Solid‐state lighting has made tremendous progress this past decade, with the potential to make much more progress over the coming decade. In this article, the current status of solid‐state lighting relative to its ultimate potential to be “smart” and ultra‐efficient is reviewed. Smart, ultra‐efficient solid‐state lighting would enable both very high “effective” efficiencies and potentially large increases in human performance. To achieve ultra‐efficiency, phosphors must give way to multi‐color semiconductor electroluminescence: some of the technological challenges associated with such electroluminescence at the semiconductor level are reviewed. To achieve smartness, additional characteristics such as control of light flux and spectra in time and space will be important: some of the technological challenges associated with achieving these characteristics at the lamp level are also reviewed. It is important to emphasise that smart and ultra‐efficient are not either/or, and few compromises need to be made between them. The ultimate route to ultra‐efficiency brings with it the potential for smartness, the ultimate route to smartness brings with it the potential for ultra‐efficiency, and the long‐term ultimate route to both might well be color‐mixed RYGB lasers.

No abstract

Inductively coupled plasma (ICP) etch rates for GaN are reported as a function of plasma pressure, plasma chemistry, rf power, and ICP power. Using a Cl2/H2/Ar plasma chemistry, GaN etch rates as high as 6875 Å/min are reported. The GaN surface morphology remains smooth over a wide range of plasma conditions as quantified using atomic force microscopy. Several etch conditions yield highly anisotropic profiles with smooth sidewalls. These results have direct application to the fabrication of group-III nitride etched laser facets.

We report the growth of InGaN/GaN multiple-quantum-well blue light-emitting diode (LED) structures on Si(111) using metalorganic vapor phase epitaxy. By using growth conditions optimized for sapphire substrates, a full width at half maximum (FWHM) (102) x-ray rocking curve of less than 600 arcsec and a room-temperature photoluminescence peak at 465 nm with a FWHM of 35 nm was obtained. Simple LEDs emitting bright electroluminescence between 450 and 480 nm with turn-on voltages at 5 V were demonstrated.

The competition between band gap and the 2.2 eV ͑yellow͒ luminescence of epitaxial GaN is studied for excitation densities ranging from 5ϫ10 Ϫ6 to 50 W/cm 2. The ratio of the peak intensities of the band gap-to-yellow luminescence changes from 4:1 to 3000:1 as the excitation density is increased by 7 orders of magnitude. At room temperature, the band gap luminescence linewidth is 2.3kT, close to the theoretical minimum of 1.8kT. A model is developed describing the intensity of the two radiative transitions as a function of the excitation density. This model is based on bimolecular rate equations and takes into account shallow impurities, deep levels, and continuum states. The theoretically predicted dependences of the two different luminescence channels follow power laws with exponents of 1 2 , 1 and 3 2. Thus the intensity of the yellow luminescence does not saturate at high excitation densities. These dependences are in excellent agreement with experimental results. The relevance of the results for optoelectronic GaN devices is discussed. It is shown that the peak intensity of the yellow luminescence line is negligibly small at typical injection currents of light-emitting diodes and lasers.