Today's Guest Bloggers are Gregory Dale Smith, Ph.D., the IMA's Otto N. Frenzel III Senior Conservation Scientist, and Michael Columbia, Ph.D., Sabbatical Leave Research Fellow - IPFW
It is an uncomfortable truth that in showing you an artwork in a museum, we are potentially destroying it. As a conservation professional, it feels wrong to admit that, but it is true. Every photon, or packet of radiant energy, that strikes the surface of an art object has the potential to do damage, and we most often see that as a negative change in the artwork’s aesthetics: darkening, fading, yellowing, chalking, crosslinking, etc. It’s an unstoppable phenomenon, but one that proceeds at a variety of rates. Certainly color change is one of the most notable alterations that light can cause in an artwork, and so we must dole out the expected lifetime of an object using an informed and rational approach. Conservators and collections managers go to great pains to protect artwork by limiting its exposure to light. This can take the form of reducing light intensity, restricting its spectral output, or limiting the duration of an exhibition. These stewards of the collection get additional insight and data from scientists who study the fading behavior of artists’ materials.
For the past several months the IMA has been conducting a condition survey of its photograph collection, over 800 objects that span the history of the medium. This program is sponsored by a generous grant from the Institute of Museum and Library Services (IMLS), a wing of the federal government that supports museum and conservation activities. In addition to the inventory and conservation assessment of each artwork, the grant has also funded a study of the lightfastness of the contemporary color photographs in the collection using a technique called microfade testing (MFT), or microfadeometry. The goal of the study is to determine the susceptibility to color change for the highest priority color photographs in the collection and to determine patterns of lightfastness among the many photographic processes. This data in turn informs our exhibition, loan, and lighting guidelines for the collection.
Figure 1. Watercolor paint outs after artificial light aging.
In the past, the kinetics of fading were determined using surrogate materials, not actual artworks, and these were assumed to reflect the behavior of the real objects. These so-called mock-ups were subjected to intense irradiation in light aging chambers to accelerate their natural fading rates to fit a laboratory timescale. Periodically the color of the samples would be measured, and a fading curve would be created that would predict the amount of color change accompanying a certain dose of light. Called the “rule of reciprocity,” fading in intense light for a short duration (weeks or months) should mimic the effects of dimmer light over a much longer time span (years or centuries). Figure 1 shows watercolor samples of a geranium lake pigment in the museum’s light aging chamber. The aluminum foil that covered a portion of the sample during the test has been pulled back to reveal the fading effects of simulated sunlight through window glass. This short but intense irradiation in the laboratory would be the equivalent of just a few years of ambient exposure in a sunny room. Geranium lake, which contains the synthetic dye eosin, was unfortunately often used by van Gogh during the last years of his life and has resulted in significant color change to some of his paintings.
Unfortunately, the vagaries of artworks are often hard to predict when assembling a mock-up. The medium, color density, substrate, admixtures, and previous fading history of a work can all change the rate of fading of a pigment or dye. Although acceleration is present in the mock-up experiments, the correlation to real world behavior is often imperfect. To overcome these difficulties, a microfade tester (MFT) was assembled by Dr. Paul Whitmore, conservation scientist at Carnegie Mellon University in the late 1990s. The MFT uses a powerful xenon arc lamp and filters to focus an intense light beam through a fiber optic to a small spot – the width of a few human hairs – on the surface of the actual artwork. A second fiber optic collects the reflected light — what we would sense as color — and delivers it to a miniature spectrometer. As the lamp fades the surface of the artwork, the resulting color change is monitored by the spectrometer. Because of the instrument’s sensitivity, the fading experiment can be stopped before the sample location reaches a color change perceivable to the human eye. No mock-up experiments are needed, and all the complexities and previous history of the sample are included in these in situ measurements. The MFT is transportable, so artworks of any size or shape can be measured, often without even having to be removed from the gallery wall. Figure 2 shows visiting professor Mike Columbia from Indiana University Purdue University Fort Wayne (IPFW) conducting a fade test on a large Cibachrome print, Dante’s Inferno, by Andres Serrano.
Figure 2. MFT analysis of Andres Serrano’s Dante’s Inferno
(1998.39, created 1990, Gift of the Contemporary Art Society)
The fading curves from the yellow and red areas of the photograph are shown in the color coded graph below, Figure 3, along with 3 fading references. These standards, Blue Wools 1 through 3, are used in the textile industry to test light fading, and each fades at about half the rate of the previous one. Delta E (DE) is the color science unit of color difference, and you can see it rising slowly for all of the curves with increasing duration of exposure to the xenon light source. The dyes used in Dante’s Inferno exhibit a sensitivity that falls between the fading rates of Blue Wool 2 and Blue Wool 3, earning them a ranking of “fugitive” and suggesting that continual exhibition under normal room lighting conditions would result in a color shift in these areas in less than 20 years. Obviously tight controls on lighting levels and limited exposure periods are warranted to protect this artwork into the future.
Figure 3. MFT fading curve showing color change (Delta E) as a function of exposure time (minutes)
for several red and yellow areas of the Dante’s Inferno and Blue Wool reference standards 1 through 3.
The graph was created with GCI’s SpectralViewer Software.
In other instances, the MFT data suggests some photographs could benefit from occasional exhibition. Dark storage of prints and photographs often results in yellowing of the organic substrates, residual processing chemicals, and media. This yellowing changes the tonal relationship of the highlight areas, which are often just the uncolored white substrate of the photograph, with the rest of the color image. Figure 4 shows a chromogenic print of New York City by Kenji Nakahashi, Good Morning Empire, which was fade tested in the white areas of the reflective puddle of water. The resulting color change graph, Figure 5, shows a sharp decay in the yellowness-blueness parameter, called b* in color science parlance. The small positive b* value indicates a slight yellow tone that falls off smoothly suggesting that the photograph is being photo-bleached to a purer white color. Occasional display of the photograph under normal museum lighting conditions will restore the original white highlights of the image.
Figure 4. Good Morning Empire, 1992.73, created 1987, printed 1991,
Kenji Nakahashi, The Carl H. Lieber Memorial Fund.
The IMLS microfading study is nearing completion and we are learning a lot about individual photographs in the collection. Fortunately there have been no “superfaders” (colorants that would dramatically change appearance over the course of even a single exhibition) discovered so far. However, the results indicate that many of the early color photographic process and even some modern digital printing processes still call for caution in terms of museum lighting and exhibition scheduling. We hope that going forward our photographic images will not burn out, but only fade gracefully. . . albeit very, very slowly.