Navigation – Plan du site

AccueilNuméros39Paint Composition and Delaminatio...

Paint Composition and Delamination Concerns in Ole Schwalbe’s Portrait of a Painting III, 1962

Louise Decq, Steven Saverwyns, Filiz Kuvvetli et Elizabeth Baadsgaard

Résumés

Cet article examine les défis liés à la préservation du Portrait d’une Peinture III (1962) d'Ole Schwalbe. Cette œuvre, représentative du constructivisme abstrait minimaliste, présente une délamination sévère de la peinture. Des techniques d'analyse telles que le SEM-EDX, le THM-GC/MS, la spectroscopie micro-Raman et la FTIR ont été utilisées pour analyser la composition de la peinture et identifier les causes de son instabilité. Les résultats suggèrent que des processus de durcissement incomplets et une dégradation oxydative dans la couche épaisse de peinture noire ont entraîné l'accumulation de savons de zinc dans la couche de préparation sous-jacente, composée de blanc de zinc. La réduction du volume des couches supérieures, combinée au gonflement des couches inférieures, a contribué à la dégradation observée. Cette recherche apporte des perspectives nouvelles sur la conservation des œuvres abstraites d'après-guerre.

Haut de page

Texte intégral

Introduction

1Constructivism, rooted in the Russian revolutionary period and manifesting in Western art through movements like De Stijl, for instance, in the 1920s, played a pivotal role in the evolution of twentieth-century art. This movement marked a departure from figurative art, embracing a new visual language that invited personal experience and emotional engagement from the viewer. Artists such as Ole Schwalbe and Gilbert Decock embraced minimalist abstract constructivism, focusing on geometric forms and limited colour palettes.

2The Belgian painter Gilbert Decock (Knokke, 19282007) was a pioneer in the post-war growth of abstract constructivism in Belgium. With a formal language of geometric forms, mainly circles and squares, and a primarily black-and-white palette, Decock's oeuvre contrasted with the chaotic outside world. It showed striking parallels in time and formal language with the work of the Dane Ole Schwalbe (Brønshøj, 19291990), emphasizing an austere arrangement of geometric forms and a minimalist colour scheme. Schwalbe limited his palette to black and white, sometimes supplemented by a single contrasting colour. These parallels illustrate the international sphere of influence within which these avant-garde artists worked. Schwalbe's Portrait of a Painting III exemplifies the production of an international output by artists with common sources of inspirations and innovative practices.

  • 1 Helou-de La Grandière, Thoury and Simonot 2023; Learner et al. 2007.
  • 2 Helou-de La Grandière, Thoury and Simonot 2023.

3In works of minimalist abstract constructivism, the surface aspects of the paint layers are paramount. Where the simplest black-and-white areas and lines form the composition, the material aspects of paint are essential. Subtle nuances in gloss, rheology, thickness and texture can play a leading role in such paintings. These properties can be influenced by the thickness of the layers, the layer structure and the composition of the paint layers, controlled by adding extra binder, filler, pigment, dryers, texturing agents, etc. A huge supply of paints and artist materials became available in the post-war industrial context to inspire these innovative, searching artists.1 However, the long-term effects of these sometimes experimental styles were difficult to foresee and can be challenging for conservators today.2 The search for solutions to material-related conservation problems in post-war art is a relatively new field within restoration, where analysis plays a crucial role. This study aims to examine the paint layers and identify the factors contributing to their instability, providing essential insights for the preservation of similar artworks.

About the artist: Ole Schwalbe (19291990)

  • 3 ‘Ole Schwalbe’, in Trap Danmark. Lexdk; Købke Sutton and Harding 2023.

4Ole Schwalbe was a Danish painter, graphic artist and sculptor who made his debut at the Artists' Autumn Exhibition in 1945 at the age of 16.3 He later joined the experimental Danish artist group Linien. A self-taught painter, Schwalbe is known for his Concrete Art, which involves creating images that do not depict specific subjects but are intended to be appreciated for their intrinsic qualities. Initially, Schwalbe's works were highly ‘geometric’, featuring few or no colours, incorporating elements such as squares, lines and ovals. The title Portrait of a Painting III ('Portræt af et maleri III') suggests a portrait of a painting, indicating Schwalbe's interest in exploring the essence of painting itself rather than representing external subjects.

The painting Portrait of a Painting III

5The painting was executed on a canvas support, stretched on a keyed stretcher. The composition consists solely of black paint areas with varying hues and consistencies (fig. 1), accented with delicate white lines. The signature on the back reads ‘Schwalbe 1962’.

Fig. 1

Fig. 1

Ole Schwalbe, Portrait of a Painting III, 1962, alkyd paint and oil paint on canvas, 100.5 × 80.5 cm (Denmark, ARoS Art Museum of Aarhus). Photo taken under normal symmetrical visible lighting in 2023, before the conservation treatment.

© Silkeborg, Kunstkonserveringen.

  • 4 Portræt af et maleri III, in Kunstindeks Danmark & Weilbachs Kunstnerleksikon, accessed 2 February (...)

6The artist applied cumulative layers of paint to specific areas, creating varied thicknesses and textures, thereby introducing subtle variations. Surrounding the central black composition intersected by white lines, the canvas texture and white underground are visible through a translucent layer of black paint in yet another matt hue. This work exemplifies Schwalbe's commitment to Concrete Art, evident in its strong black-and-white contrast and emphasis on geometric forms. The use of a limited colour palette and geometric forms employed in the painting align with the principles of Concrete Art.4

Current condition of the painting

7In 2023, while still on display in a gallery of the museum that owns the painting, conservators observed extensive flaking and delamination in the paint layer of Portrait of a Painting III. The painting was in a critical state of preservation, with two square areas particularly affected (see fig. 2).

8The most severe flaking has occurred where the paint layers are applied more thickly, creating a textured surface with small, even brushstrokes. These damaged areas feature small and large cracks that could potentially lead to further flaking and paint loss. The delaminating paint flakes are black, with some traces of the white underlayer still attached. These layers are relatively soft compared to the rest of the painting and can easily be repositioned with a light touch. The flakes tend to curl up, unveiling in the cracks a white layer described as thick and possibly structured.

Fig. 2

Fig. 2

Sections indicating different conservation conditions, as described. The red and yellow areas are affected with delamination. The other areas are well conserved. The stratigraphy and composition of the area indicated in blue were not examined. Normal incident lighting. Photo taken in 2023.

© Silkeborg, Kunstkonserveringen.

Fig. 3

Fig. 3

Details, under normal and UV lighting, preservation condition in 2023 (fig. 3a and b); the green square indicates the detail shown in fig. 3c˗f. The lower images illustrate the characteristics of the delamination in the same area, showing the paint layer’s condition in 2001, 2010 and 2023 (fig. 3c, d,-e). Fig. 3f shows the same detail under UV lighting, preservation condition in 2023.

© Silkeborg, Kunstkonserveringen.

9The square marked in red (fig. 2) is the most severely damaged, with extensive flaking (fig. 3a and b). This area is characterised by a more intense, more saturated black colour. It shows a slightly whitish appearance and a glossier surface compared to the other sections. The vertical and horizontal white lines crossing this area are intact. The square marked in yellow (fig. 2) also shows significant flaking, with clear delamination in the upper right corner. This area appears duller and has a slightly different intensity of black colour. In contrast, the section outlined in blue remains undamaged. This area appears smoother, lacking the even brushstrokes found in the section marked in red, and the black colour is lighter in tone. The borders of the painting are covered with a thin layer of black paint, which appears dark grey due to its translucency. There are no visible signs of damage in these areas.

Conservation history of the painting

  • 5 Kunstkonserveringen is Denmark’s Art Conservation Center, founded in 1978 by an association of Dani (...)

10The condition of the painting has been a concern for decades. It has undergone several conservation treatments at regular intervals, all conducted at Kunstkonserveringen in Silkeborg, Denmark.5 The first treatment took place in 2001, the second in 2010, and the most recent in 2023 (fig. 3). These conservation efforts were primarily prompted by the recurrent delamination of paint layers, particularly in the two thickly applied black areas on the left side, despite the painting being kept under stable museum preservation conditions.

Fig. 4

Fig. 4

Condition in 2001. At this point, the painting was in a fair condition, with minor flaking of the paint layer.

© Silkeborg, Kunstkonserveringen.

  • 6 Filed as # 576 in a Microsoft Access Database at Kunstkonserveringen with the name of the artist, t (...)

11In 2001 the initial conservation treatment addressed damages in the black paint layers, in the same areas as mentioned above (fig. 4). The consolidation treatment involved Beva 371, a 40% solution based on Gustav Berger’s original formula, purchased from a conservation materials supplier and diluted with petroleum ether at a 1:3 ratio. The adhesive was applied locally beneath the flaked paint using a fine brush. After the solvent evaporated, the lifted paint flakes were carefully repositioned using a hot spatula under controlled heating. Notably, the conservator emphasised that the painting’s condition should be regularly monitored due to the apparent tension in the paint layers, evidenced by small and large cracks. Additionally, as documented in the conservation report in Kunstkonserveringen’s archives,6 the conservators cautioned about the potential for repeated delamination to occur.

Fig. 5

Fig. 5

Preservation condition in 2010 under normal symmetric visible lighting, before treatment. Nine years after the first conservation treatment, the state of preservation had worsened. The affected area was now larger, though still in the ‘red’ selection. This worsened condition led to a comprehensive structural conservation treatment that included a total impregnation with Beva 371 from the rear.

© Silkeborg, Kunstkonserveringen.

12In 2010, when the painting was brought back to Kunstkonserveringen for treatment, its condition had significantly worsened (fig. 5). Additional flaking and open cracks had emerged in the areas addressed during previous conservation treatment. The observations raised questions about the long-term viability of the artist's chosen materials, underscoring the challenges faced by conservators in preserving the artwork.

13An extensive conservation treatment was again urgently required. The comprehensive consolidation of the painting’s layered structure was deemed necessary to prevent further delamination, as previous localised treatments had proven insufficient. Given the extent of the delamination, it was thought safer to apply low pressure evenly across the entire structure rather than perform multiple localised interventions from the front. Additionally, the conservators identified a risk of creating surface unevenness due to the malleable nature of the paint layers when using tools for localised treatments.

14Various consolidants were reviewed and tested for their compatibility with the properties of the paint layer. Due to the paint layer’s high sensitivity to polar solvents, the use of moisture and organic polar solvents was avoided. Moreover, the selected consolidant required strong adhesive properties to counteract the tensions in the affected areas, which limited the choice of suitable adhesives. The painting was removed from its stretcher and was impregnated from the back with a 15% Beva 371 solution dissolved in petroleum ether (boiling point 100-140°C). Hot Beva solution was applied to the entire back with a brush, and the consolidation took place on a low-pressure table at 67-70°C and 50 mbar pressure. Subsequently, the painting was remounted to its stretcher. A year of observation in the workshop followed to assess the treatment’s impact.

15Given the persistent flaking, the decision was made to reinforce the painting's structure by adding a support canvas to the back. The painting was again dismantled, and the back was impregnated with a 20% Beva 371 solution, a combination of Elvax 150 resin (ethylene-polypropylene copolymer), Laropal K-80 (cyclohexanone keton resin), A-C 400 (copolymer ethylene vinyl acetate-acrylate), Cellolyn 21 (phtalate ester of hydroabietyl acid) and paraffinic hydrocarbons. Then, the secondary support, a Cubenfiber, was applied to the rear of the painting with Beva 371 film. This support is a synthetic fabric known for its non-flexible nature, absence of fibre direction and high strength compared to the original canvas. The restoration was completed satisfactorily, but again, restorers emphasised that the painting should undergo regular inspections to monitor and address any new damages.

16In 2023 recurring damage in the same areas brought the museum to the conservation studio again. Before initiating new treatments, the decision was made to thoroughly analyse the artist's materials and the overall structure of the painting. The objective was to gain a better understanding of the materials involved and the potential reasons for the recurring damages, in the hope of formulating a sustainable preservation plan for the painting based on these insights. This method of in-depth study to enable an informed and effective decision on persistent issues illustrates the commitment of the restoration studio to aim for better long-term preservation of the artworks.

Materials and methodology

17To investigate the cause of paint delamination in Portrait of a Painting III, samples were taken from three locations, one in each affected area, and one from the non-affected border area (fig. 6). Stratigraphic analysis was performed using light microscopy and scanning electron microscopy (SEM) coupled with energy-dispersive X-ray detection (EDX), and complemented with dispersive micro-Raman spectroscopy (MRS). In parallel, binder analysis was performed with pyrolysis gas chromatography mass spectrometry (Py-GC/MS), with and without thermally assisted hydrolysis and methylation with tetramethyl ammonium hydroxide (TMAH). This analysis was complemented with Fourier-transform infrared spectroscopy (FTIR) on one sample. The technical set-up is described at the end of the article.

Fig. 6

Fig. 6

Position of sample locations, 2023.

© Silkeborg, Kunstkonserveringen.

Results

Stratigraphy

Fig. 7

Fig. 7

Three cross sections. Left column, a sample taken at the non-affected black border area (resp. polarised light, UV, SEM-EDX mapping); middle column, affected upper black square (area indicated in red in fig. 2; resp. polarised light, UV, SEM-EDX mapping); right column, affected lower black square (area indicated in yellow in fig. 2; resp. polarised light, UV, SEM-EDX mapping). The insets in the upper row show the interface sides of the samples, 200× original magnification. Image 100002010000007A0000001A9289887B74CA8B49.pngP for preparation layers, B for black layers, T for transparent layers. T in the first cross section is thought to relate to T2 in the other sections, due to the common presence of aluminium (aluminium hydrate?).

© Kunstkonserveringen, Silkeborg and Brussels, KIK-IRPA (sections K.K. OS-1a, OS-2, OS5; KIK-IRPA P252.040, P252.041, P252.041).

Fig. 8

Fig. 8

Stratigraphic scheme in the non-affected border, and with the affected upper and lower squares indicated in fig. 2. In vivid blue, the layers coloured with bone black; in pale blue, the layers coloured with carbon black; in violet, a pigment identified as aniline black; in orange, the inclusions of clay.

18The analyses reveal the presence of two white preparation layers on the canvas (fig. 7 and 8).

19The first (P1), believed to be an industrial coating, is rich in aluminium (maybe aluminium hydrate), with elongated, angular potassium-rich inclusions and some quartz crystals, measuring approximately 100 µm in thickness. The second preparation layer (P2), thought to be the artist’s preparation, contains zinc white, barium sulphate and possibly beeswax (possibly a Beva 371 signal, only detected in the lower sample), with a variable thickness around 10-70 µm. Neither preparation layer contains chalk or gypsum, and their binding mediums remain undetermined, though oil is traditionally expected. Black top layers of the composition (indicated as B, B1-B4) are applied on this zinc white preparation layer, both in affected and stable areas. The painting is covered with an aluminium rich transparent layer, possibly aluminium hydrate, whose originality is unknown (T2).

20The stratigraphy in the affected areas was studied using delaminated paint material. This implies that only the parts above the delamination area were investigated. Traces of the second preparation layer based on zinc white with barium sulphate still adhere to those flakes (P2). The paint is disconnecting at the surface of the white preparation layer (P2), leaving small white spots up to 70 µm thick on the black paint flakes.

  • 7 Karakaş et al. 2011; Learner et al. 2007; Buyondo, Kasedde and Kirabira 2022.
  • 8 La Nasa et al. 2019.

21In the affected areas, a succession of several black but differently composed layers illustrate the artist’s subtle play with different aspects of black paint. On top of the white layer P2, three or four black layers were applied. The first black layer (fig. 7, B1; ca. 30 µm in the upper square, 60-145 µm in the lower square) is rich in clay minerals. Clay, such as kaolin, has traditionally been added as a filler or extender to paint to adjust its rheology and brilliance, and is still in use today.7 In this case the artist chose a dark-coloured clay, maybe so as not to influence the deep-black colour of his paint. This first layer was probably pigmented with carbon black, as suggested by two broad carbon signals in the MRS measurement of this layer (at ca 1300 and 1600 cm-1). The thicker second black layer (B2, 200 µm thick in the upper square, ca 300 µm in the lower square), is pigmented with bone black, again with rough clay inclusions measuring up to 120 µm in thickness. The heterogeneity of the mixture and the use of identical clay in different layers may indicate that the mixture was handmade by the artist. More scarcely dispersed, aluminium containing fluorescent inclusions are seen (probably aluminium hydrate8). The third black layer B3 is a thinner, finer-grained layer of almost identical composition but with more binding medium (10-15 µm thick). Next, an identical black layer B4 follows (10-15 µm). However, a fluorescent thin layer (T1) resembling an oil medium appears between B3 and B4 (up to 4 µm in the upper square, 10 µm in the lower). As will be further discussed, two options remain open: the deliberate application of an oil or a spontaneous accumulation of free fatty acids. An aluminium-containing varnish (T2) is spread over the complete surface. Locally, two layers have been applied: an additional, very thin, semi-transparent layer with aluminosilicate material (a filler?, 0-7 µm), and an additional thin aniline black layer (1-5 µm). Those are both considered non-original.

Analysis of the non-embedded samples

22The sample material for binder analysis was limited to two flakes of delaminated paint. As they are likely composed of several layers, including several black layers and minor parts of the second preparation layer, it is not possible to determine the binding medium of the separate layers.

Fig. 9

Fig. 9

THM-GC/MS chromatogram of black paint flake sampled in the upper black square. ME: methyl ester; FA: fatty acid; DME: dimethyl ester. Partially oxygenated fatty acids indicate ongoing degradation. The position of three alkyl phenyl alkanoates (APA), markers for heat-bodied oil, is indicated.

© Brussels, KIK-IRPA, file 2023.15147.

Fig. 10

Fig. 10

FTIR spectrum of the sample taken in the affected zone (upper square) in Portrait of a Painting III. The spectrum shows a composite spectrum of bone black, binder and some indications of clay minerals (mainly 3570 cm-1 and 663 cm-1). The peak around 1456 could be associated with regular zinc stearate (Zn(stearate)2). Around 1585 a broad absorption band could indicate a zinc-sodium soap of mixed fatty acid composition (ZnNa2(palmitate/stearate)4).

© Brussels, KIK-IRPA, file 2023.15147

  • 9 Schilling et al. 2016.
  • 10 Bronken 2014.
  • 11 Standeven 2011.

23The chromatographic detection (fig. 9) of pentaerythritol and phthalates, and a spectrum of generous quantities of mono- and di-carboxylic fatty acids in the chromatographic binder analysis of both samples, combined with the FTIR spectrum (fig. 10) with a carbonyl band at 1732, indicates the presence of an alkyd medium and a drying oil. An alkyd resin is formed by the esterification of a polybasic acid (e.g. phthalic acid) with a polyhydric alcohol (e.g. glycerol, pentaerythritol). To form oil-modified alkyds with suitable drying properties (long oil alkyds), fatty acids, particularly drying and semi-drying oil acids, are added. Being a condensation product of a polyol and polycarboxylic acid, with an inclusion of at least 55% of drying oil moieties, one could consider a long oil alkyd resin as a synthetic oil medium. The analyses indeed show the presence of a heated drying oil. The relatively weak signal of phthalates, mostly detected as dimethyl phthalate, and pentaerythritol compared with the strong signal for drying oil and glycerol, indicates the presence of a drying oil paint medium. The presence of alkylphenylalkanoates (APA) suggests a prepolymerised stand oil was used,9 although the relation between the heat bodied oil and the alkyd components is not determined. Comparison with similar cases suggests that such prepolymerised oils may have been added to modify the rheology of paints.10 The presence of abietanes with 2,4,5,7-tetramethylphenanthrene demonstrates an addition of heated colophony, often added to house paint.11 As the presence of the second white preparation layer was limited in these samples, the alkyd and oil medium are likely to originate both from black or transparent upper layers. Taking into account the possibility of added metal soaps, mixed oils, successive layers and possible migration and evaporation of fatty acid fractions, the use of fatty acid ratios to determine the nature of the oil(s) is considered unrealistic.

24In the sample taken in the upper square, limited indications for a polyurethane polymer (isophorone diisocyanate IPDI), paraffin wax and a ketone resin were additionally detected. These may be related to previous conservation treatments.

  • 12 Zviagina, Drits and Dorzhieva 2020.
  • 13 Eastaugh et al. 2004, p. 180–181.

25Only one sample was subjected to FTIR analysis (fig. 2, red area). This measurement further explains the composition of the black layers (fig. 10). The FTIR spectrum shows, apart from the highly evident presence of bone black and oily/alkyd binding medium, more discrete indications of the presence of the expected clay minerals. While the Si-O bending region (~600–400 cm−1) is not visible with the applied conditions and the Si-O stretching regions (~1200–700 cm−1) is masked by the dominant bone black signal, the OH-stretching region (~3700-3500 cm-1) does show a signal at 3570 cm-1 that must be interpreted as an additional indication for the presence of a clay mineral. Although the information is too weak and incomplete to determine the exact minerals, comparing the FTIR spectrum, the SEM-EDX measurement and the published reference data, an iron-rich dioctahedral mica clay such as glauconite ((K,Na)(Fe,Al,Mg)2 (Si,Al)4O10(OH)2) (or celadonite) would be most appropriate.12 Both glauconite and celadonite are known to be found in artist materials, especially in green earths.13

Discussion

  • 14 Helou-de La Grandière 2008; Helou-de La Grandière 2017; Bronken 2014; Helou-de La Grandière 2017; H (...)
  • 15 Bronken 2023.

26The recurring damage observed in the painting under discussion shows similarities to the well-documented cases of paintings by contemporary painters Pierre Soulages (19192022) and Jean-Paul Riopelle (19232002).14 After several decades of apparent stability, the paint in certain areas begins to soften and increase in gloss. The surface often starts to deform and to lose its binding medium. This process can lead to cracking and delamination between the preparatory and paint layers. Similar cases of degradation have been reported for other paintings of the Cobra movement.15 In the following discussion, the results obtained on Portrait of a Painting III by Ole Schwalbe will regularly be compared with the findings on those other paintings.

27Stratigraphic research indicates that the flaking occurs between the second, zinc-containing preparation layer and the first black layer, with the rupture removing part of the second preparation layer. The fragments of lifted preparation layer are usually a few micrometres thin, but at places they can measure up to 75 µm, considered potentially the full thickness of the zinc-containing preparation layer. Various factors could contribute to the painting's instability, necessitating further investigation. This discussion explores the possibility of saponification issues in the second preparation layer, binder-curing and binder-pigment interactions in the flaking black upper layers, and the role of thickeners in those black upper layers.

  • 16 Casadio et al. 2019; Bronken, Boon and Steindal 2021.
  • 17 La Nasa et al. 2018.
  • 18 Casadio et al. 2019.
  • 19 Bronken 2014.

28The presence and migration of metal soaps, typically lead, zinc or aluminium soaps, within the oil paint is a recurring explanation for similar delamination issues in paintings from this period.16 Soaps can be added to the medium, most commonly Zn and Al stearates, or are formed as a product of a reaction between carboxylic acid moieties released by the hydrolysis of triglycerides or the oxidative cleavage of unsaturated fatty acids, and the metal ions contained in certain pigments.17 Depending on circumstances, metal soaps can coagulate, protrude or migrate to the surface or to an interface, causing paint layers to potentially soften, delaminate or change in brilliance, transparency and colour. This phenomenon often starts after years of (visual) stability.18 Reported characteristics include exceptionally high azelate content19 (azelate palmitate ratio), metal ion accumulation at the interface, and increased UV fluorescence. An estimation of the presence of these carboxylates can therefore provide useful information.

  • 20 Bronken, Boon and Steindal 2021.

29Since the organic composition of the zinc-containing (second) preparation layer was not separately analysed in Portrait of a Painting III, conclusions about azelate distribution or other soap indicators in this layer are lacking. In the non-affected areas, the zinc-containing (second) underlayer shows an even, vivid greenish fluorescence, without visible increase at the interface. Unfortunately, an integral image of the preparation layers in the affected areas is not available; the adhered remnants of the zinc-rich preparation layer seem uniform, maybe slightly transparent. However, the very similar case of Peinture (1964) by Pierre Soulages shows that the accumulation of soaps in the underlayer (or at the interface) must be regarded as a potentially significant contributor to the painting's instability.20

  • 21 Graczyk et al. 2019; Bronken and Boon 2014; Bronken, Boon and Steindal 2021.
  • 22 Bronken, Boon and Steindal 2021; Bronken 2014; La Nasa et al. 2020; Helou-de La Grandière 2019.
  • 23 Fuster-López et al. 2016; Burnstock et al. 2017; Banti et al. 2018; La Nasa et al. 2020.

30In the studied Schwalbe painting, the borders are well-preserved, while two central, thick squares are specifically affected. The uneven degradation across the surface of the painting suggests a potential relationship with the thickness of the paint layer, as thicker paint layers may contribute to the incomplete curing of the oil paint.21 Furthermore, the repeated observation in similar works – that black-pigmented areas, often using bone black, are significantly more prone to delamination22 – raises the important issue of pigment interference (or general paint composition) with the oil curing process.23

  • 24 Similarly, long oil alkyd resins highly depend for their qualitative polymerisation on the fatty ac (...)
  • 25 Bonaduce et al. 2019.

31Linseed oil is a mixture of triglycerides, mainly containing (mono-, di-,tri-)unsaturated C18 fatty acid chains, and to a lesser extent, saturated C16 and C18 fatty acid chains. The double bonds in the poly-unsaturated fatty acids enable polymerisation.24 During oil drying, double bonds form peroxide radicals (ROO°) under the influence of light and air. Through a chain reaction, initial oligomers form. To achieve complete curing, the pathway favouring polymerisation must be followed, consuming virtually all double bonds. This process is optimal under light irradiation, with dryers, and in the presence of sufficient oxygen. In less favourable polymerisation conditions, slower oxidative degradation occurs. Remaining double bonds may undergo oxidative degradation.25 Less polymerised oil paints can, over time, and especially in combination with hydrolysis of the triglyceride bonds, cause paints to soften and even liquify, showing glossy and tacky surfaces, losing binding medium and dripping.

  • 26 Ferreira, van der Horst and Boon 2005; Bronken and Boon 2014.
  • 27 Bronken and Boon 2014.

32The obscuration due to black pigmentation, and the thick layer thickness in the Schwalbe painting examined here, are in themselves not optimal factors for a complete oil polymerisation. Additionally, it is suspected that bone black (and potentially coal black) acts as a strong antioxidant within the paint layer, limiting the formation of fatty acid peroxide radicals and consequently slowing down the drying process.26 The lack of stabilising ions such as lead, zinc, copper and even calcium to anchor the free polar degradation products in the black layers further leads to high chances of softening of the paint.27 Some signals indicate indeed that a softening degradation process is ongoing in the affected areas of the Ole Schwalbe painting under research. This type of degradation explains the glossy, soft texture with white haze of the lifting layers, their malleability, and the increased vulnerability of thicker paint layers.

  • 28 Schilling, Carson and Khanjian 1999; Bronken and Boon 2014.

33Chromatographic analysis was performed on the full thickness of two black paint flakes. The obtained fatty acid profiles are thus an integration over different areas, potentially evening out different phenomena at the interfaces. In paints from this period, fatty acid profiles must be interpreted with care: the initial fatty acid profile can vary significantly due to the addition of (semi-drying) oils and stearate (or palmitate) salts. The low palmitate-to-stearate ratio in the lower square sample (P/S 0.9) could suggest added stearates in the paint, although soap mobility and fatty acid evaporation may also play a role.28 The unknown starting situation limits the interpretation of the values. The quantities of azelate and other dicarboxylic acids appear as expected for a dried oil (Az/P 1.4 and 1.6), although they provide little conclusive evidence due to the sample's complex composition. Present alkylphenyl alkanoates (APA) indicate the oil has been heated, although suberate/azelate (0.3) is not substantially raised. The presence of pentaerythritol and phthalates indicates the use of an alkyd paint, presumably rich in fatty acids (long oil). The relationship between the present oil(s), heated colophony and alkyd are unknown.

  • 29 Bonaduce et al. 2018.
  • 30 Languri 2004, p. 145.

34Nevertheless, finding relatively large amounts of peroxide fatty acids in the flaking paint (ROOH oxirane-octanoic acid 3-octyl ME trans + cis), which are formed during a critical intermediate stage in the early phases of drying, indicates that polymerisation was at least partially inhibited.29 Another indication of incomplete or regressed curing of the oil is the residual amount of oleic acid (C18:1) in the sample. During the drying of oil, polyunsaturated fatty acids (C18:3 and C18:2) are consumed first; monounsaturated C18 fatty acids react approximately 1000 times more slowly than polyunsaturated C18 fatty acids.30 The remaining amount of oleic acid is remarkably high (C18:1 / C18:0 = 0.26 for the lower square sample and 0.73 for the upper square sample) after a sixty-year drying period.

  • 31 La Nasa et al. 2019.

35When the oxidative degradation proceeds, the liquid fraction finds a way out from the still more apolar polymeric matrix. It may be absorbed by the canvas support, it can form a tacky and medium-rich glossy surface, or start running as drippings on the paint surface. When absorption, evaporation or extrusion are hindered, degradation products can also accumulate. The fluorescent layer initially interpreted as a deliberately applied thin organic layer on the cross sections (T1 in fig. 7 and 8) could equally be interpreted as an accumulation of polar degradation products under the better polymerised surface film (B4).31

  • 32 Rogala et al. 2010; Bronken, Boon and Steindal 2021.
  • 33 Hermans et al. 2014.
  • 34 Bronken, Boon and Steindal 2021.

36However, in the case of the observed delamination phenomenon, an important part of the degradation products seems to migrate to the zinc-rich preparation layer P2. The free mono- and dicarboxylic acids generated by oxidative degradation in the upper black layers are unable to form metal soaps in the upper layers due to the absence of ionisable cations. However, as soon as they can reach the zinc-rich preparation layer, they readily react with the zinc oxides to form metal soaps. The degrading donor layer actually fuels extensive saponification of the metal oxides in the well-polymerised zinc white preparation layer.32 Although the FTIR measurement was carried out on a mainly black sample, a small and broad band around 1580 in the FTIR spectrum (fig. 10) may indicate the presence of some soap in the sample.33 With a confirmed presence of Na, the band is likely to be attributed to a zinc sodium soap of mixed fatty acid composition (ZnNa2(palmitate/stearate)4). A more pronounced absorption at 1456 could be associated with ‘regular’ zinc stearate (Zn(stearate)2).33 The stratigraphic distribution of the soap accumulation can only be hypothesised. In this case, the delamination tears off many, sometimes rather thick fragments of the probably swollen preparation layer. It seems likely that the zinc-rich preparation layer is, to some extent, affected in depth, and the soaps are not just accumulated at the interface between black and white. Although involving the formation of lead soaps, a phenomenon was observed in a similar case that could be interpreted as a network of micro-cracks filled with (saponified) binder in the upper region of the preparation layer.34

  • 35 Helou-de La Grandière 2019; Bronken, Boon and Steindal 2021.

37In this way, the preparation layer P2 swells, attracting additional mono- and dicarboxylic fatty acids, while the upper black layers lose volume and shrink. The tension builds up, until the surface breaks. The upper layer tends to retract, lifting up the edges of the impasto paint, curling back the soft, waxy black surface without breaking. Meanwhile, the swollen white paint underneath causes the impasto layers to rise like a bubble raised from underneath.35

  • 36 La Nasa et al. 2019.
  • 37 La Nasa et al. 2019.
  • 38 Modugno et al. 2019

38The impasto black paint in the affected areas was thickened with dark clay inclusions. Initially, in the fresh paint, these rather voluminous inclusions were properly encapsulated in an apolar oil medium. With ongoing degradation, however, the clay inclusions may turn into an extra vulnerability. Oxidative degradation increases water sensitivity due to the generation of rather polar exudates, but also the remaining polymer fraction increases in polarity.36 It is reasonable to suggest that water dampness may penetrate and diffuse more easily into the less cross-linked and increasingly polar paint film.37 When reaching the hygroscopic clay inclusions, local swelling and new tensions are likely to build up, further disrupting structural integrity. A conservation environment with stable and rather low relative humidity will also decrease the oxidative degradation processes in the oil.38

Conclusion

39Analyses shed new light on the stability issues in Ole Schwalbe’s Portrait of a Painting III. The flaked paint particles contain a succession of different black layers in carbon black and bone black on a (second) preparatory layer of zinc white and barium sulphate. The incomplete polymerisation in the black paint layers of the painting, combined with the availability of Zn in the underlayer, are key factors in the paint instability observed.

40These findings align with previous research on the conservation challenges of post-war abstract art, where a similar build-up often leads to unforeseen long-term issues like recurring delamination. The results underscore the need for tailored conservation strategies, including regular monitoring and specific treatment protocols to address material-specific vulnerabilities. The clay added to the black paint risks further increasing the water sensitivity when surrounded by a more polar and fragile matrix.

Haut de page

Bibliographie

Banti et al. 2018 D. Banti, J. La Nasa, A. Lluveras Tenorio, F. Modugno, K.J. van den Berg, J. Lee, B. Ormsby, A. Burnstock and I. Bonaduce, A Molecular Study of Modern Oil Paintings: Investigating the Role of Dicarboxylic Acids in the Water Sensitivity of Modern Oil Paints, in RSC Advances, 8, 11 (2018), 6001–6012 (https://doi.org/10.1039/C7RA13364B).

Bonaduce et al. 2019 I. Bonaduce, C. Duce, A. Lluveras-Tenorio, J. Lee, B. Ormsby, A. Burnstock and K.J. van den Berg, Conservation Issues of Modern Oil Paintings: A Molecular Model on Paint Curing, in Accounts of Chemical Research, 52, 12 (17 December 2019), 3397–3406 (https://doi.org/10.1021/acs.accounts.9b00296).

Bronken 2014 I. Bronken, Softening paint and drip formation in paintings by Jean-Paul Riopelle (1923–2002): improving their condition with metal coordinating Ions, in J. Bridgland (ed.), ICOM-CC 17th Triennial Conference Preprints, Melbourne, 15–19 September 2014, Paris, 2014, art. 1303.

Bronken 2023 I. Bronken, Soft Paint and the Care of Paintings, London, 2023.

Bronken and Boon 2014 I. Bronken and J. Boon, Hard Dry Paint, Softening Tacky Paint, and Exuding Drips on Composition (1952) by Jean-Paul Riopelle, in Issues in Contemporary Oil Paint, 1 January 2014, p. 247–262 (https://doi.org/10.1007/978-3-319-10100-2_17).

Bronken, Boon and Steindal 2021 I. Bronken, J. Boon and C. Steindal, Delamination of the paint structure in Peinture, Dated 1954 by Pierre Soulages. A study on soap formation in the paint layer interface, December 2021, p. 51–64.

Burnstock et al. 2017 A. Burnstock, L. Bay, J. Lee, B. Ormsby and K.J. van den Berg, Water Sensitivity of Modern Oil Paintings, in ICOM Committee for Conservation 18th Triennial Meeting Copenhagen Denmark 48 September 2017, Copenhagen, 2017 (https://www.icom-cc-publications-online.org/1778/Water-sensitivity-of-modern-oil-paintings).

Buyondo, Kasedde and Kirabira 2022 K.A. Buyondo, H. Kasedde and J.B. Kirabira, A Comprehensive Review on Kaolin as Pigment for Paint and Coating: Recent Trends of Chemical-based Paints, their Environmental Impacts and Regulation, in Case Studies in Chemical and Environmental Engineering, 6, 1 December 2022, article 100244 (https://doi.org/10.1016/j.cscee.2022.100244).

Casadio et al. 2019 F. Casadio, K. Keune, P. Noble, A. van Loon E. Hendriks, S. Centeno and G. Osmond (eds.), Metal Soaps in Art (Cultural Heritage Science), New York, 2019.

Eastaugh et al. 2004 N. Eastaugh, V. Walsh, T. Chaplin and R. Siddall, The Pigment Compendium : A Dictionary of Historical Pigments, Amsterdam/Boston, 2004.

Ferreira, van der Horst and Boon 2005 E.S.B. Ferreira, J. van der Horst and J.J. Boon, Chemical Aspects of the Binding Media of the Oranjezaal Ensemble: An Insight into 17th-Century Netherlandish Materials and Methods, in Preprints of ICOM-CC, 14th Triennial Meeting, the Hague, 2005, vol. 2, 2005, p. 774–782.

Fuster-López et al. 2016 L. Fuster-López, F.C. Izzo, M. Piovesan, D. J. Yusá-Marco, L. Sperni and E. Zendri, Study of the Chemical Composition and the Mechanical Behaviour of 20th-Century Commercial Artists’ Oil Paints Containing Manganese-based Pigments, in Microchemical Journal, 124, 1 January 2016, p. 962–973 (https://doi.org/10.1016/j.microc.2015.08.023).

Graczyk et al. 2019 A. Graczyk, P. Helou-de La Grandière, A. Phenix and S. Mirabaud, Oil Paint Straight from the Tube: Paint-specific Deterioration in Works by Alexis Mérodack-Jeaneau, 1910–1913, in K. van den Berg et al., Conservation of Modern Oil Paintings, New York, 2019, p. 229–243 (https://doi.org/10.1007/978-3-030-19254-9_17).

Helou-de La Grandière 2008 P. Helou-de La Grandière, Recherche sur le rôle des savons métalliques dans le développement des clivages des peintures de Pierre Soulages de 1959, Rapport Cnap 2008, February 2008 (https://doi.org/10.13140/2.1.1961.5528).

Helou-de La Grandière 2017 P. Helou-de La Grandière, Dégradations des œuvres de la fin des années cinquante : cas emblématique d’une œuvre de Pierre Soulages (1959, Toulouse), in Patrimoines du sud, 6, 7 September 2017, 161–176 (https://doi.org/10.4000/pds.2329).

Helou-de La Grandière 2019 P. Helou-de La Grandière, A Montparnasse Disease? Severe Manifestations of Metal Soaps in Paintings by Pierre Soulages from Around 1959 to 1960 (Delaminating Oil Paint Layers, Medium Exudates, Discolorations), in F. Casadio et al.Metal Soaps in Art (Cultural Heritage Science), New York, 2019, p. 393–412 (https://doi.org/10.1007/978-3-319-90617-1_23).

Helou-de La Grandière, Le Hô and Mirambet 2008 P. Helou-de La Grandière, A.-S. Le Hô and F. Mirambet, Delaminating Paint Films at the End of 1950s: A Case Study on Pierre Soulages, in ICOM-CC Working Group Paintings, Conference, January 2008, p. 156–162.

Helou-de La Grandière, Thoury and Simonot 2023 P. Helou-de La Grandière, M. Thoury and L. Simonot, Les Noirs de Pierre Soulages, in Technè, 55, 16 November 2023, p. 36–49 (https://doi.org/10.4000/techne.17208).

Hermans et al. 2014 J. Hermans, K. Keune, A. van Loon, M. Stols, R. Corkery and P. Iedema, The Synthesis of New Types of lead and Zinc Soaps: A Source of Information for the Study of Oil Paint Degradation, in J. Bridgland (ed.), ICOM-CC 17th Triennial Conference Preprints, Melbourne, 15–19 September 2014, Paris, 2014, article 1603.

Karakaş et al. 2011 F. Karakaş, G. Pyrgiotakis, M. Çelik and B. Moudgil, Na-Bentonite and MgO Mixture as a Thickening Agent for Water-Based Paints, in KONA Powder and Particle Journal, 29, 1 December 2011, p. 96–106 (https://doi.org/10.14356/kona.2011012).

Købke Sutton and M. Harding 2023 G. Købke Sutton and M. Harding, Ole Schwalbe, in Dansk Biografisk Leksikon, 23 April 2023 (https://biografiskleksikon.lex.dk/Ole_Schwalbe).

La Nasa et al. 2018 J. La Nasa, A. Lluveras, F. Modugno and I. Bonaduce, Two-step Analytical Procedure for the Characterization and Quantification of Metal Soaps and Resinates in Paint Samples, in Heritage Science, 6, 26 September 2018, article 57 (https://doi.org/10.1186/s40494-018-0222-1).

La Nasa et al. 2019 J. La Nasa, J. Lee, I. Degano, A. Burnstock, K.J. van den Berg, B. Ormsby and I. Bonaduce, The Role of the Polymeric Network in the Water Sensitivity of Modern Oil Paints, in Scientific Reports, 9, 1, December 2019, article 3467 (https://doi.org/10.1038/s41598-019-39963-z).

La Nasa et al. 2020 J. La Nasa, L. Nodari, F. Nardella, F. Sabatini, I. Degano, F. Modugno, S. Legnaioli, B. Campanella, M.K. Tufano, M. Zuena, P. Tomasin, Chemistry of Modern Paint Media: The Strained and Collapsed Painting by Alexis Harding, in Microchemical Journal, 155, June 2020, article 104659 (https://doi.org/10.1016/j.microc.2020.104659).

Languri 2004 G.M. Languri, Molecular Studies of Asphalt, Mummy and Kassel Earth Pigments: Their Characterisation, Identification and Effect on the Drying of Traditional Oil Paint, University of Amsterdam, 2004.

Learner et al. 2007 T.J.S. Learner, P. Smithen, J.W. Krueger and M.R. Schilling (eds.), Modern Paints Uncovered: Proceedings from the Modern Paints Uncovered Symposium, Los Angeles, 2007 (http://hdl.handle.net/10020/gci_pubs/paints_uncovered).

Modugno et al. 2019 F. Modugno, F. Di Gianvincenzo, I. Degano, I. Dorothé van der Werf, I. Bonaduce and K.J. van den Berg, On the Influence of Relative Humidity on the Oxidation and Hydrolysis of Fresh and Aged Oil Paints, in Scientific Reports, 9, 1, December 2019, article 5533 (https://doi.org/10.1038/s41598-019-41893-9).

‘Ole Schwalbe’, in Trap Danmark. Lexdk ‘Ole Schwalbe’, in Trap Danmark. Lexdk, 1 January 1994 (https://trap.lex.dk/Ole_Schwalbe).

Portræt af et maleri III, in Kunstindeks Danmark & Weilbachs Kunstnerleksikon, accessed 2 February 2024 (https://www.kulturarv.dk/kid/VisVaerk.do?vaerkId=244649).

Rogala et al. 2010 D. Rogala, S. Lake, C. Maines and M. Mecklenburg, Condition Problems Related to Zinc Oxide Underlayers: Examination of Selected Abstract Expressionist Paintings from the Collection of the Hirshhorn Museum and Sculpture Garden, Smithsonian Institution, in Journal of the American Institute for Conservation, 49, 2, September 2010, p. 96–113 (https://doi.org/10.1179/019713610804489937).

Schilling, Carson and Khanjian 1999 M.R. Schilling, D.M. Carson and H.P. Khanjian, Gas Chromatographic Determination of the Fatty Acid and Glycerol Content of Lipids. IV. Evaporation of Fatty Acids and the Formation of Ghost Images by Framed Oil Paintings, in ICOM Committee for Conservation 12th Triennial Meeting Lyon 29 August3 September 1999, vol. 1, Lyon, 1999, p. 242–247.

Schilling et al. 2016 M.R. Schilling, A. Heginbotham, H. van Keulen and M. Szelewski, Beyond the Basics: A Systematic Approach for Comprehensive Analysis of Organic Materials in Asian Lacquers, in Studies in Conservation, 61, no. sup. 3, 1 August 2016, p. 3–27 (https://doi.org/10.1080/00393630.2016.1230978).

Standeven 2011 H.A.L. Standeven, House Paints, 1900–1960: History and Use, Los Angeles, 2011.

Zviagina, Drits and Dorzhieva 2020 B.B. Zviagina, V.A. Drits and O.V. Dorzhieva, Distinguishing Features and Identification Criteria for K-Dioctahedral 1M Micas (Illite-Aluminoceladonite and Illite-Glauconite-Celadonite Series) from Middle-Infrared Spectroscopy Data, in Minerals, 10, 2, 11 February 2020 (https://doi.org/10.3390/min10020153).

Haut de page

Annexe

Analytical procedures and equipment

Optical microscopy and scanning electron microscopy coupled with energy-dispersive X-ray detection (SEM-EDX)

The stratigraphic sections were observed with an optical microscope (AxioImager M1, Zeiss) under polarised light and UV light and photographed. The sections were then analysed by scanning electron microscopy with energy-dispersive X-ray detection (SEM-EDX, Zeiss EVO 15 LS SEM coupled to BSE and EDX (X-Max80, Oxford Instruments), operated under low vacuum mode (50 Pa, 15kV). In the scanning electron microscope, a stream of primary electrons is focused onto the sample surface, resulting in a number of different particles or rays being emitted (secondary electrons, back-scattered electrons, X-rays, photons, Auger electrons, etc.). The secondary and backscattered electrons are used for imaging while the X-rays give characteristic chemical information of the emitting atoms. The probed depth in EDX analysis is around 1-3 µm. SEM-EDX only produces information on inorganic elements present in the pigments.

Micro-Raman spectroscopy (MRS)

Since SEM-EDX analyses only give information on the elemental composition of inorganic compounds, additional measurements on the cross sections were carried out by dispersive micro-Raman spectroscopy to obtain more detailed molecular information on some components. Raman is a light-scattering technique, whereby a molecule scatters incident light from a high intensity laser light source. Most of the light is scattered at the same wavelength as the laser source and does not provide useful information. A small fraction, however, is scattered at different wavelengths, which depend on the chemical structure of the pigment. By comparing the registered Raman spectrum with Raman spectra in a Raman reference pigment database, pigments can be accurately identified. However, not all pigments are Raman-active.

A Renishaw inVia instrument with a diode laser at 785 nm (Innovative Photonic Solutions) in combination with a 1200 l mm-1 grating was used for the measurements. The Raman signal was registered by a Peltier cooled (203 K) NIR enhanced deep depletion CCD detector (576 x 384 pixels). Cross sections were analysed with the aid of the Leica DMLM microscope at a magnification of 500x. Laser power was reduced to ca. 0.5 mW to avoid any possible damage. Under the conditions set, the laser spot has a diameter of a few µm and a very limited depth of focus, restricting the measurement to the pigments under investigation and avoiding interference from the surrounding pigments. Identification of the Raman spectrum was done by comparison with in-house compiled reference spectra.

Binder analysis with Py-GC/MS, with and without online methylation

In order to determine the binder(s) in the samples, the samples are analysed two times by pyrolysis gas chromatography mass spectrometry (Py-GC/MS), with and without derivatisation. In the pyrolysis unit (Frontier Lab EGA-PY3030D), a small sample particle is quickly heated in a steel pyrolysis cup, and in the absence of oxygen, to a temperature of 550°C. Larger molecules are broken down in smaller molecules that are volatile enough for gas chromatographic analysis or can be made volatile by chemical derivatisation. With this set-up, the small sample can be volatilised and analysed as such, without the need to dissolve first.

In the case of oils, natural resins and many other organic structures, chemical derivatisation during pyrolysis is necessary to obtain less polar compounds that are volatile enough to be separated in the GC system. Therefore, a second sample particle was analysed at 480°C, in the presence of tetramethyl ammonium hydroxide (TMAH) 5% in methanol.

A helium carrier gas brings these different compounds formed during pyrolysis to the inlet of the chromatographic part (GC) of the system, to be separated (Thermo Trace GC). The pyrolysis-unit is directly coupled to the chromatographic column without making use of the classic injector. The gas flow is programmed in a ramped pressure mode to optimise resolution. The following oven temperature programme was used: 1 min at 35°C, followed by heating at 10°C min-1 until 240°C, followed by heating at 6°C min-1 until 315°C; this end temperature is held for 5 min.

At the end of the column (Supelco SLB-5ms column, length 20 m, internal diameter 0.18 mm, film thickness 0.18 µm), the separated compounds are detected by the mass spectrometer (quadrupole MS, Thermo ISQ7000, scanned between 29 and 650 amu, MS at 240°C, transfer line at 290°C), providing additional mass spectrometric information on the compounds, hence making possible their identification. Based on the different compounds identified, the medium composition can be deduced.

The resulting pyrograms were manual as semi-automatically searched for marker compounds, originating from natural and synthetic resins, oils, waxes or proteins. Extensive reference databases were used; the semi-automatic searching of the pyrogram was carried out using AMDIS (Automated Mass Spectral Deconvolution and Identification System).

Fourier-transform infrared spectroscopy (FT-IR)

After the analyses using Py-GC/MS, SEM-EDX and MRS, an additional analysis with Fourier-transform infrared spectroscopy was performed on one sample. Detecting an overview of the chemical bonds in the sample, FT-IR spectroscopy is another technique for the identification of unknown polymers, resins, residues and other organic and inorganic compounds. FT-IR spectra were acquired with a Bruker Hyperion 3000 in transmission mode. By averaging 64 spectra (4000 to 650 cm-1) with a resolution of 4 cm-1 for sample measurements and background, the Mercury-Cadmium-Telluride (MCT) detector provides an adequate signal-to-noise ratio. The spectra were acquired and processed using OPUS 7.2 FTIR software.

Haut de page

Notes

1 Helou-de La Grandière, Thoury and Simonot 2023; Learner et al. 2007.

2 Helou-de La Grandière, Thoury and Simonot 2023.

3 ‘Ole Schwalbe’, in Trap Danmark. Lexdk; Købke Sutton and Harding 2023.

4 Portræt af et maleri III, in Kunstindeks Danmark & Weilbachs Kunstnerleksikon, accessed 2 February 2024.

5 Kunstkonserveringen is Denmark’s Art Conservation Center, founded in 1978 by an association of Danish art museums. With well-equipped workshops and skilled staff, they provide counselling, preservation and restoration for Danish art museums and private collectors. Their primary objective is to contribute to the optimal preservation and storage of Danish cultural heritage objects.

6 Filed as # 576 in a Microsoft Access Database at Kunstkonserveringen with the name of the artist, the title of the artwork and the inventory number of the museum, AAK 459.

7 Karakaş et al. 2011; Learner et al. 2007; Buyondo, Kasedde and Kirabira 2022.

8 La Nasa et al. 2019.

9 Schilling et al. 2016.

10 Bronken 2014.

11 Standeven 2011.

12 Zviagina, Drits and Dorzhieva 2020.

13 Eastaugh et al. 2004, p. 180–181.

14 Helou-de La Grandière 2008; Helou-de La Grandière 2017; Bronken 2014; Helou-de La Grandière 2017; Helou-de La Grandière, Le Hô and Mirambet 2008.

15 Bronken 2023.

16 Casadio et al. 2019; Bronken, Boon and Steindal 2021.

17 La Nasa et al. 2018.

18 Casadio et al. 2019.

19 Bronken 2014.

20 Bronken, Boon and Steindal 2021.

21 Graczyk et al. 2019; Bronken and Boon 2014; Bronken, Boon and Steindal 2021.

22 Bronken, Boon and Steindal 2021; Bronken 2014; La Nasa et al. 2020; Helou-de La Grandière 2019.

23 Fuster-López et al. 2016; Burnstock et al. 2017; Banti et al. 2018; La Nasa et al. 2020.

24 Similarly, long oil alkyd resins highly depend for their qualitative polymerisation on the fatty acids double bonds.

25 Bonaduce et al. 2019.

26 Ferreira, van der Horst and Boon 2005; Bronken and Boon 2014.

27 Bronken and Boon 2014.

28 Schilling, Carson and Khanjian 1999; Bronken and Boon 2014.

29 Bonaduce et al. 2018.

30 Languri 2004, p. 145.

31 La Nasa et al. 2019.

32 Rogala et al. 2010; Bronken, Boon and Steindal 2021.

33 Hermans et al. 2014.

34 Bronken, Boon and Steindal 2021.

35 Helou-de La Grandière 2019; Bronken, Boon and Steindal 2021.

36 La Nasa et al. 2019.

37 La Nasa et al. 2019.

38 Modugno et al. 2019

Haut de page

Table des illustrations

Titre Fig. 1
Légende Ole Schwalbe, Portrait of a Painting III, 1962, alkyd paint and oil paint on canvas, 100.5 × 80.5 cm (Denmark, ARoS Art Museum of Aarhus). Photo taken under normal symmetrical visible lighting in 2023, before the conservation treatment.
URL http://journals.openedition.org/kikirpa/docannexe/image/4120/img-1.jpg
Fichier image/jpeg, 324k
Titre Fig. 2
Légende Sections indicating different conservation conditions, as described. The red and yellow areas are affected with delamination. The other areas are well conserved. The stratigraphy and composition of the area indicated in blue were not examined. Normal incident lighting. Photo taken in 2023.
Crédits © Silkeborg, Kunstkonserveringen.
URL http://journals.openedition.org/kikirpa/docannexe/image/4120/img-2.jpg
Fichier image/jpeg, 351k
Titre Fig. 3
Légende Details, under normal and UV lighting, preservation condition in 2023 (fig. 3a and b); the green square indicates the detail shown in fig. 3c˗f. The lower images illustrate the characteristics of the delamination in the same area, showing the paint layer’s condition in 2001, 2010 and 2023 (fig. 3c, d,-e). Fig. 3f shows the same detail under UV lighting, preservation condition in 2023.
Crédits © Silkeborg, Kunstkonserveringen.
URL http://journals.openedition.org/kikirpa/docannexe/image/4120/img-3.jpg
Fichier image/jpeg, 224k
Titre Fig. 4
Légende Condition in 2001. At this point, the painting was in a fair condition, with minor flaking of the paint layer.
Crédits © Silkeborg, Kunstkonserveringen.
URL http://journals.openedition.org/kikirpa/docannexe/image/4120/img-4.jpg
Fichier image/jpeg, 400k
Titre Fig. 5
Légende Preservation condition in 2010 under normal symmetric visible lighting, before treatment. Nine years after the first conservation treatment, the state of preservation had worsened. The affected area was now larger, though still in the ‘red’ selection. This worsened condition led to a comprehensive structural conservation treatment that included a total impregnation with Beva 371 from the rear.
URL http://journals.openedition.org/kikirpa/docannexe/image/4120/img-5.jpg
Fichier image/jpeg, 217k
Titre Fig. 6
Légende Position of sample locations, 2023.
Crédits © Silkeborg, Kunstkonserveringen.
URL http://journals.openedition.org/kikirpa/docannexe/image/4120/img-6.jpg
Fichier image/jpeg, 299k
Titre Fig. 7
Légende Three cross sections. Left column, a sample taken at the non-affected black border area (resp. polarised light, UV, SEM-EDX mapping); middle column, affected upper black square (area indicated in red in fig. 2; resp. polarised light, UV, SEM-EDX mapping); right column, affected lower black square (area indicated in yellow in fig. 2; resp. polarised light, UV, SEM-EDX mapping). The insets in the upper row show the interface sides of the samples, 200× original magnification. P for preparation layers, B for black layers, T for transparent layers. T in the first cross section is thought to relate to T2 in the other sections, due to the common presence of aluminium (aluminium hydrate?).
URL http://journals.openedition.org/kikirpa/docannexe/image/4120/img-7.jpg
Fichier image/jpeg, 433k
URL http://journals.openedition.org/kikirpa/docannexe/image/4120/img-8.png
Fichier image/png, 8,5k
Titre Fig. 8
Légende Stratigraphic scheme in the non-affected border, and with the affected upper and lower squares indicated in fig. 2. In vivid blue, the layers coloured with bone black; in pale blue, the layers coloured with carbon black; in violet, a pigment identified as aniline black; in orange, the inclusions of clay.
URL http://journals.openedition.org/kikirpa/docannexe/image/4120/img-9.jpg
Fichier image/jpeg, 178k
Titre Fig. 9
Légende THM-GC/MS chromatogram of black paint flake sampled in the upper black square. ME: methyl ester; FA: fatty acid; DME: dimethyl ester. Partially oxygenated fatty acids indicate ongoing degradation. The position of three alkyl phenyl alkanoates (APA), markers for heat-bodied oil, is indicated.
URL http://journals.openedition.org/kikirpa/docannexe/image/4120/img-10.jpg
Fichier image/jpeg, 94k
Titre Fig. 10
Légende FTIR spectrum of the sample taken in the affected zone (upper square) in Portrait of a Painting III. The spectrum shows a composite spectrum of bone black, binder and some indications of clay minerals (mainly 3570 cm-1 and 663 cm-1). The peak around 1456 could be associated with regular zinc stearate (Zn(stearate)2). Around 1585 a broad absorption band could indicate a zinc-sodium soap of mixed fatty acid composition (ZnNa2(palmitate/stearate)4).
URL http://journals.openedition.org/kikirpa/docannexe/image/4120/img-11.jpg
Fichier image/jpeg, 63k
Haut de page

Pour citer cet article

Référence électronique

Louise Decq, Steven Saverwyns, Filiz Kuvvetli et Elizabeth Baadsgaard, « Paint Composition and Delamination Concerns in Ole Schwalbe’s Portrait of a Painting III, 1962 », Bulletin de l’Institut royal du Patrimoine artistique / Bulletin van het Koninklijk Instituut voor het Kunstpatrimonium [En ligne], 39 | 2024, mis en ligne le 02 octobre 2024, consulté le 04 décembre 2024. URL : http://journals.openedition.org/kikirpa/4120 ; DOI : https://doi.org/10.4000/12iwx

Haut de page

Auteurs

Louise Decq

Louise Decq holds a degree in bioengineering and obtained a master’s in art history from KU Leuven and a PhD in science: chemistry from Ghent University. Since 2010 she has worked as a chemist in the Painting Lab of the Royal Institute for Cultural Heritage (KIK-IRPA), analysing artworks using various scientific methods (with an emphasis on chromatography) and interpreting results historically in collaboration with restorers and art historians.

Steven Saverwyns

Steven Saverwyns holds a PhD in science: chemistry from Ghent University. He currently heads the Painting Lab of the Royal Institute for Cultural Heritage (KIK-IRPA), providing analytical support primarily for paintings. His expertise spans various analytical methods, including macro-X-ray fluorescence and micro-Raman spectroscopy. His research focuses on classical and modern paintings, with particular emphasis on material-technical examination in the context of authenticity studies.

Filiz Kuvvetli

Filiz Kuvvetli graduated as a paintings conservator in 1999 with a BSc. in art conservation and obtained an MSc. in art conservation in 2005 from the School of Conservation of the Royal Danish Academy. She worked as a paintings conservator in Amsterdam (1999–2001) and The Hague (2002) and subsequently at various conservation studios in Denmark until 2007. Since then, she has been a paintings conservator at the Kunstkonserveringen (Art Conservation Center). In 2014 she was a visiting researcher at the Getty Conservation Institute (US).

Elizabeth Baadsgaard

Elizabeth Baadsgaard graduated as a paintings conservator with an MSc. from the School of Conservation of the Royal Danish Academy in 2002 and obtained a BA in art history in 2006. She worked as a paintings conservator in the Netherlands from 1992 to 1996 and subsequently at various conservation workshops in Denmark and abroad. Since 2010, she has been Head of Conservation of the Western Department of Kunstkonserveringen (Art Conservation Center).

Haut de page

Droits d’auteur

CC-BY-4.0

Le texte seul est utilisable sous licence CC BY 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search