Master’s Thesis
MA Conservation – Restoration of Cultural Heritage
Book & Paper Specialisation
Adhesive-induced degradation of early
20th-century coloured paper:
A case study of photo album TM-ALB-2211 from the
Tropenmuseum Amsterdam
Student: Andrada Danila (11354046) Thesis supervisor: Bas van Velzen (UvA) Second reader: Femke Prinsen (UvA)
External advisors: Suzan de Groot (RCE) and Maarten van Bommel (UvA) Module coordinator: Maartje Stols-Witlox (UvA)
Abstract
This thesis investigated TM-ALB-2211, an early 20th-century photo album from the Dutch
East Indies currently in the Tropenmuseum collection in Amsterdam. The album’s coloured paper displays areas of local discolouration caused by the adhesive used for mounting photographs. A multidisciplinary approach was required in order to identify and understand the materials involved in the deterioration mechanism. Firstly, technical analyses were carried out on damaged and undamaged paper fibres as well as on two different types of adhesive residues. UHPLC determined that the coloured paper contained synthetic dyes of the triarylmethane and azo classes. FTIR, Py-GC/MS, and SEM-EDX found starch-based and silicon-based compounds present in different proportions in each adhesive sample, suggesting that the different residue types represent several stages in the adhesive’s degradation process. Secondly, historical research into adhesive recipes collected from sources published between 1916 and 1944 demonstrated that starch or dextrins were mixed with sodium silicate (waterglass), a highly alkaline inorganic adhesive used for manufacturing corrugated board. Starch-sodium silicate mixtures were formulated for various purposes such as adhering paper labels to metal, but not intended for use on office items or photographic material. Thirdly, a simple two-part experiment on historic coloured paper samples sourced from German and English editions of an early 20th-century paper
dyeing manual confirmed that pastes prepared with starch or dextrin and sodium silicate cause deterioration similar to that displayed by TM-ALB-2211. The research therefore identified the damaging adhesive as a mixture of starch-based compounds and sodium silicate, which was inappropriately used on the photo album. The degradation phenomenon is not fully understood but was likely caused by the strong alkalinity of the adhesive, the formation of oxidative radical species, and the high RH and temperature characteristic of the tropical climate from which the album originated. Although the object appears stable in a museum environment, conditions of high RH and temperature may reinitiate the generation and migration of reactive species particularly in those adhesive residues which retain tack and may therefore still contain sufficient sodium silicate to cause damage. A passive conservation approach focused on maintaining archival-standard conditions is therefore recommended for this object.
Samenvatting
Deze scriptie onderzocht TM-ALB-2211, een vroeg 20e-eeuws fotoalbum uit Nederlands-Indië dat momenteel deel uitmaakt van de Tropenmuseum-collectie in Amsterdam. Het gekleurde papier van het album vertoont verschillende gebieden met verkleuringen die zijn veroorzaakt door de lijm die werd gebruikt voor het inplakken van foto's. Een multidisciplinaire aanpak was vereist om de achteruitgang van de gebruikte materialen te identificeren en te begrijpen. Ten eerste werden technische analyses uitgevoerd op beschadigde en onbeschadigde papiervezels en op twee verschillende soorten lijmresten. UHPLC stelde vast dat het gekleurde papier synthetische kleurstoffen van de triarylmethaan- en azo-klassen bevatte. FTIR, Py-GC/MS en SEM-EDX vonden op zetmeel gebaseerde en op silicium gebaseerde verbindingen, in verschillende verhoudingen, in elk lijm monster, wat suggereert dat de verschillende residu-typen verschillende fasen van het afbraakproces van de kleefstof vertegenwoordigden. Ten tweede, historisch onderzoek naar recepten verzameld uit bronnen gepubliceerd tussen 1916 en 1944 toonde aan dat zetmeel of dextrines werden gemengd met natriumsilicaat (waterglas), een sterk alkalische anorganische kleefstof die wordt gebruikt voor het vervaardigen van golfkarton. Zetmeel-natriumsilicaatmengsels waren bedoeld voor verschillende doeleinden, zoals het hechten van papieren etiketten aan metaal, maar niet voor gebruik op kantoorartikelen of fotografisch materiaal. Ten derde bevestigde een eenvoudig tweedelig experiment met historisch gekleurd papier, afkomstig uit Duitse en Engelse edities van een handboek voor verven uit de vroege 20e eeuw, dat kleefstoffen bereid met zetmeel of dextrine en natriumsilicaat veroorzaakte degradatie vergelijkbaar met die gevonden in TM-ALB- 2211. Het onderzoek identificeerde daarom de schadelijke lijm als een mengsel van op zetmeel gebaseerde verbindingen en natriumsilicaat, dat niet geschikt was om op het fotoalbum te worden gebruikt. Het afbraakverschijnsel is niet volledig te achterhalen maar is waarschijnlijk veroorzaakt door de sterke alkaliteit van de kleefstof, de vorming van oxidatieve radicaalsoorten en de hoge RV en temperatuurkarakteristiek van het tropische klimaat waaruit het album afkomstig is. Hoewel het object stabiel lijkt in een museumomgeving, kunnen omstandigheden als hoge RV en temperatuur de voortbrenging en migratie van reactieve soorten opnieuw initiëren, in het bijzonder die in kleefstofresten die kleverigheid behouden en daarom nog steeds voldoende natriumsilicaat kunnen bevatten om schade te veroorzaken. Daarom wordt voor dit object een passieve behoudsbenadering aangeraden die is gericht op het aanhouden van standaardinstellingen voor archieven.
Acknowledgments
I extend my gratitude to my supervisor, Bas van Velzen (UvA), for his guidance, unwavering support, and patience through the entirety of this project.
This research would not have been possible without the help of Ingeborg Eggink (Tropenmuseum) who kindly enabled me to browse the Tropenmuseum photo album collection at length until my serendipitous discovery of photo album TM-ALB-2211. Thanks also to Cindy Zalm (Tropenmuseum), Fiona MacKinnon (Tropenmuseum), and Martijn de Ruijter (Tropenmuseum) for allowing me to closely investigate this special object.
Special thanks to Suzan de Groot (RCE) for carrying out the FTIR analyses and offering so much of her valuable time and expertise. My gratitude also goes to Maarten van Bommel (UvA) for his insightful advice and the UHPLC analyses. Thanks to Henk van Keulen (RCE) for running several Py-GC/MS analyses and patiently discussing the results. I am also thankful to Ineke Joosten (RCE) for carrying out SEM-EDX, to Maartje Stols-Witlox (UvA) for coordinating the thesis module and helping me with XRF, and to Bob Pirok (UvA) and Giacomo Moro (UvA) for their contribution to the advanced UHPLC analyses.
A warm word of thanks goes to René Peschar (UvA) for his accessibility and useful suggestions. I also offer my gratitude to my tutors, Elizabet Nijhoff-Asser (UvA) and Femke Prinsen (UvA), for their instruction and tremendous support over the past two years. Thanks to Ella Hendriks (UvA) for her consultations.
My research was enriched by Klaas Jan van den Berg (RCE) who granted me the opportunity to reach out to the adhesive industry. Thanks to Coos van Waas (Royal Talens) for offering insight into the formulation and production of Gluton.
I am very grateful to Birgit Reissland (RCE) and Han Neevel (RCE) for their interest in this project and for their many helpful ideas.
Many thanks to Clara von Waldthausen (UvA) for improving my understanding of photographic techniques and to Martin Jürgens (Rijksmuseum) for his expert advice. I am thankful to Edy Seriese (IWI), Liane van der Linden (IWI), and Pamela Pattynama (IWI) for sharing their knowledge on the IWI photo album collection.
My appreciation also goes to Karin Scheper (University Library Leiden) for her suggestions and help in the early stages of this project.
Last but not least, thanks to Leanne de Wit for translating my abstract and to all my classmates in the Conservation – Restoration of Cultural Heritage department for providing good company throughout this process. Special thanks to my parents and family for encouraging all my adventures. Immense gratitude also goes to Pablo, my partner in good and bad times. Thanks for sticking around, I could not have done this without you.
Table of Contents
Abstract ...
1Samenvatting ...
2Acknowledgments ...
3Introduction ...
61. Object description and condition assessment...
91.1. Overview ... 9
1.2. The coloured paper ... 11
1.3. The adhesives ... 15
1.4. The photographs... 18
1.5. The transparent interleaves ... 23
1.6. Provenance and dating ... 24
Conclusion ... 29
2. Technical investigation ...
302.1. The coloured paper ... 30
2.2. The adhesives ... 40
Conclusion ... 47
3. Technological source research ...
493.1. The coloured paper ... 49
3.2. The adhesives ... 54 Conclusion ... 62
4. Experimental investigation ...
64 4.1. Hypotheses... 64 4.2. Experiment part 1 ... 65 4.3. Experiment part 2 ... 72 Conclusion ... 805. Results and discussion ...
815.1. The deterioration event ... 81
5.2. The deterioration mechanism ... 81
5.3. Future deterioration & recommendations for care ... 84
Bibliography ...
86Appendices ...
93Appendix 1. Object documentation ... 93
Appendix 2. Full condition assessment of photographs in TM-ALB-2211 ... 126
Appendix 3. Visual examination ... 150
Appendix 4. Spots tests and pH measurements ... 151
Appendix 5. Ultra-High Performance Liquid Chromatography (UHPLC) analyses .... 153
Appendix 6. Fourier-transform Infrared Spectroscopy (FTIR) analyses ... 180
Appendix 7. Pyrolysis-gas chromatography-mass spectrometry (Py-GC/MS) analyses ... 185
Appendix 8. Scanning electron microscopy coupled with energy-dispersive X-ray spectroscopy (SEM-EDX) analyses... 189
Appendix 9. Experimental investigation ... 216
Appendix 10. Historic starch-based adhesive recipes (1916-1944) ... 236
Abbreviations
CWP chemical wood pulp POP printing-out photographic print
DOP developing-out photographic print RCE Rijksdienst voor het Cultureel Erfgoed IWI Indisch Wetenschappelijk Instituut RH relative humidity
MWP mechanical wood pulp UV ultra-violet
OBA optical brightening agent UvA University of Amsterdam
Notes
All references to geographical and topographical names were reproduced with the spelling given by their respective historic sources rather than the forms commonly used today (i.e., Bandoeng instead of Bandung or Batavia instead of Jakarta).
All photographs and illustrations included were taken or created by the author, unless otherwise stated.
This thesis contains 17748 words, including abstracts and footnotes, excluding appendices, image captions, and tables.
Introduction
Research focus
An early 20th-century photo album with the inventory number TM-ALB-2211 from the Tropenmuseum collection in Amsterdam, the Netherlands displays an unusual degradation phenomenon that has resulted in local discolouration of the coloured paper pages as well as stains on the photographs and on the transparent paper interleaves. This deterioration appears to be related to the adhesive(s) used to mount the photographs on the album pages.
In order to design a preservation approach suited to this object and to determine whether any intervention may prove beneficial, it is necessary to establish how the coloured paper and photographs in the album arrived to their present condition and whether they will continue to degrade in this manner. Moreover, to assess the potential risks posed by adhesives to coloured papers which may currently appear unaffected, it would be useful to gain a clearer understanding of the deterioration process and of the involved agents – early 20th-century
adhesives, synthetic colourants, and paper. Although the photographic prints and the transparent paper will also be visually examined in search for clues on the identity of the damaging adhesive, they will not undergo technical analysis and their deterioration will not be discussed in detail. It is hoped that future projects will undertake this objective.
The goal of this research is therefore to investigate the materials present in the photo album, with particular focus on the coloured paper and the adhesives, including their chemical structure, their interaction with external factors, their manufacture, as well as their commercial availability in the object’s place of origin, the Dutch East Indies. This research project will then aim to identify the cause for the degradation exhibited by the paper, to describe its mechanism, and to find if certain climatic conditions may catalyse the process. Lastly, this thesis hopes to determine whether the object’s condition is stable and to understand why comparable adhesive-induced discolouration does not affect other albums from the Tropenmuseum.1
Relevance to the conservation field
Although no other examples of the phenomenon displayed by TM-ALB-2211 have been found so far,2 coloured paper is commonly encountered in many early 20th-century photo
1 A survey conducted by Elizabet Nijhoff-Asser in 2009 was used to determine whether the other photo albums
in the Tropenmuseum collection display adhesive-induced discolouration. The results of this survey, which assessed 2177 objects out of a total of 3272, were searched using the keywords “verkleuring,” “lijm,” “fotolijm,” “adhesief,” and “vlek.” The artefacts whose damage descriptions contained any of these words were visually examined. No object with damage similar to the studied phenomenon could be found.
2 In a preliminary attempt to locate objects affected by the studied phenomenon in other collections, specialists
from several institutions in the Netherlands, including the Rijksmuseum, University Library Leiden, Nederlands Fotomuseum Rotterdam, Nationaal Museum voor Wereldculturen, the Eye Museum, and Het Scheepvaartmuseum, were shown images of TM-ALB-2211. None of the consulted specialists had encountered
albums as well as in other objects with cultural and historical value. For instance, coloured paper has been used as support for drawings, as writing paper for official documents, or as endleaves in books.
The fading of synthetic dyes /under the influence of light is well-known and documented.3
In contrast, information on the deterioration of coloured papers in contact with other substances is scarce. Much of the limited literature on paper dyeing belongs to technical publications of the paper industry and does not address the topic of degradation. The few recent studies on coloured paper published within the conservation field4 are largely
concerned with the technical identification of synthetic dyes. Two sources briefly mention that exposure to light, alkalis, alcohol, and aqueous treatments may damage coloured papers but do not explore the subject any further.5 Conservators and collection specialists in charge
of coloured paper objects would undoubtedly benefit from more knowledge on the sensitivities of such artefacts. This thesis aims to collect preliminary findings on this under-researched topic that will hopefully be explored by more extensive studies in the future.
Methodology and structure
To understand the deterioration mechanism affecting the coloured paper in photo album TM-ALB-2211, this research project will rely on a combination of technical analyses, primary source research, and an experiment on historic samples.
Chapter one will provide a detailed condition assessment of the photo album, based on visual examination with visible light, UV light, and reflected light stereo microscopy. It will also investigate the provenance of the object. This information will help to date the album and hopefully offer some insight into the conditions it has been exposed to during its lifetime. Chapter two will present the results of the technical analyses which identified the materials present in the object and directed the subsequent stages of the research.
Chapter three will explore the historical context of the Dutch East Indies of the 1920s-1930s in which the photo album was likely compiled, with particular focus on the commercial and industrial landscape. Technical literature will be consulted in order to investigate the manufacture of the materials used in photo albums. Primary sources such as advertisements,
similar objects. In early January 2018, an inquiry was submitted to the Conservation Distribution List which currently has around 11,000 subscribers worldwide. The post was distributed as part of issue 2, volume 32. No answers were received.
3 The photodegradation of dyestuffs has long been an issue for manufacturers, even before the introduction of
synthetic dyes in the late 19th-century. Methods for quantitively understanding light sensitivity have indeed
been under development for centuries. For a history of lightfastness standards, see Giles and McKay 1963: 528-37. In Erfurt’s paper colouring manual from 1901, lightfastness is one of the most frequently mentioned properties of the described dyestuffs, which demonstrates its importance in the assessment of paper. Most of the abundant literature on the fading of dyes relates to the textile industry. Some recent research is also concerned with the photodegradation of dyestuffs used in colour photography and in the composition of inks.
4 The most recent articles found are Casadio et al 2010: 885-899 and Montagner et al 2011: 1669-1680. See
catalogues, and newspapers articles will provide insight into the distribution of these goods within the Dutch East Indies.
Chapter four will describe the experimental investigation carried out on historic samples of paper coloured with dyes similar to those found in TM-ALB-2211.
Finally, chapter five will discuss the paper degradation process displayed by the case study object. It will also provide ideas for further research and conclude with brief recommendations for care.
1. Object description and condition assessment
1.1. Overview
Photo album TM-ALB-22116 consists of 24 sheets (48 pages) of machine-made wove paper
that was given a dark, olive green-grey colour using synthetic dyes added to the pulp in the wet-end of the manufacturing process. The album also contains transparent paper interleaves probably produced using intensively beaten chemical wood pulp. The album is a cord binding, a loose-leaf mechanism of great popularity after the First World War (fig. 1.1). This type of binding most frequently contained dark-coloured multiply papers7 as exemplified by
TM-ALB-2211.
164 photographic prints have been pasted directly onto the pages using at least two adhesives, as demonstrated by the differences in the appearance of the solid residues present throughout the album as well as by their impact on the coloured paper (figs. 1.2-1.3). Two of the several loose photographs have been remounted using triangular corners with
6 See Appendix 1 for inventory information and full photographic documentation.
transparent acetate facings. Captions handwritten in blue ink directly onto the coloured paper describe nine photographs. The album pages are numbered in white crayon.
The photographs were produced on fibre-based baryta-coated supports using two main photographic techniques, gelatin printing-out (gelatin POP) and gelatin silver developing-out (gelatin silver DOP), with several variations in toning and surface treatment.8
8 The photographic techniques were identified based on consultation with Clara von Waldthausen and the
guidelines in Reilly 1986.
Figure 1.2. Album pages 2-3 with gelatin POPs, gelatin silver DOPs, and transparent interleaf (no magnification, visible light)
Figure 1.3. Album pages 24-25 with gelatin POPs, gelatin silver DOPs, and transparent interleaf (no magnification, visible light)
1.2 . The coloured paper
1.2.1. Visual examination in visible light
The first 22 coloured leaves of the album display damage seemingly caused by the interaction between the colourants in the paper and the adhesive used to mount the photographs. The deteriorated paper is disfigured by circular areas of local discolouration which appear yellow towards the centre and fade to reddish-brown towards the edge (fig. 1.4). The transition from the yellow centre to the reddish edge is gradual rather than distinctively linear, suggesting that the deterioration process was catalysed by a gaseous compound rather than an aqueous solution, which would have resulted in tideline formation. Each area of discolouration is visible on both sides of the paper and, in a few instances, also on adjacent pages (fig. 1.5), which indicates that the damaging reactants could migrate through several paper layers.
The last 25 pages of the album do not exhibit any mountant-induced discolouration and therefore demonstrate the use of a different, non-damaging adhesive (fig. 1.6).
Figure 1.4. Album page 1 showing discolouration induced by adhesive applied on recto (R) and verso (V) sides (no magnification, visible light)
V
V
Figure 1.5. Album page 23 showing adhesive-induced discolouration that migrated from page 22 through photographic layers and a transparent interleaf (no magnification, visible light)
1.2.2. Visual examination in UV light
The coloured paper leaves appear dark in UV light (365nm). The damaged areas do not show significant fluorescence (fig. 1.7).
1.2.3. Stereo microscopy in reflected light
Microscopic examination of the paper in both reflected and raking light shows a lack of differences in the texture, size, and appearance of the fibres in the discoloured and undamaged areas (figs. 1.8-1.9).9
9 Visual examination of the object via stereo microscopy was carried out by the student. See Appendix 3 for
Figure 1.8. Discoloured paper area on album
page 7 (25x magnification, reflected light) Figure 1.9. Discoloured paper area on album page 7 (50x magnification, raking light) Figure 1.7. Album page 7 (no magnification, UV light - 365 nm)
Microscopic examination of the handwritten captions supports the identification of the used writing instrument as a ballpoint pen,
based on the tubular, even shape of the depressions caused by the application of the blue ink onto the paper surface (fig. 1.10). Since ballpoint pens came into use in the 1940s,10 the captions
were likely added after TM-ALB-2211 arrived to the Netherlands.11
10 Although patents for ballpoint pens were filed as early as the 19th century, the first commercially viable type
was introduced by László Bíró in 1938. See Patent GB498997A, filed on April 25, 1938.
11 See Section 1.6 for TM-ALB-2211’s biography.
Figure 1.10. Album page 23 with ballpoint pen caption (no magnification and 40x magnification, visible light)
1.3. The adhesives
1.3.1. Visual examination in visible light
The first half of the album contains white solid residues deposited on the paper surface. Given their location directly corresponding to areas of discolouration and underneath loose photographs, these residues are most likely components of the adhesive which induced the studied degradation phenomenon. They may also
be products of the reaction between the adhesive and the paper colourants.
Many of the photographs mounted on the first ten pages have become loose or completely detached, demonstrating the adhesive’s decreased performance. Two photos12 have been
reattached using triangular photo corners with transparent acetate facings (fig. 1.11). In the second half of the album, much fewer residues may be observed, particularly because the photographs are securely attached, indicating the higher adhesive strength or better condition of the mountant applied on these pages compared to that used in the damaged first half.
1.3.2. Visual examination in UV light
None of the adhesive residues fluoresces in UV light (365 nm). The photo corners that support two of the loose photographs display very bright white fluorescence (fig. 1.12),
12 Album pages 6 and 10.
Figure 1.11. Photo corner on album page 6 (6x magnification, reflected light)
Figure 1.12. Album page 10 with highly fluorescent photo corners and non-fluorescent adhesive residues (no magnification, UV light – 365 nm)
suggesting that their paper base contains optical brighteners (OBAs)13 and therefore that they
were most likely added to the object after it arrived from Indonesia to the Netherlands.14
1.3.3. Stereo microscopy with reflected light
Microscopic examination of the solid residues on the discoloured pages reveals that the damaging adhesive can undertake at least three forms in its current conditions:
1) a matte white solid that is opaque and prone to glass-like cracks (fig. 1.13); 2) a glossy transparent solid thinly deposited, partially absorbed (fig. 1.14);
3) a semi-glossy yellowed solid that seems more viscous than the first type (fig. 1.15).
13 OBAs came into regular use in the paper industry in the 1950s. Mustalish 2013: 133-136. 14 See Section 1.6 for TM-ALB-2211’s biography.
Figure 1.14. Adhesive residue type 2 on album page 1 (20x magnification, reflected light) Figure 1.13. Adhesive residue type 1 on album page 10 (8x magnification, reflected light)
The residue types may represent phases in an ageing process whereby the third is the intermediate in the transition from the second to the first forms. Types 2 and 3 are found much less frequently than type 1, perhaps due to a higher adhesive content which keeps the adjacent photographs still securely attached, thereby obscuring the solid residues. A low amount of adhesive in residues of type 1 would explain their presence particularly underneath loose photographs. As the glue deteriorated, it lost its adhesive properties and deposited its non-tacky ingredients on the paper surface. Visual investigation alone is not sufficient to understand the mechanism behind this hypothetical phase-separation.
The presence of distinctive residue types may suggest that two or several different adhesives have been applied on the discoloured pages, perhaps at different points in the object’s lifetime. However, the fact that all the adhesive residues precisely correspond in location and shape to areas of paper discolouration refutes this idea.
Figure 1.15. Adhesive residue type 3 on album page 2 (40x magnification, reflected light)
The very few residues found in the undamaged half of the album do not resemble those present in the damaged half, but appear thin, glossy, transparent, and slightly yellowed under microscopic examination (fig. 1.16). This observation reinforces the idea that a different adhesive was used in this part of the album.
1.4. The photographs
1.4.1. Visual examination in visible light
Although understanding the degradation of the photographs in TM-ALB-2211 is beyond the scope of this project, their condition may offer some hints regarding the identity of the reactants involved in the paper discolouration process.
All 164 photographs in TM-ALB-2211 were individually assessed in order to establish the production method and to record damages related to the studied phenomenon. The assessment revealed that photographs mounted on the first 22 album pages (80 of 164) have suffered alterations in areas of contact with the adhesive.15 These photographs manifest
distinctive local damages, including stains of various shades and texture changes. Some photographs display reduced deterioration compared to the rest of their surface (table 1.1).
Table 1.1. Some of the effects of the damaging adhesive
Effect on photo Type of photo Image
Silver mirroring Gelatin POP
Yellow stains Gelatin silver DOP
15 One exception is the gelatin silver DOP at the top-right of page 1. The photograph was adhered with a
non-damaging adhesive at points that do not coincide with the areas of paper discolouration and solid residues, which suggests it may have been remounted at a later date.
Effect on photo Type of photo Image
Light blue stains Gelatin silver DOP
Brown stains/reduced
fading Gelatin POP
Reduced photograph
deterioration Gelatin silver DOP
Texture change on
To structure the visual information in a digestible format, the following “stain areas” were established as criteria for consistently describing the condition of each photograph (fig. 1.17):
• Stain area 1 – area of direct contact with adhesive residues
• Stain area 2 – area of former contact with the unaged, functional adhesive
• Stain area 3 – area of contact with the adhesive “environment,” or the conditions created by the presence of the adhesive, for instance local pH or moisture vapour
The observations were compiled in an Excel workbook.16 Each of the stain areas was
assigned one colour per item. Small tonal differences were not taken into account. In some cases, key words for specific local effects such as texture changes or silver mirroring were also recorded. The data was processed and visually represented using charts (table 1.2). The assessment results suggest that the majority of gelatin silver DOPs seems minorly affected by contact with the adhesive and its “environment.” In contrast, most of the gelatin POPs have undergone significant local changes. One explanation for this difference is the higher sensitivity of POPs to oxidative-reductive degradation, the most important process that can affect silver-based photographs.17 It involves a series of moisture-catalysed
reactions whereby metallic silver oxidizes to unstable silver ions only to be reduced back to metallic silver in a different location. This migration through the gelatin layer, enabled by high relative humidity, is responsible for damage such as silver mirroring.
The variety of effects within the POP category could perhaps be attributed to the use of toners, which, depending on the substances used, either change metallic silver into silver halides or replace it with other metal ions resulting in different degrees of image stability.18
The adhesive may have been a source of halides, which would explain the effect of locally reduced degradation frequently seen in DOPs. The halides would react with the metallic silver available in DOPs, thereby protecting areas of direct contact with the solid adhesive either from its own damaging “environment” or from the other aging processes taking place in the photograph.19
16 See Appendix 2.
17 While POPs contain very small discrete “photolytic” silver particles, DOPs consist of high-density clumps
of “filamentary” silver, which are more resistant to sulfur-attack and oxidation. Reilly 1986: 14-22 and Lavédrine 2009: 144.
18 Lavédrine 2009: 136 and Reilly 1986: 22-24. 19 Hendriks 2010: 234.
Figure 1.17. Schematic representation of stain areas used as guidelines for the photograph condition assessment
Table 1.2. Overview of the distribution of adhesive-induced changes on the affected photographs
Gelatin silver DOPs Gelatin POPs
Area 1
Area 2
In addition to the visual effects described by table 1.2, textural changes observed in many of the glossy photographs indicate that the adhesive was sufficiently moist to swell the gelatin in the image layer.20
1.4.2. Visual examination in UV light
In UV light (365 nm), the great majority of both gelatin POPs and gelatin silver DOPs appear very brightly fluorescent in stain area 2 and mildly fluorescent in stain area 3. Around half of the damaged photographs additionally display stains that fluoresce orange-yellow in areas where their verso had direct contact with the solid residues (fig. 1.18).
20 Reilly 2010: 43.
V
R
Figure 1.18. Album page 1 showing fluorescence induced by adhesive applied on recto (R) and verso (V) sides (no magnification, UV light - 365 nm)
1.5. The transparent interleaves
1.5.1. Visual examination in visible light
The transparent interleaves adjacent to the damaged paper are locally discoloured and cockled around the areas of contact with the coloured paper and with the adhesive “environment” (fig. 1.19). This damage suggests that conditions of high moisture or RH were involved in the deterioration process.
1.5.2. Visual examination in UV light
The areas of damage on all of the transparent interleaves adjacent to the damaged paper appear brightly fluorescent in UV light (365 nm) (fig. 1.20). The only exception is the first interleaf which inexplicably does not display any deterioration.
Figure 1.19. Transparent interleaf adjacent to album page 7 (no magnification, visible light)
1.6. Provenance and dating
TM-ALB-2211 belongs to the Insular South-East Asian collection of the Tropenmuseum in Amsterdam, which is part of the Nationaal Museum voor Wereldculturen. The object was acquired in 2006 from the Indisch Wetenschappelijk Instituut (IWI),21 which donated its
collection of 567 photo albums to the museum after digitization.22
It is difficult to establish the provenance of the object before the accession date due to the complicated formation of the IWI institute and the lack of proper documentation of its holdings. The Tropenmuseum has dated the album to ca. 1927 for unclear reasons. The only clues on the album’s life between 1927 and 2006 are provided by the object itself. Ballpoint pen captions added after the 1940s23 identify the following individuals: Mebius Hendrikus
Rooswinkel, Frau Gretel, and Mimi, John, and Rolf Hulstein. Using these names to search through primary source databases, it was possible to piece together a fascinating, but partial picture on the album’s owners, their occupation, travels, and living situation. The aim of the provenance research was not just to contextualize the origins of the artefact, but also to date the photographs and by extrapolation the mountant adhesives. While the information gathered was not sufficient to precisely determine the year when the album was compiled, it proved that some images may have been taken after 1929 and others as early as 1919.
1.6.1. The photo album in the Dutch East Indies
M.H. Rooswinkel (fig. 1.21) was a Dutch national born on May 31st 1886 in Leeuwarden in
the Friesland province. According to the municipal records of Leeuwarden,24 Rooswinkel
moved to Hamburg at an unrecorded date and was/came in possession of a foreign passport on August 16th 1919. He was already in the Dutch East Indies by April 1915, when his name
was included in a list of passengers departing Batavia towards the Netherlands on the ship SS Rembrandt. In April 1919, he traveled on SS Jan Pieterszoon,25 one of the two
identifiable ships featured in TM-ALB-2211, with the direction Surabaya. He undertook several journeys again in the following year, including a departure to Amsterdam on October 26th 1920 on SS Johan de Witt,26 the other of the two identifiable ships shown in
TM-ALB-2211. A newspaper announcement from 1921 records that Rooswinkel traveled to and from Amsterdam accompanied by a spouse. In the years 1923 and 1924, M.H. Rooswinkel appears on several passenger lists traveling within the South-East Asian region, including
21 The accession information of album TM-ALB-2211 was provided by Ingeborg Eggink (Tropenmuseum). 22 The images are available at http://www.iwicollectie.nl/iwicollectie.aspx (accessed February 26, 2018). 23 See Section 1.2.3.
24 See the digitized copy of the Geboorteregister 1866 in Gemeentebestuur van Leeuwarden 1811-1941, deel:
3435, archief 1002, inventaris-nummer 3435, aktenummer 368 available at
https://www.openarch.nl/show.php?archive=frl&identifier=5433ea26-c4ae-e8c8-cb30-9d0e6c200eac&lang=nl&six=3 (accessed February 26, 2018) and the digitized copy of the Leeuwarden Bevolkingsregister 1904 - 1922, deel: 4897, archief 1002, inventaris-nummer 4897, folio 1133 at
https://www.openarch.nl/show.php?archive=frl&identifier=ef071f19-70f5-ffa4-93fc-79841ddc0243&lang=nl&six=2 (accessed February 26, 2018).
25 Album pages 1, 3, and 5. See Appendix 2 for a full photographic documentation of TM-ALB-2211. 26 Album page 6.
an eight-day trip from Java to Sumatra in November 1924. It is possible but not certain that thirteen photographs in TM-ALB-2211 which depict a Hotel Brastagi27 likely located near
Berastagi, a village in Sumatra, document this voyage.
By that time, M.H. Rooswinkel had already permanently settled in Java. Advertisements published in a Batavian newspaper every Saturday from late August 1924 to January 1925 indicate that Rooswinkel had a commercial space at Rijswijk 7a in Batavia, where he offered various goods imported from Hamburg. Every Wednesday from early February until late July 1925, the same Batavian newspaper published Rooswinkel’s ads for “Elly” raincoats, tested and approved for life in the tropics (fig. 1.22). Very little information on Rooswinkel’s professional activity between 1925 and 1939 could be found. At an unknown date in the late 1920s or early 1930s, he became manager of N. V. Ingenieurs- & Handelsbureau, a business importing furniture and mattresses at Dagoweg 119 in Bandoeng, Java.28
27 Album pages 16-18.
28 Information recorded in the digitized copy of a telephone book from Bandung, dated January 1936 and
Figure 1.21. Gelatin POP on album page 8 captioned "Mebius Hendrikus Rooswinkel met Frau Gretel" in blue ball-point pen
No information on Frau Gretel or the Hulstein family was found. Of the three Hulsteins identified by ball-point pen captions, Mimi is the most frequently photographed, particularly in images mounted in the second half of the album, where the lack of damage suggests a change in the adhesive used. The different choice for mountant along with Mimi Hulstein’s prominent presence in the photos in this section may well be interpreted as evidence for a shift in the album’s ownership from the Rooswinkels to the Hulsteins. However, the photographs pasted in the second half of the album, including a candid portrait of M.H. Rooswinkel and Mimi Hulstein,29 show a
sustained connection between the families. According to ball-point
pen captions, three images on album pp. 23-24 show the Rooswinkel residence in Bandoeng (fig. 1.23). Since these photographs, added halfway through the album and precisely coincident with the change in adhesive use, were likely taken after Rooswinkel had already relocated from Batavia to Bandoeng, they suggest that all succeeding images date from the late 1920s-early 1930s or after.
However, other photographs complicate this timeline. For instance, two prints of the same image of a storefront bearing Rooswinkel’s name (fig. 1.24) were mounted at different locations in the album: the first on page 23, the same page that shows Rooswinkel’s house in Bandoeng; the second print on page 34, amidst several other family snapshots portraying
29 Album page 35.
Figure 1.22. Advertisement for the "Elly" raincoats sold engros by M. Rooswinkel. Source: “Advertentie,” Het nieuws van den dag voor
Nederlandsch-Indië (Batavia), April 14, 1925.
Figure 1.23. Gelatin POP of the Rooswinkel residence in Bandoeng on album page 24 captioned “Fam. Rooswinkel en Bedienden” (eng.
the Hulsteins, and the gravestone of a J. H. Oscar Lüders (d. March 1, 1929). It is possible that the store was Rooswinkel’s business in Bandoeng from the 1930s period. However, it is unclear why two identical photographic prints were included at disparate moments in the album’s timeline.
Another photograph that raises further questions depicts a storefront through which a poster advertising “Elly” raincoats may be distinguished (fig. 1.25), most likely the shop managed by Rooswinkel in Batavia during the mid-1920s. Although this image should expectedly appear before photographs from the 1930s, it is in fact the very last in the album.
Figure 1.25. Gelatin silver DOP on album page 44 Figure 1.24. Gelatin silver DOP on album page 23
The discrepancies between the pieces of information gathered from archival sources and those offered by the object itself suggest that the photographs were not mounted in a strictly chronological order. Given the use of two different adhesives, the photographs may have also been added in bulk in several sessions, rather than gradually and only a few at a time. Moreover, even though many of the images could be grouped based on their subject matter as well as stylistic and technical criteria into sections that seemingly correspond to trips, leisure activities, or social occasions, the relative randomness30 otherwise apparent
throughout the album makes the precise dating of the photographs a very difficult, if not impossible task.
1.6.2. The photo album after the Second World War
The album was probably brought to the Netherlands in 1948-1949 by the Ministry of Overseas Territories within international negotiations to retrieve the belongings of Dutch nationals taken as prisoners during the Second World War.31 The Indies Institute’s Press and
Publicity Department attempted to locate the owners of the returned objects until 1959.32
When these efforts ceased, Marie Jeanne Hillerström, who was working with the Department of Foreign Affairs at that time, took the photo albums and continued to search for their families until her death in 1978.
The albums then passed on to Lilian Ducille who donated some to the Tropenmuseum within the two following years. The rest were kept in the office of the Tong Tong magazine (nowadays the Moesson magazine) until 1985 when they were given to the newly-founded IWI institute and collection,33 headquartered at the same address in The Hague. As the IWI
collection grew during the second half of the 1980s, the institute called volunteers to catalogue, organise, glue loose photos, and “take care”34 of the albums. The ballpoint pen
captions (fig. 1.10) and the transparent-facing, OBA-containing photo corners (fig. 1.11) in TM-ALB-2211 may serve as evidence that such interventions took place at that stage. IWI ran into financial difficulties in the 1990s. Consequently, the photo albums were moved to two different locations before they were given to the Tropenmuseum in 2006 where they remain to this date. It is possible but not certain that TM-ALB-2211 followed the trajectory illustrated below (fig. 1.26).
30 Mounting photographs in a non-chronological order was not uncommon at that time. Many albums in the
Tropenmuseum collection were in fact used as scrapbooks, with postcards, newspaper clippings, letters, envelopes, and other memorabilia randomly pasted, occasionally even on the interleaves. The Dutch living in the colonies would also sometimes mount photographs of their relatives in Europe amidst their own snapshots to strengthen and remember family ties. See Dijk et al 2012: 26.
31 The early history of the Tropenmuseum photo album collection is presented in Dijk et al 2012: 52-59. 32 613 albums and thousands of photographs were returned to approximately 2000 owners.
33 The history of the IWI and its collection is described in Seriese 2011: 1-91. More information was obtained
via email communication with persons involved in the IWI management: Edy Seriese (February 21, 2018), Liane van der Linden (February 20, 2018), and Pamela Pattynama (February 25, 2018).
Figure 1.26. Proposed provenance timeline for TM-ALB-2211 (not drawn to scale)
This provenance research reveals that the case study object has traveled from a tropical to an oceanic temperate climate and has been kept in circumstances presently not considered appropriate for paper and photographic artefacts of cultural significance. The extreme and fluctuating environmental conditions the object has been exposed to may have acted as catalysts for deterioration.
Conclusion
Detailed visual examination of the object has shown that two adhesives were used to mount the photographs in TM-ALB-2211, one of which affected the coloured paper. The different types of solid residues visible on the discoloured paper could represent the damaging adhesive’s several stages of aging/deterioration.
The damaging adhesive either contained or produced oxidizing and reducing agents which attacked the album’s silver-based photographs. The adhesive-induced deterioration mechanism resulted in significant enhancement of the photographs’ fluorescence. Some fluorescent products have also transferred to the transparent interleaves. Conditions of high RH as well as moisture introduced by the adhesive probably enabled the migration of deterioration agents through several paper and photographic layers.
Provenance research suggests that TM-ALB-2211 was compiled within the approximative range of 1920-1939. Interventions into the album, including attaching and/or reattaching loose photos, may have taken place at any point from the 1920s all through the early 2000s.
2. Technical investigation
Visual examination and provenance research alone could not explain the discolouration mechanism affecting the paper in TM-ALB-2211. Technical analysis was necessary to identify the materials present in the case study object.
2.1. The coloured paper
2.1.1. Fibre composition analysis via optical microscopy
Fibres from undamaged and discoloured areas of the paper were analysed via optical microscopy with lambda-retarded polarized light.35 To sample the undamaged paper, fibres
were tweezed from an inconspicuous interior page edge. To sample the discoloured paper, fibres were tweezed from an area on page 2 where a presently missing photograph had skinned the paper surface while coming loose (fig. 2.1).
Figure 2.1. Sampling location for the discoloured paper on album page 2 (8x magnification in raking light)
The analysis found a mixture of mechanical and chemical wood pulp of coniferous source (figs. 2.2-2.3), with no detectable differences between the undamaged and damaged samples.
35 The fibre analysis was carried out by the student and Bas van Velzen (UvA). See Appendix 3 for the
Figure 2.2. Mechanical wood pulp of coniferous source (50x magnification in polarized transmitted light with lambda retarder)
Figure 2.3. Chemical wood pulp of coniferous source (100x magnification in polarized transmitted light with lambda retarder)
2.1.2. UHPLC analysis of the paper colourants
Ultra-High-Performance Liquid Chromatography coupled to Photo-Diode Array detection (UHPLC-PDA)36 identified the colourants present in the paper. To sample the undamaged
paper, fibres were taken from an interior page corner. To sample the discoloured paper, fibres were tweezed from the skinned area on page 2 that was also used for fibre analysis (fig. 2.1). Two analyses were carried out, the second using mass spectrometry in addition to PDA (UHPLC-PDA-MS).
The first analysis found the following synthetic dyes in the undamaged paper sample: • Metanil Yellow (C.I. 13065)
• Diamond Green B (C.I. 42000) • Fuchsine (C.I. 42510)
• Ponceau G (C.I. 16150) – trace amounts
The same dyestuffs were identified in the discoloured paper sample. However, Diamond Green B and Fuchsine were found in significantly lower concentrations, which suggests that they partially decomposed in their interaction with the adhesive while Metanil Yellow remained more or less intact.
The second analysis identified the following synthetic dyes: • Metanil Yellow (C.I. 13065)
• Diamond Green B (C.I. 42000) and Diamond Green G (C.I. 42040) • Methyl Violet (C.I. 42535)
• Crocein Orange G (C.I. 15970) – trace amounts
Both the undamaged and damaged samples contained Metanil Yellow in similar amounts, but Diamond Green and Methyl Violet dyestuffs were present in much lower concentrations in the latter compared to the former.
One important discrepancy between the two analyses is the identification of Fuchsine vs. Methyl Violet/Crystal Violet. The structural similarities between these dyestuffs, which belong to the same class, complicates the results: Crystal Violet is the fully methylated version of the triaminotriarylmethane structure common to all known forms of Methyl Violet, while Fuchsine has a methyl group attached directly onto one of the aromatic rings (table 2.1). The paper dyes were also possibly impure and may have contained several analogous dyestuffs in low concentration in addition to the main component, which could explain the coexistence of closely related compounds.
The interpretation is further hindered by the presence of various demethylated and de-ethylated products, most of which were only found in the undamaged paper and thereby
36 UHPLC was carried out by Maarten van Bommel (UvA), Bob Pirok (UvA), and Giacomo Moro (UvA). See
likely resulted from processes other than the adhesive-induced deterioration (for instance, photooxidation). This indicates that the studied phenomenon caused molecular breakdown more advanced than demethylation/de-ethylation.
The two analytical results are otherwise almost entirely consistent, demonstrating that the paper contained mainly dyes of the triarylmethane and azo classes (table 2.1.), all of which were affected by the adhesive-induced deterioration mechanism apart from Metanil Yellow.
Table 2.1. Main synthetic dyes found in TM-ALB-2211 (Images source: Maarten van Bommel and Sigma-Aldrich)
Common name
Other
names Class Molecular structure
Metanil Yellow
Acid Yellow
36
Anionic monoazo acid dye Diamond Green B Basic Green B, Malachite Green, Brilliant Green Cationic basic diaminotriarylmethane dye Methyl Violet Basic Violet 1 Cationic basic triaminotriarylmethane dye Fuchsine Basic Violet 14 Cationic basic triaminotriarylmethane dye
The second analysis indicated that additional dyes could be present in low concentrations (table 2.2):
• Victoria Blue B/R (C.I. 44045/C.I. 44040) • Rhodamine B/6G (C.I. 45170/C.I. 45160)
These compounds showed very low detection levels during UV-visible-spectroscopy so their identification was based on molecular masses detected via mass-spectrometry. However, because the mass values could represent compounds other than the above dyestuffs, these results remain inconclusive.
Table 2.2. Synthetic dyes possibly present in TM-ALB-2211 (Images source: Sigma-Aldrich)
Common name
Other
names Class Molecular structure
Victoria Blue B Basic Blue 4, Pigment Blue 2 Cationic basic triaminotriarylmethane dye Rhodamine B Basic Violet 10, Basic Red 1, Solvent Red 49 Cationic diaminoxanthene dye 2.1.3. pH measurement
The pH of paper fibres sampled from the same locations used for UHPLC and the fibre composition analysis was measured with a flat-electrode meter37with the following results:
• Undamaged paper fibres – pH 5.7 • Discoloured paper fibres – pH 6.1
Measurements of the surface pH of undamaged and damaged areas of the paper (on album page 7) yielded the results:
• Undamaged area – pH 4.7 • Discoloured area – pH 6.9
Each measurement was only taken once so the results lack precision. Moreover, it is difficult to accurately measure the pH of paper, particularly in a non-destructive manner.38
Nevertheless, the relation between the values is consistent across both fibre and surface measurements, indicating that the damaged areas have a higher pH compared to the undamaged paper.
To prevent damaging the artefact, the water-sensitivity of the paper was first tested with deionized water drops applied with a micro-pipette. Although this preliminary micro-spot test showed no risk of damage, the flat-electrode pH measurement, which required fairly large water drops, caused a visible change in the discoloured area tested on the paper surface (fig. 2.4). Applying a brush and blotter technique used for diminishing tidelines, it was possible to minimize the impact on the area (fig. 2.5). The undesirable impact of the pH measurement and the success of the treatment demonstrated that some of the dye degradation products in the paper are water soluble (figs. 2.6-2.7).
38 The cold extract method as described in ISO 6588-1:2012 is an accurate method of pH measurement.
However, it requires a significant amount of paper (2 grams), which is not acceptable in a cultural heritage context. See ISO 6588-1: 2012, Paper, board and pulps -- Determination of pH of aqueous extracts -- Part 1: Figure 2.5. Discoloured area on album page 7 after
tideline treatment (no magnification, visible light) Figure 2.4. Discoloured area on album page 7 after
2.1.4. Spot tests
Chemical spot tests39 were used to investigate the internal sizes of the coloured paper.
Iodine/Potassium iodide test for starch
The iodine/potassium iodide spot test was carried out on paper fibres sampled from an undamaged area in an inconspicuous location near the spine of the album.
39 The spot tests were carried out by the student. See Appendix 4 for details on the procedure.
Figure 2.7. Tideline caused by pH measurement on discoloured area on album page 7 (50x
magnification in reflected light) Figure 2.6. Tideline caused by pH measurement on
discoloured area on album page 7 (50x magnification in reflected light)
Figure 2.9. Undamaged paper sample during potassium iodine/iodide test (50x magnification,
reflected light) Figure 2.8. Undamaged paper sample before
potassium iodine/iodide test (50x magnification, reflected light)
The test had negative results: no colour changes were observed, indicating the absence of starch sizing in the album paper (figs. 2.8-2.10).
Raspail test for rosin using sulfuric acid
The Raspail spot test was carried out on fibres samples from undamaged and damaged areas of the coloured paper. The fibres were sampled from the same locations used for UHPLC and optical microscopy.
After the test, the paper samples changed to a dark red colour, indicating the presence of rosin in both the undamaged and the discoloured areas (figs. 2.11-2.16). This spot test could not quantitively establish whether the samples contain different amounts of rosin but demonstrated the use of alum-rosin sizing, as expected of early 20th-century paper.
Figure 2.10. Undamaged paper sample after potassium iodine/iodide test (50x magnification, reflected light)
Figure 2.11. Undamaged paper sample before
Figure 2.13. Undamaged paper sample after Raspail test (32x magnification, reflected light)
Figure 2.14. Damaged paper sample before Raspail
Test for aluminium ions using aluminon
The aluminon spot test was carried out on undamaged and damaged paper fibres sampled from the same locations used for optical microscopy and UHPLC.
The samples turned bright red, indicating some aluminium ions may be present in both. However, the bright red precipitate appeared to intertwine with the undamaged fibres (2.17-2.18), whereas the damaged fibres seemed bright red only at the edges of the droplet tideline but not across the entire sample (2.19-2.20). More aluminium ions are probably present in the undamaged than the damaged paper, which suggests that the alum-rosin size was affected by the adhesive-induced deterioration process.
Figure 2.16. Damaged paper sample after Raspail test (32x magnification, reflected light)
Figure 2.17. Undamaged paper sample before aluminon test (50x magnification, reflected light)
Figure 2.18. Undamaged paper sample after aluminon test (50x magnification, reflected light)
2.2. The adhesives
2.2.1. FTIR analysis
Fourier-transform Infrared Spectroscopy (FTIR)40 was used to investigate the adhesives
present in the case study object. Attempts to analyse the object using FTIR-ATR (Fourier-transform Infrared Spectroscopy – Attenuated Total Reflection), a non-invasive technique, proved unsuccessful, so it was necessary to use the FTIR-microscope which requires minute samples.
To identify the adhesive used in the undamaged second half of the album, a sample was taken from the glossy yellowed film visible on page 42 (fig. 1.16). The analysis determined the residue contains a natural gum of undetermined vegetable source.
To identify the adhesive used in the damaged first half of the album, two of the three types of solid residues described in Section 1.3.3 were investigated: the matte white residue described as type 1 (fig. 1.13) and the semi-glossy yellowed residue described as type 3 (fig.1.15). Samples were taken from residues on page 10 and on page 2, respectively. The analyses yielded the following results:
• Residue type 1 – inorganic material, possibly silicon-based
• Residue type 3 – modified starch, possibly Gluton – an adhesive formerly produced by Talens
An important limitation of the FTIR-microscope technique is the instrument’s inability to detect components present in proportion of less than 3%. The analysis would therefore fail to detect any other ingredients, such as additives, plasticizers, preservatives, that may exist in very small proportions in either of the investigated samples. Moreover, the technique is
40 FTIR was carried out by Suzan de Groot (RCE). See Appendix 6 for spectra and details on the procedure.
Figure 2.19. Damaged paper sample before aluminon
not efficient in analyzing mixtures. As for any of the technical analyses used for this project, the aging of the residues undoubtedly also renders the results more difficult to interpret.
2.2.2. Py-GC/MS analysis
Given the limited knowledge provided by FTIR, Pyrolysis–gas chromatography–mass spectrometry (Py-GC/MS)41 was used to further investigate the nature of the damaging
adhesive. The same locations and residue types were sampled (figs. 1.13 and 1.15). The analyses led to the following results:
• Residue type 1 – inconclusive: 58.8% starch, 23.8% skin grease/fat (?), 9.1% protein/not specific (?), 5.2% colophony, 2.3% phosphate, 0.7% wax
• Residue type 3 – 95.4% starch (amylopectin), 4.6% fatty acids and glycerol, possibly a non-aqueous solvent
The chromatographs generated by residue type 3 were compared to chromatographs generated by an analysis of a recently purchased Pritt stick glue. Both compounds contain chloromethane and s-triazine. Caprolactam, fatty acids (FA-C14-FA-C16) and some glucosides were found in Pritt, but not in the photo album adhesive. Furan was found in the photo album adhesive, but not in Pritt. These differences may be due to the aging of the photo album adhesive. Further research is required to determine whether Pritt may have been used in TM-ALB-2211.
2.2.3. Further FTIR analysis
FTIR42 was also carried out on a Pritt stick glue purchased post-2012. The FTIR analysis
showed that Pritt contains starch-related products but could not identify any silicon-based products. These results neither confirm nor deny the Py-GC/MS analysis.
2.2.4. SEM-EDX analysis
The damaging adhesive was further investigated via scanning electron microscopy coupled with energy-dispersive X-ray spectroscopy (SEM-EDX).43 Residue types 1 and 3 were
sampled from the same locations used for FTIR and Py-GC/MS (figs. 1.13 and 1.15). This analysis revealed the significant differences in the topographies of the two adhesive residues: • Residue type 1 – irregular, amorphous, and chalk-like surface texture suggesting a
heterogenous composition (fig. 2.21)
• Residue type 3 – smooth, crystalline, and glass-like surface texture suggesting a homogenous composition (fig. 2.22)
43 SEM-EDX was carried out by Ineke Joosten (RCE). See Appendix 8 for the procedure and detailed results
and interpretation, including the method for categorizing sample areas. Figure 2.21. Micrograph of adhesive
residue type 1 sample (350x magnification, 5.00 kV, GAD). Photo:
Ineke Joosten (RCE)
Figure 2.22. Micrograph of adhesive residue type 3 sample (350x magnification, 1.00 kV, ETD). Photo: Ineke Joosten (RCE)
The analysis also determined the elemental composition of 12 points on the sampled residue type 1 and 12 points on the sampled residue type 3. The points were grouped into 5 categories per residue sample, based on the established composition as well as the local appearance. Darker areas were associated with lower atomic numbers (Z) and brighter areas with higher Z. The average elemental ratios of the categories (or single values, in the case of categories containing only one point) are visually represented in charts 2.1 and 2.2.
Chart 2.1. Elemental composition in compound percentage of several points on sampled residue 1. Error bars represent standard deviation.
Chart 2.2. Elemental composition in compound percentage of several points on sampled residue 3. Error bars represent standard deviation.
These results should be taken as merely approximative for several reasons: instrumental error, sample sensitivity to the electron beam (the residue type 3 sample was partially destroyed by the analysis), the relatively low number of measured points, inconsistencies in the selection of the measured points, and sample contamination. However, the values do offer insight into the adhesive composition.
The “Higher Z” and “Highest Z” categories of each residue type may be discounted due to the very high amount of iron which indicates these points may contain impurities. Moreover, for both samples, the “High Z” category is distinguished from either “Medium Z” or “Low Z” only by the presence of aluminium: for residue type 1, the sole differences between the “High Z” and “Medium Z” categories are the inversely proportional quantities of silicon and aluminium; for residue type 3, the negligible proportion of aluminium in the “High Z” category justifies regrouping the results within the “Low Z” category. These adjustments reveal that both samples mostly consist of two major compounds represented by the “Low Z” and “Medium Z” categories:
• Residue type 1:
o an inorganic, silicon- and oxygen-rich compound containing traces of sodium and aluminium
o an organic, carbon-based compound containing high amounts of silicon, some sodium and aluminium
• Residue type 3:
o an organic, carbon-rich compound containing some sodium and trace amounts of silicon
o an inorganic, silicon- and oxygen-rich compound containing sodium and aluminium
It is not possible to determine the proportions of these compounds. However, given the abovementioned sample topographies (figs. 2.21-2.22), one could deduce that residue type 1 contains more inorganic, silicon-rich compounds than organic matter, while residue type 3 consists mostly of organic matter with some silicon-based additives.
2.2.5. Spot tests44
Iodine/potassium iodide test for starch
The iodine/potassium iodide spot test was carried out on adhesive residues sampled from the same locations used for FTIR, Py-GC/MS, and SEM-EDX (figs. 1.11 and 1.13) with the following results:
• Residue type 1 – no colour change, apart from some paper fibres adhered to the sample which turned blue (figs. 2.23-2.24)
• Residue type 3 – turned dark red-brown (figs. 2.25-2.26)
The spot test determined that no starches or dextrins responsive to iodine are present in the residue type 1, apart from the minute amounts that adhered the paper fibres to the solid sample. Dextrins which remain colourless in reaction with iodine may nevertheless be present. Also known as achrodextrins, these represent an advanced stage of molecular breakdown in starch degradation.45
According to Py-GC/MS, residue type 3 contains amylopectin, known to form red-brown complexes with iodine, which indicates the presence of starch products that had been converted/depolymerized, possibly sufficiently enough to be considered dextrins.46 The spot
test therefore supports the results of the technical analyses.
45 Decomposition of starch results in the formation of dextrins, the first of which appear reddish in reaction
with iodine. After further depolymerization, the dextrins do not change colour in contact with iodine. The “Achromic point” is the point at which colour-formation no longer occurs. See Cornell and Hoveling 1998: 231.
Figure 2.26. Residue type 3 after iodine test (50x magnification in reflected light)
Figure 2.24. Residue type 1 after iodine test (50x magnification in reflected light)
Figure 2.23. Residue type 1 before iodine test (50x magnification in reflected light)
Figure 2.25. Residue type 3 before iodine test (50x magnification in reflected light)
Water solubility test
To gain more information on the properties of the damaging adhesive, a simple water solubility test was carried out on residues sampled from the same locations used for the previous analyses (figs. 1.13 and 1.15), using drops of room-temperature deionized water. The following results were observed in transmitted light via optical microscopy:
• Residue type 1 – became slightly swollen within 60 seconds and lost some of its shape which had been influenced by contact with paper fibres (figs. 2.27-2.28); no further visible changes occurred after the first minute
• Residue type 3 – became swollen within 60 seconds (figs. 2.29-2.30) and gradually dissolved within the following 10 minutes (fig. 2.31); an additional tested sample which was observed via stereo microscopy 24 hours after the test demonstrated that the residue eventually dissolved completely and deposited in tidelines after water evaporation (fig. 2.32).
Figure 2.27. Residue type 1 before water-solubility test (100x magnification in polarized transmitted
light with lambda retarder)
Figure 2.28. Residue type 1 during water-solubility test (50x magnification in polarized transmitted light
with lambda retarder)
Figure 2.29. Residue type 3 before water-solubility test (50x magnification, polarized transmitted light
with lambda retarder)
Figure 2.30. Residue type 3 during water-solubility test (50x magnification, polarized transmitted light
