In the subsequent sections, the results of the study will be reported and discussed as follows. Firstly, the Ernst’ painting techniques used in the different paintings will be explained according to the outcomes from the visual observation, thanks also to the use of microphotography. The multispectral imaging, instead, allowed to emphasize hidden details, preparatory drawings and/or overpainting. Later, the findings about the materials (used by the artist or alteration products) obtained by means of spectroscopic investigations (XRF, Raman and ER-FTIR) will be presented. The common features, such as the composition of the substrate, the used binder and the alteration products will be discussed first. Once these aspects have been clarified, the palette of each artwork will be illustrated separately and, finally, its evolution through time will be described.
Imaging techniques
On The Kiss (KI) a frottage base was likely applied over the entire surface of the painting. The author used either wood or other heavy material to produce this texture as some parts of the artworks suggest—such as the black shadow to the left of the central figure (see Fig. 2a, b) [7]. The principal lines composing the oeuvres were likely created through the use of a string left falling on the preparatory surface and subsequently partly repainted with brush (see Fig. 2c, e) [15]. The blue tone was applied once the rest of the composition had dried completely (see Figs. 2c, e and 3a–c) [7].
Details from The Kiss (a–e), Zoomorphic Couple (f, g), The Entire City (h, i) and The Antipope (j–m). The dark texture of wood is visible in (a) and (b) while in the particulars due to the use of a string are clear in (c) and from the digital microscope images (d) and (e) (100X of magnification). The superimposition of different paint layer, attributable to different painting time, are evident especially in (a), (c) and (e). The frottage and dripping techniques can be recognized in (f) and (g). The pattern obtained by using textile blocks is highlighted in (h) while the details in (i) show the effect of abrasion on design and texture. From (j) to (m) the patterns obtained by printing of leaf, of the use of a sponge and decalcomania are illustrated
The Kiss (a), detail: comparison between RGB (b) and Chroma DI images (c). The analysis suggests that, probably, the blue tone was applied once the rest of the composition had dried completely. The different shades indicate, in fact, differences in the density, and thus in the quantity, of pigments used
Zoomorphic couple (ZC), on the other hand, was created with a complex technique mostly based on frottage and dripping [7]. Ernst prepared a base of pastel colours on which several paints were subsequently spattered, blown and dripped (see Fig. 2f, g). In particular, the dark parts were created by placing a paint-laden rope or string over the canvas and spraying over it [6]. Then the cream background was probably applied to define the figures and the outer contours [7] (Fig. 4a).
a Chroma DI (1050–650, 650–550, 550–450 nm) imaging of Zoomorphic Couple: in the upper part is clearly visible (in light green) a stain in the sky, due to a discoloration, whereas the different shades suggest an inhomogeneity of the background; b Chroma DI (1050–650, 650–550, 550–450 nm) detail of the Garden Airplane Trap. The floral detail was clearly done after painting the underside, as seen by the visible striped outlines. Some construction lines are also detectable, as highlighted by the different shades in the chroma DI details
In Garden Airplane Trap (GA), the chroma DI image (1050–650, 650–550, 550–450 nm) [16, 17] can provide information about the technique used by the artist (Fig. 4). In some parts, a pencil preparatory drawing is visible. On the background, created with the parallelepipeds, Ernst overpainted the airplanes and, over them, other elements (flowers, fruit and leaves) are clearly overpainted (see Fig. 4b).
In The Entire City (EC), in the upper left of the sky, the false colour image reveals a clearer stain (Additional file 1: Figure S7) that could be related to a moon/sun, which was successively covered during the painting process. In most of the works belonging to this series, a moon/sun is present in the sky, moreover, to achieve a continuous pattern of the city walls, Ernst employed decorated relief blocks (for textile printing). The flattening and the abrasion of the surface of this specific Entire City obliterate both design and texture, making it difficult to observe the rich technical process that is present in other examples (see Fig. 2h, i).
The Antipope (AP) was built with a combination of techniques such as decalcomania, traditional brushwork, and the printing of leaves into wet paint (Fig. 5). The lower landscape with birds, fossil-like vegetation, moss, shells, frogs, and rocks was created with a decalcomania base on which Ernst then added brushwork (Fig. 4a, b). Other elements such as the red dress, the horse-headed figure with an owl mask on the left, and the skeletal figure in the centre were obtained with similar technical combinations [7]. The complexity of the execution technique is highlighted in photographic details reported in Fig. 2j-m.
Results regarding the Attirement of the bride (AB) are thoroughly discussed in a paper dedicated to it: the imaging shows some preparatory drawing, such as prospective construction lines and afterthoughts (pentimenti) [9].
Materials characterization
Common features
In general, XRF reveals the presence of Zn, S, Ba, and Pb distributed overall in works studied. This evidence suggests the presence of a ground layer composed by lithopone, BaSO4 + ZnS, and lead white, 2PbCO3∙Pb(OH)2. Signals due to overtones and combination bands of SO4= in barite (2ν3 at ≈2230 cm−1, ν1 + ν3 at ≈2194–2137-2065 cm−1 and 2ν1 at ≈1965 cm−1 [18]) were detected by ER-FTIR (Fig. 6a, b) in all the canvases, except for KI. In this painting the absence of barite’ signals could be ascribed to a thicker pictorial layer than the one present in the other paintings. Another common feature, highlighted by XRF, is the ubiquitous presence of Ca. Ca can be associated with the presence of calcite, as confirmed by Raman (ν1at 1086 cm−1) and, except for ZC, by ER-FTIR (ν1 + ν4 at ≈1796 cm−1), as illustrated in Fig. 6. In all the paintings, the paint binder can be easily identified in ER-FTIR spectra (Fig. 6) by the presence of a well-resolved doublet, at 4320 and 4260 cm−1, attributable to the stretching and bending combination bands (νa/s + δa/s) of the methylene units in drying oils [19]. These signals are combined with the ester asymmetric ν(C = O), at ≈1740 cm−1, which shape suffers the distortions due to the physical features of the pictorial layer (i.e., the execution technique) and to the superimposition with other organic materials (i.e., varnishes from past restoration).
The results above reported underline some common features present in all the paintings. In addition to the use of a drying oil as a binder, the results show that the ground layers, probably commercially applied, are composed of lithopone, and lead white in all instances. Respect to lead white, which was used in Europe from the Roman period onwards as the principal white pigment and employed both as a preparation layer [20] and frequently to give a certain opacity in the flesh tones [21, 22], lithopone was commercialized only since 1874 [20]. In some cases, calcium carbonate was also detected in different areas. However, it might have been present in the paint tube as an extender. The presence of a preparatory layer is confirmed by a previous micro-sampling from AB, which showed the presence of a white ground layer [9]. µ-Raman performed on these micro-samples revealed the presence of lithopone and lead white in one case and lithopone and calcium carbonate in a second case. The use of a commercially applied ground layer is confirmed by an old conservation report [23], indicating the use of a commercial primed canvas, and supported by literature data [5].
State of conservation
Another interesting aspect, unearthed by ER-FTIR, concerns the presence of alteration products. As matter of fact, except for the GA and the EC, all paintings show evidence to metal soaps and oxalate. In KI, both zinc oxalate (ZnC2O4·2H2O) and zinc metal soaps are revealed in different tones (Additional file 1: Figure S12). Similar evidence were identified in AB and AP (Additional file 1: Figure S6). In these paintings, both metal soaps and oxalate are uniformly distributed over the surface and the signals due to zinc oxalate are in general more intense in comparison with the stearate ones. ZC presents a different condition: the signals due to ZnC2O4·2H2O seem to be more intense in the light tones while the Zn soaps ones are more evident on the dark hues. All the spectra collected in the light tones show the insurgence of a new absorption, centered at ≈1700 cm−1, that can be attributed to the νa(CO) in zinc oxalate. Usually, this signal appears as a derivative band [24], but, probably as consequence of the frottage, the rough surface enhances the volume reflection modifying the shape of the absorption.
It should be observed that zinc white was detected only on KI. In all the other artworks the presence of Zn could be related to the use of ZnO as drier [25] or with the presence of zinc stearate as fast wetting and dispersant for pigments [26, 27] in the paint tube formulations. Considering this hypothesis, zinc soaps can be present in the original paint formulation, as observed by Izzo et al. [26] or can be originated from the reaction between ZnO and the free fatty acids generated through ester hydrolysis of triglycerides as proposed by Hermans and co-workers [25]. The formation of zinc stearate can be promoted also by the presence of aluminum stearate, a very common rheology modifier as suggested by Osmond and co-workers [28]. On the contrary, metal oxalates are definitely alteration products: for instance, zinc oxalate was also detected in Pollock’s Alchemy in the holdings of the Peggy Guggenheim Collection in Venice [29].
Max Ernst’s palette
The evolution of the palette, with an overview of the pigments used, is hereafter reported, highlighting the distinctive results. Table 1 reports a synoptic overview of the pigments.
The Kiss (1927)
The blue-azure background and part of the vest are realized with Prussian blue; in fact, Raman spectra (Additional file 1: Figure S8a) show the stretching vibration of the lattice Fe–C at 532 cm−1 [30] and the stretching vibration of C≡N group at 2093 and 2154 cm−1 [31]. The ER-FTIR spectra in the blue-azure areas (Additional file 1: Figure S8b), confirm the presence of Prussian blue; moreover, they are characterized by a strong absorption at ≈512 cm−1 (Additional file 1: Figure S2b), tentatively attributed to ν(ZnO) [32]. This hypothesis is supported by XRF data that show an intense peak due to Zn in correspondence to the blue tones. On the brown-yellow paints, ER-FTIR analysis reveals the presence of silicates, probably kaolin, with the characteristic inverted ν(SiO) at 1030 cm−1 and the ν(OH) at 3695 cm−1 [18] (Additional file 1: Figure S9). This evidence, together with the intense peak of Fe, detected by XRF, supports the use of ochre [33]. In the same tones, XRF also indicates the presence of Cd and S, suggesting the use of cadmium yellow, CdS, to impart the yellow shades. On the contrary, in the brown-orange areas, Raman spectrum (Additional file 1: Figure S10) shows the chromate symmetric stretching (ν1) at 340, 822 and 845 cm−1 that, corresponding to Cr and Pb detected by XRF, are ascribable to the use of a chrome yellow orange (PbCrO4·PbO) [34,35,36]. Finally, the black lines correspond to carbon black with typical Raman absorption bands at 1336 and 1590 cm−1, D and G bands respectively; the first one is often called the ‘disorder’ band, attributed to a mode with A1g symmetry while the second one is assigned to the in-plane stretching vibration mode (E2g) [37, 38].
Zoomorphic Couple (1933)
Raman and XRF analyses on the white-yellow background show the presence of lithopone. S, Zn and Ba were detected, together with the Raman symmetric stretching of sulphate group (988 cm−1), also in the yellowish shades of the tentacles in the central part of the canvas. In the same areas, XRF reveals the presence of Sr and Cr, indicative of the use of strontium yellow, SrCrO4 (Fig. 7a). The blue shade of the tentacle is due to Prussian blue, confirmed by both Raman and ER-FTIR techniques, whereas the black tone corresponds to carbon black, identified by Raman as well. Finally, in the red shade on the tentacle, vermilion (HgS) is identified by Raman (Fig. 7b) [39, 40] and confirmed by the presence of Hg peak in the XRF spectrum. ZC presents an evident pinkish stain in the upper left probably due to discoloured reintegration [7]. The chroma DI image (1050–650, 650–550, 550–450) highlights the reintegration (Fig. 3a) together with an evident inhomogeneity in the background, probably due to the use of different shades of pigment. Raman measurements on the stain identify lead white, by the symmetric stretching CO3= group at 1050 cm−1 [39], and hematite, Fe2O3, by the Fe–O symmetric bending at 225 and 291 cm−1 (figure S11) [41]. At first glance, ZC may appear painted with a limited palette based on light/dark contrast; however, the analyses confirm the use of different pigments to render yellow, blue, and red shades on the dark tentacles of the figures.
Garden Airplane Trap (1935–1936)
The paint is dominated by light brownish-yellow tones, whose elemental composition reveals the ubiquitous presence of Sr and Cr and, in some points, of Cd and S. These elements suggest the use of strontium and cadmium yellow, respectively. The palette of the bluish areas is more complex: in the blue areas ER-FTIR and Raman clearly highlight the use of Prussian blue probably together with cobalt blue, as suggested by the Co signal in the XRF spectra and according to the literature [1]. Greenish areas show the use of Prussian blue mixed with strontium yellow and/or cadmium yellow, as suggested by the presence of Sr, Cr, Cd and S in the XRF spectra. The presence of Mn, together with silicates, detected in the ER-FTIR spectra, suggest the use of raw umber in some brownish-green shades. The red shade of the flowers is due to cadmium red, CdS1-xSex, as suggested by the signals of Cd, S and Se in the XRF spectra (Fig. 7c).
The Entire City (1937)
The painting presents an upper part (the sky), characterized by a shade from a lighter green to a white/yellowish tone and a lower part, where the tangle vegetation invades the step/pyramidal ruins, mostly composed by red, brown, and dark green tones. In the upper part, XRF evidences the presence of Cr and Sr; therefore, we can suppose the use of strontium yellow. However, since this part is mostly green and no blue pigment is evidenced, also the use of a chrome-based green (Cr2O3 or Cr2O3·2H2O) cannot be excluded. In the white-yellowish hues, XRF identifies the presence of Cd and S along with Cr and Sr; therefore, the use of both strontium and cadmium yellow can be assumed. Added to these pigments, the Raman technique identifies calcium carbonate and barium sulphate probably to give a lighter shade. The darker green tones in the lower part of the painting are associated with Prussian blue, showed by ER-FTIR analysis, added to strontium yellow and cadmium yellow recognized by XRF. Moreover, the presence of the signals of Fe and Mn in XRF spectrum and silicates peaks in ER-FTIR could be attributed to the presence of a raw umber, also identified by XRF in some red paints. In other red paints, instead, XRF technique reveals both Cd and Se associated with the use of cadmium red.
Attirement of the Bride (1940)
A detailed characterization of the AB is reported in literature [9]. In this painting, two different blue pigments are used: copper phthalocyanine is identified in the main blue tones, while Prussian blue is mixed with other pigments to obtain several shades (green, violet, flesh and yellow). Two different red tones are recognized: an orange-red one with iron oxide and a more intense tone with an azo β-naphthol pigment. In both cases, also lead white and lithopone are identified. The lead white present in red tone could have been used as a pigment or it could be referred to the white ground layer (see above). The violet tones are a mixture of Prussian blue and iron oxide, while lithopone and calcium carbonate are probably added to give a lighter shade. Furthermore, XRF reveals the presence of cobalt. The yellow tones are obtained with barium chromate and in some points also ochre is recognised. The green paint is probably due to a mixture of Prussian blue and a yellow pigment, likely barium chromate, with the addition of a cobalt-based pigment. The white paints are a mixture of anatase (TiO2), lithopone, and calcium carbonate. In the flesh tones several pigments are identified: lithopone and/or calcium carbonate, silicates, iron oxide and a raw umber are mixed for the base colour, while several shades are obtained by adding Prussian blue and a cobalt-based pigment. Lastly, the black paints are obtained with carbon black, while the brown paints are a mixture of iron oxide and a raw umber, identified for the presence of Mn.
The Antipope (1941–1942)
In the flesh tones titanium dioxide, detected by Raman in its anatase form [35] is used along with cadmium red, identified by the presence of Cd and Se in the XRF spectra of the abdomen of the figure on the right, or with iron oxide, revealed by Raman (Fig. 8a). In the legs of the figure on the right, XRF also shows Sr and Cr; therefore, strontium yellow could have been added to give a particular shade. Cadmium red is probably present also in some red paints such as the dress of the figure on the left since XRF identifies Se and Cd. In addition to cadmium red, iron oxide and titanium dioxide [35, 42] were identified by Raman, as reported in Fig. 8a. It is worth noting that the peaks at 1228, 1393 and 1609 cm−1 are due to luminescence of coprecipitated titanium white pigment [43] and this is a direct implication for dating the oeuvre, being not present in pigments produced by later preparations.
Spectral features of the ZC and in GA: a XRF identifying strontium chromate in a yellow shade of ZC and b in-situ Raman spectrum identifying vermillion in a red shade of ZC with the typical bands located at 251 cm−1, attributed to δ(S-Hg-S), and 340 cm−1, ascribed to ν(HgS). c XRF spectrum of a red paint in GA identifying cadmium red
In two pink tones XRF analysis shows the presence of As and Co; this finding suggests the presence of Co3(AsO4)2 (cobalt violet) [44]. In the orange tone of the central figure, XRF shows the presence of Fe and Mn, so the presence of a raw umber can be assumed. Unlike the previously works, Prussian blue is not present and other blue pigments are revealed. In the blue/green background in the upper part of the painting, XRF identifies Sn, Co and Cr; therefore, a cerulean blue (CoSnO3) and a chrome-based green pigment can be supposed. The use of cerulean blue is also evidenced by XRF results also in some parts of the blue figure on the right. Instead, in the lower part of the background characterized by a different blue/green shade, in the blue figure on the right and at the base of the spear, Raman analysis identifies a copper phthalocyanine. As illustrated in Fig. 8b, the main intense Raman absorption peak at 1526 cm−1 corresponds to the main macrocycle stretching vibration [45], while the signals at 680 and 747 cm−1 are due to the breathing and deformation vibration of the macrocycle [46] while the band at 1340 cm-1 is due to the C–C stretching [46, 47].
In some areas, titanium white was detected, probably used to mitigate the intense colour of copper phthalocyanine. Moreover, Raman analysis identifies the copper phthalocyanine also in some green paints.
The analyses also highlighted an elegant use of both traditional and synthetic pigments to obtain hues, shades, and vibrant colours. As a matter of fact, the Ernst’ palette evolves over the years (see Table 1), according to the availability of new pigments, but at the same time, the artist continues to use the traditional ones. For instance, in KI, Ernst uses both traditional pigments, e.g., carbon black and ochre, and synthetic ones such as Prussian blue, firstly synthetized in the eighteenth century [20, p. 315] and cadmium yellow, developed during the nineteenth century [48, p. 65] or chrome yellow orange, first used in the nineteenth century [35]. In ZC, Ernst uses vermillion, carbon black, iron oxide and lead white, with lithopone, strontium yellow and, again, Prussian blue. In GA vermillion is substituted by the cadmium red, introduced in the market at the beginning of the twentieth century [20, p. 76] and the blue tones are still obtained by Prussian blue together with a Co-based pigment, used to obtain particular shade. Corresponding to the palette used in the previous works, strontium and cadmium yellow [5] are still present for the yellow paints, even if a traditional raw umber is also used. This palette is close to the one used in EC, which is dated once year later. The latter one is enhanced by the use of a chrome-based pigment mixed with strontium chromate the green paints [5], and barium sulphate for the white paints. In AB the extensive use of Prussian blue is almost completely substituted by the recently commercialized phthalocyanine blue, synthetized in 1907 [49] but commercialized as monastral fast blue only in the 1930s [50]. AB palette is composed also by titanium white, introduced in Europe in the 1920s [51], azo pigments, developed from the second half of nineteenth century [52], and barium chromate, commercially diffused at the end of the nineteenth century [20, p. 42]. In AP, Ernst introduces the use of cerulean blue and cobalt arsenate, both introduced in the nineteenth century [20, p. 96], [53], to obtain some light blue and light violet tones, respectively. It is worthy to observe, studying the corpus of masterworks conserved at the PGC, the evolution in using Prussian blue: if in the earlier paintings Prussian blue was used to create blue tones, or mixed with yellow to obtain green hues, after the 1941 it was only used in shades, as in AB, or completely eliminated, as in AP. Similarly, Ernst seems to prefer, in both AB and AP, the use of the “new” titanium white in place of the consolidated white pigments. On the contrary, all the black areas were painted with the traditional carbon black, underlining the high Ernst‘s mastery in the experimentation with traditional and new materials.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Disclaimer:
This article is autogenerated using RSS feeds and has not been created or edited by OA JF.
Click here for Source link (https://www.springeropen.com/)