Enlightening Technologies

Lighting is an energy hog. It accounts for nearly 20% of global electricity consumption, according to the International Energy Agency. Solid-state lighting technologies such as light-emitting diodes (LEDs) and organic LEDs (OLEDs), which are much more efficient than conventional incandescent light bulbs, have the potential to slash that electricity consumption.

But LEDs and OLEDs are still a long way from their maximum efficiency. In a symposium sponsored by the Division of Physical Chemistry at last month’s American Chemical Society national meeting in Indianapolis, researchers described efforts to improve the efficiency of LEDs and OLEDs for lighting applications. Some of these methods involve nanowires; others manipulate light-emitting features of organic materials. The technologies have come a long way, researchers said, but there’s plenty of room for improvement.

The characteristics of LEDs and OLEDs make them complementary, said Stephen R. Forrest, vice president for research and professor of electrical engineering, physics, and materials science and engineering at the University of Michigan. “LEDs and OLEDs form a perfect marriage,” he said. “Both are needed.”

For example, LEDs are intense, focused light sources, whereas OLEDs are more diffuse sources and can be incorporated into more-flexible substrate materials. “Architects love OLEDs because they can design all kinds of lighting that is attractive for interiors,” Forrest said. “If you think about the old incandescent bulb, you stick it in a socket and then you have to put a lampshade around it. The lampshade cuts out all kinds of light. With an OLED, you could make the lampshade itself a light bulb.”

Before that lampshade can go mainstream, however, the efficiency of LEDs and OLEDs must be further improved. In their quest for efficiency, researchers are experimenting with new nanostructured architectures for making LEDs and on better ways to extract light out of OLED devices. Other researchers are showing how organic materials that emit light through a mechanism different from that of typical OLEDs can also be used to make efficient devices.

Solid-state lighting systems can be classified by their light-emitting material. LEDs are typically made from the inorganic semiconductor gallium nitride, whereas OLEDs are made from a variety of carbon-rich ­compounds.

Although the underlying mechanisms differ, researchers use the same terminology to describe how LEDs and OLEDs work. Electrons and positively charged holes are injected from opposite sides of each device by applying a voltage. The electrons and holes travel through charge transport layers to a light-emitting active region, where they recombine. This recombination results in the emission of a photon. The wavelength of that photon depends on the material and the electronic levels involved. But the similarities end there.

To make LEDs, crystalline gallium nitride is generally deposited onto a substrate such as sapphire. The gallium nitride contains layers that are doped to make them effective at transporting either electrons or holes. These layers sandwich an emissive region made of an alloy of gallium nitride and indium nitride.

“Current LEDs are fabricated by depositing layers of materials over a flat substrate,” said P. Daniel Dapkus, director of the Center for Energy Nanoscience at the University of Southern California. “The nature of the crystal that actually grows is governed by the particular substrate on which you grow these layers.”

It turns out that sapphire substrates have only one crystal face that’s good for growing LEDs. The crystal lattices of sapphire and gallium nitride are mismatched enough that the assembly is prone to defects, called dislocations, which decrease the efficiency of the devices.

Researchers are focusing on nanostructured materials as a way to minimize such defects. For example, George T. Wang, a researcher at the Solid-State Lighting Science Energy Frontier Research Center at Sandia National Laboratories, is making LEDs with gallium nitride nanowires. The small structures can help reduce strain in the assembly and allow better incorporation of the indium needed to produce visible light.

White-light LEDs can be made by mixing red, green, and blue light. Gallium nitride-based LEDs need a little help to be able to emit such colors, but it can be done.

On its own, gallium nitride emits ultraviolet light. Adding indium shifts its emission to blue light. Coating it with phosphors converts some of the blue light to red and yellow-greenish wavelengths, and these emissions can then be mixed to make white.

A disadvantage of phosphors is that they waste some of the light emitted by LEDs and thus reduce their efficiency. Without recourse to phosphors, red-emitting LEDs can be made from materials other than gallium nitride, but making green LEDs is a significant challenge that researchers are trying to overcome.

The problem is that the strain from the amount of indium that must be added to gallium nitride to produce green LEDs causes so many defects that the material has low efficiency, Wang said. Nanowires could help get around this not-easy-being-green problem. Nanowires’ small size and comparatively large surface area allow a greater proportion of indium to be incorporated, Wang said. “A nanowire has a large free surface, so it can elastically relax without necessarily needing to generate dislocations.”

Nanowires can be made either by growing them individually, referred to as “bottom-up” growth, or by etching them out of planar structures, a “top-down” approach. The resulting wires can be structured axially, with the active regions stacked on top of each other in the direction of wire growth, or radially, with emissive layers wrapped around a central gallium nitride core.

“The advantage of the radial structure is that your indium gallium nitride is a much bigger area,” Wang said. “You have this high-surface-area shell around each wire. Your active region can be much larger” than in a conventional non-nanowire-based LED structure.

Dapkus is also making LEDs with radial nanostructures. Another advantage of such structures, he said, is that the active layers grow on a different crystalline face from the one dictated by the underlying sapphire substrate. The different crystal planes may reduce the lattice mismatch and lead to fewer efficiency-decreasing dislocations.

Dapkus first deposits an amorphous material on top of a sapphire substrate. He then creates openings through which the nanowires can grow. The holes dictate the size and location of the nanorods.

That patterning could lead to another way to overcome the green problem. Changing the spacing of the holes, and thus of the nanorods, can control the color of the emitted light, Dapkus said. In the long term, he and his coworkers hope to be able to make white-light LEDs that don’t require phosphors, he told C&EN. Different portions of the chip will emit different colors of light, and the small size of the chip and blending of light by the chip packaging will enable it to produce white light without phosphors, he said.

Zetian Mi, an associate professor of electrical and computer engineering at McGill University, in Montreal, is also working toward phosphor-free nanowire LEDs as a more efficient way to achieve white light. He does that by making indium gallium nitride form islands rather than layers in the nanowires. These islands are equivalent to semiconductor nanocrystals known as quantum dots.

At first, growing the islands was difficult. “We figured out that we had to optimize the growth rate and other conditions so that the formation of the quantum dots is driven by strain between the indium gallium nitride and gallium nitride,” Mi said. “The dots-in-a-wire structure allows you to tune the emission by adjusting the composition and dot size.”

In OLEDs, electrons are injected into organic-molecule-based diodes, producing electronic states of those molecules that act as electrons and holes. “When a negative charge recombines with an electron hole on the same molecule, you get light out,” Forrest said.

OLEDs can be made with either fluorescent or phosphorescent materials, but phosphorescent molecules are preferred. “With fluorescent materials, usually you lose about 75% of the charge you inject into the device,” said Malte C. Gather, a physicist who recently moved from the Technical University of Dresden, in Germany, to the University of St. Andrews, in Scotland. “With phosphorescent materials, you are at least in principle able to use all the charge you inject.”

OLEDs can have internal quantum efficiencies approaching 100%. That means that in the best case, all of the electrons injected result in the formation of a photon. But that’s inside the device. For those photons to be used for lighting, they have to be extracted from the device, a process known as “outcoupling.” Poor extraction means that overall device efficiency is often as low as 20%.

In the symposium, Gather described several ways that his group is improving the extraction efficiency of OLEDs. In OLED fabrication, the organic material of the OLED is usually layered atop a transparent indium tin oxide electrode on a glass substrate. The refractive indexes of the components are so different that many photons get reflected back into the device.

Gather’s group adds a film of TiO₂ nanoparticles between the glass and the indium tin oxide. The nanoparticles scatter the light, changing the angle at which it hits the interface and making it more likely to escape the device (J. Appl. Phys. 2013, DOI: 10.1063/1.4807000). By combining the nanoparticle films with other light-extraction techniques, Gather and coworkers achieve external quantum efficiencies of 46%.

Another way to improve efficiency, Gather said, is by aligning the organic emitter molecules. “We have developed a technique to measure the orientation of the emitter molecules in the material and to estimate how much efficiency can be gained” by aligning them.

Gather’s team demonstrated the method with two phosphorescent iridium complexes, one with three phenylpyridine (ppy) ligands and the other with two ppy ligands and an acetylacetonate (Appl. Phys. Lett. 2012, DOI: 10.1063/1.4773188). In the former, the molecular dipoles are oriented isotropically (equally in all directions). In the latter, 77% of dipoles are horizontally aligned. OLEDs incorporating these emitters achieved external quantum efficiencies of 18.3% and 21.7%, respectively, demonstrating that the aligned emitters get an efficiency boost.

“Ideally, we would be able to take a super­good, superefficient material and tell it to become oriented,” Gather told C&EN. “But probably in reality, we will have to design molecules with chemical structures that help the molecules to become oriented.”

Fluorescent materials make such poor OLEDs because 75% of the injected electrons end up going into nonemissive states. But Chihaya Adachi, head of the Center for Organic Photonics & Electronics Research at Kyushu University, in Japan, and his coworkers have figured out a way to greatly improve the efficiency of fluorescent OLEDs through a process called thermally activated delayed fluorescence.

In this process the energy from a triplet excited state that doesn’t emit light is transferred to a singlet excited state that does. But such energy transfer is usually considered to be forbidden in quantum mechanical terms and usually requires heavy metals.

Adachi gets around this problem without using heavy metals by designing emitters with a small energy difference between triplet and singlet states. He described his group’s work with a family of compounds based on carbazolyl dicyano­benzene (Nature 2012, DOI: 10.1038/nature11687). They made OLEDs using compounds that emitted green, orange, and sky-blue light. The green OLED had an external quantum efficiency of more than 19%. The orange and blue OLEDs were less efficient, at 11% and 8%, respectively.

The various approaches for improving LEDs and OLEDs have the potential to make an already efficient form of lighting even more efficient. “As we consume more and more energy for other things, we’ve got to save energy wherever we can,” Dapkus told C&EN. “That’s why I feel passionate about wanting LEDs to work.”