Patent Application
20110132410
The present invention relates to GaxIn1-xN (0≦x≦1) substrates, and to methods of cleaning GaxIn1-xN (0≦x≦1) substrates.
Among GaxIn1-xN (0≦x≦1) substrates, GaN (gallium nitride) substrates, on account of their having an energy bandgap of 3.4 eV and high thermal conductivity, have gained attention as materials for such semiconductor devices as short-wavelength emitting light sources and power electronic devices.
Hydride vapor phase epitaxy (HVPE) is one typical technique for growing GaN crystal, and it allows GaN substrates to be manufactured from the GaN crystal. Then, by growing various epitaxial films superficially onto the GaN substrates, optoelectronic and other semiconductor devices can be obtained. As substrates onto which GaN epitaxial films free of surface roughness can be grown, in Patent Document 1, for example, nitride crystal substrates, such as GaN substrates, whose surface carrier density is uniform have been proposed.
Nevertheless, even when substrates described in Patent Document 1 are employed, low-quality epitaxial films have at times been grown, in which, not reflecting the uniformity in carrier density along the substrate surface, defects and haze not related to it appear. Because the device characteristics of semiconductor devices employing low-quality epitaxial films of this sort are harmed, there is a strongly felt need for stably growing high-quality epitaxial films having minimal defects and haze.
Patent Document 1: Japanese Unexamined Pat. App. Pub. No. 2005-101475
Prior art literature making mention of the extent to which particles and organic matter clinging to the surface of GaxIn1-xN substrates such as GaN substrates is removed in order to stably grow high-quality epitaxial films having minimal defects and haze being nil, however, the criteria are unclear, and a consequent problem has been that variations in GaxIn1-xN substrate surface condition have as such led directly to variations in epitaxial film quality.
An object of the present invention is to make available GaxIn1-xN substrates onto which high-quality epitaxial films can be stably grown, and cleaning methods for obtaining the GaxIn1-xN substrates.
One aspect of the present invention is a GaxIn1-xN substrate in which the number of particles of not less than 0.2 μm particle size present on the GaxIn1-xN substrate surface is 20 or fewer, given that the GaxIn1-xN substrate diameter is 2 inches. Herein, in the present description, “GaxIn1-xN substrates” refer to nitride crystal substrates including at least one of gallium (Ga) and indium (In).
The present invention is also a GaxIn1-xN substrate cleaning method of cleaning a GaxIn1-xN substrate by immersing it in a cleaning solution being any one substance selected from the group consisting of aqueous ammonia, ammonium hydroxide-hydrogen peroxide-water mixture, and aqueous organic alkali solutions, while applying ultrasonic waves to the solution, to bring the number of particles of not less than 0.2 μm particle size present on the GaxIn1-xN substrate surface to 20 or fewer, given that the GaxIn1-xN substrate diameter is 2 inches.
Herein, in the GaxIn1-xN substrate cleaning method of the present invention, as the cleaning solution, any one of aqueous ammonia in which the ammonia concentration is 0.5% by mass or more, ammonium hydroxide-hydrogen peroxide-water mixture in which the aqueous hydrogen peroxide concentration is 0.1% by mass or more, with the ammonia concentration being 0.1% by mass or more, and an aqueous organic alkali solution in which the organic alkali concentration is 0.5% by mass or more is preferably utilized.
Furthermore, in a GaxIn1-xN substrate cleaning method of the present invention, the aqueous organic alkali solution is preferably a solution in which an organic alkali being one of either tetramethylammonium hydroxide or trimethyl-2-hidroxyethyl ammonium hydroxide is dissolved in water.
Additionally, in a GaxIn1-xN substrate cleaning method of the present invention, the GaxIn1-xN substrate immersion time is preferably 30 seconds or more.
The present invention further is a GaxIn1-xN substrate in which in a photoelectron spectrum along the GaxIn1-xN substrate surface by X-ray photoelectron spectroscopy at take-off angle of 10°, the ratio between the peak areas of the C1s electron and N1s electron (C1s electron peak area/N1s electron peak area) is 3 or less.
The present invention is furthermore a GaxIn1-xN substrate cleaning method of immersing a GaxIn1-xN substrate in an acid solution to bring the ratio between the peak areas of the C1s electron and N1s electron (C1s electron peak area/N1s electron peak area) to 3 or less in a photoelectron spectrum along the GaxIn1-xN substrate surface by X-ray photoelectron spectroscopy at a take-off angle of 10°.
Herein, in the GaxIn1-xN substrate cleaning method of the present invention, the acid solution is preferably composed of either at least one substance selected from the group consisting of fluoric acid, hydrochloric acid, and sulfuric acid, or an admixed solution of aqueous hydrogen peroxide and at least one substance selected from the group consisting of fluoric acid, hydrochloric acid, and sulfuric acid.
Furthermore, in a GaxIn1-xN substrate cleaning method of the present invention, it is preferable that in implementations in which the acid solution is composed of at least one substance selected from the group consisting of fluoric acid, hydrochloric acid, and sulfuric acid, the total concentration of fluoric acid, hydrochloric acid, and sulfuric acid in the acid solution is 0.5% by mass or more, and in implementations in which the acid solution is composed of an admixed solution of aqueous hydrogen peroxide and at least one substance selected from the group consisting of fluoric acid, hydrochloric acid, and sulfuric acid, the total concentration of fluoric acid, hydrochloric acid, and sulfuric acid in the acid solution is 0.1% by mass or more, with aqueous hydrogen peroxide concentration being 0.1% by mass or more.
Additionally, in a GaxIn1-xN substrate cleaning method of the present invention, the GaxIn1-xN substrate immersion time is preferably 30 seconds or more.
The present invention affords GaxIn1-xN substrates onto which high-quality epitaxial films can be stably grown, and cleaning methods for obtaining the GaxIn1-xN substrates.
Below, a description of modes of embodying the present invention will be made. It should be understood that in the drawings for the present invention, identical reference marks indicate identical or corresponding parts.
The present invention is a GaxIn1-xN substrate in which the number of particles of not less than 0.2 μm particle size present on the GaxIn1-xN substrate surface is 20 or fewer, given that the diameter of the GaxIn1-xN substrate is 2 inches. It being that, as a result of concentrated investigative efforts, the present inventors discovered that if the number of particles of not less than 0.2 μm particle size that are on the GaxIn1-xN substrate surface is controlled as just noted, high-quality epitaxial films having fewer defects can be grown.
Herein, the number of particles on the GaxIn1-xN substrate surface is calculated by counting all of the particles of not less than 0.2 μm particle size present globally across the GaxIn1-xN substrate surface, and converting the counted number of particles to a value hypothesized for when the diameter of the GaxIn1-xN substrate is 2 inches. Accordingly, in the present invention, the size of the GaxIn1-xN substrate is not limited. For example, inasmuch as the surface area of a GaxIn1-xN substrate whose diameter is 4 inches is four times that of one that is 2 inches, in implementations employing GaxIn1-xN substrates whose diameter is 4 inches, ¼ of the total number of particles present on their surface will be what herein is termed the particle count. It should be noted that particles are counted employing equipment such as hitherto publicly known light-scattering substrate-surface inspection devices. Furthermore, the particle substances are not, in particular, limited.
The present invention is also a method of cleaning a GaxIn1-xN substrate by immersing it in a cleaning solution being any one substance selected from the group consisting of aqueous ammonia, ammonium hydroxide-hydrogen peroxide-water mixture, and aqueous organic alkali solutions, while applying ultrasonic waves to the solution, to bring the number of particles of not less than 0.2 μm particle size present on the GaxIn1-xN substrate surface to 20 or fewer, given that the GaxIn1-xN substrate diameter is 2 inches.
Herein, “ammonium hydroxide-hydrogen peroxide-water mixture” is an admixed solution of aqueous hydrogen peroxide and aqueous ammonia, thus not being limited to admixed solutions in which aqueous hydrogen peroxide has been added to aqueous ammonia. Further, “aqueous organic alkali solutions” are solutions in which an organic alkali has been dissolved in water, wherein either of tetramethylammonium hydroxide expressed by the following structural formula (1), or trimethyl-2-hidroxyethyl ammonium hydroxide expressed by the following structural formula (2), is preferably utilized as the organic alkali.
Furthermore, in implementations in which aqueous ammonia is utilized as the cleaning solution, the ammonia concentration with respect to the cleaning solution as a whole is preferably 0.5% by mass or more. And in implementations in which ammonium hydroxide-hydrogen peroxide-water mixture is utilized as the cleaning solution, the aqueous hydrogen peroxide concentration with respect to the cleaning solution as a whole is preferably 0.1% by mass or more, with the ammonia concentration being 0.1% by mass or more. Still further, in implementations in which an aqueous organic alkali solution is utilized as the cleaning solution, the organic alkali concentration with respect to the cleaning solution as a whole is preferably 0.5% by mass or more. Thus regulating concentration in the cleaning solution tends to make the particle count on the GaxIn1-xN substrate surface more stable, making it possible to control the count in the manner described above.
Further, the period for which the GaxIn1-xN substrate is immersed in the cleaning solution is preferably 30 seconds or more. In this case, the fact that the GaxIn1-xN substrate is sufficiently immersed in the cleaning solution tends to make the particle count on the GaxIn1-xN substrate surface more stable, making it possible to control the count in the manner described above. Herein, the GaxIn1-xN substrate immersion time is the time from the point at which ultrasonic waves are applied to the cleaning solution.
The present invention further is a GaxIn1-xN substrate in which in a photoelectron spectrum along the GaxIn1-xN substrate surface by X-ray photoelectron spectroscopy (XPS) at a take-off angle of 10°, the ratio between the peak areas of the C1s electron and N1s electron (C1s electron peak area/N1s electron peak area) is 3 or less. It being that, as a result of concentrated investigative efforts, the present inventors discovered that if the ratio between the C1s electron peak area and the N1s electron peak area in a photoelectron spectrum along the GaxIn1-xN substrate surface by X-ray photoelectron spectroscopy at a take-off angle of 10° is controlled as just noted, high-quality epitaxial films free of haze can be grown. Herein, the ratio between the C1s and N1s electron peak areas in a photoelectron spectrum along the GaxIn1-xN substrate surface by X-ray photoelectron spectroscopy (XPS) at a take-off angle of 10° indicates the amount of organic matter on the GaxIn1-xN substrate surface with respect to nitrogen in the near-surface of the GaxIn1-xN substrate, wherein controlling the proportion in the manner just noted enables growing haze-free high-quality epitaxial films.
Herein, the C1s electron refers to 1s-orbital electrons in carbon (C), and the N1s electron refers to 1s-orbital electrons in nitrogen (N). Then, as illustrated in
The present invention is furthermore a cleaning method of immersing a GaxIn1-xN substrate in an acid solution to bring the ratio between the peak areas of the C1s electron and N1s electron (C1s electron peak area/N1s electron peak area) is 3 or less in a photoelectron spectrum along the GaxIn1-xN substrate surface by X-ray photoelectron spectroscopy at a take-off angle of 10°.
Herein, the acid solution is preferably composed of either at least one substance selected from the group consisting of fluoric acid, hydrochloric acid, and sulfuric acid, or an admixed solution of aqueous hydrogen peroxide and at least one substance selected from the group consisting of fluoric acid, hydrochloric acid, and sulfuric acid. These implementations tend to make the just-noted ratio between the C1s electron peak area and the N1s electron peak area more stable, making it possible to bring it to 3 or less.
Furthermore, in implementations in which the acid solution is composed of least one substance selected from the group consisting of fluoric acid, hydrochloric acid, and sulfuric acid, the total concentration of fluoric acid, hydrochloric acid, or sulfuric acid in the acid solution is preferably 0.5% by mass or more. On the other hand, in implementations in which the acid solution is composed of an admixed solution of aqueous hydrogen peroxide and at least one substance selected from the group consisting of fluoric acid, hydrochloric acid, and sulfuric acid, the total concentration of fluoric acid, hydrochloric acid, or sulfuric acid in the acid solution is preferably 0.1% by mass or more, with the aqueous hydrogen peroxide concentration being 0.1% by mass or more. These implementations tend to make the just-noted ratio between the C1s electron peak area and the N1s electron peak area further stable, making it possible to bring the proportion to 3 or less.
Also, the period for which the GaxIn1-xN substrate is immersed in the acid solution is preferably 30 seconds or more. In this case, the fact that the GaxIn1-xN substrate is sufficiently immersed in the acid solution tends to make the just-noted ratio between the C1s electron peak area and the N1s electron peak area more stable, making it possible to control the proportion in the way just described.
First, 50 GaN substrates of 2 inch-diameter were prepared, the GaN substrates being produced by specular-polishing a GaN crystal grown by HYPE, and then by removing the layer damaged by the specular-polishing. Herein, the 50 GaN substrates each were 400 μm in thickness, and their surfaces were the faces oriented at 2° with respect to the (0001) orientation.
Next, employing a cleaning apparatus as illustrated in the schematic sectional diagram in
Then, in each of the GaN substrates after cleaning, with a light-scattering substrate surface-inspecting device, the numbers of particles present on the GaN substrate surfaces, having a particle size of 0.2 μm or more were counted.
After that, a 1 μm-thick epitaxial film composed of GaN crystal was grown onto the surface of each of the 50 GaN substrates by metalorganic vapor phase epitaxy (MOVPE) under the same conditions. Then, employing the same light-scattering substrate surface-inspecting device as the above device, the number of defects in the epitaxial film was counted.
Results of the experiment were plotted and shown in
As is clear from
Furthermore, a GaN substrate in which the number of particles present on the substrate surface, having a particle size of 0.2 μm or more was 20 or fewer was a substrate cleaned with tetramethylammonium hydroxide concentration with respect to the cleaning solution as a whole being brought to 0.5% by mass or more, and with the time required to immerse the GaN substrate being made 30 seconds or more.
It should be understood that although GaN substrates were employed in Experimental Example 1, the same results are believed to be obtained if other GaxIn1-xN substrates, apart from GaN substrates, are employed. Additionally, GaN substrate thickness and plane orientation are not limited to the above example, and even if the thickness and plane orientation are arbitrary, the same results as in Experimental Example 1 will be obtained.
First, as in Experimental Example 1, 50 GaN substrates of 2 inch-diameter were prepared, the GaN substrates being produced by specular-polishing a GaN crystal, and then removing a layer damaged by the specular-polishing. Herein, the 50 GaN substrates each is 400 μm in thickness, and the GaN substrate surfaces are the faces oriented at 2° with respect to the (0001) orientation.
Next, employing a cleaning apparatus as a schematic sectional diagram in
Then, in each of the GaN substrates after cleaning, photoelectron spectrum on the GaN substrate surfaces at a photoelectron take-off angle of 10° was measured by an X-ray photoelectron spectrometer in which Kα-rays of Mg and Al is utilized as X-ray source, and based on the peak area of an O 1s electron (1s-orbital electron in oxygen), a ratio of the peak area of a C1s electron to the peak area of a N1s electron (C1s electron peak area/N1s electron peak area) was calculated.
After the calculation, a 1 μm-thick epitaxial film composed of GaN crystal was grown onto the surface of each of the 50 GaN substrates by MOVPE under the same conditions. Then, whether haze occurred or not in each of the grown epitaxial films was visually evaluated by the following standards, and epitaxial films in which haze occurred were counted in each of sections represented by (C1s electron peak area/N1s electron peak area) shown in the table. The results are set forth in the table.
Furthermore, a GaN substrate in which the ratio between the peak areas of the C1s electron and N1s electron (C1s electron peak area/N1s electron peak area) along the GaN substrate surface was 3 or less was a substrate cleaned with hydrochloric acid whose concentration with respect to the cleaning solution as a whole had been brought to 0.5% by mass or more, and whose GaN substrate immersion time was 30 seconds or more.
It should be understood that although GaN substrates were employed in Experimental Example 2, the same results are believed to be obtained if other GaxIn1-xN substrates, apart from GaN substrates, are employed. Furthermore, GaN substrate thickness and plane orientation are not limited to above example, and even where the thickness and plane orientation are arbitrary, the same results as in Experimental Example 2 will be obtained.
The presently disclosed embodiments and implementation examples should in all respects be considered to be illustrative and not limiting. The scope of the present invention is set forth not by the foregoing description but by the scope of the patent claims, and is intended to include meanings equivalent to the scope of the patent claims and all modifications within the scope.
The present invention can be advantageously be exploited in the manufacture of semiconductor devices employing GaxIn1-xN substrates.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2006-214537 | Aug 2006 | JP | national |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/JP2007/062075 | 6/15/2007 | WO | 00 | 2/6/2009 |
