A Cloudy Appearance: An Investigation into the White Haze on Jacobus van Looy’s Kinderen

(1)

A Cloudy Appearance:

An Investigation into the White Haze on Jacobus van

Looy’s Kinderen

Master thesis

Christa Molenaar, 10633847

Thesis supervisor: mw. dr. Emilie Froment

Second reader: mw. Kate van Lookeren Campagne Contacts RCE: mw. ing. Saskia Smulders-de Jong and prof. dr. Klaas Jan van den Berg

University of Amsterdam, Amsterdam June 2019

(2)

Preface and Acknowledgements

This master thesis is written within the programme of the Master Conservation and Restoration of Cultural Heritage, Paintings Specialisation, at the University of Amsterdam (UvA). The research was conducted over the course of three months under the supervision of Emilie Froment (UvA), whom I would like to thank for her support, feedback and valuable advise. I am also thankful that Lidwien Speleers, conservator of the Dordrecht Museum, suggested the thesis subject and provided me the opportunity to investigate the painting in her studio. She also assisted me with photographing the object, which is much appreciated.

I would furthermore like to express my gratitude to Maartje Stols-Witlox (UvA) for the SEM-EDX analysis and the guidance with in the interpretation of the measurements, together with Ineke Joosten who also provided additional advice. I am also thankful to the other involved UvA teachers Maarten van Bommel, Ella Hendriks and Rene Peschar for their guidance and advice.

This research owes much to the contribution of the Cultural Heritage Agency of the Netherlands (RCE). Many thanks to Saskia Smulders-de Jong, Conservation Scientist at the RCE, for carrying out the GC/MS analysis and to Klaas-Jan van den Berg, Senior Researcher at the (RCE) and Professor of the Chemical Aspects of the Conservation and Restoration of Cultural Heritage at the UvA, for his advise and guidance with regard to the interpretation of the results. The input of Aviva Burnstock, head of the Department of Art Conservation & Technology at the Courtauld Institute in London, on the interpretation of the SEM SE images was furthermore very much appreciated.

Last but not least, many thanks to my lovely fellow students Ivana Jerdoneková, Jolijn Schilder, Sofie Dubbeldam and Steinunn Harðardóttir, for their great support and help during this project!

(5)

Summary

This master thesis describes the investigation into a white haze that is visible on a substantial part of Kinderen’s surface. This oil painting is made by Jacobus van Looy (1855-1930) in 1888 and is currently part of the Dordrecht Museum’s collection. The haze gives the painting a blanched appearance, due to optical light scattering. This degradation phenomenon disturbs the image to such a degree that it is no longer considered suitable for exhibition. The aim of this study is to determine the nature of the whitish haze and will furthermore attempt to connect this phenomenon to the material composition of Van Looy’s painting and his employed painting technique. Based on several preliminary tests and a brief literature study the hypothesis is formulated that the white haze is likely to be caused by the migration of either free fatty acids or wax.

A literature research was carried out into the occurrence of white hazes caused by either fatty acids or wax, as well as factors that could have influenced their migration towards the paint surface. The nature of the white haze was investigated with visual examination, GC/MS and SEM SE. The material composition and painting technique were studied with several analytical techniques, such as optical microscopy and SEM-EDX. This provided insight into the materials present in Kinderen as well as Van Looy’s working methods, and their relation to the white haze.

The research indicated that the migration of free fatty acids is probably responsible for the appearance of the white haze. Some of the materials found in Kinderen, as well as certain aspects of Van Looy’s painting technique seem to have played a profound role in the formation of the surface whitening on the painting. Other aspects could have stimulated the blooming of fatty acids as well, such as the ambient

climatological conditions and previous conservation treatments, but they were not investigated in this study.

(6)

Samenvatting

Deze masterscriptie onderzoekt een witte waas die zichtbaar is op een groot gedeelte van het oppervlak van Kinderen. Dit betreft een olieverfschilderij dat

gemaakt is door Jacobus van Looy (1855-1930) in 1888 en momenteel onderdeel is van de collectie van het Dordrecht Museum. De waas geeft het schilderij een wittige gloed, als gevolg van optische lichtverstrooiing. Dit fenomeen verstoort de afbeelding zodanig, dat het werk niet langer als geschikt voor tentoonstelling wordt beschouwd. Deze studie beoogt de aard van de witte waas te bepalen en zal ook pogen het fenomeen te relateren aan het materiaalgebruik en de techniek van Van Looy. Op basis van enkele voorlopige tests en een bondige literatuurstudie is de hypothese geformuleerd dat de witte waas waarschijnlijk is veroorzaakt door het migreren van vrije vetzuren of was.

Een uitvoerige literatuurstudie is gedaan naar het optreden van witte wazen

veroorzaakt door ofwel vrije vetzuren ofwel was. Daarnaast zijn factoren onderzocht die invloed kunnen hebben uitgeoefend op het migreren van beiden naar het

verfoppervlak. De aard van de waas is geanalyseerd met visueel onderzoek, GC/MS en SEM SE. Het materiaalgebruik en de schildertechniek zijn onderzocht met onder andere optische microscopie en SEM-EDX. Hierdoor kon inzicht verkregen worden in de materialen die aanwezig zijn in Kinderen en Van Looy’s schildertechniek, als ook hun relatie tot de witte waas.

Het onderzoek toonde aan dat de migratie van vrije vetzuren waarschijnlijk

verantwoordelijk is voor het verschijnen van de witte waas. In Kinderen zijn materialen gevonden die invloed kunnen hebben uitgeoefend op het ontstaan van de waas, en Van Looy’s schildertechniek heeft hier waarschijnlijk ook aan bijgedragen. Andere factoren kunnen bovendien de uitbloei van vetzuren hebben gestimuleerd, zoals de klimatologische omstandigheden en voorgaande restoraties, maar zijn in deze studie niet onderzocht.

(7)

Chapter I: Introduction

This thesis describes the investigation into a painting owned by the Dordrecht

Museum, that is painted by Jacobus van Looy (1855-1930) in 1888 (fig. 1). The oil on canvas painting is called Kinderen, it is 96 centimetres in height and 66

centimetres in width. The signed painting 1

depicts two sisters, who posed for Van Looy in his studio. The eldest sister Cor is sitting in a chair, whilst her younger sister Jo is sitting on her lap. The girls are 2

depicted with long, dark blond hair and big dark blueish eyes. Both are wearing a reddish hat piece, as well as cloths in a red and black colour palette. The more or less even background has a golden hue. To fit a frame, Van Looy cut a piece of the top part of the canvas that originally depicted a large plane with incoming sunlight. According to Van Looy, this made the children look finer than they appear now. 3

The back of the painting holds two labels, which inform us on the provenance of the painting.

1

The inscriptions include: ‘Geschenk aan mijn lieve vrouw bij ons huwelijksfeest 23 september 1889 gevierd te Brussel’ and ‘zie: het stukje “Kinderen” door Jac. van Looy in de Nieuwe Gids van 1 december 1890.’ After Roodenburg bought the painting in 1889, the heirs of the family donated the painting to the Dordrecht Museum in Dordrecht, in 1915.

Bel et al., Jacobus van Looy 1855-1930 : niets is zoo mooi als zien…, eds. Joyce van der

Smit-2

Meijer and Chris Will (Zwolle: Waanders uitgevers, Haarlem: Frans Halsmuseum and Haarlem: Stichting Jacobus van Looy, 1998), 94.

Ibidem.

3

Fig. 1: Jacobus van Looy, Kinderen, 1888. Oil on canvas, 96x66cm. Collection Dordrecht Museum, inv. nr. DM/915/435. Photo from website Dordrecht Museum (taken in 2013).

(8)

Kinderen currently suffers from a whitish haze, which covers a substantial part of the

painting’s surface. The haze gives the painting a blanched appearance, due to optical light scattering. This degradation phenomenon disturbs the image to such a degree that it is no longer considered suitable for exhibition. This diagnostic research aims to investigate what is causing the appearance of this white haze. The Dordrecht

Museum is eager to learn whether the haze can safely be removed, for which it is essential to determine its nature. Improved understanding of the phenomenon might also help to determine possible future preventive conservation measures.

Van Looy’s painting materials and technique have already been the topic of research in the past by Boitelle. He found that many of Van Looy’s paintings suffer from 4

issues that are related to a deficiency in the drying process of the oil paint. This 5

research aims to expand our knowledge of Van Looy’s working methods with visual and technical examination of Kinderen. It will furthermore try to connect the white haze to the material composition of Van Looy’s painting and his employed technique. Other aspects could have influenced the appearance of the white haze as well, but are not investigated due to the limited scope of in this study.

1.1 Research question and hypothesis

The following research question is formulated to investigate the white haze and the possible connection with Van Looy’s working methods:

How can the whitish haze on the surface of Jacobus van Looy’s Kinderen (1888) painting be characterised, and can this phenomenon be related to Van Looy’s use of materials and painting technique?

In 2014, the conservator of the Dordrecht Museum, Lidwien Speleers, carried out two small tests to examine the characteristics of the white haze. These tests showed that the haze disappeared upon heating with a hair drier and that the haze did not

René Boitelle conducted research into Van Looy’s painting technique at the Frans Hals museum

4

in 1993-94, a part of his research is published in: Bel et al., Jacobus van Looy 1855-1930. Bel et al., Jacobus van Looy, 51.

(9)

dissolve in saliva. Based on these observations, as well as a brief literature study, the 6

hypothesis is formulated that the white haze is likely to be caused by the migration of either free fatty acids or wax.

1.2 Methodology and structure

The research methodology includes visual examination of Van Looy’s Kinderen with the naked eye, as well as with the stereomicroscope. The painting is moreover investigated with ultraviolet (UV) light and Infrared reflectography (IRR). Furthermore, instrumental analysis of scrapings and paint samples are included to gain

understanding of the nature of the white haze, and to get insight in Van Looy’s painting technique and materials. Also, a literature study of relevant publications is carried out and comparisons are made with other works by Van Looy. Particularly interesting is Mevrouw Van Looy lezend bij petroleum lamp (1895), which seems to suffer from the same phenomena (See Appendix IV). 7

After the nomenclature on white hazes in conservation (science) is described in the following paragraph, Chapter II will outline the current scientific knowledge of Van Looy’s bibliography, his materials and painting technique. This chapter will

furthermore include the scientific knowledge on white hazes caused by fatty acids or waxes and will give a theoretical model on fatty acids in drying oils, as well as the drying/curing process of oil paints.

Chapter III will describe the investigation carried out into the nature of the white haze that is visible on Kinderen. This chapter will start with a description of the appearance and characteristics of the haze, followed by a section on the applied analytical

techniques. Both the results, as well as an interpretation of the outcomes will be given, which will help explain the nature of the haze in the discussion of this chapter.

See Appendix III for Lidwien Speleers her notes of the preliminary tests.

6

See Appendix IV for images of the painting’s referred to in the text, other than Kinderen by Van

7

(10)

Chapter IV will try to connect the white haze to Van Looy’s choice of materials, as well as his painting technique. For this purpose, the painting’s layer build-up will be discussed from the canvas support up to the varnish and the possible influence of his used materials will be elaborated on. Also, several examples aim to demonstrate that Van Looy’s methods might have influenced the appearance of the white haze. The outcomes of the research are brought together to answer the main research question in the conclusion. A reflection on the research methods, suggestions for further research, as well as treatment options and recommendations for preservation are also given.

1.3 Nomenclature white hazes

In the field of conservation, many terms have been used to describe a whitish appearance on the surface of easel paintings: whitening, blanching, hazing, blooming, efflorescence, chalking, clouding, sweating, crazing, fading, bleaching, whitening, crystallisation, micro-cracking/fissures/voids, and blushing. Van Loon 8

describes that most of these terms are quite specific and relate to the mechanism of whitening. For instance, fading or bleaching refers to discolouration or colour loss of 9

a pigment, and blanching, interchangeably used with the terms micro-cracking/ fissures/voids and crazing, indicates a physical break-up of the paint surface. However, the definition of terms is not always as straightforward.

Van Loon et al. describe blanching as a general term which often refers to any sort of optical surface whitening. This is already different from what Van Loon describes in 10

her dissertation, were the term is more specifically related to the physical break-up of the paint surface. Van Loon states in her dissertation that when the exact mechanism of the whitening is unknown, it is “safer and clearer to use a general term to describe

Annelies van Loon, “Color changes and chemical reactivity in seventeenth-century oil

8

paintings” (PhD diss., University of Amsterdam, 2008), 122. Ibidem.

9

Annelies van Loon, Petria Noble, and Aviva Burnstock, "Aging and Deterioration of Traditional

10

Oil and Tempera Paints,” in Conservation of Easel Paintings, eds. Joyce Hill Stoner and Rebecca Rushfield (London: Routledge. 2012), 234.

(11)

the surface phenomenon, such as whitening, hazing or white surface depositions (exudates/accretions), precipitates or crusts (encrustations).” 11

Blooming and sweating may be used for the process of migration and deposition of extractable components from the oil paint on the surface. Both terms are typically related to whitish deposits of organic origin, such as free fatty acids, waxes, or soaps. Efflorescence on the other hand “may also imply, and include, the presence 12

of crystalline (inorganic) deposits on the paint surface, usually water-soluble salts.” 13

Van Looy describes the term in her dissertation as: “more specific for crystalline (inorganic) deposits on the surface, water-soluble salts.” 14

Ordonez et al. point out that the terms blanching, chalking and blooming “have been used inconsistently to describe the various phenomena that result in irregular hazy patches on surfaces.” This article introduces the use of the term efflorescence in 15

architectural conservation, the building industry and geology. Here, efflorescence is used to describe inorganic salts that are found on the surface of walls or rocks, which migrated through the substrate towards the surface. Ordonez et al. have 16

chosen to use this term in a broader manner to describe the migration of organic and inorganic compounds to the painting’s surface.

Akerlund describes that the term efflorescence is used in conservation literature for organic components and their migration processes. It is not completely understood 17

where this statement is precisely founded on, since Akerlund refers to Ordonez et al.,

Loon, “Color changes and chemical reactivity in seventeenth-century oil paintings”, 122.

11

Loon, "Aging and Deterioration of Traditional Oil and Tempera Paints,” 234 / Loon, “Color

12

changes and chemical reactivity in seventeenth-century oil paintings”, 122. Loon, "Aging and Deterioration of Traditional Oil and Tempera Paints,” 234.

13

Loon, “Color changes and chemical reactivity in seventeenth-century oil paintings”, 122.

14

E. Ordonez, and J. Twilley, “Clarifying the haze: efflorescence of works of art,” Analytical

15

Chemistry, vol. 69 (1997): 416A.

Ibidem.

16

Luciana Akerlund, "Efflorescence in paintings and the role of moisture,” Kleurhistorisch

17

Platform, January 18, 2013,

(12)

who actually use the term for the migration of both organic and inorganic

compounds, and Boon et al. who do not seem to mention the term efflorescence in their article. 18

In summary, whitening, hazing and white surface depositions are more general terms, that can be used if the mechanism of the whitening is unknown. Both blooming and sweating are quite consistently used for organic, free fatty acid deposits and their associated mechanisms. In the broadest definition efflorescence can also be used for this purpose, although some authors use it more specifically for inorganic

components. The nature of the haze is under investigation in this thesis and unknown upon writing. Therefore, the more general terms whitening and hazing will be used interchangeably in the text.

Jaap J. Boon, Frank Hoogland, and Katrien Keune, “Chemical Processes in Aged Oil Paints

18

(13)

Chapter II: Current scientific knowledge

2.1 Biography Van Looy

The main body of written work about the Dutch writer and painter Van Looy, discusses either his written oeuvre or outlines his biography. Although some 19

passages of these publications provide us with valuable information about for

instance Van Looy’s artistic training, these publications do not (or very limited) inform us about the materialistic side of his work, nor his working methods. An exception is a catalogue that accompanied an exhibition on Van Looy at the Frans Hals Museum in 1998. In this work Bel et al. provide a great deal of information on both Van 20

Looy’s artistic life, as well as his painting technique.

From these art historical sources we learn that Van Looy moved into the Gereformeerde Burgerweeshuis in Haarlem at the age of five, after his parents passed away. His noticeable drawing talent gave him the opportunity to follow drawing lessons at the Burgeravondschool. In order to learn a profession, Van Looy received training as a house and carriage painter. His artistic talent enabled him to subsequently study at the Rijks-Academie van Beeldende Kunsten in Amsterdam, from 1877 onwards. Here, Van Looy followed classes by the painter August Allebé 21

and developed friendships with some of his fellow students, like Maurits van der Valk

Sources that describe Van Looy’s biography and the art historical side of this painted oeuvre

19

include:

- Mischa Andriessen et al., Rumoer in de stad: de schilders van tachtig, ed. Frouke van Dijke (Den Haag: Gemeentemuseum Den Haag and Zwolle: WBOOKS, 2017).

- M.J. Brusse, Jacobus van Looy over zijn werk (Rotterdam: W.L. & J. Brusse, 1930).

- F.P. Huygens, Wie dronk toen water! Bloemlezing uit de briefwisseling met August Allebé

gedurende zijn Prix de Rome-reis 1885-1887 verzorgd door F.P. Huygens (Amsterdam:

Meulenhoff, 1975).

- George Slieker, Van heiligland tot hout: plaatsen uit Jacobus van Looy's Jaapje-cyclus en zijn

persoonlijke leven in foto's en ansichtkaarten (Koog aan de Zaan: Leesmijnboek.nl, 2007).

- Peter J. A. Winkels, “Jacobus van Looy: schilder van huis uit, schrijver door toevallige omstandigheden: een biografische schets” (PhD diss., University of Amsterdam, 1981). Bel et al., Jacobus van Looy.

20

Slieker, Van heiligland tot hout, 9.

(14)

and Willem Witsen. In 1882 Van Looy travelled to Paris with Witsen where he visited the Salon. 22

An interesting source of information entails a handwritten correspondence between Van Looy and his teacher Allebé, that has been made available for research online. 23

In 1884 Van Looy won the Prix de Rome, an incentive prize for young and promising artists. This offered him the chance to go on Grand Tour to Italy and Spain between 1885-1887. During his journey Allebé and Van Looy wrote regularly about the 24

drawings and paintings he created, although only seldom about his materials.

In 1975 Huygens wrote an anthology on the correspondence between Allebé and Van Looy. Although this publication only contains a selection of the correspondence 25

between the two artists, it is much better readable in comparison to the handwritten letters. Van der Smit-Meijer studied this correspondence in 1998 to learn about Van Looy’s artistic views. She concluded that originality and independence were 26

important values for Van Looy’s art. He derived inspiration from nature in all its

greatness, had interest in the craft side of artistry and admired Rembrandt and Millet. The letters were also studied by Boitelle, who studied Van Looy’s painting

technique. 27

In April 1887 Van Looy settles in Amsterdam. A successful period in his lifetime has arrived, as his Spanish drawings and paintings are exhibited in the Rijks-Academie. 28

In this period Van Looy painted his Kinderen painting, the topic of this thesis. A few

Bel et al., Jacobus van Looy, 175.

22

"Correspondentie Van De Schilder En Schrijver Jacobus Van Looy,” Geheugen Van Nederland,

23

accessed February 5, 2018, https://www.geheugenvannederland.nl/nl/geheugen/pages/collectie/ Correspondentie van de schilder en schrijver Jacobus van Looy.

Slieker, Van heiligland tot hout, 9.

24

F.P. Huygens, Wie dronk toen water! Bloemlezing uit de briefwisseling met August Allebé

25

gedurende zijn Prix de Rome-reis 1885-1887 verzorgd door F.P. Huygens (Amsterdam:

Meulenhoff, 1975).

Bel et al., Jacobus van Looy, 11-20.

26

Ibidem, 51-60.

27

Ibidem, 176.

(15)

years later, Van Looy received negative reviews on his first solo exhibition in 1901 at Arti et Amicitiae in Amsterdam. He decided to never exhibit as a painter again. The impact of the criticism he received becomes apparent in an interview six years later, where he refuses to discuss his visual arts: “Van den schilder moesten we liever zwijgen… want dat raakt me aan zoo veel intiems, en ‘k wil niet bitter worden…”. 29

Although Van Looy was predominantly undervalued in his own lifetime, a renewed appreciation arose from the 1960’s onwards. The aforementioned anthology by Huygens was published in 1975, followed by a biographical overview in 1982, a publication on his double sided talent in 1988 by Will and Winkels and the exhibition catalogue on Van Looy in 1998. Furthermore, Van Looy’s paintings are now 30

frequently exhibited by museums and his work is being sold in auction houses. This renewed interest resulted in the recognition for Van Looy’s painted oeuvre by the public, museums and auction houses.

2.2 Van Looy’s materials

Only one publication is found that describes Van Looy’s painting materials and/or technique. Boitelle conducted research at the Frans Hals museum in 1993-94. All 31

paintings in the collection of the Stichting Jacobus van Looy received a condition report and two paintings were subsequently treated. The RCE investigated several 32

paint samples with analytical techniques for this project. Since Kinderen is not part of the investigated collection, it was not researched. Van Looy’s palette was

investigated in the context of the MolArt-project (Molecular Aspects of Ageing in Painted Works of Art, a project from 1995 to 1999), a multidisciplinary research project into the ageing of artworks, among which poorly drying oil paints in nineteenth century paintings.

M.J. Brusse, Jacobus van Looy over zijn werk (Rotterdam: W.L. & J. Brusse, 1930 ), 10.

29

Bel et al., Jacobus van Looy, 34-35.

30

Ibidem, 51-60.

31

‘Stichting Jacobus van Looy’ was founded in 1949, nine years after the death of Van Looy’s

32

widow Titia van Looy-Gelder. The foundation became the owner of approximately 40 paintings, several hundreds of drawings, and many manuscripts and letters by Van Looy. This collection is nowadays located at the Frans Hals Museum in Haarlem.

(16)

2.2.1 Paint tubes

Boitelle described that Van Looy, together with several other artists, signed a

statement in 1900 recommending the paint tubes from Claus & Fritz. The analysis of the paints on Van Looy’s palette showed fine and evenly distributed pigment grains, indicating Van Looy indeed used paint tubes. Van de Laar et al. investigated the paints from Claus & Fritz, by analysing the inorganic pigments and binding media of 41 different paints from the 1921 line. Drying oil appeared to be the main 33

component. The precise nature of the oil could not be determined, because

classification is normally based on saturated fatty acids in hardened paint, while the samples had to be taken from not yet hardened paint. However, in some cases “irregularities were found that would indicate unusual oil components.” 34

Paint manufactures could add multiple materials to tube paints, such as non-drying oils, waxes, and balms. This enhanced the shelf-life and improved the smoothness of the paints. Additives could also have downsides, for instance wax tended to induce 35

drying issues and required a high oil content to reach a saturated, deep colour. The addition of more oil to the paint, could subsequently result in a fatty substance which would dry more slowly. In some tubes investigated by Van de Laar et al., minimal 36

amounts of beeswax and triterpenoid resin were found. However, the presence of these materials could not be confirmed in Van Looy’s paints with the analysis carried out during Boitelle’s research. 37

It should be noted that the investigated paints by Van de Laar et al. were produced some 30 years later than Kinderen was painted. It is therefore likely that the

composition of the Claus & Fritz’s paints changed within this timeframe. For instance, Van de Laar et al. describe that the cadmium colours only came into commercial

Michel van de Laar, and Aviva Burnstock, “"With Paint from Claus & Fritz": A Study of an

33

Amsterdam Painting Materials Firm (1841-1931),” Journal of the American Institute for

Conservation 36, no. 1 (Spring, 1997), 1-16.

Ibidem, 9.

34

35

Ibidem.

36

Ibidem, 54-58.

(17)

production around 1910. This means that when Van Looy painted Kinderen, the 38

cadmium colours were not yet available. The type and amount of paint additives were possibly also modified during these 30 years, indicating some limitations when interpreting Van de Laar et al.’s article.

2.2.2 Van Looy’s palette

Boitelle describes Van Looy’s palette as “heavily used”, with many paint blobs and wipes, and unmixed paint only at a few spots. An overview of the results of the 39

analysis of the materials on the palette has been published by Boitelle et al. (Table 1). Most samples contained linseed oil, in some cases with some pre-polymarised 40

oil.

Determining the elements did not always unambiguously identify the pigments in the paint. Especially the identification of the yellow and orange pigments was difficult, according to Boitelle. The colours could either be mixed on the palette from paints originating from different tubes, or the paint came from the tube in the way it was found on the palette. In all paint samples barium sulphate was found, except for the black, white and blue (2) paint. Barium sulphate was used as a filling material in tube paints. 41

Table 1: Analysis of Van Looy’s palette with: XRF, SEM-EDX and XRD.

Colour Elements Interpretation of the pigment

Black Pb, Zn, Ca, Fe brown-black ocher

Light yellow Pb, Cr, Zn, Ba, S barium chromate

Orange Pb, Cr, Zn, Cd, S, Ba cadmium orange and/or chrome

yellow

Green Pb, Cr, Zn, Ba chrome oxide green

Bright orange Pb, Zn, Cd, S, Ba cadmium yellow or cadmium orange

Van de Laar, “With Paint from Claus & Fritz”, 7.

38

39

Ibidem, 60.

40

Ibidem.

(18)

It is unknown during which period Van Looy used this palette, but it is certain that the cadmiums on the palette were not available when van Looy painted Kinderen. It is also noticeable that no red paint is visible on the palette, a very prominent colour in

Kinderen.

2.3 Van Looy

’s painting technique

Boitelle found that many of Van Looy’s paintings suffer from issues that are related to a deficiency in the drying process of the oil paint. This can result in the formation of 42

(extensive) drying cracks and deformations in the paint, and both phenomena are often encountered in his paintings. Boitelle furthermore describes the formation of fine transparent grains on the surface of one painting he investigated Mevrouw Van

Looy lezend bij petroleumlamp. The transparent grains were found to be fatty acids,

that migrated from the paint layer towards the surface of the film. According to Boitelle, this phenomenon is also likely to be a consequence of the bad drying properties of the paint. 43

In 1999 Schilling investigated the formation and composition of ghost images, i.e. hazy films that may appear on the inside surface of protective glasses over framed oil paintings. One mechanism responsible for ghost image formation is the evaporation of free fatty acids. He found that this phenomenon is related to the quantity of free 44

Yellow Pb, Cr, Zn, Cd, S, Ba lead chromate, cadmium yellow,

barium chromate

Blue (1) Pb, Si, Ca, Fe, Zn, S, Ba Synthetic ultramarine

Blue (2) Pb, Si, Fe, Zn, S Synthetic ultramarine

White Pb, Ca, Zn lead white, zinc white

Colour Elements Interpretation of the pigment

42

Ibidem, 58.

43

M.R. Schilling, Carson, D.M., and Khanjian, H.P., “Gas chromatographic determination of the

44

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 August-3 September 1999: Preprints, ed. J. Bridgeland (London: James & James,

(19)

fatty acids in paints, which in turn is affected by the paint’s composition. The type 45

of pigments and oil, as well as the amount of oil in the paint influences the paint’s drying properties. Schilling found that slow drying paints are particularly susceptible to the formation of whitish surface deposits of fatty acids, confirming Boitelle's findings. 46

Boitelle furthermore found that Van Looy’s painting technique greatly influenced the drying problems of the paint. He tended to work and retouch with binder-rich paint, on areas which did not yet dry completely. He was often insecure about his work, which led him to rework his paintings wet-in-wet multiple times. This practice prevented his paintings to dry properly. 47

2.4 White haze formation due to the evaporation of free fatty acids or wax

On the Conservation Wiki website, Gridley describes the following physical, chemical and biological processes that can cause the appearance of a white haze: dust and atmospheric pollutants, a physical break-up of the surface film causing light

scattering, mould growth, migration of free fatty acids or wax, saponification and complex salt formation, and lastly the formation of epsomite/magnesium sulfate heat hydrate. As already described, the hypothesis is formulated that the white haze is 48

likely to be caused by the migration of either fatty acids or wax. This paragraph will therefore focus on these mechanisms.

Burnstock et al. mention in Issues in Contemporary Oil Paint that with the application of new analytical techniques it is possible to distinguish between the different causes and classes of optical whitening, which can be of “organic, inorganic and

Schilling et al., “Gas chromatographic determination,” 246.

45

Ibidem.

46

47

Mary Gridley, "White Surface Hazes,” White Surface Hazes, last modified November 7, 2017,

48

(20)

organometallic (soaps) origin”. The characterisation of the type of surface whitening 49

can inform conservation practice, of which Hinde et al.’s investigation of surface whitening in a series of twentieth century Spanish paintings provides a good

example. Hinde closely examined the causes of whitening and used the results for 50

their choice of the most appropriate treatment for the paintings.

Van Loon et al. describe in The Conservation of Easel Paintings, a comprehensive text on the history, philosophy, and methods of treatment of easel paintings, that free fatty acids are capable of migrating through paint layers. The composition of the 51

deposits on the surface is furthermore not static, since they can become thicker and spread over time. Akerlund conducted a research project that focussed on fatty acid efflorescence on paintings from the nineteenth up to the twenty-first century. She 52

discusses several case-studies, among which one varnished painting. She describes that the fatty acid crystals seem to emerge from the thick, crack-free varnish layer. This indicates that fatty acids can also migrate through varnish layers. Akerlund cautiously concluded that the “presence of a natural resin component in the varnish may affect the ease of migration of free fatty acids through an unbroken film.” 53

2.4.1 Appearance of fatty acid deposits

Akerlund describes the appearance of fatty acid deposits as ‘tiny needles’ or

‘crystals’ on the surface of paintings, which appear white due to their light scattering effect. They range in size from 1.5 mm to 10 mm and are usually easily removed with mechanical action. Both Akerlund and Hinde revealed the morphology of free fatty

Aviva Burnstock, and Klaas Jan van den Berg, “Twentieth Century Oil Paint. The Interface

49

Between Science and Conservation and the Challenges for Modern Oil Paint Research,” in Issues

in Contemporary Oil Paint (Heidelberg: Springer International Publishing, 2014), 4.

Laura Hinde, Klaas Jan van den Berg, Suzan de Groot and Aviva Burnstock, "Characterisation

50

of Surface Whitening in 20th Century European Paintings at Dudmaston Hall, United Kingdom,” in

ICOM-CC Postprints, Paintings (Lisbon, 2011), 1-10.

Van Loon et al., “Aging and Deterioration”, 235-36.

51

Akerlund, "Efflorescence in paintings and the role of moisture.”

52

Ibidem.

(21)

acid deposits with scanning electron micrograph (SEM). They describe the 54

appearance with SEM as clusters of shard like structures of 10-20um in length. 55

2.4.2 Sources of free fatty acids

Akerlund, Hinde et al. and Ordonez et al. researched the blooming of fatty acids on nineteenth and/or twentieth century paintings. They described the following possible sources of free fatty acids: linseed oil, and even higher levels of free fatty acids in linseed stand oil, the addition of metal stearate and/or wax to the paint, and previous conservation treatments. Other factors that can explain the distribution of fatty acid 56

blooming on a painting, are the pigment composition and the physical structure of the painting (i.e. the number and thickness of the paint layers). Noble et al. 57

investigated the formation of whitish deposits on the inside of the microclimate box of a seventeenth-century painting. They describe that the migrating free fatty acid

components originated from the oil matrix, where they are released through

hydrolysis of the glyceride ester link and oxidation reactions at double bonds. They 58

also point to the role of the paint composition.

2.4.3 Pigments

Noble and Loon investigated the evaporation of free fatty acids from a seventeenth century painting by Dou. They describe that pigments that retard the drying of oil and are highly oil-absorbing are more often related to areas in paintings that exhibit free fatty acid migration. Among these pigments are mentioned: black carbonised

Ibidem / Hinde, "Characterisation of Surface Whitening”, 1-3.

54

Hinde, "Characterisation of Surface Whitening”, 1-3.

55

Ordonez, “Clarifying the haze”, 418-19 / Akerlund, "Efflorescence in paintings and the role of

56

moisture.” / Hinde, "Characterisation of Surface Whitening”, 4. Ordonez, “Clarifying the haze”, 420.

57

Petria Noble and Annelies van Loon, “Evaporation of fatty acids and formation of whitish

58

deposits on the inside of the glass/microclimate boxes: a case study in the Mauritshuis,” in

PROPAINT Improved Protection of Paintings during Exhibition, Storage and Transit Final Activity Report, edited by Elin Dahlin (Kjeller: Norwegian Institute for Air Research, 2010), 158-59.

(22)

products, Kassel earth, ochres, ultramarine, vermilion and alizarin. These pigments 59

furthermore lack polyvalent metal ions which can interact with the carboxylic acid groups of the free fatty acids, to create a stable cross-linked structure; a topic that will be discussed further in Chapter 2.5.

Ordonez et al. also found that certain pigments were more likely to be afflicted by fatty acid efflorescence. Carbon black and cobalt blue were often affected, whilst 60

lead white and zinc white were usually not. Like Noble and Loon they conclude that paints that lack pigments with ionisable cations are more prone to fatty acid

migration. Other factors that affect the physical nature of the paint film - and thereby influence the chances of fatty acid blooming - are the oil absorption properties of pigments and the extent to which the pigment enhances the drying of the linseed oil. 61

2.4.2 Additives

Artists could mix a variety of substances into their oil mediums to enhance the working properties of the paint, like resins, wax, egg, starch, silica and gelatine. 62

Carlyle describes that stiffly ground colours could be thinned to a nice consistency, opaque paint could become more transparent and paints could impart a buttery quality through the addition of these materials. However, for instance resins and 63

wax provoked detrimental effects on the drying properties of the paint, as well as causing the paint to darken and discolour. 64

Noble and Loon, “Evaporation of fatty acids”, 159.

59

Ordonez, “Clarifying the haze”, 420.

60

Ibidem.

61

Leslie Carlyle, The Artist’s Assistant: Oil Paintings Instruction Manuals and Handbooks in Britain

62

1800-1900, with Reference to Selected Eighteenth-century Sources (London: Archetype

Publications Ltd, 2001), 101. Ibidem.

63

Joyce H. Townsend, Leslie Carlyle, Aviva Burnstock, Marianne Odlyha, and Jaap J. Boon,

64

“Nineteenth-Century Paint Media: The Formulation and Properties of Megilps,” In Painting

Techniques History, Materials and Studio Practice, edited by Ashok Roy and Perry Smith,

(23)

A few drops of brownish transparent binding medium were found on Van Looy’s palette, which coloured greenish in UV-light. Analysis indicated it likely to be a dammar resin mixed with oil. Boon wrote in a letter to Boitelle that these resins 65

samples could be megilps. Megilps are a type of ‘gelled mediums’, which enhance 66

the working properties of the paint. Drying oil and varnish are their main ingredients, 67

but other materials could be used for this purpose as well. 68

Driers, or siccatives, could also be added to the oil medium. Driers are chemical compounds that accelerate the polymerisation of oils through the uptake of

oxygen. They could be added during the preparation of the oil, or whilst painting. 69

Carlyle describes that compounds of lead, zinc, copper, manganese, or cobalt were most often used for this purpose. Artists could add them easily, because by the end of the nineteenth century these metal ions were available in fluid form. Already in 70

Van Looy’s own time, warnings were expressed for the usage of too much siccative, since it “renders oil saponaceous, is inimical to drying, and injurious the permanent texture of the work”. Boitelle wrote that Van Looy might have used driers, but its 71

use by Van Looy could not be confirmed with analysis. 72

2.4.5 Climatological influences

The role of moisture on the formation and migration of fatty acids is discussed in several articles. Reducing the exposure of the paint film to moisture is a key preventive measure, according to Akerlund. She subsequently mentions several 73

Townsend et al., ““Nineteenth-Century Paint Media,” 205.

65

Letter from Jaap Boon to René Boitelle, dated 22-04-1997. See Appendix III.

66

Carlyle describes that megilps were commonly used in the nineteenth century, based on many

67

recipes that she found in British artist handbooks and manuals. Carlyle, The Artist’s Assistant, 101.

68

Ibidem, 41.

69

G. Mulder, De Scheikunde der droogende oliën en haar toepssing (Rotterdam, 1865), 220-235.

70

George Field, Chromatography; or, A Treatise on Colours and Pigments, and of their Powers in

71

Painting (London, Charles Tilt: 1835), 56.

Bel et al., 53-54.

72

(24)

ways in which this can be accomplished. Noble et al. investigated a Dutch seventeenth-century painting that was kept in a micro-climate box. They 74

researched the influence of temperature and relative humidity fluctuations on the evaporation of fatty acids, which deposited on the inside of the glass and

subsequently reacted with the sodium in the glass to form sodium soaps. The climate conditions of the exterior wall on which the painting hung were monitored and high temperature and relative humidity fluctuations were found. In certain conditions this can result in a temperature gradient inside the microclimate box, which has a profound influence on the ambient relative humidity. Noble et al. 75

discuss that moisture and temperature gradients appeared to be the driving force begin the mobilisation of fatty acids. 76

2.5 Fatty acids and the drying/ageing of oil paint

To gain understanding of free fatty acids and their migration, the nature of oil paint and the complex drying processes will

shortly be explained. Drying oil consists of triglycerides, which are esters of one molecule of trihydric alcohol glycerol with three straight chain acids (fig. 2). The following five 77

fatty acids are predominantly found in drying oils: linoleic, linolenic, oleic, palmitic and stearic acids.

The drying properties of oil paints are principally determined by two of these fatty acids, namely linoleic and linolenic. Linoleic is di-unsaturated, meaning that the compound has two double bonds, whilst linolenic is tri-unsaturated with three double bonds in its structure. The presence of two

Noble, “Evaporation of fatty acids,” 152.

74

Ibidem, 156-57.

75

Ibidem, 160.

76

J. S. Mills, and R. White, The Organic Chemistry of Museum Objects (Leiden, Taylor & Francis

77

Ltd: 1987), 31.

Fig. 2: The general chemical formula of triglycerides (triester of glycerol). Source: https:// bit.ly/2KIDtqD

(25)

or more double bonds makes compounds more prone to oxidation. Through this process, oxygen is built into the 78

matrix, which creates a dried, cross-linked film (fig. 3). Since the other fatty acids only have one or no double bonds, linoleic and linolenic are principally responsible for the drying properties of oil paints.

It is unknown to what extent the monounsaturated oleic acid takes part in the polymerisation reactions, but it does gradually oxidise. The saturated palmitic and 79

stearic acids are not involved in the cross-linking process, due to their non-reactive state. The ratio 80

between these saturated fatty acids (described as P/S ratio) is commonly used for the identification of drying oils. Linseed, nut and poppy oil are examples of drying

oils. Figure 4 gives the different ratios of fatty acids in these oils. Linseed oil has the largest amount of double bonds and therefore exhibits the best drying properties. 81

P/S rations between one and two are indicative for linseed oil. 82

Mills and White, The Organic Chemistry of Museum Objects, 32.

78

Ibidem, 35.

79

Ordonez, “Clarifying the haze”, 417A.

80

Bel et al., 52.

81

Susan F. Lake, Willem de Kooning: The Artist's Materials (Los Angeles: Getty Trust

82

Publications, 2010), 73.

Fig. 3: Schematic representation of the chemistry of the drying of oil paint, through the built in of oxygen between

neighbouring chains. Source: https://en.wikipedia.org/wiki/ Drying_oil

Fig. 4: Table with the different fatty acid ratios in four types of drying oils. Source: Mills and White, The Organic Chemistry of Museum

(26)

Van den Berg et al. have designed a model of oil paint in which three processes are described that occur during the drying and ageing of oil paint. Oxidative 83

polymerisation creates a cross-linked network and causes the paint to dry. This process occurs in competition with oxidative degradation, which leads to the formation of smaller molecules which do not participate in polymerisation. The third mechanism that takes place is hydrolysis. This occurs gradually from the start of the drying process and continues on the longer term. Hydrolysis reduces the molecular weight of the binder matrix over time. Free fatty acids are separated from the triglycerides through hydrolysis. This results in a relatively flexible paint film that contains mobile components, like free fatty acids.

In the absence of a sufficient amount of polyvalent metal ions, such as lead, copper and zinc, the paint becomes more vulnerable to the extraction of low molecular weight fragments by solvents. In the presence of polyvalent ions however, a more rigid and resistant paint film is formed. This is due to the fact that metal ions can combine with the free fatty acids in the cross-linked structure. 84

Burnstock and Van den Berg, “Twentieth Century Oil Paint,” 5.

83

Ibidem, 5-6.

(27)

Chapter III: The appearance, characteristics and nature of the

white haze

This chapter will start with a description of the appearance and characteristics of the surface whitening on Kinderen. To verify the formulated hypothesis several analytical techniques have been employed. The results will be described in this chapter, as well as an interpretation of the findings.

3.1 Description of the white haze

The white haze exhibits itself as a translucent, greyish/whitish layer that covers several parts of the painting’s surface (fig. 5). Many of the reddish areas are affected, such as the girls’ hats and clothing, as well as the brownish tones in their hair and the wooden parts of the chair (see fig. 6 for a mapping of the haze). The leather-like part of the chair is partly covered with the whitening effect, although not very extensive. It is striking that the girls’ faces, as well as the golden-hued background and the bottom, darkest part of the painting are not showing any signs of hazing. In some areas the whitening follows certain coloured shapes in the painting, like the hat of the eldest girl. In

other areas the shape is much more unpredictable and shows peculiar forms, like in the same girl’s hair. It furthermore differs per area how densely and pronounced the hazing appears. It seems to create a tideline-effect in the hat of the eldest girl, where

Fig. 5: Large detail photo of Kinderen in normal light, on which the appearance of the haze is visible.

(28)

the edges are covered with a much

thicker greyish/whitish layer in comparison to the main body of the hat. This pattern repeats itself in other parts as well.

Examination with a stereomicroscope reveals the white haze to consist of many tiny crystal-like, rounded structures that migrate through the varnish. This is

especially clear in areas where the haze is not densely grown together, here the protruding ’crystals’ are individually visible (fig. 7, see Appendix II). Some are only 85

very small (0.1mm), whilst others have grown significantly bigger (0.5mm). Both sizes are smaller than Akerlund described for fatty acid blooming on unvarnished paintings, which ranged in size from 1.5 mm to 10 mm. 86

It appears like the crystals show different levels of migration through the varnish, although this aspect was very difficult to assess. Because it was not possible to investigate the level of migration through the varnish with a paint cross-section, this question could not be investigated further.

Appendix II shows the location of the microscope images on the painting.

85

86

Fig. 6: Photograph of Kinderen with a mapping of the white haze in light grey colour.

Fig. 7: Stereomicroscope photo made with normal light (the scale bar measures one millimetre per line). The image depicts the appearance of the whitening, the arrows point to the separate crystal-like shapes.

(29)

The haze seems to grow in rounded shapes, growing inwards from the outside. Ring-shaped forms, i.e. donut shapes, are here and there visible, showing whitening at the outlines with unaffected inner parts (fig. 8). In other areas, the ‘crystals’ are densely packed. Individual crystal-like shapes are not visible anymore and the whitening visually resembles something like softened ice on a brook (fig. 9A and B). In some areas the whitening appears quite opaque through the microscope, whilst in other parts it shows more translucency.

Several other characteristics of the haze were assessed with a few small tests. Speleers found in 2014 that the haze melted upon heating with a hair drier and that it did not dissolve in saliva. During examination in February 2019 it was furthermore noticed that the haze is quite soft, since it can easily be dislocated with a wooden stick. In June 2019 Speleers conducted preliminary solubility tests on the white haze. The fact that the haze is not an isolated layer, but seems to be intermingled with the varnish layer, complicated the interpretation of the results. For the tests, solvents were

chosen that could dissolve free fatty acids

Fig. 8: Stereomicroscope photo made with normal light. The arrows point to the ring-like shapes, which indicate that the rounded ‘crystals’ grow from the outside inwards.

Figs. 9A and B: Two stereomicroscope photos made with normal light (A above, B below). Both depict how the hazing appears in areas where it appears more densely grown together.

(30)

or wax, which in isolated form would be expected to dissolve in aromatic

hydrocarbons. Both the haze and varnish were not affected by white spirits with 87

18% aromatics. With the solvent xylene however, a difference was noticed between intact varnish and by the haze affected varnish. The haze, presumably consisting of fatty acids or wax, seems to have altered the solubility of the varnish. The affected varnish with haze was removed much quicker than the unaffected varnish. 88

Since it is known that fatty acids can travel in two directions, both to the front and to the back of the painting, the painting’s back was examined as well. Tiny 89

shimmering crystals were indeed visible.

3.2 Results analysis of the whitish haze

The nature of the whitish haze was investigated with two analytical techniques: Gas Chromatography Mass Spectrometry (GC/MS) and Scanning Electron Microscopy (SEM, see Appendix VII and VIII). The results of both techniques will be discussed in 90

this paragraph.

3.2.1 GC/MS

Three different samples were collected for GC/MS analysis. Small shimmering 91

crystals from the back of the canvas, as well as scrapings of both the white haze and the (upper) varnish layer without the haze visually present (see Appendix VI). This 92

last sample was taken, since it was likely that the scraping of the white haze would be contaminated with the varnish. The haze’s scraping could therefore be a mixture of materials and comparison with the results of the varnish layer without hazing could support the interpretation of the results. The samples were analysed with Thermally

87

Lidwien Speleers, Email messages to author, June 17 2019.

88

Ibidem. / Hinde, "Characterisation of Surface Whitening”,2.

89

Appendix VII gives a complete overview of the GC/MS results and Appendix IX an overview of

90

the SEM SE images.

Emilie Froment (UvA) and Lidwien Speleers (Dordrecht Museum) assisted with the sample

91

collecting for GC/MS analysis.

Appendix VI elaborates on the sample taking and sample locations.

(31)

assisted Hydrolysis and Methylation GC/MS (THM-GC/MS), combined with Pyrolysis (Py), which can give specific information concerning the nature of organic materials like oils, resins, and waxes (see Appendix V). 93

The crystals from the paintings reverse showed mainly mono carboxylic fatty acids, as well as a small amount of animal glue and some sulphur components. Smulders-de Jong, who performed the GC/MS analysis, concluSmulders-ded the following: “The P/S and A/P ratio (resp. 5.00 and 0.26) and the very small amount of glycerol are atypical for oil, but do substantiate the possibility of the presence of free fatty acids.” The 94

animal glue could well originate from either the sizing of the canvas or an animal glue containing ground layer. According to Smulders-de Jong the sulphur components are most likely from sulphur containing amino acids in the glue.

Analysis of the upper varnish revealed mainly oil, natural resins (pinaceae/pine and dammar) and very small amounts of beeswax and cyclohexanol (fig. 10). These last two components can be neglected due to their low content. The presence of natural

The GC/MS analysis was carried out by Saskia Smulders-de Jong, a Cultural Heritage Scientist

93

from the RCE. See Appendix V for more details about the specifics of the employed technique. Barbara H. Stuart, Analytical Techniques in Material Conservation (Chichester: John Wiley & Sons, 2007), 304-06.

The P/S ratio reflects the ratio between the fatty acids palmitic and stearic. The A/P ratio refers

94

to the ratio between the fatty acids azelaic and palmitic. See Chapter 2.5 for additional information on fatty acids in drying oils.

Fig. 10: The GC/MS results in piecharts of the upper varnish layer. It reveals mainly oil, pine and dammar and very small amount of beeswax and cyclohexanol.

(32)

dammar corresponds to Jos Deuss’ conservation report from 1996, which describes the appliance of a dammar varnish to diminish the effect of the whitish haze. 95

Pinaceae refers to trees from the pine family, of which its resin is referred to as colophony, rosin or violin rosin. This resin is the residuum from the distillation of 96

turpentine and is both soluble in spirit and oil. Its presence in the sample implies 97

that a pine resin varnish was applied, most likely underneath the dammar varnish. White and Kirby studied nineteenth and early twentieth century varnish compositions from paintings in the National Gallery in London. They describe that pine resin was widely used in the production of cheap varnishes. It may also have been an

adulterant of other, more expensive resins. Carlyle, who studied artists’ handbooks, 98

manuals and treatises on oil painting published in Britain between 1800 and 1900, found pine resin in several varnish recipes; for instance mixed with mastic resins or as an ingredient for turpentine varnishes. 99

The oil in the sample could originate from different sources, or perhaps of a

combination of two or more. It could for instance come from the pine resin varnish. Although pine resins were used as solved based, spirit varnishes, drying oils were regularly added “to reduce bloom or chill and to enhance their protective qualities and durability.” Phenix and Townsend describe that many natural resin, spirit 100

See Appendix III for a scan of the brief report by Jos Deuss.

95

Lance Mayer, "II. Traditional Artists' Varnishes,” Conservation Wik, published December 1995,

96

accessed May 10, 2019, https://www.conservation-wiki.com/wiki/ II._Traditional_Artists'_Varnishes.

Carlyle, The Artist’s Assistant, 78.

97

Raymond White and Jo Kirby, “A Survey of Nineteenth- and Early Twentieth-Century Varnish

98

Compositions found on a Selection of Paintings in the National Gallery Collection,” National

Gallery Technical Bulletin, vol. 22 (2001): 75.

Ibidem, 83.

99

Alan Phenix and Joyce Townsend, “A brief survey of historical varnishes,” in Conservation of

100

Easel Paintings, edited by Joyce Hill Stoner and Rebecca Rushfield (London: Routledge. 2012),

252-53. In this chapter solvent-based varnishes are describes as following: “spirit varnishes, essential oil varnishes – solutions of tree resins, as well as some other resinous materials (such as shellac), in a solvent which evaporates during drying. The spirit varnishes are based on the more volatile solvents such as alcohol (spirits of wine, aqua vitae ), naphtha (petroleum distillate, olio di sasso ) and essential oils such as those derived from turpentine or spike lavender. Spirit varnishes are less resistant to physical degradation, but are inherently more easily removable when aged since, in general, they do not form the insoluble cross-linked network characteristic of oils.”

(33)

varnishes on paintings also contain a drying oil component. An example of such a 101

recipe is given by Livache in a treatise from 1899. He includes a recipe for a linseed oil/colophony varnish, for which 100 parts of colophony are combined with 50 parts of linseed oil and 80 parts of spirit of turpentine. Another option would be that the 102

oily components were extracted from the paint layer during the appliance and drying of the varnish solutions. Tsang and Erhardt describe that the solvent evaporates from the surface of the varnish solutions, but that some also diffuses into the paint film. 103

“Solvent that diffuses into the paint film may dissolve soluble components. As this solvent eventually diffuses back into and through the varnish layer and evaporates from the surface, it may carry the soluble oil components with it.” Lastly, over time 104

fatty acids can also migrate independently through the paint layers into the varnish.

The analysis of the haze showed mainly oil, the P/S and A/P ratio (resp. 1.92 and 1.86) indicate specifically linseed oil. The relatively low glycerol signal, reflected in the quite high P/G ratio of 14.43, indicates that free fatty acids could also be present in the sample. However, the ratio between bonded and free fatty acids in this sample 105

can not be determined with GC/MS. It is furthermore striking that no resin was found, whilst it was found in the scraping of the varnish without visible hazing.

3.2.2 SEM

Scanning electron microscopes employ accelerated electrons in stead of light. When the surface of a sample is scanned with a beam of energetic electrons several signals are produced, such as backscattered electrons (BSE), secondary electrons (SE) and

Phenix and Townsend, “A brief survey of historical varnishes,” 533.

101

A. Livache, ‘The Manufacture of Varnishes, Oil Crushing, Refining and Boiling and kindred

102

Industries’, trans. J.G. Mcintosh (London, Scott Greenwood: 1899), 49-52.

Jia-Sun Tsang and David Erhardt, “Current Research on the Effects of Solvents and Gelled

103

and Aqueous Cleaning Systems on Oil Paint Films”, Journal of the American Institute for

Conservation, vol. 31, no. 1 (Spring 1992): 89.

Ibidem.

104

The P/G ratio is short for the ratio between palmitic acid and glycerol. See Chapter 2.5 for

105

(34)

X-ray fluorescence. Stuart describes these different types of signals in an 106

understandable way in the book Analytical Techniques:

As the electron beam strikes a sample, some electrons are elastically scattered by the sample atoms without a significant loss of energy. These electrons retain slightly less energy than the beam and are known as backscattered electrons. Other electrons produce an inelastic collision and may cause the sample atom to ionise. These electrons lose more energy and are known as secondary electrons. Before the backscattered electrons leave the sample they may result in inelastic collisions and generate secondary

electrons. Electrons from outer energy atomic shells fill the holes produced by the emission of secondary electrons from the inner energy shells. During this process, energy is released in the form of a characteristic X-ray. 107

The secondary electrons are very useful to depict the surface topography of a sample in high resolution, as they have a low energy and can therefore only escape from the top layer of the sample’s surface. Fatty acids tend to have a typical 108

surface morphology, often described as ‘shard-like crystals’, and can therefore often be recognised with this technique. 109

To correctly interpret SEM SE images it is essential to understand how to read them. The contrast in the signal depends on the number of secondary electrons reaching the detector, which is depending on the inclination to the beam. Although this 110

mechanism is quite complicated, it in short means that steep surfaces and edges are commonly more brightly depicted than flat surfaces, which tend to appear darker. This is profoundly different from BSE images, where heavy elements backscatter

Stuart, Analytical Techniques, 92.

106

Ibidem.

107

Ibidem, 91-93. / Aviva Burnstock and Chris Jones, “Scanning electron microscopy techniques for

108

imaging materials from paintings,” In Radiation in Art and Archeometry, eds. D.C. Creagh and D.A. Bradley (Elsevier Science: Amsterdam, 2000), 204.

109

Gilberto Artioli, Scientific Methods and Cultural Heritage: An introduction to the application of

110

(35)

electrons stronger than light elements, resulting in a brighter appearance on the image. The contrast in the SEM SE image is therefore dependent on the surface 111

morphology of the sample, whilst the SEM BSE image detects contrasts between areas of different chemical composition.

Two flakes have been collected from the painting’s haze, one appeared more densely affected than the other (see Appendix V and VI). A stereomicroscope photo in normal light was taken, as well as SEM BSE and SEM SE images. The images of the less dense crystal-like shapes can be found in figure 11, the other images are located in Appendix IX. The SEM BSE images give no indication for the presence of lead, as no

Artoli, Scientific Methods and Cultural Heritage, 67.

111

Fig. 11: Upper left: Stereomicroscope photo in normal light, of the flake from the haze. Upper right: SEM SE image of the same flake (100x magnification). Lower left: SEM BSE image of a detail from the flake (350x magnification). Lower right: SEM SE image of the same detail (350x magnification).

(36)

bright/whitish areas are visible. Together with the low melting point of the haze, this indicates the haze is not caused by lead soaps.

For the interpretation of the SE images, Aviva Burnstock was consulted. She is a specialist in the interpretation of SEM images of paintings. She recognised 112

‘hexcells’ in the images, which are rounded cells that are characteristic of a paint drying defect produced by the loss of volatile materials from the paint film. 113

The SEM SE image in figure 12, shows an example of the formation of hexcells on another nineteenth century painting. Hess describes that these hexagonal structures result from so-called ‘Vortex ring action’, which is produced and maintained by the evaporation of the volatile constituents of paints or

varnish. Keck explains this turbulence 114

in freshly applied paints as follows: “The pigment particles and the vehicle are set in motion as thinner [or solvent] from

She wrote her PhD thesis on The application of scanning electron microscopy (SEM) to the

112

examination painting materials, with special reference to cleaning and blanching.

Aviva Burnstock and Chris Jones, “Scanning electron microscopy techniques for imaging

113

materials from paintings,” In Radiation in Art and Archeometry, ed. D.C. Creagh and D.A. Bradley (Elsevier Science: Amsterdam, 2000), 220. / Aviva Burnstock, Email message to Klaas Jan van den Berg, June 17 2019.

M. Hess, Paint Film Defects: Their Causes and Cure, second edition, ed. M. Hess, W.A.

114

Edwards, T.W. Wilkinson, N.A. Bennett, G.N. Hill and W. Philips (Chapman and Hall Ltd: London, 1965), 82.

Fig. 12: SEM SE image of a nineteenth century painting, on which so-called ‘hexcells’ are depicted. This image was taken with a Hitachi S530 in high vacuum, whilst the SEM SE images of Van Looy’s painting were made in low

vacuum. Source: Burnstock and Jones,

Fig. 13: Photograph of vortex cells in enamel paint in reflected light. It shows that the centres and walls are less thickly than the area in between. Source: Keck, “Mechanical Alteration of the Paint Film”, 13.

(37)

below the surface emerges to evaporate. A centre of evaporation slightly depressed in plane is surrounded by a raised doughnut-like ring composed of pigment particles and medium moving upward and outward.” Keck furthermore describes that the 115

centres and the hexagonal walls are less thick and contain fewer pigments than the area in between (fig. 13).

It seems that alongside these hexagonal shapes, that occurred upon drying of the paint, the fatty acids could migrate more easily towards the painting’s surface. This could explain the more or less rounded or hexagonal shapes in which the blooming appears on the surface.

3.3 Discussion

Visual examination of the white haze in combination with preliminary tests to assess its characteristic and technical analysis, has given insight into the haze’s nature. The deposits are soft and easily dislocated, which points towards either wax or fatty acid blooming. The low melting point of the haze eliminates the possibility that metal soaps are present, which is further supported by the SEM BSE images that lack bright/whitish areas. The employed analytical techniques indicated that the surface whitening is likely to be caused by the migration of free fatty acids, in stead of wax - as no wax was found in the haze’s scraping.

The GC/MS analysis of the crystals on the painting’s reverse point quite clearly to the presence of free fatty acids, and the scraping on the front is likely to contain them as well. It is furthermore often encountered that fatty acids migrate both towards the front and back of paintings. The SEM SE images indicate that ‘hexcells’ formed upon drying of the paint, of which the centres and edges are thinner than the adjacent areas. It seems that the fatty acids could migrate more easily through the thin regions. It is furthermore striking that in the haze, ring-like shapes were recognised with the stereomicroscope. The formation of ‘hexcells’ could perhaps explain the

Sheldon Keck, “Mechanical Alteration of the Paint Film”, Studies in Conservation, vol. 14, no.

115

A Cloudy Appearance: An Investigation into the White Haze on Jacobus van Looy’s Kinderen