Phosphorescent White OLEDs: Lighting the Way

As the sizes of flat-panel televisions continue to increase, the amount of power required to light them also increases. However, white organic light-emitting-device (WOLED) technology holds the promise of a low-power full-color lighting source for flat-panel TVs, along with other solid-state-lighting applications.

by Brian D'Andrade, J-Y. Tsai, Chun Lin, Peter B. Mackenzie, and Michael S. Weaver

INTEREST in the application of white organic light-emitting-device (WOLED) technology for flat-panel-display (FPD) and general solid-state-lighting applications has been steadily increasing for the past 10 years.1 The drive toward large flat-panel television sets is pushing the power consumption of these devices to new levels. In 2001, it was estimated that color televisions were responsible for 2.9% of the energy used in households, while lighting was responsible for 8.8%.2Therefore, OLED research has significant cross-cutting energy-savings benefits.

Overall, lighting consumes ~765 TWh of electricity each year in the U.S., or nearly 30% of all electricity produced for buildings, which corresponds to 18% of total building energy consumption. In terms of total primary energy consumption, lighting accounts for 8.3% of all the energy used in the U.S., or about 22% of all the electricity produced. The cost for consumers to light their homes, offices, streets, and factories amounts to almost $58 billion a year.3

Given these statistics, it is clear that increasing the efficiency of illumination sources has the potential to generate tremendous savings in both cost and energy use. Incandescent lamps were developed more than 100 years ago and still account for 42% of the electrical energy consumed for illumination. The total power efficiency (ηt) of a typical incandescent light bulb is 12–17 lm/W, whereas OLED laboratory demonstrations already have achieved ηt = 30–65 lm/W,4,5 suggesting that there are considerable advantages to be gained by using OLEDs in this application.

Brian D'Andrade and Michael S. Weaver are with Universal Display Corp., 375 Phillips Blvd., Ewing, NJ 08618; telephone 609/671-0980 x292; e-mail: bdandrade@universal


Universal Display Corp.

Fig. 1: White organic light-emitting-diode (WOLED) panel

Questions of whether WOLED luminance, cost, and reliability can all meet the targets of the lighting and display industries are being actively studied worldwide in both the academic and industrial communities. The answers to these questions are critical because the benefits are significant. For example, the U.S. Department of Energy is anticipating that solid-state lighting in the form of WOLEDs, shown in Fig. 1, and white LEDs will decrease national energy consumption by 29% by 2025. Provided that solid-state-lighting devices achieve their projected price and performance targets, the U.S. could accumulate energy savings of $125 billion from 2005 to 2025, defer the construction of 40 1000-MW power plants, and create solid-state-lighting market revenues of $10-billion/year nationally. Initially, WOLED products may be introduced in niche lighting applications, such as seasonal decorative lights (for example, the star on the top of the tree as shown in Fig. 2). In time, WOLED performance will improve, and these devices may become standard illumination sources.

The development of OLED displays is also supported by the unique features offered by this technology, many of which surpass those of active-matrix liquid-crystal displays (AMLCDs), particularly for mobile applications. Perhaps the most important characteristic is power consumption. Active-matrix OLED (AMOLED) technology is a power-on-demand technology, so a factor of 5 in energy savings can be realized by using OLED displays versus LCDs that require backlights that are constantly on.6 Additionally, displays that employ phosphorescent OLED (PHOLED) technology can consume significantly less power than their fluorescent OLED counterparts because PHOLEDs can have 100% internal quantum efficiency (IQE). As shown in Fig. 3, ηt for a WOLED has steadily increased from less than 1 lm/W to more than 60 lm/W during the past 10 years.

WOLED performance advances coupled with color filters have made WOLED use in large-area displays very attractive due to their manufacturing simplicity. Currently, the viability of WOLED technology has been clearlyestablished by the demonstration of prototypes and by the introduction of products into the marketplace, but further development is required in order to meet manufacturing requirements.

The incorporation of state-of-the-art outcoupling methods and PHOLEDs with 100% IQE should theoretically lead to external quantum efficiencies (EQE) in excess of 30%. For example, by using microlenses, the EQE is 1.8 times greater than that for a standard flat-glass substrate. Other outcoupling enhancement methods reported include resonant cavity, excitation of surface plasmons, and the use of periodic structures placed in the optically active layer to introduce Bragg scattering normal to the substrate plane.

One challenge facing WOLED technology is the small full-width-at-half-maximum (FWHM) emission from organic materials compared to the visible spectrum range. Emission from organic emitters typically only spans about one-third of the visible spectrum. Color-tuning molecules to emit in the blue, green, or red portions of the visible spectrum can be accomplished with a variety of molecular structures and their derivatives, and optimizing OLED structures to efficiently produce primary-colored emission can be very effectively accomplished. However, it requires more time, effort, and infrastructure to incorporate two or three emitters into an optimized WOLED structure to achieve 100% IQE. Although WOLED optimization is more challenging than OLED development, the demand for WOLEDs encourages significant work in this field.


Universal Display Corp.

Fig. 2: The star at the top of this Christmas tree is illuminated by WOLED technology.


Universal Display Corp.

Fig. 3: This chart shows the power efficiency of phosphorescent and fluorescent OLEDs.

WOLED Characteristics

As an example, an all-phosphorescent WOLED without outcoupling enhancement was recently demonstrated7 and achieves 14% EQE (or 25% with outcoupling enhancement), corresponding to 27.5 cd/A at 1000 cd/m2 [see Fig. 4(a)]. The EQE decreased at high luminance, thus EQE = 11% at 5000 cd/m2. The device CIE coordinates negligibly varied from (0.47, 0.45) at 1000 nits to (0.46, 0.46) at 5000 nits; 97% of the WOLED emission power was between 470 and 750 nm, and local maxima in the emission spectrum were at 486, 540, and 606 nm. The operational lifetime to 500 cd/m2 from 1000 cd/m2 was projected to be ~21,000 hours.

The normalized luminance versus time is plotted in Fig. 4(b). There was no initial large drop in luminance during the first several hours of operation, so these WOLEDs were not burned-in before recording the normalized luminance versus time. The longevity of PHOLEDs requires that large (>40 mA/cm2) drive current densities (J) be used to fully degrade devices in a reasonable amount of time. The longest-lived red PHOLEDs are driven at J > 200 mA/cm2. If PHOLED stability continues to increase, drive currents of 1 A/cm2 will potentially be required to age devices. At 1 A/cm2, PHOLED technology may start competing with inorganic LED stability and brightness.

The electroluminescence (EL) spectrum varied slightly with luminance as shown in Fig. 4(c). The height of the green peak at 540 nm increased relative to the blue (486 nm) and red (606 nm) peaks as the luminance was increased from 1000 to 5000 nits; therefore, the CIE abscissa decreased and the ordinate increased by 0.01 over this luminance range. The small changes in CIE were well within the requirements for displays; however, smaller CIE changes are possible for different device structures. The ratio of the three peak heights can be adjusted by varying the concentration of the three phosphorescent emitters and/or varying the thickness of the two emitting layers (EMLs). A CIE of (0.40, 0.47) was calculated to be possible when all three peaks have the same radiance. The EL spectrum varies slightly during operation. At 80% of the initial luminance (L0), the CIE at 10 mA/cm2 was the same as the CIE at L0; however, at 0.5L0 , the CIE at 10 mA/cm2 changed from (0.46, 0.45) to (0.45, 0.46). Again, this change is well within design specifications.

To generate the primary colors for a full-color display using the emission spectra shown in Fig. 4(c), one can conceivably use a short-pass filter that cuts-off above 495 nm to obtain a blue CIE of (0.10, 0.15), a long-pass filter that cuts-on at 595 nm to generate a red CIE of (0.68, 0.32), and a band-pass filter to obtain a desired green CIE. Further development of the WOLED EL spectrum strongly depends on the filters that will be combined with the WOLED.

The device operating voltage is between 8 and 11 V for the range of luminance shown in Figs. 4(a) and 4(c) because resistive electron- and hole-transport layers were employed in the device. Given the state of the art in conductive transport layers,8 one may expect the voltage to be reduced by 2–3 times if highly conductive transport layers are employed in this device. Additionally, the power efficiency can be further increased by a factor of at least 1.5 if the IQE is increased to 100%, where monochromatic PHOLED EQEs have been reported.


Power consumption continues to be an issue as the sizes of flat-panel televisions continue to increase every year. As demonstrated here, WOLEDs with all-phosphorescent emitters are able to achieve a power savings while maintaining proper color performance. Thus, WOLEDs hold great promise as a lighting source for flat-panel televisions as well as other solid-state-lighting applications.


The authors thank the U.S. Department of Energy for their support of this work under SBIR contract DE-FG02-05ER84263.


Universal Display Corp.

Fig. 4: WOLED characteristics: (a) Luminance vs. quantum efficiency (solid curve) and luminance vs. luminance efficacy (4' curve); (b) normalized luminance vs. time; and (c) intensity vs. wavelength.


1B. W. D'Andrade and S. R. Forrest, Adv. Mater. 16, 1585 (2004).

2Energy Information Administration, Office of Energy Markets and End Use, Forms EIA-457A-C, E, and H of the 1997 RECS; Energy Information Administration, Office of Energy Markets and End Use, 1993 and 2001 Residential Energy Consumption Surveys.

3U.S. Department of Energy, "National Lighting Inventory and Energy Consumption Estimate," Vol. 1 (U.S. Government Printing Office, Washington, DC, 2001), page x.

4B. W. D'Andrade, B. Alleyne, M. Hack, R. Hewitt, and J. J. Brown, "White phosphorescent organic light emitting devices for lighting applications," Proc. SPIE 6333 (2006).

5J. Kido, N. Ide, D. Tanaka, Y. Agata, and T. Takeda, "High-Efficiency White OLEDs," presented at the Polymer Processing Society 22nd Annual Meeting, SP7.01, Yamagata, Japan (2006).

6L. F. Weber, "The Race for TVs with Higher Luminous Efficiency," IMID '05 Digest.

7Y. Sun and S. Forrest, J. Appl. Phys. (to be published).

8T. Shiga, H. Fujikawa, and Y. Taga, J. Appl. Phys93, 19-22 (2003).

9J. Feng, T. Okamoto, and S. Kawata. Appl. Phys. Lett. 87, 2411109 (2005).

10J. M. Lupton, B. J. Matterson, I. D. W. Samuel, M. J. Jory, and W. L. Barnes, Appl. Phys. Lett. 77. 3340-3342 (2000).

11B. D'Andrade, M. Weaver, and J. Brown, Proc. 14th IDW (2007).

12T. Canzler, M. Burghart, S. Murano, J. Blochwitz-Nimoth, B. D'Andrade, M. Hack, and J. J. Brown, Proc SPIE 6333 (2006). •