Panel III:
Organic Light Emitting Diodes


Patrick Windham

Windham Consulting

Mr. Windham opened by suggesting that organic light emitting diodes (OLEDs) had an “enormous potential to help replace area lighting, and to do so at a low cost.” He mentioned the likelihood that large-area OLEDs will some day be made cheaply by roll-to-roll processing, like newsprint. Like other areas of technology, he said, there are technological challenges to overcome before this can happen, and therefore opportunities for universities, national labs, and industries to work together to improve the technology and make it more cost effective.


Mark Thompson

University of Southern California

Dr. Thompson began by showing the audience a box of red, green, and blue OLEDs. (Unlike the inorganic LED demonstrated earlier, the OLEDs were safe to examine closely.)

The Advantages of OLEDs
The Simplicity of OLEDs

He said that OLEDs got their start in display applications, where they are not only safe but have many benefits that also apply to lighting applications. The first

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement

Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 66
Partnership for Solid-State Lighting: Report of a Workshop Panel III: Organic Light Emitting Diodes INTRODUCTION Patrick Windham Windham Consulting Mr. Windham opened by suggesting that organic light emitting diodes (OLEDs) had an “enormous potential to help replace area lighting, and to do so at a low cost.” He mentioned the likelihood that large-area OLEDs will some day be made cheaply by roll-to-roll processing, like newsprint. Like other areas of technology, he said, there are technological challenges to overcome before this can happen, and therefore opportunities for universities, national labs, and industries to work together to improve the technology and make it more cost effective. AN INTRODUCTION TO OLEDs Mark Thompson University of Southern California Dr. Thompson began by showing the audience a box of red, green, and blue OLEDs. (Unlike the inorganic LED demonstrated earlier, the OLEDs were safe to examine closely.) The Advantages of OLEDs The Simplicity of OLEDs He said that OLEDs got their start in display applications, where they are not only safe but have many benefits that also apply to lighting applications. The first

OCR for page 66
Partnership for Solid-State Lighting: Report of a Workshop benefit is that the device structures themselves are simple. They consist of a transparent conductor on which lies some number of organic layers, usually four. The total thickness of the organic layer is only about 2,000 angstroms, and that layer is capped with a thin metal cathode. The Advantage of Flexibility Another benefit is that all the organic layers are amorphous and therefore bendable. This differentiates them from semiconductors, which are formed by the crystal-growing process of epitaxy2 and are therefore rigid. OLEDs are so flexible that they can be bent around any reasonable radius and grown as a passive matrix display on virtually any medium, such as plastic sheets or rolls. They can be applied easily to ceilings, walls, or other large surfaces or even embedded in fabrics or other soft elements. They can also be located on or in firm surfaces such as glass, metal, or silicon. Straightforward Processing Because organic LEDs are amorphous, the methods for preparing them are relatively straightforward, inexpensive, and can be scaled to large areas. All of the deposition tools for large-scale processing are commercially available. Finally, the lights can be readily tuned in terms of color and electronic properties using chemical means. There are two types of materials used for making OLEDs: small molecules and polymeric thin films. Electronically the two materials are very similar, although there are some differences in preparation techniques. For small molecules, vacuum deposition is used for both organic thin films and the metal electrodes; the organic compounds are deposited from a heated source directly onto the substrate in a vacuum. Another technique is organic vapor phase deposition, which gives good control over film thickness and composition. For polymers, solution processing (spin coating) is the most common technique. Another technique that Dr. Thompson’s laboratory has been developing involves not a vacuum but vapor deposition. This type of process is much more convenient than standard deposition, although the quality of the polymer layers is typically low. The Advantage of Transparency Another benefit of these devices is that they are virtually transparent to their own radiation, because the organic layers themselves are very thin. If the cathode 2   Epitaxy is the growth of the crystals of one mineral on the crystal face of another mineral, so that the crystalline substrates of both minerals have the same structural orientation. Semiconductors are grown from such solid crystalline substances as silicon and germanium.

OCR for page 66
Partnership for Solid-State Lighting: Report of a Workshop is thin enough, one can make devices that are 75 percent or more transparent. By replacing the metal electrode with an electron injection layer, devices have been made that are 95-percent transparent. Because the OLEDs themselves are transparent, different-colored OLEDs can be stacked to combine colors; placing green on top of blue on top of red can create an OLED “sandwich” that gives white light when all three of the devices are turned on. In addition, transparency allows OLEDs to be overlaid on windshields or other transparent substrates where the light can be turned on as needed; when it is off, the view is clear. Transparency is also valuable in terms of function. Dr. Thompson’s laboratory has demonstrated approximately 80-percent efficiency for monochromatic lights. In certain applications, such as games, they have achieved external luminosities as high as 60 lumens per watt for monochromatic green. Lamp lifetimes of 10,000-hour lifetimes are becoming common, and ultimate lifetimes of approximately 100,000 hours (~12 years) are anticipated. Some devices are very bright, with demonstrations as high as a million candelas per square meter, compared with 100 candelas per square meter for a cathode ray tube and 800 candelas for a fluorescent panel. Turn-on voltages as low as 3 volts have been demonstrated. OLEDs are Specialized No single device has all of those desirable characteristics; it is unlikely that any single device has even two of them. More typically, devices are designed around particular parameters, reflecting decisions about whether long-lifetime, high-efficiency brightness or some other feature is primary. There are different choices of structure and materials and within the device itself one can use various electrodes, transporting layers, and emissive layers. For the application of lighting, efficiency is paramount, and other qualities may have to be sacrificed. OLED Function Dr. Thompson offered a simple description of OLED function, displaying a device package with a two-layer structure. When a fixed voltage is applied, electrons pile up on one side of the organic interface, electron holes3 on the other. The recombination of electrons and electron holes at the interface leads to a formation of excitons, or excited states, consisting of bound electrons and holes. When the exciton energy is transferred from the host material to the dopant, the dopant emits radiation in the decay process to the ground state.4 3   In physics a hole (or electron hole) is a vacant position in a substance left by the absence of an electron, especially a position in a semiconductor that acts as a carrier of positive electric charge. 4   A dopant is a substance, such as phosphorus, added in very small amounts to a semiconductor or an OLED to improve the quantum efficiency of the material.

OCR for page 66
Partnership for Solid-State Lighting: Report of a Workshop The Complication of Excitons At the same time, this light-emitting system encounters another problem. Unlike inorganic devices, OLEDs form two different types of excitons. About 75 percent of excitons form in a triplet-excited state (i.e., three bound excitons), which produces phosphorescence, and about 25 percent occur in a singlet-excited state, which produces fluorescence. Phosphorescence (emission from triplets) is a process that is forbidden in OLEDs by physical laws and thus is typically inefficient. Another way to characterize this is to say that singlet-state excitons quickly return to the ground, or non-excited, state when voltage is removed; the triplet excited state is “forbidden” to return to the ground state, continuing to emit light. This long lifetime of organic phosphors is why a glow-in-the-dark Frisbee can keep glowing for many minutes. While long-glowing phosphors are perfect for nighttime games, they are unacceptable in lighting devices. Most OLEDs produce light by fluorescence (singlets only), thus wasting the majority of the excitons. Physicists began to tackle this problem through a growing understanding of the photophysics of inorganic compounds. They found that by incorporating such heavy metals as iridium, platinum, and gold, they could get very strong “state mixing” that combines the relaxation of the singlet state with the forbidden relaxation of the triplet state. This dopant-induced state mixing produced very high efficiencies for the relaxations, which largely solved the problem of the triplets. The first proof of this solution came in the late 1990s, when investigators using iridium obtained quantum efficiencies up to around 9 percent and very good efficiencies out to the fairly high brightness of 1,000 candelas per square meter. By further optimizing the matrix into which the dopants are placed, the external quantum efficiency has been raised to 15.5 percent, which is achieved by an internal quantum efficiency of 80 percent. Luminous efficiencies are on the order of 40 lumens per watt. Glass Mesas for Better Outcoupling Another issue in the development of OLEDs is internal versus external light, which is also an issue for inorganic LEDs. In a typical organic LED only about a fifth of the light can be directed in the forward direction; four-fifths of the light comes out the sides of the substrate. By extensive experiments in patterning or shaping the substrate, Dr. Thompson’s group has shown that they can outcouple a significantly higher fraction of light. They have now raised the outcoupling efficiency from about 20 percent to almost 50 percent, which increases the external efficiency from 15 percent to about 30 percent. They have arrived at fairly complicated substrate shapes called glass mesas. These mesas do not have to be on the front side of the substrate; they work equally well on the backside. Careful work here is necessary for lighting applications, which require

OCR for page 66
Partnership for Solid-State Lighting: Report of a Workshop every possible photon from a device for acceptable outcoupling efficiency. To date, even the best substrate designs have not yielded high enough outcoupling for lighting applications. A Chemical Exercise In many ways, designing different OLEDs is more of a chemical than an engineering exercise. Different devices are made by choosing dopants that create green, red, or yellow light, rather than by optimizing physical parts of the system. An early question was whether the different chemistries would all work to make reasonably effective devices. It turns out that they will, although light output does vary by color. For green, applying one milliamp of current per square centimeter gives about 440 candelas of light per square meter, or about 18 lumens per watt. Output for yellow is not quite as good as for green because of the overlapping of photopic response. For the same reason, output for red also drops to 2.2 to 2.5 lumens per watt. Needed: More Brightness In addition, said Dr. Thompson, all of the phosphor or dopants have relatively short lifetimes. They are relatively dim, but no intrinsic decay mechanism has been seen during the first 3,000 hours of operation. An iridium-based system gives a reasonable brightness on the order of 300 candelas and an extrapolated lifetime of about 20,000 hours. While such systems are reasonably well suited for displays, they are about an order of magnitude from the output that will be needed for lighting applications, such as large panel lights. Mixing Colors The blending of colors in OLEDs turns out to be manageable, with a number of strategies for mixing colors to produce white light. Dyes can be mixed in the emissive layer of the OLED to convert, in specific contexts, other colors to white, without showing significant changes in color. Individually addressable R-G-B components (side-by-side or stacked) will allow users to set their own color balance for soft or hard lighting. Another approach avoids having to mix colors in a single device. R-G-B pixels can be arranged side-by-side or stacked, similar to the techniques used to make flat-panel displays. A benefit of this technique is that users can change the color balance themselves. For example, if one color ages more than the others over time, they can correct the balance as necessary. For transparent devices, colors can be combined by stacking pixels one on top of the other. A simpler system would stack large sheets of transparent R-G-B to produce white.

OCR for page 66
Partnership for Solid-State Lighting: Report of a Workshop The Advantage of Synergy Dr. Thompson pointed out that his chemistry group in California worked in close collaboration with electrical engineers in New Jersey at a small, aggressive company. “We need to have this kind of synergy in all of this work,” he said. “The best way to move forward is to build as many teams like this as we can that can work on several fronts simultaneously in an integrated fashion.” Poor Packaging He concluded by saying that while the available options for OLED substrates were becoming clear, the same is not true for the packaging of OLEDs. This continues to present a major challenge, just as it does with inorganic LEDs. Engineers still have not found good ways to achieve a hermetic seal for the substrate or a way to package the substrate itself. OLEDs FOR GENERAL ILLUMINATION Steve Duclos General Electric Corporate Research and Development A Vision for OLEDs Dr. Duclos began by offering his vision of “where we see OLED technology taking us” in terms of the benefits, key markets, and R&D challenges. He predicted that OLED technology would lead to the use of large-area, white-light, flat-panel emitting devices. These devices might have one, two, or three layers of organic materials between conducting contacts. The organics could emit in the red, green, and blue, which combine to produce white. Manufacturing Goals For manufacturing, the goal will be to mount them on a flexible substrate or film that will enable the use of efficient roll-to-roll processing. The organic materials, because they are amorphous or disordered and not crystalline, can be printed or “splatted” onto the substrate using known methods. These advantages suggest a route to efficient solid-state lighting at inherently low cost. Large-Area Applications and Energy Savings OLEDs differ in applications from inorganic LEDs. Inorganic LEDs are best suited to high-brightness point sources of light, such as spotlights, traffic lights, light filaments, and projection lamps. OLEDs will have large-area, diffuse appli-

OCR for page 66
Partnership for Solid-State Lighting: Report of a Workshop cations, such as backlight, signage, and, most important, light panels for general illumination that can replace fluorescent lamps in ceilings. This is significant for energy consumption because fluorescent lamps are the largest user of electricity in the lighting sector, which is dominated by the commercial and industrial sectors. Based on fluorescent lamp distribution, it is estimated that OLEDs could penetrate 50 to 70 percent of these sectors and that degree of penetration would save more than 1 quad of energy in the United States. With such savings come environmental benefits: 16 million fewer metric tons of carbon dioxide emissions and the elimination of some of the mercury (used in fluorescent lamps) that contaminates landfills. The commercial and industrial sectors understand the economics of lighting and are likely to respond to such potential savings. If OLEDs reach a light output of 120 lumens per watt, there is likely to be a huge market for them. Some Technical Challenges Before OLEDs are ready for this market they must satisfy the technical requirements that will bring the desired color, efficiency, brightness, lifetime, and cost. Color. Correct white color is critical to market acceptance. This includes color temperature, Color Rendering Index, color uniformity, and color maintenance over the lifetime of the device. White light can be generated by multiple methods, and it is still not clear which will be most effective. Efficiency. Efficiency is measured in two ways: lumens per watt and external efficiency. Incandescent bulbs achieve about 13 to 15 lumens per watt (lm/W) and fluorescent bulbs achieve about 100 lm/W. However, fluorescents must be placed in a fixture where they lose some light to reflection, reducing efficiency to about 70 lm/W. OLEDs require no fixtures. The goal of OLEDs is 120 lm/W, and to achieve this in acceptable white light they must have an external efficiency (optical watts divided by electrical watts) of 40 percent. Today the efficiency range is 1 to 4 percent for blue, green, and red. An order-of-magnitude increase is needed, which is comparable to the improvement of the last ten years. Brightness and lifetime. These two qualities are linked: a brighter light usually has a shorter lifetime. Commercial and industrial customers demand lifetimes of at least 20,000 hours, during which the light has to retain a constant color. OLEDs today last about 1,000 hours, so an order-of-magnitude increase is needed here, too. This must be done while solving the problem of differential decay. Cost. Commercial and industrial customers demand a two-year payback on energy-saving devices. To get there, OLEDs must have an efficiency of 120 lm/W and a cost of $6/kilolumen. Given the economies of roll-to-

OCR for page 66
Partnership for Solid-State Lighting: Report of a Workshop roll manufacturing, they appear to have the capability of meeting this requirement. Such economics are persuasive for large customers. OLEDs would provide a savings of about $35 over the lifetime of each typical fluorescent fixture, which holds 3 to 4 bulbs. For large retailers, which can use nearly 5 million such fixtures, total annual savings would be roughly $30 to 40 million. Hard Work Ahead in R&D Dr. Duclos concluded by listing a series of R&D challenges that will be needed to raise the efficiency of OLEDs to 120 lm/W, including materials, device design, and large-area processing. Material. More stability is needed in encapsulation materials, along with better understanding of how to limit major side reactions caused by permeation of air and water into these materials. Fundamental understandings of some of the limiting efficiencies of materials are also needed. Design. Engineers need to know much more about internal and external quantum efficiencies, how to extract light and reduce light trapping, deposition techniques, and surface texturing. They also need better electrode materials, better transparent materials, and higher conductivity materials. Processing. Although the potential for bulk roll-to-roll processing is apparent, many engineering questions must be answered, especially how to keep air and moisture out of the device. He suggested that to accelerate the development of OLEDs, improvements in light efficiency and lifetimes must be improved by approximately an order of magnitude and cost must be lowered substantially. Addressing the three primary R&D challenges (materials, design, and processing) would have to be accomplished in parallel. This, he said, would require the collaboration of industry, academia, and the national labs. If an effective collaboration could be devised, he said, OLEDs have a “real potential for being a very low-cost, high-efficiency replacement for fluorescent lamps” while bringing significant energy savings and environmental benefits. CRITICAL R&D CHALLENGES Steve Van Slyke Eastman Kodak Company Mr. Van Slyke began by describing OLEDs as multilayer structures for which each layer has a particular function. An entire OLED film is only 200 nanometers

OCR for page 66
Partnership for Solid-State Lighting: Report of a Workshop or so thick, consisting of anode, hole-injection, hole-transport, emissive, electron-transport, and cathode layers. Light is emitted when a current is passed through the layers; the emissive layer can be modified to give red, white, or blue, for example. Current Applications of OLEDs Many companies and universities around the world are working with OLED technologies (see Table 1). Kodak holds over 60 OLED patents and has licensed them broadly. The first company to bring the technology to market was Pioneer, of Japan, making passive matrix displays for such applications as after-market car stereos and CD players. Recently the company brought out larger car stereo displays with a wide-viewing angle, higher-power efficiency, and compatibility with existing electronics. Motorola incorporated a Pioneer OLED display in its cell phones. Another company, TDK, is selling an OLED display to Alpine for car stereos. This display is essentially a white emitter with color filters. Kodak has collaborated with Sanyo to produce a passive matrix for cell phones; its emissions can be patterned with various colors that are “pleasing to the eye and easy to see.” Recently Kodak demonstrated full-color active matrix displays using a technique similar to that for liquid crystal displays; the substrate was made of thinfilm transistors and each pixel had an R-G-B emitter associated with it. TABLE 1 Companies Involved in OLED Activities Europe North America Asia Avecia Agilent Lexell Canon Samsung Aventis Alien 3M Denso Sanyo CDT Dow Siemens Idemsu Seiko-Epson Covion Du Pont Three-Five LG Sony Opsys Kodak Uniax Mitsubishi Stanley Electric IBM eMagin UDC NEC Sumitomo Philips IBM Xerox Nippon Seiki TDK   Lucent Pioneer Toshiba     Ritek   Difficult Challenges Color Tuning OLEDs have the potential of covering large areas; in a typical office half the ceiling might be covered with a white-light OLED. Kodak has experimented with changing emission colors by adjusting the composition of a single layer of

OCR for page 66
Partnership for Solid-State Lighting: Report of a Workshop the OLED. For yellow and blue emitters, for example, the concentration of the two colors can be tuned to change from blue to reddish yellow, with white in the middle. While this device is effective, it is not yet efficient. He said that simultaneously developing all of these key qualities would require the well-planned collaboration of industry, academic, and government-lab researchers. The Degradation of Light Quality Another challenge is to halt the degradation of light quality over the operational lifetime of the OLED. The typical white OLED, operating at an initial power of 500 candelas per square meter, suffers a “steady, monotonous, irritating degradation” that eventually causes lighting efficiency to decline by about half. The display industry could tolerate such a degree of degradation, but the lighting industry could accept a degradation of less than 5 percent over a lifetime of about 10,000 hours. It has proven difficult to understand the causes of instability, and considerable R&D is needed to elucidate the mechanisms. The Drive for 100 Lumens Improving efficiency may prove to be less problematic, although the improvements needed are indeed large. Fluorescent bulbs produce about 100 lm/W (before losses to the fixture); by comparison, current white OLEDs achieve only 3 to 5 lm/ W, indicating the need for a 20- to 30-fold improvement. The “carrot” in this pursuit is that green OLEDs already yield more than 30 lm/W. “My opinion,” said Mr. Van Slyke, “is that you’re going to get to 100 lumens. It just depends on how much time and money goes into the effort.” He added that it is very difficult to extract enough light out of a device. Only about 20 percent of the light generated in an OLED display emits from the front; the rest leaks out the edge of the display. Solving this challenge will require a great deal of device engineering. Encapsulation Is Inadequate Another R&D challenge for OLEDs (which mirrors that of inorganic LEDs) is to develop better encapsulation techniques. All the devices now on the market and being tested have an overly complex encapsulation system. With organic layers deposited on a glass substrate, the anode or the cathode layer is a very reactive metal, requiring a stainless steel “can,” a glued perimeter seal, and a desiccant inside the seal to capture any invading moisture. This system is too complex and expensive for the competitive lighting industry. The next-generation device is entirely thin-film encapsulation on a plastic substrate. This system places great demands on the film substrate, and new encapsulation methods must be developed to prevent any moisture from entering and reacting with the cathode metal.

OCR for page 66
Partnership for Solid-State Lighting: Report of a Workshop Benefits of a Consortium He concluded by describing some of the benefits of a government-supported collaboration. One is its ability to marshal diverse expertise in different locations. For example, the national labs have broad expertise in lighting design and lighting engineering, with specific expertise at Pacific Northwest National Laboratory in encapsulation methods that use thin-film technology. For materials and device research, much work is being done by industry and academia; manufacturing technology requires collaboration between industry and the national labs to drive the cost of fabrication as low as possible. DISCUSSION Current Federal Support In response to a question about the level of federal effort, Mr. Van Slyke said that DARPA (the Advanced Research Projects Agency of the Defense Department) was spending $10-12 million per year on LED research, and that the Department of Energy supports a program at General Electric on OLEDs for lighting. The Degradation Process Another questioner asked about the nature of the degradation mechanism that shortens the lifetime of OLEDs. Mr. Van Slyke said that one cause is the presence of impurities. Researchers have gradually learned that “purity” in the sense used by organic chemists is not enough; OLEDs need the degree of purity used in the electronics industry. Another mechanism may have to do with hole injection, which is a positive charge injection into one of the layers that creates a cation. Finally, the cathode itself degrades. Dr. Duclos said that the material with the shortest lifetime is the blue and speculated that this may be because the light is emitted close to the band edge of the materials. Dr. Thompson emphasized that addressing degradation is difficult because the matrix of OLEDs is inherently complex. In any device, there are different materials that all have to be optimized for both lifetime and efficiency. “We don’t know enough about the degradation mechanisms to know which things we need to avoid,” he said. “Even worse, we don’t know what combinations of things we have to avoid. One material may function well in one device, but when it is coupled with another (electron transporter, for example), you may get degradation. It’s a complicated issue.” Another questioner raised the issue of dark spots, which Mr. Van Slyke said are caused when moisture leaks into a device. He reiterated the need for better encapsulation. Referring again to the work of Pacific Northwest National Laboratory, Dr. Thompson said that researchers there had made OLED films with per-

OCR for page 66
Partnership for Solid-State Lighting: Report of a Workshop meability so low they could not measure it, and even those films could not keep moisture out. “The OLED itself,” he said, “is actually a much better permeability test than anything they have for measuring permeability.” Private-Sector Investment in LEDs Dr. Wessner asked how much is being invested in research by the firms represented at the workshop, by the U.S. government, and by other countries. Mr. Van Slyke said that Kodak had no allocation for OLED lighting and that all of its R&D is directed at display applications. This amounts to some 50 to 60 people, with a budget of over $10 million a year, “still a small amount.” Dr. Duclos said that General Electric has a three-year contract with the Department of Energy to demonstrate OLED lighting with higher efficiency than incandescent bulbs. A Manufacturing Strategy Dr. Duclos commented on work being done abroad, citing intense activity in Asia over the last three years. He said that Japanese firms had applied for some 8,000 patents related to OLEDs with the expectation that OLED displays would capture a significant fraction of the $50 billion display industry in the next 5 to 10 years. He said that one way to maintain leadership in OLEDs is to push the development of roll-to-roll manufacturing and flexible substrates, where the United States is the leader. He called for a government partnership in this area. Dr. Thompson added that the cost of substrates must come down before the industry can make an economic argument for OLED lighting. Intense Activity in Asia Dr. Bergh expanded on the subject of competitiveness. He agreed that in the last two years Japanese firms had issued some 9,000 patents in OLEDs. He cautioned that this figure could not be compared one to one with American or European patents, because the Japanese patent system permits only one claim per patent. To compare them, one has to divide the Japanese number by approximately three. Nonetheless, this figure is far higher than the number of patents obtained by European (400) and American (500) firms in the last year. The Need for an Outside Force Dr. Bergh also emphasized the importance of the differing requirements of the lighting and display sectors. Research on the two sectors can be combined only if an outside force guides the industry toward broader objectives. Industry will not voluntarily commit itself to lighting research if reliability and efficiency still have to be improved by more than an order of magnitude. Industry can only

OCR for page 66
Partnership for Solid-State Lighting: Report of a Workshop move toward general lighting if the government steps in, coordinates the forces in the industry, and leads it to that goal. Lighting requires research programs that are longer term than industry can afford. Dr. Thompson agreed that the industry is doing the best it can, given the technical challenges. He added that display and lighting applications differ in costs. While the display industry wants lower costs, the lighting industry absolutely has to have them. OLEDs versus Solar Panels A questioner asked whether the lighting industry use available thin-film solar cell technology, especially for OLEDs. Dr. Thompson agreed that the best OLEDs are based on the same concepts as thin film solar cells, but he noted an important difference. If an OLED loses 5 percent of its efficiency to creeping decay and dark spot formation, it is ruined; if a solar cell loses 5 percent if its efficiency, it simply produces 5 percent less electricity. The Argument for a Government Role Mr. Trimble commented that if the United States wanted to create the success for OLEDs that it achieved for the semiconductor and computer industries, the government must play a similar role; that is, the government must generate a market for panels by announcing large procurement goals for lighting panels, say, a million square feet. This, he said, would do more than anything else to focus attention on this valuable technology. Dr. Duclos agreed on the need for a government-led effort like SEMATECH, as well as more infrastructure research in the national labs and basic research by academic scientists and engineers. He noted that General Electric is working with the help of its Department of Energy grant to produce a 2-foot-square OLED panel on glass with the same efficiency as incandescent light. He expressed gratitude to the Department of Energy for its support. Dr. Chipalkatti and Mr. van Slyke agreed that even the largest private companies cannot afford all the research necessary to bring OLED lighting to market quickly, and that a “lighting SEMATECH” is required.