A professor and his students work at three levels to realize the promise of light-emitting diodes. Light-emitting diodes (LEDs), says Nelson Tansu, are poised to play a breakthrough role in the world’s energy future, and sooner than you might think.
The LED revolution is already encroaching into a global lighting market that is estimated at more than $230 billion. LEDs are showing up in the headlamps of luxury automobiles and in new, brighter traffic signals on roadways. They are used in DvD and Tv remotes, in billboard advertising, in video displays, in sensors and in a host of other applications.
Meanwhile, nearly 20 percent of the electricity generated in the United States is consumed to light homes and businesses, and much of this illumination is provided by incandescent bulbs and fluorescent tubes. Both can be cheaply manufactured, but neither technology, according to the U.S. Department of Energy, is expected to improve significantly in efficiency in the next 20 years.
Incandescent bulbs are particularly wasteful, says Tansu, the Class of 1961 Associate Professor of Electrical and Computer Engineering. They convert 5 percent of the energy they consume into light and dissipate the remainder as heat – which can drive up cooling costs.
LEDs are 10 times more efficient than incandescent bulbs, and their lifespan, potentially, is more than 10 years, compared with 1,000 hours for the traditional light bulb. LEDs are also longer-lasting and slightly more efficient than fluorescent lighting, but more expensive.
LEDs are attracting interest as the surging demand for power worldwide matches the growing prosperity of China, India, Indonesia and other developing nations.
“Developing countries have a combined population of almost 3 billion,” says Tansu, a native of Indonesia. “As their economies grow, this puts real burdens on generating capacity.”
Building more power plants to meet growing energy demand is costly and threatens the environment, says Tansu. “It’s cheaper to save energy than to generate it. Building a power plant to generate a megawatt of electricity is an order of magnitude more expensive than conserving that megawatt.”
Tansu and his students use the state-of-the-art facilities in Lehigh’s Center for Photonics and nanoelectronics to design, grow, characterize, fabricate and test LED devices. Their work is performed at three levels:
- at the nanoscale to improve the efficiency of light generation in the so-called active region of semiconductor materials;
- at the microscale to improve light extraction from the semiconductor into free space;
- at the macroscale to cut costs and enable other practical applications for LED technology. These include high-power transistors, solar energy conversion and storage, and laser technologies for biomedical and environmental applications.
Engineering at the atomic scale
LEDs are a specialized type of diode, an electronic component that allows current to flow in one direction but not the other. An LED’s semiconductor material includes regions called n-type and p-type layers whose carriers are electrons and electron holes.
N- and p-type layers are formed by incorporating impurities in semiconductors that cause high concentrations of free electrons and holes in the two layers. When a current is switched on, the electrons and holes travel in opposite directions and recombine in a semiconductor’s active regions.
These active regions consist of several ultra-thin semiconducting layers with thickness of 2 to 3 nanometers, or 6 to 10 layers of atoms. The active regions, with appropriate barrier layers, are embedded between the n- and p-type layers.
The recombination of electrons and holes causes the electron to drop to a significantly lower energy level and to release the lost energy as a photon or light. This recombination can also lead to the loss of energy, in the form of other non-radiative processes, without any light being generated.
To improve the brightness and intensity of LED light, says Tansu, it is critical to engineer the active layer to achieve a light-generation rate that is significantly higher in value than the generation rate for non-radiative processes.
The transition energy and the light-generation rate in the active layer depend on the material energy bandgaps, heterostructures, and composition and thicknesses of the n- and p-type layers.
The transition energy (the energy difference between the quantum-mechanical electron and holes in the active layer) determines the frequency, or color, of the light emitted. This energy can be engineered by controlling the composition and thicknesses of the n- and p-type layers. By engineering the heterostructures and design of the active layer, the light-generation rate can be increased.
Tansu’s group works mainly with indium gallium nitride (Ingan) LEDs that produce blue, green and white light. The researchers use metal organic chemical vapor deposition (MOCvD) instruments in the Smith family Laboratory for Optical Technologies to form precise layers of Inga and n atoms whose growth rates and design structures are optimized to increase light generation at the desired frequency from the active regions.
By using transmission electron microscopy, the researchers distinguish Inga and n atoms as a series of distinct alternating straight lines that are deposited by MOCvD equipment that is capable of laying down one row of atoms at a time. By adjusting this nanostructure, the researchers can precisely control the energy that will be “lost” when an electron and hole recombine, thereby engineering the color and intensity of the light released.
Spatial separation and efficiency droop
One challenge for the group is presented by green-emitting LEDs, whose light-generating efficiency is relatively lower due to the increased spatial separation of electrons and holes in the active layer.
“We are engineering the nanostructures in the active layer so that the spatial separation of these electrons and holes is significantly reduced,” says Tansu. “This leads to high internal quantum efficiency in the devices.”
Another obstacle – blue LEDs achieve relatively high efficiency at low current density, but this efficiency ‘droops’ at high current density. Tansu’s group is working to identify the factors that cause efficiency droop in LEDs, while developing new types of barrier layers to suppress droop.
Reliable, efficient LED devices must be built on substrates with low dislocation density. In a collaboration with Helen Chan and Richard vinci, professors of materials science and engineering, Tansu has developed a new method of growing gan templates on nanopatterned sapphire substrates.
Engineering a timely escape
Generating photons efficiently is important; making that light visible outside the LED itself is another challenge.
“Once you generate light in semiconductor chips, you have to be able to engineer microstructures that allow more of that light to escape,” Tansu says. Most LEDs are produced by a process that results in planar LEDs at the interface between the semiconductor and free space.
“For planar LEDs, approximately 4 percent of the light rays will be within the escape cone of the devices,” Tansu says. “Photon recycling increases total light extraction as much as 20 percent.”
The remainder of the light is scattered back and absorbed into the material itself. Engineers have developed ways to increase efficiency by roughening the surface of the material to allow more escape angles for the light. Researchers have also formed periodic microstructures by using highly precise electron-beam lithography to increase light extraction from LEDs.
Tansu’s group is working with James gilchrist, associate professor of chemical engineering, to improve optical efficiency with low-cost and practical methods of light extraction. They have developed arrays of microlenses to boost light extraction; their method also enables greater control of the angular distribution of the light output.
The colloidal microlenses measure less than 1 micron in diameter; 100 of them would fit inside a human hair. viewed microscopically, the lenses look like a single layer of ping-pong balls spread in a highly ordered and close-packed configuration across a table. Tansu and gilchrist cover the balls with varying thicknesses of polymeric materials in order to alter their optical properties.
“We can tune the thickness of the polymer layers and form them with different aspect ratios,” says Tansu. This enables the researchers to control whether the light emitted from an LED active layer is intense and focused or soft and diffuse.
Rather than a row of balls packed closely together, some samples look like an array of children’s building blocks with rounded prongs, while others appear more like an array of sunny-side-up eggs with small, yolk-like protrusions.
The LED potential
The state of the art in quantum efficiency and optical effectiveness already enables prodigious lighting capacity, says Tansu.
“You can get the same light output as a traditional light bulb, and this light is generated from an active layer with a total thickness of only 30 layers of atoms, or approximately 1/10,000 of the thickness of your hair,” Tansu says.
Currently, LEDs are packaged like traditional bulbs, with familiar structures to ensure electrical contact and light transmission in existing fixtures.
In the developing world, where new construction is the norm, there is the potential to reimagine what the light bulb looks like and how it functions, Tansu says. Over the next 30 years, it is expected that 60 percent of houses in developing countries will be newly built.
This, says Tansu, offers abundant opportunities to reinvent entire lighting systems.
“In warm areas, the heat loss from traditional lighting causes additional power to be consumed in order to cool residential and office buildings. The energy efficiency gained from using LED lighting will also enable cooling costs to be cut.”
Opportunities for creative interior decoration will also open up, as permanent incandescent light fixtures, with their successions of bulbs, give way to arrays of smaller, brighter LED assemblies integrated smartly into new construction.
And it may one day be possible, in developed and developing nations alike, to engineer LED sensors that detect biological pathogens in the air and also emit deep Uv light that renders them harmless.
Meanwhile, Tansu’s students, taking advantage of the fact that solid-state LEDs can be switched much faster than traditional bulbs, are working to send and receive e-mails by using visible LED light modulation as proof-of-concept experiments.
Costs must drop, however, before LEDs achieve significant market penetration, says Tansu. new methods are required to scale up the low-cost production of LEDs. Engineers must still learn how to make current flow evenly through devices and how to handle the heat that is generated when photons are absorbed back in the semiconducting material.
Still, says Tansu, it is only a matter of time – and not much time – before LEDs become common in households.
“It’s not a matter of if LED light sources will come to your home, but when. In the near future, houses for the general population will be lit only with LED lighting.”