GaN single crystal substrate and method of producing same

Details GaN single crystal substrate and method of producing same GaN single crystal substrate and method of producing same

1999-06-15

2002-07-02

Le, Hoa T. (Department: 1773)

Stock material or miscellaneous articles

Web or sheet containing structurally defined element or...

Physical dimension specified

C117S952000, C423S409000

Reexamination Certificate

active

06413627

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a GaN single crystal substrate for making blue light LEDs (light emitting diodes) or blue light LDs (laser diodes) based upon III-V group nitride (GaN-like) semiconductors.
This application claims the priority of Japanese Patent Applications No.10-171276(171276/1998) filed on Jun. 18, 1998 and No.10-183446(183446/1998) filed on Jun. 30, 1998, which are incorporated herein by reference.
2. Description of Prior Art
FIG. 1
shows the ratios of the lattice constant and the thermal expansion coefficient of the candidate materials for the substrate of growing GaN to those of GaN. Here, candidate substrates for GaN growth are sapphire (Al
2
O
3
), silicon carbide (SiC), silicon (Si), gallium arsenide (GaAs) and zinc oxide (ZnO). The nitride semiconductor light source devices, or GaN-group light source devices have been produced by piling GaN or GaN-like films epitaxially upon a sapphire substrate (Al
2
O
3
). The sapphire substrate is endowed with high chemical stability and high heat resistance. Sapphire is rigid. Although the lattice constant of sapphire is different from that of GaN by about 16%, GaN crystal can grow epitaxially on a sapphire substrate by interposing a GaN buffer film. For this reason, sapphire is exclusively used as a substrate for the GaN crystal growth. Since there is no means of eliminating the sapphire substrate, the GaN-group LEDs having the sapphire substrate are utilized for practical purposes. The current GaN-LEDs (more strictly; GaInN-LEDs) are complex devices including GaN and sapphire. The GaN/sapphire LEDs are already practical devices on the market at present. It is said that GaN/sapphire LDs (laser diodes) will be sold on the market in near future.
Sapphire (Al
2
O
3
) is different from GaN in the lattice constant. In spite of the large difference of the lattice constant, a sapphire substrate enables GaN crystal to grow epitaxially on it. The reason is the occurrence of gradual lattice relaxation.
FIG. 2
shows the relation between the film thickness and the lattice constant of GaN films on sapphire. At the limit of 0 thickness, the lattice constant is equal to that of sapphire. The lattice constant of GaN film on sapphire slowly decreases to that of GaN as the GaN film thickness increases. At present, sapphire is the best material for the substrate of growing GaN films. All the GaN-LEDs on sale have the structure of GaN/Al
2
O
3
. The conventional structure of GaN/sapphire is explained, for example, by;
{circle around (1)} Japanese Patent Laying Open No. 5-183189 (183189/'93) and
{circle around (2)} Japanese Patent Laying Open No. 6-260680(260680/'94).
The best sapphire substrate still has a problem. The defect density in the GaN film on a sapphire substrate is very large. The defect density is about 10
9
cm
−2
in the GaN on sapphire. The high density of defects may originate from the misfit of the lattice constant between GaN and sapphire. Defects prevail in the GaN. However, GaInN-LEDs are gifted with a long lifetime. The high density of defects may be a problem on crystallography but is not a practical problem in the GaN-LEDs. The sapphire substrate has another drawback from the mechanical viewpoint. Sapphire is chemically stable and physically rigid. The high stability against chemicals is a drawback as well as a merit. Any chemicals can not remove only the sapphire substrate, keeping the GaN films intact. The most inconvenient matter is the lack of cleavage. Since sapphire has no cleavage planes, the GaN/sapphire wafer is diced by pushing blades upon the sapphire wafer forcibly. Sometimes the wafer is broken. The yield of dicing is low. The lack of cleavage and the rigidity enhance the difficulty of dicing a GaN/sapphire wafer into plenty of device chips. Rigidity and non-cleavage are the most serious difficulty of sapphire substrates.
{circle around (3+L )} S. Nakamura et al., “High-Power, Long-Lifetime InGaN/GaN/AlGaN-Based Laser Diodes Grown on Pure GaN Substrates”, Jpn. J. Appl. Phys. Vol.
37 (1998)pp.L309-L312, shows an experiment of obtaining a GaN wafer by eliminating the sapphire substrate by polishing. However, this example is only on a laboratory scale.
Somebody tried to replace the sapphire substrate by silicon carbide SiC having cleavage planes. A GaN/SiC device was proposed by;
{circle around (4+L )} A. Kuramata et al., “InGaN Laser Diode Grown on
6H-SiC Substrate Using Low-Pressure Metal Organic Vapor Phase Epitaxy”, Jpn. J. Appl. Phys. Vol.36(1997) pp.L1130-L1132.
SiC, however, has drawbacks. SiC has so high chemical stability that high temperature more than 1500° C. is required for producing SiC single crystals. Namely, difficulty lies at the production of SiC crystals. The SiC substrate itself is still expensive. The SiC will raise the cost of GaN/SiC devices. In practice, the GaN/SiC devices have not been made on a large scale yet. SiC is not matured to a practical material of substrates.
Prior GaN-LEDs are produced on sapphire substrates. The sapphire substrate cannot be eliminated. The sapphire substrates accompany the GaN devices in use at present. The GaN-LED devices at present may be called a “GaN/sapphire complex” device.
The substrate must be heated at a temperature higher than 1000° C. in a furnace for growing GaN-type films epitaxially on the substrate. The vapor phase reaction requires such a high temperature. When the GaN-type epitaxial films have been grown on the substrate, the substrate with the epitaxial films is cooled to room temperature for getting out of the furnace. Cooling causes undesirable influences on the GaN films due to the difference of thermal expansion coefficients between the GaN films and the substrate. Strictly speaking, the thermal expansion coefficients are not constants but variables as a function of temperature. Ignoring the small temperature dependence, rough estimation teaches us that GaAs has thermal expansion coefficient of about 1.08 times as big as GaN. In the normalized unit GaN=1, the thermal coefficients are 1.08 for GaAs, 0.87 for SiC and 1.36 for sapphire (Al
2
O
3
).
The difference of thermal expansion coefficients between the films and the substrate causes a first problem of occurrence of thermal stress in the GaN films. The thermal stress induces microcracks and other defects in the GaN films. A further problem is the fact that the thermal expansion coefficient difference invites distortion of the substrate in the cooling process. The whole epitaxial wafer having the substrate and the films deforms due to the thermal expansion discrepancy. A third problem is impossibility of making a large complex GaN/sapphire substrate. The complex GaN/sapphire might be called a GaN substrate. But, large distortion and big stress prevent manufactures from producing large GaN/sapphire wafers. Someone reported a success of making a GaN crystal of a several millimeter square which is not available to a mass production on industrial scale.
Many attempts had been made long years before to make GaN crystals on GaAs substrates. As shown in
FIG. 1
, GaAs has a thermal expansion coefficient nearly equal to GaN. GaAs had been a promising candidate as a substrate of GaN growth. The GaAs substrate, however, had a serious drawback. The high temperature at the growth forced As to evaporate from the surface of the GaAs crystal due to the high dissociation pressure of As at high temperature. The GaAs substrate reacts with ammonia NH
3
. These reasons forbade a GaAs substrate from growing a good GaN crystal thereupon. All the trials had failed in making a good GaN on a GaAs substrate wafer. Then, the GaN growth on GaAs had been deemed to be entirely impracticable.
Only GaN/sapphire survives now as a GaN device. Thus, one way of improvement is directed to sophistication of the sapphire substrate method. It is said that GaN/sapphire LEDs have a long lifetime despite the high density defects. However, if the defects were reduced, the lifetime of GaN LEDs would be prolonged further. Besides, GaN LDs don&a

Affiliated with

Also associated with

Rating

GaN single crystal substrate and method of producing same does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with GaN single crystal substrate and method of producing same, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and GaN single crystal substrate and method of producing same will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2880180