4. Conductivity control of nitride semiconductors
4.1. Discovery of p-type conduction in GaN and realization
of GaN p–n junction blue LED
Many groups attempted to produce p-type GaN but
without success. With the successful control of the crystal
quality of GaN, we could start to work on p-type doping.
Using LT-AlN buffer layers, the residual donor density in
GaN was also drastically decreased as mentioned above.
But in spite of repeated efforts on Zn-doping, it was not
possible to produce p-type GaN. In 1987, we found that
intensity of Zn-related luminescence greatly increased when
high quality Zn-doped GaN grown with the LT-AlN buffer
layer was irradiated with electron beams during cathodoluminescence
(CL) measurements [21]. We thought that
this phenomenon (named EB effect) [21] might be closely
related with activation of Zn-acceptors and hence with ptype
conduction. But the crystals did not show p-type
conduction. Meanwhile in 1988, we noticed that Mg might
be an acceptor shallower than Zn, since its electronegativity
was larger than that of Zn [22]. In 1989, we succeeded in
Mg-doping in high quality GaN using Cp2Mg or MCP2Mg
as a Mg-dopant while maintaining the high crystalline
quality by using the LT-AlN buffer layer technology [23].
Then the Mg-doped GaN samples were irradiated with
electron beams in the same way as the Zn-doped samples.
We found greatly increased blue luminescence of such Mgdoped
GaN samples (the EB effect) as well as the samples
being low resistivity p-type GaN [24]. Immediately, we
realized the world’s first GaN p–n junction blue/UV LED
with encouraging I–V characteristics in 1989 [24], as shown
in Fig. 5. We achieved p-type AlGaN in 1991 [25] and p-type
GaInN in 1995 [26] in the same manner. In 1992, ptype
GaN was also produced by thermal annealing of Mgdoped
GaN grown with the LT-GaN buffer layer by
Nakamura et al. [27]. Afterwards, p-type GaN was
obtained by UV [28] or electro-magnetic wave irradiation
[29,30] at elevated temperatures below 400 1C.
To realize p-type nitrides, it is essential to activate Mg
acceptors by releasing hydrogen [31,32]. But, we should
first drastically reduce the residual donors before solving
problems related to the hydrogen passivation of acceptors.
不是我不想给分，我也很急，可是都是那么不通顺，我都交不出去了。