Tricks of the Shade: Preservation of Chemically Developed Shading Papers Based on DuoShade Samples

ABSTRACT

From the late 1920s through the early 2000s, artists sometimes utilized chemically developed shading papers, or “chemigraphic” papers, to add dimensional effects to their drawings. These papers were precoated with patterns that were selectively revealed by applying a liquid developer. Drawings on these types of papers often show signs of deterioration, such as discoloration and fading of the developed patterns and staining of the paper support in areas where the developer was applied. This study investigates samples made from Grafix DuoShade chemigraphic papers and developers manufactured in the 1990s. These samples were analyzed to better understand their chemical makeup, deterioration, and preservation needs. The samples were characterized using macro X-ray fluorescence scanning and complementary spectroscopic techniques. To determine optimal storage environments, samples were artificially aged in different storage enclosures and colorimetry was used to compare the artificially aged samples to unaged controls. To determine optimal display environments, the light sensitivity of the samples before and after artificial aging was measured using microfadeometry. Based on the results, preservation recommendations for handling, storage, and display are provided.

RÉSUMÉ

De la fin des années 1920 jusqu'au début des années 2000, certains artistes ont utilisé des papiers tramés développés chimiquement, ou papiers « chimigraphiques », dans le but d'ajouter des effets dimensionnels à leurs dessins. Ces papiers étaient préencollés avec une émulsion contenant une trame prédéterminée et développée de manière sélective par l'application d'un révélateur liquide. Les dessins sur ces types de papiers présentent souvent des altérations comme le jaunissement et la décoloration de la trame développée, tandis que le papier de support est taché là où le révélateur a été appliqué. Cette recherche se concentre sur des échantillons des papiers et révélateurs chimigraphiques Graphix DuoShade fabriqués dans les années 1990. Ces échantillons ont été analysés pour comprendre leur composition chimique, leurs mécanismes de détérioration et comment les préserver. Ces échantillons ont été analysés en utilisant la cartographie de fluorescence des rayons X ainsi que d'autres méthodes spectroscopiques complémentaires. Pour déterminer l'environnement optimal de stockage, les échantillons ont été vieillis artificiellement dans un certain nombre de boîtes de conditionnement et des mesures de colorimétrie ont été utilisées pour comparer les échantillons vieillis avec ceux non-vieillis servant de control. Pour déterminer les meilleures conditions d'exposition, la sensibilité à la lumière des échantillons avant et après vieillissement artificiel a été mesurée grâce à la micro-décoloration. Basées sur ces résultats, des recommandations en termes de manipulation, stockage et exposition sont fournies. Traduit par Sophie Barbisan.

RESUMO

Do final da década de 1920 até o início dos anos 2000, os artistas às vezes utilizavam papéis de sombreamento desenvolvidos quimicamente, ou papéis “quimigráficos”, para adicionar efeitos dimensionais aos seus desenhos. Esses papéis foram pré-revestidos com padrões que foram revelados seletivamente pela aplicação de um revelador líquido. Os desenhos nestes tipos de papéis geralmente apresentam sinais de deterioração, como descoloração e desbotamento dos padrões desenvolvidos e manchas no suporte do papel nas áreas onde o revelador foi aplicado. Este estudo investiga amostras feitas de papéis quimiográficos Grafix DuoShade e reveladores fabricados na década de 1990. Essas amostras foram analisadas para melhor compreender sua composição química, deterioração e necessidades de preservação. As amostras foram caracterizadas por macrofluorescência de raios X e técnicas espectroscópicas complementares. Para determinar os ambientes de armazenamento ideais, as amostras foram envelhecidas artificialmente em diferentes compartimentos de armazenamento e a colorimetria foi usada para comparar as amostras envelhecidas artificialmente com os controles não envelhecidos. Para determinar os ambientes de exibição ideais, a sensibilidade à luz das amostras antes e depois do envelhecimento artificial foi medida usando microfadeometria. Com base nos resultados, são fornecidas recomendações de preservação para manuseio, armazenamento e exposição. Traduzido por Beatriz Haspo.

RESUMEN

Desde finales de la década de 1920 hasta principios del 2000, los artistas han utilizado ocasionalmente el llamado papel de sombreado de revelado químico o “Quimiagrafías” para agregar efectos dimensionales a sus dibujos. Este tipo de papeles contaban con una capa de patrones que se revelaban de manera selectiva al aplicar un líquido revelador. Los dibujos sobre este tipo de papeles suelen mostrar signos de deterioro tales como decoloración y desvanecimiento de los patrones revelados y manchas del soporte de papel en las zonas donde se aplicó el revelador. En esta investigación se estudiaron muestras de los papeles Grafix DuoShade y reveladores fabricados en la década de 1990 y se analizaron para comprender mejor su composición química, deterioro y sus necesidades de preservación. Dichas muestras se caracterizaron mediante el escaneo de fluorescencia de rayos X macro y técnicas espectroscópicas complementarias. Con el fin de determinar las condiciones ambientales óptimas para su almacenamiento, las muestras se sometieron a pruebas de envejecimiento en diferentes tipos de guardas de almacenamiento, y se utilizó colorimetría para comparar las muestras envejecidas artificialmente con las muestras de los controles no envejecidos. Posteriormente para determinar las condiciones óptimas de exhibición, se midió la sensibilidad a la luz de las muestras antes y después del envejecimiento artificial mediante microfadeometría. A partir de los resultados obtenidos, se ofrecen recomendaciones de conservación para la manipulación, el almacenamiento y la exhibición. Traducción Mirasol Estrada y Mónica Pérez; revisión Amparo Rueda e Irene Delaveris.

KEYWORDS:

• DuoShade
• chemigraphic
• MA-XRF scanning
• Raman
• colorimetry
• microfadeometry
• artificial aging
• artist materials

1. Introduction

From the late 1920s through the early 2000s, artists sometimes utilized chemically developed papers to create shading effects in their drawings. The umbrella term “chemigraphic” (drawn with chemistry) was developed by the authors as a means to describe these papers without the use of proprietary labels. Line drawings were done on papers that were precoated with unobtrusive, non-actinic patterns. Left undeveloped, these patterns would remain invisible to reprographic photography. The patterns could be selectively revealed by artists by applying a liquid developer with a brush to create textural shading in the form of repeated lines and dots (Figure 1). Chemigraphic patterns provided distinct advantages to artists, including cartoonists, graphic designers, and architects. They reproduced well in subsequent halftone printing – as photo reproductions in newspapers, comic books, magazines, and architectural plans – and achieved in an instant what would have otherwise taken a long time to render by hand.

Figure 1. Drawing with revealed chemigraphic line patterns (parallel lines and crosshatching). Note the yellow-brown staining of the paper where chemigraphic developers were applied and the discoloration of some areas of the crosshatching pattern. Detail: Pat Oliphant, [Mr. Fischer seems to be ready now … shall we commence, Mr. Spassky?], 1972, ink on chemigraphic paper, sheet 30 × 45 cm, Lot 15350 no. 17, Library of Congress Prints and Photographs Division.

Drawings on chemigraphic papers often show signs of deterioration, such as discoloration and fading of the developed patterns and staining of the paper support in areas where the developer was applied. Found in libraries, archives, and museums, yet understudied and often unidentified, chemigraphic materials require deeper investigation. This research starts that investigation by analyzing samples made using DuoShade chemigraphic products manufactured in the 1990s by Grafix.

1.1. History

In 1924, representatives of Craftint Manufacturing Company filed a patent for photographically based chemigraphic papers. The patent application was ultimately forfeited due to the impracticality of the process (Craftint Mfg. Co. v. Baker 1938), however by 1927 a simplified method was introduced. According to the patent for Mistograph (Baker 1929), the technique involved producing a silver-based photographic image of one repeated line or dot pattern on paper, using traditional photographic developers and fixers in the process. The pattern was rendered completely invisible by treating it with mercuric chloride, allowing an artist to draw freely on the paper. Areas of the pattern could then be selectively revealed by an artist with sodium hydroxide to create areas of shading. The patent does not elaborate on the exact chemical reactions that rendered the pattern invisible or allowed for its redevelopment.

In 1928, a non-photographic method to produce chemigraphic papers was introduced by Craftint (Maier and Swaysland 1930). In this process, a pattern was printed on paper with an ink composed of white lead sulfate, which was virtually invisible on white paper. The artist selectively revealed the pattern with a soluble sulfide, which reacted with the white lead sulfate to form black lead sulfide. Any mistakes could be corrected with hydrogen peroxide to revert the black lead sulfide back to white lead sulfate.

In 1935, a non-photographic process for developing two patterns to create two tones on the same sheet was introduced (McIntosh 1936). To produce this paper, two colorless patterns were printed on the surface with sensitizers; one pattern could be revealed with a “light” developer or both patterns simultaneously revealed using a “dark” developer. This allowed artists to create two distinct tones (Figure 2). The patent describes mercurous chloride and lead sulfate as the sensitizers for the two patterns and ammonium hydroxide and a soluble sulfide as the developers to reveal the patterns, respectively. Although McIntosh (1936) considered these the most practical chemicals for this use at the time, he also described several other chemical variations.

Figure 2. Two colorless patterns printed on paper and activated with a “light” developer that reveals one pattern (dots; above) and a “dark” developer that simultaneously reveals both patterns (dots and rectangles between them creating crosshatched lines; below). Detail from McIntosh 1936.

From 1935 until at least 1953, improvements and variations were made on one- and two-tone processes (Isler 1940; McIntosh 1940a; McIntosh 1940b; McIntosh 1942a; McIntosh 1942b; McIntosh 1944; McIntosh 1945; McIntosh 1947; McIntosh 1952; McIntosh 1953; McLellan 1935; Sanders 1939; Sanders 1940; Sanders 1944; Sanders 1946; Swaysland 1935). The possible variations of chemicals used for preparing the papers and for developing the patterns described in the patents from this time period (1935-1953) are too numerous to describe in detail here. Further, patents often stated that many sensitizers and developers were interchangeable, where one could be used as the other. Although countless variations were introduced, non-photographic methods dominated patent applications during this period.

There were several notable developments that emerged during this early phase of invention. Two patents introduced a transparent film or paper overlay, printed with an invisible chemigraphic pattern, which was attached onto a line drawing on a non-chemigraphic support, then selectively developed by the artist (McLellan 1935; Swaysland 1935). Other patents described achieving different tones by varying the length of time a developer was in contact with the pattern, where longer times resulted in larger developed areas. This was achieved by printing a pattern with a varnish on a sensitized paper, which would slow the penetration of the developer solution. One could either blot the developer to arrest the intensity at the desired level (McIntosh 1940a) or use a “slow” and “fast” developer to produce variations in tone (McIntosh 1940b). Multicolor chemigraphic patterns were introduced by 1940 (Sanders 1940; Sanders 1944). In the 1940s, registration between two tones was improved by printing the two patterns as intersecting sets of parallel lines at 90 degrees (McIntosh 1945) or as small dots superimposed by larger dots (McIntosh 1944). In both cases, a light tone was created by developing one pattern and a dark tone by simultaneously developing both patterns.

Chemigraphic papers made by Craftint Manufacturing Company, which were available from the late 1920s until 1963, were branded under the proprietary names SingleTone and DoubleTone. In 1963, Ohio Graphic Arts Systems was founded with one product line: chemigraphic papers branded with the proprietary names UniShade and DuoShade (Grafix 2021). These could be rebranded versions of the earlier Craftint papers, assuming Ohio Graphic Arts Systems bought out the Craftint chemigraphic product line, but further study is needed to confirm a consistency between these two manufacturers’ formulations. Ohio Graphic Arts Systems changed its name to Grafix in 1990 (sometimes referred to as Graphic Arts Systems Inc.) (Grafix 2021). According to Grafix representative Jill Walters (pers. comm., 1996), their method of DuoShade production never changed since the product was introduced in 1963, although some of the components may have changed due to discontinuation or change in manufacturer. As noted by the current president of consumer products at Grafix, Hayley Prendergast (pers. comm., 2021), Grafix stopped production of the UniShade and DuoShade products in 2007 and sold its remaining stock by 2009. The products had become obsolete in the digital age (Gardner 2009).

Both manufacturers, Craftint and Ohio Graphic Arts Systems / Grafix, offered single pattern and double pattern papers producing a black image tone when developed. Universally, the single pattern papers were developed using one solution, while double paper patterns came with two distinct solutions: a “light tone” developer to reveal a single pattern or a “dark tone” developer to reveal two intersecting patterns simultaneously. Left undeveloped, the patterns are colorless or light blue. For UniShade, the undeveloped pattern is colorless. For DuoShade, the undeveloped pattern revealed by the light and dark tone developers is colorless, while the undeveloped pattern revealed by the dark developer is tinted blue so that it is identifiable to artists but still invisible to photography.

Multiple patterns and paper supports were available from Grafix. Each pattern had its own identification number and varieties included regular and irregular black dots, lines, and stipples (Figure 3). For UniShade and DuoShade, “the number of the pattern refers to the number of lines or dots per inch. For example, UniShade #32D means a dot pattern of 32 per inch. DuoShade #232 means two patterns, each with 32 lines per inch. The light pattern is the equivalent of approximately 25% black; the dark pattern is about 50% black.” (Grafix 1996). By the 1990s, UniShade and DuoShade were manufactured on 44.5 cm x 61 cm (17.5 in. x 24 in.) supports that had a reproduction area of 42 cm x 58.5 cm (16.5 in. x 23 in.). Paper supports included translucent and opaque papers of different weights and textures, such as 3-ply hot-press opaque paper similar to cardstock (called “Bristol Board”); translucent paper (called “Tracing Vellum”); 2-ply cold-press opaque paper (called “Strathmore”), and 3-ply cold-press opaque paper (called “Rough”) (Grafix 1996).

Figure 3. Several Grafix DuoShade and UniShade patterns available in 1996. X = discontinued patterns. Detail from Grafix 1996.

1.2. Condition

Drawings on chemigraphic papers can exhibit multiple signs of deterioration. These observations were made based on the hundreds of catalogued chemigraphic drawing papers in the Prints and Photographs Division of the Library of Congress. DuoShade papers were the focus of condition observations, as DuoShade samples were used in this study. DuoShade papers were identified based on their production period, 1963–2007, and the distinct patterns noted previously. Common condition issues from this period, and often found in earlier (pre-1963) examples of chemigraphic papers, include (Figure 4):

1. Yellow-brown staining of the white paper in areas where either the light or dark developer was applied (Figure 4a).

2. Discoloration of the developed pattern from a more neutral black to a warmer, browner tone. Usually the pattern that shifts is that which was revealed by only the dark developer (Figure 4b).

3. Fading of the revealed pattern from black to gray or colorless. This fading is typically uneven and can follow the pattern of brushstrokes. Usually the faded pattern is in the “light tone” (Figure 4c).

Figure 4. Common condition issues in a DuoShade example, where the light tone is a regular dot pattern and the dark tone is the regular dot pattern surrounded by irregular stipples: a) yellow-brown staining of the white paper where the light and dark developers were applied; b) discoloration of the irregular stipple pattern revealed by the dark developer; c) fading of the regular dot pattern revealed by the light developer. Detail: Pat Oliphant, [And, being a non-Communist junta, we can count on automatic U.S. support!], 1973, ink on chemigraphic paper, sheet 31 × 45 cm, Lot 15351 no. 47, Library of Congress Prints and Photographs Division.

Less commonly, the brownish pattern can appear in areas not originally developed by the artist. These areas are most often around the edges of the paper support, where the objects are handled. Direct contact of the support with masking tape can also develop the pattern unintentionally. Contrasting the frequently poor condition of the reproduction area, chemigraphic paper supports are usually in stable condition.

Renowned political cartoonist Pat Oliphant (1935-) observed the instability of chemigraphic products, which he used from the 1950s through the early 1980s, according to an interview with the authors in 2019. He noticed brown staining in some areas where the developer was applied, subsequent interference with the legibility of his drawings, and that the developer solution itself could turn from colorless to brown within days, a year, or several years. Experiments were done in the late 1980s by book conservator Don Etherington and researcher Bruce Humphrey of Union Carbide, working with Oliphant, in order to design a conservation treatment for chemigraphic papers (Humphrey 1984). These experiments led to the consolidation of approximately 50 of Oliphant’s drawings on chemigraphic papers with parylene in 1987 (Etherington, pers. comm.). The identity of the drawings that received this treatment and where they reside is unknown to the authors, therefore the effect of the parylene treatment on chemigraphic papers could not be studied.

2. Methods

2.1. Samples

Samples made from Grafix DuoShade papers and developers manufactured in the 1990s were used for this study. The materials were obtained contemporaneously from the manufacturer in an “Introductory Kit” and include four 21.6 x 27.9 cm (8.5 x 11 in.) DuoShade papers and two developing solutions, labeled “Light Tone Developer No. 1” and “Dark Tone Developer No. 2.” The light developer was colorless and the dark developer was discolored yellow at the time they were used for this study. No expiration dates were listed on the products. The papers represent four patterns: 224, 232, 240, and 270. Patterns 224, 232, and 240 are composed of lines, where the light developer reveals a set of parallel lines while the dark developer produces crosshatching. Pattern 270 is composed of dots and stipples, where the light developer produces regular dots and the dark developer reveals these dots surrounded by irregular stipples. On the four papers, the undeveloped pattern revealed by the light and dark tone developers is colorless, while the undeveloped pattern revealed by the dark tone developer is light blue. Patterns 232, 240, and 270 have the same supports: a smooth, white opaque paper similar to cardstock, approximately 0.5 mm thick, and is probably what was referred to by the company as “Bristol board.” The support for Pattern 224 is a smooth, translucent white paper, approximately 0.1 mm thick, and is probably what was referred to by the company as “tracing vellum.” Although not representative of all types of chemigraphic papers throughout their history, the samples are a starting point to better understand the process.

The Material Safety Data Sheet (MSDS) for the materials was provided contemporaneously from the manufacturer (Grafix 1991). The 1991 MSDS notes that the drawing supports contain mercury and copper, but contains no other chemical information about the supports. The ingredients for the developing solutions from the 1991 MSDS are listed in table 1. To the knowledge of the authors, the information in this MSDS has not been previously published or widely distributed. Although a few of the patents refer to some of the ingredients listed in the developers, their exact functions are not fully explained and are uncertain.

Table 1. Ingredients in DuoShade light and dark tone developers from the 1991 MSDS.

Light Tone Developer No. 1		Dark Tone Developer No. 2
Ingredients	% Composition	Ingredients	% Composition
Zinc nitrate, Zn(NO₃)₂	<10	Sodium tungstate, Na2WO4	<10
Thiourea, SC(NH2)2	<10	Thiourea, SC(NH2)2	<10
Sodium tungstate, Na2WO4	<15	Ammonium hydroxide, NH₄OH	<5
Sulfuric acid, H2SO4	<10	Diethylcarbamodithioic acid, sodium salt, C5H11NNaS2	<15
Distilled water, H₂O	>40	Distilled water, H₂O	>40

The samples for this study were made by cutting the four papers into strips. The light developer was applied lengthwise on the left of each strip (henceforth referred to as the “light-developed area”) and the dark developer was applied lengthwise on the right of each strip (henceforth referred to as the “dark-developed area”) with a brush. The two developers overlapped in the center of each strip to see if this caused a change in appearance; it did not, the areas of overlap appeared similar to the areas only developed with the dark developer. The strips were then cut into equal rectangles, and the top quarter of each sample was cut to save as a control (Figure 5) and the bottom three-quarters of each sample was artificially aged. This resulted in eighteen control samples that were not aged and eighteen samples that were artificially aged, for each of the four types of patterns.

Figure 5. Control samples of DuoShade patterns 224 (top left), 232 (top right), 240 (bottom left), and 270 (bottom right). The light tone developer was applied on the left and the dark tone developer was applied on the right of each sample.

Making the samples clarified two of the common observations related to DuoShade. In all of the control samples, the pattern revealed by only the dark developer was brown and the dark developer stained the paper yellow-brown as soon as it was applied. Therefore, these issues may be in part due to deterioration and discoloration of the dark tone developer solution itself prior to application.

2.2. Macro-XRF

Four control samples, one of each pattern type (224, 232, 240, and 270), underwent scanning macro-X-ray fluorescence spectroscopy (MA-XRF) to help visualize and clarify the elemental layout of the samples with very small, complex patterns and to confirm major elemental components of the patterns and developers listed in the MSDS. MA-XRF non-invasively identifies the elements present, their relative amounts, and their distribution across the samples. Non-destructive analytical techniques were chosen for this study as they also can be used on historical objects for comparison. Due to limitations of XRF and the instrumentation, elements lighter than aluminum could not be identified nor mapped in this study.

The four control samples were mapped together using a Bruker M6 Jetstream large area macro-XRF scanner with a rhodium X-ray tube voltage of 50 kV and current of 300μA. The samples were supported on a sheet of acrylic (6.35 mm thick) that was raised so that at least 50 mm of air was behind it during XRF measurements. The integration time for each pixel was 125 ms. The X-ray spot size was 100μm with a step size of 75μm. The helium flow for the instrument was kept to 1L/min to enhance the signal of lighter elements during the mapping measurements. Bruker M6 Jetstream software version 1.5.2.47 was used for data collection, processing, and generating elemental maps. Elemental maps were generated with the “fast deconvolution” feature in the Bruker software that resolves peak overlaps by appointing intensity counts in every channel to X-ray lines based on the probability that a count belongs to one of the selected elements. Sum spectra of selected pixel regions of interest from the maps were used to verify the presence of the elements.

2.3. Point XRF

Point X-ray fluorescence spectroscopy (XRF) was used to acquire spectra on one control sample (Pattern 240), as the MA-XRF results showed similar distributions for all of the control samples of different pattern types. These representative spectra can be used as references to help identify chemigraphic papers in collections in the future. Spectra were obtained from several areas, including from the two developed patterns, as well as from spaces on the paper support between the patterns in an undeveloped area, light-developed area, and dark-developed area. Point spectra were acquired using the Bruker M6 Jetstream large area macro-XRF scanner with a rhodium X-ray tube voltage of 50 kV and current of 300μA. The X-ray spot size was 100μm and the integration time was 300s. A helium flush was not used for these single-point XRF measurements due to a helium shortage during the time of acquisition. Bruker M6 Jetstream software version 1.5.2.47 was used for data collection and Artax software version 7.2.5.0 was used for data evaluation.

2.4. Raman spectroscopy

To further characterize the patterns, a Bruker Senterra Raman spectrometer coupled with an Olympus microscope was used to acquire spectra with a 785 nm laser excitation. Measurements were taken in situ­ using the 50x objective, with a laser power of 1-10 mW for 25–45 secs for 1–5 coadditions depending on the sample. Mercury (I) chloride (Hg₂Cl­₂), 99.5% (Thermo Scientific, CAS 10112-91-1), and mercury (II) chloride (HgCl₂), 99.5% (Thermo Scientific CAS 7487-94-7), were used as references.

2.5. Artificial aging

The samples reserved for aging (18 for each of the four pattern types) were artificially aged in three typical archival storage enclosures: buffered paper, unbuffered paper, and polyethylene terephthalate film, hereafter referred to as polyester film. To create the enclosures, the buffered and unbuffered papers were folded into non-adhesive envelopes and the polyester film was heat-welded into a U-shape. Samples were placed in individual enclosures and hung from the top of the aging chamber using clips. The aging took place over 14 days at 50°C and 60%RH in a Caron Environmental Chamber Model 7000-33-1.

Permalife buffered paper (20 lb.) which passed Library of Congress Specification 100–100 for Buffered Paper Stock for the Storage of Artifacts (Library of Congress 2016a), stipulating a calcium carbonate alkaline reserve of 2-5% and pH 8-9.5, was used for the buffered paper enclosure. CT (formerly Chartham) Translucent Vellum (24 lb.), a smooth, translucent paper used to store works on paper with friable media, was chosen for the unbuffered paper enclosure. Although the smooth and translucent qualities of the paper were not required for the purposes of this project, this paper was selected because it passed Library of Congress Specification 100–101 for Unbuffered Paper Stock for the Storage of Artifacts (Library of Congress 2016b), stipulating no alkaline reserve and a neutral pH. Polyester film (3 mil), which passed Library of Congress Specification 500–500 for Polyester: Poly(ethylene-terephthalate) Film for the Storage of Artifacts (Library of Congress 2016c), was used for the polyester film standard.

2.6. Colorimetry

Average color measurements were calculated for the light-developed (left side) and dark-developed (right side) areas on each of the control and aged samples. These measurements were made using digital photography. The control and aged samples for each pattern type (224, 232, 240, and 270) were simultaneously photographed on a copy stand, along with a Robin Myers Imaging color target. All variables, including light source, light intensity, light angle, camera and lens used, height of the camera from the samples, and exposure parameters, were consistent throughout capture. The photographs were processed in Adobe Photoshop. As raw digital negatives (DNG), they were white balanced using the N8 gray patch of the color target placed next to the samples. As tagged images (TIFF), the light-developed and dark-developed areas of each control and aged sample were cropped separately. The average of all colors in each area was taken with the Blur > Average filter. This average color was read with the color sampler tool as CIELAB coordinates. The effect of different storage enclosures was reflected in the overall color difference, ΔE (CIE 2000), between the averaged light-developed areas of unaged and aged samples and between the averaged dark-developed areas of unaged and aged samples. The average of each area was evaluated based on the unevenness of the developer and complexity and small sizes of the patterns. In addition, although historic examples show lightening, darkening, and hue changes, all of which would not be possible to clarify through averages, the samples studied here only exhibited darkening and hue shifts after aging, discussed later in the results.

2.7. Microfadeometry

Microfade test exposures used visible wavelengths (400-700 nm). Each test was five minutes in duration at 600 mlm, approximating the same visible light exposure as would be received in five years of exhibition in a gallery at 50 lux. Tests used a custom-made Paul Whitmore microfade tester, comprised of an Ocean Optics HPX-2000 xenon light source, UV filter, fiber optic cables arranged in a 0°/45° measurement geometry, Control Development Silicon PDA (photo-diode array) spectrometer, and accompanying CDI Spec32 data acquisition software. Measurements used a standard illuminant D65, 2° standard observer, and change was measured in ΔE (CIE 1994). Results were compared to Blue Wool Standards measured simultaneously. Five different areas were tested on one control and one aged sample of Patterns 224, 232, and 240. The spaces between the revealed patterns in both the light- and dark-developed areas were tested, as were the patterns themselves, including the pattern revealed by the light and dark developers (in both the light- and dark-developed areas) and the pattern revealed by the dark developer only (in the dark-developed area). For one control and one aged sample of Pattern 270, only four tests were done on each (on the spaces between the revealed patterns and on the patterns in the light- and dark- developed areas) as it was difficult to distinguish between the dot and stipple patterns in the dark-developed areas. Microfade testing was completed following materials characterization and colorimetry. As the results of materials characterization and colorimetry were consistent across the four patterns, an adequate sample size for microfadeometry was achieved through testing the same kinds of areas in all four patterns.

3. Results and discussion

3.1. Sample characterization

The 1991 MSDS for these samples do not identify the exact compounds comprising the patterns beyond stating they contain mercury and copper. MA-XRF showed the elemental differences in the two patterns on each sample (Figure 6). The presence of these elements in the patterns is supported by point XRF spectra (Figure 7). In all four control samples, the pattern developed by both the light and dark developers has strong mercury (Hg) and chlorine (Cl) signals (Figure 6b-c). The variation in the chlorine map may be due to the presence of other elements in the developers attenuating or blocking the signal from chlorine in the pattern. The co-localization of mercury and chlorine suggests the presence of either mercury (I) chloride, Hg₂Cl₂, or mercury (II) chloride, HgCl₂, in the pattern developed by both the light and dark developers. Raman spectroscopy was used to distinguish which form was present, as Hg₂Cl₂ has characteristic peaks at 167 and 276 cm⁻¹, while HgCl₂ has characteristic peaks at 73, 125, and 313 cm⁻¹. Raman measurements were taken of the undeveloped colorless pattern on the edge of control samples of Patterns 224, 240, 270. Spectra from these areas produced sharp peaks at 167 and 276 cm⁻¹, indicating mercury (I) chloride was used to create the undeveloped colorless pattern. A developed area of this pattern was measured on a control sample of Pattern 270 and the resulting spectrum has weak peaks at 167 and 276 cm⁻¹, suggesting some remaining Hg₂Cl₂ is present, but most of the material has reacted to a different form that could not be identified with Raman and would benefit from further research.

Figure 6. Reference image (a) and MA-XRF elemental maps (b-c) of control samples elucidating the makeup of the patterns: a) samples in visible light; b) elemental map of mercury (yellow) and copper (blue); c) elemental map of chlorine (pink). In each image and map, the control samples are positioned in the same orientation: 224 (top left), 232 (top right), 240 (bottom left), and 270 (bottom right), with the light tone developer applied on the left and the dark tone developer applied on the right of each sample.

Figure 7. Point XRF spectra comparing the two patterns, including the paper support for reference. Black: paper support between undeveloped patterns, with no developer solution. Blue: pattern developed by both the light and dark tone developers, containing mercury and chlorine. Orange: pattern developed only by the dark tone developer, containing copper.

In all four control samples, the pattern which is only developed by the dark developer has strong signals of copper (Cu) (Figure 6b). The full copper compound in undeveloped and developed areas of this pattern could not be distinguished with XRF or Raman. Three patents include copper-based patterns (McIntosh 1940a; McIntosh 1947; McIntosh 1953). In the first patent to refer to copper, McIntosh (1940a) lists “zinc di-n-butyl di-thiocarbamate” as a possible sensitizer and “copper compounds” as its developer, and then notes that the sensitizer and developer may be interchangeable. In McIntosh (1947), the patent mentions a pattern of “an insoluble copper compound developed with a soluble sulfide solution or a solution of a soluble salt of dialkyl dithiocarbamic acid such as sodium diethyl dithiocarbamate.” Interestingly, McIntosh (1947) also details a separate mercurous chloride pattern on the same sheet, developed with an aqueous solution of thiourea. In McIntosh (1953), the patent suggests a copper-based pattern, but does not elaborate. As these patents do not help clarify this pattern, further analysis is necessary to identify the undeveloped and developed copper compounds in the samples more completely, as well as to identify the binders for the patterns, but is beyond the scope of this study.

Results of MA-XRF (Figure 8) and point XRF (Figure 9) support the compositions of the developer solutions listed in the MSDS. Elemental maps show a sensitivity to the thickness of application and pooling of the developer solutions so conclusions regarding the relative amounts of the compounds in each developer cannot be made. Analysis supports the presence of zinc nitrate (ZnNO₃) in the light tone developer and sodium tungstate (NaWO₄) in both developers, as listed in the MSDS for the solutions. This is due to the increased signals of zinc throughout the light-developed areas (Figure 8b) and of tungsten in areas with both developers (Figure 8c) compared to the paper support. Additionally, an increase in sulfur signals, to varying degrees across the samples, correspond to developed surfaces (Figure 8d), consistent with sulfur-based compounds being present in both developer solutions. There also seems to be sulfur present in the patterns in both developed and undeveloped areas. The presence of sulfur is challenging to see in the single point XRF spectra (Figure 9), likely because a helium flow could not be used to enhance the signal of those measurements. Further research, such as destructive testing of the developing solutions, could definitively confirm the compositions of the solutions, but is beyond the scope of this study.

Figure 8. Reference image (a) and MA-XRF elemental maps (b-d) of control samples supporting the makeup of the liquid developers as stated in the MSDS: a) samples in visible light; b) elemental map of zinc (white); c) elemental map of tungsten (purple); d) elemental map of sulfur (orange). In each image and map, the control samples are positioned in the same orientation: 224 (top left), 232 (top right), 240 (bottom left), and 270 (bottom right), with the light tone developer applied on the left and the dark tone developer applied on the right of each sample.

Figure 9. Point XRF spectra comparing areas between patterns. Black: paper support between undeveloped patterns, with no developer solution. Blue: paper support with light tone developer and no pattern, containing tungsten, zinc, and sulfur. Orange: paper support with dark tone developer and no pattern, containing tungsten and sulfur.

3.2. Storage enclosures

After artificial aging in different storage enclosures (polyester film, unbuffered paper, and buffered paper), color measurements of the aged samples were compared with control (unaged) samples. Colorimetry results are summarized in table 2, where each ΔL*, Δa*, Δb*, and ΔE*₀₀ value is based on an average of six samples per pattern. See Appendix 1 for all colorimetry measurements.

Table 2. Average color differences between control samples and samples aged in polyester film, unbuffered paper, and buffered paper.

	Light-developed areas				Dark-developed areas
Pattern	ΔL*	Δa*	Δb*	ΔE*00	ΔL*	Δa*	Δb*	ΔE*00
	Polyester film
224	1.67	0.67	1.50	2.41	10.83	0.50	1.17	10.65
232	5.40	0.60	2.00	4.89	14.00	1.60	1.40	12.85
240	4.00	0.83	2.50	4.06	17.17	0.67	0.67	16.79
270	3.33	0.83	2.33	3.25	13.67	1.33	2.17	11.58
Average across patterns	3.60	0.73	2.08	3.65	13.92	1.03	1.35	12.97
	Unbuffered paper
224	2.17	0.17	1.17	2.14	3.83	0.33	1.67	3.97
232	2.17	0.67	2.33	2.79	7.17	1.00	0.17	6.74
240	2.00	1.00	2.50	3.24	6.83	0.83	1.00	6.43
270	1.67	0.83	1.67	2.39	7.17	0.33	0.83	5.97
Average across patterns	2.00	0.67	1.92	2.64	6.25	0.62	0.92	5.78
	Buffered paper
224	1.33	0.17	1.00	1.53	1.83	0.67	1.67	2.38
232	1.17	1.00	2.33	2.73	2.17	0.50	1.50	2.58
240	1.83	0.83	3.00	3.04	1.67	0.17	1.33	1.93
270	0.67	0.50	1.67	1.79	2.00	0.17	1.50	2.07
Average across patterns	1.25	0.63	2.00	2.27	1.92	0.38	1.50	2.24

Artificial aging resulted in measurable color change. The different patterns responded similarly to aging in each type of enclosure. Dark-developed areas changed more than light-developed areas, except for samples aged in buffered paper where the ΔE*₀₀ values of light- and dark-developed areas were similar. The ΔE*₀₀ for the dark-developed areas was generally a result of decreases in L* values, becoming darker. Differences in a* and b* values in dark-developed areas were minimal. In light-developed areas, the ΔE*₀₀ was generally a result of decreases in L* values, becoming darker, and increases in b* values, shifting towards yellow. Differences in a* values in light-developed areas were minimal.

Overall, aging in polyester film resulted in the most color change, especially in dark-developed areas. This was clearly apparent upon visual observation (Figure 10) and confirmed with color measurements. Aging in unbuffered paper also resulted in significant color change in dark-developed areas. The samples aged in buffered paper changed the least, however still had a ΔE*₀₀ over 2 which exceeds the threshold of perceivable color difference (Beltran et al. 2021). Based on these results, the authors recommend storing the materials in buffered paper enclosures.

Figure 10. Select control and aged samples of DuoShade pattern 240. The top quarter of each sample is an unaged control. The bottom three-quarters of each sample was aged in different enclosures: buffered paper (left), unbuffered paper (middle), and polyester film (right).

3.3. Light sensitivity

Of the 76 total areas tested with microfadeometry (including developed patterns and spaces between developed patterns on control and aged samples), 73 had ΔE₉₄ values below 1 and three had ΔE₉₄ values between 1 and 2. Generally, a ΔE between 1 and 2 indicates a Just Noticeable Difference (JND) and below this indicates no perceivable difference (Beltran et al. 2021). See Appendix 2 for all microfadeometry measurements. All three tests with a ΔE₉₄ over 1 indicate slight darkening, were on samples that had been aged, and were on patterns in light-developed areas. For reference, the ΔE₉₄ of Blue Wool 1 measured 5.3, the ΔE₉₄ of Blue Wool 2 measured 3.8, and the ΔE₉₄ of Blue Wool 3 measured 0.4. The average ΔE₉₄ value of the 76 DuoShade sample readings was 0.38 (with a standard deviation of 0.32), closest to Blue Wool 3.

Overall, it is difficult to draw conclusions regarding trends based on the microfadeometry results. Of the developed patterns tested, a slight majority (28 out of 44) showed no change during microfadeometry, while 13 slightly darkened (all in the pattern revealed by both the light and dark developers – eleven in light-developed areas and two in dark-developed areas), and only three slightly lightened. Of the spaces tested between the patterns, a slight majority (17 out of 32) lightened during microfadeometry, while eight showed no change, six darkened, and one both lightened and darkened during testing. The microfadeometry results for control and aged samples were not significantly different.

These results indicate that the materials are moderately light sensitive. The authors recommend conservative light levels during display similar to other works on paper (Library of Congress 2022), especially considering the sensitivity of other media often drawn on these materials.

4. Conclusion

This study begins to explore the complex nature of chemigraphic papers through the analysis of DuoShade samples from the 1990s, understanding that the samples may not be representative of all types of chemigraphic papers through their history.

Sample characterization using MA-XRF and Raman spectroscopies supported the presence of components in the light and dark tone developers provided in the MSDS, including zinc nitrate, sodium tungstate and sulfur-based compounds, and expanded on the material makeup of the patterns. Mercury (I) chloride was identified in the undeveloped colorless pattern (the pattern developed by both the light and dark tone developers), while copper was confirmed in the undeveloped light blue pattern (the pattern developed by the dark developer only). The copper compound could not be distinguished fully using these techniques and would benefit from further study. Out of caution for the mercury component and due to the condition issues noted from handling, the authors recommend wearing nitrile gloves when working with these items.

Microfadeometry results indicate that unaged and aged DuoShade samples are moderately light sensitive and the authors recommend heeding general light exposure guidelines for display of works on paper. Colorimetry results indicate that the storage enclosure is an important consideration for these materials, with buffered paper resulting in the least color change and polyester film resulting in the greatest color change after artificial aging of the DuoShade samples. However, aging in all types of storage enclosures resulted in perceivable color change, indicating that the condition issues noted may be primarily due to the inherent chemistry of the materials and suggests that these items would benefit from cold storage to slow their deterioration.

Chemigraphic materials are most likely underrepresented in collection databases due to several factors, including the previously limited availability of information regarding their history and materials characteristics, the lack of identifying information on the drawing supports, and confusion over different brand names. This study aims to help custodians identify and better understand chemigraphic materials, to provide preservation guidelines on display and storage, and to serve as a foundation for further investigation. Future research could include more invasive analyses of the developer solutions and samples, non-invasive analyses of historic chemigraphic drawings in collections, and analyses of chemigraphic materials beyond the brand and period that were the focus of this study.

Acknowledgements

The authors would like to thank: Sara Duke and Helena Zinkham (Prints and Photographs Division, Library of Congress); Aniko Bezur, Katherine Schilling, and Paul Whitmore (Institute for the Preservation of Cultural Heritage, Yale University); Don Etherington, Hayley Prendergast, Pat Oliphant and Susan Conway. With great foresight, Holly Krueger (Conservation Division, Library of Congress) obtained the DuoShade sample materials used in this research in the 1990s and made them available through the Center for Heritage Analytical Reference Materials (CHARM) in the Preservation Research and Testing Division, Library of Congress. The authors would also like to thank Elmer Eusman, Yasmeen Khan, and Andrew Robb (Conservation Division, Library of Congress) for their support.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Gwenanne Edwards

Gwenanne Edwards is a senior paper conservator at the Library of Congress. She received a Master of Arts and Certificate of Advanced Study in Art Conservation from SUNY Buffalo State College in 2012. Address: Library of Congress, 101 Independence Ave. SE, Washington, DC 20540, USA. Email: [email protected].

Adrienne Lundgren

Adrienne Lundgren is a senior photograph conservator at the Library of Congress. She received a Master of Science in Art Conservation from the Winterthur / University of Delaware in 2001. Address: Same as Gwenanne Edwards. Email: [email protected].

Marcie B. Wiggins

Marcie B. Wiggins is an assistant conservation scientist and Diana Luv Chen fellow at the Yale Institute for the Preservation of Cultural Heritage. She received her doctorate in analytical chemistry from the University of Delaware in 2019. Address: Institute for the Preservation of Cultural Heritage, 300 Heffernan Dr., Bldg 900, West Haven, CT 06477, USA. Email: [email protected].

Colette Hardman-Peavy

Colette Hardman-Peavy is an associate photograph conservator at the Virginia Museum of Fine Arts (previously paper and photograph research conservator at the Yale Institute for the Preservation of Cultural Heritage). She received her Master of Art Conservation from Queen’s University in 2018. Address: Virginia Museum of Fine Arts, 200 N Arthur Ashe Blvd., Richmond, VA 23220, USA. Email: [email protected].

Appendix 1

Table A1. Colorimetry measurements for samples aged in polyester film.

Sample	Light-developed areas										Dark-developed areas
	Control			Aged							Control			Aged
	L*	a*	b*	L*	a*	b*	ΔL*	Δa*	Δb*	ΔE*00	L*	a*	b*	L*	a*	b*	ΔL*	Δa*	Δb*	ΔE*00
224.p.a	76	0	−1	72	1	0	4	1	1	3.44	63	6	18	54	5	17	9	1	1	8.11
224.p.b	70	0	0	66	1	1	4	1	1	3.61	58	5	14	44	5	15	14	0	1	13.96
224.p.c	68	1	0	68	1	1	0	0	1	0.96	55	5	17	44	6	16	11	1	1	11.07
224.p.d	72	2	1	72	1	3	0	1	2	2.36	58	6	15	51	6	16	7	0	1	6.72
224.p.e	70	1	0	70	2	1	0	1	2	1.65	57	6	15	46	7	16	11	1	1	10.99
224.p.f	71	2	−1	69	2	1	2	0	2	2.44	57	6	17	44	6	15	13	0	2	13.03
Average							1.67	0.67	1.5	2.41							10.83	0.5	1.17	10.65
STDEV							1.97	0.52	0.55	1.02							2.56	0.55	0.41	2.79
232.p.a	73	0	2	75	1	5	2	1	3	3.27	64	4	14	52	6	15	12	2	1	11.07
232.p.b	77	1	3	72	1	5	5	0	2	4.05	64	4	13	52	5	15	12	1	2	10.98
232.p.c	81	1	4	69	1	5	12	0	1	8.80	72	4	15	51	6	16	21	2	1	18.18
232.p.d	72	1	3	67	2	6	5	1	3	4.76	62	5	15	48	7	16	14	2	1	13.43
232.p.e	71	1	3	68	2	6	3	1	3	3.60	60	6	14	49	7	16	11	1	2	10.58
232.p.f	Unavailable			75	3	6	N/A			N/A	Unavailable			50	7	16	N/A			N/A
Average							5.4	0.6	2	4.89							14	1.6	1.4	12.85
STDEV							3.91	0.55	0.89	2.26							4.06	0.55	0.55	3.19
240.p.a	79	0	3	71	1	6	8	1	3	6.49	68	4	13	46	5	13	22	1	0	20.25
240.p.b	73	0	3	67	1	6	6	1	3	5.43	59	4	12	43	5	12	16	1	0	15.99
240.p.c	70	1	4	68	2	7	2	1	3	3.08	56	5	14	40	6	14	16	1	0	15.85
240.p.d	71	2	5	69	2	7	2	0	2	2.24	56	6	15	40	6	13	16	0	2	15.87
240.p.e	70	1	5	69	3	7	1	2	2	3.03	56	6	15	40	6	13	16	0	2	15.87
240.p.f	75	3	5	70	3	7	5	0	2	4.09	57	7	13	40	6	13	17	1	0	16.90
Average							4	0.83	2.5	4.06							17.17	0.67	0.67	16.79
STDEV							2.76	0.75	0.55	1.62							2.40	0.52	1.03	1.74
270.p.a	84	0	3	80	1	5	4	1	2	3.17	78	3	13	64	5	14	14	2	1	10.92
270.p.b	83	0	3	79	1	6	4	1	3	3.93	74	4	12	54	6	16	20	2	4	16.94
270.p.c	80	0	3	76	1	5	4	1	2	3.57	69	6	18	55	7	19	14	1	1	12.03
270.p.d	80	1	4	78	2	6	2	1	2	2.46	68	5	16	55	5	17	13	0	1	11.21
270.p.e	81	1	4	77	2	6	4	1	2	3.45	66	7	14	56	6	17	10	1	3	9.07
270.p.f	80	2	3	78	2	6	2	0	3	2.89	70	6	14	59	7	17	11	1	3	9.30
Average							3.33	0.83	2.33	3.25							13.67	1.33	2.17	11.58
STDEV							1.03	0.41	0.52	0.52							3.50	0.75	1.33	2.86

STDEV = standard deviation. Samples are identified as follows: [pattern number].[p = polyester film enclosure].[sequential letter].

Table A2. Colorimetry measurements for samples aged in unbuffered paper.

Sample	Light-developed areas										Dark-developed areas
	Control			Aged							Control			Aged
	L*	a*	b*	L*	a*	b*	ΔL*	Δa*	Δb*	ΔE*00	L*	a*	b*	L*	a*	b*	ΔL*	Δa*	Δb*	ΔE*00
224.ub.a	76	0	1	74	0	−1	2	0	2	2.45	63	7	23	58	7	19	5	0	4	5.08
224.ub.b	72	0	−1	68	1	0	4	1	1	3.57	56	7	19	48	7	17	8	0	2	7.96
224.ub.c	68	1	−1	70	1	0	2	0	1	1.84	50	8	21	49	7	19	1	1	2	1.60
224.ub.d	73	1	0	72	1	0	1	0	0	0.75	51	7	20	47	8	20	4	1	0	4.16
224.ub.e	70	1	0	72	1	1	2	0	1	1.81	53	8	21	50	8	20	3	0	1	3.02
224.ub.f	72	2	−1	74	2	1	2	0	2	2.40	54	8	19	52	8	20	2	0	1	2.03
Average							2.17	0.17	1.17	2.14							3.83	0.33	1.67	3.97
STDEV							0.98	0.41	0.75	0.93							2.48	0.52	1.37	2.34
232.ub.a	80	0	3	76	0	5	4	0	2	3.30	71	5	15	64	4	14	7	1	1	5.70
232.ub.b	74	0	3	72	0	5	2	0	2	2.26	65	4	14	58	4	14	7	0	0	6.03
232.ub.c	75	1	4	75	2	7	0	1	3	2.65	60	5	14	54	5	14	6	0	0	5.51
232.ub.d	75	1	4	73	2	7	2	1	3	3.03	59	9	14	53	5	14	6	4	0	7.35
232.ub.e	73	2	4	71	3	6	2	1	2	2.47	60	6	14	52	7	14	8	1	0	7.52
232.ub.f	72	2	3	69	3	5	3	1	2	3.06	61	7	14	52	7	14	9	0	0	8.33
Average							2.17	0.67	2.33	2.79							7.17	1	0.17	6.74
STDEV							1.33	0.52	0.52	0.40							1.17	1.55	0.41	1.15
240.ub.a	79	0	4	73	1	4	6	1	0	4.57	71	4	15	63	4	16	8	0	1	6.44
240.ub.b	72	0	3	71	2	7	1	2	4	4.23	61	4	12	50	5	14	11	1	2	10.45
240.ub.c	72	1	4	73	2	8	1	1	4	3.40	58	4	15	54	5	14	4	1	1	4.03
240.ub.d	71	1	5	72	2	7	1	1	2	2.12	57	5	14	52	6	13	5	1	1	5.00
240.ub.e	72	2	5	75	2	8	3	0	3	3.26	56	6	14	52	7	15	4	1	1	4.02
240.ub.f	76	2	6	76	3	8	0	1	2	1.84	59	6	14	50	7	14	9	1	0	8.64
Average							2	1	2.5	3.24							6.83	0.83	1	6.43
STDEV							2.19	0.63	1.52	1.09							2.93	0.41	0.63	2.63
270.ub.a	84	0	3	81	1	4	3	1	1	2.61	78	5	16	69	5	15	9	0	1	6.72
270.ub.b	82	0	3	80	1	5	2	1	2	2.57	70	5	17	62	6	18	8	1	1	6.53
270.ub.c	81	0	3	78	1	4	3	1	1	2.66	67	7	19	58	6	18	9	1	1	7.73
270.ub.d	78	0	4	78	1	5	0	1	1	1.63	62	7	17	57	7	18	5	0	1	4.49
270.ub.e	79	1	4	79	2	7	0	1	3	2.65	65	7	18	58	7	17	7	0	1	6.08
270.ub.f	81	2	3	79	2	5	2	0	2	2.20	65	7	16	60	7	16	5	0	0	4.25
Average							1.67	0.83	1.67	2.39							7.17	0.33	0.83	5.97
STDEV							1.37	0.41	0.82	0.41							1.83	0.52	0.41	1.35

STDEV = standard deviation. Samples are identified as follows: [pattern number].[ub = unbuffered paper enclosure].[sequential letter].

Table A3. Colorimetry measurements for samples aged in buffered paper.

Sample	Light-developed areas										Dark-developed areas
	Control			Aged							Control			Aged
	L*	a*	b*	L*	a*	b*	ΔL*	Δa*	Δb*	ΔE*00	L*	a*	b*	L*	a*	b*	ΔL*	Δa*	Δb*	ΔE*00
224.b.a	76	0	−1	74	0	1	2	0	2	2.45	61	7	23	58	7	20	3	0	3	2.66
224.b.b	71	0	−1	71	0	0	0	0	1	0.98	57	5	14	53	5	14	4	0	0	3.79
224.b.c	71	1	0	71	1	0	0	0	0	0.00	54	5	14	53	6	14	1	1	0	1.47
224.b.d	71	1	0	71	1	0	0	0	0	0.00	54	6	14	53	7	17	1	1	3	2.13
224.b.e	71	2	−1	73	1	0	2	1	1	2.23	57	7	17	55	7	17	2	0	0	1.87
224.b.f	71	2	−1	75	2	1	4	0	2	3.53	56	6	16	56	8	20	0	2	4	2.35
Average							1.33	0.17	1	1.53							1.83	0.67	1.67	2.38
STDEV							1.63	0.41	0.89	1.44							1.47	0.82	1.86	0.80
232.b.a	80	0	3	77	0	5	3	0	2	2.71	68	5	16	64	5	17	4	0	1	3.29
232.b.b	74	0	3	73	1	6	1	1	3	3.18	63	4	14	58	4	15	5	0	1	4.41
232.b.c	72	0	4	72	2	6	0	2	2	3.16	57	4	14	56	5	15	1	1	1	1.48
232.b.d	71	1	3	72	2	6	1	1	3	2.84	57	6	14	56	6	16	1	0	2	1.46
232.b.e	72	2	4	72	3	6	0	1	2	1.95	59	6	14	59	7	16	0	1	2	1.30
232.b.f	72	2	3	70	3	5	2	1	2	2.53	59	6	14	57	7	16	2	1	2	3.23
Average							1.17	1	2.33	2.73							2.17	0.5	1.5	2.58
STDEV							1.17	0.63	0.52	0.46							1.94	0.55	0.55	1.29
240.b.a	77	0	3	73	1	6	4	1	3	4.06	65	4	13	60	4	15	5	0	2	4.39
240.b.b	73	1	4	72	2	7	1	1	3	2.76	58	4	14	56	4	15	2	0	1	1.94
240.b.c	70	1	4	67	2	7	3	1	3	3.55	56	5	14	55	5	14	1	0	0	0.94
240.b.d	69	1	4	70	2	8	1	1	4	3.41	54	5	13	53	5	14	1	0	1	1.12
240.b.e	69	2	4	70	3	7	1	1	3	2.71	54	6	12	54	6	14	0	0	2	1.50
240.b.f	74	3	5	75	3	7	1	0	2	1.78	56	6	13	57	7	15	1	1	2	1.66
Average							1.83	0.83	3	3.04							1.67	0.17	1.33	1.93
STDEV							1.33	0.41	0.63	0.80							1.75	0.41	0.82	1.26
270.b.a	83	0	3	82	1	4	1	1	1	1.78	74	4	16	71	4	16	3	0	0	2.25
270.b.b	81	0	3	82	0	5	1	0	2	1.83	71	5	14	68	5	17	3	0	3	2.88
270.b.c	79	0	3	79	0	5	0	0	2	1.69	66	6	17	65	6	18	1	0	1	1.10
270.b.d	77	0	3	78	1	5	1	1	2	2.29	62	6	17	63	7	19	1	1	2	1.54
270.b.e	78	1	4	78	2	6	0	1	2	2.02	60	7	17	63	7	17	3	0	0	2.58
270.b.f	78	2	4	79	2	5	1	0	1	1.10	64	7	15	65	7	18	1	0	3	2.06
Average							0.67	0.5	1.67	1.79							2	0.17	1.5	2.07
STDEV							0.52	0.55	0.52	0.40							1.10	0.41	1.38	0.66

STDEV = standard deviation. Samples are identified as follows: [pattern number].[b = buffered paper enclosure].[sequential letter].

Appendix 2

Table A4. Microfadeometry measurements for control samples and samples aged in buffered paper, unbuffered paper, and polyester film enclosures.

Pattern	Developed area	Location	Control (unaged)			Aged - buffered paper			Aged – unbuffered paper			Aged – polyester film
			L* a* b* start	L* a* b* end	ΔE*94	L* a* b* start	L* a* b* end	ΔE*94	L* a* b* start	L* a* b* end	ΔE*94	L* a* b* start	L* a* b* end	ΔE*94
224	light	line	75.39 −1.46 6.03	75.49 −1.41 5.92	0.3	59.69 1.94 8.23	59.20 2.06 8.26	0.5	60.80 1.31 5.53	60.18 1.44 5.65	0.7	50.30 2.26 5.84	50.16 2.32 6.01	0.2
	light	space	81.65 −0.54 −2.98	81.50 −0.52 −3.06	0.2	80.57 −0.80 −1.96	80.43 −0.78 −2.01	0.1	87.40 −0.46 −2.15	87.34 −0.47 −2.20	0.1	80.88 −1.26 −2.94	80.91 −1.27 −3.09	0.2
	dark	line	64.30 0.71 24.57	64.25 0.71 24.43	0.2	53.73 6.00 37.74	53.66 6.00 37.59	0.2	51.45 5.22 30.83	51.42 5.23 30.76	0.1	50.64 3.34 27.93	50.58 3.33 27.98	0.1
	dark	space	67.06 −1.36 3.88	66.85 −1.28 3.85	0.3	80.31 −0.93 4.17	80.27 −0.84 3.96	0.3	72.35 −0.19 16.48	72.33 −0.16 16.35	0.1	66.79 −0.01 13.06	66.78 0.03 12.91	0.2
	dark	line	59.57 1.89 13.88	59.46 1.93 13.87	0.1	63.37 0.77 10.78	63.38 0.79 10.73	0.1	45.32 1.83 10.95	45.30 1.83 10.94	0	44.74 1.45 2.97	44.87 1.42 2.96	0.2
232	light	line	64.11 1.93 13.19	63.26 2.27 13.32	0.9	66.75 2.29 11.81	66.54 2.36 11.66	0.3	54.98 2.23 9.80	54.88 2.26 9.71	0.1	65.74 1.37 11.00	63.67 2.08 11.42	2
	light	space	90.75 −1.32 2.99	90.34 −1.11 2.52	0.7	85.45 −1.74 3.66	85.39 −1.67 3.47	0.2	87.98 −1.16 5.10	87.94 −1.09 4.87	0.3	85.033 −1.27 3.73	84.97 −1.21 3.54	0.2
	dark	line	63.23 1.25 16.69	63.15 1.31 16.49	0.2	62.64 3.56 21.55	62.68 3.57 21.38	0.2	55.45 6.59 25.16	55.38 6.61 25.01	0.2	49.79 5.27 22.02	49.69 5.32 22.13	0.1
	dark	space	89.43 −1.07 10.44	89.42 −0.94 9.95	0.6	82.64 −1.41 9.98	82.51 −1.16 9.17	0.6	82.99 −1.73 10.71	82.92 −1.56 10.22	0.5	71.39 −0.77 13.72	71.36 −0.74 13.51	0.2
	dark	line	50.62 2.46 12.64	50.57 2.48 12.56	0.1	45.69 3.41 10.73	45.60 3.44 10.76	0.1	48.86 2.99 10.96	48.71 3.02 10.98	0.2	55.97 2.27 13.60	55.89 2.30 13.70	0.1
240	light	line	71.52 1.52 10.50	71.41 1.58 10.24	0.3	63.75 3.34 17.07	62.75 3.65 17.53	1.2	72.87 0.91 11.47	72.03 1.17 11.47	0.6	51.53 3.79 10.40	50.39 4.09 10.29	1.2
	light	space	93.53 −1.87 4.20	93.44 −1.75 3.71	0.5	79.40 −1.06 7.01	79.33 −1.01 6.89	0.2	85.19 −0.63 8.86	85.05 −0.57 8.62	0.3	84.18 −1.51 5.52	84.13 −1.47 5.29	0.3
	dark	line	70.27 1.74 17.11	70.22 1.81 16.86	0.3	59.18 4.32 25.11	58.96 4.45 24.81	0.4	53.54 5.58 23.70	53.32 5.64 23.85	0.3	48.95 3.99 15.90	48.92 4.00 15.82	0.1
	dark	space	80.36 −2.22 12.31	80.34 −2.10 11.95	0.4	81.27 −1.89 12.90	81.05 −1.65 12.18	0.9	68.89 −1.27 10.19	68.51 −1.04 9.83	0.5	61.34 1.61 14.14	61.34 1.62 14.07	0.1
	dark	line	63.08 2.23 15.11	63.02 2.29 14.92	0.2	62.10 5.00 23.96	61.91 5.12 23.62	0.4	55.28 3.09 14.07	54.87 3.23 14.47	0.6	40.99 2.59 6.99	40.81 2.63 7.24	0.3
270	light	dot	80.68 1.30 13.37	80.55 1.39 13.02	0.4	86.18 −0.50 7.76	85.92 −0.41 7.63	0.3	63.32 3.35 16.69	62.82 3.51 17.01	0.7	55.46 2.27 9.64	55.14 2.36 9.53	0.4
	light	space	91.93 −1.63 3.75	91.91 −1.51 3.33	0.4	93.36 −1.02 6.35	93.33 −0.95 6.01	0.4	94.13 −0.78 7.36	94.08 −0.69 6.92	0.5	87.65 −0.55 8.07	87.64 −0.50 7.71	0.4
	dark	dot	70.48 2.86 19.22	70.45 2.99 18.78	0.5	69.05 5.33 31.02	69.00 5.37 30.65	0.4	47.52 5.79 21.86	47.42 5.83 21.64	0.3	54.70 5.92 24.16	54.66 5.95 24.06	0.1
	dark	space	81.42 −1.42 8.41	81.35 −1.17 7.77	0.8	84.76 −0.17 11.52	84.71 −0.02 10.98	0.6	82.90 0.07 18.23	82.89 0.27 17.41	0.9	70.70 1.38 15.52	70.65 1.44 15.22	0.3