the use and properties of titanium white pigments, 2) understanding and monitoring the degradation of titanium white containing oil paints and 3) predicting degradation caused by titanium white pigments. Combining the results of these three parts leads to a risk management strategy for modern art collections presented as a conclusion of this thesis. Titanium white pigments were introduced in the 20th century as an alternative for lead white and zinc white. The pigments underwent a gradual development resulting in a large variety of pigments available throughout history. This variety ranges from very photocatalytic or ‘bad’ pigments, which severely speed up degradation, to photostable or ‘good’ pigments which can protect their environment from UV irradiation. Both ‘good’ and ‘bad’ pigments found their way into artist materials and thus into paintings. Hence the question ‘Titanium white, Friend or Foe?’. The pigment’s photocatalytic activity is highly dependent on the pigment’s crystal structure (rutile or anatase) and (inorganic) surface treatment. When a pigment is photocatalytic, radicals can form upon UV irradiation, which attack the oil binding medium and break it down to volatile components, leading to an effect called chalking: the pigment is unbound on the paint surface. Interestingly, while we know that ‘bad’ TiO2 pigments were used in modern oil paints, photocatalytic degradation problems have not been widely reported thus far. Several hypotheses for the lack of problems in collections are presented and investigated in this thesis. Additionally, characterization tools, predictive tools, and monitoring tools are developed, in order to provide solutions before it is too late. Simultaneously, analytical methods and research approaches, that are uncommon in the field of conservation science are explored to evaluate their applicability in answering cultural heritage research questions. (OLD) MSE-