Tag Archives: Painting

The Many Shades of Naples Yellow: Experimental Re-Working and the Influence of Fluxes on Pigment Colour

By Umberto Veronesi, Mario Bandiera, Marcia Villarigues, Andreia Ruivo, Marta Manso,  Susana Coentro

The ChromAz project tells the story of sixteenth- to eighteenth-century Portuguese azulejos from the perspective of colour technology. Tiles are a crucial piece of Portuguese national heritage, and artists had a broad palette of colours at their disposal. We started with one of the most complex to achieve: a pigment known as Naples yellow (lead antimonate), a product of human craft with a long history.

A panel of tiles show a pattern of blue, yellow, and white.
Figure 1. A panel of tiles that shows different shades of yellow. Image courtesy by Umberto Veronese and courtesy of Museu Nacional do Azulejo, a partner in our project.


As one of the very first artificial materials, Naples yellow was employed from the Late Bronze Age (ca. 1500 BC) and initially linked to the production of opaque yellow glass. After disappearing from the European tradition for over a millennium, the pigment resurfaced in the sixteenth century, used by Venetian glassmakers, in oil paintings and in decorated glazed ceramics.

Around this time, technical texts begin to feature recipes for the manufacture of Naples yellow. Besides the lead-antimony base, authors list a number of other reagents, but only vaguely (if at all) describe the chromatic effects brought by the additions of these ingredients. So, to gain insights into such an important aspect of the early modern artistic process, we replicated eight recipes, summarised in Table 1.

A table that summarizes eight experiments with historical pigments.

Four recipes (Mariani I and II, Piccolpasso and Marmi 126) are simple, binary lead antimonates to which salt or tartar (or both) were added as fluxes to facilitate the reaction. The remaining four (Mariani III, Marmi 136, Darduin and Danzica) are ternary variants of the pigment. They can come with or without fluxes but, importantly, they contain what the authors refer to as tutty or tuccia, which scholars have variably interpreted as either zinc or tin oxide. To evaluate differences, we made two sets, one with the former and one with the latter, bringing the total to twelve re-worked recipes.

After thoroughly mixing the ingredients, we fired the raw pigments at 950°C for five hours and left them in the furnace to cool down to room temperature. Finally, we collected the mixtures, which had turned yellow, and ground them to a fine powder to homogenise the colour (Figure 2).

Two top panels show ground pigment powders, in grey and amber tones. Two panels below show pigments after firing, in vivid yellow and amber tones.
Figure 2. Top: some of the pigment mixtures before (left) and after (right) firing; bottom: two mixtures being ground showing very different colours due to recipe variations. Left: Mariani’s potters’ yellow III with tin; right: Danzica with zinc (Photos by the authors).


As the images show, variations in the recipes resulted in remarkably different hues, ranging from pale to bright yellow and all the way to dark orange. Adding zinc tends to darken the colour, while tin makes it lighter, as previous works already indicated (Figure 3A). The one yellow with calcina (lead and tin calcined together) is also made somewhat paler, due to the action of tin (Figure 3B).

Five cylinders show different hues of historical recipes, all of which are yellow or amber in tone.
Figure 3.A: The recipe by Valerio Mariani in the binary version (left), with zinc (centre) and with tin (right), showing different hues; B: Two very similar recipes showing how the addition of calcina (left) makes the pigment lighter (Photos by the authors).


However, it was more interesting to find out that fluxes have an equally crucial role in defining the chromatic characteristics of the pigment. We found out that salt invariably makes the pigment lighter in colour. On the other hand, the recipes containing tartar as the main flux display a darker, reddish hue (Figure 4).

A panel of five images shows the different yellow tones of the pigments, depending on the salt and tartar content.
Figure 4. Five Naples yellow recipes with different amounts of salt and tartar, showing the general lightening effect of the former and the darkening effect of the latter. From left to right: Mariani I (binary), Mariani III (with zinc), Marmi 126, Piccolpasso and Mariani II. (Photos by the authors).

Re-working lead antimonate recipes provided access to the wide chromatic palette that Renaissance artists could rely on. The pigment could be adjusted to be proper yellow or more orange-red, depending on users’ needs. Importantly, this flexibility also meant that recipes could be tweaked to accommodate issues of materials’ availability. Our work is still very much in progress, but it shows how performative methods complement information from texts and material culture and give them new context, throwing more decisive light on the act of art-making in the past.


Key references

Dik, J., Hermens, E., Peschar, R. and Schenk, H. “Early production recipes for lead antimonate yellow in Italian art”, Archaeometry 47, 3 (2005): 593-607.

Hermens, E. “A seventeenth-century Italian treatise on miniature painting and its author(s)”, in A. Wallert, E. Hermens and M. Peek (eds), Historical painting techniques, materials and studio practice: preprints of a symposium, University of Leiden, the Netherlands, 26–29 June 1995. Marina Del Rey (CA): Getty Conservation Institute, 1995: 48–57.

Marmi, D. The Ceramist’s Secrets. Edited by Fausto Berti. Translated with an introductory note by David P. Bénéteau. Montelupo Fiorentino: Aedo, 2005.

Moretti, C., Salerno, C.S. and Tommasi Ferroni, S. Ricette Vetrarie Muranesi. Gasparo Brunoro e il Manoscritto di Danzica. Firenze: Nardini Editore, 2004.

Piccolpasso, C. The Three Books of the Potters’ Art. Vendin-le-Vieil: Editions la Revue de la céramique et du verre, 2007.

Rosi, F., Manuali, V., Miliani, C., Brunetti, B.G., Sgamellotti, A., Grygar, T. and Hradil, D. 2008. “Raman scattering features of lead pyroantimonate compounds. Partt I: XRD and Raman characterization of Pb2Sb2O7 doped with tin and zinc”, Journal of Raman Spectroscopy 40(1): 107-111.

Wainwright, I., Taylor, J.M. and Harley, R.D. “Lead antimonate yellow”, in R.L. Feller (ed.), Artists’ Pigments. A Handbook of their History and Characteristics, vol. 1. Cambridge: Cambridge University Press, 1986: 219–254.

Zecchin, L. Il Ricettario Darduin. Un Codice Vetrario del Seicento Trascritto e Commentato, Venezia: Stazione Sperimentale del Vetro, 1986.

How to avoid a bad buy and angry patrons: a recipe for pigment testing

As we have seen in other blog posts before, recipes for counterfeits, imitations, and ersatz products were fairly common in the early modern period. On the other hand, there were recipes to detect the results of such deceit, such as this recipe to determine whether wine had been sweetened with lead or not. This testing of products was not only relevant for consumers and apothecaries, but also for artists, as appears from an eighteenth-century artist’s manuscript I have been studying recently.

Mattheus Verheyden, portrait of Otto Frederick Houttuyn. Was the robe painted with a cochineal-based red?
Mattheus Verheyden, portrait of Otto Frederick Houttuyn. Was the robe painted with a cochineal-based red?

Mattheus Verheyden or Verheijden (1700-1776) was a successful Dutch portrait painter. The son of the painter Franck Pietersz Verheyden was orphaned at an early age, but thanks to his late mother’s fortune he received training as an artist and soon he became a sought-after painter in the upper classes, and made many portraits of regents, clergymen and royals. In the Rijksmuseum Research Library, a manuscript ‘art and recipe book’ attributed to him is kept (call number 319H17). Written in a neat hand in an unruled notebook meant for music notation, Verheyden recorded almost two hundred pages of recipes and instructions for artists. On the title page he wrote: “Art and Recipe book, for painters and etchers &c., brought together over time and with diligence, by M.V.H.” Some of the instructions come from well-known artists handbooks of the time, while the origin of others is not entirely clear.

From the order of the contents of the book, it becomes clear that the quality of his materials was very important for Verheyden. The first ten pages are devoted to various recipes for making ultramarine paint based on lapis lazuli. Yet they are followed by a recipe that does not describe how to make a paint, but how to test the quality of pigments. It reads:

Verheyden's pigment test. Source: Rijksmuseum Research Library, call number 319H17.
Verheyden’s pigment test. Source: Rijksmuseum Research Library, call number 319H17.

Test, to discover whether the lacquer one uses to paint is good &c. If the lacquer is made of cochineal, it never changes. And if it is made of Brazil wood, &c., it will becomes an orange colour over time. To discover this, take a small grain of lacquer, and rub it with your nail on a piece of white paper, and put a drop of lemon juice on it. Rub it through the lacquer with your finger. If the lacquer remains unchanged, it is good, and made of cochineal. But if it turns an orange colour, it is unstable, and made of Fernambuk, or Brazil wood, and will evaporate over time, and is no good.

Cochineal in crude and powdered form
Cochineal in crude and powdered form

Powdered brazil wood extract. It is not difficult to imagine the confusion.
Powdered brazil wood extract. It is not difficult to imagine the confusion.

Red pigment made of brazil wood was – and is – much cheaper than that made of cochineal. Even today, the latter is four to thirteen times more expensive than the former – crude cochineal sells at about ten euros for 25 grams, and carmine red (purified cochineal) costs about thirty euros per 25 grams, while the same amount of brazil wood pigment costs only about 2.25 euros. The relative instability of brazil wood as a red pigment had been known for centuries, but this recipe gave a quick and relatively easy way to determine whether what was on offer was cochineal or brazil wood.

For an artists like Verheyden, whose reputation depended partly on the quality of the source materials he used for his paints, this was understandably very important. Commercial, ready-to-use paints only became available in the nineteenth century. Although Verheyden’s recipe book never appeared in print as far as we know, it is reasonable to assume that the kind of testing and quality-control described in this recipe was a desirable commodity for painters and other artists who worked with pigments that they bought from apothecaries, herbalists, chemists, or traveling salesmen.