Enclosures and Air Pollution in Image Preservation

Transcription

1 Final Report to the Office of Preservation National Endowment for the Humanities Enclosures and Air Pollution in Image Preservation Grant # PS Edward Zinn, James M. Reilly, and Douglas W. Nishimura Image Permanence Institute November 24, 1997

2 Table of Contents Executive Summary...1 Acknowlegments...3 Introduction...4 Ozone and Objects Indoors... 5 Sulfur Dioxide and Nitrogen Oxides Indoors... 7 Enclosures and Air Pollution... 8 Test Apparatus Overview of Test Methodology Phase I: Study of Pure Oxidizing Gases at Ambient Conditions...12 Phase II: Study of Synergistic Effects...14 Phase III: Study of Enclosures...15 How Incubation Conditions Relate to Real Life Results...22 Phase I: Study of Pure Oxidizing Gases at Ambient Conditions ppm Ozone...25 Color...25 Black-and-White Photographic Materials ppm Nitrogen Dioxide...26 Color...26 Base Staining of Print Supports...27 Staining of Plastic Enclosures...28 Black-and-White Photographic Materials...28 Phase II: Study of Synergistic Effects Oxidizing/Reducing Gas Mixtures...30 Effects of 0.1 ppm NO 2 /0.05 ppm SO 2, 0.1 ppm O 3 /0.05 ppm SO 2, and 0.1 ppm NO 2 /0.05 ppm H 2 S...30 Color Materials...30 Black-and-White Silver Materials...31 Oxidant Gas Mixtures...31 Effects of 0.1 ppm O 3 /0.1 ppm NO 2 (Six Months)...31 Color Materials...31 Black-and-White Silver Materials...32 Effects of 0.1 ppm NO 2 /0.1 ppm O 3 /0.05 ppm H 2 S (6 Months)...32 Color Materials in Mixed NO 2 /O 3 /H 2 S Atmosphere...33 Phase III: Study of Enclosures Paper and Plastic Sleeves...34 Paper and Plastic Sleeve Permeability...36 Multiple Layers of Sample Material in a Microfilm Storage Box...36

3 Passive Ozone Monitor Choosing a Housing...38 The Detector...39 Detectors in Test Incubations...40 Detectors Distributed in the Office Space...40 Conclusions/Plans for Development...40 Conclusions...42 References...44 Bibliography...46

4 Executive Summary This three-year project, Enclosures and Air Pollution in Image Preservation (PS ), built upon the work completed in a previous grant funded by the National Endowment for the Humanities (NEH), Air Pollution Effects on Library Microforms (PS ). Although the earlier project outlined in broad strokes the relative sensitivity of several commonly used micrographic products to high levels of pollutant gases, it did not attempt to evaluate the potential threat posed to common imaging materials color and black-and-white photographs and films by ambient levels of pollution in libraries and archives. The earlier project also did not address mixtures of gases or protection offered by common storage enclosures. The current project was broken into three phases to address these issues. Phase I was a logical continuation of the original NEH-funded program in which pure oxidants at high concentrations were used to establish susceptibility of various photographic materials to oxidation. In this new study, pure oxidant gases O 3 (ozone) and NO 2 (nitrogen dioxide) were used in a low concentration of 0.1 parts per million (ppm) at ambient temperature (25 C) and humidity (50% RH) to quantify the effects of plausible real-life indoor concentrations of these gases. Although some of the very sensitive color print materials were severely faded by these incubations, most of the chromogenic color films and prints were not affected. Furthermore, aside from some formation of redox blemishes on the direct duplicating film and edge mirroring on the RC prints, most of the silver products resisted attack by these oxidants. While the previous project showed that filamentary silver is relatively resistant to the effects of pure ozone, nitrogen dioxide, and sulfur dioxide, there are numerous real-world examples of silver images showing image discoloration and/or fading when exposed to an environment having elevated pollutant levels. In all likelihood, such damage is the result of a combination of gases and exposure for extended periods of time. Phase II of this study investigated the interaction of mixed pollutant gases at low concentrations and long incubation periods to determine if there were synergistic effects as a result of the interaction of two or more pollutants over time. Mixing these gases did produce effects but did not result in the increased activity that was expected. On the contrary, in most of the mixed-gas experiments, a reduced amount of fading of the color samples was noted. Apparently an interaction between the individual gases resulted in a net reduction in the total gas concentrations. This project did not measure the pollutants absorbed by the samples and did not attempt to study the long-term effects that these absorbed gases might present. Phase III sought to evaluate the potential of enclosures for mitigating the effects of pollutants in storage atmospheres. Several commonly available enclosure materials in different 1

5 configurations were studied to determine the usefulness of each system in blocking the detrimental effects of gaseous attack on the photographic materials stored in them. This phase incorporated higher concentrations of pollutant gases, both pure gas and mixed atmospheres, and shorter periods of gas exposure, in order to produce results within a reasonable time. The results of this work indicate that multiple layers of packaging do impede the diffusion of pollutant gases. Also, the study showed that enclosures made of paper or board do not offer any long-term protection against high levels of pollutants in the storage environment. This agrees with work by Charles Guttman at the National Institute of Standards and Technology performed on behalf of the National Archives. 1 The atmosphere inside a cardboard enclosure will reach equilibrium with the outside air in a relatively short time. Enclosure configurations that block the diffusion of outside air (encapsulation in plastic or sealed metal cans) will protect the enclosed object from pollutants outside, but they will also trap any deterioration by-products released by the objects. This could increase the rate of autocatalytic forms of deterioration. This study also investigated the use of Kodak Ektatherm dye diffusion print material as the basis for a passive ozone monitor for libraries and archives. Samples of this material were added to all incubations in the project. A prototype passive ozone monitor was developed incorporating a disk of the thermal print material as an ozone detector. This medium proved to be a good indicator of high oxidant levels, 0.1 ppm or more, but more work needs to be done to calibrate it for the to 0.1 ppm concentrations typically found in use environments. Also, the product is no longer available; any further work will require finding a new sensitive detector. The housing that was used to isolate and mount the detector should be adaptable to other materials. Archivists and librarians can use the data gathered in this project to interpret air quality measurements made in their photographic storage areas and to guide engineering professionals in the design of new storage facilities. Black-and-white images require very low levels of airborne contaminants, but the data show that color photographs are more tolerant of longterm exposure to low levels of airborne oxidants. While multiple levels of enclosures (plastic sleeves, folders, boxes, and closed cabinets) can together mitigate somewhat the attack by pollutants, archivists should not rely on enclosures for protection against pollutant effects. Gas phase filtration equipment still has an important role to play in photograph, cinema, and microfilm storage. 2

6 Acknowlegments This research project was sponsored by the Division of Preservation and Access, National Endowment for the Humanities (NEH), by the New York State Program for the Conservation and Preservation of Library Research Materials, and by Rochester Institute of Technology. Direct duplicating film and diazo microfiche were provided by University Microfilms, microfilm boxes by Gaylord Brothers, RC print material and processing chemistry by Ilford Photo Corporation, and color print material by Eastman Kodak Company. 3

7 Introduction The grant project Enclosures and Air Pollution in Image Preservation was a three-year research study that investigated the threat posed by the ingress of polluted air into library and archival storage environments. This work adds to the knowledge gained in an earlier study, Air Pollution Effects on Library Microforms, 2 which dealt with the effects of four common air pollutants (NO 2, O 3, SO 2, and H 2 S) on microfilm and other types of photographic materials. Whereas the original study evaluated the gross sensitivity of photographic materials to pollutant gases, one of the prime focuses in the current work was to explore in depth those sensitivities at test conditions more in line with real life. In the complex situation of a real archives or library, the contribution that air pollution makes to the fading of images is not easy to measure or observe. The effects are subtle and occur slowly, blending in with the action of light, heat, humidity, and the inherent stability (or instability) of each individual type of photograph. We are not able to say what percentage of the fading of a color print is due to air pollution. We know it causes some fading, but the thermal fading characteristics of color materials are probably the major factors. Because of the chemical nature of images (including in that term products commonly used in still photography, cinema, microforms, and color hard copy output from computers), it is a reasonable expectation that airborne pollutants would act to destroy them. Color images are usually comprised of organic dyes, which depend on complex structural arrangements within the dye molecule to absorb light. Disturb that precise structure at the right spot, and a yellow dye becomes colorless, or a cyan dye becomes a weak yellow. Black-and-white photographic images are comprised of metallic silver, which must first become oxidized before it can degrade; but, with silver, the response to atmospheric oxidants is complex and depends on more than just the presence of an oxidant. 4

8 Ozone and Objects Indoors In recent years, considerable investigation has been done on the effects of atmospheric ozone on modern and traditional artists pigments. 3,4,5,6,7 Ozone is a concern because it is such a strong oxidizing agent, especially reactive toward the double bonds in organic dyes, which often are part of the key chromophoric (color-generating) groups on dye molecules. The research (much of which was done at an O 3 concentration of 0.40 ppm, the approximate peak outdoor ozone levels in a polluted area) found that colorants vary in susceptibility, and that some are quite sensitive. A very important determinant of susceptibility to attack was not only the dye structure itself but also whether or not the dye was protected by a binder material, for example, gum arabic or linseed oil. Other related research (much of which resulted from a broad initiative on environmental effects conducted by the Getty Conservation Institute) attempted to both directly measure and predict from a model the ozone concentrations that would actually be encountered inside southern California museums. 8,9,10 These studies and others outside the conservation field 11 have found that indoor ozone concentrations can be as much as 70% of outdoor concentrations and that relatively simple models can predict indoor ozone concentrations. Leaving aside for the moment any air purification system or internal sources of ozone such as copiers, the amount of ozone in indoor air will be determined by: (1) outside levels, (2) the rate at which outdoor air is admitted to the indoor environment, and (3) the area and nature of surfaces indoors. Ozone is decomposed by reaction with any surface (floors, walls, etc.) so the ratio of indoor surfaces to the volume of air circulating through the space becomes important. Items freely exposed to indoor air in a building could accumulate very large ozone doses over time. Most images, however, are encased in boxes or other forms of storage enclosure. Based on this model of the behavior of a building, there are two reasons why ozone damage would be mitigated inside a storage box. First, the air exchange rate between the inside and the outside of the box is normally low, so the total accumulated ozone exposure inside would be less than outside. Second, there is also the chance that ozone that does enter the box would react with the walls and not the contents. Ozone as an outdoor air pollutant is most effectively produced when abundant sunlight acts on nitrogen oxides and hydrocarbons. The latter two are characteristic of areas with lots of cars in use. Though ozone first came to prominence in southern California, many other urban and nonurban areas have elevated ozone levels. 12 Most locations east of the Mississippi are subject to peak ozone concentrations of 0.04 ppm to 0.12 ppm. 13 Local concentrations (primarily in or near cities) can be higher. Ozone concentrations in cities rise and fall each day 5

9 because they are generated largely by the action of sunlight, but Chicago, Washington, and New York at peak levels have about two-thirds of the 0.33 ppm found in Los Angeles. 14 In all this work, the main points are that ozone can exist indoors in concentrations high enough to cause some colorants to fade and that the effects of ozone are cumulative over time and serious enough to cause damage with only a few months exposure. Moreover, the effects are directly related to the product of concentration times duration of exposure. 10 In other words, two weeks at 0.40 ppm ozone gives the same result as four weeks at 0.20 ppm or six days at 1 ppm. However, the basis for this conclusion was work done on unprotected colorants (colorants without a binder). Binders complicate this simple concentration times time formula and make extrapolations from high concentration difficult. 6

10 Sulfur Dioxide and Nitrogen Oxides Indoors Many of the studies of these acid gases have been devoted to the effects of acid rain on outdoor stone and metal objects. 15,16,17,18,19 However, levels of both of these oxides are relatively high in museums, raising concern about books, paper, and works of fine art. Hackney made measurements of sulfur dioxide at the National Gallery of Ireland, the Victoria and Albert Museum, and the Tate Gallery. 20 Nitrogen dioxide levels were also measured at the Tate Gallery. Hackney found significant levels of nitrogen dioxide in unconditioned galleries, which correlated well with the sulfur dioxide measurements. Levels of SO 2 were also significant in all three museums and varied irregularly with weather conditions. Levels fell rapidly after a rain, although winter weather brought levels up. Grosjean and Hisham similarly found that indoor levels of sulfur dioxide and nitrogen dioxide in southern California museums were a significant proportion of the outdoor levels. 21,22 Limited studies have been performed on the effects of air pollutants on cellulose, cotton, and paper, but these experiments have concentrated on the acid gases, sulfur dioxide and nitrogen dioxide. 23,24,25,26 Of particular significance to accelerated aging research have been studies by Havermans on the influences of sulfur dioxide and nitrogen oxides on accelerated aging of paper. 27 Havermans showed that in the presence of sulfur and nitrogen oxides, acid hydrolysis of the paper was the dominant deterioration mechanism. Buffered paper tended to perform much better than nonbuffered paper although measureable amounts of nitrates and sulfates were found only in buffered paper. 7

11 Enclosures and Air Pollution Enclosures are a very important aspect of storing photographs. Collection managers cannot change the nature of photographic materials to make them fade and stain more slowly or to be more physically robust. Nevertheless, among the things that collection managers can control are handling practices, environmental conditions, and the choice of storage enclosures. Although temperature and RH are the primary factors in the survival of images, enclosures and handling practices are next in importance. Whether for still photography or motion pictures, enclosures add to the useful life of collections by providing physical protection, by buffering rapid environmental changes, and by being inert, harmless containers during years of storage. There is no one ideal enclosure design for a particular type of photograph; the handling patterns, storage environment, use environment, and economic circumstances of every collection are different. Enclosures are always part of a multilevel system where each object has an individual enclosure, which in turn is matched to a box or folder, which finally is placed in a third level consisting of shelving or cabinets. Ease and safety in handling and the ability to integrate with the other levels should be designed into the overall system from the beginning. Photographs are found in many different sizes and types prints, negatives, transparencies, roll films, and sheet films and all are physically delicate. Whatever style of design and construction is used, the most important thing is that the enclosure be able to physically perform the tasks desired of it. For example, motion picture film cans need to be strong enough to withstand the weight of other cans. Transparencies might need clear enclosures so that they can be examined without being directly handled. One other aspect of enclosures is significant: their chemical inertness toward photographs stored inside them. Enclosures should not emit harmful gases or leach out substances that would cause dye fading, staining of gelatin, or other problems. The glues, inks, labels, and coatings used in enclosure manufacture should all be inert and nonreactive with photographs. There are many examples in real-life collections of damage from inappropriate enclosures and adhesives; most people are familiar with the stains and residues left by older cellophane tape, decomposing rubber bands, rusty paper clips, etc. The inertness of photographic storage enclosures can be evaluated by use of the Photographic Activity Test (PAT), an accelerated-aging method described in ANSI Standard IT , American National Standard for Imaging Media Photographic Activity Test. Passage of the PAT is not a guarantee that a storage enclosure will be satisfactory in all aspects, merely that it will not cause staining or fading of photographs because of substances emitted by the enclosure. The protective benefits of enclosures with respect to air pollutants and photographic materials (or other types of archival materials) have not been directly studied using full enclosures 8

12 with actual materials inside. In recent years, the National Institute of Standards and Technology (NIST) has undertaken research (sponsored by the National Archives) to examine the issues relating to the microclimate inside a typical acid-free document box. The initial research, conducted by Elio Passaglia, 28 sought to develop a general model for containers that would predict various aspects of behavior, including how well the document box would insulate its contents from pollutant gases. Using available information, the model suggested that transport of gases through the box walls was slow and that equilibration would occur principally through the gaps in the box and would require a few months. 28 More recently, new work has been conducted by Dr. Charles Guttman of NIST to actually measure the rate at which SO 2 diffuses through the walls of a typical acid-free document box. He has found that diffusion through the box wall is substantially faster than earlier estimates, and now Passaglia s model predicts that inside/outside SO 2 equilibration is complete within a few days. 1 As others have found, the NIST work also shows that the acid-free box wall absorbs SO 2 (about 0.2% by weight). NIST s preliminary data with NO suggests very little absorption, but NO diffuses through the wall at about the same rate as SO 2. Other work by NIST measured the NO 2 diffusion and absorption rates with conventional acid-free and buffered lignin-free box boards. The NIST research was aimed at refining the Passaglia model by providing more accurate diffusion rate constants. It did not, however, attempt to measure how much pollutant enters a box under real-life conditions. The IPI research reported here did involve direct exposures of complete boxes, sleeves, and envelopes and contents to ambient levels of pollutants. It did not quantify how much of each gas made it through the enclosure, but the resulting effects on the imaging materials stored inside them were quantified. 9

13 Figure 1: Layout of IPI Pollution Studies Lab. (A) Computer terminal. (B) Pollution chamber. (C) Gas monitoring equipment. (D) Cutaway view of free-hanging film samples within pollution chamber. (E) Gas canisters. (F) Sample preparation area. Test Apparatus Figure 1 is a graphic view of the air pollution testing facility at IPI. This equipment was acquired with the aid of NEH under a grant proposal (Air Pollution Effects on Library Microforms, PS ) in The specially designed incubation chambers are one of a kind and are designed to withstand the rigors of constant exposure to varying levels of corrosive gases and extremes of temperature and humidity. These chambers have a large capacity (21 cubic feet) and are designed to maintain constant temperature, Figure 2: Interior view of incubation chamber. humidity, and gas concentration while providing recirculation and adequate mixing of the test atmosphere. Figure 2 shows the inside of one of the incubation chambers. All wetted components of the chamber are constructed of Teflon, polyvinylidene fluoride (PVDF), or stainless steel. The chamber is depicted without the removable rear panel to show the circulation fans, heating coil, gas and humidity inlets, sample outlet, and cooling and dehumidification coils. Test conditions are maintained from a desktop PC that is networked with two ANAFAZE 8 10

14 Figure 3: Schematic drawing of ozone injection system used in project. Figure 4: Exploded view of ozone generator. loop PID controllers that allow operator input and monitoring of PID function. The test gases are injected into the chambers from large gas cylinders. NO 2, SO 2, and H 2 S are supplied on demand via the controllers and dedicated Teflon solenoid valves for each gas. The ozone injection system is outlined in Figure 3. Ozone can be regulated by air volume and variable output from the PID controller. Ozone is generated on demand using a high-voltage discharge ozone generator (Figures 3 and 4). Monitoring of the chamber atmospheres is accomplished by withdrawing a sample from the chamber outlet and analyzing it with the appropriate gas analyzer. Monitoring equipment (all manufactured by Monitor Labs) includes the following: Ozone Nitrogen dioxide Sulfur dioxide Hydrogen sulfide Model 8810 U.V. Photometric Ozone Analyzer Model 8840 Chemiluminescent Nitrogen Oxides Analyzer Model 8850 Fluorescent SO 2 Analyzer Model 8850 Fluorescent SO 2 Analyzer with Model 8770 H 2 S Converter The monitoring equipment is mounted in an instrumentation rack placed between the two chambers and is calibrated by a Monitor Labs 8500 calibrator. All of the equipment is interfaced with the controllers and subsequently with the computer. 11

15 Overview of Test Methodology This study was divided into three individual phases. In many cases, the phases overlapped each other and materials used in one phase were also included in one or both of the other phases. Additional enclosure configurations were added to some tests in order to explore a particular hypothesis or interest, these will be discussed in the text of individual sections. The work on the passive ozone monitor encompassed the entire project; these monitors were added to all incubations. Monitors were also placed in various positions around IPI s lab, for instance, over the photocopier, over the laser printer, in the collection storage area, etc., in an attempt to look at their behavior in as many microenvironments as possible. The specific differences in each phase of the project are discussed in the overview that follows. The experimental plan was divided into the following three phases: Phase I: Study of Pure Oxidizing Gases at Ambient Conditions Phase II: Study of Synergistic Effects Phase III: Study of Enclosures Development of Passive Ozone Monitor (all test conditions) Phase I: Study of Pure Oxidizing Gases at Ambient Conditions In this phase, pure oxidants were run at low concentrations for an extended time period. Only the pollutants ozone and nitrogen dioxide were studied at the 0.1 ppm (100 parts per billion) concentration, because the previous study showed that these are the two more aggressive oxidant gases. The tests were run for one year at 25 C, 50% RH. To separate the effects of the pollutants from normal aging, all samples were also incubated for the same time period in a humidity-controlled chamber at 25 C, 50% RH without any pollutant gases present. Specimens were removed from the pollution chamber every four months, measured for density, and then replaced in the chamber. Samples were suspended in the chamber and freely exposed to the chamber atmosphere. This approach simplified the specimen preparation, requiring only two specimens of each material type per chamber. A wide and broadly representative variety of photographic materials were tested. This maximized the information that could be obtained and was highly desirable in view of the long incubation times required and limited time blocks available. 1. Negative motion picture film (Kodak 5272). This product showed significant ozone-induced color fading in the original study, particularly of the yellow dye. 2. Positive motion picture film (Kodak 5384). This proved to be the most ozonesensitive chromogenic film that was evaluated in the first study. 3. Conventional chromogenic color print (Ektacolor). This is a very important mate- 12

16 rial of great interest to archives, and it has a fairly high ozone sensitivity. 4. Diazo microfilm (Xidex). This product is an important micrographic film for user and service copies. 5. Source document microfilm (Fuji HR2). Although silver images did not prove to be as sensitive as color images to the effects of pure oxidizing agents, they can be prone to such defects as silver mirroring in the presence of some combinations of pollutant gases. This is typical of films used as master negatives in preservation microfilming. 6. Source document microfilm, polysulfide treated (Fuji HR2). Polysulfide treatment protects the silver image against oxidants by converting the silver to silver sulfide, which is more peroxide-resistant. Verification of the resistance of sulfided silver to prolonged exposure to low levels of ozone and nitrogen dioxide and to pollutant combinations was the purpose for including this material. 7. Direct duplicating microfilm (Kodak 2468). This film is widely used in the microfilm industry. In the original study it showed slightly greater changes than the silver camera microfilm. 8. Colloidal silver film (detector used in the ANSI Photographic Activity Test). Although this is not a commercial or user product, colloidal silver has proven to be much more sensitive to the atmospheric environment and pollutants than filamentary silver images. It is for this reason that colloidal silver is used in the Photographic Activity Test to evaluate enclosure materials. In the first pollutant program at IPI, colloidal silver was much more sensitive to the effects of ozone and H 2 S than silver camera microfilm. 9. Ilfochrome color micrographic film (Cibachrome CMM). Although this product is not as widely used as the chromogenic color materials, the dyes are inherently more stable. It is being used in a number of color preservation microfilming applications. 10. Resin-coated black-and-white paper print (Ilford Polycontrast). There are many practical examples of these prints showing silver mirroring, fading, or microblemishes when exposed to pollutant atmospheres. Their inclusion in this study should provide the archivist with practical information about this important imaging material. 11. Fiber-base black-and-white paper print (Kodak Polyfiber). This is similar to the previous product, but its paper base can more readily absorb atmospheric pollutants. 12. Thermal dye diffusion print (Kodak Ektatherm). Although this color hard copy material and similar products from Sony, Tektronix, and Mitsubishi is not yet widely stored, it likely will be in the future. Consequently, a comparison of the relative stability of some of these newer color hard copies with that of the traditional chromogenic 13

17 photographic materials is of considerable importance. This comparison is believed to be more critical in light of results from the first pollution study. The dyes in this material are not protected by a gelatin binder and proved to be very susceptible to ozone fading. IPI believes also that this material is so extraordinarily sensitive to ozone that it may be the basis for a practical long-term ozone monitor for libraries and archives. 13. Electrophotographic color print (Canon Color Copier). This product will also become increasingly important to the archivist. Studies to date have shown it to be much more stable in the presence of pollutants than the thermal dye diffusion transfer material. 14. Color inkjet print (Hewlett Packard Paint Jet). This product is also finding application as a printout copy. No information is available on its stability to pollutants. All black-and-white samples had a step wedge of seven density blocks ranging from Dmin to Dmax, a range of approximately 0.05 to 2.0 blue density, including 1.0 above Dmin. Color materials had three sets of color patches (cyan, magenta, and yellow dyes), each color having four density steps. Status A blue densities were measured on the seven black-and-white (or neutral tone) materials, while blue, green, and red densities were measured on the seven color images. Density change and visual inspection for silver mirroring or redox blemish formation were the only criteria for this phase. Phase II: Study of Synergistic Effects The original pollutant program showed that filamentary silver is relatively resistant to the effects of pure ozone, nitrogen dioxide, and sulfur dioxide. However, there are numerous practical examples of silver images showing mirroring and/or fading when exposed to an environment having elevated pollutant levels. It is now believed that this is the result of the interaction of two or more pollutants. It was considered important to determine if this synergistic effect takes place under more practical conditions, particularly in the case of NO 2 and H 2 S. Four pollutant gases were used in this phase in the combinations listed in the table below. Interactions between the gases were expected, and the increase or mitigation of attack by one gas in the presence of other gases was the prime concern. Ozone is a strong oxidant, and its oxidizing effect is greater in the presence of acid. Nitrogen dioxide is also a strong oxidizing agent; it is strongly acidic as well. Sulfur dioxide is a weak reducing agent and produces acid in the presence of moisture. Hydrogen sulfide is also a weak reducing agent but is a strong sulfiding gas. The effects of oxidant combinations were expected to be seen primarily with color images, while oxidants and sulfur gases were expected to affect primarily silver images. 14

18 Five gaseous combinations were used in Phase II and are summarized below: Time Period Pollutant Concentration 1 year 0.1 ppm NO 2 and 0.05 ppm H 2 S 6 months 0.1 ppm O 3 and 0.1 ppm NO 2 6 months 0.1 ppm O 3 and 0.05 ppm SO 2 6 months 0.1 ppm NO 2 and 0.05 ppm SO 2 6 months 0.1 ppm NO 2, 0.1 ppm O 3, and 0.05 ppm H 2 S The 0.1 ppm concentrations of ozone and nitrogen dioxide are the same as in Phase I. The 0.05 ppm of hydrogen sulfide and sulfur dioxide are also plausible indoor pollutant levels for an urban archives. All concentrations are at least a decade higher than the lower detectability level. Two specimens each of the same fourteen materials studied in Phase I were studied in Phase II, using the same density measurements. Densities were determined at intervals of two or four months, resulting in three sets of determinations for the six-month studies and three sets for the one-year study. After each measurement the specimens were returned to the chamber for further incubation. All specimens were suspended in the chamber by stainless steel rods, as shown in Figure 5, so that they were freely exposed to the pollutants at 25 C, 50% RH. Phase III: Study of Enclosures Under practical storage conditions, control of pollutants is frequently impossible because of the expense of installing and operating the necessary equipment. However, photographic materials are usually placed in enclosures such as sleeves or boxes. It has been shown that these enclosures will mitigate the effect of pollutants. The unanswered questions were by how much, and what types of enclosures are to be preferred with each type of photographic material. These practical considerations were the primary focus of Phase III. Four photographic materials were chosen for this study: 1. Thermal dye diffusion print (Kodak Ektatherm). This material was selected because of its very high susceptibility to oxidant pollutants. Figure 5: Diagram of sample-hanging arrangement within pollution chamber. 2. Positive motion picture film (Kodak 5384). This is a much more common film which also has high susceptibility to oxidant gases, although not as high as the thermal 15

19 dye diffusion print material. 3. Direct duplicating microfilm (Kodak 2468). This film is of wide interest in the microfilm industry where long-term storage is important. 4. Resin-coated black-and-white paper print. This is another example of a widely stored silver image. The commonly available enclosures studied were: 1. Buffered paper envelope (Light Impressions Straight Cut Envelopes made from 80-lb. Apollo Paper TM ). 2. Nonbuffered paper envelope (Light Impressions Straight Cut Envelopes made from 80-lb. Renaissance Paper TM ). 3. Polyethylene sleeve (Print File 35-1M Polyethylene Sleeve). 4. Polypropylene sleeve (Light Impressions 3-mil FoldLock TM Sleeve). 5. Polyester sleeve (Light Impressions 3-mil Mylar D sleeve). 6. Metal-edge box made from lignin-free buffered cardboard (Conservation Resources Lig-free Type I 35mm Archival Microfilm Boxes). 7. Polyester sleeve inside a metal-edge box (Light Impressions 3-mil Mylar D sleeve). 8. Buffered paper envelope inside a metal-edge box (Light Impressions Straight Cut Envelopes made from 80-lb. Apollo Paper TM ). One sample was inserted into each envelope or sleeve. Three replicates of each material configuration were prepared, one sample for each measurement interval. The first configuration in this phase was designed to mimic sleeved negatives or prints that were placed vertically in a drawer. In real life, photographs in this type of configuration can be placed loosely in the drawer or be packed tightly together. In this test, in order to limit the number of samples required, a normal filing cabinet drawer was not used. Samples were instead prepared to fit into a collapsible 3½ computer floppy disk storage box. This box allowed the three samples of each material to be placed in a separate compartment in the box. Placing three samples in each compartment caused a slight pressure to be placed on the samples and restricted air flow into the package. This configuration is seen in Figure 6. Sleeved prints and negatives are also stored horizontally in boxes where the box itself could act as an impediment to attack from pollutants in the atmosphere. The second configuration consisted of one sleeved sample placed in a roll microfilm storage box Figure 6: Collapsible floppy disk storage box simulated conditions in a packed filing cabinet drawer by placing slight pressure on and restricting air flow around samples. 16

20 Figure 7: Single sleeved sample in roll microfilm storage box. Figure 8: Arrangement of boxed samples in incubation chamber. Spacing boxes a half-inch apart on raised rack permitted free flow of air around samples. Figure 9: Stack of five sample sheets without sleeves in roll microfilm storage box. (Figure 7). These boxes are not well-sealed. Therefore, it was presumed that they offered little protection from ingress of the outside atmosphere, but the extent or lack of benefit was unknown. Three individual boxes were required for each sleeve/sample material combination. The boxed samples were placed on a perforated stainless steel rack at the bottom of the chamber (Figure 8). The rack raised the boxes approximately one inch off the bottom of the chamber and the perforations allowed air flow around the samples. A half-inch space was left on all sides of each box to separate it from its neighbor and allow free air movement. Another box configuration included in this phase was intended to explore the commonly seen stack of unsleeved negatives or prints in a box. In this configuration, five separate samples of each of the four materials were stacked on top of each other and placed in the same roll microfilm storage box as above (Figure 9). In a normal storage situation there would be many more objects stacked together typically whatever the box would hold. This represented a problem, in that the more objects there are in the box, the more weight and subsequent pressure there is placed on objects in the stack. Creating large stacks of samples in each box for testing was impractical, so this configuration included only five samples, and only the top, middle, and bottom samples were measured. It was felt that, while the heavy stack of materials found in real life could not be recreated easily, at least the sample-to-sample contact, characteristic of a stack, could be observed. Three boxes of each sample were prepared for each test. In conjunction with Phase III, the sleeve permeability test was devised to evaluate a sleeve s diffusion resistance or pollutant-trapping capacity (see Figure 10). In this test, 2 x 2 pieces of the thermal print material and the Kodak 2468 direct duplicating film were arranged on a 24 x 30 sheet of clear polycarbonate. Each sample was then covered by a 2½ x 2½ piece 17

21 Figure 10: Sleeve permeability test. of one of the five sleeve materials. There were three separate samples of each sleeve material for both the thermal print and the 2468 film. The sleeve materials were affixed to the sheet with aluminized polyester tape. The samples were thus completely covered and sealed from the environment. The pollutant gases could affect the samples only by diffusing through the plastic or paper sleeve material. All samples in Phase III were made as one uniform density patch of approximately 1.0 density. The single field of 1.0 density was used to track edge effects, as the gases were expected to penetrate and cause fading only at the edges of the samples inside the sleeves. Color patches of only cyan dye, with a density of 1.0 above Dmin, were used for the two color materials. The cyan dye had previously been shown to be very sensitive to oxidants for both the movie-print film and the thermal print material. A single patch with a blue density of 1.0 above Dmin was used for the two black-and-white samples. Measurements were taken at five different positions on each sample before incubation. A template of clear polyester was used to ensure measurement of the same areas on each sample. Figure 11 shows the areas that were measured. Since removing the specimens from the enclosure would disturb the gaseous distribution, it was necessary to use separate specimens for each time period. Single specimens were prepared and measured for each time period. Figure 11: Samples were measured in the five areas indicated. The duration of each test was eight weeks, and there were three sample time periods: two and one half weeks, five weeks, and eight weeks. Since the primary purpose of this phase was to determine the relative degree of improvement afforded by the various enclosures, tests were conducted at 5 ppm of the various pollutants at 25 C, 50% RH. It was recognized that these conditions were more severe than those found in practice, but longer incubation times at lower concentrations were not possible within the time allowed. It was felt, based on previous 18

22 studies, that this concentration would give meaningful changes in shorter time periods and allow comparisons between enclosures. Time Period 8 weeks 5 ppm O 3 8 weeks 5 ppm NO 2 Pollutant Concentration 8 weeks 5 ppm O 3 and 5 ppm NO 2 8 weeks 5 ppm O 3 and 5 ppm H 2 S 8 weeks 5 ppm NO 2 and 5 ppm H 2 S 8 weeks 5 ppm O 3, 5 ppm NO 2, and 5 ppm H 2 S No t e: To obtain a more practical evaluation of the enclosure benefits at lower pollutant concentration levels, the same eight enclosures were also placed in the chambers for the Phase I and II studies. Since the size of the chambers limited the number of metal-edge boxes to 36, time periods were limited to three for each of the seven atmospheres. The forty-two separate enclosure configurations are listed on the following page. 19

23 Polyethylene sleeve/thermal print Polyethylene sleeve/direct dupe film GROUP A Sleeve Permeability Thermal print/direct dupe sealed under sleeve materials Polyester sleeve/thermal print Polyester sleeve/direct dupe film Polypropylene sleeve/thermal print Polypropylene sleeve/direct dupe film Buffered paper envelope/thermal print Buffered paper envelope/direct dupe film Nonbuffered paper envelope/thermal print Nonbuffered paper envelope/direct dupe film GROUP B Sensitive materials in buffered, lignin-free, metal-edge boxes Box/thermal print, stack of 5 Box/color positive film, stack of 5 Box/b&w RC print, stack of 5 Box/direct dupe film, stack of 5 Box/thermal print in buffered paper envelope Box/color positive film in buffered paper envelope Box/b&w RC print in buffered paper envelope Box/direct dupe film in buffered paper envelope Box/thermal print in polyester sleeve Box/color positive film in polyester sleeve Box/b&w RC print in polyester sleeve Box/direct dupe film in polyester sleeve Polyethylene sleeve/thermal print Polyethylene sleeve/color positive fill Polyethylene sleeve/b&w RC print Polyethylene sleeve/direct dupe film GROUP C Sensitive materials in sleeves and envelopes Polyester sleeve/thermal print Polyester sleeve/color positive film Polyester sleeve/b&w RC print Polyester sleeve/direct dupe film Polypropylene sleeve/thermal print Polypropylene sleeve/color positive film Polypropylene sleeve/b&w RC print Polypropylene sleeve/direct dupe film Buffered paper envelope/thermal print Buffered paper envelope/color positive film Buffered paper envelope/b&w RC print Buffered paper envelope/direct dupe film Nonbuffered paper envelope/thermal print Nonbuffered paper envelope/color positive film Nonbuffered paper envelope/b&w RC print Nonbuffered paper envelope/direct dupe film 20

24 How Incubation Conditions Relate to Real Life The test conditions used in this study were planned with different goals in mind. The shortterm, high-concentration tests were meant to acquire as much data as possible in a short-term experiment. Relying on data from previous IPI research, we were confident that measurable density change was possible at the 5-ppm gas concentration in the proposed time frame. In the long-term tests, the purpose was to simulate as closely as possible real-life exposures. Tables 1 and 2 show a general relationship between real-life concentrations of pollutant gases and the project test conditions. Table 1 compares the two conditions used in the current study and shows how long it would take at real-life conditions (assuming an average level of ppm) to attain the same exposure received in both of the test conditions. The.025 ppm gas concentration represents a typical average concentration that might be found in a storage space that has no filtering or HVAC control. The comparison is hypothetical and based on ozone only; it assumes a linearity of concentration versus time and does not consider the variability of the test materials. According to Table 1, the short-term tests in this project would be equal to approximately 33 years at room conditions. The long-term tests would require about four years at ambient indoor conditions to reach the test exposure levels. Table 1: Relationships between test atmospheres and exposure times. Ozone Concentration Elapsed Time (days) Total Exposure (ppm hours) Equivalent Exposure at.025 ppm (years) 0.05 ppm ppm* ppm ppm ppm ppm* ppm * Concentrations used in current study Table 2: Comparison of the two test atmospheres in the current study. Ozone Concentration 0.10 ppm 5 ppm Elapsed Time (days) 365 (3000 days would be needed to equal 5 ppm exposure) 60 (7.3 days would be equal to 365 days at 0.10 ppm) Total Exposure (ppm hours) Equivalent Exposure at ppm (years)

25 Results Table 3, on the following page, shows an overview of the sensitivity of each sample material used in the project to each of the incubation conditions that they were exposed to. Only four sample materials were used in the short-term experiments, so data is seen for only those four materials in incubations H through M. The legend for the table lists the test atmosphere and duration for each incubation and provides a key for the various levels of attack and any other manifestations of the exposure. The table shows that the filamentary silver film materials were only slightly affected by these gaseous atmospheres. In the real world, examples of filamentary silver fading are everywhere, but this project s use of plausible real-life storage atmospheres failed to produce much effect in the time allowed. The mixed-gas incubations sought to mimic the oxidation/reduction cycle that is known to be the mechanism of silver image fading, but they did not reproduce real-life fading. The one variable that was not equivalent to real life was time. As noted above, the low-concentration incubations would only approximate four or five years exposure to real-life conditions. Nevertheless, silver mirroring and redox blemish formation in only a year s time at room temperatures and ambient gas concentrations have been observed in the IPI lab. 22

26 Table 3: Relative sensitivity of the sample materials to test conditions. A B C D E F G H I J K L M N Material BS/CBS 0 0BS/CBS 1+BS/CBS 0BS/CBS 0 0 Ektacolor 0BS 2-0BS 0BS 0 0BS 0BS 0 Xidex 2-BS 2-1-BS 2-BS 2-2-BS 2-BS 0 HR HR2 Toned RB 0 0 RB RB Colloidal silver CMM B&W RC 0BS/M 0 0BS 0BS/M 0 0BS 0BS 0M/BS M/BS 0BS 0BS 0BS 0 0 B&W Fiberbased 0BS 0 0BS 0BS 0 0BS 0BS 0 Ektatherm 4-BS/CBS 4-3-BS/CBS 4-BS/CBS 4-2-BS/CBS 2-BS/CBS 4-BS/CBS 4-4-BS/CBS 4-BS/CBS 4-BS/CBS 3-BS/CBS 0 Electrophotographic Ink jet 1-BS 3-2-BS 2-BS 3-1-BS 1-BS 0 Legend for Gas Atmospheres Gas Concentration Test Duration Level of Attack A 0.1 ppm NO 2 1 year 0 = No density loss (-) or gain (+) B 0.1 ppm O 3 1 year 1 = Slight density loss (-) or gain (+) C 0.1 ppm NO 2 and 0.05 ppm H 2 S 1 year 2 = Moderate density loss (-) or gain (+) D 0.1 ppm O 3 and 0.1 ppm NO 2 6 months 3 = Severe density loss (-) or gain (+) E 0.1 ppm O 3 and 0.05 ppm SO 2 6 months 4 = Very severe density loss (-) or gain (+) F 0.1 ppm NO 2 and 0.05 ppm SO 2 6 months G 0.1 ppm NO 2, 0.1 ppm O 3, and 0.05 ppm H 2 S 6 months H 5 ppm NO 2 8 weeks I 5 ppm O 3 8 weeks RB = Redox blemishes J 5 ppm NO 2 and 5 ppm O 3 8 weeks M = Mirroring K 5 ppm NO 2 and 5 ppm H 2 S 8 weeks DY = Dmin yellowing L 5 ppm NO 2, 5 ppm O 3, and 5 ppm H 2 S 8 weeks BS = Base stain M 5 ppm O 3 and 5 ppm H 2 S 8 weeks CBS = Color balance shift N Control (no gas) 1 year 23

27 Phase I: Study of Pure Oxidizing Gases at Ambient Conditions These two one-year incubations were designed to explore the long-term effects of the oxidant gases nitrogen dioxide and ozone on the photographic materials in the study. The gases were used individually in separate chambers at the 0.1 ppm concentration level at 25 C and 50% RH for a period of one year. These incubations represent a total pollutant exposure approximately equal to four years in a normal collection environment, assuming a yearly average of ppm of either of the gases. The ventilation air that comes into the collection environment may have passed through a multiple-bed chemical filtration system that removes a high percentage of some pollutant gases, or else it may have come in through open windows, causing the indoor air to closely track the outdoor pollutant levels. The actual level of oxidizing contaminants in collection environments can vary greatly, but ppm is a reasonable figure to represent real-life average concentrations for O 3 and NO 2. The currently recommended levels of pollutant gases for museums are very low (<0.001 ppm). 12 At these levels, our test atmospheres represent a 100-year exposure, but the reality is that these levels are lower than can be accurately measured by most instrumentation. In other words, it is not possible to measure such low concentrations, and the equipment needed to even come close is very expensive and not practical for most institutions. Given this fact, the purpose of this study was to look at how dangerous the ambient real-life levels of oxidizing pollutants are and to determine the urgency of control measures for photographic storage areas. The fourteen different free-hanging photographic materials used in this test series are listed in Introduction, Phase I: Study of Pure Oxidizing Gases. ( Free-hanging means that materials were not in enclosures; they hung freely in the chamber.) Descriptions of their importance and of the sample preparation and measurement are found there. The four sensitive materials used in Phase III (see Phase III: Study of Enclosures), in their respective enclosure configurations, were also included in this phase. According to the data from this study, most of the common photographic materials tested would not show any effect from ambient levels of oxidizing pollutants for many decades. The exceptions were those color output materials that deposit their imaging material directly on the support surface with no binder layer or surface lamination to protect it. These exceptions included thermal diffusion print material and ink-jet prints. Also included in the materials sensitive to oxidation were the Xidex diazo microfiche and the colloidal silver indicator strip. Both of these materials were also badly oxidized in previous IPI research using much higher concentrations of gas and higher temperatures and humidities. The significance here is that even at low concentrations and room temperature and humidity these materials will likely display substantial fading within a decade. While the colloidal silver is not found in modern 24