After the research efforts of Pankove and co-workers had ended, work on GaN virtually ceased. In 1982 only a single paper was published on GaN. However, Isamu Akasaki and co-workers in Nagoya, Japan, refused to give up, and in 1989 they demonstrated the first true p-type doping and p-type conductivity in GaN. The stubborn Mg acceptors were activated by electron-beam irradiation (Amano et al., 1989). It was later shown that a high-temperature post-growth anneal of Mg-doped GaN also activates Mg dopants in GaN (Nakamura et al., 1994a). Superlattice doping (Schubert et al., 1996) further enhances the activation efficiency of deep acceptors. These p-type doping breakthroughs opened the door to efficient p-n junction LEDs and laser diodes. Today, Mg-doping of GaN is the basis for all nitride-based LEDs and laser diodes.

Subsequent to the attainment of p-type doping, the first GaN p-n-homojunction LED was reported by Akasaki et al. (1992). The LED that emitted light in the ultraviolet (UV) and blue

spectral range, was grown on a sapphire substrate. The result was presented at the “GaAs and Related Compounds” conference held in Karuizawa, Japan in 1992. The LED had an efficiency of approximately 1%. This was a surprisingly high value for the highly dislocated GaN material grown on the mismatched sapphire substrate. It was also the first demonstration that nitride LED efficiencies are not affected by dislocations in the same adverse manner as III-V arsenide and phosphide light emitters.

A name closely associated with GaN LEDs and lasers is that of the Nichia Chemical Industries Corporation, Japan. A team of researchers that included Shuji Nakamura and Takashi Mukai has made numerous contributions to the development of GaN growth, LEDs, and lasers. Their contributions included the demonstration of the first viable blue and green GaInN double­heterostructure LED (Nakamura et al., 1993a, 1993b, 1994b) that achieved efficiencies of 10% (Nakamura et al., 1995), and the demonstration of the first pulsed and cw GaInN/GaN current injection blue laser operating at room temperature (Nakamura et al., 1996). Initially, a particular design, the two-flow organometallic vapor-phase epitaxy (OMVPE) growth-system design was used (Nakamura et al., 1991). However, the use of two-flow OMVPE at Nichia Corporation has been discontinued (Mukai, 2005). Detailed accounts of the team’s contributions were given by Nakamura and Fasol (1997) in the book The Blue Laser Diode and by the Nichia Corporation in the booklet Remarkable Technology (Nichia, 2004).

Blue LEDs made by the Nichia Corporation are shown in Fig. 1.11. A common application of high-brightness GaInN green LEDs is traffic signals as shown in Fig. 1.12. The earlier mentioned GaP:N green LEDs are not suited for this application due to their much lower brightness.

In 1990, when Nakamura entered the field of GaN devices while working for the Nichia Corporation, he was a 36-year-old engineer without a Ph. D., not a single publication, and no conference contribution (Nakamura and Fasol, 1997). At the end of the 1990s, he had become a Professor at the University of California in Santa Barbara and a consultant for the Cree Lighting Corporation, a fierce competitor of Nichia. In 2001, he strongly criticized the Nichia Corporation and Japanese society. In the book entitled Breakthrough With Anger, Nakamura (2001) stated, “There is something wrong with this country. Industry and universities are terribly sick.”

The GaInN material system is also suited for white LEDs. There are different approaches to white LEDs, including white LEDs based on phosphor wavelength converters (see, for example, Nakamura and Fasol, 1997) and on semiconductor wavelength converters (Guo et al., 1999). Much progress is expected in the area of white LEDs, since they have the potential to deliver a substantially higher luminous efficiency compared with conventional incandescent and fluorescent light sources. Whereas conventional light sources have typical (demonstrated) luminous efficiencies of 15-100 lm/W, white LEDs have the potential for luminous efficiencies exceeding 300 lm/W.