The fate of neutral verdigris in wall paintings: chromatic, micro-morphological, mineralogical and chemical transformations

Abstract

This study fills the gap regarding the behaviour of neutral verdigris pigment (Cu(CH₃COO)₂·H₂O) in traditional wall-painting techniques, an area far less studied than its degradation in manuscripts or easel paintings. Verdigris was applied using three historical methods—a secco (egg yolk or rabbit glue), mezzo-fresco (lime water), and fresco (lime paste)—and complementary pigment–binder mixtures were prepared to isolate binder effects. Multimodal analyses (stereomicroscopy, spectrophotometry, OM, SEM, XRD, ATR-FTIR) revealed a striking contrast between acidic and alkaline environments: verdigris remained chemically and microstructurally stable in a secco but underwent rapid and extensive alteration in lime-based techniques. These transformations produced marked colour shifts, micro-textural heterogeneity, and micrometric secondary compounds such as copper and calcium acetates (e.g. paceite), and tenorite. The findings provide new insights into historical wall-painting practices as they re-examine the use of verdigris with certain painting techniques, also with significant implications for conservation strategies involving verdigris-containing artworks.

Introduction

Verdigris, a term umbrella for copper(II) acetate-based compounds, is one of the earliest known synthetic pigments used in art, with references dating back to the 3rd century BCE^1,2. Its manufacturing process was widely described since antiquity, most often involving the exposure of copper or copper alloys to vinegar vapours, subsequently scraping off the resulting corrosion that formed on the surface^1,3,4; other organic substances—wine lees, milk, urine, or honey—were also reported to yield mixed compounds^1,3,4. These methods yielded various verdigris phases which differ in colour, crystal structure and stability, and can be distinguished by their chemical composition, expressed as the molar ratio xCu(CH₃COO)₂·yCu(OH)₂·zH₂O (the x–y–z phase)⁵. Broadly, verdigris phases can be grouped into two categories: (i) neutral (sometimes distilled or recrystallised) and (ii) basic verdigris⁵. The first group comprises bluish-green copper(II) acetates—both hydrous (1–0–1 phase, Cu(CH₃COO)₂·H₂O) and anhydrous (1–0–0 phase, Cu(CH₃COO)₂)—valued for their relative stability^{5, 6, 7, 8–9}. The second group encompasses copper(II) acetate hydroxide hydrates—up to seven distinct phases—which are more susceptible to alteration^{5, 6, 7, 8, 9–10}.

The use of this pigment was widespread across various substrates during the Middle Ages—including stone, wood, canvas, panel, paper, and parchment—particularly in 15th- to 17th-century easel painting¹. Its popularity, however, declined in the early 20th century with the advent of alternative green pigments, such as emerald green (Cu(C₂H₃O₂)₂·3Cu(AsO₂)₂) or viridian (Cr₂O₃·H₂O)^1,11,12. Despite its extensive historical use in painting, artists and treatises had long acknowledged its instability^{13, 14, 15–16}. Cennino Cennini (14th c.)¹⁷, for instance, recommended its application on paper with egg yolk but cautioned that the colour would not last. Indeed, copper acetates undergo gradual colour changes depending on the binding medium: with resins, they form copper resinates; with oils, copper oleates; and with proteins, copper-protein complexes⁴. Other historical sources explicitly discouraged its use in painting due to its high reactivity with atmospheric pollutants such as sulphur dioxide¹⁸ or with other pigments, particularly those sulphur-based^19,20. As a result, substantial research has been devoted both to reproducing historical recipes and various forms of verdigris^6,9,10,21,22, and to its degradation when combined with organic binders such as gum arabic^{15,16,23, 24, 25–26}, protein-based binders (i.e. egg yolk or rabbit glue)^{25,27, 28, 29–30} and linseed oil^{26,28, 29, 30, 31, 32–33}.

In contrast to its well-documented behaviour in paper, parchment and easel painting, the stability of verdigris—whether neutral or basic—in wall paintings remains poorly understood. Verdigris is known to be highly sensitive to alkaline environments, rendering it unsuitable for application on walls due to the inherent alkalinity of mortars²⁰. Nonetheless, verdigris has been identified in Romanesque wall paintings in France³⁴, in several examples of Gothic wall paintings in England³⁵, and in various wall paintings in Italy and Spain that appear to have been executed as a fresco^{36, 37–38}. While alternative techniques such as mezzo-fresco may also explain these occurrences, no securely documented cases are known. Moreover, the term ‘fresco’ has been used loosely in the literature, potentially contributing to misattributions of technique and obscuring the original conditions under which verdigris was originally applied³⁹. Notably, historical authors such as Cennino Cennini (14th c.), Giorgio Vasari (16th c.) or Andrea Pozzo (18th c.) all explicitly warned against using verdigris in fresco painting⁴⁰.

This apparent contradiction—clear historical prohibitions vs repeated archaeological identifications—highlights a significant gap in current knowledge. The alteration process of verdigris in wall-painting environments remains insufficiently studied, and its degradation pathways are still a matter of debate. A recent investigation⁴¹ evaluated the stability of several copper-based pigments in fresco conditions and found copper acetate to be the most unstable, showing heterogeneous discolouration and morphological changes distinct from the homogeneous blackening observed in copper carbonates and silicates. These findings indicate that verdigris undergoes a more complex, poorly understood alteration trajectory in alkaline media—one that is not satisfactorily explained in existing research.

To address this gap, the present study explicitly aims to evaluate the alteration process of neutral verdigris in alkaline wall-painting environments—from the earliest stages of its degradation to its typical transformation into black copper oxides such as tenorite. Specifically, we investigate how both the binding medium and the painting technique influence its degradation. A bluish-green neutral copper(II) acetate monohydrate with the formula Cu(CH₃COO)₂·H₂O was used as pigment. To isolate the pigment-binder interactions, verdigris was mixed in petri dishes with egg yolk, rabbit glue, lime water and lime paste separately. To evaluate the role of the painting technique, the pigment was applied a secco, where the pigment is mixed with an organic binder (with egg yolk or with rabbit glue) and applied onto a dry mortar surface and fixed by polymerisation of the binder¹⁹. The pigment was also applied using a mezzo-fresco technique, which involves mixing the pigment with lime water and applying it onto a semi-dry mortar, fixed by reactivation of the carbonation process¹⁹. Lastly, the pigment was also applied a fresco: the pigment is mixed with water, applied onto a freshly laid lime-based mortar and fixed to the surface by carbonation of the lime¹⁹.

A physical evaluation of the pigment-binder mixtures (petri dishes) and the paintings (on a lime-based mortar substrate) was conducted by digital photography, stereomicroscopy, and colour spectrophotometry, followed by detailed micro-morphological, mineralogical and chemical analyses using optical and scanning electron microscopy with energy dispersive X-ray spectroscopy (SEM-EDS), X-ray diffraction (XRD) and Attenuated Total Reflectance Fourier-transform Infrared Spectroscopy (ATR-FTIR). By combining controlled pigment-binder experiments with systematically prepared wall painting samples, this study provides the first integrated evaluation of how neutral verdigris degrades across traditional wall painting techniques. The finding shed new light on the complex morphological, mineralogical and chemical transformations induced by alkaline environments—transformations that have remained largely uncharacterised until now and that have long hindered the accurate interpretation, conservation and attribution of wall paintings containing verdigris.

Methods

Materials

A calcitic lime paste, stored under water for more than 20 years, was used as a binder in the mortar manufacture. It was mainly composed of portlandite (Ca(OH)₂) and small amounts of calcite (CaCO₃) due to partial carbonation. The lime paste was mixed with three aggregates: coarse (1–2 mm) and fine silica (0.4–0.8 mm), and marble powder (<0.7 mm). The silica aggregates, composed of quartz (SiO₂), sodium/potassium feldspars and rutile (TiO₂), were supplied by a local hardware store (Granada, Spain). The marble powder, composed of calcite, quartz, and dolomite (CaMg(CO₃)₂), was supplied by CTS S.L. (Madrid, Spain). The pigment selected for this study was commercialised as a basic copper(II) acetate monohydrate (Ref. 44450), with formula C₄H₆CuO₄·H₂O and Cu(CH₃COO)₂·[Cu(OH)₂]₃·2H₂O, supplied by Kremer Pigments GmbH & Co. KG (Aichstetten, Germany). The mineralogical characterisation of the raw materials (lime paste, aggregates and pigment) was obtained from ref. ⁴². As for the binders used for each painting technique, the eggs were purchased at a local supermarket (Pontevedra, Spain), and the rabbit glue (Ref. 63028) and the lime water (Ref. 31808) were supplied by Kremer.

Paint mock-up preparation

To isolate and evaluate the influence of pigment-binder interactions independently from the mortar substrate, an admixture of the pigment mixed in a 1:1 (pigment:binder) ratio in volume with egg yolk (VER-EY hereinafter), rabbit glue (VER-RG hereinafter), lime water (VER-MF hereinafter) and lime paste (VER-F hereinafter) were prepared in petri dishes.

The manufacture of the paint mock-ups, composed of two mortar layers (arriccio and intonaco), was carried out following the methodology described in ref. ⁴¹. The mock-ups took the form of thick, circular discs (≈8 cm diameter and ≈2.5 cm tall), in which the base structure of historical wall paintings was reproduced⁴³. Four sets of paint mock-ups were prepared (paint mock-ups were identified with a -P after the identifier of the painting technique: EY, RG, MF and F):

• a secco with egg-yolk (VER-EY-P hereinafter): the yolk was separated from the white by pouring it back and forth in the half shells. It was then rolled onto a paper towel to remove any remaining egg white. The skin of the yolk was then punctured with a needle, pouring the resulting liquid into a glass jar. The yolk was then mixed with verdigris in a 1:1 (pigment:binder) ratio in volume and applied onto a complete carbonated dry mortar (after 6 months).

• a secco with rabbit glue (VER-RG-P hereinafter): 8 g of rabbit glue were soaked in 100 mL of deionised water for 24 h, under periodic stirring. Afterwards, the mixture was heated below 50 °C in a water bath and stirred to obtain a homogeneous mixture. The liquid was then mixed with verdigris in a 1:1 (pigment:binder) ratio in volume and applied onto the lime substrate, as carried out with VER-EY-P.

• a mezzo-fresco (VER-MF-P hereinafter): the pigment was mixed with lime water and applied onto the semi-dry mortar (intonaco), after 3 days of applying the intonaco.

• a fresco (VER-F-P hereinafter): verdigris was mixed with water until the right consistency (it flows easily from the brush) was achieved and then applied onto a wet and freshly made lime-based mortar (intonaco).

For both secco techniques, painting was carried out after 6 months of mortar carbonation, ensuring minimal portlandite content—and therefore low alkalinity. XRD analysis indicated that the intonaco contained approximately 8% portlandite, a residual amount consistent with the slow carbonation process, which may remain incomplete even after two years under laboratory conditions⁴⁴. Thus, we can state with confidence that our a secco samples were applied to well-carbonated mortars. After paint application, all samples (both pigment-binder mixtures and painted mock-ups) were left to dry for one month under controlled laboratory conditions (22 ± 2 °C and 65 ± 5% relative humidity).

Analytical techniques

The following analytical protocol was followed:

• The pH of the binders (egg yolk, rabbit glue, lime paste and lime water) was measured using a Crison® PH-25 portable potentiometer (I.C.T., S.L., Madrid, Spain), with a precision of ±1 mV (±0.01 pH unit). An automatic calibration of the instrument was carried out with buffer solutions at 2.00, 4.01, 7.00, 9.21 and 10.90 pH.

• Digital photographs of the paint samples and the mixtures in the petri dishes were taken with a Nikon D3400 digital camera (Shinagawa, Tokyo, Japan) equipped with a Nikon Nikkor Zoom Lens, DX 18–105 mm.

• A Nikon SMZ 1000 stereomicroscope (Melville, New York, United States) was used to examine the textural and chromatic features of the raw pigment, the surface of the paint mock-ups, the petri dish samples, and the alteration products.

• The colour of the alteration products taken from the samples was characterised using CIELAB and CIELCH colour spaces^45,46, measuring L* (lightness), a* and b* (colour coordinates) and C*_ab (chroma) values using a Minolta CM-700d spectrophotometer (Tokyo, Japan). L* represents lightness, varying from 0 (black) to 100 (white). The other two parameters are chromaticity coordinates: a* goes from red to green (where +a* is red and −a* is green) and b* from yellow to blue (+b* is yellow and −b* is blue). C*_ab is calculated according to the following formula: C*_ab = (a² + b²)^1/2. The measurements were made in specular component excluded (SCE) mode, for a spot diameter of 8 mm, using illuminant D65 at observer angle 10°. A total of 15 measurements were made for each mock-up, and their average values and standard deviations (STD) were computed. Colour parameters were also measured for the pigment powder, after first flattening it against a glass support to obtain a smooth surface.

• The paint thin cross sections were studied by polarised light microscopy (PLM) with reflected light (RL) in plane-polarised light (analyser removed) or crossed polarisers (analyser inserted) using a Carl Zeiss Jenapol U microscope (Oberkochen, Germany). For this, the specimens were embedded in epoxy resin (EpoThin 2 Epoxy Resin and EpoThin 2 Epoxy Hardener) and then cut and polished to a mirror shine.

• The microtexture and elemental composition of the raw pigment, the pigment-binder admixtures, and the paint mock-ups (surface and cross-sections) were evaluated using a FEI Quanta 400 ESEM FEG (Hillsboro, Oregon, USA) with energy-dispersive X-ray spectroscopy (EDS) in both secondary (SE) and backscattered electron (BSE) detection modes. Observation conditions included a working distance of ~10 mm, accelerating potential of 20 kV and specimen current of ~60 mA.

• The mineralogical composition of the raw pigment and the alteration products obtained from the paint mock-ups was determined using XRD with a XPert PRO (PANalytical B.V., Almelo, The Netherlands), according to the random-powder method. Analyses were performed using Cu-Kα radiation, Ni filter, 45 kV voltage, and 40 mA intensity. The exploration range was 3° to 60° 2θ, and the goniometer speed was 0.05° 2θ s⁻¹. The mineral phases were identified using X’Pert HighScore software.

• The molecular composition of the pigment and the alteration products was obtained by Attenuated Total Reflectance Fourier–Transform Infrared Spectroscopy (ATR-FTIR), using a Thermo Nicolet 6700 (Thermo Fisher, Massachusetts, United States) at a 2 cm⁻¹ resolution over 100 scans in the mid (4000–400 cm⁻¹) and far (600–200 cm⁻¹) infrared spectral regions. Samples were individually analysed by placing them on a diamond crystal and pressing the plunger against them, exerting the maximum pressure permitted by this instrument so as to ensure the best possible contact with the diamond crystal.

Results

Characterisation of the verdigris pigment and the binders

The bluish-green acetate pigment (VER) corresponds chemically and structurally with the mineral phase hoganite, as confirmed by XRD (Supplementary Fig. 1). Its composition, Cu(CH₃COO)₂·H₂O, identifies it as a neutral hydrous verdigris (the 1–0–1 phase), considered a pure phase^5,9. A small additional peak at 2θ = 11.53° (d spacing = 7.68 Å) was also detected. Following Bette et al.⁹, this peak can be associated with the anhydrous 1–0–0 phase, Cu(CH₃COO)₂ (marked with a red arrow in S1.1). Interestingly, although the supplier labels the pigment as a basic verdigris and reports a composition that includes both the neutral phase (C₄H₆CuO₄·H₂O) and a basic copper acetate hydroxide (Cu(CH₃COO)₂·[Cu(OH)₂]₃·2H₂O), our analysis shows that the pigment consists predominantly of the neutral hydrous phase. SEM observations (Supplementary Fig. 2) reveal that the pigment particles are mainly euhedral crystals with prismatic habit, accompanied by a smaller proportion of subhedral crystals. The chemical characterisation by EDS analysis revealed that all particles were composed of variable amounts of carbon (C), oxygen (O) and copper (Cu). Regarding the pH measurements of the binders, egg yolk and rabbit glue were slightly acidic (6.54 and 5.82 pH units, respectively) compared to the high alkalinity detected in lime water (pH = 11.56) and lime paste (pH = 11.92).

Physical evaluation

Figure 1 shows digital photographs and stereomicrographs of the powder pigment, the pigment-binder admixtures (left column) and of the painting mock-ups (right column). It was observed that a secco samples, both pigment-binder admixtures (VER-EY and VER-RG) and paint mock-ups (VER-EY-P and VER-RG-P), were homogeneous in terms of colour, though different between them, as observed in Fig. 1b, c, respectively. Egg yolk samples manifested a greenish hue due to the colouration effect of the yolk (orange in colour), whilst rabbit glue ones retained a similar bluish hue to that of the raw pigment. As for mezzo-fresco (VER-MF and VER-MF-P) and fresco (VER-F and VER-F-P) samples, a highly heterogeneous surface was observed due to the presence of different compounds exhibiting various colourations, as shown in Fig. 1d, e, respectively. As opposed to EY and RG samples, MF and F ones showed evident differences between the pigment-binder admixtures and the paint mock-ups. Different coloured compounds were depicted on the surface, ranging from light to dark blues. Greyish-black compounds were also observed, especially in the fresco paint (VER-F-P).

[See PDF for image]

Fig. 1

Preliminary assesment by stereomicroscopy.

a Stereomicrographs of neutral verdigris raw pigment (VER). b–e Digital photographs (in the circle) and stereomicrographs (in the rectangle) of the pigment-binder admixtures prepared in petri dishes (left column) and the paint mock-ups (right column) of (b) a secco with egg yolk (-EY), ca secco with rabbit glue (-RG), da mezzo-fresco (-MF) and ea fresco (-F).

The colour parameters for the pigment-binder admixtures and for the paint mock-ups were obtained (included as Supplementary Table 1). By comparing L* and C*_ab parameters (Fig. 2a), it was observed that mezzo-fresco (VER-MF and VER-MF-P) and fresco (VER-F and VER-F-P) samples showed higher L* compared to the secco ones. This was related to the whitening effect of the carbonation process of lime water and lime paste. Moreover, the high STDs observed in MF and F, as opposed to EY and RG, were due to the wide range of different-coloured forms observed (Fig. 1d, e). Figure 2b represents a* (red-green) and b* (yellow-blue) coordinates. It was observed that EY-based samples showed higher b* (yellow) and a* (red) components, thus orange in appearance, as mentioned before, due to the influence of the binder. The bluest samples were RG-based ones (lower b*), resembling a similar bluish hue to that of the raw pigment (represented with a black star in Fig. 2b). As for MF and F samples, they generally showed a greenish-blue hue. However, we must consider the heterogeneity in colour observed by stereomicroscopy in MF and F samples; thus, while in a secco paint, a* and b* parameters showed STDs of c.a. 1 unit, in the case of mezzo-fresco and fresco samples, it was generally c.a. 5 units. Therefore, the latter showed higher colour variations (especially in VER-F-P).

[See PDF for image]

Fig. 2

Chromatic parameters by spectrophotometry.

a Scatter plot representing L* (lightness) and C*_ab (chroma) parameters and b polar plot representing a* (green to red) and b* (blue to yellow) parameters of neutral verdigris powder pigment (VER), the pigment-binder admixtures prepared in petri dishes and the paint mock-ups (-P) of a secco with egg yolk (-EY), a secco with rabbit glue (-RG), a mezzo-fresco (-MF) and a fresco (-F) samples.

A total of 13 different patterns were collected from the pigment-lime water (VER-MF) and pigment-lime paste (VER-F) admixtures, and from the mezzo-fresco (VER-MF-P) and fresco (VER-F-P) paint mock-ups. The selection criteria was based on colour —bluish-green (VER-MF_1), light blue (VER-MF_2, VER-MF-P_5, VER-MF-P_6, VER-F_9, VER-F_10, VER-F-P_11 and VER-F-P_12), dark blue (VER-MF_3, VER-F-P_4, VER-F_7 and VER-F-_8) and greyish-black (VER-F-P_13)—and morphology—globular (VER-MF_1, VER—MF-P_4-6, VER-F_7-8, VER-F_10 and VER-F-P_11), acicular (VER-MF_2-3 and VER-F-P_12) and laminar (VER-F_9. In short, these 13 samples were selected as potential alteration forms to be further characterised. Their appearance under stereomicroscopy is included in Fig. 3.

[See PDF for image]

Fig. 3

Identification of the alteration forms by stereomicroscopy.

Identification of the alteration forms by stereomicroscopy. Micrographs of the 13 alteration forms obtained from the pigment-binder admixtures with lime water (VER-MF) and with lime paste (VER-F), and from the paint mock-ups, amezzo-fresco (VER-MF-P) and bfresco (VER-F-P).

Micro-morphological characterization

SEM-EDS observations enabled the identification of different morphologies and significant micro-textural and micro-morphological changes in mezzo-fresco and fresco samples as opposed to secco. The study of the cross-sections by PLM showed a homogeneous paint layer in both secco paints, greener in VER-EY-P and bluer in VER-RG-P (Fig. 4a, c, respectively), as opposed to the heterogeneous paint layer observed in mezzo-fresco and fresco paints (Fig. 4eg, respectively). Besides different bluish shades, brownish-black areas were observed on both MF and F paints. The evaluation by SEM of the surface of a secco paint allowed to identification of the characteristic euhedral prismatic habit of hoganite crystals in both egg yolk (VER-EY-P, Fig. 4b) and rabbit-glue (VER-RG-P, Fig. 4d). No micro-textural nor morphological changes related to a possible degradation of the pigment were observed. By SEM, both the paint cross-section and paint surface in mezzo-fresco (VER-MF-P, Fig. 4e, f) and fresco paints (VER-MF-P, Fig. 4g, h) were highly irregular and heterogeneous. Both paint layers were composed of varying amounts of C, O, Ca and Cu, as observed by EDS mapping analysis.

[See PDF for image]

Fig. 4

Cross-section and surface evaluation of the paint mock-ups by optical and electron microscopies.

PLM micrographs of the paint mock-ups: a a secco with egg yolk (VER-EY-P), c a secco with rabbit glue (VER-RG-P), e a mezzo-fresco (VER-MF-P), and g a fresco (VER-F-P). SEM micrographs of the surface of b VER-EY-P, d VER-RG-P, e, f VER-MF-P, and g, h VER-F-P.

The microscopic study of the surface of the pigment-binder admixtures and of the paint mock-ups by SEM-EDS allowed the identification of a great number of neoformed particles with different habits. Figure 5 includes the different micro-morphologies identified in the pigment-binder admixtures (VER-MF and VER-F), and Fig. 6 includes those identified in the paint mock-ups (VER-MF-P and VER-F-P). The micrographs obtained from the samples where the pigment was mixed with lime water (VER-MF) suggested the beginning of an alteration process since the morphology of hoganite was visible albeit with noticeable micro-textural changes in the form of micro-pitting patterns (marked with a white arrow in Fig. 5a). Many subhedral crystals with prismatic habit, not identified previously in the raw pigment characterization, were identified grouped together, as observed in Fig. 5b. EDS analysis showed they were composed of varying amounts of C, O and Cu suggesting they could be related to neoformed hoganite or to other acetates. Bigger agglomerates of particles with similar prismatic habit merged together were also observed (marked with white arrows in Fig. 5c). Other euhedral crystals with similar chemical composition were depicted, some of them with a hexagonal habit (marked with a white arrow in Fig. 5d). Fibrous morphologies could be observed ranging from the surface of some of the particles in VER-MF (marked with red arrows in Fig. 5c–e) suggesting an ongoing transformation. Finally, other ribbon-shaped morphologies conformed by Cu-rich needle-like morphologies were identified (Fig. 5f). The samples where the pigment was mixed with lime paste (VER-F) were predominantly different to those observed in VER-MF. In VER-F, none of the crystals showed the prismatic habits observed in VER-MF (Fig. 5a–e). In turn, the formation of crystals with tabular habit, composed of varying amounts of C, O, Cu and Ca, and grouped forming rosettes was widespread (Fig. 5g). A laminar corrosion process was clearly visible at higher magnifications in similar particles (Fig. 5h). Lastly, Cu-rich needle-like morphologies, similar to those observed in VER-MF_3 (Fig. 5f) were sporadically seen (Fig. 5i), together with other acicular (Fig. 5j) and laminar morphologies (Fig. 5k).

[See PDF for image]

Fig. 5

SEM micrographs of alteration forms observed in the pigment and binder admixtures with lime water (VER-MF) and lime paste (VER-F).

The samples selected are shown in Fig. 3 from the samples in Fig. 2d, e.

[See PDF for image]

Fig. 6

SEM micrographs of the alteration forms observed on the surface of the paint mock-ups when applying a mezzo-fresco (VER-MF-P) and a fresco (VER-F-P).

The paint mock-ups are shown in Fig. 2d, e.

The study of the surface of mezzo-fresco (VER-MF-P) and fresco (VER-F-P) paint mock-ups showed completely irregular surfaces, with the presence of numerous neoformed micro-morphologies with heterogeneous micro-texture. The ones presented in Fig. 6 were those more widespread. In the case of MF paint, different Cu-rich acicular morphologies were identified with clear differences in length and thickness (Fig. 6a, c). All micro-morphologies showed varying amounts of C, O, Cu and Ca. Although these acicular shapes were the most widespread, other flower-like morphologies were depicted (Fig. 6d), composed of higher amounts of Ca than the previous particles. Lastly, large areas of the mezzo-fresco paint showed fibrous morphologies with a “spider web” appearance (Fig. 6e). As for the fresco paint, the morphologies were more different, generally composed of higher contents of Ca than the morphologies identified in the mezzo-fresco paint. Still, similar Cu-rich needle-like morphologies were sporadically identified (Fig. 6f). Other acicular (Fig. 6g–i) and spherical (Fig. 6j) morphologies and flake-like agglomerations (Fig. 6k) were also present. Other areas of the fresco paint surface were highly cracked and fissured (Fig. 6k, l); no particles with specific morphology were identified.

Mineralogical characterization

The mineralogical characterisation by XRD of the secco paintings (VER-EY-P and VER-RG-P), and of the 13 alteration products selected from mezzo-fresco (MF) and fresco (F) samples, are presented in Table 1. Overall, hoganite (Cu(CH₃COO)₂·H₂O)—the main mineral phase of the raw pigment—remains the dominant phase in the secco samples (VER-EY-P and VER-RG-P), where no alteration products were detected. Hoganite was also present in all MF-based samples (VER-MF_1, VER-MF_2, VER-MF_3, VER-MF-P_4, VER-MF-P_5 and VER-MF-P_6), accompanied by additional acetate phases associated with the alteration process. The XRD patterns of the verdigris–lime water mixtures (VER-MF_1, VER-MF_2, VER-MF_3, see Supplementary Figs. 3–5) display an additional peak at approximately 2θ = 9.29° (marked with red arrows). Previous studies^6,10 have attributed this peak to basic verdigris phases 1–3–2 (Cu₂(CH₃COO)(OH)₃·H₂O) and 2–1–5 (Cu₃(CH₃COO)₄(OH)₂·5H₂O). Notably, the intensity ratio between the hoganite phase and the 2θ = 9.29° peak varies among the three samples, with the basic phases appearing more prominent in VER-MF_2 and especially in VER-MF_3. Paceite (CaCu(CH₃COO)₄·6(H₂O), as well as calcium acetate in its either monohydrate (Ca(CH₃COO)₂·H₂O) or hemihydrate (Ca(CH₃COO)₂·(H₂O)_0.5) form, was also frequently identified in both MF and F samples. In addition, anhydrous copper acetate (Cu(CH₃COO)₂) was detected in several fresco samples. Since this phase was already present in small amounts in the raw pigment, it cannot be considered an alteration product. Copper oxide (tenorite, CuO) was also identified in sample VER-F-P_13, which was also the only sample showing distinctly black alteration. Some XRD patterns also revealed a sharp, well-defined unassigned peak that did not correspond to any phase in our database. This peak—listed in Table 1—appeared exclusively in the verdigris–lime paste admixtures (VER-F). Samples VER-F_7, VER-F_8 and VER-F_9 (see red arrows in Supplementary Fig. 9-11) displayed a strong peak at approximately 2θ = 10.30°. Finally, calcite (CaCO₃) was present in most samples, occasionally together with portlandite (Ca(OH)₂). Portlandite was exclusively present in verdigris–lime paste samples, consistent with the slow nature of the carbonation process. Sporadic detection of quartz (SiO₂), illite (KAl₄(Si, Al)₈O₁₀·4H₂O) or anorthoclase ((Na, K)AlSi₃O) is attributed to the siliceous sand used for the mortar preparation.

Table 1. XRD mineralogical characterisation of a secco with egg yolk (VER-EY-P) and a secco with rabbit glue (VER-RG-P) paint mock-ups, and of the 13 alteration forms obtained from the pigment-binder admixtures with lime water (VER-MF) and with lime paste (VER-F), and from the mezzo-fresco (VER-MF-P) and fresco (VER-F-P) paint mock-ups

Samples	Mineralogical characterisation by XRD
	Assigned to the raw pigment	Assigned to the mortar	Assigned to alteration products	Unassigned phases*
VER-EY-P	Hoganite, Cu(CH3COO)2·H2O	Calcite, CaCO3 Quartz, SiO2	-	-
VER-RG-P	Hoganite, Cu(CH3COO)2·H2O	Calcite, CaCO3 Quartz, SiO2	-	-
VER-MF_1	Hoganite, Cu(CH3COO)2·H2O	Calcite, CaCO3 Quartz, SiO2	Cu2(CH3COO)(OH)3·H2O/Cu3(CH3COO)4(OH)2·5H2O	-
VER-MF_2	Hoganite, Cu(CH3COO)2·H2O	Quartz, SiO2	Cu2(CH3COO)(OH)3·H2O/Cu3(CH3COO)4(OH)2·5H2O	-
VER-MF_3	Hoganite, Cu(CH3COO)2·H2O	Calcite, CaCO3 Quartz, SiO2	Cu2(CH3COO)(OH)3·H2O/Cu3(CH3COO)4(OH)2·5H2O Paceite, CaCu(CH3COO)4·6(H2O)	-
VER-MF-P_4	Hoganite, Cu(CH3COO)2·H2O	-	Paceite, CaCu(CH3COO)4·6(H2O) Ca(CH₃COO)₂·(H₂O)0.5/Ca(CH₃COO)₂·H₂O	-
VER-MF-P_5	Hoganite, Cu(CH3COO)2·H2O	Calcite, CaCO3 Quartz, SiO2	-	-
VER-MF-P_6	Hoganite, Cu(CH3COO)2·H2O	Calcite, CaCO3 Quartz, SiO2 Illite, KAl4(Si,Al)8O10·4H2O	Ca(CH₃COO)₂·(H₂O)0.5/Ca(CH₃COO)₂·H₂O	-
VER-F_7	Cu(CH3COO)2	Portlandite, Ca(OH)2 Quartz, SiO2	Ca(CH₃COO)₂·(H₂O)0.5/Ca(CH₃COO)₂·H₂O	2θ = 10.30°
VER-F_8	Cu(CH3COO)2	Portlandite, Ca(OH)2 Quartz, SiO2 Anorthoclase, (Na,K)AlSi3O	Paceite, CaCu(CH3COO)4·6(H2O) Ca(CH₃COO)₂·(H₂O)0.5/Ca(CH₃COO)₂·H₂O	2θ = 10.30°
VER-F_9	Cu(CH3COO)2	Portlandite, Ca(OH)2 Quartz, SiO2	Ca(CH₃COO)₂·(H₂O)0.5/Ca(CH₃COO)₂·H₂O	2θ = 10.30°
VER-F_10	Cu(CH3COO)2	-	Ca(CH₃COO)₂·(H₂O)0.5/Ca(CH₃COO)₂·H₂O	-
VER-F-P_11	Cu(CH3COO)2	Calcite, CaCO3	-	-
VER-F-P_12	-	Calcite, CaCO3 Quartz, SiO2	Paceite, CaCu(CH3COO)4·6(H2O) Ca(CH₃COO)₂·(H₂O)0.5/Ca(CH₃COO)₂·H₂O	-
VER-F-P_13	-	Calcite, CaCO3 Quartz, SiO2	Tenorite, CuO	-

The alteration forms selected are shown in Fig. 3.

*The XRD patterns (included in Supplementary Figs. 3–15) showed additional unassigned mineral phases. The 2-theta (2θ) position for the main intensity peaks unassigned are included.

Chemical characterization

For comparative purposes, the FTIR spectrum of neutral verdigris pigment (sample VER) is shown in Figs. 7 and 8 alongside those of the paints and alteration products. The characteristic IR features of the pigment are highlighted with blue rectangles. The spectrum of the raw pigment—composed mainly of the hydrous 1–0–1 phase (Cu(CH₃COO)₂·H₂O), with a minor contribution from the anhydrous 1–0–0 phase (Cu(CH₃COO)₂)—does not perfectly match either of the spectra reported by Bette et al.⁹. Three large bands at 3450, 3365 and 3270 cm^–1, assigned to –OH stretching vibrations related to the water molecules associated to the acetate ion, confirm the presence of the hydrous phase⁹. Additional bands characteristic of this phase include a signal at 1355 cm^–1 (CH₃ rocking vibration) and bands at 1050 and 1035 cm^–1 (C–H tensions)^7,9,22,47. The copper carboxylate (COO^–) functional groups are represented by bands at 1600 and 1420 cm^–1 (asymmetric and symmetric stretching, respectively), as well as a rocking vibration around 630 cm^{–17,9,22,47,48}. The carbonyl group (C=O) was present as a low intensity band at 1260 cm^–1 (symmetric stretching)^7,22,47,48. O–C–O deformation (690 cm^–1) and in-plane O–C–O rocking vibrations (590–420 cm^–1) were also present^7,22,47,48. However, the raw pigment spectrum also displayed additional low-intensity methyl group (CH₃) bands at 2990 and 2945 cm^–1, typical of acetate groups^7,22,47,48 but not characteristic of either the hydrous or anhydrous phases described in ref. ⁹. Their presence suggests that minor amounts of other acetate compounds may also be present. Finally, the spectra obtained in the far-IR region (600–200 cm^–1, right spectrum) showed a broad band assigned to O–C–O rocking vibrations around 590–420 cm^{–17,22,47,48}. In addition, high intensity bands at 370 cm^–1 (ν(Cu–C)), 330 cm^–1 (ν(Cu–OH)), 270 cm^–1 (δ(Cu–CO)), 250 and 230 cm^–1 (δ(O–Cu–OH)) were identified^7,49.

[See PDF for image]

Fig. 7

FTIR spectra of neutral verdigris raw pigment (VER) and the alteration products identified in the pigment-binder admixtures with lime water (VER-MF) and from the a mezzo-fresco (VER-MF-P) paint mock-ups.

Spectra obtained in the (a) mid-IR (4000–400 cm⁻¹) and (b) far-IR (600–200 cm⁻¹) regions. The samples selected are shown in Fig. 3. Marked in red are the characteristic bands assigned to acetate, which underwent great changes compared to the ones in sample VER.

[See PDF for image]

Fig. 8

FTIR spectra of neutral verdigris raw pigment (VER) and the alteration products identified in the pigment-binderadmixtures with lime paste (VER-F) and from the fresco (VER-F-P) paint mock-ups.

Spectra obtained in the (a) mid-IR (4000–400 cm−1) and (b) far-IR (600–200 cm−1) regions. The samples selected are shown in Fig. 3. Marked in red are the characteristic bands assigned to acetate, which underwent great changes compared to the ones in sample VER.

On the one hand, in secco paint samples, no significant changes were observed (see Supplementary Figs. 16 and 17). The previously mentioned characteristic bands of verdigris were observed on both egg yolk (VER-EY-P) and rabbit glue (VER-RG-P) paint samples. Slight changes were detected due to the overlap between the CH₃ of verdigris (2990 and 2945 cm^–1) and the CH₂ stretching from long-chain fatty acids in egg yolk (at 2920 and 2850 cm^–1) and rabbit glue (at 2918 cm^–1) paint samples^{50, 51, 52–53}. A shift to lower wavenumbers was observed in the carboxylate group of verdigris (COO^– stretching at 1600 cm^–1 and 1420 cm^–1) due to the overlap with bands assigned to egg yolk—at 1625 cm^–1 (C=O stretching from amide I) and at 1520 cm^–1 (C=O bending from amide II)^50,51—and to rabbit glue—at 1635 cm^–1 (C=O stretching from amide I) and around 1533 cm^–1 (C–N stretching and N–H bending from amide II)^52,53. As for the spectra in the far-IR region, VER-EY-P and VER-RG-P showed practically the same shape as that observed in the pigment spectrum (VER). Therefore, no chemical changes were observed.

On the other hand, major differences were observed when comparing the pigment spectrum with that of the 13 samples collected from mezzo-fresco (MF) and fresco (F) samples (Figs. 7 and 8, respectively). VER-MF_1 and VER-MF_2 showed a similar spectra to that of the pigment —suggesting that hoganite was still present— (highlighted with blue rectangles in Fig. 7). In turn, the acetate characteristic bands assigned to carboxyl (COO^–), methyl (CH₃) and carbonyl (C=O) groups shifted to higher or lower wavenumbers and/or changed their intensity ratios to a greater or lesser extent in the other 11 samples (marked in red in Figs. 7a and 8a, respectively). In fact, Supplementary Table 2 displays the IR vibrational position of each of the characteristic acetate bands present in the 13 samples. It is clear that the results indicate that –OH and COO^– underwent considerable shifts when compared to those of the raw pigment (sample VER). Also, the characteristic bands assigned to ν(Cu–C), ν(Cu–OH), δ(Cu–CO) and δ(O–Cu–OH) either disappeared—appearing other bands assigned to Cu–O and Cu–OH bonding (highlighted with pink rectangles in Figs. 7b and 8b)—or changed in shape. All these changes were more prominent in F samples (Fig. 8). Moreover, all samples (except VER-MF_1 and VER-MF_2) presented broad bands assigned to hydroxyl groups (–OH) in the region 3600–2800 cm^–1 (Figs. 7a and 8a). A low intensity band at 3740 cm^–1 was exclusively present in the spectra of fresco-based samples (marked with a black arrow in Fig. 8a). This band is related to –OH groups associated with metals (ν(M–OH))⁵⁴. For that band to be assigned to M–OH, a band around 1500 cm^–1 (H–O–H scissoring mode of H₂O molecules) should also be present⁵⁵. However, this band was not visible since it was probably hidden by the bands assigned to ν(COO^–). Calcium carbonate (CaCO₃) characteristic bands were occasionally present, especially in F samples. A black dotted rectangle in Fig. 7 and Fig. 8 highlights the vibrational modes of CaCO₃: at 1795 cm^–1 (C=O), 1390 cm^–1 ν(CO₃^2–), 875 cm^–1δ_asym(CO₃^2–), 710 cm^–1δ_sym(CO₃^2–) and at 330–250, 216 and 215 cm^–1 (lattice modes of Ca²⁺ and/or CO₃²⁻)^{51,56, 57–58}.

Discussion

Results demonstrate that wall painting techniques, through the properties of their binders and associated pH conditions, strongly influenced the micro-morphological and chemical features of the pigment. A neutral verdigris pigment was employed, primarily composed of the hydrous 1–0–1 phase (Cu(CH₃COO)₂·H₂O), with minor presence of the anhydrous phase 1–0–0 (Cu(CH₃COO)₂). Our results did not align with the supplier’s reported composition, which stated that the pigment was composed of both neutral (C₄H₆CuO₄·H₂O) and basic phases (Cu(CH₃COO)₂·[Cu(OH)₂]₃·2H₂O). Gražėnaitė⁸ likewise identified the hydrous neutral phase—contradictory to the manufacturer’s claim—together with other additional unassigned peaks (2θ = 19.76° and 2θ = 43.12°), suggesting the pigment was not entirely pure. Still, the phases reported by the manufacturer may be present in low proportion and therefore below the detection limit of the XRD technique (3 wt%). In fact, the ATR-FTIR spectrum of the powder pigment showed bands (2990 and 2945 cm^–1) which are ascribable to the presence of other acetate compounds^7,22,47,48; thus, the presence of other phases in minor amounts cannot be excluded.

In a secco painting, where egg yolk and rabbit glue were used as binders (6.54 and 5.82 pH units, respectively), the pigment remained physically, mineralogically and chemically stable. However, it is well established that verdigris can react with the fatty acids present in the binder—such as stearic, palmitic, oleic or linoleic acids—to form carboxylates (metal soaps) through a saponification process⁵⁹. In fact, soap formation is part of the natural drying process of this type of paint and is therefore always expected to form⁶⁰. They play a significant role in degradation processes with an undesirable impact on appearance (darkening) and stability (cracking)^25,59,61. Still, our analysis found no evidence of such compounds. The characteristic FTIR band around 1600–1500 cm^–1, associated with asymmetric COO^– stretching in free fatty acids, which would suggest soap formation^{60, 61, 62–63}, was absent from the spectrum. Nonetheless, the absence of soaps in our secco paints may indicate that saponification has not yet occurred—given that it is a long-term process in paintings⁶¹—or that the soaps are present only in trace amounts, insufficient to cause immediate colour changes. Still, the chemistry behind metal soaps is a complicated issue as they can adopt structures ranging from amorphous to crystalline⁶⁴.

Quite so, the alkaline environment of mezzo-fresco and fresco techniques (lime water and lime paste showed 11.56 and 11.92 pH units, respectively) promoted the conversion of bluish-green neutral verdigris into a range of secondary compounds. In the verdigris–lime water mixtures, copper acetates previously described as basic forms of verdigris—specifically phases 1–3–2 (Cu₂(CH₃COO)(OH)₃·H₂O) or 2–1–5 (Cu₃(CH₃COO)₄(OH)₂·5H₂O)^6,10—were identified. Although these phases have very similar XRD patterns, they can be distinguished by their morphology: Bette et al.¹⁰ reported a layer-like morphology for phase 1–3–2 and a needle-like morphology for phase 2–1–5. Our observation by stereomicroscopy (Fig. 3a) and SEM (Figs. 5f and 6a) of samples VER-MF_2 and VER-MF_3 revealed a needle-like morphology consistent with that reported for phase 2–1–5¹⁰. Furthermore, because phase 1–3–2 is metastable under conditions of high hydroxide supersaturation—such as the ones created by lime water (Ca(OH)₂)—it may also form as an intermediate phase before transforming into the 2–1–5 phase⁶. Taken together, these observations suggest that Cu₃(CH₃COO)₄(OH)₂·5H₂O (the 2–1–5 phase) likely forms as an alteration product of neutral verdigris in strongly alkaline environments. Indeed, it has been previously reported that neutral verdigris can transform into basic verdigris phases^15,16. A dark-blue calcium-copper acetate hexahydrate known as paceite (CaCu(CH₃COO)₄·6(H₂O)) was also identified in some mezzo-fresco samples. This compound has been reported as an alteration product of neutral verdigris in fresco painting, forming through the reaction with calcium hydroxide⁴². In support of this mechanism, Li et al. ⁶⁵ successfully synthesised paceite by reacting calcium hydroxide or calcium carbonate with copper acetate (hoganite) in the presence of acetic acid.

Calcium acetate—either monohydrate (Ca(CH₃COO)₂·H₂O) or hemihydrate (Ca(CH₃COO)₂·(H₂O)_0.5)—was identified in both MF and F samples. Although no studies have directly described the hemihydrate phase, related work has demonstrated that the monohydrate (Ca(CH₃COO)₂·H₂O) can be synthesised from mussel shells and acetic acid⁶⁶. These exhibited thin, needle-like crystals similar to those identified in our samples (e.g. Figs. 5f and 6e). This suggests that the phase present in our samples may correspond to the monohydrate phase; however, calcium acetates are known to adopt diverse hydrated forms with comparable morphologies⁶⁷. While we cannot determine the exact hydration state of the calcium acetate in our samples, its identification represents, to our knowledge, the first report of this compound as an alteration product of verdigris.

In contrast, only the fresco samples (-F) showed the presence of anhydrous verdigris (Cu(CH₃COO)₂), a phase already present in trace amounts in the raw pigment. As it is known that temperatures exceeding 90 °C are required for its formation, it cannot be considered as an alteration of neutral verdigris⁹. Tenorite (CuO)—long recognised as a final blackening product in copper-based paintings⁴¹—was also identified and is known to form through decomposition of neutral verdigris (Cu(CH₃COO)₂·H₂O)⁶⁸. Tenorite particles display morphologies comparable to fibrous nano-ribbons⁶⁹, micro-tubes⁷⁰ and micro-spheres⁷¹ morphologies, similar to those observed in Figs. 5f and 6f, j, respectively. Finally, some XRD patterns of the F samples showed an unassigned XRD peak at 2θ = 10.30°, which likely corresponds to an unidentified metal carboxylate. This suggests the presence of additional copper-, calcium-, or mixed copper-calcium acetate phases of yet undetermined composition. The ATR-FTIR results support this interpretation. The presence in the FTIR spectra of broad hydroxyl (–OH) bands, different in each spectrum, confirms an ongoing modification and the presence of hydrated compounds. This could be related to the aforementioned compounds (various hydrated acetates and/or copper hydroxides) as suggested by the widespread presence of acicular and ribbon-like morphologies. Moreover, the shift of COO^– asymmetric stretching to lower wavenumbers, as observed in our spectra (shifts of up to −70 cm^–1), suggests a mixture of compounds is formed⁴⁷. This was further supported by the changes observed related to Cu–O and Cu–OH bonding.

Regarding morphological transformations, the results obtained in this study point to a diverse and complex set of alteration products. The needle-like morphologies (Figs. 5i and 6a, f–h) are most plausibly associated with the following compounds identified in this study: Cu₃(CH₃COO)₄(OH)₂·5H₂O, diverse hydrated forms of calcium acetates (C₄H₆CaO₄·nH₂O), or copper oxides such as tenorite (CuO). Nevertheless, needle-like forms were diverse, and the presence of other compounds not detected by XRD cannot be excluded. For instance, it is well-known that verdigris pigments can transform into blue copper hydroxides, such as spertiniite (Cu(OH)₂), in the presence of alkalis as an intermediate step in copper oxide formation^1,55. This compound can also form with a needle-like (or acicular) morphology^{72, 73–74}. Moreover, nano-ribbon morphologies have also been reported for Cu(OH)₂, which highly resembles the morphology observed in Fig. 5f^{72, 73–74}. Finally, the tabular, rosette-like habits observed (Fig. 5g, h) are most likely linked to other forms of copper acetate since similar morphologies to ours have been reported by Raza et al.⁷⁵ who synthesised Cu(II) carboxylates with monoclinic symmetry.

In summary, this study highlights the pronounced instability of neutral verdigris when applied a mezzo-fresco and a fresco, leading to the formation of a wide variety of secondary compounds. These findings contrast with its documented presence in historical works, raising important questions. It is likely that this discrepancy arises from misattributions to the original painting technique. Even so, future research should investigate whether verdigris can be stabilised in such alkaline environments to enable its practical use in painting. Moreover, unresolved XRD peaks and infrared spectral shifts suggest the presence of additional carboxylates or copper oxides/hydroxides, indicating a more complex alteration pathway than previously recognised. The need for advanced micro-analytical methods, such as synchrotron radiation micro-X-ray powder diffraction (SR-μPXRD), to fully resolve the complex formation of secondary phases arising from verdigris alteration, is highlighted.

Acknowledgements

Daniel Jiménez-Desmond was supported by the ED481A-2023/086 predoctoral contract through “Programa de axudas á etapa predoutoral da Xunta de Galicia” cofinanced by the European Union within the framework of the FSE+ Galicia 2021–2027 programme. Anna Arizzi was funded by the Research Project PID2023-146405OB-100 (2024–2027, funded by MICIU/AEI/10.13039/501100011033 and ERDF, EU), and by the Junta de Andalucía Research Group RNM179. XRD and SEM-EDS were carried out at the Centro de Apoio á Investigación Científico-Tecnolóxica from the University of Vigo. Funding for the open access charge was provided by the Universidade de Vigo/CISUG. This study was funded by the Xunta de Galicia ED431F 2022/07 research project.

Author contributions

D.J.-D.: writing—original draft. J.S.P.-A. and A.A.: writing—review and editing. D.J.-D.: data curation and investigation. D.J.-D, J.S.P.-A., and A.A.: formal analysis. J.S.P.-A. and A.A.: funding acquisition, project administration and supervision. All authors reviewed the manuscript.

Data availability

All data needed to evaluate the conclusions are present in this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s40494-025-02291-9.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Kühn, H. Verdigris and copper resinate in artists pigments. A Handbook of Their History and Characteristics Vol. 2 (ed. Roy, A.) 131–156 (Archetype Publications, 1993).

2. Eastaugh, N., Walsh, V., Chaplin, T. & Siddall, R. Pigment Compendium: a Dictionary and Optical Microscopy of Historical Pigments (Elsevier Ltd., 2007).

3. Scott, DA; Taniguchi, Y; Koseto, E. The verisimilitude of verdigris: a review of the copper carboxylates. Stud. Conserv.; 2001; 46, pp. 73-91.

4. Scott, D. A. Copper and Bronze in Art. Corrosion, Colorants, Conservation (Getty Publications, 2002).

5. Bette, S; Eggert, G; Dinnebier, RE. Crystalline materials in art and conservation: verdigris pigments—what we know and what we still don’t know. Phys. Sci. Rev.; 2023; 9, 8.

6. Bette, S et al. On verdigris, part I: synthesis, crystal structure solution and characterisation of the 1-2-0 phase (Cu₃(CH₃COO)₂(OH)₄). Dalton Trans; 2017; 46, pp. 14847-14858. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29043336]

7. Prati, S et al. ATR-FTIR microscopy in mapping mode for the study of verdigris and its secondary products. Appl. Phys. A Mater. Sci. Process.; 2016; 122, pp. 1-16.

8. Gražėnaitė, E. Inorganic Green Pigments: Investigation of Historical and Synthesis of Novel Pigments by Sol-Gel Method Doctoral dissertation (Vilnius University, 2018).

9. Bette, S et al. On verdigris, part III: crystal structure, magnetic and spectral properties of anhydrous copper(II) acetate, a paddle wheel chain. ZAAC; 2019; 645, pp. 988-997.

10. Bette, S; Kremer, RK; Eggert, G; Dinnebier, RE. On verdigris, part II: synthesis of the 2-1-5 phase, Cu3(CH3COO)4(OH)2·5H2O, by long-term crystallisation from aqueous solution at room temperature. Dalton Trans; 2018; 47, pp. 8209-8220. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29870036]

11. Švarcová, S; Hradil, D; Hradilová, J; Čermáková, Z. Pigments—copper-based greens and blues. Archaeol. Anthropol. Sci.; 2021; 13, 190.

12. Aceto, M. Pigments—the palette of organic colourants in wall paintings. Archaeol. Anthropol. Sci.; 2021; 13, 159.

13. Hofmann, C et al. Stabilization of Verdigris. J. Pap. Conserv.; 2016; 17, pp. 88-99.

14. Bhowmik, SK. A Note on the use and deterioration of verdigris in indian watercolour painting. Stud. Conserv.; 1970; 15, pp. 154-156.

15. Brostoff, LB; Connelly Ryan, C. Tracing the alteration of verdigris pigment through combined Raman spectroscopy and X-ray diffraction. Part I. Restaurator; 2020; 41, pp. 3-30.

16. Brostoff, LB; Connelly Ryan, C; Black, I. Tracing the alteration of verdigris pigment through combined Raman spectroscopy and X-ray diffraction, part II: natural ageing. Restaurator; 2020; 41, pp. 177-203.

17. Cennini, C. Libro Dell’arte (Akal, 1998).

18. Da Vinci, L. Tratado de Pintura (Ediciones Akal S.A., 2004).

19. Mayer, R. Materiales y Técnicas del Arte (Hermann Blume Ediciones, 2005).

20. Doerner, M. The Materials of the Artist and Their use in Painting, with Notes on the Techniques of the Old Masters (Houghton Mifflin Harcourt, 1998).

21. Buse, J; Otero, V; Melo, MJ. New insights into synthetic copper greens: the search for specific signatures by Raman and infrared spectroscopy for their characterization in Medieval artworks. Heritage; 2019; 2, pp. 1614-1629.

22. San Andres, M. et al. Verdigris pigment: a mixture of compounds. Input from Raman spectroscopy. J. Raman Spectrosc. 41, 1468–1476 (2010).

23. Hofmann, C et al. Studies on the conservation of verdigris on paper. Restaurator; 2015; 36, pp. 147-82.

24. Ahn, K; Hartl, A; Hofmann, C; Henniges, U; Potthast, A. Investigation of the stabilization of verdigris-containing rag paper by wet chemical treatments. npj Herit. Sci.; 2014; 2, 12.

25. Punis, R; Zoleo, A. Copper complexes in verdigris painting mixtures: an electron paramagnetic resonance characterization. Restaurator; 2023; 45, pp. 2-3.

26. De la Roja, JM; San Andrés, M; Cubino, NS; Santos-Gómez, S. Variations in the colorimetric characteristics of verdigris pictorial films depending on the process used to produce the pigment and the type of binding agent used in applying it. Color Res. Appl.; 2007; 32, pp. 414-423.

27. Zoleo, A et al. Characterization of pigment and binder in badly conserved illuminations of a 15th-century manuscript. Archaeometry; 2014; 56, pp. 496-512.

28. Ortiz-Miranda, AS et al. Analyzing chemical changes in verdigris pictorial specimens upon bacteria and fungi biodeterioration using voltammetry of microparticles. Herit. Sci.; 2017; 5, 8.

29. Punis, R; Veronese, M; Meneghetti, M; Zoleo, A. Copper soaps formation in verdigris-linseed oil painting mixtures: a multispectroscopic characterization. Inorg. Chem.; 2024; 63, pp. 4108-4119. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38372641]

30. Punis, R; Zoleo, A. Exploring the early stages of verdigris-linseed oil mixture curing by EPR spectroscopy. J. Mol. Struct.; 2026; 1350, 144016.

31. Santoro, C et al. New highlights on degradation process of verdigris from easel paintings. Appl. Phys. A; 2014; 114, pp. 637-645.

32. Ioakimoglou, E et al. Thin-film study on the oxidation of linseed oil in the presence of selected copper pigments. Chem. Mater.; 1999; 11, pp. 2013-2022.

33. Alter, M et al. Alteration markers of green copper bimetallic pigments in easel paintings by an inter-comparison of analogues and historical samples. Microchem. J.; 2025; 209, 112541.

34. Rollier-Hanselmann, J. D’Auxerre à Cluny: technique de la peinture murale entre le VIII et le XII siècle en Bourgogne. Cahiers de Civilis. Médiév.; 1997; 40, pp. 57-90.

35. Howard, H. Pigments of English Medieval Wall Painting (Archetype Publications, 2003).

36. Damiani, D; Gliozzo, E; Memmi, IT. The ‘Madonna and Child Enthroned with Saints’ of Ambrogio Lorenzetti in the St. Augustine Church (Siena, Italy): Raman microspectroscopy and SEM-EDS characterisation of the pigments. Archaeol. Anthropol. Sci.; 2014; 6, pp. 363-371.

37. Jáudenes Ruiz de Atauri, Í. Estudio científico de la pintura mural al fresco de Mariano De Cossío. Aportaciones al conocimiento de los murales del siglo XX en Canarias. Doctoral dissertation (Universidad de La Laguna, 1999).

38. Beltrán, Plaza. M. Submerged villages. Recovering wall paintings from the church of Atance (Guadalajara, Spain). Technical study, exhibition, and 3D display. Int. J. Conserv. Sci.; 2016; 7, pp. 1031-1046.

39. Botticelli, G. Metodologia di Restauro Delle Pitture Murali (Centro Di, 1992).

40. Merrifield, M. P. The Art of Fresco Painting in the Middle Ages and the Renaissance (Dover Publications Inc., 2004).

41. Jiménez-Desmond, D. et al. Blackening of copper pigments in wall paintings: impact of the fresco technique and the chemical composition of the pigments. npj Herit. Sci. (UNDER PEER REVIEW) (2025).

42. Jiménez-Desmond, D; Pozo-Antonio, JS; Arizzi, A; López-Martínez, T. Physico-chemical compatibility of an aqueous colloidal dispersion of silica nano-particles as binder for chromatic reintegration in wall paintings. Applied Sciences; 2025; 15, 3690.

43. Jiménez-Desmond, D; Pozo-Antonio, JS; Arizzi, A. The fresco wall painting techniques in the Mediterranean area from Antiquity to the present: a review. J. Cult. Herit.; 2024; 66, pp. 166-186.

44. Arizzi, A; Cultrone, G. The influence of aggregate texture, morphology and grading on the carbonation of non-hydraulic (aerial) lime-based mortars. QJEGH; 2013; 46, pp. 507-520.

45. CIE S014-4/E:2007. Colorimetry-part 4: CIE 1976 L* a* b* Colour Space (International Standard, 2007).

46. Commission Internationale de l’Eclairage (CIE) (1986) Colorimetry, 2nd edn (Publication CIE No. 15.2. Commission Internationale de l’Eclairage, Vienna).

47. Salvadó, N et al. Shades of green in 15th century paintings: combined microanalysis of the materials using synchrotron radiation XRD, FTIR and XRF. Appl. Phys. A Mater. Sci. Process.; 2013; 111, pp. 47-57.

48. Tackett, JE. FT-IR Characterization of metal acetates in aqueous solution. Appl. Spectrosc.; 1989; 43, pp. 483-489.

49. Vahur, S; Teearu, A; Leito, I. ATR-FT-IR spectroscopy in the region of 550-230 cm^-1 for identification of inorganic pigments. Spectrochim. Acta A Mol. Biomol. Spectrosc.; 2010; 75, pp. 1061-1072. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20061180]

50. Nodari, L. & Ricciardi, P. Non-invasive identification of paint binders in illuminated manuscripts by ER-FTIR spectroscopy: a systematic study of the influence of different pigments on the binders’ characteristic spectral features. npj Herit. Sci. 7, 7 (2019).

51. Mazzeo, R et al. Attenuated total reflection micro FTIR characterisation of pigment-binder interaction in reconstructed paint films. Anal. Bioanal. Chem.; 2008; 392, pp. 65-76. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/18454281]

52. Cardell, C et al. Pigment-size effect on the physico-chemical behavior of azurite-tempera dosimeters upon natural and accelerated photo aging. Dyes Pig.; 2017; 141, pp. 53-65.

53. Pellegrini, D et al. Fourier transform infrared spectroscopic study of rabbit glue/inorganic pigments mixtures in fresh and aged reference paint reconstructions. Microchem. J.; 2016; 124, pp. 31-35.

54. Socrates, G. Infrared and Raman Characteristic group frequencies: Tables and Charts (John Wiley & Sons LTD, 2004).

55. Soroka, IL; Shchukarev, A; Jonsson, M; Tarakina, NV; Korzhavyi, PA. Cuprous hydroxide in a solid form: Does it exist?. Dalton Transactions; 2013; 42, pp. 9585-9594. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23673918]

56. Fernández-Carrasco, L., Torrens-Martín, D., Morales, L. M. & Martínez-Ramírez, S. Infrared spectroscopy in the analysis of building and construction materials. In Proc. Infrared Spectroscopy: Materials Science, Engineering and Technology (ed. Theophanides, T.) 369–381 (InTech, 2012).

57. Derrick, M. R., Stulik, D. & Landry, J. M. Scientific Tools for Conservation. Infrared Spectroscopy in Conservation Science (The Getty Conservation Institute, 1999).

58. Galván-Ruiz, M; Hernández, J; Baños, L; Noriega-Montes, J; Rodríguez-García, ME. Characterization of calcium carbonate, calcium oxide, and calcium hydroxide as starting point to the improvement of lime for their use in construction. J. Mater. Civ.; 2009; 21, pp. 694-698.

59. Russo, S; Brambilla, L; Thomas, JB; Joseph, E. But aren’t all soaps metal soaps? A review of applications, physico-chemical properties of metal soaps and their occurrence in cultural heritage studies. npj Herit. Sci.; 2023; 11, pp. 1-18.

60. Salvadó, N et al. Identification of reaction compounds in micrometric layers from gothic paintings using combined SR-XRD and SR-FTIR. Talanta; 2009; 79, pp. 419-428. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19559899]

61. Svarcová, S et al. Saponification in egg yolk-based tempera paintings with lead-tin yellow type I. J. Cult. Herit.; 2019; 38, pp. 8-19.

62. Gunn, M; Chottard, G; Rivière, E; Girerd, JJ; Chottard, JC. Chemical reactions between copper pigments and oleoresinous media. Stud. Conserv.; 2002; 47, pp. 12-23.

63. Poli, T; Piccirillo, A; Nervo, M; Chiantore, O. Interactions of natural resins and pigments in works of art. J. Colloid. Interface. Sci.; 2017; 503, pp. 1-9. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28500934]

64. Hermans, J. J. et al. The synthesis of new types of lead and zinc soaps: a source of information for the study of oil paint degradation. In ICOM-CC 17thTriennial Conference, Building Strong Culture Through Conservation 17–19 (International Council of Museums, Melbourne, 2014).

65. Li, S et al. Mechanochemical synthesis, accelerated aging, and thermodynamic stability of the organic mineral paceite and its cadmium analogue. ACS Omega; 2019; 4, pp. 5486-5495. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31459711][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6649266]

66. Murphy, J. N., Morgan, M. A., Christian-Robinson, S., Fitzgerald, M. M. & Kerton, F. M. Optimized process to produce calcium acetate from waste blue mussel shells and its use as a de-icer. Can. J. Chem. 103, 913–922 (2025).

67. Bette, S et al. Crystal structure, polymorphism, and anisotropic thermal expansion of α-Ca(CH₃COO)₂. Cryst. Growth. Des.; 2020; 20, pp. 5346-5355.

68. Naktiyok, J; Özer, AK. Synthesis of copper oxide (CuO) from thermal decomposition of copper acetate monohydrate (Cu(CH₃COO)₂·H₂O). NOHU J. Eng. Sci.; 2019; 8, pp. 1292-1298.

69. Li, H et al. Copper-based nanoribbons fabricated on a copper substrate by a liquid-solid reaction and their corrosion performance. Mater. Chem. Phys.; 2020; 246, 122839.

70. Balassone, G et al. Copper minerals at vesuvius volcano (Southern Italy): a mineralogical review. Minerals; 2019; 9, 730.

71. Renuga, D; Jeyasundari, J; Shakthi Athithan, AS; Brightson, Arul; Jacob, Y. Synthesis and characterization of copper oxide nanoparticles using Brassicaoleracea var. italic extract for its antifungal application. Mater. Res. Express; 2020; 7, 045007.

72. Hyam, RS; Lee, J; Cho, E; Khim, J; Lee, H. Synthesis of copper hydroxide and oxide nanostructures via anodization technique for efficient photocatalytic application. J. Nanosci. Nanotechnol.; 2012; 12, pp. 8396-8400. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23421221]

73. Ma, M et al. Controllable hydrocarbon formation from the electrochemical reduction of CO₂ over Cu nanowire arrays. Angew. Chem. Int. Ed.; 2016; 55, pp. 6680-6684.

74. Chiguma, JE; Jones Junior, W. Template-free synthesis of aligned polyaniline nanorods/tubes and copper/copper hydroxide nanowires for application as fillers in polymer nanocomposites. AMPC; 2018; 8, pp. 71-88.

75. Raza, H et al. Synthesis of a 2D copper(II)-carboxylate framework having ultrafast adsorption of organic dyes. J. Colloid. Interface Sci.; 2021; 602, pp. 43-54. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34118604]