Method of growing single crystal GaN, method of making...

Details Method of growing single crystal GaN, method of making... Method of growing single crystal GaN, method of making...

2000-09-27

2002-10-22

Kunemund, Robert (Department: 1765)

Single-crystal, oriented-crystal, and epitaxy growth processes;

Forming from vapor or gaseous state

With decomposition of a precursor

C117S095000, C117S097000, C117S106000, C117S952000, C117S953000

Reexamination Certificate

active

06468347

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a GaN single crystal substrate, a method of growing a GaN single crystal and a method of making a GaN single crystal substrate which is used for producing light emitting devices, for example, light emitting diodes (LEDs) and laser diodes (LDs).
This application claims the priority of Japanese Patent Application No.11-273882 (273882/1999) filed Sep. 28, 1999 which is incorporated herein by reference.
2. Description of Related Art
Light emitting devices based upon group III-V nitride semiconductors (GaN, GaInN) have been put into practice, in the field of blue light LEDs (light emitting diodes). Since wide gallium nitride (GaN) substrates cannot be produced, the nitride semiconductor devices (blue light LEDs) have been produced upon sapphire substrates (Al
2
O
3
). Sapphire crystal (Al
2
O
3
) belongs to hexagonal symmetry group. The c-plane (0001) has six-folding rotation symmetry. Thin films of GaN or GaInN are heteroepitaxially grown upon the sapphire substrates for making GaInN type blue light LEDs. The GaN films or GaInN films grown on the sapphire substrates are suffering from a large number of dislocations of about 10
9
cm
−2
. Despite the high dislocation density, the GaInN/sapphire LEDs made by piling GaN films and GaInN films on the sapphire substrates exhibit high blue light power and a long lifetime as blue light LEDs. The high density of dislocations in the GaN or GaInN films is not a hindrance to the GaInN/sapphire LEDs. Sapphire is sturdy and strong chemically and physically and refractory (high heat-resistance). The sapphire substrate is a very hard and stable material. The advantages allow sapphire to exclusively serve substrates to the GaInN-type blue light LEDs.
The sapphire substrates have still drawbacks. The sapphire substrate has no cleavage plane which would enable device makers to cut a processed wafer into individual chips in exact orientations along the natural cleavage planes without difficulty. Unlike traditional semiconductor wafers, the lack of cleavage forces the device makers to dice the sapphire wafer lengthwise and crosswise with a dicing machine. The dicing step raises the cost of production for making LEDs. In the case of making laser diodes in future, the lack of cleavage will prohibit the makers from forming a pair of mirrors as a resonator by natural cleavage. The resonators made by polishing incur problems in quality and raise the cost of manufacturing. A further drawback is the fact that the sapphire substrate is an insulator.
The insulating substrate invites various problems on electrode fabrication. Unlike a conventional conductive substrate, the bottom of the sapphire substrate cannot be an electrode. In stead of the bottom, a part of a middle layer is exposed by polishing partially the top of a GaN chip for serving a room for an n-electrode. Connection between the LED electrodes and leads requires wirebonding two times per chip. The electric current flows in the horizontal direction in the intermediate GaN layer having the lower electrode. The GaN intermediate layer should be allocated with a sufficient large thickness for decreasing electric resistance. Two electrodes made on the top require a wide area for the LED chip. The GaN devices on the sapphire substrates are suffering from high cost.
Silicon carbide (SiC) substrates have been proposed for the GaN devices for solving the problems accompanying the sapphire substrates. SiC has natural cleavage planes which allow device makers to cut a SiC wafer along the natural cleavage lines in exact orientations without difficulty. The SiC substrate would solve the problems of the dicing step and the resonators of LDs. SiC has good electric conduction which allows a substrate bottom to become a lower electrode. The bottom electrode can reduce the space occupied by the electrodes. A single wirebonding connects the top electrode to the lead. In spite of the convenient properties, SiC is far more expensive than sapphire. SiC is difficult to obtain due to poor supply of SiC. Poor supply of SiC would invite high cost and instability in quality. Further, a GaN film grown upon the SiC substrate still has a problem in quality which is not solved at present. The high cost still inhibits the SiC substrate from bringing GaInN/SiC blue light LEDs into practice.
Crystallographical problems should be pointed out. Heteroepitaxial growth of GaN single crystal films upon sapphire substrates or silicon carbide substrates introduces many defects, e.g., dislocations into the film crystals due to the misfit of lattice structures between the films and the substrates. Different lattice constants degrade the property of the grown crystals. Indeed, it is said that the epitaxial layers of GaN or GaInN upon sapphire substrates on sale should include very high density of dislocations of about 10
9
cm
−2
.
In the case of the silicon carbide substrates, they say that the GaN or GaInN layers should contain high dislocation density of about 10
8
cm
2
.
Such high density of dislocation would fully deprive Si or GaAs devices of all the desired functions in the case of the Si or GaAs semiconductor devices. Defects are fatal for Si or GaAs crystals. Device fabrication on state of art requires dislocation-free crystals for Si and low dislocation density crystals for GaAs. Low dislocation density or non-dislocation density is indispensable for the preceding Si devices or GaAs devices.
To our surprise, GaN(GaInN) blue light LEDs function quite well despite of such high dislocation density. Many dislocations do not impede the device makers from bringing the GaN type LEDs into practice. Plenty of dislocations do not invite degradation of the LEDs. The GaInN type blue light LEDs are not annoyed at the high dislocation density at present.
The high dislocation density of the crystals induces few problems on the GaN LEDs, because the current density is small in the LEDs. However, such high density of dislocations will induce difficulties in the case of LDs (laser diodes) which require far larger current density than LEDs. The high current density facilitates the degradation originating from the defects in a short time. Current GaInN blue light LDs made upon sapphire substrates are still suffering from a short lifetime. The GaInN blue light LDs have not attained to the practical level yet due to the rapid degradation and the short lifetime in contradiction to the GaInN LEDs. 10
9
cm
−2
dislocations seem to reduce the lifetime of the GaInN LDs. The sapphire substrates originate plenty of dislocations in the GaN and GaInN layers grown on the substrates.
The Inventors of the present invention think that the best substrate for the GaN devices should be a GaN single crystal substrate. The adoption of GaN single crystals as the substrates should entirely solve the problem of the mismatching of the lattice constant. The GaN single crystal has cleavage unlike sapphire. Natural cleavage will facilitate to cut a processed wafer into individual device chips. Natural cleavage will replace the dicing step. The cleavage planes will be assigned as the mirrors of resonators in the case of making GaInN laser diodes (LDs). Unlike sapphire, GaN crystal has electric conductivity which simplifies the structure of electrodes by allowing the crystal to assign an n-electrode on the bottom. GaN single crystal is the best candidate for the substrate for growing the GaN or GaInP layers. However, the GaN single crystal has not been used as the substrate for the GaN-type LEDs or LDs. Why has GaN not been adopted yet? The reason is that large GaN single crystals have never been made. The difficulty of making the GaN single crystal forces the adoption of sapphire for the substrate of the GaN devices.
When solid GaN material is heated, the GaN is directly sublimed instead of giving a GaN melt. It is impossible to make a GaN melt which would allow the ordinary Czochralski method. It is said that ultrahigh pressure would be able to make an equilibrium state be

Affiliated with

Also associated with

Rating

Method of growing single crystal GaN, method of making... does not yet have a rating. At this time, there are no reviews or comments for this patent.

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

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3000280