Colour Analysis: An Introduction to the Power of Studying Pigments and Dyes in Archaeological and Historical Objects

Nature offers a myriad of colours and the desire to replicate them is intrinsic to human nature. People from past civilisations have, in fact, searched for natural sources of colour since prehistoric times and have looked for ways to apply them to surfaces and make them last as long as possible. These ancient people found out that some minerals could be ground to obtain a fine powder and that some plants, insects and molluscs yielded coloured solutions upon immersion in hot water. The former group of materials is referred to as mineral or inorganic pigments, whereas the latter group of materials is referred to as natural or organic dyes.

Pigments are completely or nearly insoluble in water and must be applied to surfaces by mixing them with a binding medium. Dyes can be applied to fibres in several ways in a water solution. The use of mordanting agents is often required. Additionally, organic dyes can be processed into insoluble pigments by precipitating them onto an inorganic substrate, thus creating the so-called lake pigments.

Since their discoveries and first applications, pigments and dyes have contributed to some of the most spectacular forms of art and craftsmanship, from dyed textiles to painted masterpieces. Painters, weavers, dyers, artists and craftsmen have developed their skills over millennia to respond to their own creativity, stimuli from the natural world, availability of local resources, as well as to everyday life needs, introduction of new materials, market requests and even fashion. Terms such as Tyrian purple, lapis lazuli and cochineal, to mention only a few, resonate in the collective imagination, evoking images of priceless royal dresses, tireless caravanserai and fearless explorers.

This constant evolution and exchange of knowledge between different peoples have led to a plethora of natural materials used to create and apply colour. These can drastically vary based on geographical regions and time periods, so that the study of these materials requires a high level of expertise and knowledge from the researchers tasked with their identification. The scenario becomes even more complex in the second half of the 19th century, when scientific advances revealed that colour was intrinsically related to the structure of organic molecules and to crystallographic arrangements in minerals. Driven by the desire to chemically replicate natural dye molecules, scientists embarked on a journey that led to the discovery of synthetic dyes in 1856, when William Perkin accidentally synthesised the coloured molecule mauveine (or aniline purple). By the end of the 19th century, more than 400 synthetic dye formulations were patented, and the numbers grew exponentially during the first decades of the 20th century, as the invention of new molecules accompanied and partially drove the industrial revolution. However, it is not correct to say that synthetic molecules replaced the natural ones, as there is evidence that traditional methods and materials continued to be used and are still in use today.

The 15 articles selected for this Special Issue represent a window onto this very complex research scenario, for which materials knowledge is often not sufficient to obtain the desired results. This leads to highly inter-disciplinary studies, in which scientists, conservators, historians, art historians and archaeologists join forces to answer research questions on colour.

Sotiropoulou et al.’s review [1] on the use of molluscan purple (or “true” purple—in Greek, porphyra = πορφύρα) as a pigment in wall paintings from the Aegean area, summarises the research carried out on probably the most prestigious organic colourant in history, and integrates it with new scientific evidence. The article reports on the use of calcium carbonate from ground mollusc shells as the substrate on which to precipitate the organic dye, thus pointing to the recycling of these shells, which were considered as waste products from the “purple industry”. Connections between textile dyeing and pigment production are also supported by archaeological evidence, as numerous loom weights were found in the sites under consideration. The wall paintings themselves, in which the purple pigment is used to depict several details in women’s costumes, implicitly reinforce such a connection. The review ultimately showcases how science, archaeology and art history come together to offer new keys for the interpretation of a topic that has fascinated generations of researchers and scholars.

Pigments in wall paintings are also the main subject treated by Rampazzi et al. in their investigation on 16th-century Italian wall paintings [2]. In addition to summarising evidence of a traditional palette of inorganic pigments, such as red and yellow ochres, ultramarine blue, bianco Sangiovanni, cinnabar/vermilion and azurite, other pigments, such as clinochlore, Brunswick green and ultramarine yellow were detected in the upper layers of samples that also showed complex stratigraphy. This is reported as the first evidence for the use of such pigments in wall paintings. Furthermore, as these pigments were all synthesised in the 19th century, their presence provides dating information for the execution of re-painting campaigns. The authors further the investigation, by discussing the ubiquitous identification of aragonite as an indication of the use of shells as an aggregate in the mortar, a practice that was known from the Roman period, but that has never been attested in later wall paintings. These results, together with the report on the poor conservation state of the paintings, enable a fuller picture to be drawn of the undocumented history of the church that houses them and will hopefully foster conservation interventions.

Sixteenth-century artist materials are also the topic of Balbas et al.’s contribution [3]. The study brings the reader to Mexico, and focuses on a polychrome maize stem sculpture. The identification of the pigments is only one part of a detailed investigation aimed at reconstructing the production/conservation history of the sculpture. Traditional Mexican materials, such as maize, paper made of cotton fibres, and colorín wood are present alongside materials traditionally encountered in polychrome objects of European origin (gypsum and animal glue for the ground; vermillion, cochineal lake and lead white for the red shades), thus reflecting the convergence of indigenous and European artistic traditions in this type of objects. The historical research and stylistic analysis led to hypothesise the exact workshop in which the sculpture was produced. Computer tomography and stratigraphic analysis of the polychromy highlighted original elements and alterations. All the observations come together in a captivating discussion.

In the studies introduced so far, samples were taken and analysed to answer some of the research questions. However, in some cases, sampling is simply not an option, as the works of art under investigation are too precious or too fragile to allow for samples, even very tiny ones, to be removed from them. This is particularly true for artworks on parchment and paper, as highlighted in the contributions by Agostino et al. [4] and Dill et al. [5], respectively. The development of non-invasive and portable spectroscopic techniques is one of the most active research areas in heritage science. Agostino et al. show how an analytical workflow based on optical microscopy, fibre optic reflectance spectroscopy (FORS), fibre optic molecular fluorimetry (FOMF), X-ray fluorescence spectrometry (XRF) and micro-Raman spectroscopy can unveil the palette used to decorate Medieval illuminated manuscripts [4]. This information was then used to underline the fact that certain pigments and colourants were used in a hierarchical way, thus providing further insight into the modus operandi of the artist. On the other hand, Dill et al.’s contribution showcases the potential of non-invasive imaging techniques as opposed to point analysis [5]. Macro-XRF (MA-XRF), hyperspectral imaging (HSI), photometric stereo imaging and transmitted light imaging were used to investigate three hand-coloured prints showing the work of artist-naturalist Maria Sibylla Merian (1647–1717) and answer questions related to the attribution and dating of the prints. The observations enabled the prints and paper to be attributed to a posthumous edition of the artist’s work published as a French translation in 1771, whereas the colouring was carried out after 1920, as shown by the presence of cadmium red and titanium white pigments, probably by an art dealer or collector trying to add value to the prints.

Modern pigments and colourants are the topic of several articles in this collection. These materials represent a challenge for scientists and conservators for many reasons, including the limited information available on their chemical composition and degradation. Hence, characterisation studies, such as the contribution by Pause et al. [6], are fundamental to building databases and enhancing our knowledge of these materials. In their study, the authors test the use of a handheld Raman device for the non-invasive identification of synthetic organic pigments in historic pre-1950 varnished paint-outs from manufacturer Royal Talens. Historic books and catalogues from colour manufacturers represent an important source of information, as highlighted in other studies as well [7]. In fact, the confusion around nomenclature and chemical formulae of synthetic organic pigments and dyes is a challenge in itself. In addition to reporting on the feasibility of the identification technique, the article contributes to a better understanding of artists’ materials used between 1932 and 1950.

Building this foundation of knowledge is the first step towards more focused studies, such as the ones reported by Lizun et al. [8] and Martins et al. [9]. Lizun et al. take on the complex task of exploring the painting technique and materials adopted by Singapore artist Liu Kang during his period in Paris (1929–1932), by analysing 14 paintings with a wide array of non- and micro-invasive analytical techniques [8]. The detailed technical examination is coupled with historical information derived from archives and contemporary catalogues. The research leads the authors to reveal the use of a relatively restricted palette of pigments, such as ultramarine blue, viridian green, chrome yellow, iron oxides, organic reds, lead white, and bone black, with some additional pigments used sporadically, such as cobalt blue, Prussian blue, emerald green, cadmium yellow, cobalt yellow, and zinc white, while the painting technique was in line with the exploratory phase of the artist, who was developing his own style during this early stage of his career.

Martins et al. focus on an iconic work by Henri Matisse with the intention of identifying the pigments used in the gouaches of the illustrated book, “Jazz” [9]. Three different copies were investigated, and the results were consistent, revealing the use of 39 pigments, including both mineral and synthetic organic pigments. The article also discusses the use of multivariate statistical analysis for the treatment of large datasets and the limitations of certain non-invasive techniques for the identification of organic molecules. But most importantly, the research was able to inform on the condition of the prints, in order to support conservation strategies. In particular, similar colour shades were obtained using very different materials with different light sensitivities, thus impacting conservation decisions.

In Haddad et al.’s contribution, Alexander Calder’s “Man-Eater with Pennants” is at the centre of an investigation to reveal the original paint colours hidden beneath layers of repainting [10]. In addition to showing an application on a sculpture rather than a two-dimensional object, the article reports on the difficulty of distinguishing original materials and subsequent applications in modern objects, and on the different degradation pathways occurring during outdoor exposure. The original paint layers were identified as containing Prussian blue, parachlor red and chrome yellow, whereas the many layers of overpaint contained titanium white, molybdate orange, several β-naphthol reds, red lead, and ultramarine blue. Interestingly, the initial intention to use a maquette model of the sculpture to confirm the original palette of the artist was discarded, as the analyses showed that the maquette was painted after “Man-Eater” was first installed. The results ultimately informed the restoration campaign that led to the sculpture once again being exhibited in the Museum of Modern Art’s sculpture garden in New York City.

The implication of scientific investigation on conservation and display decision-making is also treated in Rayner et al.’s contribution [11], which focuses on an ancient Greek terracotta krater treated with selective in-painting with cadmium orange (CdSSe). Some of the restauration areas showed alteration after one year from the application of cadmium orange. These areas were associated with the presence of chlorine, and, upon recreation of the stratigraphy in mock-up samples, it was observed that the cadmium orange degrades in the presence of chlorine and light, forming selenium-rich structures. The research provides new information on the degradation of this pigment and led to the removal of the altered paint layer and re-treatment of the object with non-cadmium containing pigments.

Modern synthetic dyes are discussed in Ciccola et al.’s contribution [12], which reports on an interesting case study focusing on the analysis of two Indian leather puppets made in the 1970s. The spectroscopic analysis by reflectance spectroscopy and surface enhanced Raman spectroscopy (SERS) led to the identification of malachite green dye, crystal/methyl violet, rhodamine B and eosin Y, although some azo dyes and some complex mixtures were also present but difficult to disclose using only a spectroscopic approach. In fact, organic dyes often have complex molecular compositions that require the integration of a chromatographic approach, in order to separate the molecular components and accurately identify the dye source(s). Doherty et al. address this point in their contribution [13]. Following the recent structural elucidation of urolithin C, one of the main degradation products of the brazilwood dye [14], this article describes the suitability of SERS to identify this molecular marker in a variety of textile samples. The correlation of the data obtained by SERS with those obtained by high performance liquid chromatography (HPLC) with both a diode array detector (DAD) and a mass spectrometry detector (MS) enabled the results to be confirmed. Additionally, other dyes were identified in some of the samples, highlighting the possibility to detect brazilwood by SERS even in the presence of other dyes.

Weld, another well-known natural dye, is the focus of the contribution by Veneno et al. [15]. The authors made reconstructions of five 19th-century recipes of weld lake pigments discovered in the Windsor & Newton archive database, and then characterised them by FTIR and HPLC-DAD. The article discusses the steps necessary for the commercial preparation of the pigments, provides new information on FTIR bands specific for the identification of weld lake pigments, and shows the effects of the different matrices deriving from the different manufacture of the pigments.

Finally, specific attention to dyed textiles is given in the contributions by Campos Ayala et al. [16] and Łucejko et al. [17]. In their article, Campos Ayala et al. address the characterisation of the dyes used to produce primary (red, blue, and yellow) and secondary (purple, orange, and green) colours in Peruvian textiles spanning five major civilizations: the Paracas Necropolis, the Nazca, the Wari, the Chancay, and the Lambayeque [16]. With the intention to validate the use of ambient ionisation MS techniques, such as direct analysis in real time (DART-MS) and paper spray MS, the authors report on the results obtained from both reference samples and archaeological textiles, thus providing the reader with an important database. Comparative results obtained by HPLC-DAD integrate the results and support the conclusions. The article clearly shows the variety of dye sources available in the Andean region and the difficulties in identifying yellow dyes, once again underlining the complexity of conducting this type of research, as the biodiversity of the natural world becomes a challenge in itself. From South America to Northern Europe, the contribution by Łucejko et al. addresses the investigation of textile fragments, including rare silk embroideries, from the Viking Gokstad ship grave in Norway, dating to ca. 900 AD [17]. The severe fading of the colours made the analytical work challenging, but thanks to the application of HPLC-MS/MS, molecular markers for the presence of madder (probably from Rubia tinctorum) were detected in most samples. Additionally, hypotheses are made about the possible use of different dye recipes to obtain different colour shades (from yellowish to red) using only madder as dye source.

The overview provided by this selection of articles shows only a small fraction of the potential of colour analysis in heritage science. Analytical challenges related to the growing need for non-invasive analyses are further driving technological developments in this research field. However, obtaining accurate molecular information from samples is sometimes the only way to provide the desired answers. Through scientific analysis, our knowledge of pigments and dyes, as well as their degradation, grows year on year. As a result, researchers are offered new tools to address their questions in a stimulating and cross-disciplinary environment.

I hope that this collection of articles will inspire future research on pigments and dyes to go towards the direction of not just looking at colour in terms of wavelengths, electronic transitions, chemical components or molecular interactions, but to focus more on the artists and craftsmen who made and used colours both in ancient and more recent times. Switching the point of view from the colours themselves to the hands and minds that produced them, processed them, experimented with them, and applied them has the potential to unlock new answers, and certainly new questions, on this fascinating research topic.