Progress in modeling complex dye mixtures: a case study for cultural heritage

Abstract

Unlike synthetic dyes that produce a single hue, natural dyes offer a rich spectrum of colors. In old times, artisans controlled these colors through their expertise, making the results dependent on each dyer's skill and subjectivity. Nowadays, modern computational methods allow precise predictions of visible absorption spectra, enabling a more standardized color control. However, most studies focus on individual molecules, while natural dyes are made of complex mixtures. We address this gap by proposing improved methods for predicting the spectra of a mixture of natural dyes, focusing on the madder plant, a renowned source of complex colorants. Molecular compositions were determined using liquid chromatography, and the visible spectrum of the mixture was computed, converted to standard color coordinates, and compared with experimental data. Our approach achieved high accuracy, with deviations within accepted theoretical error margins, and successfully handled large molecules with sugar groups. Our workflow can be used for other various chemical systems, notably dye mixtures.