Longevidade do metacrilato: Processo e testes de produção

Este é o trabalho do meu amigo Martin C. Jürgens, que hoje trabalha no museu Rijksmuseum em Amsterdam

O Martin cedeu a sua monografia sobre os testes que ele fez no metacrilato. Ainda não deu para traduzir, mas estamos quase lá.

Silicone Rubber Face-Mounting of Photographs to Poly(methyl methacrylate): Process, Structure, Materials, and Long-Term Dark Stability

A report submitted to the Department of Art
in conformity with the requirements for the degree of

Master of Art Conservation Queen’s University, Kingston, Ontario

Martin C. Jürgens, May 2001

Abridged Version: Appendices IV and V omitted.

Abstract

This research project investigates face-mounting processes for colour photographs that use an acetic acid curing, one-part silicone sealant to permanently adhere the emulsion side of a print to a sheet of poly(methyl methacrylate). The original procedure was patented in the early 1970’s under the name of Diasec® by its Swiss inventor Heinz Sovilla-Bruhlhardt, and has since caused imitators to develop similar processes. The stability of prints mounted in this fashion compared with that of unmounted prints has not yet been examined or tested in any published studies. In this project, samples of both unmounted and face-mounted prints from three laboratories, which use different sets of silicone sealant and primer coating as well as double-sided adhesive film, were subjected to an analysis of their physical structure by examination of cross-sections with a microscope, a comparison of FT-IR spectra of the primers and sealants, and artificial ageing in dark storage conditions following ANSI/NAPM IT9.9-1996, with additional quantification of acetic acid off-gassing by gas chromatography. Infrared spectroscopy showed the unlabeled proprietary sealant used in the licensed process to be of similar composition to commercially available ones used by non-license laboratories. Exposure of the samples to intense environmental conditions showed the face-mounting process to slow the yellow staining of the minimum density areas and in two cases to retard colour shifts in the neutral grey areas of the prints. The application of standard testing methods was evaluated in terms of its appropriateness for face-mounted prints.

Acknowledgements

Dieter Jung, Manager, Grieger GmbH + Co. KG, Düsseldorf
Rob McCallum, Camera Kingston, Kingston, Ontario
Prof. Alison Murray, Art Conservation Program, Queen’s University, Kingston, Ontario Douglas Nishimura, Image Permanence Institute, Rochester, NY
Sylvie Pénichon, Art Institute of Chicago, Chicago, IL
Dr. Allison Rutter, Dr. Graham Cairns, and Dr. Sandra Graham, Analytical Services Unit,

Environmental Studies, Queen’s University, Kingston, Ontario
Dr. H.F. Shurvell, Professor Emeritus of Chemistry at Queen’s University, Kingston, Ontario

I would also like to thank:
Vince Abraham, Technical Consultant, GE Silicones; Deutscher Akademischer Austauschdienst

(DAAD), Bonn, Germany; Dorothee Farr and Nigel Barnett, Agnes Etherington Art Centre, Kingston, Ontario; German Department, Queen’s University, Kingston, Ontario; Amanda Gray, paintings conservator, Kingston, Ontario; Ron Hancock, Royal Military College, Kingston, Ontario; Heinz and Marjorie Jürgens; Lloyd Kennedy, Pathology Laboratory, Kingston General Hospital, Kingston, Ontario; Wayne Lyons, Anatomy and Cell Biology Department, Queen’s University, Kingston, Ontario; John McElhone, Photograph Conservator, National Gallery of Canada, Ottawa, Ontario; Peter Mustardo, The Better Image, Pittstown, NJ; Kimberly Schenk, Baltimore Museum of Art, Baltimore, Maryland; Preston Schiedel, photographer, Kingston, Ontario; Sophie Schmalriede, Düsseldorf; Hans Schuurmans, President, DuroTech Corporation, Richmond, VA; Markus Seewald, Geo Magazine, Hamburg; Scott Williams, Canadian Conservation Institute, Ottawa, Ontario.

List of Figures ……………………………………………………. v
List of Tables …………………………………………………….. vii
List of Abbreviations ……………………………………………. viii
1. Introduction ……………………………………………………. 1
2. Process Description ………………………………………….. 14 2.1.
History …………………………………………………………….. 14 2.2.
Mounting Procedures and Problems ………………………….. 18 2.3.
Materials …………………………………………………………… 24
2.3.1. Photographic Papers …………………………………….. 24 2.3.2.
Silicone Sealants …………………………………………………. 25 2.3.3.
Primers ……………………………………………………………. 33 2.3.4.
Poly(methyl methacrylate) …………………………………….. 35 2.3.5.
Double-sided Adhesive Films ………………………………….. 36
2.4. Laminate Structure ………………………………………… 37
3. Experiment I: Chemical Analysis ………………………….. 42 3.1.
Experimental …………………………………………………… 42 3.1.1.
Materials and Equipment …………………………………….. 42 3.1.2.
Procedure ……………………………………………………….. 44 3.2.
Results and Discussion ………………………………………… 45
iii
4. Experiment II: Dark Ageing in Intense Environmental Conditions ………………………. 59 4.1.
Experimental …………………………………………………… 59 4.1.1.
Materials and Equipment …………………………………….. 59 4.1.2.
Procedure ………………………………………………………. 62 4.2.
Results and Discussion ………………………………………. 69
5. Experiment III: Determination of Acetic Acid Offgassing …………………………………….. 83 5.1.
Experimental …………………………………………………. 83 5.1.1.
Materials and Equipment ………………………………….. 83 5.1.2.
Procedure ……………………………………………………. 83 5.2.
Results and Discussion …………………………………….. 87
6. Conclusions…………………………………………………. 94
7. Further Research…………………………………………… 97
8. References …………………………………………………. 103
Appendices …………………………………………………….. 115 I.
Equipment and Material Specifications ……………………. 115 II.
Photographic Laboratory and Mounting Studio Specifications …..117 III.
Density Measurement Terminology and Calculations ……………..118 IV.
Densitometry Measurement Data ………………………………… 120 V.
Acetic Acid Off-gassing GC-FID Data …………………………….. 139 VI.
Curriculum Vitae ……………………………………………………… 160

List of Tables

Table 3.1. Sealants and primers for FT-IR analysis ………………….. 42
Table 3.2. Labelling system for FT-IR Samples ……………………………. 45
Table 3.3. Individual sample preparation of sealants for FT-IR analysis … 46
Table 3.4. Individual sample preparation of primers for FT-IR analysis … 47
Table 3.5. Assignment of IR frequencies in the FT-IR spectrum of sample BS-1, uncured GE Silicone 1201
sealant ………………………………………………………………………………. 52
Table 4.1. Unmounted samples for accelerated ageing ………………… 62
Table 4.2. Face-mounted samples for accelerated ageing and acetic acid off-gassing tests ………………………….. 63
Table 4.3. Labelling system for samples for accelerated ageing and acetic acid off-gassing tests …………………. 65
Table 5.1. AD-Strip readings after ageing for 38 days ………………… 88
Table 5.2. Results from GC-FID measurements and calculations ……….. 89

List of Abbreviations

∆: difference between values
ANSI: American National Standards Institute
B: blue colour patch
C: cyan colour patch
°C: degrees Celcius
d: symbol for measured density
D: symbol for density corrected for dmin
dpi: dots per inch
FT-IR: Fourier transform infrared spectroscopy
G: green colour patch
GC-FID: gas chromatography with flame ioniser detector GE: General Electric
IR: infrared
LDPE: low density polyethylene
LED: light emitting diode
M: magenta colour patch
min: area of minimal density
mm: millimeter
MSDS: material safety data sheet
n: refractive index
N: neutral grey colour patch
nm: nanometer
PDMS: poly(dimethyl siloxane)
PE: polyethylene

viii

PMMA: poly(methyl methacrylate) ppm: parts per million
PVAC: poly(vinyl acetate)
R: red colour patch

RC: resin coated
RH: relative humidity
RTV: room temperature vulcanising t: elapsed time
Tg: glass transition temperature
UV: ultra violet
Y: yellow colour patch

1. Introduction

Since their conception, colour photographs have been more sensitive to degradative influences than their black and white contemporaries, given the premise of correct processing. Although image and support stability have benefited greatly from technological advances, the same deterioration mechanisms induced by exterior influences that troubled early colour prints are still valid today. These include mainly changes in the image forming dyes due to oxidation, be it thermally, light, ultraviolet radiation (hereafter UV), or chemically induced. Pollutants in the atmosphere such as peroxides, ozone, and sulphur dioxide attack the synthetic organic dyes found in modern colour prints. In addition, physical damage to the emulsion and support surface can damage and disfigure a print. Water damage can lead to staining and mould growth.

Next to its stability, of primary importance to a photograph is of course also its appearance and form of presentation. A print is usually made to be seen and appreciated. Unfortunately, this need will often compromise the preservation aspects. A photograph conceived to be displayed mounted to a board but otherwise unframed, as favoured by many contemporary artists, for example, is exposed to physical and chemical dangers. Thus, the quest for presenting and simultaneously protecting prints has always been one that has necessitated inventiveness and imagination based on a sound physical and chemical understanding of the materials involved.

In the late 1960’s and early 1970’s Heinz Sovilla-Bruhlhart, of Switzerland, put considerable effort into developing mounting methods that would not only serve to preserve a photograph but also make it presentable. In 1972 he applied for a patent for his newly developed process, given the name Diasec®, of using a moisture-curing silicone rubber as an adhesive between the photographic emulsion of a print and a sheet of clear acrylic (Sovilla-Bruhlhart 1973). The main advantage of the use of a silicone rubber over that of conventional adhesives was that the curing process could occur spontaneously in the air-tight space that was present between the print and the acrylic sheet by utilising the moisture in the gelatine emulsion of the photograph. Silicone rubber is chemically stable and inert. It is also known as a material that can

withstand great changes in temperature and relative humidity (hereafter RH) as well as age without yellowing. The photographic emulsion is essentially sealed in an airless environment that should not harm it chemically and should protect it from fluctuations of RH and atmospheric pollutants in the environment. It is not obvious, however, that the sealing of a colour photograph in such a laminate structure can only be regarded as beneficial. Residual chemicals in a poorly processed print may be trapped in the process of face-mounting, and it has been stated that the complete elimination of air penetration to a colour emulsion may increase the rate of cyan dye fading (Kodak Professional Division 1998).1 At the present state of knowledge, the face- mounting process is considered irreversible.

The acrylic sheet to which the print was adhered was not specified further in the patent. Today, Plexiglas®, a brand of poly(methyl methacrylate) (hereafter PMMA), is mostly used, and it often will have incorporated UV filters. It serves to protect the emulsion from UV radiation, abrasion, and mechanical damage due to its structural stability. In terms of light stability of the face-mounted print, Sovilla-Bruhlhart claimed that, on a scale of increasing lightfastness from 1-8, his mounting process would enhance Kodak materials from 3.5 to 5, and Ilford Cibachrome from 5 to 7.2 This data appears to stem from a report of lightfastness tests carried out at the Eidgenössische Materialprüfungs- und Versuchsanstalt (EMPA) in St. Gallen, Switzerland (1987), in which Diasec® mounted colour and greyscales were subjected to a xenon light source alongside a blue wool standard. The reasons for these results are not absolutely clear, however, since it would appear that, as most colour papers since the 1980’s have been manufactured with a protective UV filtering surface coating (Wilhelm 1993, 111, 145), the necessity of UV filters in the acrylic sheet would be superfluous. It is more likely the tendency of Plexiglas® to absorb a portion of visible light that is of importance here.

Rochester, NY, personal communication, May 9, 2001).

This statement, however, has not been corroborated by experiments (Doug Nishimura, Image Permanence Institute,

Diasec Sovilla. December 12, 1990. Fax to Metropolitan Museum of Art, New York.

Next to the protection this face-mounting technique is designed to offer, the aesthetic function is important. In contrast to a conventionally matted or non-matted framed photograph, with face-mounting, the acrylic sheet and the print have fused to become one object. The Diasec® effect, as it might be termed today, consists mainly of applying a hard, thick, transparent, and highly glossy surface to a photographic print, a process that results in a similar effect reached by applying varnish to a painting. The same effect can be seen on a beach, where wet stones by the water will appear more brilliant and colourful than the dry ones further up the shore. In effect, the contrast is heightened, the colours are darker, and the perceived saturation of the hues is enhanced. Poly(dimethyl siloxane), the main component of silicone rubber, has a refractive index
(hereafter n) of 1.43, which is relatively close to that of poly(methyl methacrylate), of which Plexiglas® is made, at 1.49. The photographic emulsion, consisting of the protein gelatine in a highly purified form, has a refractive index in the vicinity of 1.54 (Seferis 1989, VI/454-455). The similarity of these materials in terms of their interaction with light, their intimate contact with each other, and the exclusion of air in the face-mounted laminate leads to a minimal reflection and refraction of light rays at the interfaces between the materials. In this manner, light scattering that would be present in a layer of air between the print surface and the glazing in a conventional frame is eliminated. The original surface of the photographic print cannot be seen, independent of the viewing angle or distance. Instead, light reflects from the surface of the Plexiglas®, behind which is a deep “space” of colour, namely the thickness of the PMMA sheeting. When an unframed Diasec® print is viewed from the side, the colour of the photograph can be seen in the Plexiglas® sheet edges, resulting in a print that appears to be of thick, solid colour.

It is these characteristics primarily that have fuelled the popularity of the Diasec® face- mounting technique for chromogenic colour prints on resin coated paper (hereafter RC paper) specifically in fine art circles in Germany and other European countries. In a review of German photographer Matthias Hoch’s work, Hans Dieter Huber points out that it is the surface of a print that determines its reception as a work of art. He goes on to remark that “the traditional method

of framing a photograph with a passepartout and ordinary glass creates a greater distance to the viewer’s eye which is not only perceived spatially but which may also be apprehended as an emotional, intuitive or semantic distance. The way a photograph is presented – its surface, its materiality and its optical distance – plays a significant part in the specific form of the meaning which then derives from the image” (1998).

The format of a Diasec® print is restricted only by the size of the photographic paper and of the acrylic sheet it is adhered to (Figure 1.1). Prints can be found as wide as 1.8 metres, the width of the paper roll, with lengths of greater dimensions. At such large formats, it becomes important to secure the structural stability of a photograph. Face-mounting offers a stiff but slightly flexible support to a print, and has led to a growing trend of displaying images without frames.

®3 Figure 1.1. Thomas Ruff, “Portrait 1988”, chromogenic print and Diasec , Art Institute of Chicago

As the prices for art that is face-mounted with silicone rubber go up, the need for research into the history, materials, and long-term stability of these mounted prints increases. To the knowledge of this author, there have been no published studies on silicone rubber face-mounting.

Image: Sylvie Pénichon, Art Institute of Chicago, 2000.

Next to the test run at EMPA (1987), only sporadic testing of lightfastness has been carried out at Ilford Switzerland.4 In terms of the conservation of these prints, even the most basic research has not yet been undertaken.

In general, the experience of curators and conservators with Diasec® prints has been positive.5 The adhesion seems to be permanent, and only few cases of delamination, discolouration, or unexpected accelerated fading of the dyes or staining of the substrate have been reported. These are explored in part in more detail in section 7, “Further Research”. Face- mounted prints have been treated with the same precautions in terms of exposure to light as have unmounted colour photographs. The main problem seems to be that of the susceptibility of the Plexiglas® surface to abrasion and scratching, as well as its tendency to build up static charge and attract dust. Despite the mainly positive aspects of the seemingly ideal preservation and unique presentation strategy that face-mounting photographs presents, the need for a precise and critical examination is valid, as these prints have only been available for about 20 years, and a number of problematic questions arise upon closer examination of their production process. Furthermore, for the prospective buyer, who is willing to spend high amounts for Diasec® mounted images at auctions and galleries, but also for the artists themselves and their collectors, evidence of the inherent stability of the process is called for.

Some of the questionable issues mentioned above are based on the nature of the materials of the process. The silicone rubber that is used as an adhesive produces acetic acid as it cures, but it is not clear in which quantity it is present and to which extent this is harmful for the acrylic sheet, gelatine, photographic dyes, and colour couplers that make up the emulsion. It is also not known where the acetic acid is deposited, or if it migrates through the print support and the PMMA. Furthermore, the effect of the primer, and specifically the deposition of solvents from

Nora Kennedy, Metropolitan Museum of Art, New York; John McElhone, National Gallery of Canada, Ottawa; Sylvie Pénichon, Art Institute of Chicago, personal communication.

Jean-Noel Gex, Ilford Technischer Dienst, Switzerland, personal communication, August 2000.

the primer on the emulsion and the acrylic sheet should be further investigated. An unsolved problem in the production of the Diasec® and related mounting processes is the occasional forming of magenta stains and the fading of dyes, which mostly seems to occur within the first 48 hours after the application of the silicone rubber.6

Literature Review
The bulk of literature on silicon compounds is found in the area of their chemistry and their

use as an adhesive and a construction material, for example as a sealant for joints between glass and concrete or other materials. Iler describes the complex chemistry of silica, which is used as a strengthening filler in most silicone rubbers and other adhesives (1979). Plueddemann presents the chemistry of silane coupling agents, found in primers for silicone sealants, in extensive detail (1991). The general chemistry of silicon compounds is the topic of Noll’s (1968) and Brook’s (2000) books. The specific literature on the adhesive properties of silicones includes Cook (1970) and Wake (1982). Lewis (1962) and Lynch (1978) describe the technology and science of silicone products, and a text prepared for the internet site of GE Silicones (2000d) gives specific information and chemistry regarding the products often used in non-licensed face-mounting. Rochow presents a concise and inclusive historical overview of mankind’s use and development of silicon materials (1987).

Next to literature on the use of silicones in technology, the application of silicone rubbers in conservation has also been described in various contexts. The Canadian Conservation Institute examined a GE Silicones sealant to determine its suitability for adhering coins to a backing for a display (Analytical Research Services 1989). Infrared spectroscopy showed the material to be poly(dimethyl siloxane), a typical silicone rubber polymer. As it released acetic acid vapours during curing, the sealant was not recommended for its intended use. Fieux presents results on his

Dieter Jung, Grieger Düsseldorf, personal communication, August 2000.

investigation into the use of pressure sensitive silicone adhesives for the production of a lining material for paintings on canvas (1984). He also describes the characteristics and benefits of clear silicone varnishes that might be considered for surface coating paintings. These materials do not seem to be in widespread use in actual painting conservation practice, however.7 Byrne (1984) examines the ability of fumed colloidal silica particles to modify the characteristics of various rubber and adhesive types. This filler material is also found in silicone sealants that are used in face-mounting.

Silicone rubbers have been used extensively for moulding of artefacts at archaeological sites. The ability of a room temperature vulcanising (hereafter RTV) silicone rubber to cure under water has been found to be beneficial in the moulding of shipwreck timbers (Daley and Murdock 1984). A silicone rubber was used in the consolidation of fracturing dinosaur trails in sedimentary rock in Queensland, Australia (Agnew 1984). White quartz sand was mixed with the RTV rubber, which, once cured, demonstrated an elasticity and hydrophobicity that could compensate for the rapid contraction of the rock and its thorough wetting during summer rain storms. These are environmental conditions similar to those in which sealants serve as a construction material in joints on buildings. Maish has determined the staining of terracotta surfaces during moulding to be due to the presence of silicone oils in RTV compounds (1994). A silicone resin similar to that used as a primer for the face-mounting process has been described in applications as a consolidant for archaeological waterlogged wood (Yashvili 1978) and stone sculptures (Charola et al. 1984). The resin is advantageous for this use in that it has a high ability to penetrate stone and is chemically stable due to its strong silicon-oxygen bonds. The experiments showed that a satisfactory curing of the resin could only be achieved in a 30-50% RH range. This factor might correlate with the observations of fluctuations in ambient RH as a possible cause for magenta staining reported by the face-mounting studios, which is discussed in more detail in section 2.3, “Mounting Procedures”.

Amanda Gray, paintings conservator, Kingston, Ontario, personal communication, February 2001.

Another application of an organosilane resin, a compound containing both organic functional groups and silicon, was tested by Wagner (1989). The resin was used to re-adhere delaminating negative emulsions to their glass plate supports. As a bifunctional molecule, the organosilane is able to bond both to the chemically closely related glass and to an organic polymer such as gelatine. A similar scenario is present in the face-mounting process, in which a silane primer is applied both to the surface of the gelatine emulsion and to that of the PMMA. Its organic group is able to bond to each polymer, and the silane end has a high affinity to the silicone sealant that is applied between the two surfaces.

In terms of the stability of the PMMA to incident chemicals, it has been demonstrated that it is liable to swell, dissolve, craze, or turn opaque when brought into contact with certain organic solvents in liquid form (Sale 1988). The primers that are brushed onto the Plexiglas® prior to application of the sealant contain various mixtures of organic solvents, including toluene, alcohols, and acetone, which, according to Sale, all are capable of affecting the plastic. Fenn has shown that, among other plastics, PMMA is capable not only of absorbing organic solvent vapours, such as those from acetic acid, but subsequently off-gassing them over a long period of time (1995). In Oddy tests, metal coupons were corroded by the presence of plastic materials that had previously been exposed to low concentrations of organic solvent vapours. It was pointed out that display cases and book stands made of PMMA may be suspected to be sources of corrosive vapours for other objects in their vicinity. The sealant used in face-mounting processes is an acetic acid curing rubber. The acid vapours that are produced are suspected to diffuse into the gelatine emulsion and into the Plexiglas®, which, following Fenn’s observations, might then serve as a source for acetic acid vapours for many years. This possibility is further explored in the discussion of the acid vapour measurements in sections 5.2, “Results and Discussion” and 7, “Further Research”.

Experimentals
It was felt that a preliminary investigation into the Diasec® process should primarily be

concerned with the mounting process, basic structure, materials, and long-term stability of face- mounted prints. The mounting process and problems that have been observed connected with it are described in detail. The materials relevant to face-mounting are then examined individually in terms of physical and chemical structure and characteristics: photographic chromogenic papers, silicone sealants, silane primers, and acrylic sheets. A cross-sectional analysis of face-mounted samples explores the structure of Diasec® prints.

The sets of unmounted and face-mounted print samples were produced in four laboratories: Grieger Düsseldorf, a further German laboratory that has requested to remain unnamed and carries the pseudonym “Lab X” in this report, a laboratory in New York named “Lab Y” for this report, and Gamma in Chicago.8 The first is a studio that has a license from Diasec Sovilla and produces the true Diasec®. It supplied two sets of samples, those printed digitally to photographic paper, and those printed digitally to a negative, then analogue to a second photographic paper. The second and third laboratories are non-licensed studios that offer face-mounting as a part of their services. The fourth laboratory, Gamma, made test targets which were then sent to New York for face-mounting at Lab Y.

The prints that were face-mounted at Lab Y were found not to have a layer of silicone rubber as an adhesive between the photograph and the PMMA. Instead, the cross-section showed the mounting material to be a transparent and colourless plastic film with an adhesive on each side. This is a technique regularly used by mounting studios, and the result cannot be visually distinguished from silicone rubber face-mounted prints. This discovery compromised the original scope of the experimentals, but the time schedule prohibited the new production of samples with silicone rubber mounting. New samples were solely made for the experimentals of Sylvie Pénichon, who was collaborating as described in “Accelerated Ageing” below.

Chemical Analysis
The long-term stability of face-mounted photographs is mainly governed by inherent

material stability, lightfastness, and sensitivity to moisture. In Experiment I, “Chemical Analysis”, the ingredients of the adhesives involved in the face-mounting process were compared to each other by performing a basic chemical analysis of four sealants and five primers used by mounting studios. Rather than obtaining a precise analysis of the individual ingredients of each of the sealants and primers, in this case the Fourier transform infrared (hereafter FT-IR) analysis allowed for a general comparison of the organic composition of the no-name sealant and primer with that of the brands for which the ingredients can be found in material safety data sheets (hereafter MSDS).9 Of those examined, only the GE Silicone sealants and primer have MSDS available that list the full composition. The technical datasheets of the other known materials either have some ingredients listed as “trade secret” (DuroTech Corp. 1997, 1999) or omit ingredients (Jenny Co. AG 1999b, 2000). Through this analysis it was determined that the no- name products used in the patented Diasec® process, for which, due to license agreements, no information could be acquired, are similar to those used by non-licensed labs. The specific identification of the unknown ingredients was not possible.

Accelerated Ageing
The major concern expressed by those dealing with the face-mounted prints is their

unknown long-term inherent stability or lack thereof. Experiment II examined the colour stability of face-mounted prints. Most chromogenic dyes used in colour photographic papers for exposure from a negative are known to be inherently unstable: with time, they tend to fade or undergo a shift of hue. The rate of these reactions is determined not only by the chemical make-up and

Specifics of these laboratories are listed in Appendix II.

stability of the dyes and their immediate environment, but also by the temperature and RH of the environment as well as the degree of atmospheric pollution present in the air in which the photograph is kept (ANSI 1996, ii). As the temperature and humidity content of the gelatine binder decrease, chemical reactions that may lead to the destruction of the dye molecules become less common. As the values increase, the reaction rate also increases. This will hold true until the environmental conditions cannot support the physical state of the materials any longer. In contrast to light fading, in which dyes are destroyed by photochemical reactions, this type of colour fading is termed “dark fading”, indicating that this degradation will occur without the influence of visible light or UV radiation. Therefore, a colour photograph will change in any case and in any environment. The dark fading test as described in the standard ANSI/NAPM IT9.9- 1996, Stability of Color Photographic Images – Methods for Measuring (1996) can be interpreted as a method for determining the inherent stability of the image-forming dyes and undeveloped colour couplers in chromogenic papers.

Experiment II was part of a joint research project: it was performed in collaboration with Sylvie Pénichon, photograph conservator at the Art Institute of Chicago, who carried out light exposure tests and dark ageing tests of free-hanging unmounted and face-mounted prints in temperature and humidity controlled ovens at the facilities of the Image Permanence Institute in Rochester, NY. In the part of the project carried out at Queen’s University in Kingston, Ontario, prints of the same batches as those used in Rochester were dark-aged in temperature controlled ovens while sealed in bags. The goal was to compare the dark stability of face-mounted photographs to that of the prints that remained unmounted. It was determined that, for the most part, the face-mounted prints underwent deteriorative changes to the image at a slower rate than

A detailed analysis of the materials would require the separation of the individual ingredients, for example the isolation of the polymer backbone, as described by Angelotti and Hanson (1974, 65), that is beyond the scope of this project.

their unmounted counterparts. In conclusion to these tests, the application of the standard test method ANSI/NAPM IT9.9-1996 as used for this specific material was evaluated.

A method of prediction for the long-term stability of materials based on accelerated ageing, the Arrhenius extrapolation, had been considered for the evaluation of the ageing tests. The measured data did not result in curves appropriate for this type of extrapolation, however, since, in most cases, they were not nearly linear and showed pronounced irregularities. As a result, it was decided to merely compare the measured results of the aged samples and discuss their differences, without attempting to predict the long-term stability of the materials in specific numbers of years versus storage temperature.

Acetic Acid Measurement
It is known that the sealants used in face-mounting are of the type that produce acetic acid

during curing. Various methods for measuring amounts of acetic acid emissions have been described. Tétreault used pH indicator strips saturated in a glycerin-water solution to determine the presence and amount of acidic volatile products in a closed environment (1992). In a relevant application, it was shown that at least 25 days are required for the acetic acid emissions of one gram of an acetoxy silicone rubber to slow down almost to a stop, leaving a relatively inert rubber product. Nicholson and O’Loughlin make use of a simpler detector of acidic vapours in their experiments with AD-Strips10 (1996). The material under examination was placed in a stoppered jar with an AD-Strip, and the colour change of the indicator was read after a set amount of time. This procedure was considered a quick and cheap precursor to an Oddy test, which requires a more elaborate preparation time.

AD-Strips are manufactured and distributed by the Image Permanence Institute, Rochester, NY. They consist of a strip of paper impregnated with the indicator bromcresol green, which changes its colour when in the presence of acidic gases. AD-Strips were developed primarily for the detection of acidic off-gassing from cellulose acetate film base, for which the developers received a Technical Achievement Award (“Oscar”) from the Academy of Motion Picture Arts and Sciences in 1997.

In Experiment III, the off-gassing of the face-mounted prints was primarily determined on a general level by the inclusion of AD-Strips in the sealed bags, and then quantified in more detail by analysis of the air in the bags by gas chromatography with flame ioniser detection (hereafter GC-FID). Next to the samples that were aged at elevated temperatures in Experiment II, off- gassing was also examined in samples kept at room temperature, to simulate an exhibition or conventional storage environment, and in those at –15° C, to simulate cold storage for colour photographs. It was found that the rate of off-gassing was dependant on the storage temperature.

Further research and investigation into the Diasec® and related processes will be inevitable if more interest is shown in this mounting technique, and it is anticipated that this research may be a starting point on a long endeavour.

2. Process Description 2.1. History

The initial historical development of the silicone rubber face-mounting process can be followed in the respective patent literature, which has been compiled by Sylvie Pénichon, Art Institute of Chicago. The goal of the inventions described in the patents is clearly the long term protection of photographic images by their removal from damaging environments. The first patent obtained by Heinz Sovilla-Bruhlhart, of Renens, Switzerland, in 1970, describes a method of vacuum-sealing a photographic print between two glass plates, at least one of which is transparent (Sovilla-Bruhlhart 1970a). The borders of the plates are sealed, and the air is removed. The ensuing removal of moisture from the print due to the vacuum is thought to increase its resistance to UV-related fading. To further enhance the protection against UV radiation, an inert liquid such as paraffin oil may be introduced into the sealed enclosure. In another patent, that was submitted later but granted before the one described above, Sovilla- Bruhlhart proposes the deposition of a thin coating of fine varnish droplets over the surface of the slide material to be sealed before the vacuum is applied (1969). The resulting matte film prevents the formation of Newton rings.11 In a third variation on the sealed sandwich technique, a 1970 patent describes the introduction of an inert material such as argon, neon, helium, krypton, xenon, nitrogen, or other gases without reducing potential between the glass plates (Sovilla-Bruhlhart 1970b). In a patent applied for in mid 1971, but only granted in 1973, Sovilla-Bruhlhart describes a new system that seems easier and presumably cheaper to carry out and which more closely resembles the process that is used today. A photographic or photomechanical print is face- mounted to a sheet of glass or acrylic with an adhesive coating that has a protective release paper.

Newton rings are a form of interference of light waves often encountered in photographic applications. They appear mainly as concentric bands of colour around the point of contact between a film of gelatine or plastic and a sheet of glass. “Anti-Newton glass” has a smooth side and a finely etched side that is in contact with the film and in this manner avoids the formation of light interference bands.

Once the paper is removed, the print is applied to the sheet with pressure. The single sheet provides enough rigidity to dispense the need of a backing (Sovilla-Bruhlhart 1973a).

On January 13, 1971, the inventor applied for a further patent in Switzerland, which was granted three years later. In this document he describes a solution to a number of difficulties encountered with the use of pressure sensitive adhesives (Sovilla-Bruhlhart 1974). Adhesive problems include the cockling and the delamination of the print from the acrylic sheet when a water-based adhesive is used. An adhesive based on solvents, however, smudges printing inks and may not set due to poor permeability of the plastic layers used in photographic paper to solvents. In addition, no adhesive was found that would not yellow with age. As a solution, Sovilla-Bruhlhart introduces the use of a silicone sealant as an adhesive, which he admits, works “surprisingly”12 well for permanently and perfectly mounting prints and transparent slides. It is ideal in that it eliminates the problems associated with the adhesives described above, as it sets using the moisture contained in the gelatine layer of the photograph. The use of a primer is mentioned as being advantageous to the overall adhesion. This patent names the brands of silicone rubber and primer that were used and contains instructions for the mounting process that are principally still valid today. It can therefore be considered a breakthrough document. The same technique, described in more detail, was deposited for patent in Germany in 1972. The patent was granted one year later (Sovilla-Bruhlhart 1973b).

With dates that slightly conflict with the patent information, Sovilla-Bruhlhart describes his invention (sic): “At midnight 31.12.1971 i succeeded the first correct and useful DIASEC FACE mounting with help of a very thick PVC glue. Very shortly after, discussing with a college friend about different items he told me about a strange experiment he had observed at the Plexiglas makers plant Röhm in Germany, the experiment was using silicone gum for producing plates of

Translation of “überraschenderweise” from the German version of the Swiss patent (Sovilla-Bruhlhart 1973b, 3).

Plexiglas elastically bonded together with great adhesivity to the material… The 7.2.72 i put my first ‘silicone’ Swiss patentclaims”.13

A further German patent, also applied for in 1972, took almost five years to be accepted (Sovilla-Bruhlhart 1977). This document describes in detail the use of the same silicone rubber and primer to permanently mount the back of a photographic material to a sheet of plastic. In it, Sovilla-Bruhlhart refers to a German book, Die Technologie der Klebstoffe, written by C. Lüttgen in 1959, in which the use of a silicone adhesive to mount images on various supports to plastic sheeting is mentioned. Sovilla-Bruhlhart describes the advantages of this method, which include stability to high temperatures14 and humid environments.

Later attempts at improving on Sovilla-Bruhlhart’s system were made, as is apparent from further patent literature, but none of them seem to be practiced by mounting studios today. Ludwig Gminder received a patent on his use of an acrylic dispersion adhesive that is applied to the Plexiglas® surface, then left to partially dry (1983). The photograph is pressed onto the acrylic adhesive and Plexiglas®, and the residual moisture from the adhesive diffuses through the photograph substrate over time. This system requires no primer. A later Australian patent describes a method of entirely embedding a print into an acrylic resin, which in turn is adhered to a sheet of acrylic such as Plexiglas® or sandwiched between two sheets (Attila 1989). The resin, which necessitates a hardener, must be centrifuged to remove all air bubbles prior to its application to the acrylic sheet and photograph. Once all layers have been pasted together, the whole laminate object must be heated gradually to 80 °C, then left to cool. A UV absorber such as Tinuvin, manufactured by Ciba-Geigy, may be added to the resin. Almost ten years after this invention, Claudio Cesar received a patent for his method of encapsulating silver dye bleach transparencies between two glass plates with a polyurethane resin (1997). The use of a UV

In an experiment, the author heated mounted images up to 100 °C and found the adhesion to be stable (Sovilla- Bruhlhart 1977, 4). No details of the experimental are given.

Heinz Sovilla-Bruhlhart, fax to Nora Kennedy, The Better Image, March 24, 1992.

inhibitor, PE-399, manufactured by Morton International, increased the lightfastness of the transparencies in testing carried out by the inventor.

The company Diasec Sovilla SA was founded in Cossonayville, Switzerland, in 1969 by Heinz Sovilla-Bruhlhart.15 The firm offered Diasec® and the related Durasec® along with other printing and mounting processes. License contracts with strict regulations have been sold to a small number of mounting studios throughout the world, and the Diasec® process has always been highly secretive. The license terms include the payment of 5% royalities for each Diasec® mounting performed.16 In December 1992 Heinz Sovilla-Bruhlhart died, and in 1996 the mounting business was taken over by a close co-worker of Sovilla-Bruhlhart and Diasec® licensee, Jean-Marc Trimolet.17 The new business, named Diasec JMT after the current owner’s initials, is one of ten licensees worldwide. The licensed mounting studios are found in Israel, England, Holland, France, Belgium, Germany, and Austria. Sovilla-Bruhlhart’s widow and son have taken over sales of the license, production, and distribution of the special primer, the solution that enhances adhesion between the silicone rubber, the acrylic sheet, and the emulsion. Production is sporadic, and the recipe is kept in three safes in different locations in Switzerland. The ingredients of the primer is the main secret of the Diasec® process, and Mrs. Sovilla referred to the recipe as her “Coca-Cola formula”. The silicone rubber is produced by an undisclosed German company to which the licensees are directed.18

Since the use of Diasec® face-mounting has increased greatly over the past ten years, a number of unlicensed photographic printing and mounting studios have begun to offer virtually the same process, but most are careful to respect the license agreements and are selling it under

JMT, Cossonayville, Switzerland, and the inventor’s widow Mrs. Sovilla, February 8, 2001.

16 ®
303 Gallery, New York, NY, fax of Diasec advertising material to unknown recipient, December 13, 1990.

JMT, Cossonayville, Switzerland, and the inventor’s widow Mrs. Sovilla, February 8, 2001.

Ibid.

Sylvie Pénichon, Art Institute of Chicago, personal communication with Jean-Marc Trimolet, owner of Diasec

different names. Although the results may appear similar to the original, the materials used in the mounting process are from other manufacturers. Methods of achieving similar aesthetic results with different means have become widespread, most notably the use of transparent plastic films with a pressure-sensitive acrylic adhesive on each side applied between the photograph and the Plexiglas®. At present, the process is mainly used for face-mounting photographs for advertising and trade-fair design, but it is also gaining popularity in the fine art photography sector. Especially in Europe, and more specifically in Germany, it has been a favourite finishing and presentation method for contemporary photographers since about 15 years. The current internationally best selling German photographers, Thomas Ruff, Thomas Struth, and Andreas Gursky, usually have their prints made and Diasec® face-mounted at the mounting studio Grieger Düsseldorf. Only seldom does Grieger face-mount prints that were produced somewhere else. An internet search for the term “Diasec” resulted in a majority of German photographers using the technique, or at least the term. Although a protected name, “Diasec has become a word like ‘Xerox’ or ‘Kleenex’ and is being freely used to describe face-mounted photographs, even if they haven’t been mounted by a licensee”.19

2.2. Mounting Procedures and Problems

The mounting procedure for the Diasec® process is described below as it was witnessed at the Grieger Düsseldorf studio in August 2000. The mounting was carried out on a chromogenic photographic print that had been produced at the same studio. It was performed in a large room without climate control that had enough space for handling very large format prints. The equipment consisted of three to four large tables to support large prints, rubber gloves and a chamois cloth for application of the primer, pressure sensitive tape, a cutter, a two metre wide electric metal cylinder press with variable pressure adjustment, pressured air guns for dispensing

Sylvie Pénichon, Art Institute of Chicago, personal communication, October 2000.

silicone rubber from their tubes, and a grindstone for smoothing the Plexiglas® edges. Figure 2.1 shows the mounting process in a cut-through schematic.

Figure 2.1. Cut-through schematic of the cylinder press and mounting technique

The colour print must have ca. 10 cm wide margins on each side that are later cut off. The Plexiglas® must be cut to its end format. It is placed on a brown paper sheet ca. 20 cm larger at three edges, and the top protection film is peeled from its surface, which is subsequently wiped down with the primer using a chamois cloth. The solvent vehicle evaporates quickly, leaving the Plexiglas® surface dry. The print surface is also coated with primer with chamois cloth and left to dry. The evaporation of the solvents can take up to 15 minutes. On large prints, the primer is only applied to the border areas of the Plexiglas® and print for economical reasons. The print is placed face down onto the Plexiglas® and one of the short edges is taped around the equivalent edge of the Plexiglas® to form a hinge. The package is then brought to the press with the taped edge butting up against the slit between the cylinders. The print is lifted up and draped over the top cylinder to reveal the open hinge. Silicone rubber is dispensed uniformly into the crack to form a

line across the Plexiglas®. Following an intial manual push, the package is automatically pulled through the electric press, with the top metal cylinder in contact with the print verso through the brown paper, and the bottom one in contact with the bottom side of the Plexiglas®. By this action, the print is firmly pressed onto the Plexiglas® and the silicone rubber is evenly and thinly squeezed out between the two materials to form a film. Should the line of silicone at the cylinder crack become too thin, extra sealant can be dispensed. The pressure of press is crucial, and is usually set at 2-3 mm for a 4 mm Plexiglas® sheet. When the end of the package is close to being pulled through the cylinders, the protruding edge of the brown paper is folded over the print and the Plexiglas® edge to protect the cylinder from coming into contact with excess silicone that is squeezed out of the end of the sandwich.

Once free from the press, the mounted print is turned over. For the shipping of prints face- mounted for trade fair and advertising clients, the protective film is usually left on the front surface of the Plexiglas®. In the case of mounting of artwork, where a more precise visual quality control is necessary, the film is removed immediately after mounting, so that in the case of adhesion defects or obvious trapped dust a new print can be made and mounted. The silicone rubber is left to cure for 48 hours, after which the image is checked for any staining that may have developed. Finally, the print edges and plastic backing are cut flush to the Plexiglas®. In more recent years, Grieger Düsseldorf has resorted to adhering the verso of the face-mounted print to a 1-2 mm poly(vinyl chloride) hard foam sheet with double-sided pressure sensitive adhesive film in another press. This step is designed to give the photograph additional protection from damage to the verso. If the print is not to be framed, but rather hung free, the Plexiglas® edges are polished with a grindstone. The face-mounted print is then packed in a crate and sent to the client. Artists have prints picked up by art transport companies who often make individual crates.

The mounting process at Lab X20 does not differ much from that of Grieger Düsseldorf, except that presently only the Plexiglas® is wiped with primer, as the technicians do not deem it necessary to do both. Until the summer of 2000, a primer was applied to both acrylic sheet and photograph emulsion. The electric cylinder press is fitted with two hard rubber cylinders with a metal core. In addition, there is no time for a quality control step after 48 hours of resting due to the fast production workflow. As an alternative to silicone rubber, prints and slide materials can also be mounted to Plexiglas® with a highly transparent film coated on each side with a pressure sensitive acrylic adhesive. The result is visually indistinguishable from the silicone rubber process.

Certain problems have been noted sporadically at both German mounting studios pertaining to the mounting process. After mounting, yellow-magenta stains can form on a print, either locally or overall. The colour can be slight or vivid. These stains do not seem to appear at a later point if they have not already formed after 48 hours. Both studios have noted that the staining occurs more frequently when there is high relative humidity in the mounting room, especially in the summer. The manager of Grieger Düsseldorf presumes that the staining may be caused by a locally restricted area of excess of silicone rubber, where the sealant is too thick to set completely. It is thought that this may be due to insufficient pressure during mounting or to thinner areas of non-uniform Plexiglas® causing silicone rubber “pockets”. Another suggested cause is poor processing of the photographic print, which indicates the presence of chemical residues.21 This case is examined in more detail in section 7, “Further Research”.

In addition to local staining, Lab X has sometimes experienced an overall shift of the image hue towards magenta, and all colours can fade. These changes can appear after three to five days or as late as two weeks after mounting. The cause presumed by the technicians is the dissolution of topmost emulsion dye layer, namely cyan. This might also explain the overall

The mounting process was witnessed in August 2000.
Sylvie Pénichon, Art Institute of Chicago, personal communication, October 2000.

colour fading, as a part of the image colour is being lost. Lab X has also been troubled by a slight graduated staining of the white borders around the exposed print area, as if colours have bled from the image area to the non-image area. This phenomenon can be observed in all face- mounted prints, however, and may be a merely optical effect of internal reflections of light between the top and bottom edges of the Plexiglas®. This effect is also mentioned in section 7, “Further Research”.

Perhaps the most likely cause of the changes in dye density and location, causing staining, may be the penetration of certain organic solvents into the gelatine emulsion, which, according to Eastman Kodak Co., is most pronounced in humid conditions such as those described above by both mounting studios (Kodak Professional Division 1998). Primarily solvents of the alcohol, ester, ketone, and carbitol classes enhance the mobility of emulsion components such as dyes and colour couplers between the cyan, magenta, and yellow layers of the emulsion. Of the primers examined, all have solvents of the given categories as one or more parts of the ingredients. According to the technical literature, the major components of the two Sallmetall primers previously used by Lab X are isopropanol (DuroTech Corp. 1997, 1999). The GE Silicone primer used by Lab Y, New York, contains 80-99% ethyl acetate (GE Silicone 2000c), one of the esters specified by Kodak as being able to penetrate moist gelatine. Finally, the “butanon” (sic), presumably butanone, which is methyl ethyl ketone, content of the Gurisil PR 435 primer specified in the Diasec® patent is listed as over 20% (Jenny Co. AG 2000b). In conclusion, all examined primers seem capable of having an effect on the layered structure and chemical system of the dyes and colour couplers in the photograph emulsion.

Among the various effects that Kodak has specified, such as a blue shift due to a reduction in yellow dye stability, the formation of cyan spots, and the production of yellow colourants, the most fitting to the observations made during face-mounting is the formation of red stains. This is caused by the penetration of solvents into the top image-forming layer of the emulsion containing the cyan dye. Due to increased mobility, the dye molecules and other layer components will come

into contact with each other and undergo thermal reactions that lead to a fading of the cyan colourants. This reaction unbalances the absorption of light by the three colour layers, and is apparent as red staining. According to the observations of the Kodak technicians, strong pressure applied to a print previously treated with the solvents specified above may further accentuate the cyan dye fading, as the reactive molecules will increasingly be brought into contact with each other. Kodak has also warned that silicone rubbers that release acetic acid should not be used on colour photographic prints because the acid can adversely affect the yellow and cyan dye stability, but the mechanism is not described. In the case of the face-mounting process, inappropriate solvents and high pressure are applied to the emulsion, which is additionally brought into direct contact with a sealant that gives off acetic acid. These may be contributing factors to the formation of red staining as observed at the mounting studios, but it remains unclear why the staining only occurs sporadically.

Lab X has described further mounting problems, such as a non-uniform adhesion of the print to the Plexiglas®, resulting in delamination at the corners. Furthermore, the rubber cylinder used in the face-mounting process has some kinks that impress themselves into the verso of the print. These indentations are not visible from the print recto. Without acknowledging that prints mounted with the Diasec® licensed materials may also stain, Sovilla-Bruhlhart described problems found in non-licensed mounted prints (sic): “Without our basic DIASEC products you produce something but you never know what troubles will come, like: peel-off -bubbles and air bladders – cristals – redening or yellowing of the colour prints and so on. And the nicest is, our competitors know never if and when and why this may happen tothem, sometimes ther bad reactions beginn after only six months delay”.22

Heinz Sovilla-Bruhlhart, fax to Nora Kennedy, The Better Image, Washington, DC, March 24, 1992.

2.3. Materials
2.3.1. Photographic Papers
The colour photographic papers discussed here are of the class named “chromogenic” due

to the nature of the colour formation in the emulsion. Colourless leuco dyes and colour couplers are present in the gelatine and only form cyan, magenta, or yellow dyes during the development of the paper. Chromogenic papers are typically exposed from a negative image. This may be either a photographic negative or a negative image generated from a digital file. In the first case, termed analogue exposure, white light is projected through the colour negative onto the paper. The second technique, digital exposure, involves red, green, and blue lasers or LEDs that are controlled by a computer. In both methods the papers are developed and fixed following conventional chemical processes. Next to paper, negative material can also be exposed by lasers and LEDs. In this project, the Durst Lambda printer was used to expose the digital papers at the studios Grieger Düsseldorf and Lab X. The second batch of prints produced at Grieger Düsseldorf was made on an analogue paper from a photographic negative that had been exposed from the digital file with the Lambda printer.

Structurally, contemporary chromogenic papers are of the RC type. The paper base is coated by extrusion on each side with a thin layer of polyethylene. The bottom coat is clear, and the top coat has incorporated white pigments that form the white base of the print. The photographic gelatine emulsions are coated onto this layer in individual steps. The emulsion of the colour photograph shown in Figure 2.9 was temporarily swelled with a drop of distilled water under a cover glass (Figure 2.2). This allowed for the discernment of the individual layers, which are, from the surface downwards: cyan, magenta, and yellow. The layers are interspersed with thinner layers of gelatine that contain filters which control the action of light on each emulsion. The top protective overcoat, seen as a whitish layer on the cyan layer, consists of hardened gelatine (Swan 1987, 274) with an incorporated UV barrier (Wilhelm 1993, 111, 145). Under magnification, both the papers for analogue and for digital exposure showed the same order of

layers (Figures 2.9 and 2.10). They differ mainly in their parameters of sensitivity to the type of light to which they will be exposed.

Figure 2.2. Cross-section of ca. 3x swelled emulsion of Kodak Professional Ultra III Paper in reflected light

Though not further examined in this project, Ilfochrome paper, manufactured by Ilford, and black and white prints have also been face-mounted with silicone rubber and adhesive films. Ilfochrome, prior known as Cibachrome, is exposed from a positive colour transparency instead of from a negative. It is able to accommodate for the high contrast commonly found in photographic slide material, and usually has a very high surface gloss. The dyes are not chromogenic, but instead are present in the emulsion prior to exposure and are bleached out in the highlight areas during development. The base is usually of pure plastic. Black and white prints may be on RC paper or fibre-based paper. Currently, not much experience has been gathered with the face-mounting of ink jet or other digital prints,23 although it is assumed that these materials will find greater use in the future.

2.3.2. Silicone Sealants

Silicone sealant, also termed mastic, caulking, gum, or elastomer, is a material based mainly on silicon, carbon, and oxygen that forms polysiloxanes. These polymers are irreversibly converted by crosslinking, or curing, from their highly viscous plastic state to an elastic state, in

Dieter Jung, Grieger Düsseldorf, and the manager of Lab X, Düsseldorf, personal communication, August 2000.

which the material is called silicone rubber. The major component of silicone rubber is the linear polymer poly(dimethyl siloxane) (hereafter PDMS). It is used in a number of construction applications due to characteristics that make it suitable for sealing exterior joints between metal and concrete panels, as well as in glazing and joining metal to glass. Silicones can withstand a number of environmental conditions destructive to other polymers. These include oxygen and ozone attack, ultraviolet radiation, corona, many chemicals and solvents, and water.

Solvents for polyorganosiloxanes, as listed by Fuchs, include aromatic and chlorinated hydrocarbons and esters. Solvents specifically for PDMS may be: amyl acetate, chlorobenzene, chloroform, cyclohexyl acetate, dichlorobenzene, 1,2-dimethoxy ethane, ethyl acetate, ethyl bromide, hydrogenated xylene, isopropyl acetate, methyl ethyl ketone (>20°C), octylamine, o- fluoro-toluene, phenetol (>13°C), and trichloroethylene, (1989, VII/391). Noll states that PDMS can be swelled in a small number of materials, including benzene, carbon tetrachloride, diethyl ether, ethyl acetate, gasoline, methyl ethyl ketone, trichloroethylene, and xylene, and that strong acids and alkalis will destroy silicone rubbers (1968, 512-514). Horie demonstrates swelling characteristics of silicone rubber in the Teas chart (1997, 220). Internal chemical degradation of silicone rubbers appears to occur only at very high temperatures. Due to the oxidation of PDMS, formaldehyde may form at 150° C upwards (Jenny Co. AG 1999b).

Despite their high chemical resistance, methyl silicone rubbers are many times more permeable to gases than are organic rubbers, presumably due to their microporous structure. They are also permeable to liquids, but to a substantially lesser degree (Noll 1968, 514). In terms of permeability P, silicone has the value 270 for carbon dioxide permeation compared to polyethylene at 1.4 and natural rubber at 13.1 (Lewis 1962, 1262). At –125° C, silicones have a low glass transition temperature (hereafter Tg) and therefore tend to pick up and hold dust and dirt (Horie 1997, 159 and 185). Depending on their ingredients, cured silicone rubbers remain stable and flexible in temperatures ranging from ca. -130° to +150° C (Noll 1968, 495 and 501). This indicates that environments recommended for contemporary archival storage for photographs,

such as sub-zero temperatures for colour prints(McCormick-Goodhart 1996, 19), may not be problematic for this component of face-mounted prints. Horie warns that silicone rubbers may contain 2-6% of unreacted silicone oil that may migrate or creep out of joints onto exposed surfaces (1997, 161-162). To the knowledge of the author this phenomenon has not yet been reported in the case of face-mounted prints, but it may be a sign of deterioration that one should be aware of.

The rubber described in the following segments is restricted to the type used in the face- mounting processes, a one-part room temperature vulcanising, or RTV, material that is available as an uncured polymeric compound in an airtight container. Upon exposure to air, it cures rapidly with the help of atmospheric moisture, thereby releasing acetic acid, giving it its classification as an acetoxy RTV sealant. There is minimal or no shrinkage in this process. The cured material is a rubbery, permanently flexible, chemically inert substance that exhibits release properties.24 Due to its relatively pure and simple make-up, it is for the most part biocompatible and non-hazardous to humans. Silicone rubber formulations typically include a polymeric backbone (60-80%), a crosslinking agent(1-5 %), a reinforcing filler (10-30%), and processing aids (1-5 %).25 Sealants used in face-mounting photographs have no incorporated pigmentation, and are thus colourless and transparent.

Polymer (60-80%):
The main ingredient of a typical silicone sealant, responsible for its inherent chemical

properties, is the silicone26, or siloxane, polymer. It is a clear and colourless linear polymer that forms the skeleton of the rubber. It is obtained from the processing of silica containing sand. The

This term, though not compatible with current chemical terminology, has persisted in its use until the present. It was originally introduced under the supposition that compounds of the formula RR’SiO were analogous to ketones.

Other silanes are applied to paper and polyester sheets to form silicon release materials.
The percentages given here relate to the composition of GE Silicones sealant SCS 1200 (GE Silicones 1998b).

polymer is made up from silicon-oxygen monomers, which often have organic side groups attached, the most common of which is methyl. The major component of the two GE Silicones sealants, SCS 1000 and SCS 1200, which serve here as exemplary compounds, is PDMS, as shown in Fig. 2.3. The uncured polymer, which is terminated with hydroxyl groups at each end, has a molecular weight range of 300,000 to 800,000 (GE Silicones 2000d). It may contain a certain degree of branching (Noll 1968, 388), which serves to increase its low temperature flexibility and influence its degree of networking.

Figure 2.3. Poly(dimethyl siloxane)

The high bonding strength between silicon and oxygen is mainly responsible for the excellent thermal and chemical stability of silicones. In comparison to the bond energy of 85 kcal/mole between two carbon atoms, the bond between silicon and oxygen has an energy value of 106 (Lewis 1962, 1240). In addition, all bonds in siloxanes are saturated, and thus less reactive than unsaturated C=C bonds, which are not only prone to chain scission, but also form conjugation in molecules that can cause discolouration upon absorption of radiation. Silicone rubbers have high colour stability.

Crosslinking Agent (1-5 %):

The RTV sealants used in face-mounting photographs are one-part compounds. They contain all of the components needed for crosslinking, but rely on moisture vapour from the atmosphere to start the reaction. As long as the sealant is preserved in an airtight container, it will

remain uncured. It can be assumed that, in the case of face-mounting processes, in which the silicone sealant is pressed into a thin film in an airtight sandwich between the print and the PMMA, the moisture necessary for the cure is available from the gelatine of the photographic emulsion27 (Sovilla-Bruhlhart 1973, 3), and, to a lesser degree, from the acrylic sheet, which holds adsorbed moisture to its surface unless it has been treated with heat.28 In the case of acetoxy sealants, the crosslinking agent methyltriacetoxysilane serves as a centre for moisture to initiate the combination of up to three polyorganosiloxanes at one point (Figure 2.4). The acetoxy groups are hydrolytically cleaved from the silicon atom by water diffusing into the sealant, and acetic acid vapour is released. The evaporation of the acetic acid causes a disruption in the educt- product equilibrium (Noll 1968, 399), which drives the reaction to complete full hydrolysis of the methyltriacetoxysilane. The hydroxyl groups at the polymer ends and on the crosslinking agent molecule then react to form water, thereby creating a siloxane bond. The released water can initiate further cross-linking reactions in areas of the sealant which are not exposed to the original source of moisture. In this manner the sealant cures from the outside edge inwards.

At 22° C and 50% RH, gelatine contains ca. 14% by weight of water (McCormick-Goodhart 1996, 9). Jean-Noel Gex, Ilford Technischer Dienst, Switzerland, personal communication, August 2000

Figure 2.4. Reaction sequence in an acetoxy RTV sealant moisture cure (Noll 1968, 399, and Brook 2000, 283)

A second type of crosslinking agent is of interest, as it was found to be present in the RTV sealant originally mentioned in the Diasec® patent of 1974 (Sovilla-Bruhlhart), namely Gurisil

575, manufactured by Jenny Co. AG. According to the literature, this sealant contains “tris- alkylamino-silan” (Jenny Co. AG 1999b), which is assumed to have three alkylaminos on a silane base, forming either the structure (RNH2)3Si- or (R2N)3Si-, with “R” as an alkyl group. The silane functions in a manner similar to that of the methyltriacetoxysilane molecule to cure the silicone rubber. In contact with atmospheric water, the alkylamino groups are split off from the silane to form an amine gas. The freed silane is then able to bond with the hydroxy-terminated siloxane backbone by giving off water, as in the acetoxy cure reaction. This reaction is depicted in a simplified schematic in Figure 2.5.

Figure 2.5. Simplified reaction sequence in an amine RTV sealant moisture cure

Reinforcing Filler (10-30%):

Crosslinked silicone polymer alone forms a relatively weak material, so it is compounded with a filler material, the most common of which is silica. Silica has the empirical formula SiO2 and is commonly encountered in tetrahedral form, in which the silicon atom at the centre forms single bonds with four oxygen atoms around it. Pyrogenic, or fumed, silica is produced by burning chlorosilanes in the presence of hydrogen and oxygen. This process forms amorphous, extremely fine silica particles in the range of 5-20 nm in diameter that have a surface area of 50- 600 m2/g. These particles are encountered in chain-like aggregates with a rough and irregular surface (Brook 2000, 311). Due to their large surface area, these aggregates are well suited to interact chemically with the silicone backbone. This interaction serves to reinforce the rubber, thereby enhancing the tensile strength of the cured material by a factor of up to ten times (Brook 2000, 320). In conservation, amorphous fumed silica has also been used for modifying adhesive

characteristics such as adhesive and cohesive strength, thixotrope formation, and gel formation of solutions (Byrne 1984).

Process Aids (1-5 %):

The process aids serve to control the viscosity of the rubber by modifying the degree of bonding that occurs between the filler and the rubber polymer. Pendant silanols (SiOH) on the silica surface are able to form hydrogen bonds with oxygen in the siloxane backbone (Figure 2.6). This may occur to such an extent that an uncured rubber will crosslink by hydrogen bonding between polymer and filler alone, which alters its handling characteristics. In order to avoid this premature hardening, the silica is treated by reacting it with silicone oils or an alkoxysilane that renders its surface more hydrophobic by partially replacing the hydroxyl sites with non-reactive alkoxy groups, thus reducing its affinity to the silicone polymer (Figure 2.7). In this manner, the ability of the treated fumed silica to disperse in the rubber can be improved by the addition of a cyclic polyorganosiloxane such as octamethylcyclotetrasiloxane (Noll 1968, 403), a silicone oil, to the rubber compound. This additive is present in both GE Silicones sealants tested here (GE Silicones 1998b and 2000b) and may also serve as a plasticiser and lubricant. A further process aid that is found in the sealant examined here is a siloxane with methylsilsesquioxane. It has cross-linking controlling functions and also modifies the characteristics of the rubber.29

Vince Abraham, Technical Consultant, GE Silicones, personal communication, April 24, 2001.

Figure 2.6. H-bonding between fumed silica and polysiloxane (adapted from GE Silicones 2000d)

Figure 2.7. Reduced H-bonding between treated silica and polysiloxane following surface treatment (adapted from GE Silicones 2000d)

2.3.3. Primers

It is common to pre-treat the surfaces of high surface-energy adherends with a substance that is able to enhance its affinity to the adhesive, thereby promoting the fusing of two materials via an intermediate adhesive. These substances are called primers or coupling agents. The use of a primer is recommended for most silicone sealants by GE Silicones, and different primers are formulated and manufactured for different combinations of substrates and sealants. A primer is often of a similar chemical structure to the adhesive it is designated for, but it is applied in a

solution of a much lower viscosity or greater dilution so that it can flow over a the surface of the substrate easily and uniformly (Brewis 1983, 242-243). Primers used in silicone rubber face- mounting contain organo-functional silicones30 or silane compounds31 which are delivered in an organic solvent mixture. The solvents provide for a low surface tension vehicle that will wet the substrate surface well and displace adsorbed air and water vapour which would otherwise hinder the viscous adhesive from coming into intimate contact with the adherend. Each face-mounting process utilises a specific primer or set of primers that was chosen for a specific sealant. The primer is applied to the surfaces of the PMMA and the photographic print and the solvents are allowed to evaporate.

As hybrid molecules with both organic and silicon functions, organosilane compounds, the active ingredients of primers, are able to form primary bonds with both the organic adherend and the silicon based adhesive. Instead of having to bond with the surface of the adherend, the sealant adheres mainly to the thin coat of primer (Cook 1970, 60), which thus acts in an mediating function. The primer examined here, GE Silicones SS 4179, is that which is recommended for the preparation of an acrylic sheet surface for the GE Silicones sealants SCS 1000 and 1200 by the manufacturer (GE Silicones 2000a). It contains a long and complex organic chain and trimethoxysilyl, and undergoes complicated reaction mechanisms. Among other groups, the organic rest contains 2-propenoic acid (CH2=CH-COOH), which is acrylic acid. This may have an affinity for the PMMA it is designed to adhere to. The trimethoxysilyl group can form primary bonds with hydroxyls on the gelatine, PMMA, and silica filler and siloxanes in the rubber by forming methanol. The primer may cross-link internally and at the chain ends when the solvent vehicle evaporates.32 In short, the primer interacts both chemically and physically, in that it firstly

functional groups.

Siloxanes with an organic rest group.
Compounds with the general formula R-Si-(OR’)3, where R is an organo-functional group, and -OR’ are alkoxy

Vince Abraham, Technical Consultant, GE Silicones, personal communication, April 24, 2001.

enhances primary bonding, and secondly enhances the efficiency of bonding by providing the adhesive and the substrate with a material that is similar in composition.

2.3.4. Poly(methyl methacrylate)

Cast sheets of poly(methyl methacrylate) (PMMA) were commercially introduced on a large scale by Rohm and Haas, Darmstadt, Germany, in 1936, with the trade name Plexiglas® (Hochheiser 1986, 59). The transparent, hard plastic found immediate use in the war industry, specifically for aeroplane cockpit enclosures. Since then it has been used in almost every aspect of modern life. Further trade names have included Perspex®, Oroglas®, Acrylite®, and Lucite®.

PMMA is manufactured by polymerisation of methyl methacrylate (Figure 2.8). It is extremely clear and transmits approximately 92% of visible light. Optical qualities that permit edge lighting effects, exploited in Plexiglas® sculptures, are also apparent in unframed face- mounted prints, in that the colours of the photograph in the vicinity of the edge of the print are visible in the edge of the Plexiglas® when the mounted object is viewed from the side. With a Tg of 110° C, PMMA is a tough but brittle material, and is fairly rigid at room temperature. This makes it an ideal material for face-mounting, in that it gives even very large format photographs physical support and protects the print surface while not shattering easily as would a sheet of glass. Its main problem is its high susceptibility to scratching of the surface. It is available with various degrees of incorporated UV-absorbing filter materials.

Figure 2.8. Poly(methyl methacrylate) (Hochheiser 1986, 210)

PMMA sheets can be produced by casting or by extrusion, the latter of which incorporates high levels of stress during manufacture. To reduce internal stresses, the sheets may be annealed prior to their use (Sale 1995). The PMMA used by Lab X, Plexiglas® XT, is extruded. Upon ageing, it may be subject to stress-related cracking (Fenn 1995), accentuated by the presence of solvents in liquid or vapour form. Cast Plexiglas®, as is used by Grieger Düsseldorf, is less susceptible to this type of degradation.

Plexiglas® is moderately resistant to most chemicals, light, ozone, and is biologically stable. In the context of cleaning and the eventual treatment of prints face-mounted to PMMA, information on its solubility is given. Solvents for PMMA include: acetic acid, benzene, chloroform, chlorobenzene, cyclohexanol (hot), cyclohexanone, cyclohexyl acetate, dioxane, ethanol/water, ethanol/carbon tetrachloride, ß-ethoxy ethanol, ethyl acetate, formic acid, isobutanol (hot), isobutyric acid, isopropanol/methyl ethyl ketone 1:1 (>25°C), methylene chloride, methyl ethyl ketone, nitroethane, and xylene (Fuchs 1989, VII/382).

2.3.5. Double-sided Adhesive Film

The batch of samples that were face-mounted at the studio Lab Y in New York were shown to have a double-sided adhesive film between the photograph and the PMMA. The film used for mounting the tested samples was PermaTrans® IP2100, a product of MACtac USA.33 It was designed specifically for face-mounting and back-mounting photographic transparencies to clear substrates such as Plexiglas®. According to the technical literature, the film aids in protecting the mounted image from UV exposure, but the means are not explained. The polyester core film is coated on each side with a clear acrylic pressure-sensitive adhesive. The film with adhesive has a refractive index of 1.49. It is able to transmit 98% of white light and is stable in a temperature

Sylvie Pénichon, Art Institute of Chicago, personal communication, February 2001.

range of -29° to 93° C. The pH of the acrylic adhesive is neutral. The shelf life of the film is given as two years when stored at 24° C and 50% RH or less (MACtac, unknown date).

2.4. Laminate Structure

A cross-sectional analysis was undertaken for one of each of the four sample batches. Slivers of the face-mounted prints were sawn with an Isomet low speed diamond saw from the neutral patch located at CMY/5.34 The samples were embedded in Ward’s Bio-Plastic resin, which was allowed to cure. The samples were then cut and wet-ground to reveal thin-sections of the face-mounted prints. Observations were made with a Nikon S-KT stereo microscope at four magnifications: 50x, 125x, 250x, and 500x. For transmitted light observation, Köhler illumination was used with plane-polarized light. For reflected light examination, a fibre-optic lamp was used for illumination. Photomicrographs were taken in transmitted and reflected light with a Nikon Coolpix 990 digital camera. The silicone rubber tended to retract from the exposed cut edge of the cross-section, which formed a gap between the emulsion and the Plexiglas of the width of the silicone rubber layer. Debris from the sanding process was caught in this gap and adhered easily to the silicone rubber. It is partially visible in the cross-sections as white particles that are in or out of focus.

Figures 2.9. to 2.12. are reflected light photomicrographs of the four cross-sections of the face-mounted photographs. The visible layers are labelled and their approximate thickness is given. The first two are Diasec® prints that differ only in the choice of photographic paper, the first (Figure 2.9) being a paper for analogue exposure from a negative (Kodak Professional Ultra III Paper), and the second (Figure 2.10) a digital paper for exposure in a laser-diode printer such as a Lambda printer (Kodak Professional Digital III Color Paper). Both cross-sections show a typical laminate structure of a photographic RC paper, which consists of, from top to bottom:

See Figure 4.3 for details.

photographic emulsion, white pigmented polyethylene, paper base, and clear polyethylene. In direct contact with the emulsion and the bottom edge of the Plexiglas® sheet is the layer of silicone rubber, which in the first sample has a thickness of approximately 70 μm and in the second sample 100 μm. The primer on the surface of the emulsion and the Plexiglas® cannot be discerned as a layer, presumably because it is colourless and is applied very thinly. Along the bottom of the Plexiglas® a lighter band beneath a straight line can be discerned in Figures 2.9 and 2.10. This band was examined in reflected and transmitted light (Figure 2.13) and is thought to be an artefact of the reflection or refraction of light along the plastic edge. This interpretation cannot be confirmed, however no other explanation for its cause could be determined.

The digital paper mentioned above can also be observed in the third and fourth cross- sections. Figure 2.11 shows a print from the batch face-mounted with silicone rubber without a Diasec® license at Lab X in Düsseldorf. The photographic paper appears different in structure, but it has the same manufacturers markings on the verso and has been confirmed by the manager of the mounting studio to be Kodak Professional Digital III Color Paper.35 Its thick layer of pigmented polyethylene may be an irregularity in this area of the print. As apparent in this cross- section, the silicone rubber differs from the Diasec® prints in its bulk. Its approximate thickness of 300 μm may be a result of the use of less pressure on the cylinder press during mounting.

The fourth cross-section is from a print mounted at Lab Y in New York (Figure 2.12). It shows that the photograph was not mounted with silicone rubber as had been requested, but rather with a double-sided adhesive film. The adhesive can be seen adhering to the polyester core, and it proved to be tacky and viscous to the touch, indicating that it does not dry or set quickly, if at all. In the technical literature the thickness of the film with adhesive is given as 132 μm (MACtac, unknown date), however, in this cross-section it is measured at ca. 90 μm. This reduction in thickness is perhaps due to the pressure used in mounting the photograph.

Manager of Lab X, personal communication, January 2001.

Plexiglas®

silicone rubber photographic emulsion pigmented polyethylene

paper base

polyethylene

plastic sheeting (thickness: 4 mm)

adhesive (thickness: 70 μm)

photographic paper (thickness: 260 μm)

Figure 2.9. Cross-section of Diasec® mounted analogue photographic paper from Grieger Düsseldorf, reflected light

Plexiglas®

silicone rubber

photographic emulsion pigmented polyethylene

paper base

polyethylene

plastic sheeting (thickness: 4 mm)

adhesive (thickness: 100 μm)

photographic paper (thickness: 245 μm)

Figure 2.10. Cross-section of Diasec® silicone rubber mounted digital photographic paper from Grieger Düsseldorf, reflected light

Plexiglas®

silicone rubber

photographic emulsion

pigmented polyethylene

paper base

polyethylene

plastic sheeting (thickness: 4 mm)

adhesive (thickness: 300 μm)

photographic paper (thickness: 260 μm)

Figure 2.11. Cross-section of non-Diasec® silicone rubber mounted digital photographic paper from Lab X, reflected light

Plexiglas®

acrylic adhesive polyester film core

acrylic adhesive photographic emulsion pigmented polyethylene

paper base

polyethylene

plastic sheeting (thickness: 3 mm)

double-sided film (thickness: 90 μm)

photographic paper (thickness: 245 μm)

Figure 2.12. Cross-section of digital photographic paper mounted with double-sided adhesive film from Lab Y, New York, reflected light

line caused by refraction of light (?)

Very clear in transmitted light, and less obvious in reflected light, a distinct line is apparent along the bottom edge of the Plexiglas®, as can be seen in this view of the edge of the cross-section shown in Figure 2.10. The line is visible to a lesser extent along the top and side edges of the Plexiglas®, and it may be due to refraction of light along its edge. In cross-polarised light it does not appear different from the bulk of the Plexiglas®, and it can be observed in all four cross-sections. A final interpretation of this “layer” was not reached.

Figure 2.13. Cross-section of Diasec® mounted digital photographic paper from Grieger Düsseldorf, transmitted light

3. Experiment I: Chemical Analysis

3.1. Experimental
3.1.1. Materials and Equipment
Four silicone sealants and five primers used by three mounting laboratories were analysed

by FT-IR. These are listed in Table 3.1. It was not possible to obtain the sealant and corresponding primer from each mounting studio. Grieger Düsseldorf donated one tube of sealant, but was not willing to part with a sample of the primer due to its Diasec® license agreements. According to the manager, it is the Diasec® primer that is the protected secret ingredient of the mounting process36. It can be bought only under license conditions from Heinz Sovilla-Bruhlhart’s widow, Mrs. Sovilla, who also supervises its manufacture. It is available in unlabeled one litre plastic bottles, smells of organic solvents, and is a yellow liquid. The silicone rubber is available in unlabeled plastic tubes. It is colourless, translucent, and smells of acetic acid. It is manufactured in Germany by an undisclosed company recommended to licensees by Mrs. Sovilla.37

Source	Materials for Mounting
Grieger Düsseldorf Lab X, Germany Lab Y, New York, NY ® Diasec patent	Sealant: unlabeled silicone rubber Sealant: GE Silicone Construction SCS-1201 Primers: – unlabeled primer – Sallmetall “Duramount”, article no. 299000310 – Sallmetall “Catalyst”, article no. 5375700100000 Sealant: GE Silicone Contractors SCS-1001 Primer: GE Silicone SS4179 Sealant: Gurisil 575 (Jenny Co.) Primer: Gurisil PR435 (Jenny Co.)

Table 3.1. Sealants and primers for FT-IR analysis

Dieter Jung, Grieger Düsseldorf, personal communication, August 2000. Sylvie Pénichon: Personal communication with Mrs. Sovilla, February 2001.

The second German mounting studio, named Lab X in this research project, does not operate with a Diasec® license and consequently uses mainly commercially available products. The sealant is GE Silicone Construction SCS-1201, a transparent, colourless, acetic acid curing, one-part RTV silicone rubber which is commonly used to seal building joints for weatherproofing (GE Silicones 1998a). Three samples of primer were donated by Lab X. The first is for use only on the surface of the Plexiglas®. It is a yellow liquid that smells of solvents which is sold by an undisclosed provider in unlabeled one litre plastic bottles similar in shape to those of the Diasec® primer. Its name was not known by the manager of Lab X. The use of the second and third primers was discontinued by the mounting studio at the beginning of the year 2000. Both are products of the Dutch company Sallmetall38, which was bought by the Philadelphia based company Hunt Manufacturing Co. in March 1997. The products were part of a silicone rubber mounting set for photographs named “Duramount”. This set was later bought by DuroTech Corp., based in Richmond, Virginia, where it is now sold, presumably modified, as the “Durosil” system. The second primer, applied to the Plexiglas® sheet only, is labelled as Sallmetall “Duramount”, article no. 299000310. The president of DuroTech Corp. has stated that it is “most likely”39 identical to his company’s Primer WA-3, for which an MSDS is available (DuroTech Corp. 1997). The third primer, designated for application to the photographic emulsion only, is the Sallmetall “Catalyst”, article no. 5375700100000, which may be identical to DuroTech Catalyst CP-48-R,40 for which an MSDS can also be consulted (DuroTech Corp. 1999).

Lab Y, a photographic laboratory and mounting studio in New York, NY, uses GE Silicone products for face-mounting prints. The sealant used is GE Silicone Contractors SCS-1001, which has identical ingredients to the sealant SCS-1201 used at Lab X, but differs in the percentage

manufacturing equipment, paper, boards, vacuum presses, cutters, laminators, and adhesive films for mounting and

laminating.

Sallmetall was a laminating systems and materials company that sold accessories for frames, moulding

Hans Schuurmans, President, DuroTech Corp., personal communication, December 6, 2000. Ibid.

ratio of its components (GE Silicones 2000b). It is designed for general glazing and sealing applications (GE Silicones 1997a). To prepare both the surface of the photograph and the acrylic sheet, the colourless primer GE Silicone SS4179 is used.

For the purpose of comparison, samples of the sealant and primer specified in the 1977 patent by Sovilla-Bruhlhart were acquired from the manufacturer, Jenny Co. AG, Switzerland. These are the one-component RTV sealant Gurisil 575 and the primer Gurisil PR435 (Sovilla- Bruhlhart 1977). Their composition may have been altered since their use in the 1970’s. As opposed to the other examined sealants, Gurisil 575 contains an amine moisture curing PDMS.

The samples were analysed on the Avatar 320 TM E.S.P.TM FT-IR spectrometer with a transmission baseplate in the laboratories of the Art Centre Extension Building at Queen’s University. Specifics are detailed in Appendix I. The spectrometer was set to average 16 scans at a resolution value of 4 in the transmission mode.

3.1.2. Procedure

The samples were prepared and examined by transmission analysis in four test runs. In the first, the sealants were examined in an uncured state, which allowed a direct comparison of the IR spectra of the sealants of unknown ingredients to those of the GE Silicone sealants. Each of the four sealants was examined with transmission spectroscopy before it cured. A small bead of the sealant was placed on a sodium chloride (hereafter NaCl) plate and spread to a thin film with a glass rod, then rapidly analysed. The plate was set aside overnight to allow the sealant to cure fully for the second test run, in which the cured sealant was analysed (Figure 3.1). In the third test run a drop of each primer was sandwiched between NaCl plates to form a thin film for transmission analysis. The fourth run consisted of an analysis of the active ingredients of the primers. One drop of each solution was placed on separate NaCl plates. The solvents were allowed to evaporate for 30 minutes, leaving a solid residue on the plate. To enable their

differentiation, the samples were labelled with a code that contains three separate parts. These are explained in Table 3.2.

Figure 3.1. Cured silicone rubber film on NaCl plate

Source	Type		Sample Preparation and FT-IR Run
A = Grieger B =Lab X C = Lab Y ® D =Diasec patent	S = sealant P =primer (P1-3 for B only)	–	1 = uncured sealant film on or between NaCl plates 2 = cured sealant film on NaCl plate 3 = solvent primer film between NaCl plates 4 = evaporated primer film on NaCl plate

Table 3.2. Labelling system for FT-IR Samples

3.2. Results and Discussion
As was expected, the spectra were slightly difficult to interpret in great detail as the analysed

materials contained complex mixtures of different substances that cannot easily be differentiated with IR spectroscopy, as stronger absorption bands may obscure weaker ones. The individual samples, their descriptions and analysis methods are listed in Tables 3.3 and 3.4. Some general statements can be made about the FT-IR spectra, which are shown in Figures 3.2 to 3.5. The spectra that have been baseline-corrected are marked with an asterix (*). No Si-H bonds, which typically show strong and sharp absorption bands at 2130 cm-1, could be detected in either the

primers or sealants. Silicon-chlorine bonds could not be detected, because their absorption bands are mainly below 600 cm-1, and the NaCl plates used for the analysis cut off the IR spectrum at ca. 650 cm-1. Interpretation of the spectra was possible with the help of Dr. H.F. Shurvell, Professor Emeritus of Chemistry at Queen’s University, and the references in Noll (1968), Anderson (1974), and Sigma-Aldrich (1997).

Table 3.3. Individual sample preparation of sealants for FT-IR analysis

Code	Sample description	41 Ingredients from MSDS, if available
AS-1 BS-1 CS-1 DS-1	Grieger uncured no-name sealant film Lab X uncured GE SCS-1201 sealant film Lab Y uncured GE SCS-1001 sealant film ® Diasec patent uncured Gurisil 575 sealant film	unknown methyltriacetoxysilane: 1-5% octamethylcyclotetrasiloxane: 1-5% polydimethylsiloxane silanol/stdp: 60-80% silanol/stdp siloxane w/me silsqxns: 10-30% treated fumed silica: 10-30% methyltriacetoxysilane: 1-5% octamethylcyclotetrasiloxane: 1-5% polydimethylsiloxane silanol/stdp: 60-80% silanol/stdp siloxane w/me silsqxns: 5-10% treated fumed silica: 5-10% tris-alkylaminosilane: <5%
AS-2 BS-2 CS-2 DS-2	Grieger cured no-name sealant film Lab X cured GE SCS-1201 sealant film Lab Y cured GE SCS-1001 sealant film ® Diasec patent cured Gurisil 575 sealant film	polymerised AS-1 polymerised BS-1 polymerised CS-1 polymerised DS-1

The ingredients are listed as found in the literature consulted: DuroTech Corp. 1997, 1999; GE Silicones 1998b, 2000b, 2000c; Jenny Co. AG 1999b, 2000b.

Table 3.4. Individual sample preparation of primers for FT-IR analysis

Code	Sample description	42 Ingredients from MSDS, if available
BP1-3 BP2-3 BP3-3 CP-3 DP-3	Lab X no-name primer in solvent Lab X Sallmetall Duramount primer in solvent Lab X Sallmetall Catalyst primer in solvent Lab Y GE SS4179 primer in solvent ® Diasec patent Gurisil PR 435 primer in solvent	unknown toluene: <20% isopropanol: <50% perchloroethylene: <15% active ingredients (trade secret): <10% toluene: <15% isopropanol: <45% acetone: <20% perchloroethylene: <10% n-butanol: <5% active ingredients (trade secret): <15% benzene: < 1 ppb methanol: 1-5% toluene: < 1 ppm ethyl acetate: 80-99% 2-propenoic acid, 2-methyl-, methyl-ester, polymer w/ 3-(trimethoxysilyl) propyl 2-methyl, 2-propenoate: 5-10% silicone resin: 1-2% butanone: >20% toluene: >12.5% active ingredient unknown
BP1-4 BP2-4 BP3-4 CP-4 DP-4	Lab X evaporated no-name primer residue Lab X evaporated Sallmetall Duramount primer residue Lab X evaporated Sallmetall Catalyst primer residue Lab Y evaporated GE SS4179 primer residue ® Diasec patent evaporated Gurisil PR 435 primer residue	unknown active ingredients (trade secret) active ingredients (trade secret) 2-propenoic acid, 2-methyl-, methyl-ester, polymer w/ 3-(trimethoxysilyl) propyl 2-methyl, 2-propenoate (polymerised?) silicone resin: 1-2% active ingredient unknown

Ibid.

The spectra of the uncured sealants, analysed in the first run, all showed certain similarities (Figure 3.2). Especially the top three spectra, those of the acetoxy-curing sealants, differ only in degree of absorption, not in the position of the absorption bands. This may be due to either the varying thickness of the samples, varying amounts of ingredients in each, or both. The major peaks of these spectra, as exemplified for sample BS-1, the uncured GE Silicone 1201 sealant, are summarised in Table 3.5. The main chemical groups that would make up the compounds listed in Table 3.3 are present. These are hydroxyl and methyl groups, silicon-carbon bonds and siloxanes. Typical for organosiloxanes are a sharp peak around 1260 cm-1 (SiCH3) and a broad band that often has two peaks in the region of 1130-1000 cm-1 (Si-O-Si). A carbonyl and a methyl group characteristic of acetyl groups were also found, and were interpreted to be a part of the crosslinking agent methyltriacetoxysilane.

-1 Region or peak (cm )	Assignment	Comments
3760-3580 2963, 2905 1745 1445 1412 1371 1262 1130-1000 864, 797 701, 664	OH stretch CH3 stretch C=O stretch CH3 asym. deformation SiC CH3 sym. deformation Si-CH3 Si-O-Si stretch SiC CH3 rock	hydroxyl groups attached to the silanes, siloxanes, and fumed silica methyl groups attached to the silanes and siloxanes carbonyl of acetyl group in methyltriacetoxysilane methyl groups attached to the silanes and siloxanes methyl groups attached to the silanes and siloxanes methyl group attached to the acetyl group in methyltriacetoxysilane methyl groups attached to the silanes and siloxanes siloxane backbone of PDMS polymer methyl groups attached to the silanes and siloxanes methyl groups attached to the silanes and siloxanes

Table 3.5. Assignment of IR frequencies in the FT-IR spectrum of sample BS-1, uncured GE Silicone 1201 sealant (second from top in Figure 3.2)

The fourth spectra, that of DS-1, uncured Gurisil 575 sealant, differs from the other three mainly in that the carbonyl group at 1745 cm-1 is not present. As it is an amino-curing PDMS, acetyl groups were not expected. The presence of an amino group could not be seen, presumably due to the fact that the tris-alkylaminosilane contains a tertiary amine which does not have a

characteristic vibration. A very slight shoulder at 1156 cm-1 may be due to a C-N bond. Although the uncured sealant smells of ammonia, no evidence of NH3, an absorption band in the region of 3300-3200 cm-1, is apparent in the spectrum. The very strong bands due to Si-O-Si stretching at 1093 and 1021 cm-1 indicate a linear PDMS polymer (Anderson 1974, 277).

Figure 3.6. Comparison of the FT-IR spectra of BS-1, uncured, with BS-2, cured GE Silicones 1201 sealant

Figure 3.7. Comparison of the FT-IR spectra of BP1-3, no-name primer in solvent, with BP3-3, Sallmetall “Catalyst” primer in solvent

The spectra of the cured sealants are shown in Figure 3.3. The only difference to the spectra of the uncured acetoxy samples, AS-1, BS-1, and CS-1, is the absence of the peaks in the cured samples at 1745 and 1371 cm-1, corresponding to a carbonyl and a methyl group respectively. These are associated with the acetic acid that has been given off during curing. A comparison of the uncured with the cured GE Silicone 1201 sealant is given in Figure 3.6. The methyl group at 1371 cm-1 in DS-1, the uncured amino-curing sealant, is also not present in its cured form, DS-2 (Figure 3.3). In addition to this, the C-N shoulder at 1156 cm-1 is not visible, indicating the loss of an alkylamino compound. This would correspond to the assumed curing mechanism of the sealant as described in section 2.3.2, “Silicone Sealants”.

Typical absorption bands for silica, those of Si-O-Si at 1065 and 806 cm-1, could not be seen in any of the sealant spectra. It was assumed that the peaks were hidden under those of the PDMS. In general, the spectra of all sealants closely resembled that of PDMS as shown in the Aldrich collection (Sigma-Aldrich Co. 1997, 4635C). As the spectrum of the no-name sealant used by Grieger Düsseldorf in the Diasec® process matched those of the sealants with known ingredients, it may be assumed that it is fundamentally the same material, though it may vary in content by small amounts of additives or ratios of ingredients.

As has been shown in section 2.3.3, “Primers”, the compounds found in silane primers are very complex. FT-IR was not expected to enable full characterisation of these materials. The spectra of the five primers in solvent, as listed in Table 3.4, are given in Figure 3.4. No absorption bands relating to bonds with silicon were found in these spectra, presumably because the quantity of solvent vehicle greatly exceeded that of the active ingredients. The spectra of the silanes were obscured by that of the solvent mixture. The spectra of the Sallmetall primers, coded BP2-3 and BP3-3, showed no sign of toluene content, as there are no peaks indicating aromatic C-H bond structures just above 3000 cm-1. It had been speculated that the Sallmetall “Duramount” primer was identical to the DuroTech Primer WA-3, and that the Sallmetall “Catalyst” primer was the same as DuroTech Catalyst CP-48-R. The DuroTech products both contain toluene, however, as

specified in the MSDS (DuroTech Corp. 1997 and 1999), so it was concluded that the Sallmetall materials differ from the DuroTech primers. The spectrum of BP2-3 matches almost perfectly that of iso-propanol (Sigma-Aldrich Co. 1997, 170B), with the exception of the small sharp peak at 713 cm-1, which may correspond to a C-Cl bond found in a chloroethylene. The characteristic strong C-Cl peak around 910 of tetrachloroethylene (Sigma-Aldrich Co. 1997, 145C), however, is not present. BP3-3 also contains an alcohol.

The spectrum of CP-3, the GE Silicones SS4179 primer in solvent, is a perfect match for ethyl acetate (Sigma-Aldrich Co. 1997, 948A). This result confirms the ingredients as listed in the MSDS (GE Silicones 2000c), namely 80-99% ethyl acetate. Here again, it is assumed that the active ingredients, which make up 5-10% of the primer, and are apparent in the spectrum of the evaporated primer (CP-4 in Figure 3.5) are merely obscured by the strong bands of the solvent in CP-3. The spectrum of the Gurisil primer in solvent, DP-3, shows similarities to those of both toluene, with aromatic C-H groups above 3000 cm-1 and characteristic peaks at 731 and 696 cm-1, and methyl ethyl ketone (Sigma-Aldrich Co. 1997, 1626B and 640C), both of which are specified as solvent vehicles in the list of ingredients (Table 3.4).

Of the spectra of the four known primers described above, the spectrum of the no-name primer BP1-3 most closely matches that of BP3-3, the Sallmetall “Catalyst”. The two primers in solvent are compared in Figure 3.7. The peak at 1710 cm-1 in the no-name primer BP1-3, which is not present in BP3-3, may be the carbonyl group of a ketone, which is usually around 1720 cm-1. The peak at 910 cm-1 may indicate the presence of a chloroethylene solvent. In conclusion, it is reasoned that the no-name primer is very similar to the Sallmetall “Catalyst” and contains an alcohol and perhaps a ketone and a chloroethylene.

The spectroscopic analysis of the residues of the evaporated primers gave the results shown in Figure 3.5. Due to the technique of sampling the products, the residue film was relatively thick compared to the primers in solvent, as one or more large droplets were placed on the NaCl plate and allowed to evaporate. On the other hand, the primers in solvent were held between two NaCl

plates, resulting in an extremely thin “film” of material for analysis. This difference in quantity may be the reason for the invisibility of the silicon-related peaks in the spectra of the primers in solvent, and their strong absorption when analysed as a residue.

Comparison of BP1-4, the residue of the no-name primer, with BP3-4, the Sallmetall “Catalyst”, shows that both have identical spectra, differing only in the degree of absorption. Both are relatively simple compounds that contain siloxanes (Si-O-Si), as indicated by the double peaks between 1130 and 1000 cm-1, and methyl groups attached to silicon, with peaks around 1270 and 768 cm-1. The smaller absorption peaks just below 3000 cm-1 also indicate aliphatic C- H groups. Slight but broad absorption centred around 3400 cm-1 indicates the lingering presence of the hydroxyl groups of the alcohol solvent that was so strongly apparent in the spectra of the primers in solvent (Figure 3.4).

The spectrum of BP2-4, the Sallmetall “Duramount” primer after evaporation, showed no major absorption after one drop had been applied and allowed to evaporate, following the experimental procedure. Only after 15 drops at the same location with intermediate evaporation did some peaks begin to appear on the spectrum. As can be seen in Figure 3.5, however, these are simply aliphatic C-H groups that are most likely residual traces of solvent. This primer contains no active ingredients, but is instead purely a solution of mainly iso-propanol possibly with small amounts of other solvents. In this, it differs from the ingredients given for DuroTech Primer WA- 3, thought to be identical with it, which have listed “active ingredients (trade secret): <10%” (DuroTech Corp. 1997).

The residues of CP-4, GE Silicones SS4179 primer, contain no Si-O-Si, as the typical broad peaks above 1000 cm-1 are not present. The MSDS of this product lists the active ingredient as the complex compound “2-propenoic acid, 2-methyl-, methyl-ester, polymer w/ 3- (trimethoxysilyl) propyl 2-methyl, 2-propenoate” (GE Silicones 2000c). The strong peak at 1731cm-1 may be the carbonyl group of the 2-propenoic acid, and the absorption at 1271 cm-1 is the Si-CH3 encountered in all primer residues but that of BP2-4, the pure solvent mixture. The

DP-4 residues, those of the Gurisil PR 435 primer, still have traces of toluene, even after the prolonged evaporation time of 24 hours. Aromatic peaks at 3072 and 3050 cm-1 as well as small peaks at 1962, 1891, and 1823 cm-1, overtones characteristic of a monosubstituted benzene, and peaks at 700 and 743 cm-1 indicate toluene. A Si-O-Si peak is present at 1135 cm-1, the form of which indicates a sesquisiloxane (CH3SiO3/2) (Anderson 1974, 277).

4. Experiment II: Dark Ageing in Intense Environmental Conditions

4.1. Experimental
4.1.1. Materials and Equipment
All samples tested in this experiment consist of chromogenic photographic paper that has a

grid of colour patches printed on it. The origin of the image is a 400 dpi digital file in RGB colour space produced with the software Adobe Photoshop 6.0. The prints were exposed, developed, and fixed conventionally, then one half of the batch was face-mounted and the other half was left unmounted. The image was printed multiple times onto larger sheets which had to be cut down to individual size. The sets of samples were produced in four laboratories43 in December 2000, and were sent to the Art Institute of Chicago for cutting before they were then sent on to Queen’s University. The unmounted prints were cut with a knife and straightedge on a cutting mat (Figure 4.1), and the face-mounted prints were sawn through the print side using a table saw (Figure 4.2). Immediately following their production, the face-mounted samples were left to set for a minimum of 24 hours, then packed and shipped. Although each set of prints was packed in an individual plastic bag during most of this time, the samples were removed periodically for cutting, cleaning, and counting, which allowed them to air and offgas to a certain degree. The precise dates of the individual printing of the samples are unknown, but on average the face-mounted samples were bagged and sealed for ageing two months after their production.

Specifics of these laboratories are listed in Appendix II.

Figure 4.1. Cutting unmounted test samples

Figure 4.2. Sawing face-mounted test samples

The first laboratory, Grieger Düsseldorf, supplied two sets of Diasec® samples. As most fine art photographers that employ the services of Grieger Düsseldorf ask for their prints to be made from photographic negatives rather than digital files, a set of test targets was made on analogue paper from a colour negative that had been exposed digitally from the original file. The second set were printed directly digitally to a photographic paper on a Durst Lambda printer. Both materials are RC papers with type F (glossy) surface manufactured by Kodak, but they differ in their sensitivity and printing characteristics. The digital paper is Kodak Professional

Digital III Color Paper, and the analogue paper is Kodak Professional Ultra III Paper. Both were developed in a Sitte Constamat R processor with the Agfa process AP 94. The Diasec® samples were mounted to a cast 4 mm (± 0.4 mm) thick PMMA sheet with an incorporated UV barrier of either 85% or 100% effectivity. The brand name and manufacturer of the acrylic sheeting was not disclosed. A total of 29 unmounted and mounted samples of the analogue paper, and 29 unmounted and 33 mounted samples of the digital paper were needed for the experiments.

The second German laboratory, given the pseudonym “Lab X”, is a non-licensed studio that offers photographic printing and face-mounting as a part of its services. The samples were printed on Kodak Professional Digital III Color Paper with type F (glossy) surface with a Durst Lambda 130 printer, then developed in an Autopan or a Hostert Pro processor following Kodak recommendations.44 The prints were mounted to 4 mm thick extruded Plexiglas® XT. The manufacturer is Röhm GmbH, Darmstadt, Germany. For the experimentals, 29 unmounted and 33 mounted samples of the Lab X samples were needed.

The last set of samples was produced jointly by two US-American studios. Gamma, a Chicago based laboratory, printed the targets, and they were then sent to New York for face- mounting at the studio Lab Y. The targets were printed on Kodak Professional Digital III Color Paper with type F (glossy) surface with an undisclosed printer. They were face-mounted to 1/8 inch (3 mm) thick Acrylite® FF sheeting, manufactured by Cyro Industries, a US-American daughter company of Röhm GmbH. Acrylite® is the registered brand name for Plexiglas® in the USA. For the experimentals, 29 unmounted and 33 mounted samples were used. Due to in-house miscommunication, the samples were mounted with MACtac PermaTrans® IP2100 double-sided adhesive film instead of silicone rubber. Specifics of this film are listed in section 2.3.5, “Double- sided Adhesive Film”.

For density measurements, an X-Rite 810 (status A, reflected light) densitometer, situated at the photographic supply store Camera Kingston, was used. The laminate bags were sealed with

a Clamco Corp., Cleveland, OH, heat sealer at the Anatomy Laboratory, Queen’s University. The samples were aged in ovens with recirculating forced air and controlled temperature in the laboratories of the Art Centre Extension Building at Queen’s University. These were: National Appliance Company (Napco) Model 58401, Fisher Isotemp® Oven 200 Series, and a Lab-Line Imperial II Radiant Heat oven. The control samples were frozen in a Viking refrigerator (unknown model) with a freezer compartment. The utensils included 252 18×13 cm nylon- aluminum-polyethylene laminate bags, manufactured by MACO Bags, Inc., Victor, NY, 112 14×8 cm sheets of ArchivArt silicone release film C2S (0.015 caliper), a Mannix digital thermohygrometer, a Sanford Sharpie black permanent marker, cotton gloves, binders’ board, and PVAC glue.

4.1.2. Procedure

ANSI standard IT9.9-1996 gives precise instructions for the accelerated ageing of colour photographic materials, and these were followed in the experiment. Due to the novel physical structure of the face-mounted test samples, however, some deviations in procedure were decided upon. The deviations were of an extent that it was hoped would not influence the validity of the tests. The quantities of samples and their designated functions in the experiments are listed in Tables 4.1. and 4.2.

Table 4.1. Unmounted samples for accelerated ageing

Total:

Source

Control samples 7 Samples for ageing at 4 temperatures Total

Grieger: analogue Grieger: digital Lab X Gamma/Lab Y

1 7 x 4 = 28 29 1 7 x 4 = 28 29 1 7 x 4 = 28 29 1 7 x 4 = 28 29

116

Manager of Lab X, personal communication, January 2001.

Source

Control samples with AD- 7 Samples for ageing 1 Sample with Strip and glass vials at 4 temperatures AD-Strip and glass vial at 4 temperatures

1 Sample with Total AD-Strip and
glass vial at room
temp.

Grieger: analogue

Grieger: digital Lab X Gamma/Lab Y

1 without vial and AD-Strip (no GC)

1 1 1

7 x 4 = 28

7 x 4 = 28 7 x 4 = 28 7 x 4 = 28

1 x 4 = 4 1 x 4 = 4 1 x 4 = 4

0 29

1 34 1 34 1 34

131

Total: Table 4.2. Face-mounted samples for accelerated ageing and acetic acid off-gassing tests

Following the specifications given, the 14.5 x 9 cm test target (Figure 4.3.) was designed to have colour patches in pure yellow, magenta, and cyan dyes, a patch of minimum density showing the white of the substrate without any colourants in the emulsion, as well as patches that contained all three colours to form neutral grey (ANSI 1996, 2). Measurements in the colour patches were designated to determine the fading of individual dyes, whereas those in the neutral patches showed colour shifts caused by the uneven fading of the different dyes. Measurements in the dmin patch were designated to record the typical yellow staining found in aged colour photographs, attributed to the discolouration of unreacted magenta colour couplers in the emulsion (Wilhelm 1993, 168). The co-ordinates of a scale of 1-7 on the x-axis combined with the column names on the y-axis which indicate the colour dyes allowed for a designation of each patch. The colours were printed as a scale of seven 1.4×1.4 cm patches which, in the originating digital file, began at 12.5% colour saturation and each had an increment of 12.5% to reach a total of 100%. A 1.7 cm wide white border around the colour patches was designated for the handling of the target.

Figure 4.3. Test target design

The samples were numbered in the bottom right field with a permanent black marker with a code (Table 4.3.), then selected patches were measured with the densitometer (Figure 4.4.). An area of dmin (minimum density), coded as “CMY/0”, and the selected patches of each colour and of neutral grey that had a 1.0 ± 0.05 density units increase over dmin were used for measuring in the tests. The grey and colour patches that fit this criterion were determined by preliminary trial and error density measurements. The patches chosen for each set of prints varied, as the targets were printed with slightly differing colour casts and lightness, and the face-mounted prints were

darker than the unmounted ones due to the absorption of light by the acrylic sheeting. The designated patches and the measured values are detailed in Appendix IV, “Densitometry Measurement Data”.

Source	Type		Temperature		Sample No.	Acetic Acid Control
A1 = Grieger: analogue A2 = Grieger: digital B = Lab X C = Gamma/Lab Y AD-Strip	U = unmounted M =mounted	–	0 = control 20 = room temp. 55 = 55° C 65 = 65° C 75 = 75° C 85 = 85° C	–	1-7 = samples for ageing 8-11 = extra samples for second round of ageing at 75° C	AD = contains AD-Strip GC = for GC analysis

Table 4.3. Labelling system for samples for accelerated ageing and acetic acid off-gassing tests

Figure 4.4. Use of the densitometer on an unmounted target

In order to exclude the influence of stray ambient light penetrating the front and the edges of the Plexiglas® of the face-mounted samples during measurements, these were placed in a light- tight tray constructed of binders’ board edges adhered to the black side of a piece of the laminate bags. A cover of the same material with 7 mm2 windows cut out for the densitometer head was placed over the target for measurement (Figure 4.5).

Figure 4.5. Use of the densitometer on a face-mounted target

All samples were conditioned to 55% RH for 7 days in the storage vault of the Agnes Etherington Art Centre, Kingston, which has a controlled environment with a measured fluctuation of ± 3% RH (Figure 4.6) during this period. Following this, they were sealed individually in labelled nylon-aluminum-polyethylene laminate bags with the heat sealer (Figure 4.7).45 A sheet of silicone release film was placed on the emulsion side of the unmounted samples to prevent the gelatine from adhering to the interior of the bags during ageing.46 The two main factors that influence dark storage behaviour of colour photographs are temperature and relative humidity of the immediate environment of a print. High values of each will accelerate the degradation of the image-forming dyes (ANSI 1996, ii). As the conditioned samples were in a

The standard calls for multiple samples in one bag, but it was felt that in the hypothetical case of acetic acid off- gassing from the silicone rubber on the face-mounted prints, the amount per bag should be reduced to avoid it influencing the ageing procedure. The filler material also recommended in the standard was left out, as a “stacking” situation was not being simulated here. The samples were not double bagged as recommended in the standard due to

a significant cost factor.

This was recommended by Doug Nishimura, Image Permanence Institute, Rochester, NY.

sealed environment, the humidity content of the emulsion remained constant (ANSI 1996, 6). Therefore the only variable during ageing was the temperature.

Figure 4.6. Conditioning the samples

Figure 4.7. Sealing the samples in bags

The sealed control test targets were placed in the freezer and kept at –15° C for the duration of the tests. The samples for ageing were separated into four groups, each of which contained unmounted and face-mounted samples from all three laboratories. Each group was placed in a

separate oven set at constant temperatures of 55, 65, 75, and 85°C (Figure 4.8). A control of RH inside the ovens was not necessary, as the moisture content of the interior of the sealed bags could not change. The samples of each group were aged until they had reached their designated endpoint in density change (∆d). As a progress control, samples of each mounting process were taken from each oven consecutively over the total ageing period to open the bags and measure ∆d of the target patches. Next to the densitometry measurements, aged samples were examined visually by comparison to a control sample. They were also examined for irregularities such as warping or delamination. Once removed from the sealed bags, the samples were not re-bagged for further ageing.

Figure 4.8. Bagged samples in an oven

The degree of change that is acceptable in a colour photograph upon ageing can depend on the colour perception of individual viewers, on the image, and on the tonal values of the print (ANSI 1996, iii). In the test parameters described in the standard, the endpoint for the changes in dye densities is determined differently for each colour patch. The density terminology and calculations used in this context is explained in Appendix III. Endpoint for the neutral patch was a 30% change of DN(R), DN(G), and DN(B); for the colour patches was 30% change of DC(R),

DM(G), and DY(B); for colour balance in the neutral patch was 13%47 change of dN(R-G), dN(R- B), or dN(G-B); for minimum density patch was 0.10 density units change of dmin(R), dmin(G), and dmin(B); and for colour balance in the minimum density patch was 0.06 density units change of dmin(R-G), dmin(R-B), and dmin(G-B). Once the samples had passed their endpoints, the ageing was stopped.

4.2. Results and Discussion
During ageing, some problems were encountered that changed the scope of the

experimental. After 20 days, the samples designated for ageing at 75° C were found to be discolouring at a rate higher than that found in the samples at 85° C. Upon investigation, the air circulation in the oven set at 75° C was shown to be defect, resulting in areas of different temperatures. The end of the thermometer had been measuring 75°, whereas the bulk of the samples were in an area later measured to be 93° C. For this reason, all 75° samples were not evaluated in the experiment. Four extra samples of each batch could be labelled, bagged and sealed for a second round of ageing in a different oven set at 75° C. These samples were given the numbers 8-11 in their code to differentiate them from the void ones. The first samples in the 85° C oven were removed and measured after 20 days. Unfortunately, this delay proved to have been sufficient for the prints to have changed to such an extent that the endpoints had been passed by a great amount. For this reason, it was decided not to include the 85° C samples in the results, but to regard them as examples of extreme degradation.

The density measurement data and calculated values for the remaining samples are given in Appendix IV. The results are reported and discussed in terms of comparison of the unmounted with the face-mounted prints of each batch, and in terms of the different ageing characteristics between each face-mounted batch. As extrapolation of ageing data following the Arrhenius

This endpoint was reduced from the 15% recommended in the standard (ANSI 1996, 18), as it was felt that the visual changes were sufficient at 13% to define them as an endpoint.

method was not carried out,48 the time to reach the endpoint for each batch is only reported in the results where it is used for comparison of two or more batches. In general, samples in the oven set at the highest temperature reached their designated endpoint fastest, those at the lowest temperature slowest.

For the most part, the measured changes could be visually confirmed, as the change in appearance of the samples increased as ageing progressed. Irregular values were sometimes measured for the face-mounted prints, so the same patch was often measured multiple times to calculate an average. This irregularity may have been caused by the measuring set-up for the samples. A further possible source for irregular ageing curves for both unmounted and mounted samples was the fact that each point on a curve is measured from a different sample, since once the sample has been removed from it sealed environmnet it should not be resealed for further ageing. This procedure not only increased the number of samples needed for the experiment, but also meant that no single sample was measured in its changes throughout its ageing. A further source for irregular readings may be the fact that the samples were conditioned at 55% RH, but gelatine has been shown to change from a solid to a gel state when above 50° C and 50% RH (ANSI 1996, 6). This may have altered the way in which the photographic images deteriorated.

During ageing, it could be seen that the endpoints that were approached fastest were two types of colour shifts in the neutral density patches. A colour shift towards red, described by dN(R-G)t, is shown in graph form in Figures 4.9 to 4.12. A shift towards violet, given by dN(G-B)t, is plotted in Figures 4.12 to 4.15. These two colour shifts were visually the most obvious changes in the samples. The calculated values for dN(R-B)t showed no significant changes and were not plotted. Colour shifts due to different rates of fading of the three image- forming dyes are typical degradation manifestations for chromogenic prints stored in the dark (Wilhelm 1993, 164-165).

As explained in section 1, “Introduction”.

The extent of yellow staining of the minimum density patch is shown in Figure 4.17. The graph shows the difference in dmin(B) values after 58 days at 65° C for all batches. Increases in the dmin(B) measurements were more prominent than the measurements of dmin(R) and dmin(G). Figures 4.18 to 4.20 are graphs of changes in dye density in the cyan, magenta, and yellow coloured patches of samples aged at 65° C. Changes in saturation of the colour patches were the least visually obvious. Figures 4.17 to 4.20 are exemplary graphs in that they show only values for the 65° C batches, which were felt to be representative. It can be deduced from all graphs that the dye density appears to increase (positive % values) before it begins to decrease (lower positive % values and negative % values). This result is surprising at first, since it is assumed that the dyes will fade at high temperatures. It has been observed in many accelerated ageing tests of photographic emulsions, however. These high measurements are presumably the result of physical changes in the heated gelatine that cause both aggregation of the dyes due to their increased mobility and alterations in the emulsion surface.49 The decreasing values that follow this initial rise describe the actual fading of the dyes.

Doug Nishimura, Image Permanence Institute, Rochester, NY, personal communication, May 9, 2001.

Results for A1 Samples (Grieger Düsseldorf analogue paper)
Figures 4.9 and 4.13 show that the red and violet colour shifts in the unmounted and face-

mounted A1 samples occurred at a similar rate and to a similar degree. The curves of the unmounted samples, shown with a dotted line, are relatively close to those of the face-mounted samples, shown with a solid line. An exception is the slower red shift in the unmounted sample aged at 55° C (A1U) in Figure 4.9. The data of each temperature shows that the face-mounted prints aged slightly faster than their unmounted counterparts. The changes in Dmin(B) after 58 days at 65° C show that the white base of the unmounted prints yellowed to a greater extent, at 0.05 density units, than that of the face-mounted prints, at 0.01 units (Figure 4.17). Finally, Figures 4.18 and 4.19 indicate that there was no significant fading of the cyan and magenta dyes in the colour patches. The yellow dye, however, faded in both unmounted and face-mounted to the extent that it is visually distinguishable. This dye loss is shown in Figure 4.20. In one 85° C sample a large air bubble appeared between the print and the Plexiglas® (Figure 4.21), but no other cases of delamination were encountered.

Figure 4.21. Formation of a bubble and delamination in a sample from Grieger Düsseldorf aged for 20 days at 85° C

Results for A2 Samples (Grieger Düsseldorf digital paper)
The most obvious colour shift in the neutral grey patches of the A2 samples was towards

red (Figure 4.10). It can be seen that the unmounted prints changed more rapidly than their face- mounted counterparts at each temperature. At 75° C, for example, the unmounted print reached its endpoint of 13% after 20 days, but the face-mounted sample only after 45 days, the duration at which the 65° C unmounted print had already reached its endpoint. Shifts towards violet as shown in Figure 4.14 were inconclusive. The dmin patch of the unmounted prints yellowed to a much greater extent than that of the face-mounted prints. The unmounted sample aged for 58 days at 65° C measured a difference in dmin(B) of 0.06 density units, whereas the face-mounted sample remained at 0.01 (Figure 4.17). Dye losses in the colour patches were insignificant, as shown in Figures 4.18, 4.19, and 4.20. When aged samples are compared to the control, the colour patches cannot be visually distinguished.

Small areas of delamination that appeared “streaked” were apparent in all 75° C samples. This pattern may be a result of the method of application of the primer, in which a chamois cloth is used to distribute the solvents over the Plexiglas® surface. In this fashion, some areas may not have received any or enough primer and were therefore more likely to suffer from adhesion failure.

Results for B Samples (Lab X)
Although the neutral patches underwent both red and violet shifts (Figures 4.11 and 4.15),

the latter was dominant. Both unmounted and face-mounted samples aged rapidly at a similar rate and to a similar degree, although the shifts in the face-mounted prints were slightly more pronounced. The dmin(B) measurements of 65° C samples show yellowing in both unmounted and face-mounted prints, those of the former reaching a difference of 0.07 density units, and those of the latter 0.04 after 58 days (Figure 4.17). The 65° C face-mounted samples show highly irregular

curves for dye fading in cyan and magenta colour patches (Figures 4.18 and 4.19), but no changes are apparent in visual comparison. The unmounted prints did not suffer from significant cyan, magenta, or yellow fading. The yellow patches on the face-mounted prints faded severely, as is demonstrated in Figure 4.20 and is visually apparent when comparing the aged samples with the controls.

Migration of the yellow dye became apparent along the black lines in the face-mounted prints that were aged at 65° C and above. In those samples aged at 85° C, migration of the magenta dye is visible (Figure 4.22). The unmounted samples showed no dye migration. All face- mounted samples subjected to accelerated ageing suffered from horizontal convex warping of the Plexiglas® (Figure 4.23). The longer and hotter the samples were aged for, the stronger the migration and warping became. The distortion may be a result of the use of extruded instead of cast Plexiglas® for these samples, as it is known to have internal stresses from the manufacturing process (Sale 1995) which might make it less stable at higher temperatures. Its possible tendency to warp at lower temperatures such as those found in common storage environments would have to be tested in long-term natural ageing experiments.

Figure 4.22. Migration of the magenta dyes in a sample from Lab X aged for 20 days at 85° C

Figure 4.23. Warping of a sample from Lab X aged for 25 days at 65° C

Results for C Samples (Lab Y, New York)
The neutral patches of both unmounted and face-mounted samples shifted towards red and

violet (Figures 4.12 and 4.16), the former being more prominent. Both Figures show that the colour shift was more rapid in the unmounted than in the face-mounted prints, with the exception of the red shift in the samples aged at 55° C (Figure 4.12). The unmounted prints yellowed more rapidly than the face-mounted ones, and reached a difference in Dmin(B) value of 0.07 density units compared to that of 0.02 (Figure 4.17). As shown in Figures 4.18 to 4.20, no significant changes in colour density during ageing were measured in either the cyan, magenta, or yellow patches of the unmounted and face-mounted 65° C samples. Visually, the face-mounted samples appeared much closer to the control after ageing even at high temperatures.

The only problem encountered with the face-mounted samples was the appearance of small areas of a type of delamination in the samples aged for longer than 20 days at 55° C and in all aged at higher temperatures (Figure 4.24). This phenomenon has been termed “snowflakes”, due to the form that the blisters take on. Ilford Photo attributes these small bubbles to moisture retained on the surface of the acrylic sheet that is driven into the adhesive layer of the mounting film (1988, 14-15). As a remedy, an eight hour period of pre-heating the acrylic to remove adsorbed moisture is recommended. Long-term ageing at ambient temperatures would be necessary to determine whether this form of degradation also occurs at common storage conditions.

Figure 4.24. Formation of “snowflakes”: delamination in an adhesive film face-mounted sample from Lab Y, New York, aged at 85° C for 20 days

Comparison of the Ageing Results of all Batches
The compilation of all ageing results shows that in the A1 and B batches, the face-mounted

prints reached their endpoints by colour shift in the neutral patches more rapidly than the unmounted samples (Figure 4.25). In the A2 and C batches, on the other hand, endpoints were reached more rapidly in the unmounted samples. In the latter cases, the differences in time are more pronounced. The unmounted prints of all batches yellowed to a greater degree than the face-mounted prints (Figure 4.17).

The prints of samples A1 and A2 were both processed and mounted with the same methods and materials, but they differ in type of photographic paper, the former being sensitised for analogue exposure from a negative, the latter for exposure from a digital file. In terms of yellow stain formation, both showed similar characteristics, but the unmounted A2 prints reached their endpoints faster and the face-mounted prints reached theirs at least two times more slowly than the A1 samples. These discrepancies can only be attributed to the different characteristics of the different emulsions, as all other factors were equal. In contrast, the prints of all three groups A2,

B, and C, were made on the same digital paper, but at different laboratories. Variations in ageing speed (Figure 4.25) are great enough to show that chemical processing of the paper (Wilhelm 1993, 164) and mounting materials can have an effect on the long-term stability of these photographs.

The fastest colour changes and rapid warping were apparent in the face-mounted BM samples, those printed and mounted at Lab X. Their unmounted counterparts also aged quickly. Yellowing was significant in both unmounted and face-mounted samples. The slowest ageing was found in both the A2M (Diasec® face-mounted digital paper from Grieger Düsseldorf), and the CM face-mounted prints, those with the double-sided adhesive film from Lab Y, New York.

The only general statement on the influence of face-mounting on the stability of prints that can be made relates to the retarded yellow staining of the base. In terms of colour shifts in the

print, the long term stability seems to depend on the choice of materials, specifically the combination of photographic paper and silicone primer and sealant. The reasons remain unclear, however.

The results for the A2 and C samples would suggest that the exclusion of air from the vicinity of the emulsion, as is the case with face-mounted prints, slows down the fading of dyes. This interpretation would be corroborated by the fact that oxidation, accelerated by high temperature and RH, is one of the causes of chromogenic print degradation. In a statement that contradicts these results, Eastman Kodak technicians have written that complete isolation of a colour emulsion from air may increase cyan dye fading. Furthermore, the acetic acid acid released by silicone sealants is said to adversely affect cyan and yellow dye stability (Kodak Professional Division 1998, 3). The only concession to these statements that could be observed in Experimental II, however, was the accelerated yellow dye fading in the BM face-mounted samples, which, incidentally, also had the highest amounts of off-gassed acetic acid, as is described in section 5.2, “Results and Discussion”, of Experimental III below. These contradicting statements and observations would warrant further research.

Evaluation of the Ageing and Measuring Methods
The application of the standard test method ANSI/NAPM IT9.9-1996 as used for the

materials tested here is evaluated in the following. ANSI states that the sealed bag method is ideal for simulating storage conditions for enclosed photographic materials such as motion picture film in a metal or plastic film can. In the closed environment, acetic acid off-gassing from cellulose acetate film base will accumulate and may contribute to the fading of the dyes (1996, 5). In the case of ageing face-mounted prints it may not be desirable to seal the acidic vapours within the enclosure, since this is only one of various possible storage conditions found in collection vaults. If the accelerated ageing is to simulate conventionally found conditions, the prints might better be aged free-hanging in RH-controlled ovens. To avoid the accumulation of

acetic acid in the oven, the prints would have to be aired until there is no strong smell of acetic acid being emitted from them prior to the experiment.

A further problem encountered with the use of the sealed bag method was the fact that measurements were taken only twice from a sample: once before ageing and once when it is removed from the oven. Once evaluated, the print was not re-bagged and returned to the oven. Therefore, the curve that resulted from the measurements, which is meant to describe the gradual ageing of a sample, is actually showing the individual points of many samples. Although these were processed and finished equally and simultaneously, as far as relayed by the photographic laboratories and mounting studios, variations between the many prints may occur that will cause the curve to become irregular. Multiple measurements were taken from each patch in an effort to average the results, but if a smaller number of prints had been aged simultaneously throughout the full length of time necessary to reach an endpoint, and the results of these measurements averaged, a more reliable curve would probably result. This method would only be possible if the samples could be returned to the ovens after measurement, as is the case with ageing in temperature and RH-controlled ovens.

The use of densitometry was not entirely satisfactory for reading the values of the face- mounted prints, since the values were not always similar, differing by up to 0.10 density units depending on the exact position of the measuring head, which was positioned above the small areas that had been cut out of the cover of the measuring box. For this reason, at least three measurements were made for each designated patch on the sample. These discrepancies may have lead to irregularities encountered in the data as seen in the graphs. For future experiments, a better solution should be found for measuring the patches through the acrylic sheeting.

5. Experiment III: Determination of Acetic Acid Offgassing

5.1. Experimental
5.1.1. Materials and Equipment
The prints used in this test were additional samples of the four face-mounted batches used

in the accelerated ageing tests. They were sealed as described in section 4.1.2, “Procedure”. AD- Strips, used for the determination of acidic offgassing, are small paper strips coated with bromocresol green that changes colour in the presence of acidic vapours. The indicator is sensitive to the quantity of acidic volatile products present and changes from dark blue through green to orange when exposed to increasing levels of acidic vapours. The air samples for gas chromatography were taken with Anasorb® CSC solid sorbent glass tubes containing two sections of activated coconut shell charcoal separated by a urethane foam plug. These tubes were attached one by one by flexible rubber tubing to a sampling pump.

Further experimental materials and equipment included glacial acetic acid, formic acid, a soap bubble air flow indicator, 1 and 5 mL syringes, a 10 mL volumetric flask, 2 mL glass vials with Teflon® lined caps, a crimper, pliers, a metal file, a mat cutter, a glass pipette, 20 AD-Strips, and 18 further small glass vials. Gas chromatographic analysis was performed with a Hewlett Packard 5890 Series II at the Analytical Services Unit, Environmental Studies, Queen’s University. The column used was designated as “1m x 4 mm ID glass; Carbopack B 60/80 mesh/3% Carbowax 20M/0.5% H3PO4”. The column temperature was set at 130° C, and the run time was 10 minutes.

5.1.2. Procedure

AD-Strips and small glass vials were included in the face-mounted control sample bags of Experiment II. The strips and vials were also placed in bags with additional face-mounted samples coded with “ADGC” to be aged at each temperature, and in those to be kept at room

temperature. The original 75° samples were not evaluated due to the oven difficulties described in Experiment II. In consequence, the samples were aged for 38 days at -15°, 20°, 55°, 65°, and 85° C. The samples designated for 85° ageing were placed in a 75° oven after 25 days due to technical difficulties with the ovens. They remained at 75° for further 13 days, which made up a full 38 days.

The AD-Strips were designated to determine, on a broadly quantitative level, the presence of acetic acid off-gassing from the silicone during storage and ageing. One AD-Strip was aged alone in a sealed bag at each temperature as a method of controlling the colour stability of the bromocresol green at higher temperatures and thus ensuring the accuracy of the readings.

The interior bag air of the samples marked “ADGC” was analysed with GC-FID to detect the quantity of off-gassing acetic acid on a more precise level than the AD-Strips could provide. The protocol followed for this analysis was Acetic Acid, Method 1603, Issue 2, as provided by the NIOSH Manual of Analytical Methods (National Institute for Occupational Safety and Health 1994). The glass vials in the bags functioned as spacers to maintain a sufficient amount of air in the bag for analysis.

After ageing was completed, the air from the bag interior was removed by suction with an air pump. The flow rate of the pump had previously been calibrated to 10.5 mL/second with a soap bubble air flow indicator. The sealed ends of a solid sorbent tube were broken off with pliers, and the sorbent tube was inserted into the end of the rubber tubing connected with the pump. The pump was turned on, and a small slit immediately made in the sealed bag near the end of the incorporated vial. The tube end was inserted into the slit and stretched the laminate bag material sufficiently to form an airtight seal around it. After an average of seven seconds, the air had been removed by suction from the bag interior, as could be seen by a vacuum effect on the retracting bag sides and heard by the straining pump motor. In this duration, ca. 70 mL of air was sampled. It was assumed that the interior surface of the bag material and the Plexiglas® of the samples had adsorbed some of the offgassed acetic acid, so the corner opposite to the sampling

tube was quickly cut off, and a glass pipette inserted into the slit to open it sufficiently. The bag was then purged with air from the laboratory for five minutes (Figure 5.1).50 In this manner, any remaining acetic acid in the bag could be removed by suction and adsorbed by the activated charcoal in the sorbent tube.

Figure 5.1. Sampling the air from the bag interior with a solid sorbent tube and an air pump

Following sampling, the two ends of the tube were capped and labelled. The bag was removed from the area quickly so as not to contaminate the laboratory air for the next bag sampling. Any contamination with acetic acid would easily have been detected by the operator through its smell, since the odour threshold limit of acetic acid is extremely low, at 0.48 ppm (Hill Brothers Chemical Company 1999).51 In another location, the bag was fully cut open and the AD-Strip removed. The colour of the AD-Strip was read immediately by comparing it to the

50 Method 1603 is designed for sampling air in a room or hall, and specifies a sample size of 20 to 300 litres (1994, 2). In this experiment, it was found through a trial run that the amount sampled, just over 3 litres, was sufficient to remove much of the acetic acid from the bags. The samples did, however, still smell of acetic acid after sampling. This may be due to their continued offgassing.

In comparison, the lowest measured quantity of acetic acid in a bag was 8.8 ppm, so measurable contamination of the air could be ruled out.

colour scale and then sealed in a small glass vial to retain its colour for later re-evaluation. All bags were sampled in this manner, and three blank control samples for GC-FID analysis were made by sampling empty bags, one at the beginning, one half way through, and one at the end of the sampling session.

In preparation for GC-FID analysis, the solid sorbent tubes were broken open with the metal file, and the charcoal from each transferred to 2 mL vials. Following the addition of 1 mL formic acid, each vial was sealed with a Teflon® lined cap using a crimper (Figure 5.2). The labelled vials were agitated for one hour so that the acetic acid could be desorbed from the charcoal into the formic acid. A standard solution designated “Standard 900” was made up, consisting of 0.9 mL glacial acetic acid in 10 mL of formic acid. These steps were carried out in a fume hood.

Figure 5.2. Preparation of the sample vials for GC-FID analysis in the fume hood

Prior to analysis of the samples, the GC column was flushed four times with formic acid. The basic trace content of acetic acid in the formic acid used could be determined. The vials with the desorbed acetic acid from the samples were run, each interspersed with a formic acid blank to

clean the column from traces of the previous. The samples were run in the order from lowest assumed acetic acid content to highest, beginning with the three blanks and ending with the 85° C samples and the Standard 900.

5.2. Results and Discussion
The AD-Strip colours and their respective numbers according to the AD-Strip scale are

listed in Table 5.1. The AD-Strip literature supplied by the manufacturers indicates an increasing acid concentration on a scale from 0 to 3, with a reading of 0 as no change due to the absence of acidic vapours, and a 3 indicating the highest amount of acid in the air (Image Permanence Institute 1998). As some of the AD-Strips had a colour shift that was stronger than the maximum value of the scale, an attempt was made to extend the scale to a value of 4. The AD-Strip sealed in a glass vial at the far right of Figure 5.3. is too light and too orange to be regarded as a number 3, for example. Some AD-Strips were bluer than the blue given by the scale for a value of 0, or “no change”. These were recorded as “>0” in the results. The colours could not always be definitely categorised as one specific number, so tendencies were estimated and an interpretative reading was sometimes necessary.

Figure 5.3. Colour scale and some AD-Strips unsealed and sealed in vials after ageing

Sample code	Temperature (°C)	Time (days)	AD-Strip reading colour number
AD-Strip-0 AD-Strip-55 AD-Strip-65 AD-Strip-85	-15 55 65 85, 75	t=38 t=38 t=38 t85=25, t75=13	dark blue blue blue-green dark green	>0 0 0.5 1
A2M-0-ADGC A2M -20-ADGC A2M -55-ADGC A2M -65-ADGC A2M -85-ADGC	-15 20 55 65 85, 75	t=38 t=38 t=38 t=38 t85=25, t75=13	blue-green – green (non-uniform) dark orange-green light green dark orange-green dark orange-green	0.5 2.5 2 2.5 2.5
BM-0-ADGC BM-20-ADGC BM-55-ADGC BM-65-ADGC BM-85-ADGC	-15 20 55 65 85, 75	t=38 t=38 t=38 t=38 t85=25, t75=13	dark orange-green yellow-green orange-green (looks larger than 3) orange-green (looks larger than 3) yellow (looks like a 4)	2.5 3 3.5 3.5 4
CM-0-ADGC CM-20-ADGC CM-55-ADGC CM-65-ADGC CM-85-ADGC	-15 20 55 65 85, 75	t=38 t=38 t=38 t=38 t85=25, t75=13	dark blue dark blue blue-green dark green light green	>0 >0 0.5 1 2

Table 5.1. AD-Strip readings after ageing for 38 days

The AD-Strips that were aged alone as controls showed slight colour changes at 65° and 85° C. The bags from the higher temperature ovens had a sharp, musty odour exuding from the interior when cut open. This odour may be from the inside material, polyethylene, itself, or from an additive to the plastic. The change of the colour of the AD-Strip is either due to an instability of bromocresol green to high heat for prolonged periods of time, is a result of the musty odour, which may contain acidic vapours, or both. For the sake of simplicity, these changes were not taken into account for the evaluation of the off-gassing at high temperatures, since the AD-Strips only give a general indication of the amount of volatile acid present.

Data from GC-FID analysis, comprising of graphs and value tables, for all samples are attached in Appendix V. The relevant readings are summarised in Table 5.2. The data was evaluated by recording the value of the area beneath the acetic acid peak at ca. 4.3 minutes

retention time on the x-axis, divided by 1000. The three blank samples of empty bags resulted in values that could only have resulted from traces of acetic acid in the formic acid eluent. Their values were averaged and subtracted from the area values of the other samples. In order to determine the concentration of acetic acid in the samples, the known concentration of the Standard 900 vial, 1084 μg/mL, was divided by the area under the acetic acid peak measured for it: 2830. This calculation resulted in a conversion factor for the other area values, namely 0.383 μg/mL, with which the amounts of acetic acid in each sample vial could be determined. As each vial contained 1 mL of formic acid with the desorbed acetic acid, the resulting value, listed in the right column of Table 5.2, is valid for the amount of acetic acid in the total air content of each sealed bag.

Sample code	Temperature (°C)	Time (days)	GC-FID Data area under acetic acid area under peak minus μg acetic acid / peak blank average of 18 mL formic acid
Blank 1 Blank 2 Blank 3 Standard 900	– – – –	– – – –	18 18 17 2830	– – – –	– – – –
A2M-0-ADGC A2M -20-ADGC A2M -55-ADGC A2M -65-ADGC A2M -85-ADGC	-15 20 55 65 85, 75	t=38 t=38 t=38 t=38 t85=25, t75=13	41 92 64 70 79	23 74 46 52 61	8.8 28.3 17.6 20 23.4
BM-0-ADGC BM-20-ADGC BM-55-ADGC BM-65-ADGC BM-85-ADGC	-15 20 55 65 85, 75	t=38 t=38 t=38 t=38 t85=25, t75=13	340 1036 1970 2878 3785	322 1018 1952 2860 3767	123.3 389.9 747.6 1095.4 1442.8
CM-0-ADGC CM-20-ADGC CM-55-ADGC CM-65-ADGC CM-85-ADGC	-15 20 55 65 85, 75	t=38 t=38 t=38 t=38 t85=25, t75=13	17 18 19 24 35	-1 0 1 6 17	– 0 0.4 2.3 6.5

Table 5.2. Results from GC-FID measurements and calculations

The values in Tables 5.1 and 5.2 and the corresponding graphs in Figures 5.4 and 5.5 show that in general for each batch an increase in temperature causes an increase of acetic acid off- gassing from the samples. A direct correlation between both graphs can be seen, taking the interpretive readings of the AD-Strip colours into account. This may be evaluated as a confirmation of the accuracy of both the gas chromatography and the AD-Strip readings.

The CM samples, those mounted at Lab Y using a double-sided adhesive film, have acetic acid values that are very low. It is not clear how the acetic acid is produced, as these samples contain no silicone rubber. The A2M samples, Diasec® prints made at Grieger Düsseldorf, have moderate acetic acid levels at all temperatures. Interestingly, they show a higher reading for acetic acid at 20 degrees than at higher temperatures. The BM samples produced comparatively very large amounts of acetic acid. This may be a function either of the thickness of the silicone rubber, three times that of the Diasec® face-mounted samples, as was shown in section 2.4, “Laminate Structure”, and therefore of the exposed area along the sample edges, or of a possible tendency of this material to give off a greater amount of acetic acid when curing. It may also be a combination of both factors.

The results shown here also correlate with initial observations made upon receipt of the samples from the mounting studios. When unpacking the samples from their plastic bag enclosures, in which they had been packed for ca. one week following their production, the samples from Lab X emitted a very strong and the samples from Grieger only a slight odour of acetic acid. The samples from Lab Y had no perceivable odour. It appears that although temperature does have an influence on the amount of acid emitted, acetic acid is given off even at a temperature as low as -15° C. It should be taken into account, however, that the values may also derive from the brief time that the sealed samples were not in a -15° C environment, namely for the few hours before and after their storage in the freezer compartment.

As acetic acid is solely produced during the curing process, only a finite amount is present in the sealant, and this is being constantly offgassed. At some point the residual acid in the

sealant will have been fully given off, and the print can be considered “acid-free”. Tétreault has shown this to be the case for one gram of silicone rubber in 50 mL air in his experiments with modified indicator strips as acid vapour detectors (1999). In the case of the face-mounted prints, the sealant is sandwiched tightly between two relatively impermeable materials, each of which is capable of absorbing acetic acid, as has been demonstrated for Plexiglas® by Fenn (1995). Due to these circumstances, it may be assumed that the rate of off-gassing of acetic acid from the face- mounted prints is slower than that measured by Tétreault, resulting in objects that will retain their potential for giving off corrosive vapours for a long time.

An observation that supports this suspicion is the fact that extra samples that have been aired and bagged on and off for four months since their production still have a distinct odour of acetic acid that is especially strong along the edges. A simple experiment was conducted to confirm this observation: An AD-Strip was taped with a transparent tape (Scotch MagicTM Tape) around the edge of a BM sample that had previously been aged for 59 days at 55° C. The tape was used to exclude any acids from the atmosphere, and a control was made by taping up a second AD-Strip on both sides. Boths samples were kept in the dark. After one day, a shift to green of the AD-Strip on the BM sample could be seen originating along the area in contact with the exposed layer of silicone rubber and diffusing outwards (Figure 5.6).

As the acid escapes mainly through the edges, it would be a logical conclusion that larger prints reach their off-gassing endpoint slower than smaller prints, since the acid must migrate further from the print centre to reach an edge. This may be especially the case for those large face-mounted prints that have been backed with double-sided adhesive film and a plastic sheet, as described in section 2.2, “Mounting Procedures”, in which the possibility for the acetic acid to permeate through the print substrate is further hindered by the additional barriers. Questions as to consequences of these findings are considered in section 7, “Further Research”.

Figure 5.6. Colour shift in an AD-Strip taped around the edge of a Lab X silicone rubber mounted print after 1 day

6. Conclusions

It should be stressed that the conclusions reached here are valid only for the chromogenic photographic papers in combination with the primers, sealants, adhesive film, and acrylic sheeting that were examined and tested in the experimentals described above. Different brands of chromogenic papers and silver dye bleach papers such as Ilfochrome may react very differently during ageing, and the ageing of other silicone sealants may vary from the results given here.

Materials
1. Photographs face-mounted with silicone sealant and double-sided adhesive film are

visually indistinguishable from each other, but recently mounted prints using silicone may be detected by their odour of acetic acid.

2. The sealants commonly used by Diasec® licensees and other mounting studios are acetoxy curing, one-component, RTV silicone rubbers. The sealant from the Diasec® patent is an amine curing rubber. Once cured, the sealants have the same general mix of ingredients, but may vary by ratio and by small quantities of additives. The no-name sealant used the Diasec® process is very similar in composition to the GE Silicones sealants that were examined.

3. The primers may consist of organosilanes or siloxanes in a solvent mixture vehicle. The primer Sallmetall “Duramount” was shown to be pure solvent mixture. The no-name primer used in Lab X has the same active ingredients the Sallmetall “Catalyst”. The precise function mechanisms of the primers cannot easily be determined. Their irregular application to the acrylic sheet or photograph surface may lead to partial delamination in a warm environment.

4. The acetoxy silicone sealants give off significant amounts of acetic acid during curing, which escapes primarily through the exposed area of sealant at the edges of the face-mounted prints. The rate of off-gassing is governed by temperature and probably also by the type of sealant used. The acetic acid may be retained by the PMMA and photographic print materials and

given off over a long period of time, which may make face-mounted prints a corrosion hazard to other objects in their vicinity.

5. As acetic acid off-gasses only slowly, prints mounted with acetoxy-curing silicone rubber should be aired in a well ventilated area until they have lost their odour of acetic acid. Only then may they be wrapped or housed in impermeable enclosures. Due to the low odour threshold limit of acetic acid, smelling the recto, verso, and edges of a face-mounted print may give an indication of its degree of off-gassing and age.

Long-term Stability
Despite the oven problems and irregularities in measurements, it was felt that meaningful

conclusions could be made. It is highly recommended, however, that further testing be carried out to corroborate or contradict the results obtained here.

6. In two of the four sample batches, face-mounting the prints significantly slowed colour shifts in the image in dark storage. In the other two, it slightly accelerated these changes. Causes for these differences were not determined.

7. Compared to the unmounted prints, yellow staining of the base was notably slowed down in all face-mounted prints.

8. When processed and face-mounted identically, the digital chromogenic paper (A2M) showed better dark stability than the analogue paper (A1M). As shown by comparison with the other face-mounted sample batches (BM and CM), however, the dark ageing characteristics of the same digital paper can be influenced to a large degree by print processing, materials used in face-mounting, or both.

9. The face-mounted samples from Lab X (BM) had the highest quantity of acetic acid off- gassing, the highest degree of yellow stain formation, the fastest rate of colour shifts, and the strongest fading of the yellow dye. Correlations between these results were not determined.

10. Significant differences between the dark ageing characteristics of prints face-mounted with silicone rubber and those with double-sided adhesive film were not found. The latter may suffer from “snowflake” formation in the adhesive layer at temperatures at least above 55° C.

11. The use of extruded PMMA for face-mounting may lead to warping in storage at temperatures at least above 55° C.

Test Methods
12. The sealed bag method did not seem to be a practical and reliable method for dark

ageing face-mounted prints. The free-hanging method might better simulate actual storage and exhibition conditions. Accelerated dye fading due to trapped acetic acid in the bags could not be confirmed.

The long-term dark stability of face-mounted prints has been shown to be dependent on a number of factors that together determine how photographs react when placed in an environment such as that given by the face-mounting process. These include the type of photographic material, print processing, materials used for mounting, and exhibition or storage environment. What seemed to be an ingeniously simple process, namely adhering a photograph to a sheet of Plexiglas®, has proven to be wrought with physical and chemical variables that contribute to the end result. Their interdependency could not be widely explored in this project, but an understanding of their interactions may prove to be a key for the conservation of face-mounted photographs. Next to the experimentals presented here, which only examined some aspects of the longevity of face-mounted prints, a major factor for determining guidelines for dealing with these objects is the influence that face-mounting may have on the lightfastness. Much research must still be carried out if we are to reach a better understanding of these intricate objects.

7. Future Research

The research presented in this project may be regarded as preliminary, as the widespread use of the face-mounting process for photographs warrants more in depth research. The issues explored here, however, may have given an indication of the strengths and weaknesses of face- mounted prints, as potentially problematic areas have been uncovered. Further research needs are presented in the following, focussing on stability and conservation issues.

Physical Structure and its Impact on Aesthetics

An investigation into the adhesion mechanisms, both physical and chemical in nature, would result in a better insight into the potential problems associated with laminate objects such as differing expansion and contraction coefficients and reactions to fluctuating environmental conditions. In terms of the aesthetics of face-mounted prints, the optical impact on image colours and contrast in comparison to conventionally framed photographs could be examined theoretically and in practical experiments. A light ray such as a laser beam might be used to determine the refraction of light at the interfaces between the materials. It is not clear if the presence of a layer of air between the print surface and the cover glass of a conventionally framed photograph increases the amount of bouncing reflection of light in the layer, which may be the cause for increased exposure to light. The elimination of this layer may be beneficial in that it eliminates this source of enhanced light exposure. A light fading test comparing face-mounted and conventionally framed prints would allow more insight into these circumstances.

Materials Stability and Deterioration

Long-term dark stability tests should be re-run to confirm or contest the data obtained in Experiment II. As the results were found to be slightly inconsistent, these tests might be better conducted in ovens that allow for RH control as well as temperature. This would make the need

for sealed bags superfluous. At best, unsealed and free-hanging samples could be aged parallel to each other under the same circumstances to test for any differences in ageing characteristics. In addition, a more accurate system of measuring changes in the face-mounted prints than that used in this project would lead to more reliable results. In terms of light stability, the necessity of UV filters in the PMMA might be re-evaluated. Wilhelm has shown the presence of a UV filters in an acrylic sheet to be superfluous, as chromogenic papers already have a UV filtering surface coating (1993, 145). It may be more likely the tendency of Plexiglas® to absorb a portion of visible light that is of importance here. Light stability comparison tests of unmounted and face- mounted prints have been carried out by Sylvie Pénichon, but the results and conclusions had not been published at the date of preparation of this manuscript.

Next to light, the degree to which water is present is a critical factor for the long term stability of photographs. It is unknown how face-mounted photographs react to changes in RH and direct contact with water, but it would seem that the edges are the weak point of the system. It might be beneficial to determine the rate at which water vapour is taken up and given off by the materials involved. It is assumed that face-mounted prints are not as susceptible to mould damage as unmounted prints, as the most nutritious material, the gelatine, is sealed in an airtight environment.

In terms of the longevity of the adhesive, it has been demonstrated by the manufacturers that silicone rubbers are very stable materials. Independent tests, as they are common in conservation research, would, however, serve to better examine the ageing characteristics of these adhesives. Critical factors would include possible discolouration, changes of pH, refractive index, flexibility, hardness, tensile strength, solubility, adhesive strength, and chemical stability. A Photographic Activity Test (PAT)52 of a silicone rubber film might give an indication of its suitability for being in direct contact with a photographic emulsion.

The PAT was developed by the Image Permanence Institute in Rochester, NY. It is designed to test materials for their potential of altering silver particles such as those found in photographs.

Experimental III showed that acetic acid is present in face-mounted prints and is given off to a degree that is dependent on temperature. It is not known, however, how long a silicone sealant film as it is found sandwiched between two relatively impermeable layers will take to release all of its acetic acid. It is assumed that the acid is deposited in the gelatine and PMMA, but the extent has not yet been shown. The influence, if any, of acetic acid on the gelatine and the dyes and colour couplers of the emulsion might be examined. The amount of acid that could be regarded as critical for these materials is unknown. In terms of offgassing, many questions are still unanswered: What happens to the acetic acid and how much is detrimental? Does a frame retard the offgassing, and if so, is this an advantage or a disadvantage? NIOSH exposure limits (TWA)53 for acetic acid are given at 25 mg/m3, or 10 ppm (National Institute for Occupational Safety and Health, unknown date, 2). Is there danger for the people handling face-mounted prints? Do Diasec® prints endanger other objects, particularly metals, in their vicinity? An Oddy test54 with silicone rubber face-mounted prints might give more insight into this complex.

A further topic for future study would be possible effects of the solvents in the primer on the PMMA and photographic emulsion, in both the short and the long term. Some of the primer ingredients described in section 3.2, “Results and Discussion”, are also found in the list of solvents for PMMA in section 2.3.4, “Poly(methyl methacrylate)”.

It would be of interest to examine the use of so-called neutral or non-corrosive curing sealants alternative to those that give off acetic acid. An alkoxy curing material, such as GE Silicones TSE397 sealant, cures with atmospheric moisture by releasing methanol and ammonia (GE Silicones 1997b). The presence of an alcohol may, however, lead to crazing of the PMMA, especially if it is an extruded sheet that contains stresses. Polycarbonate has been shown to

The Oddy test (Oddy 1975) is used in conservation to test materials, such as those used in display cases, for their suitability for use in the vicinity of objects. Polished metal coupons are incubated with the materials in question and their subsequent degree of corrosion is evaluated.

Time Weighted Average

deteriorate in the presence of these reaction products(GE Silicones, unknown date a). The continued use of acetoxy sealants for face-mounting may have other reasons. Conventional alkoxy RTV sealants cure more slowly (Lynch 1978, 158), and thus would lengthen the time necessary for the mounted photograph to lie still following mounting, which is not a very attractive notion to many photographic printing and mounting studios, as it would slow down production. Modern sealants can be found with varying curing rates, however (GE Silicones 2000d). The use of the amine-curing Gurisil 575 sealant, mentioned in the patents and examined in Experimental I, was discontinued, but the reasons are unclear.

Handling, Preservation, and Storage

Guidelines specifically for handling face-mounted prints may have already been established in institutional collections. When on exhibit, special considerations may need to be implemented. Visitors at an exhibition might not be aware that the Plexiglas surface of a framed Diasec® print is an integral part of the object, and may inadvertently touch the surface without realising that they are actually touching the object and not just the glazing that is there to protect the work of art. This may lead to abrasion, the accumulation of fingerprints, dirt, and dust, and other staining of the PMMA surface. In terms of long-term preservation of face-mounted images, cold storage as commonly used for colour prints should be evaluated for potential negative effects.

Conservation and Treatment Options

The treatment of face-mounted prints has not yet been widely explored. Disaster simulation may be a way to determine the reaction of face-mounted prints to effects of fire, such as heat and smoke, water intake at edges, and impact from various angles. Dirty, scratched and damaged Plexiglas® may be one of the greatest problems faced by conservators. Simple cleaning and dusting can lead to scratches in the Plexiglas® surface, and various methods have been devised to

100

avoid these. An alternative to a duster or chamois cloth is the “PCR-Roller” sold by Grieger Düsseldorf. It consists of a slightly tacky synthetic rubber roller that is rolled across the PMMA surface to pick up dust, then subsequently over a sheet of sticky film to remove the dust and lint adhered to it. A variety of polishing pastes for the removal of scratches are available.

Further experimentation may be necessary for devising methods of infilling and inpainting losses in the photographic emulsion that have occurred by deep scratches to the unprotected print verso. In one case, scratches that penetrated through to the Plexiglas® in the area of the white border around the image area of a face-mounted print had been filled with an unknown opaque white cement material.55 It conformed to the losses, but differed not only in tone, but also in depth: as the material was in direct contact with the Plexiglas®, it appeared to “hover” over the surrounding print areas, which were set back from the plastic sheeting by the layer of gelatine emulsion. Even if only very slight, this difference in distance gave the infills a three-dimensional appearance.

Methods for reversing delamination will also be of importance. Conservator Kimberly Schenk writes of a case that she experienced with a face-mounted Cibachrome print: “The Baltimore Museum has a Laurie Simmons Cibachrome print face-mounted with Plexiglas® probably using the Diasec® method or a similar system. The print is from 1987. It was exhibited in 1997 here at the Museum. The artist wanted the works flat against the wall without frames. Her gallery at the time used pieces of Velcro to attach the artworks to the wall. When the strips were removed from our work a few years after the show, some delamination occurred resulting in bubbles […]. I was able to remove and partially reduce the remaining bubbles with weights.”56 It is interesting to note that the bubbles did not occur in a circa two inch band along the edge of the print, and, although they could be moved by applying local pressure to the print verso, they would

101

Peter Mustardo, The Better Image, Pittstown, NJ, personal communication, March 2001.
Kimberly Schenk, Baltimore Museum of Art, Baltimore, MD, personal communication, April 18, 2001.

not run into this area of high adhesion. This may be due to the practice of some mounting studios to merely apply the primer to the edges of the Plexiglas® and print in order to work economically.57 This will ensure critical adhesion along the edges, where any delamination is thought most probable. As the trapped air in the Baltimore print could not be released at the edges of the object, the print was punctured with a pin from the verso in the centre of each bubble prior to the application of weights. Finally, the pinhole was sealed with the acrylic resin Acryloid B-72.

In terms of reversibility of silicone rubber face-mounting, much is unknown. GE Silicones gives advice on removing cured sealant, but its applicability to a face-mounted print must be evaluated. After mechanically removing the bulk of the silicone with a knife or a razor, a solvent is applied to remove oily residues or remaining sealant. The solvent can be used locally or the object may be immersed and left to soak overnight. In the order of aggressiveness, the following solvents are recommended: mineral spirits, toluene, xylene, acetone, and methylene chloride (GE Silicones, unknown date b). For removal of a bronze stuck to Plexiglas® with silicone rubber, a treatment has been described using a solution of equal parts of de-ionized water, ethanol, and acetone for local immersion. This solvent mixture allowed the mechanical separation of the bronze and the Plexiglas® from the sealant.58 A commercial product, Amtex CCR Silicone Remover, has also been mentioned.59

The proposed topics for future research compiled here should be regarded as a list of possible approaches to gaining a better understanding of the permanence and conservation of face-mounted photographs. It would benefit from further input and experimentation, with the goal that, in the near future, face-mounted prints become objects we feel comfortable about handling, exhibiting, storing, and treating.

102

Manager of Lab X, Düsseldorf, personal communication, August 2000.
David Cottier-Angeli, Conservation DistList, <http://palimpsest.stanford.edu>, letter of December 19, 2000. Julie L. Sloan, Conservation DistList, <http://palimpsest.stanford.edu>, letter of December 20, 2000.

7. References

Agnew, N. 1984. The use of silicones in the preservation of a field site – the Lark Quarry dinosaur trackways. In Adhesives and consolidants: preprints of the contributions to the Paris congress, 2-8 September 1984, ed. N.S. Brommelle, E.M. Pye, P. Smith, and G. Thomson. London: The International Institute for Conservation of Historic and Artistic Works. 87-91.

Allen, K.W. 1984. Fundamental aspects of adhesion. In Adhesives in engineering and design, ed. W.A. Lees. Berlin: Springer Verlag. 129-139.

American Psychological Association. 1994. Publication manual of the American Psychological Association. 4th ed. Washington, DC: American Psychological Association.

Analytical Research Services. 1989. ARS No. 1666: General Electric silicone seal. In Commercial product analytical report. Ottawa: Canadian Conservation Institute.

Anderson, D.R. 1974. Infrared, raman and ultraviolet spectroscopy. In Analysis of silicones, ed. A.L. Smith. New York: John Wiley & Sons. 247-286.

Angelotti, N.C. and Hanson, C.L. 1974. Analysis of polymers, mixtures, and compositions. In Analysis of silicones, ed. A.L. Smith. New York: John Wiley & Sons. 41-72.

ANSI/NAPM. 1996. IT9.9-1996, Stability of color photographic images – methods for measuring. New York: American National Standards Institute.

103

Attila, K. (Ebond Australasia PTY Ltd.) 1989. Method and aparatus for mounting photographic prints. Australia. Patent WO8911681. November 30, 1989.

Brewis, D.M. 1983. Aluminium adherends. In Durability of structural adhesives, ed. A.J. Kinloch. London: Applied Science Publishers. 215-254.

Brill, T.B. 1980. Why objects appear as they do. Journal of chemical education 57(4): 259-263.

Brook, M.A. 2000. Silicon in organic, organometallic, and polymer chemistry. New York: Wiley Interscience, John Wiley & Sons, Inc.

Byrne, G.S. 1984. Adhesive formulations manipulated by the addition of fumed colloidal silica. In Adhesives and consolidants: preprints of the contributions to the Paris congress, 2-8 September 1984, ed. N.S. Brommelle, E.M. Pye, P. Smith, and G. Thomson. London: The International Institute for Conservation of Historic and Artistic Works. 78-80.

Canadian General Standards Board. 1991. National standards of Canada: Glossary of terms related to sealants, CAN/CGSB-19.28-91. Ottawa: Canadian General Standards Board.

Cesar, C. 1997. Archival photoprint laminate. USA. Patent 5,595,794. January 21, 1997.

Charola, A.E., Wheeler, G.E., and Freund, G.G. 1984. The influence of relative humidity in the polymerization of methyl trimethoxy silane. In Adhesives and consolidants: preprints of the contributions to the Paris congress, 2-8 September 1984, ed. N.S. Brommelle, E.M. Pye, P.

104

Smith, and G. Thomson. London: The International Institute for Conservation of Historic and Artistic Works. 177-181.

Cook, J.P. 1970. Construction sealants and adhesives. New York: Wiley Interscience, John Wiley & Sons, Inc.

Cox, R.J., ed. 1976. Photographic gelatin II. London: Academic Press.

Croome, R.J., and Clegg, F.G. 1965. Photographic gelatin. London: The Focal Press.

Daley, T.W. and Murdock, L.D. 1984. Underwater molding of a cross-section of the San Juan hull: Red Bay, Labrador, Canada. In ICOM Committee for Conservation, 7th triennial meeting, Copenhagen 10-14 September 1984: preprints. Paris: International Council of Museums. 84.7.1- 84.7.5.

Day, R.A. 1998. How to write and publish a scientific paper, 5th ed. Phoenix, AZ: The Oryx Press.

DuroTech Corp. 1997. Material safety data sheet: primer WA-3. January 21, 1997.

DuroTech Corp. 1999. Material safety data sheet: catalyst CP-48-R. February 22, 1999.

DuroTech Corp. unknown date a. Material safety data sheet: silicone rubber adhesive CCC 478.

DuroTech Corp. unknown date b. Instructions for mounting with silicone adhesives.

105

Edge, M. 1996. Life prediction: fact or fantasy? In Research techniques in photographic conservation: conference proceedings, 14-19 May 1995, ed. M. S. Koch, T. Padfield, J. S. Johnsen, and U. Bøgvad Kejser. Copenhagen: The Royal Danish Academy of Fine Arts. 97-100.

Eidgenössische Materialprüfungs- und Versuchsanstalt (EMPA). 1987. Untersuchungsbericht, Bestimmung der Lichtechtheit, Diasec Sovilla SA. St. Gallen, Switzerland.

Erhardt, D., & Mecklenburg, M.F. 1995. Accelerated vs. natural aging: effect of aging conditions on the aging process of cellulose. In Materials issues in art and archaeology IV: symposium held May 16-21, 1994, Cancun, Mexico, ed. P. B. Vandiver, J. R. Druzik, J. L. Galván Madrid, I. C. Freestone, and G. S. Wheeler. Materials Research Society. 247-70.

Feller, R.L. 1994. Accelerated aging: photochemical and thermal aspects. The Getty Conservation Institute.

Fenn, J. 1995. Secret sabotage: reassessing museum plastics in display and storage. In Resins: ancient and modern, ed. M. Wright, and J. Townsend. Aberdeen: Scottish Society for Conservation and Restoration. 38-41.

Fieux, R.E. 1984. Silicone polymers for relining of paintings. In Adhesives and consolidants: preprints of the contributions to the Paris congress, 2-8 September 1984, ed. N.S. Brommelle, E.M. Pye, P. Smith, and G. Thomson. London: The International Institute for Conservation of Historic and Artistic Works. 46-49.

Fuchs, O. 1989. Solvents and non-solvents for polymers. In Polymer handbook, 3rd ed. Ed. J. Brandup and E.H. Immergut. New York: John Wiley & Sons, Inc. VII/379-407.

106

GE Silicones. 1997a. Datasheet for contractors SCS1001 silicone sealant. <www.gesilicones.com>.

GE Silicones. 1997b. Material safety data sheet: TSE397 silicone electronics adhesive. October 14, 1997. <www.gesilicones.com>.

GE Silicones. 1998a. Construction SCS1200 silicone sealant, product description. <www.gesilicones.com>.

GE Silicones. 1998b. Material safety data sheet: SCS1201 RTV silicone rubber compound. November 6, 1998. <www.gesilicones.com>.

GE Silicones. 2000a. Datasheet for SS4179 ready-to-use transparent silicone primer. <www.gesilicones.com>.

GE Silicones. 2000b. Material safety data sheet: SCS1001 construction sealant. December 12, 2000. <www.gesilicones.com>.

GE Silicones. 2000c. Material safety data sheet: SS4179 silicone primer. December 12, 2000. <www.gesilicones.com>.

GE Silicones. 2000d. The chemistry of silicone rubber. <http://www.gesilicones.com/silicones/americas/business/portfolio/hce/workshops/paperpodhce0 2Part1.shtml>

107

GE Silicones. unknown date a. Datasheet for N-SIL®, RTV128 silicone modified alkoxy adhesive sealant. <www.gesilicones.com>.

GE Silicones, unknown date b. FAQ. <http://www.gesilicones.com/silicones/americas/business/answercenter/faq/rtv.shtml>

Gminder, L. 1983. Use of a dispersion adhesive for the permanent glueing of an image to a supporting sheet. Germany. Patent 32 17 932. November 17, 1983.

Hill Brothers Chemical Company. 1999. Material safety data sheet: acetic acid, glacial. August 17, 1999. <http://hillbrothers.com/msds/aag.htm>.

Hochheiser, S. 1986. Rohm and Haas – history of a chemical company. Philadelphia: University of Pennsylvania Press.

Horie, C.V. 1997. Materials for conservation. Oxford: Architectural Press.

Huber, H.D. 1998/99. Matthias Hoch – To the beat of the techno-systems. In Matthias Hoch – Speicher. Dogenhaus Galerie Leipzig, Goethe Institut Paris. Exhibition catalogue without page numbering.

Iler, R.K. 1979. The chemistry of silica: solubility, polymerization, colloid and surface properties, and biochemistry. New York: John Wiley & Sons.

Ilford Photo Corporation. 1988. Cibachrome: technical information: mounting and laminating.

108

Image Permanence Institute. 1998. User’s guide for A-D Strips: film base deterioration monitors, version 1.8. Rochester, NY: Image Permanence Institute, Rochester Institute of Technology.

Jenny Co. AG. 1997. Gurisil PR 435, Grundierung.

Jenny Co. AG. 1999a. Gurisil 575, Einkomponenten Klebe- und Dichtungsmasse.

Jenny Co. AG. 1999b. Gurisil 575, Safety data sheet. November 12, 1999.

Jenny Co. AG. 2000b. Gurisil PR 435, Safety data sheet. May 8, 2000.

Johnson, J.S. 1996. Accelerated aging: changes in swelling and melting point of photographic gelatin. In ICOM committee for conservation: 11th triennial meeting, Edinburgh, 1-6 Sept. 1996, preprints 2. London: International Council of Museums. 586-590.

Journal of the American Institute for Conservation. 1999. Guidelines for authors. Journal of the American Institute for Conservation 38(3): 485-487.

Katz, I. 1971. Adhesive materials: their properties and use. Long Beach, CA: Foster Publishing Company.

Kinloch, A.J. 1983. Introduction. In Durability of structural adhesives, ed. A.J. Kinloch. London: Applied Science Publishers. 1-42.

109

Kodak Professional Division. 1998. Effects of post-processing treatments on the image stability of color prints. Kodak Publication No. E-176. Rochester, NY: Eastman Kodak Co.

Lavédrine, B. 1996. Prediction of dark stability of colour chromogenic films using Arrhenius law and comparison after ten years of natural ageing. In Research techniques in photographic conservation: conference proceedings, 14-19 May 1995, ed. M. S. Koch, T. Padfield, J. S. Johnsen, and U. Bøgvad Kejser. Copenhagen: The Royal Danish Academy of Fine Arts. 71-76.

Lewis, F.M. 1962. The science and technology of silicone rubber. In Rubber chemistry and technology. 35(5). 1222-1275.

Lynch, W. 1978. Handbook of silicone rubber fabrication. New York: Van Nostrand Reinhold Co.

MACtac. unknown date. PermaTrans® IP2100 Performance Guide.

Maish, J.P. 1994. Silicone rubber staining of terracotta surfaces. Studies in conservation 39(4): 250-256.

McCormick-Goodhart, M.H. 1995. Moisture-content isolines of gelatin and the implications for accelerated aging tests and long-term storage of photographic materials. Journal of imaging science and technology 39(2): 157-162.

McCormick-Goodhart, M.H. 1996. The allowable temperature and relative humidity range for the safe use and storage of photographic materials. Journal of the society of archivists 17(1): 7-21.

110

National Institute for Occupational Safety and Health (NIOSH). 1994. Acetic acid: method 1603, issue 2. In NIOSH Manual of Analytical Methods, 4th ed.

National Institute for Occupational Safety and Health (NIOSH). Unknown date. Pocket guide to chemical hazards. U.S. department of health and Human Services.

Nicholson, C., and O’Loughlin, E. 1996. The use of A-D strips for screening conservation and exhibit materials. The book and paper group annual 15: 83-85.

Nishimura, D.W. 1996. The practical presentation of research studies on film stability. In Research techniques in photographic conservation: conference proceedings, 14-19 May 1995, ed. M. S. Koch, T. Padfield, J. S. Johnsen, and U. Bøgvad Kejser. Copenhagen: The Royal Danish Academy of Fine Arts. 85-92.

Noll, W. 1968. Chemistry and technology of silicones. Orlando, Florida: Academic Press, Inc.

Oddy, W.A. 1975. The corrosion of metals on display. In Conservation in archaeology and the applied arts. Pre-prints of the contributions to the Stockholm congress, June 1975. London: IIC. 235-237.

Plueddemann, E.P. 1991. Silane coupling agents, 2nd ed. New York: Plenum Press.

Reilly, J.M. 1996. Accelerated aging tests. In Research techniques in photographic conservation: conference proceedings, 14-19 May 1995, ed. M. S. Koch, T. Padfield, J. S. Johnsen, and U. Bøgvad Kejser. Copenhagen: The Royal Danish Academy of Fine Arts. 77-84.

111

Rochow, E.G. 1987. Silicon and silicones. Berlin: Springer Verlag.

Sale, D. 1988. The effects of solvents on four plastics found in museum collections: a treatment dilemma. In Preprints of contributions to the modern organic materials meeting held at the university of Edinburgh, 14-15 April 1988. Edinburg: Scottish Society for Conservation and Restoration. 105-114.

Sale, D. 1995. An evaluation of six adhesives for repairing poly(methyl methacrylate) objects and sculpture: changes in tensile strength and colour after accelerated ageing. In Resins: ancient and modern, ed. M. Wright, and J. Townsend. Aberdeen: Scottish Society for Conservation and Restoration. 17-32.

Seferis, J.C. 1989. Refractive indices of polymers. In Polymer handbook, 3rd ed. Ed. J. Brandup and E.H. Immergut. New York: John Wiley & Sons, Inc. VI/451-461.

Sigma-Aldrich Co. 1997. The Aldrich library of FT-IR spectra, 2nd ed., 1-3. Milwaukee, WI: Sigma-Aldrich Co.

Sovilla-Bruhlhart, H. 1969. Procédé pour protéger un film diapositif contre la formation des anneaux de Newton. Switzerland. Patent 483 658. December 31, 1969.
Sovilla-Bruhlhart, H. 1970a. Panneau – support d’un tirage photographique. Switzerland. Patent 489 040. April 15, 1970.

Sovilla-Bruhlhart, H. 1970b. Panneau – support de tirage. France. Patent 2.030.331. November 13, 1970.

112

Sovilla-Bruhlhart, H. 1973a. Plaque transparente destinée à un support d’un tirage photographique ou d’imprimerie. Switzerland. Patent 534 901. March 15, 1973.

Sovilla-Bruhlhart, H. 1973b. Mit einer Trägerplatte verbundenes Bild. Germany. Patent 22 56 656.9. August 23, 1973.

Sovilla-Bruhlhart, H. 1974. Procédé de collage de tirages photographiques. Switzerland. Patent 546 968. January 31, 1974.

Sovilla-Bruhlhart, H. 1977. Bild, das rückseitig mit einer Trägerplatte unter Verwendung eines vulkanisierbaren Siliconklebstoffes verbunden ist. Germany. Patent 22 65 275.1. May 12, 1977.

Swan, A. 1987. Problems in the conservation of silver gelatine prints. In Conservation of library and archive materials and the graphic arts, ed. G. Petherbridge. London: Butterworths, Society of Archivists, Institute of Paper Conservation. 273-285.

Tétreault, J. 1992. La mesure de l’acidité des prduits volatils. Journal of IIC-CG 17: 17-25. (English translation: “Measuring acidity of volatile products” at the Canadian Conservation Institute website at <http://www.cci-icc.gc.ca/>)
Wagner, S. 1989. A preliminary study: consolidation of gelatin glass plate negatives with organosilanes. In Topics in photographic preservation 3. Washington, DC: Photographic Materials Group, American Institute for Conservation. 69-85.

Wake, W.C. 1982. Adhesion and the formulation of adhesives. 2nd ed. London: Applied Science Publishers.

113

Warrick, E.L. 1990. Forty years of firsts: the recollections of a Dow Corning pioneer. New York: McGraw-Hill Publishing Company.

Wilhelm, H. 1993. The permanence and care of color photographs: traditional and digital color prints, color negatives, slides, and motion pictures. Grinnell, Iowa: Preservation Publishing Company.

Yashvili, N. 1978. Testing new transparent siliconorganic and some organic polymers for conservation of archaeological wood. In ICOM Committee for Conservation, 5th triennial meeting, Zagreb 1978: preprints (2). 78/7/3/1-78/7/3/5.