But dialing in both color and intensity is a lot harder, and more expensive, than it sounds.
LEDs inherently produce monochromatic light. An excited electron decays back to the ground state, releasing its energy in the form of a photon. The wavelength of this photon is defined by the band structure of the semiconductors used to make the LED.
While monochromatic light is fine for indicator lights, most display and general lighting applications use white light. Not only is white light a mixture of multiple wavelengths but, as previously discussed, it can be cooler or warmer depending on the balance among its red, green, and blue component wavelengths. Lighting designers need to be able to choose from a variety of “whites” depending on the application.
In theory, a lighting element could simply incorporate red, green, and blue LEDs, adjusting the intensity of each to produce the desired output spectrum. In practice, this approach is challenging because the three colors are not equally efficient. Without complex power management circuitry, the final color will tend to be dominated by the blue component. Instead, most solid-state lighting elements depend on phosphors to convert some of the light from one or more blue LEDs into other colors. Phosphors have thus assumed an important supporting role in the lighting industry.
However, phosphors introduce additional complexity of their own. The wavelength conversion process is highly efficient in photon terms: in most cases, more than 90% of the photons supplied by the LED will be absorbed and re-emitted at the phosphor wavelength. Unfortunately, a blue LED photon with a peak emission wavelength of 450nm, has much more energy than, say, a green phosphor photon with a peak emission wavelength of 535nm. Energy lost due to this Stokes shift is converted to heat. Elevated temperatures, in turn, cause the phosphor’s crystal lattice to relax. The relaxed band structure (see Figure) allows phosphor electrons to decay to the ground state without emitting photons, thereby reducing efficiency and raising the temperature still further. This “thermal quench” effect is especially challenging for high-brightness LEDs, due to the higher power at which they operate.
Figure: At elevated temperatures, overlap between the ground state and excited state energy bands due to lattice relaxation allows electrons to dissipate absorbed energy without re-emission of a photon. Image courtesy Dow Electronic Materials.
One possible solution to the problem, choosing a different LED-phosphor combination to minimize the wavelength difference, is only workable for some output spectra. To produce white light, the line between the LED wavelength and the phosphor wavelength must pass through the center of the chromaticity diagram. If the shift is too small, the white portion of the diagram simply won’t be accessible.
Another approach, and one that is being pursued by several phosphor suppliers, seeks to address the thermal quench effect at its root cause, lattice relaxation. If the lattice is stiffer, with a lower coefficient of thermal expansion, then the phosphor performance will remain stable at higher temperatures.
Garnet-based phosphors in the YAG family (yttrium-aluminum-garnet, with various dopants) offer good thermal stability and high conversion efficiency, but typically emit in the red portion of the spectrum. Intematix has developed, and recently patented (U.S. Patents #8,529,791 and #8,475,683) a family of LuAG phosphors (lutetium-aluminum-gallium, also with various dopants). According to Julian Carey, VP of marketing at Intematix, these have the same crystal structure as YAG phosphors, and therefore similar thermal characteristics, but emit in the green or yellow-green portion of the spectrum.
Dow Electronic Materials, meanwhile, is exploring improvements to nitride and oxynitride phosphors. Yongchi Tian of Dow Electronic Materials explained that these crystals, based on Si- and/or Al-centered nitride polyhedra, can be modified by substituting carbon for some of their nitrogen. The resulting carbidonitride and oxycarbidonitride phosphors (U.S. Patent #8,536,777) offer superior thermal performance in, respectively, the red and green parts of the spectrum.
Device specialists don’t usually devote much attention to packaging. The package keeps the environment out and connects the device to the larger circuit, but does not itself take an active role. In solid-state lighting, in contrast, the package becomes much more active. It might include reflectors and other elements to shape the extracted light and, as discussed here, uses phosphors to manipulate the character of that light. In lighting, the capabilities of the package are as important as those of the underlying device.