by
Douglas Bugner, Lab Head, Cut Sheet Commercialization Laboratory, Inkjet Printing Systems Division, Research and Development, Eastman Kodak Company

How long your pictures last, whether they be traditional photographic prints or digitally produced inkjet prints, is more than a question of lightfastness or resistance to ultraviolet (UV) radiation. Just as important to overall image stability is the darkfastness, that is, the stability of the print to heat, humidity, and other environmental factors, such as air quality or contact with album materials.

This paper summarizes our current understanding of these different phenomena in the context of inkjet printing. Unless otherwise noted, our primary focus will be the stability and durability of photographic quality inkjet prints, that is, prints made with photo-capable printers using "photo" inks (if available) printed onto glossy photographic quality papers or films, and intended for use in the home.

Lightfastness of Inkjet Prints. It is important to note that lightfastness is a function of both the inks and the paper that are used to make the final print. Further, it is the colorants used in the inks that have the greatest effect on the overall lightfastness of a print.

Inks. The inks used in essentially all consumer desktop inkjet printers that are optimized for photo printing contain water-soluble dyes. A minimum dye set for photo printing includes cyan (C) magenta (M), and yellow (Y). Black (K) dye-based inks are also included in most photo ink sets. In the highest quality printers, such as the Kodak Personal Picture Maker 200, six inks are used: light cyan (c) and light magenta (m) inks (same dyes, less concentrated) are used in addition to the full strength CMYK dye set to produce a smoother tone scale.

Things would be easier if all that was needed from a dye set was lightfastness. In addition to lightfastness, however, color quality is a primary attribute of the dye set. In many cases, ink manufacturers must make a trade-off between lightfastness and color quality. At Kodak, we believe that image and color quality should not be compromised. That is why we have taken great pains to develop a proprietary dye set for the Kodak Personal Picture Makers 120 and 200 that offers unsurpassed color quality and industry-leading light stability.

Papers. Papers (or films) used for photographic quality inkjet printing have a special ink-receptive coating on at least one surface. This coating is designed to absorb the inks and to optimize both the image quality and image stability of the print. Again, trade-offs are often made among the various performance attributes that the overall print must satisfy. For example, polymers and other additives that are added to the coating to enhance dry time, gloss, and/or waterfastness can often adversely affect the light stability of the dye set to varying degrees.

In addition to the lightfastness of the image itself, the overall lightfastness of the materials used in the paper substrate and the ink-receptive coating is also critical. Cheaper, non-photographic paper bases are prone to yellowing, as are some of the components that have been used in the ink-receptive coatings of some non-Kodak products. This yellowing in turn leads to objectionable color shifts in the image, even if the dyes themselves are unfaded. The resin-coated papers used by Kodak for inkjet printing are the same paper substrates that have been developed over many years for traditional standard photo prints. They are designed to withstand many generations of light exposure without yellowing or degrading.

On the other hand, for most desktop office printers, plain paper compatibility is understandably a higher priority than lightfastness. As such, the inks used for printers that are designed for use in the office typically contain dyes and other ingredients that compromise lightfastness at the expense of producing bright, colorful graphics on plain paper.

Display Conditions. Answering the question "How long will my inkjet prints last?" necessitates knowing the conditions for displaying the prints. The type of light source and duration of the exposure are key factors is how fast a given print will fade. Light sources rich in the higher energy ultraviolet end of the spectrum, such as sunlight through a window or unfiltered fluorescent lights, tend to accelerate print fade relative to incandescent sources. To complicate matters, it is not unlikely that a print will experience more than one type of light exposure in the home. For optimum print life, we recommend that a print be displayed out of direct sunlight. Light intensities typically found in the home vary both by location and time of day between about 0-250 lux, with 120 lux being a good overall estimate for typical home conditions. This estimate appears to be reasonable for both print viewing and print life.

A further complication is the fact that environmental factors such as temperature, humidity, and air quality can all affect image stability, both independently and in combination with light exposure. With certain ink and paper combinations, these environmental factors can even dominate and/or greatly accelerate light fade. A recommended range of temperature and humidity for print display is 68-77^oF (20-25^oC) and 40-60% relative humidity. Studies have shown that these ranges of temperature and humidity are commonly found in an average home. Although air quality is not usually a concern in the home, certain combinations of ink and paper have been recently flagged as being particularly susceptible to premature fade due to airborne pollutants.

Relatively Speaking. When prints are carefully prepared and displayed under optimum conditions of temperature, humidity, and air quality, light fade can become the primary limiter of print life. Although it may be difficult to precisely answer the question of how long a specific print will last, accelerated light fade tests allow us to make direct comparisons between prints under a set of controlled conditions of exposure and environment. However, the use of accelerated conditions to estimate print life for the same prints under real world conditions is extremely risky. This is owing primarily to a phenomenon known as reciprocity failure. Stated simply, reciprocity is the property of a light-induced change, e.g., print fade, such that an equivalent cumulative exposure produces a constant change independent of the intensity of the exposure. Reciprocity failure, therefore, occurs when prints fade differently under high intensity, accelerated conditions, than under low intensity real world conditions.

Standardized Methods. Even when doing relative comparisons of light fade, employinga standardized methodology is important. Perhaps the most widely cited standard for measuring light fade is ANSI IT9.9 (last updated in 1996). This standard specifies the procedures and test equipment needed to characterize the density loss from a test target that has been treated under one of several recommended exposure conditions. The exposure condition that comes closest to modeling the home environment specifies the use of glass-filtered, "cool white" fluorescent illumination at approximately 50 times the intensity of the typical home. Temperature and humidity are specified at 22-26^oC and 45-55% RH.

This standard also offers guidelines with respect to image-life parameters, that is., at what level of fade is the print judged to be at the end of its useful life. These recommendations are offered as guidelines only, and the user of the standard should specify end-points that are appropriate to the specific product and application. To estimate a useful print life, the user of the standard should specify the duration and intensity of the real world exposure condition to which the accelerated conditions are being extrapolated, assuming that the law of reciprocity is valid, which as noted may not be a good assumption.

Recent studies on inkjet photographs have shown that glass-filtered fluorescent lights may contain a higher level of ultraviolet light than is typically found in the home or office. Plexiglas-filtered fluorescent lights were found to more closely approximate the home and office environments. The same studies demonstrated that inkjet prints experience significant and highly variable reciprocity failure, depending upon the specific combination of ink and paper being tested. For these reasons, we believe that accelerated fade tests should be run under Plexiglas-filtered fluorescent lights at several intensities to better quantify the magnitude of the deviation from reciprocity. When photographs printed on the Kodak PPM 120 and PPM 200 were tested at two different intensities (67 klux and 5.4 klux), the deviation from reciprocity over this wide intensity range was found to be relatively minor and similar to traditional silver halide prints over this same range. Tests run under lower intensity conditions are ongoing, but after six months of continuous exposure at 450 lux, the rate of fade is identical to that predicted by the test done at 5.4 klux. (Note: Testing at typical home display light levels of 120 lux was simply not possible owing to the time such tests would require in relation to Kodak’s rapid product development cycles.)

Summary
To summarize, the dyes used in the inks have the greatest impact on the overall lightfastness of an inkjet print. Photo quality receivers that employ true photographic paper bases are also critical to avoid unwanted light-induced yellowing of the print. Because of the many trade-offs involved in optimizing a system, matched systems that are designed for only printing photos tend to yield the best lightfastness. Complicating matters is the fact that the nature and duration of the light exposure, in combination with the specific environmental conditions under which a print is displayed, are all-important determinants of the life of a print. A further complication to the use of high-intensity accelerated fade conditions for estimating useful print life is the reciprocity failure for inkjet prints can be significant and unpredictable.

Glossary
Intensity — a measure of the strength of the light under which the print is being displayed, usually expressed in units of klux (thousand of lux).

Cumulative exposure — the product of exposure intensity and the duration of the exposure, express in units of klux-yrs.

Reciprocity — the property of a light-induced change such that an equivalent cumulative exposure produces a constant change independent of the intensity of the exposure. For example, an exposure of 10 klux for 1 year should produce the same amount of fade as an exposure of 1 klux for 10 years (cumulative exposure = 10 klux-yrs in each case). Any deviation in the amount of change that occurs between two equivalent cumulative exposures is termed reciprocity failure.

End-point criteria — the status A density loss from a starting density of 1.0 above D_min at which a print is defined as being unacceptably faded.

Standard day — the average cumulative exposure that a home or office environment receives per 24-hour period.

Print-life estimate — assuming no reciprocity failure, the extrapolation of the cumulative exposure at the first end-point criterion reached under an accelerated condition to the length of time in years that it would take to reach the same cumulative exposure under standard conditions in the home or office.

ANSI Standard IT9.9 — last updated in 1996, the most widely used standard for determining the stability of photographic prints. This standard does not mandate specific end-point criteria, nor does it define a standard day; rather it provides guidelines and requires that any print-life estimates that claim to follow the standard clearly state the end-point criteria and standard day assumptions that were used in the projections.

Endnotes
¹ S. I. Anderson and R. J. Anderson, J. Imaging, Tech., 17: 127-132 (1991).

² S. Anderson and G. Larson, J. Imaging Tech., 13: 49-54 (1987).

³ http://www.epson.com/whatsnew/ygtsi/lightfast.html

⁴ R. E. McComb, Photo Trade News, February, 1998.

⁵ D. E. Bugner and C. Suminski, Proceedings of IS&T's NIP16: International Conference on Digital Printing Technologies, Vancouver, B.C., October 15-20, 2000.

⁶ M. J. Carmody, S. Evans, and S. Robinson, Proceedings of IS&T's NIP16: International Conference on Digital Printing Technologies, Vancouver, B.C., October 15-20, 2000.

| Back |