Light emitting diodes (LEDs) are semiconductor devices made in roughly the same way as computer chips. Their light doesn’t come from a heated wire or excited atoms in a vapor, instead it is emitted from a solid crystalline chip, which is why it’s called solid-state lighting. There are a few different processes that can be used to build semiconductors, but they all start by building a crystal.
A crystal is defined as an arrangement of atoms with long range order. Imagine a billiard table packed full of billiard balls, with no room in between. Then put on another layer of billiard balls, then another, each resting in the “pockets” where the balls on the previous level come together. If you know where the center of one billiard ball is, and the direction of the sides of the table, you know exactly how far to go to find the center of another billiard ball, in all three directions. That’s what solid state physicists mean by “long range order.” The long range order creates effects with far-reaching consequences.
Doping isn't Always Bad
Under “normal” circumstances, each atom in the crystal has just the right number of electrons around it, all in relatively low energy states called the “valence band.” Interesting things happen when the base material (whatever it is) is “doped” by the addition of atoms that have either one more or one less electron than the base material. The impurities—the dopant atoms—squeeze into the spot that a base atom would otherwise occupy, connected to the atoms around it. But that leaves either one extra electron floating around, or a “vacancy” where there should be an electron. The vacancies are typically called “holes.”
Material fabricated with extra electrons is called n-type, because it is more negative than the pure material, while that with holes is called p-type because it’s more positive. When a layer of p-type is put next to a layer of n-type it creates a region called a junction. The interesting effects are all near the junction.
Because of quantum mechanical rules, the extra electrons in n-type material have higher energy than the normal valence band electrons. The higher energy electrons are called “conduction band” electrons. Just like every other high energy electron, these conduction band electrons in the n-type material would like to give up energy and get into a lower state—but there’s nowhere for them to go. The vacancies are all in the valence band over in the p-type material.
But applying a voltage across the chip pushes electrons in the n-type and the holes in the p-type material towards each other. The electrons then drop down into the lower energy spot in the valence band. With properly selected materials, those electrons will put that energy into a photon, that is, the diode will emit light.
Engineering Light
Device engineers select the pure crystalline material, the dopants and the level of doping, the thickness of the crystal layers and other parameters that influence the lifetime, the current carrying capacity of the diode, and the energy difference between the conduction and valence bands (called the band gap). By tuning the band gap, the wavelength—the color—of the light can be selected. Unlike incandescent and fluorescent lights, the electrical energy goes directly into the electrons that will be emitting photons. That’s the source of the efficiency of solid-state lighting. Electrical energy goes right into photons of a specific wavelength.
The story is not exactly that simple. A single LED puts out light of a single color. So white LEDs either mix different color LEDs together or they put short wavelength blue light into a phosphor, which emits some reddish light. The separate colors, whether from individual LED chips or from a single LED and a phosphor, get combined to make “white.” LEDs trick the eye into thinking it’s seeing all the wavelengths that make up white light, even though the LEDs don’t put out all the different colors that make up white. Still, white light LEDs are available with specific color temperatures.
The solid-state lighting industry is still in its infancy. LEDs are now “good enough” to be competitive with incandescents and fluorescents, but they are still not as bright as traditional sources. Although they are the most efficient light source available, they still have room to improve their efficiency. And—although they are economically competitive on a cost-of-ownership basis—they still need to come down in purchase price to find wide consumer acceptance.
But the fundamental physics driving SSL is inherently more efficient than the mechanisms responsible for incandescent and fluorescent light, so expect to see LEDs emerge as the next generation general lighting source.
Previous articles in the Physics of Light series:
Not Your Grandpa’s Bulb: The Physics of Lighting
Hot Bodies—How Incandescent Bulbs Work
Zapping Atoms—How Fluorescent Lamps Work
Sources
- Nikolic, B. (N.D.). Crystalline Solids: Symmetry and Bonding. Retrieved from physics.udel.edu.
- Davidson, M. (2006). Diode Lasers. Retrieved from micro.magnet.fsu.edu.
- Anon. (2010). Solid State Lighting. Retrieved from eere.energy.gov.
Additional Resources
- Waide, P. (2010). Phase Out of Incandescent Lamps. Retrieved from iea.org.
- Koehler, K. (2003). Electronic Transitions. Retrieved from rwc.uc.edu.
- Fowler, M. (2008). Black Body Radiation. In Modern Physics. Retrieved from Galileo.phys.Virginia.edu.
- Anon. (N.D.). The Nature of Light and Color. Retrieved from motion.kodak.com.
- Ronda, C. and Srivastava, A. (2003). Phosphors. The Electrochemical Society Interface. Retrieved from electrochem.org.
- Nave, R. (N.D.) Fluorescent Lighting. Retrieved from hyperphysics.phy-astr.gsu.edu.
Join the Conversation