FR

Smalt in CENTRAL Italian Painting in Early Cinquecento:
Knowledge, Practice, and Problems of Detection from Leonardo da Vinci to Raphael

Laure Chevalier, phd

Abstract
Between 1500 and 1520, smalt occupied a limited yet significant place in painting in central Italy. In the case of Leonardo da Vinci, documentary sources attest to a detailed technical understanding of this material, even though its pictorial use has not yet been demonstrated by material analysis of surviving works. In Raphael’s practice, smalt appears only exceptionally and in stratified contexts, within a palette dominated by more stable mineral pigments. Recent hyperspectral investigations nonetheless invite a reassessment of certain analytical silences, suggesting that smalt may have functioned as an optical modulator even when partially discoloured or concealed beneath later layers. Far from closing the debate, this study argues for combined analytical approaches and for a critical reading of the sources, in which absence of evidence should not be confused with evidence of absence.

​
**
Smalt—potash glass coloured with cobalt and ground for pictorial use—occupies a distinctive position in the history of Renaissance painting techniques. Although well documented in the fire arts, glassmaking, and goldsmiths’ practice from the late fifteenth century onwards, it remains difficult to grasp within the field of Italian easel painting around 1500. This difficulty stems less from any lack of familiarity with the material on the part of artists than from a persistent discrepancy between written sources, which attest to its preparation and handling, and analytical data, which are rarely conclusive in surviving works.
This tension is particularly acute in the case of Leonardo da Vinci. The term smalto is explicitly mentioned in his notebooks, and his descriptions reveal a precise command of processes of grinding, purification, and chromatic enhancement. Yet, despite extensive analytical campaigns undertaken on his paintings, none has so far identified the presence of cobalt-bearing smalt unequivocally within autograph paint layers. Conversely, in Raphael’s case, smalt is neither attested by the textual record nor regularly revealed by analysis, even though a single instance--La Madonna dell’Impannata—has recently reopened the discussion.
Drawing on a corpus that brings together testimonial sources, scientific publications, conservation reports, and recent hyperspectral results, this study offers a synthetic and critical reading of smalt usage between 1490 and 1520. Priority is given to Leonardo da Vinci before extending the discussion to Raphael and his circle, while taking into account the methodological constraints that currently condition the identification of this pigment.
From a material standpoint, smalt differs fundamentally from the crystalline mineral pigments that dominate Italian Renaissance palettes. It is a vitreous pigment, whose identification rests on the association of a silicate matrix, an alkaline flux—principally potassium—and cobalt as the colouring agent, sometimes accompanied by trace elements derived from the ores employed. This glassy nature entails a particular vulnerability to chemical alteration: progressive leaching of potassium leads to a loss of blue colour, while cobalt may persist without any visible chromatic expression. Under these conditions, smalt used in thin layers, in mixtures, or covered by more saturated pigments may readily elude surface analyses, making any non-detection intrinsically equivocal.
Leonardo’s writings provide exceptional evidence in this respect. In the notebooks as transmitted by Jean Paul Richter’s 1888 edition, Leonardo describes in detail how smalt should be ground on tempered steel, how porphyry should be avoided to prevent mineral contamination, how the pigment may be purified using aqua fortis, and how its colour may be refined through treatment with tartar. These passages testify to a fully mastered operational knowledge and to a sustained reflection on the constraints and potentialities of the material.
To this technical mastery is added an economic dimension. In a series of building-site notes and accounts, datable to around 1513-1514 and explicitly framed as estimates of costs and working time, smalt is mentioned alongside azzurro, gold, white, indigo, and glue. It appears not in a theoretical context, but within a strictly accounting rubric, linked to the invention of the composition and the execution of painted architectural elements. Smalt is thus integrated into the workshop’s material economy, not as a marginal curiosity, but as a recognised component of pictorial expenditure. While these documents do not constitute direct proof of smalt’s effective use in the surviving works, they make it difficult to sustain the idea that Leonardo rejected the pigment or was ignorant of it on principle.
The contrast nonetheless remains striking between this documentary richness and the silence of material analysis. The most comprehensive studies—carried out notably at the C2RMF and the National Gallery, and synthesised in the 2024 Technè special issue—converge in identifying natural ultramarine and azurite as Leonardo’s blue pigments, without any associated cobalt-coloured glass matrix. Several explanations must therefore be considered jointly: occasional or experimental use, application in underlayers, later overpainting, chemical alteration of the smalt, or the inherent limitations of non-invasive methods. The result is an intermediate position: Leonardo knew smalt, prepared it, and incorporated it into the economy of his workshop; if he employed it pictorially, such use may have been localised, discreet, and thus less a legible stylistic choice than a question of material visibility.
In Raphael’s case, the logic is different. Cangiante effects are generally achieved through pigment substitution between ultramarine and azurite, through underlying red lakes, and through modulation using earth pigments and organic blacks. This strategy privileges chromatic stability and control of modelling, without systematic recourse to vitreous pigments. The study of La Madonna dell’Impannata (1511–1515) nonetheless constitutes a notable exception: analyses have revealed an azzurro di smalto characterised by the cobalt–potassium–silicon association, used alongside azurite and ultramarine within a complex stratigraphy, possibly linked to an earlier compositional state. Although isolated, this case shows that smalt could be mobilised on a punctual basis in precisely controlled contexts.
Recent hyperspectral investigations conducted on The Madonna dei garofani - Chatron provide additional insight. Certain areas of blue drapery display signatures compatible with smalt–ultramarine associations, combined with red pigments such as vermilion and lake. Analysis of spectral derivatives reveals vibrational and chromatic modulation effects consistent with moiré-like or cangiante phenomena, observable to the naked eye. A crucial methodological point lies in the explicit recognition of discoloured smalt: loss of chromatic intensity may alter or suppress certain spectral signatures without implying the material disappearance of the pigment. These observations reinforce the hypothesis of a discreet optical role for smalt, even where its chromatic contribution is no longer immediately perceptible.
Taken together, these data call for heightened caution. The risks of misinterpretation are numerous: an unwarranted assimilation of smalt to finely ground colourless glass used as an additive; the presence of modern cobalt-containing retouchings lacking a potash glass matrix; complex stratigraphies masking earlier layers; and chemical alteration of the pigment. It is now accepted that the mere detection of cobalt cannot suffice to identify smalt without rigorous microstructural and stratigraphic characterisation.

Between 1500 and 1520, smalt thus occupied a modest yet historically significant place in Italian painting. Leonardo’s writings attest to a detailed technical understanding of the material, even though its pictorial use has not yet been demonstrated in surviving works. In Raphael’s practice, smalt appears only exceptionally and in circumscribed contexts, within a palette dominated by more stable mineral pigments. Recent hyperspectral imaging nonetheless encourages a reassessment of certain analytical silences and invites consideration of smalt as a potential optical modulator, even when partially altered or concealed. Rather than closing the debate, these results underscore the need for integrated analytical approaches and for a historically informed reading of the sources, in which absence of evidence should not be confused with evidence of absence.

Bibliography (selection)
BERRIE, Barbara H., “Mining for Color: New Blues, Yellows, and Translucent Paint”, Early Science and Medicine, vol. 20, nos 4–6 (2015), p. 308–334.
BORGIA, Ilaria; SECCARONI, Claudio, “L’azzurro di smalto nella pittura e nelle fonti italiane del XV e XVI secolo”, OPD Restauro, no 17 (2005), p. 152-164.
CIATTI, Marco; CASTELLI, Ciro; FROSININI, Cecilia; GUSMEROLI, Luisa; CARTECHINI, Laura; ROSI, Francesca; SECCARONI, Claudio; MILIANI, Costanza; MOIOLI, Pietro, “Il restauro della Madonna dell’Impannata di Raffaello, delle Gallerie degli Uffizi, Galleria Palatina di Palazzo Pitti”, OPD Restauro, no 29 (2017), p. 48–90.
SPRING, Marika, “Colourless Powdered Glass as an Additive in Fifteenth- and Sixteenth-Century European Paintings”, National Gallery Technical Bulletin, vol. 33 (2012), p. 4–26.
MOTTIN, Bruno; RAVAUD, Élisabeth (éds.), « Léonard de Vinci, l’expérience de l’art », Technè, Hors-série (2024).

Testimonial sources
Leonardo da Vinci, Codex Forster II.2, fol. 4r (ex–S.K.M. II.2), V&A, Londres ; ed. J. P. Richter, 1888, p . 591.
[The Complete Notebooks of Leonardo da Vinci, vol. I, trad. J. P. Richter, Londres, 1888, p. 270; p. 591.]

The Madonna Chatron (La Madonna dei garofani) - X-ray fluorescence (XRF) - The Dress - ©AGALMATA / Ph. Walter (2020)