Development of a GC–MS strategy for the determination of cross-linked proteins in 20th century paint tubes

In this paper we present the technical study of protein-drying oil mixtures in the paint formulation of original manufacturers paint tubes from Fabbrica di colori per Belle Arti Maimeri dating back to the 1950s. The study is part of a wider research project aimed at understanding the chemistry of protein-drying oil mixtures, and how the oil curing in the presence of proteins may affects the protein analysis with mass spectrometric based techniques. Tempera grassa, according to traditional Late Medieval and Renaissance painting techniques, entails the use, as paint binder, of a mixture of a drying oil with a water miscible medium, commonly egg. The paints analysed in this study were, according to technical documentation available at the manufacturer, based on a mixture of linseed oil, ovalbumin and mastic resin. All the components of the paints were detected but proteins proved very challenging. They could not be detected by pyrolysis coupled to gas chromatography-mass spectrometry, nor by using classical procedures for gas chromatography – mass spectrometry analysis of amino acids after hydrolysis. A new analytical procedure based on two subsequent hydrolysis steps assisted by microwaves was implemented and proved to be successful. The first step with trifluoroacetic acid was aimed at a partial hydrolysis of proteins into peptides, for their efficient extraction. The second with hydrochloric acid in the vapour phase was used for a complete hydrolysis of peptides into amino acids for the GC–MS analysis. Linking experimental results on the tempera grassa paints to data obtained on the curing of oil/protein mixtures in paint mock-ups, the hypothesis that copolymerisation of proteins and the oil has taken place is proposed, and a discussion on how this may affect protein detection is carried out.