Address of all authors: Image Permanence Institute, RIT City Center, 50 West Main Street, Rochester, NY 14614, USA.
Presentation at the "Conservation in Archives" Symposium, National Archives of Canada, May, 1988. This article is the copyrighted property of the Image Permanence Institute and may not be reproduced without permission.
The stability of black-and-white photographic images (i.e., those consisting of a metallic silver image) is a significant issue for archival collections. It is important because both pictorial images as well as microfilm are included, and microfilming is now a vital technology for the preservation of deteriorating books and paper documents. Silver image stability is a complex, difficult subject which has often been misunderstood--and usually overestimated. Recently, ideas on this subject have changed dramatically, in part because of research dome by the photographic industry and the Image Permanence Institute. In this article, the history of the stability issue and the implications of the new research will be reviewed.
In recent years, new insights into the mechanisms of silver image deterioration have resulted in a change of emphasis in preservation measures. Long-held beliefs about the inherent stability of silver and the role of processing in causing deterioration have been questioned.. There are two fundamental mechanisms by which silver photographs deteriorate: they react with sulfur present in the air or left behind in processing, or else they fade because of image oxidation, in which air and moisture (or various pollutants) literally corrode the silver. Much of the literature on photographic preservation is devoted to ways of ensuring that the fixer solution (which is a sulfur compound) is completely removed during processing, and is not left behind to threaten the image in subsequent storage. Discoloration and fading due to poor processing are certainly very important concerns, but proper processing is not enough to guarantee stability. That is where the revolution in thinking has occurred (1).
Attitudes and concepts about the permanence of silver images have their beginnings in the 19th century, when photography itself was a new and little-understood technology. In the 1840s, the rapid fading of prints was a troublesome problem that threatened to completely discredit photography on paper, leaving daguerreotypy as the only durable way to make images. The basic chemistry of photography has changed surprisingly little since this early time, and then, as now, the last steps in processing were fixing in sodium thiosulfate (hypo), followed by washing in water.
A blue-ribbon committee of chemists and photographers was appointed by the Photographic Society of London in 1855 (financed by a donation from Prince Albert) to study the problem. This was the first sustained scientific inquiry into the stability of silver images. After more than a year of collecting examples and performing landmark accelerated aging experiments, the committee concluded: "The most common cause of fading, has been the presence of hyposulphite of soda [sodium thiosulfate], left in the paper from imperfect washing after fixing." (2) This was the origin of the belief, still almost universally held, that retained hypo and improper fixing are indeed the root cause of most silver image deterioration. We know now that for most types of black-and-white photographs, image oxidation due to moisture, pollution, and poor-quality enclosures (and not retained hypo) is in fact the commonest cause of image degradation.
The "Fading Committee," as it was popularly known, based its conclusions on examination of prints, the testimony of photographers about how the examples were made and stored, and on accelerated aging tests, some of the first of their kind to be performed. The committee' s methods were scientifically sound, led by the brilliant chemist T. F. Hardwich. In his separately published writings on image permanence (3), he accurately described the vulnerability of silver to oxidation, and utilized virtually all the currently used types of accelerated aging tests, including moist air, peroxide fuming, light fading tests, and sulfur dioxide fuming tests.
This pioneering research on silver stability was correct in singling out retained hypo as the cause of acute fading of early salted paper prints, but that particular problem was a limited, anomalous case from which it was wrong to generalize. Photographers deliberately left hypo in the prints and chemically decomposed the fixing bath in order to produce a more pleasing print color. Unfortunately, in the process of explaining the errors of these practices, a deep misconception and oversimplification was created which persists until the present day.
Later writers in the 19th century (4) spoke about image stability principally in terms of residual hypo, ignoring the other issues raised by Hardwich and the committee report. It is understandable that photographers could more easily relate to a simple, straightforward explanation (poor fixing and washing) rather than an unfamiliar chemical explanation (oxidation from a variety of causes) for the fading of silver images. In addition, oxidation proceeds slowly, and its symptoms cam resemble those caused by retained thiosulfate or absorption of sulfur gases from the atmosphere.
The oxidation mechanism is the one which has received the least attention over the years, but is the most important. Assuming that an image has been properly processed, its survival depends on the inertness of the metallic silver image. Though sometimes considered a "noble" (unreactive) metal, silver does react with air and moisture--it corrodes, like most metals. In the process, silver forms water-soluble species (silver ions and silver compounds) and these begin to migrate throughout the gelatin layer. Ultimately, most of the silver is redeposited in the metallic state, but at some distance and in a different physical form that it originally was. It is this physical rearrangement, together with the fact that silver compounds are largely colorless, that is the real cause of "fading." In actuality, the silver is repeatedly oxidized and redeposited, as long as the driving forces (air, moisture, or pollutants) are present.
In recent years it has been realized that there are many more sources of oxidizing gases in the storage environment than previously imagined (5), and an awareness has grown that even clean moist air produces a slow form of attack. Many kinds of woods, cosmetics, industrial processes, adhesives, plastics and other materials give off peroxides (6). The levels of oxidant pollution in general are rising, not declining. Whether attacked by oxidants at relatively high concentrations from a specific source such as storage cartons, or slowly oxidized by moist air, the longer one projects the lifespan of a silver image, the more important it becomes to protect the silver against oxidation.
The first serious concern about image stability of microfilm occurred in the early 1960s, when small, reddish circular faded areas were discovered on microfilm from several major collections, including the National Archives (7,8). The spots ranged in size from microscopic to several millimeters in diameter. Some spots had patterns of concentric light and dark rings, some occurred only in high density areas, and others only at boundaries between light and dark areas (6). There was great concern at the time because this kind of deterioration was previously unknown, and the extent of the threat was difficult to assess. They appeared primarily on the edges and first few turns of film stored on reels, and seldom were present on film stored in metal cans. Surveys of collections were done to document the extent and type of deterioration (9). Major research efforts to determine their cause and prevention were launched at the National Bureau of Standards and at Eastman Kodak Company.
Ultimately, it was determined that the spots were the result of oxidation of the silver image by oxidizing gases present in the storage environment (10). Peroxides generated by deteriorating poor-quality cardboard storage boxes were identified as a primary cause, although a wide variety of oxidants present in the atmosphere could damage film in the same way (7,9). Spotwise deterioration and overall discoloration could be produced using ozone or hydrogen peroxide in laboratory accelerated tests.
Red spots of "redox blemishes" form in still air when the right concentrations of oxidant are present, when scratches or other physical discontinuities in gelatin provide an opportunity for localized attack, and when humidities are high enough to allow gases to diffuse through the gelatin layer. The mechanism of oxidative attack has been elaborated in a series of articles by E. S. Brandt of Kodak Research Laboratories; he showed that spotwise attack proceeds by a complex and delicate electrochemical mechanism (11,12).
The threat to the viability of microfilm as an archival medium created by the discovery of redox blemishes was not catastrophic as originally thought. On the one hand the problem turned out to be less common than had been feared at first, and on the other a cheap, reasonably effective preventive measure was found and universally adopted. Surveys had pointed out the fact that some batches of film and some entire collections had virtually no microspot damage. It was discovered that those films which were processed in fixer containing a small; amount of potassium iodide were far less likely to develop redox blemishes than those fixed without iodide present. Laboratory work showed that adding 0.2 grams of potassium iodide to each liter of fixing solution provided protection against red spots, and this became (and still is) standard practice (13).
Considerable laboratory work was devoted during the mid and late 1960s to finding a way to make microfilm images more resistant to oxidation. The iodide treatment increased resistance, but accelerated aging tests showed that the protection was not absolute. Eastman Kodak strongly recommended gold toning to protect images against oxidative attack, and demonstrated that it gave better protection than iodide (14,15). But gold toning was expensive and required major changes in processing regimes; it was never adopted on a major scale, even though its benefits were quite well established. By the early 1970s the concern over microfilm stability had lessened considerably, and confidence in it as an archival recording medium was restored. In the 1980s, microfilm continues to be the preserved archival storage technology, despite advances in digital magnetic and optical systems (16).
However, most recently there seems to be a much wider scale "redox blemish" problem in library and archive microfilm collections than is generality believed. Since the Image Permanence Institute began its microfilm project in July 1987, a rather large number of institutions which have had such problems (in varying degrees of severity) have come to our attention. These include:
The prevalence and severity of such outbreaks--for example, 21% of the 500,000 rolls in the Illinois State Archives are affected--is evidence that iodide does not provide significant enough protection in a world with so many sources of oxidant pollution. Much of the Illinois film with red spots dates from within the last 20 years, well after the practice of adding iodide to the fixer was introduced.
One of the best microfilm storage facilities in the world is that of the Genealogical Society of Utah (Mormon Church) near Salt Lake City, Utah. According to Herbert White, manager of micrographic services, redox blemishes have been found on film in their collections, but so far only in cases where the film spent some part of its life in storage outside their air-purified and temperature/humidity controlled vault. Some of the affected film was stored in a converted warehouse prior to construction of the vault in the mid-1960s.
The experience of the Mormon Church collection is instructive in two ways: it shows that if the environment is controlled in all respects, then film will be at minimal risk. On the other hand, it suggests that anything less than the ideal storage environment can lead to problems. The Mormon case is rare indeed; many collections have some environmental controls, but few can afford to fulfil the ideal in as many respects (activated charcoal scrubbers to remove oxidants from the vault atmosphere, new archival cartons for every roll not produced inhouse, etc.) as they do. For most collections, the prognosis is for more and more image stability problems for their microfilm collections.
At first the problem of redox blemishes was thought to be peculiar to microfilm. They were later found to occur on many types of silver photographic materials, including normal camera and cinema fins, gelatin dry plates, auto-chromes (an early 20th century color transparency having a silver image), astronomical plates (17), resin-coated paper prints (18), and glass lantern slides. Redox blemishes are now realized to be a special case of the larger and more general problem of oxidative attack on silver images. While the spotwise attack on microfilm images can be very damaging and can occur in relatively short time periods (within five years), the more frequent form of oxidative attack is overall fading, silver mirroring, and discoloration. It is important that these hazards to photographic image preservation also be addressed.
During 1987-88 the Image Permanence Institute conducted a study of the benefits of selenium treatment of microfilm to improve its image stability. This project, which was funded by NEH and the New York State Library Preservation Grant Program, began in July 1987, and at the time of this writing (May 1988) is still in progress. It had two main purposes: 1) to develop an improved hydrogen peroxide accelerated test for image oxidation, and 2) to establish that treatment of microfilm with selenium made it resist image oxidation, without adversely affecting resolution, emulsion physical properties, information content, or image density.
The first objective, devising an improved hydrogen peroxide aging test, was met in full. The purpose of the peroxide test is to simulate the effect of oxidant pollution on photographs. Redox blemishes are not produced when microfilm is put through conventional accelerated aging utilizing heat and humidity only. There is a slow general fading caused by moist air, but in such tests red spots do not appear. Only when the test environment contains an active oxidant like hydrogen peroxide are spots produced. Therefore, peroxide tests are the best way to explore whether microfilm does or does not have the ability to resist spot formation due to air pollution.
Hydrogen peroxide tests have been used since the 1850s, but have always suffered from a lack of reproducibility. Most photographic manufacturers use such tests internally, but no two companies use the same technique. Up until now no method has been considered broadly useful or repeatable enough to be adopted as an American National Standards Institute (ANSI) standard test for silver image stability. IPI's first attempts began with a glass desiccator jar in which film samples were suspended over a piece of paper soaked in peroxide solution. This arrangement provided spots and discoloration on the film but was not repeatable. The problem was to achieve a uniform distribution of peroxide vapors on all film surfaces.
After several months of frustrating work, in which numerous internal arrangements were tried, a solution was found to the difficulty of distributing the peroxide vapors evenly, and the goals of repeatability and uniformity of action were achieved. This was done by controlling the geometry of the sample holder in a perfect circle and by use of a high-speed fan which runs only long enough to guarantee even distribution of vapors. Many other details of the technique also had to be examined and controlled to establish reproducibility, but the final result was an improved technique which overcame the most important drawback to peroxide testing, lack of reproducibility.
Attention now turned to examination of the main variables in the test itself: temperature, RH, time, and peroxide concentration. These were examined in a factorial experimental design. Humidity and peroxide concentration were shown to be critical quantities, while time and temperature were of lesser importance. The data from this experiment are the first to be obtained showing the relationship between temperature, time, humidity, and oxidant attack on microfilm. They demonstrate that if the RH of the storage area is kept at 50% or below, even high concentrations of oxidant have little effect on the film. This is strong ammunition to add to an already large body of reasons to make RH control a first priority of collections.
Four different sets of conditions for the peroxide test were identified as potentially useful in research, based on the severity of attack, and how well the appearance of actual field samples of affected microfilm was reproduced in the test. By adjusting the test conditions, almost all kinds of blemishes and discoloration observed in field samples can be recreated in the peroxide test. The single test condition most useful for everyday stability research (such as in evaluating the effectiveness of stabilizing treatments) was 18 hrs. at 50°C, 81% RH, with a (calculated) peroxide concentration of 2000 parts per million.
The improved peroxide test is a very significant advance; it is relatively easy and quick to perform, and gives reliable results. There is a likelihood that this method will become an ANSI standard test for oxidation resistance of all types of silver images. Experiments showed that the test worked well with photographic prints and other types of film. The method was discussed with technical staff of Kodak and Fuji on several occasions during development, and both companies are examining the benefits of the improved test procedure for their internal testing programs.
The must significant aspect of this new test for the microfilm community is that it reproduces the redox blemishes that are the biggest threat to microfilm collections. If in the real world red spots are the main concern, then it is an accelerated test for red spot formation which ought to guide what is done in microfilm processing to guard against this difficulty. The improved test is a more precise tool for investigating microfilm image stability than has existed before.
Once the peroxide test technique was finalized, the project plan called for "benchmarking" the test with films whose relative stability was known. This included gold toned and non-iodide treated films, as well as films intentionally "spiked" with varying levels of residual thiosulfate. The purpose of these experiments was to be sure that the peroxide test ranked the stability of these films in accordance with previous experience and expectations. This series of experiments provided no results which would undermine the validity of the test method. However, both the "benchmark" experiments and those that followed using selenium and a variety of other toners provided startling results.
The results of the IPI research led to four main conclusions:
The detailed results of these quite recent experiments will be published in appropriate journals; the discussion here will be confined to some of the specific findings which bear on the practice of preservation microfilming. First and foremost, even the mildest peroxide tests show current camera original microfilms--the ones most often used for archival copies--to be extremely vulnerable to attack from oxidant pollution. This fact, together with the serious outbreaks of redox blemishes now appearing, sends a loud and clear signal that some action is needed, before confidence in microfilm is eroded, and hard-won resources for preservation microfilming are questioned. Fortunately, there does seem to be a relatively cheap and completely effective approach to pursue, namely sulfiding treatments.
The peroxide tests at IPI showed that Kodak Rapid Selenium Toner failed to provide protection against redox blemishes, when used as suggested. If highly concentrated solutions were used, the level of protection increased, but was not complete. Such concentrated solutions are impractical for reasons of cost, excessive contrast buildup, and excessive image color change, even if they did provide enough protection. This finding conflicted with numerous published results from Kodak (17,20,21); when we spoke with Kodak personnel, they confirmed that in their own recent peroxide testing with microfilm, the selenium toner was depositing selenium, but not preventing oxidant attack, which it had done in tests performed as recently as one year ago. They suspected that small changes in formulation made by the manufacturing area were responsible, but were not clear on exactly why.
It is our strong feeling that the changes in formulation that suddenly rendered dilute selenium toner ineffective relate to the sulfiding action of minor constituents. Although the formula for Kodak Rapid Selenium Toner is proprietary, it is known to contain both sodium sulfite and hypo (sodium thiosulfate), both of which may be contaminated with small amounts of highly active sulfiding agents. Apparently insignificant manufacturing changes may have caused this active agent to be no longer present; it would still form silver selenide and achieve a toning action (in the sense of color change), but would no longer protect against peroxide. In any case, the surprising ineffectiveness of Kodak Rapid Selenium Toner, together with many other signs of the potency of sulfiding agents, pointed the way to a much different analysis of image stability and how to achieve practical protection against red spots.
One of the strongest clues to the power of sulfiding agents to protect against peroxide came from experiments with gold tuners. Kodak has recommended a formula known as GP-2 since the 1960s for the treatment of microfilm to prevent red spot attack (7,14). Because of the high cost it has seldom been used in practice, but it was always regarded as absolute protection. One of the ingredients of GP-2 is thiourea, a known sulfiding agent. In experiments at IPI, this formula was indeed completely effective in preventing peroxide attack. However, experiments with the same formula without the gold were completely effective. In both the gold toner and the selenium toner, it seemed to be the sulfiding agents, not substitution with gold or conversion to silver selenide, that was providing the bulk of the protection against oxidants.
This was confirmed in another series of experiments where gold and selenium formulas which did not contain a sulfiding agent were used. They "toned," in the sense of depositing gold or converting the silver to silver selenide, but the protection against peroxide was only in proportion to the degree of substitution. Even virtually complete conversion of the silver image to gold did not stop peroxide attack; the small amount of remaining silver discolored in the peroxide test.
There is great promise for sulfiding treatments as a way of protecting microfilm against oxidative attack. Once we had determined that the sulfiding ingredients were in fact responsible for most of the protection imparted by dilute selenium and gold toners, we began to explore the effectiveness of various compounds which might form a layer of silver sulfide at the surface of developed silver grains. This work is still under way, but already at least one simple, extremely effective approach has been identified: the use of polysulfides, as found (for example) in the commercial product "Kodak Brown Toner." This product gives complete protection in our severe hydrogen peroxide test (2000 ppm), even when used in quite dilute solution (for example, 1 part toner to 200 parts water).
It is characteristic of the sulfiding approach that only a small amount of the sulfiding agent is needed. For example, sodium sulfide solutions of 0.1 grams per liter (about 1/100th of a percent) are completely effective. However, for reasons of diminished odor, toxicity of the bulk substance, and shelf life of the solution, the polysulfides are preferable in practice to straight sodium sulfide. We have shown that Kodak Brown Toner does its work of protecting the image silver without significant change of density or image hue. The method of treatment is simple: processed microfilm of any age can be immersed in the solution for a few seconds (shorter immersion times require slightly higher concentrations than longer times), then washed and dried. Conventional processing equipment can be readily used for post-treating, with throughput rates comparable to normal processing.
NEH's Office of Preservation granted permission to IPI in May 1988 to formally change the experimental objectives of the 1987-88 "Selenium" project; the mew objectives are to explore the effectiveness of various sulfiding treatments and identify possible compounds for further evaluation in subsequent work. When this study is completed in December 1988, we will have a better understanding of which sulfur compounds are effective against peroxide attack and which are most favorable from the point of view of cost, ease of use, toxicity, density change, image hue change, and possible odor. One of the most positive aspect of sulfiding treatments compared to selenium toners is that they are far cheaper and pose less environmental hazard.
IPI is currently seeking funds to undertake a new research project whose purpose is to thoroughly explore all aspects of the protection offered by sulfiding treatments for microfilm, and all the practical issues surrounding its use. It is essential that those doing preservation microfilming have a method of protecting their collections against oxidative attack as soon as possible. Sulfiding seems to offer that protection, but there are severai important questions left to answer before it can be recommended for use. These include:
Though many silver images have been ruined by poor processing (improper fixing and washing), a more important deterioration mechanism for the majority of photographs is image oxidation. Microfilm collections are especially vulnerable to the form of oxidation known as "red spots," and outbreaks of this problem are becoming more frequent and severe. Numerous recommendations have been made for the use of gold and selenium treatments to protect silver images against all forms of oxidation (overall attack as well as red spots). Recent research at the Image Permanence Institute has shown that sulfiding treatments give excellent protection for microfilm against red spots, and that gold and selenium treatments only are effective (in the absence of sulfiding action) to the degree chat the silver image is converted to gold or to silver selenide. Planned new research may result in a redefinition of "archival processing" for all types of silver images to include sulfiding treatments to resist oxidation.
1. F. J. Drago and W. E. Lee, "Review of the Effects of Processing on the Image Stability of Black-and-White Silver Materials," Journal of Imaging Technology, Vol. 12, Feb. 1986, pp. 57-65.
2. Journal of the Photographic Society, Vol. 2, No. 36, Nov. 21, 1855, pp. 251-252.
3. T. F. Hardwich, A Manual of Photographic Chemistry, London: John Churchill Publishers, 1855, p. 153.
4. Fr. Wilde, "The Permanency of Photographs - Silver, Carbon, and Platinum," American Journal of Photography, Vol. 12, No. 134, Feb. 1891, pp. 49-58.
5. E. Weyde, "A Simple Test to Identify Gases Which Destroy Silver Images," Photographic Science and Engineering, Vol. 16, July/August 1972, pp. 283-286.
6. C. S. McCamy and C. I. Pope, "Current Research on Preservation of Archival Records on Silver-Gelatin Type Microfilm in Roll Form," Journal of Research of the National Bureau of Standards (A), Vol. 69A, Sept. 1965, pp. 385-395.
7. R. W. Henn and D. G. Wiest, "Microscopic Spots in Processed Microfilm: Their Nature and Prevention," Photographic Science and Engineering, Vol. 7, Sept. 1963, pp. 253-261.
8. E. McCrady, "The History of Microfilm Blemishes," Restaurator, Vol. 6, No. 3-4, 1984, pp. 191-204.
9. C. S. McCamy, S. R. Wiley, and J. A. Speckman, "A Survey of Blemishes on Processed Microfilm," Journal of Research of the National Bureau of Standards (A), Vol. 73A, Jan. 1969, pp. 79-99.
10. C. I. Pope, "Blemish Formation in Processed Microfilm," Journal of Research of the National Bureau of Standards (A), Vol. 72A, May 1968, pp. 251-259.
11. E. S. Brandt, "Mechanistic Studies of Silver Image Stability. 1: Redox Chemistry of Oxygen and Hydrogen at Clean and Adsorbate-Covered Electrodes," Photographic Science and Engineering, Vol. 28, Jan. 1984, pp. 1-11.
12. E. S. Brandt, "Mechanistic Studies of Image Stability. 3. Oxidation of Silver from the Vantage Point of Corrosion Theory," Journal of Imaging Science, Vol. 31, Sept./Oct. 1987, pp. 199-207.
13. R. W. Henn, D. G. Wiest, and B. 0. Mack, "Microscopic Spots in Processed Microfilm: The Effect of Iodide," Photographic Science and Engineering, Vol. 9, May 1965, pp. 167-173.
14. R. W. Henn and B. D. Mack, "A Cold Protective Treatment for Microfilm," Photographic Science and Engineering, Vol. 9, Nov. 1965, pp. 378-384.
15. R. W. Henn and D. G. Wiest, "Properties of Cold-Treated Microfilm Images," Photographic Science and Engineering, Vol. 10, Jan. 1966, pp. 15-22.
16. Preservation of Historical Records, A Report by the Committee on Historical Records of the National Materials Advisory Board, National Research Council. National Academy Press, Washington, DC, 1986.
17. W. E. Lee, F. J. Drago, and A. T. Ran, "New Procedures for Processing and Storage of Kodak Spectroscopic Plates, TYPE IIIa-J," Journal of Imaging Technology, vol. 10, Feb. 1984, pp. 22-28.
18. L. Feldman, "Discoloration of Black and White Photographic Prints," Journal of Applied Photographic Engineering, Vol. 7, Jan. 1981, pp. 1-9.
19. Y. Minagawa and M. Torigoe, "Some Factors Influencing the Discoloration of Black-and-White Photographic Prints in Hydrogen Peroxide Atmosphere," Preprints of the 1st SPSE Symposium on Stability and Preservation of Photographic Images, Ottawa, 1982, p. 43.
20. W. E. Lee and F. J. Drago, "Toner Treatments for Photographic Images to Enhance Image Stability," Journal of Imaging Technology, Vol. 10, June 1984, pp. 119-126.
21. F. J. Drago and W. E. Lee, "Stability and Restoration of Images on Kodak Professional B/W Duplicating Film / 4168," Journal of Imaging Technology, Vol. 10, June 1984, pp. 113-118.