Powerful stacked organic light-emitting diodes for improved energy output
In the United States, lighting consumes ∼765TWh of electricity each year. This represents 18% of total building energy consumption and 8.3% of total energy use.1 As one consequence, white organic light-emitting diode technology (WOLED) is under development due to its potential to achieve over 10 times the efficacy of incandescent lamps.2 However, for a WOLED to be as operationally long-lived as a fluorescent light and to emit optical power at levels comparable to a 60W incandescent bulb, its substrate area (∼0.10m2) tends to be larger than that of a typical light bulb.
Although a simple method to increase the flux of a WOLED would be to increase input power, efficacy thereby decreases and diode lifetime is inversely proportional to optical output power.3 One approach to creating a powerful WOLED light source on a reasonably small (0.02m2) substrate would seek to improve total power efficacy by enhancing the outcoupling efficiency4 through a reduction in operating voltage5 and elevation of the quantum yield.6 Given the current trend, which has seen an exponential increase in efficacy over the last 10 years, a 100lm/W device might be expected within 5 years.
We report on another solution currently in development that would make small-area WOLED lighting sources more optically powerful. This involves electrically connecting WOLEDs that are vertically stacked away from the substrate. This architecture enables maximum optical output power increases of 2 to 4 times that of a single WOLED.
Technology for the stacked organic light-emitting device (SOLED) was originally developed for full-color displays,7 and is currently being adapted for illumination applications. To fabricate a white SOLED, the first WOLED is grown on a substrate, with a second directly on top of the first. A third can be grown on top of the second, and so on. A key component is the electrical connection between stacked units. Our team has been developing a single metal connector that serves as both a cathode for one device and an anode for another.
To demonstrate this approach, a SOLED that uses a 70nm-thick aluminum floating electrode to serially connect two vertically stacked OLEDs was fabricated and characterized. A green phosphorescent OLED (PHOLED) was first grown on the substrate; then a red PHOLED was grown directly on the Al cathode of the green PHOLED. The top electrode, which is the cathode of the top red OLED, consists of a thin transparent metal layer. As shown in Figure 1, the color emissions do not mix due to the reflectivity of the connecting aluminum electrode.
The bottom green device has characteristics that are identical to a single OLED, while the thin metal electrode limits optical output for the red device to 40% due to transmission loss. Such loss can be eliminated by employing a transparent top cathode such as indium tin oxide. The next step in development is to replace the red and green OLEDs with WOLEDs to produce light sources that resemble the one shown in Figure 2.
The most significant advantage of this SOLED structure is ease of fabrication. There are no doped transport layers, and the connecting electrode is identical to a standard OLED cathode. All organic layers similarly use materials commonly employed in the OLED industry. New deposition systems, or evaporation sources to accommodate materials needed solely for the purpose of electrically connecting stacked OLEDs, are not required. Moreover, operational stabilities of individual SOLED units are expected to be similar to equivalent OLEDs.
To meet future energy demands, solid-state lighting systems containing WOLEDs may help significantly to reduce energy consumption by replacing inefficient incandescent lighting. Our work indicates that the total optical power emitted from these sources can be doubled, an increase in output power that would translate into a reduction of the active area of WOLEDs required to replace 60W bulbs. Our next step will be to develop stacks of three and four WOLEDs to further increase output power.
The authors would like to thank the United States Department of Energy for partial support of this work under grant numbers DE-FG02-03ER83812 and DE-FG02-03ER83813.
