Volume 19, Number 2 .... May 1997

Conference Review

PROGRESS IN LEATHER CONSERVATION

University of Texas, Austin
March 12-14, 1997

Sponsors: Texas Memorial Museum, Harry Ransom Humanities Research Center, Leather Conservation News, and International Academic Projects

Coordinators: Jessica Johnson and Olivia Primanis

Instructor: Christopher Calnan, Adviser on the conservation of Organic Materials, National Trust (England)

This three-day seminar was a welcome opportunity for conservators working with leather materials to hear a well-organized and comprehensive presentation on the chemistry and structure of leather, historical and modern tanning processes, causes of deterioration in leathers, and, finally, a brief history of research in leather conservation, and a chronology of treatment approaches with their associated successes and failures. The instructor for this highly informative course was Christopher Calnan, Adviser on the Conservation of Organic Material for England's National Trust. Throughout the three days, he kept up a ceaseless flow of information, which was avidly absorbed by the assembled group. Each day was divided into two parts: A morning lecture session, attended by about thirty conservators and conservation students, and an afternoon lab practical, attended by twelve participants who had registered for the full course. All sessions took place in the conservation department of the Harry Ransom Humanities Research Center, on the University of Texas campus.

The participant group was a mix of book conservators and objects conservators with a diversity of specialties from archaeological conservation to ethnographic materials, from all over the U.S. For me, as a book conservator, this was one of the main attractions. Book conservators tend to be trained with and by their own kind, and the opportunity to exchange information with objects conservators and a bona fide leather specialist was just what I had been waiting for. Many thanks to Jessica and Olivia for organizing the course, and to Chris Calnan for sharing his knowledge.

Chris opened the first morning session with a disclaimer, stating that he wasn't sure there had actually been much progress in leather conservation lately, but that we could judge for ourselves at the end of the course. He also said that at the moment there are very few people involved in new research in leather conservation, and that much of the information available to conservators today comes from the leather industry. That said, he launched into the first block of lectures, covering the molecular structure of collagen and the morphology of skins and leathers from various animals. Understanding the physical structure of a skin is essential for identification purposes - each species of animal skin shows a characteristic ratio of grain layer to corium, a distinct follicle pattern, and a typical arrangement and density of fibers. These morphological features were illustrated with slides, diagrams and samples of a myriad of skin types. Some species are easy to identify (such as pigskin, with its large, distinct follicles in triangular groups of three), while others are anomalous (such as hair sheep, which is more similar to goat than to regular wool sheep). Physical features of the skin can be altered by processing, making identification more difficult. The corium layer can be largely shaved off to produce a thinner skin, while surface treatments such as coatings or false graining can disguise a skin's true animal origin.

During the lab session, participants experienced the challenges of skin ID, using stereo microscopes to compare new leather reference samples with a set of seven "unknowns" of various types, ages and conditions. Many of the samples had indeed been processed to look like something they weren't, such as sheepskin that had been given a grain surface that simulated morocco goat, a higher quality and more expensive material. The giveaway was the ease with which the grain layer split away from the corium, a characteristic of (wool) sheepskin caused by fatty deposits that leave voids in the skin structure when the fats are removed during tanning. It was interesting to compare the deteriorated samples with the new reference samples, and to try to decipher the identity of a skin whose most obvious identifying characteristic was eroded or missing. Kristen St. John, a student in the University of Texas Preservation and Conservation Studies program, assembled an exhibit of historical bookbindings from the HRC collections; these were covered in a wide variety of skin materials, on which we could further test our identification skills.

The lecture on tanning covered numerous historical processes. The most common tannage used in Europe and Colonial America was vegetable tannage using oak bark or other plant materials to tan skins that had been previously de-haired in a series of lime baths. Through the 18th century, tanning was a very slow process, tailored to the raw materials and to the desired end product. For example, longer liming would produce a softer leather with a more open skin structure. During the 17th and 18th centuries, leather for shoe soles would be limed for 2 weeks, belting leather for 6 weeks, while three to four months might be necessary to produce very soft leathers. Additional processes such as drenching and bating (immersion in weak organic acids and dog or pigeon dung), could be used to further open the skin structure, contributing to softness, flexibility and better penetration of tannins. The whole tanning process could take over a year, and up to two years for very thick hides.

The industrial revolution brought with it the need for greater production speed, and chemical agents (sulfites and sulfates), highly concentrated or condensed vegetable tannins, and mechanical processes were introduced to achieve this end. The results were not unlike those of late 19th-century changes in papermaking: extreme susceptibility to deterioration from atmospheric pollution and other environmental stresses, very low pH levels, and an associated irreversible shortening and embrittlement of the leather fibers. These changes in the manufacturing process provide much of the explanation for the relatively good condition of many older leathers compared to modern vegetable-tanned leathers stored under similar conditions.

Today there are two tanneries surviving in England that still use traditional English oak-bark tannage, a process that takes 8 months and is used mostly for harness and saddlery. A number of other small tanneries producing goatskins for bookbinding and conservation import skins from Nigeria, where they are tanned for export using the native acacia arabica plant. Upon their arrival in England, the original tan is either partially or entirely stripped out using a mild alkali, and the skins are retanned using a vegetable material such as sumac. For at least 100 years, the Nigerian skins have had a reputation for excellent durability, but increasing industrialization may eventually bring about the introduction of harsher chemicals for tanning. There have already been instances of a persistent spew appearing on whole batches of bookbinding leathers produced in England from Nigerian skins that were loaded with cheap fats to increase their weight prior to sale to the English manufacturer. Like everything else, the quality of a finished leather is dependent on the raw materials as well as the methods used, and the continuation of traditional processes in today's world grows increasingly uncertain.

Other tannins and tanning methods were also described, including oil, chrome, aluminum, brain, brain-and-smoke, alum-tawing and "Hungarian" tanning (alum-tawing followed by the application of oil driven in by heat, used mainly for harness leather). Chrome tanning, introduced in the mid-19^th century, produces an extremely stable leather using a process that is quick and inexpensive, and it is used extensively in modern leather manufacture. Chrome-tanned leather has a high pH relative to vegetable tanned leather, absorbs less atmospheric SO₂, and is resistant to dimensional change from contact with water. However, some of the very characteristics that make it so durable also make it difficult to work with for applications such as bookbinding, which requires the skin to be pared thin and molded when wet, operations that work best with traditional vegetable-tanned and alum-tawed binding leathers. Alum-tawed skins are also quite durable, but are very sensitive to water.

In an attempt to combine the best of all worlds, manufacturers of bookbinding leathers have recently developed combination tans using vegetable and mineral processes (chrome or aluminum) on the same skin. Chris clarified the confusing terminology of these combination processes. A "semi-chrome" is skin tanned first with a vegetable tannin, followed by a final chrome-tan; a "chrome re-tan" is a chrome-tanned leather subsequently re-tanned using a vegetable process. The predominant property in terms of the leather's durability is the first tannage, so therefore a chrome re-tan could be expected to fare better than a semi-chrome in an aggressive environment. However, it may also have working properties that are less desirable than a semi-chrome or a straight vegetable tan.

The main causes of deterioration in leather were identified as follows, and illustrated with slides:

1. RH, temperature, and associated effects on moisture content The average moisture content of leather is 15%. RH above 75% can cause mold growth and swelling of the leather fibers, whereas RH below 25% can cause shrinkage of the skin. A combination of dryness and heat can cause irreversible deformation of the skin. Cycling of temperature and RH can lead to the development of stress fractures and cracking. Persistent low temperatures (below 12 degrees C) can solidify the fats present in the skin and cause formation of spew on the leather surface. High temperatures will cause dryness and embrittlement, and may also turn skin oils rancid. Water damage can cause the migration of soluble tannins to the surface of leathers, where they cause darkening and hardening of the skin. Deteriorated leathers are particularly susceptible to this effect, the fibers being more likely to collapse and stick together.

2. Chemical deterioration - oxidation and acid hydrolysis Absorption of large quantities of SO₂ by vegetable tanned leathers will ultimately lead to "red-rot", the irreversible shortening and embrittlement of leather fibers which causes the surface to fall away from the leather in the form of a fine, reddish powder. Vegetable tanned leather is second only to cotton and wool in its ability to absorb SO₂; the presence of the condensed tannins introduced in the 19^th century exacerbates this effect. Alum-tawed and chrome-tanned leathers are less affected. Photochemical attack on leather will cause bleaching, embrittlement, and even denaturation of the protein in the skin. Chemical deterioration increases the leather's vulnerability to physical damage.

3. Biological attack - insects and mold While leathers are not normally attacked directly, damage from insect larvae frequently occurs when leather is associated with paper, textiles, adhesives, etc. Mold can be supported by the fats and waxes on the surface of the leather. The organic acids and enzymes produced by mold activity can bleach the skin. Previous attack by mold can also encourage insect attack directly on the skin.

4. Physical wear and tear from improper handling or continued use

5. Inappropriate conservation treatments Chris then went on to outline the history of research into leather and its deterioration in the UK, starting in 1843 when gas lamps were identified as a major cause of leather "rot", and leading up to current research. As early as 1905, the Royal Society of Arts in London established a commission to carry out a study of leather manufacture, the results of which identified condensed tannins, acidic dye baths, and the use of gas lighting in libraries as contributors to the widespread problem of leather decay. It also recommended the use of Nigerian goatskins tanned with the hydrolyzable sumac or acacia arabica for the manufacture of bookbinding leathers.

The 1920's and -30's saw the development of the P.I.R.A. (Printing Industries Research Association) test. This test subjected leathers to repeated bouts of aggressive oxidation and hydrolysis using peroxide and sulfuric acid, in order to assess their resistance to adverse conditions. At the same time, long-term natural aging trials were set up using 500 bindings made with some 109 different skins. The books were divided into two similar groups and one sent to the British Library in London, the other to the National Library of Wales. After 40 years, 66% of the leathers passing the P.I.R.A. test had lasted well, with a very small percentage showing severe deterioration. Notably, the leathers stored in the National Library of Wales, a well-controlled and relatively pollutant-free environment, survived better than those stored in the British Library, which at the time lacked adequate environmental controls, leaving the books exposed to the full impact of urban pollutants and seasonal changes in temperature and humidity.

In the late 1970's and early '80's, Betty Haines of the British Leather Manufacturer's Research Association (BLMRA) developed recommendations for a semi-aluminum tannage (vegetable followed by aluminum) as a way to protect vegetable-tanned leathers from rapid chemical deterioration; she also investigated the use of surface treatments for red-rot.

In 1991, the EEC funded a three-year study called the STEP (Science and Technology for Environmental Protection) Leather Research Project. The project involved the collaboration of five European conservation laboratories, with additional participation of seven other institutions. The purpose of the project, in which Chris Calnan participated, was to develop meaningful artificial aging methods for vegetable tanned leathers, as well as standard tests to evaluate the potential stability of new leathers. Various historical samples were collected and tested, including samples from the P.I.R.A. test groups in London and Wales. Among the many factors used to evaluate the conditions of the samples were pH, shrinkage temperature, the amount of sulfate present, the physical state of leather fibers, and the ratio of alkaline to acidic amino acids in the samples. Samples with condensed vegetable tannins showed higher sulfate levels, lower shrinkage temperatures and lower pH levels than samples with hydrolyzable vegetable tannins or chrome-tanned samples. They also showed the greatest degree of physical deterioration. Some of the chrome-tanned leathers were virtually unchanged. (A report on the project has been published by the European Commission: Larson, R. et al. STEP Leather Project. Copenhagen, Denmark: The Royal Danish Academy of Fine Arts, School of Conservation. 1994)

In general, shrinkage temperature (Ts) and pH turned out to be the two most significant indicators of the future stability and the current condition of the leathers tested. One exception might be alum-tawed skin, which has a low Ts (55-60 degrees C) despite its proven durability over periods of hundreds of years. A new chrome-tanned skin will generally have a Ts of close to 100 degrees C, and a pH over 4. A good-quality new vegetable-tanned leather will shrink at 75-85 degrees C, and show a pH of 3.5-4.5. (The isometric point of collagen is 4.5.) A pH level of 3 or less is considered very low, and is characteristic of unstable new leathers or old, deteriorated leathers. Deteriorated and unstable leathers will also shrink at much lower temperatures than those listed above.

The lab practicals for this part of the course consisted of simple tests to identify and characterize tannins, and to establish pH and shrinkage temperature. The presence of vegetable tannins is detected with the iron tannate test; drop 1% ferric sulfate solution on the sample, which will produce a blue/black stain if the result is positive. Condensed vegetable tannins are identified using the vanillin test, in which a drop of vanillin dissolved in alcohol is applied to the sample, followed by a drop of concentrated hydrochloric acid. A bright red color indicates a positive result.

Ash tests were also carried out by incinerating skin samples in a muffle furnace or with a blowtorch, and observing the color of the ash. Green indicates the presence of chromium, while white residues identify alum. Black ash is left by a vegetable tan, oil tan, or an untanned skin. For a combination tan, the ash text may completely burn away the vegetable tanned portion of the sample, leaving only the mineral residue. Therefore, the iron tannate test should be done first to determine if vegetable tannins are present.

Shrinkage temperatures of various samples were observed simply by attaching a small strip of skin to the bulb end of a glass thermometer suspended in a beaker of water on a hot-plate, and noting the temperature at which the sample curled up. We used samples cut from our study leathers, but shrinkage temperature can also be observed by placing individual fibers on a glass slide under a microscope with a heated stage.

pH readings were taken both from the surfaces of samples with a flat-head electrode, and by reading the pH of a slurry of fine shavings in water. We also applied small bits of fiber scraped from the samples to narrow-range Whatman pH indicator strips (1.8-3.8) dampened with distilled water. All of the methods used to test pH produced similar results, and in general, the relative pH values of the aged samples were consistent with their apparent condition. One exception to the reliability of a pH reading as an indicator of stability can occur when the leather is so deteriorated that all free sulfate has been consumed and replaced by ammonium sulfate, the product of complete breakdown of the proteins by sulfate. The ammonium sulfate will give a deceptively high pH reading. Chris used this point to stress the importance of using more than one parameter to assess the stability or condition of a given sample.

The final lecture session of the course discussed leather conservation treatments, past and present. In an attempt to mitigate the effects of extreme acidity and atmospheric pollution, lactate buffers were added to new leathers, usually in the form of potassium lactate solutions. The inspiration for this approach, developed in the 1920's and 30's, came from the identification of "non-tans" found in Nigerian native-tanned skins -- salts, sugars, and other substances that were thought to have a role in the skins' exceptional stability. New skins treated with solutions of 10% or greater do seem to have survived somewhat better than untreated skins, but lower concentrations showed no benefit at all. Inevitably, these lactate solutions were also applied to historical leathers, with no discernible benefit; the salts have frequently leached back out onto the surface over time, forming hazy white deposits.

In the 1960's, ammonia vapors were explored as a way to reduce acidity in leather. In theory the ammonia would combine with free sulfates in the leather to form stable ammonium sulfate; in practice it was unsuccessful and introduced a highly alkaline substance into a system that is stable at a fairly low pH. In the mid-1980's, Chris was involved in a project at the Leather Conservation Centre in Northampton, England that explored the use of alkoxides as a means to introduce aluminum into vegetable-tanned leather. These compounds, soluble in organic solvents, could combine with vegetable tannins and also with free sulfates to form aluminum sulfate (alum) which is extremely stable. This process was found to significantly raise the shrinkage temperature of the samples treated as well as maintaining a pH of well above 3 even after aggressive artificial aging. However, after ten years, the protection of new leathers against deterioration has not been as great as was expected based on early test results, and in the case of deteriorated leathers the benefits were not significant. Chris now feels that the best use for alkoxides may be as a preventive treatment for poor quality new leathers likely to deteriorate in the future. He also noted that he and other leather conservators have grown away from the idea that eliminating acidity from leather is an appropriate or feasible goal. At this point it seems more productive to accept leather's inherent acidity and work to prevent its exposure to conditions that hasten its deterioration.

The subject of leather dressings and surface coatings was of particular interest, since most of the class deals regularly with questions from collectors, friends, members of the public and other institutions concerning what can be safely used to coat or lubricate leather objects. Most of us say "nothing", having seen some of the incredible damage that has been done in the past by over-application of these products. Chris felt, however, that adding a lubricant is appropriate under certain conditions, and with an understanding of potential drawbacks. Most leathers are stable at an oil content of about 5% by weight, and many historical leather objects have survived very well without any lubrication other than what was originally added when the skin was manufactured.

Before adding a lubricant to any leather, it is important to decide if it actually needs more flexibility, and to understand why it is stiff. If the fibers themselves are deteriorated and brittle, lubrication will not help. If the object must be used, and the fibers are still reasonably strong, addition of a lubricant can add flexibility by increasing the space between the fibers and allowing them to move against each other more freely. Neatsfoot oil is a very effective lubricant, but darkens the leather substantially. Most commercially-available neatsfoot oil is now made up of lard, whose excess fatty acids can cause spew on the surface of the treated item. If neatsfoot oil is used, it must be "cold-tested" against solidification of fatty acids at low temperatures. Since application of a concentrated oil can increase the oil content of a skin by up to 10%, Chris recommends diluting neatsfoot oil to a 5/10% solution in mineral spirits, Stoddard's solvent, or petroleum ether before applying.

Avoid aggressive solvents that might strip out the oils already present in the leather. Over-application of oils can disrupt the moisture content of the leather, by taking up space that would otherwise be occupied by water. Lanolin, another traditional component of leather dressings, can cause problems because it is hygroscopic and may rob the surrounding leather of its natural moisture content. Lubricants afford no protection against the penetration of SO₂, although they can form a sticky surface layer that attracts lint and dust.

Consolidants are also problematic, since they are usually polymers with molecules too large to penetrate well into the structure of the skin. The cellulose ether Klucel G is currently in favor, particularly among book conservators, as a relatively benign surface consolidant. It is usually applied dissolved in ethanol, which, although it is a polar solvent does not affect leather as drastically as water. Olivia Primanis of the HRC described the use of repeated applications of very dilute (0.5-1.0%) solutions of Klucel G with an airbrush as a successful method for consolidating reverse calf, though an increase in the stiffness of the surface nap was noted. Higher concentrations of Klucel can cause color change. Chris suggested that isopropanol rather than ethanol as the solvent might reduce this effect.

Other options under consideration in the field are acrylic monomers such as Lankrothane 1034 (no longer available) that cure in contact with moisture. The monomers are introduced into the leather in solution with organic solvents, forming polymers in situ as the solvent evaporates out and moisture from the air catalyzes the hardening process. Parylene and other acrylic monomers are available, but the ideal material has not yet been identified, and so far such treatments are totally irreversible.

Surface coatings can provide a barrier to SO₂ penetration. SC6000, an acrylic and wax emulsion available from the Leather Conservation Centre, has changed formulations several times and is now more highly alkaline than the original product. While it has its uses (notably to provide surface loss to fills and repairs) it probably should not be used on historical leather objects as an overall coating. Microcrystalline sax coatings like Renaissance Wax form less effective pollutant barriers than SC6000, but they can be used to good cosmetic effect and without any worries regarding pH.

As many conservators have discovered from painful experience, water-containing adhesives, even viscous polyvinyl acetate emulsions, can cause irreversible blackening, hardening and shrinking of severely deteriorated leathers. This may be the result of deterioration that is so extreme that the shrinkage temperature of the leather is just barely above room temperature, and even the slight heat reaction that takes place as leather absorbs water is enough to shrink and collapse the fibers. To avoid the hazards of wet adhesives, Chris often works with Lascaux 360HV, which can be painted onto almost any support (polyester web, paper, goldbeaters' skin) and then reactivated with acetone or a chlorinated solvent and applied as a mend. If a less tacky film is desired, Lascaux 498HV, which dries harder, can be mixed in any proportion with the 360HV. Chris has found Beva 371 to be a good adhesive for fills.

The final lecture session ended with a laundry list of questions to ask leather suppliers about their products when purchasing new skins: Is it a re-tannage? What are the tannins? Beware of half-truths! Sumac leaf extract is not the same as sumac leaf. Even though sumac extract will show as a hydrolyzable tannin when tested, it doesn't give the skin the potential benefits of the non-tans and buffers present in true sumac leaf.

The last afternoon of the class was devoted to a field trip to the Materials Conservation Laboratory of the Texas Memorial Museum, workplace of conservators Jessica Johnson, Kathy Hall, and Marilyn Lenz. What seemed like hundreds of leather and rawhide objects from the collection were on view there, including saddles, harnesses, clothing, and even a stuffed rhinoceros head. This generated a group discussion of housing and storage, environmental controls, and dealing with insect infestations. The day concluded with a sort of general wrap-up discussion, and ended with everyone suddenly feeling the after-effects of three days of information overload. We all wondered how Chris was still on two feet. All of the participants I talked to felt they had learned a great deal, and although there are still no magic bullets for dealing with some of the most pernicious problems of leather conservation, we at least felt better equipped to think about them and deal with them in the future.