Keywords

1 1917 Paintings

Four paintings by Picasso from the Museu Picasso collection (Barcelona) painted in 1917 were examined: Seated ManFootnote 1 (Fig. 18.1a), Man with Fruit BowlFootnote 2 (Fig. 18.1b), Woman on an ArmchairFootnote 3 (Fig. 18.1c) and Blanquita SuárezFootnote 4 (Fig. 18.1d) [7]. Despite presenting similarities in materials, technique and concept and having been displayed together in Picasso’s family residence in Barcelona from 1917 to 1970, the paintings show very different degrees of cracking. The hypothesis is that differences in the composition of canvas and ground layers have contributed to the degree of cracking, from one painting with deep cracks throughout to another with no cracks. Fortunately, according to the principle of proofed fluctuations [11], it is unlikely that the fracture damage will increase in the present environment since the paintings have already been through stress relaxation and cracking caused by the less controlled climate of the Picasso residence. They have in other words become less vulnerable due to this initial response to former climate fluctuations.

Fig. 18.1
figure 1

(a, b, c, d). (a) Seated Man; (b) Man With A Fruit Bowl (c) Woman In An Armchair (d) Blanquita Suárez. Pablo Picasso, Barcelona, 1917, Oil on canvas

Full size image

All four paintings have thin paint layers on top of a thick ground and a cotton canvas. Seated Man is, however, more affected by cracks than the other paintings in the series. A deep crack network covering the whole painted surface was observed in the paint, regardless of colour. A less dense crack pattern was evident in the areas coincident with the stretcher bars, suggesting that protection of the back of the canvas was afforded by the stretcher [8]. Blanquita Suárez showed similar phenomena while within the boundaries of the stretcher bars, locally extensive, deep cracks can be observed. Man with Fruit Bowl has no significant cracks whereas Woman on an Armchair exhibited varied surface phenomena and the paint shows a few cracks in the areas that are not coincident with the stretcher. The film appears more compact than in the other three paintings studied. Seated Man and Woman on an Armchair each showed evidence of local water stains on the back of the canvas, but this cannot explain the differences in overall conditions.

2 Materials and Methods

Canvas analysis was carried out by taking close-up photos of the back of the original canvases, including a scale bar, and analysing them using the open-source software ImageJ. The weave and spinning geometry were measured on ten different threads, the width of the threads and degree of spinning were determined, and the cover factor of the fabric was calculated in both weave directions. Cover factor is the combined coverage of an area with threads [9]. Identification of fibres was performed with polarized light microscopy.

Cross-section samples from the edge of each painting were taken and embedded in Technovit 2000 LC in an EasySection®. The first part of the polishing was carried out wet. The last part of the polishing was done dry using Micromesh with grits up to 12,000.

Cross-section micrographs were taken with a Leica DM 4000 M microscope with Leica Application Suite 4.1 software. Incident light was used with the filter settings Dark Field (DF), Bright Field (BF), Differential Interference Contrast (DIC), and ultraviolet (UV) filters A and I3.

A JEOL JSM 6300 scanning electron microscope operating with a Link-Oxford-Isis energy-dispersive X-ray microanalysis system was used to observe surface and cross sections of the films. Data was collected with 20 kV accelerating voltage, 2.10–9 A beam current, and 15 mm working distance.

A selection of the collected samples was analyzed by micro-Fourier transform infrared spectroscopy (μ-FTIR) in attenuated total reflectance (ATR) and transmittance mode with a Thermo Fisher Scientific Nicolet iN10 MX Ultra Fast Motorized stage with MCT Detector. The spectra were acquired between 4000 and 600 cm−1, 128 scans, with a resolution of 4 cm−1.

Due to the complex and different nature of organic materials present in modern paint samples, two different analytical procedures were used:

  • Procedure 1 -Identification of lipidic, waxy and resinous materials

    • The samples were trans estherified using (trifluoromethylphenyl) trimethylammonium hydroxide, overnight reaction. The procedure and the details are described in several publications [2,3,4].

  • Procedure 2 -Identification of proteinaceous and polysaccharidic materials

    • The samples were analysed as N,O-acetyl methyl esters after hydrolysis with trifluoroacetic acid and derivatisation. GC-MS analysis was performed using an Agilent 6890NGC instrument with a capillary Carbowax Column, 30 m, 0.20 mm, 0.5 μm interfaced with an 5973NetworkMS. The temperature program was: 80–250 °C, 20 °C/min until 210, then 2 °C/min until 250 °C. The MS was run in Full Scan mode (m/z 40–500). Mannitol and Norleucine were used as internal standards for sugars and amino acids, respectively.

3 Results and Discussion

Analyses have revealed minor but important differences related to the specific use of some materials in both canvases and grounds.

3.1 Analysis of Canvas

The canvases used for all four paintings were identified as mercerized cotton. The weave geometry showed a significant difference between the canvases (Table 18.1). Threads showed similar thickness and degree of spinning, while the canvas in Seated Man was more tightly woven with cover factors of 71% and 85% in each weave direction, whereas the other paintings only had cover factors of around 50% in each weave direction. This signifies a tight weave in the Seated Man, which is indicative of canvases that are prone to shrinkage [1]. Moreover, it has been suggested that a hygroscopic layer containing animal glue combined with a tightly woven canvas is more prone to develop stress and possibly cracking than if the canvas is a more open weave [5].

Table 18.1 Study of support (weave geometry) of the four case studies
Full size table

3.2 Analysis of Ground Layers

Table 18.2 shows optical microscopy (OM) and SEM backscattered electron (BSE) cross-section images of samples taken from the edges where mainly the ground layers are present. All four paintings have two ground layers: the OM cross-sections show a dark brown layer beneath the double ground that is highly fluorescent in UV light and includes small particles. In all four paintings the upper ground appeared more compact with slight variations in tonality and elemental composition characterised using SEM-EDX as summarised in Table 18.2.

Table 18.2 Optical microscopy (OM), SEM backscattered electron (BSE) cross-section images and SEM-EDX results of samples taken from the edges. In MPB 110.007, part of the top surface of the upper layer of this stratigraphy shows the presence of a thin white layer of a denser nature and a slightly different fluorescence, as seen in the OM and SEM BSE images. The study of the surface might indicate that it is part of the whitish paint layer of the painting
Full size table

The lower ground layer of Seated Man contains lead (Pb) and/or sulfur (S) together with relatively lower amounts of calcium (Ca), barium (Ba), potassium (K) and chlorine (Cl). The upper ground layer presents relatively high amounts of Pb and/or S and relatively lower amounts of Ba, Ca, Si and strontium (Sr). The results suggest the presence of lead white as a major component in the upper layer of this painting. This was confirmed by the micro-FTIR results that identified basic lead carbonate in the white preparation layer together with calcium carbonate and barium sulphate.

Elemental analysis of the preparation ground layers from Woman on an Armchair found that the lower layer contains Pb/S and Ca with traces of magnesium aluminium, sodium, phosphorous, potassium and chlorine while the upper ground layer contained relatively high amounts of lead and calcium and magnesium with lower proportions of silicon and sodium. Micro-FTIR analysis of the lower brown layer identified both a lipidic and proteinaceous organic compounds and inorganic materials, such as CaCO3 and silicates. In the upper whitish layer, micro-FTIR analysis characterised basic lead carbonate and calcium carbonate mixed with a lipidic binder. Thanks to the typical IR absorptions at around 1510 and 1460 cm−1, the presence of lead carboxylates was also observed, that may have affected the mechanical properties of the ground layers [13].

Elemental analysis of the ground layers of Man with Fruit Bowl showed that the lower ground included Ca, S, and Mg and relatively small amounts of Si, Al, Na and Cl and an ochre upper ground layer composed of relatively high amounts of zinc (Zn), Ba and Ca, together with Pb and/or S, and relatively lower amounts of Si, Sr, K and Mg. Micro-FTIR has also confirmed the presence of calcium carbonate and barite. It might be concluded that the upper ground contains chalk and lithopone and/or zinc white (ZnO), zinc sulfide (ZnS) and barium sulfate. ZnO containing paint layers in oil media are known to be stronger but also brittle, having a tendency to delaminate rather than to crack [12]. Although no cracks were found in Man with Fruit Bowl, no delaminations were evident either.

Different elements were identified in the lower and upper ground layers of Blanquita Suárez: in the lower brownish coloured ground layer, lead, calcium and relatively lower amounts of Si, Al and Mg were found. Furthermore, in the orange upper ground layer, relatively high amounts of Pb, together with iron (Fe) and Ca, and relatively lower amounts of Mg, Al, Si and possibly Ti were identified. Micro-FTIR confirmed the presence of lead and calcium carbonates together with alumino-silicates. These silicates (together with traces of titanium) are characteristic of natural iron oxides responsible for the orange colour of the upper ground layer.

Based on GC-MS results, in all the case studies, the grounds contained a proteinaceous binder, which was identified as animal glue indicated by the amino acidic profile and the detection of hydroxyproline. This is the specific biomarker for animal glue, typically used as a binder in the execution of ground layers. This finding points to the presence of animal glue, particularly in Seated Man and Blanquita Suárez, which exhibited characteristic fluorescence under UV illumination and the typical IR absorption bands of a proteinaceous compound. Animal glue has been previously recognized as responsible for stress in ground and paint layers due to its high response to relative humidity [6, 10]. Furthermore, a lipidic fraction (containing both mono- and di-fatty acids) was detected in all the ground layers, possibly due to the presence of a drying oil. In Man with Fruit Bowl, the oil fraction was high and its presence was also corroborated by micro-FTIR where the binder of the ground was identified as a mixture of lipidic and proteinaceous compounds, characterised by the typical IR absorption at 1740–1460–1373-1090-720 cm−1 and 3390–1620 cm−1 respectively. Finally, and considering that the paintings have not been lined, the presence of polysaccharidic compounds (glucose and galactose derivatives) identified in the grounds of all four paintings was possibly from starch that could have been mixed into the grounds or from a pre-treatment of the canvas as part of the manufacture proceedings.

4 Conclusion

This paper presents the results of the first stage of a project where the canvas and ground layers of four paintings by Picasso were analysed. The visual observations associated with diverse kinds of mechanical deterioration in the paintings suggests that the material differences in the works created by Picasso in Barcelona within the same year have caused various degrees of vulnerability to environmental and chemical deterioration.

Material characterization has revealed some key differences between the grounds of the paintings. Three of the four paintings included an upper ground layer composed mainly of lead white and other materials that result in slight differences in the tonality of the layer. Man with a Fruit Bowl is the only painting in this group that contains a Zn compound in the upper ground layer. Lead carboxylates were identified in Woman on an Armchair. All four paintings had grounds that contain animal glue, a possible drying oil, and polysaccharides as detected by GC-MS. In particular cases, some compounds are in relatively higher amounts, such as the polysaccharide compounds in Blanquita Suárez that could be related to the sizing recipes.

It is proposed that material differences in the canvas and grounds can be related to the degree of cracking observed in the paintings. For example, Seated Man exhibited the most extensive cracking, which could be explained by a high proportion of animal glue in or below the ground that responded rapidly to fluctuations in moisture while the paint was stiffer and slower to respond. Animal glue is known to shrink both at low and very high relative humidity and furthermore the tightly woven canvas support used for this painting rendered it prone to shrinkage at high humidity. Woman on an Armchair has large, shallow cracks that could be explained by the combination of animal glue and possibly the carboxylates that can make a film stiffer and somewhat brittle. Blanquita Suárez also has locally extensive cracks, again probably due to the relatively high amount of animal glue and impact from handling. Man with a Fruit Bowl has no cracks. The zinc compound detected in the upper ground layer may indicate a material that has a tendency to delaminate rather than to crack, but no evidence was found of delaminations. Further research to identify the Zn-containing compound will help to better understand the condition of this painting.

In conclusion, the mechanisms behind the cracking observed in each painting is complex and warrants further analysis. Future research will include chemical characterization of the paint layers as the combination of specific pigments and binders is known to result in paint layers with different mechanical properties, moisture sensitivities, and tendencies for cracking. The influence of fillers will also be explored.

It is unlikely that these paintings will suffer in the controlled environment of the Picasso Museum. However, if the paintings are moved to other locations or if the works are treated by conservators, the resulting response to the environment should be monitored. This would be especially important for the paintings that contain higher amounts of animal glue and the one that is painted on closely woven canvas.