



While SERS provided unequivocal identification of laccaic acid A (the main component of lac dye), microspectrofluorimetry described the chromophore in its environment in situ. Recently, we have shown that microspectrofluorimetry and surface-enhanced Raman spectroscopy (SERS) may be combined to unambiguously characterize lac dye red in medieval illuminations. Our research has demonstrated that dyes are more specific of a certain cultural milieux than inorganic pigments, and we also know that they are prone to fading-for all these reasons, they have been at the centre of our research. Colourants (inorganic pigments and dyes) as well as binding medium formulations and additives were analysed. Through this material study, we have sought to understand the medieval techniques used to produce colour. More recently, these studies have also included books of hours as well as other artworks covering a period from the twelfth to the nineteenth centuries. Over the past 10 years, we have contributed to the development of advanced techniques that allow us to characterize and rigorously quantify colour and colour loss in monastic manuscripts from the twelfth and thirteenth centuries. Illuminated manuscripts are a precious cultural heritage with specific features that include a distinctive colour palette (when compared with other painted artworks). (a) The art of colour in medieval illuminations This article is part of the themed issue ‘Raman spectroscopy in art and archaeology’. Our findings in fourteenth–sixteenth century manuscripts confirm the hypothesis that this dye was lost during the High Middle Ages, to be later rediscovered. Orcein purple, also known as orchil dye, has been previously identified in medieval manuscripts, dated from the sixth to the ninth centuries. Complementing this identification, through microspectrofluorimetry and micro-Fourier transform infrared spectroscopy, it was possible to propose a complete paint formulation by comparison with our database of references the dark pink hues, in the three case studies, were produced by combining brazilwood pigment lakes and gypsum in a protein- and gum arabic-based tempera. To the best of our knowledge, this is the first time that brazilein, the main chromophore in brazilwood lake pigments, has been unequivocally identified by surface-enhanced Raman spectroscopy in an illuminated work (the Dourado Atlas). This was possible by combining Raman microscopy and surface-enhanced Raman spectroscopy with microspectrofluorimetry. All three works were on a parchment support. Brazilwood pigment lakes and orcein purple colours were unequivocally identified in illuminated manuscripts dated by art historians to be from the thirteenth to the fifteenth centuries and in the Fernão Vaz Dourado Atlas (sixteenth century). In this study, we successfully addressed the challenges posed by the identification of dyes in medieval illuminations.
