CN1716545A - AlxGayIn1-x-yN substrate and its cleaning method, AlN substrate and its cleaning method - Google Patents

Abstract

An Al x Ga y In 1-x-y N substrate in which particles having a grain size of at least 0.2 m on a surface of the Al x Ga y In 1-x-y N substrate are at most 20 in number when a diameter of the Al x Ga y In 1-x-y N substrate is two inches, and a cleaning method with which the Al x Ga y In 1-x-y N substrate can be obtained are provided. Further, an Al x Ga y In 1-x-y N substrate (51) in which, in a photoelectron spectrum of a surface of the Al x Ga y In 1-x-y N substrate (51) by X-ray photoelectron spectroscopy with a detection angle of 10 DEG , a ratio between a peak area of C 1s electrons and a peak area of N 1s electrons (C 1s electron peak area / N 1s electron peak area) is at most 3, and a cleaning method with which the Al x Ga y In 1-x-y N substrate can be obtained are provided. Still further, an AlN substrate (52) in which, in a photoelectron spectrum of a surface of the AlN substrate (52) by X-ray photoelectron spectroscopy with a detection angle of 10 DEG , a ratio between a peak area of Al 2s electrons and a peak area of N 1s electrons (Al 2s electron peak area / N 1s electron peak area) is at most 0.65 and a cleaning method with which the AlN substrate can be obtained are provided.

Description

AlxGayIn1-x-yN substrate and its cleaning method, AlN substrate and its cleaning method

technical field

The present invention relates to an Al _x Ga _y In _1-xy N (0<x≤1, 0<y≤1, x+y≤1) substrate and the Al _x Ga _y In _1-xy N (0<x ≤1, 0<y≤1, x+y≤1) cleaning method of the substrate. The present invention also relates to an AlN substrate capable of stably growing an epitaxial film at a low haze level, and a cleaning method for the AlN substrate. In this specification, Al _x Ga _y In _1-xy N (0<x ≤ 1, 0 < y ≤ 1, x+y ≤ 1) is abbreviated as Al _x Ga _y In _1-xy N.

Background technique

The _{AlxGayIn1 -xyN substrate} can be suitably used as a substrate of various semiconductor devices, such as a substrate of optical devices and/or electronic devices.

A representative growth method of Al _x Ga _y In _1-xy N crystal is the HVPE (Hydride Vapor Phase Epitaxy) method, and Al _x Ga _y In _1-xy N crystal can be produced from Al _x Ga _y In _{1- xy} N substrate. Semiconductor devices such _as optical devices and/or electronic devices can be obtained by growing various epitaxial films on the surface of _{AlxGayIn1 -xyN} substrates.

Among Al _x Ga _y In _1-xy N substrates, AlN substrates have an energy band gap of 6.2eV, thermal conductivity of about 3.3WK ^-1 cm ^-1 , and high electrical resistance, so they are used as various semiconductors Substrates for devices, such as substrates for optical and/or electronic devices, have attracted attention.

AlN substrates can be fabricated from AlN crystals grown by the HVPE method or the sublimation method. By growing various epitaxial films on the surface of an AlN substrate, semiconductor devices such as optical devices and/or electronic devices can be obtained.

For example, Toshio Nishida et al. in Appl. Phys. Lett., 2003, Volume 82, No. 1 "GaN-free transparent ultraviolet light-emitting diodes" disclose a method by growing an AlGaN film on an AlN substrate, etc. obtained light-emitting diodes. Also, "The Characteristics of UV-LED Grown on Bulk AlN Substrate Under Large Current Injection" by Toshio Nishida et al., 51st Spring Meeting of the Japan Society of Applied Physics and Related Societies, Extended Abstract, March 2004, p. 409, also A light emitting diode formed on a bulk AlN substrate is disclosed.

Contents of the invention

Summary of the invention

When an epitaxial film is grown on the surface of an _{AlxGayIn1 -xyN} substrate, a low- _quality epitaxial film with a large number of defects and/or tarnishes is sometimes grown. Semiconductor devices using such low-quality epitaxial films have poor device characteristics, so it is necessary to stably grow high-quality epitaxial films with few defects and/or tarnishing.

Therefore, in order to stably grow a high- _quality epitaxial film with few defects and/or tarnishing, particles and/or organic substances attached to the surface of the _{AlxGayIn1 -xyN} substrate have been removed by cleaning. However, since there are no conventional technical references _dealing with the degree of removal of particles and/or organics from the surface of _{AlxGayIn1 -xyN} substrates, and their criteria are _unclear , there are _{AlxGayIn1 -xyN} substrates Changes in surface conditions directly lead to the problem of changes in the quality of the epitaxial film.

In addition, when an epitaxial film is grown on the surface of an AlN substrate, an epitaxial film with a high haze level is sometimes grown. A semiconductor device using such an epitaxial film has poor device characteristics, so it is required to stably grow an epitaxial film with a low haze level.

Therefore, in order to stably grow an epitaxial film with a low haze level, the AlN substrate is already cleaned. However, since the standard of AlN substrate surface that can stably grow epitaxial film with low turbidity level is not clear, there is a problem that the change of _{AlxGayIn1 -xyN substrate} surface condition directly leads to the change of epitaxial film quality.

An object _of the present invention is to provide an _{AlxGayIn1 -xyN} substrate capable of _stably growing high-quality epitaxial films, and a cleaning method for obtaining the _{AlxGayIn1 -xyN} substrate. Another object of the present invention is to provide an AlN substrate capable of stably growing an epitaxial film at a low haze level, and a cleaning method for the AlN substrate.

The present invention relates to an _{AlxGayIn1 -xyN substrate ,} wherein when the diameter of the _{AlxGayIn1 -xyN} substrate is 2 inches, _one surface of the _{AlxGayIn1 -xyN} substrate The number of particles having an upper grain size of at least 0.2 μm is at most 20. Here, in this specification, the _{AlxGayIn1 -xyN substrate} refers to a nitride crystal substrate containing aluminum (Al), and it may contain gallium (Ga) and / or Indium (In).

In addition, the present invention relates to _a method for cleaning an _{AlxGayIn1 -xyN} substrate, wherein the _{AlxGayIn1 -xyN} substrate is immersed in a method _selected from ammonia water, ammonium hydroxide (ammonia hydroxide) )/hydrogen peroxide mixture and a solution in an aqueous organic alkali solution while ultrasonically sonicating, so that when the Al _x Ga _y In _1-xy N substrate has a diameter of two inches, the Al _x Ga _y The number of particles having a crystal grain size of at least 0.2 µm on one surface of the In _1-xy N substrate is at most 20.

Here, in the cleaning method of the _{AlxGayIn1 -xyN} substrate of the present invention, it is preferable to use ammonia water with an ammonia concentration of at least 0.5% by weight, a hydrogen _peroxide solution with a concentration of at least 0.1% by weight and an ammonia concentration of at least 0.1 weight percent ammonium hydroxide/hydrogen peroxide mixture, and an organic alkali aqueous solution with an organic alkali concentration of at least 0.5 weight percent are used as cleaning solutions.

In the _{AlxGayIn1 -xyN} substrate cleaning method of the present invention, preferably the aqueous organic base solution is an organic base dissolved in _water , and the organic base is tetramethylammonium hydroxide and 2-hydroxyethyltrimethylammonium hydroxide one of the.

In addition, in the _{AlxGayIn1 -xyN} substrate cleaning _method of the present invention, the soaking time of the _{AlxGayIn1 -xyN substrate} is at least 30 seconds.

In addition, the present invention relates to an Al _x Ga _y In _1-xy N substrate, wherein the photoelectron energy at the surface of the Al _x Ga _y In _1-xy N substrate obtained by X-ray photoelectron spectroscopy at a detection angle of 10° In the spectrum, the ratio between the peak area of C _1s electrons and the peak area of N _1s electrons (C _1s electron peak area/N _1s electron peak area) is at most 3.

In addition, the present invention relates to a cleaning method of an Al _x Ga _y In _1-xy N substrate, wherein the Al _x Ga _y In _1-xy N substrate is soaked in an acid solution, so that the X-ray photoelectron spectroscopy Method _{The ratio between the peak area of C 1s electrons} and the peak area _of N _1s electrons ( _{C 1s} electron peak area/N _1s electron peak area) is at most 3.

Here, in the cleaning _method of the _{AlxGayIn1 -xyN} substrate of the present invention, it is preferable that the acid solution is a solution made of at least one acid selected from hydrofluoric acid, hydrochloric acid and sulfuric acid, or is A solution made of a mixture of at least one acid selected from hydrofluoric acid, hydrochloric acid and sulfuric acid and a hydrogen peroxide solution.

Furthermore, in the cleaning method of the _{AlxGayIn1 -xyN} substrate of the present invention, when _the acid solution is a solution made of at least one acid selected from hydrofluoric acid, hydrochloric acid and sulfuric acid, the acid The total concentration of hydrofluoric acid, hydrochloric acid and sulfuric acid in the solution is at least 0.5% by weight, and when the acid solution is a solution made of a mixture of at least one acid selected from hydrofluoric acid, hydrochloric acid and sulfuric acid and a hydrogen peroxide solution When, preferably the total concentration of hydrofluoric acid, hydrochloric acid and sulfuric acid in the acid solution is at least 0.1 weight percent, while the concentration of hydrogen peroxide solution is at least 0.1 weight percent.

In addition, in the _{AlxGayIn1 -xyN} substrate _cleaning method of the present invention, it is preferable that the soaking time _of the _{AlxGayIn1 -xyN} substrate is at least 30 seconds.

The present invention relates to an AlN substrate, wherein in the photoelectron spectrum of the AlN substrate surface obtained by the X-ray photoelectron spectroscopy method at a detection angle of 10°, the peak area of Al _2s electrons and the peak area of N _1s electrons are between The ratio (Al _2s electron peak area/N _1s electron peak area) is at most 0.65.

In addition, the present invention relates to a cleaning method of an AlN substrate, wherein the AlN substrate is immersed in an acid solution so that the photoelectron spectrum of the AlN substrate surface obtained by the X-ray photoelectron spectroscopy method at a detection angle of 10° , the ratio between the Al _2s electron peak area and the N _1s electron peak area (Al _2s electron peak area/N _1s electron peak area) is at most 0.65.

Here, in the AlN substrate cleaning method of the present invention, it is preferable that the acid solution is an acid solution made of at least one acid selected from nitric acid, phosphoric acid, and acetic acid.

Furthermore, in the AlN substrate cleaning method of the present invention, it is preferable that the concentration of the acid solution is at least 0.5% by weight.

Furthermore, in the AlN substrate cleaning method of the present invention, it is preferable that the soaking time of the AlN substrate is at least 40 seconds.

According to the present invention, an _{AlxGayIn1 -xyN} substrate capable of stably growing high-quality epitaxial films and a cleaning _method for obtaining the _{AlxGayIn1 -xyN substrate} can be provided.

Furthermore, an AlN substrate capable of growing an epitaxial film at a low haze level and a cleaning method for obtaining the AlN substrate can be provided.

The above and other objects, features, aspects and advantages of the present invention will become more apparent with the following detailed description of the present invention in conjunction with the accompanying drawings.

Description of drawings

1 and 2 are schematic diagrams illustrating an example of an X-ray photoelectron spectrum with a detection angle of 10° according to the present invention.

FIG. 3 is a schematic cross-sectional view of a cleaning apparatus used in Example 1. FIG.

FIG. 4 shows the relationship between the number of particles and the number of defects in the epitaxial film grown on the surface of the AlN substrate in Example 1. FIG.

FIG. 5 is a schematic cross-sectional view of a cleaning apparatus used in Example 2. FIG.

FIG. 6 is a schematic cross-sectional view of a cleaning apparatus used in Example 3. FIG.

Fig. 7 shows the ratio (Al _2s electron peak area/N _1s electron peak area) between the Al _2s electron peak area and the N _1s electron peak area on the AlN substrate after cleaning in embodiment 3 and grows on this AlN substrate The relationship between the haze level of the epitaxial film on the surface.

DESCRIPTION OF THE PREFERRED EMBODIMENT

The present invention relates to an _{AlxGayIn1 -xyN substrate ,} wherein when the diameter of the _{AlxGayIn1 -xyN} substrate is 2 inches, _one surface of the _{AlxGayIn1 -xyN} substrate The number of particles having an upper grain size of at least 0.2 μm is at most 20. This is based on the present inventors' discovery that when particles having a grain size of at least 0.2 µm on the surface of an _{AlxGayIn1 -xyN} substrate are controlled within the above-mentioned _number , few defective High quality epitaxial film.

Here, the number of particles on the surface of the _{AlxGayIn1 -xyN substrate} is calculated as follows: For all particles _on the surface of the _{AlxGayIn1 -xyN} substrate with a grain size of at least 0.2 μm Counts were performed and the number of particles counted was converted to a value assuming a 2 inch diameter _{AlxGayIn1 -xyN substrate} . Therefore, in the present _invention , there is no limitation on the size of the _{AlxGayIn1 -xyN} substrate. For example _, an _{AlxGayIn1 -} xyN substrate with a diameter of 4 inches has four times the area of an _{AlxGayIn1 -xyN substrate} with a diameter of 2 inches. Therefore, when using an _{AlxGayIn1 -xyN} substrate with a diameter of 4 inches, 1/4 of the _total number of particles on its surface corresponds to the number of particles used here. It should be noted that particle counts are performed using well-known substrate surface inspection instruments such as light scattering schemes. In addition, there is no particular limitation on the material of the particles.

The present invention relates to a _cleaning method, wherein an _{AlxGayIn1 -xyN} substrate is soaked in a cleaning solution selected from ammonia water, ammonium hydroxide/hydrogen peroxide mixture, and an aqueous organic alkali solution while ultrasonically performing, thereby When the _{AlxGayIn1 -xyN} substrate has a diameter of two inches, the number of _particles having a crystal grain size of at least 0.2 µm on one surface of the _{AlxGayIn1 -xyN substrate} is at most 20.

Here, the ammonium hydroxide/hydrogen peroxide mixture refers to a mixture of hydrogen peroxide solution and ammonia water. In addition, the organic base aqueous solution refers to an organic base dissolved in water, and it is preferable to use one of tetramethylammonium hydroxide represented by the following structural formula (1) and 2-hydroxyethyltrimethylammonium hydroxide represented by the structural formula (2) as Organic base:

When ammonia water is used as the cleaning solution, it is preferred that the ammonia concentration is at least 0.5% by weight relative to the entire cleaning solution. When an ammonium hydroxide/hydrogen peroxide mixture is used as the cleaning solution, preferably the hydrogen peroxide solution has a concentration of at least 0.1 weight percent and the ammonia concentration is at least 0.1 weight percent relative to the total cleaning solution. When an organic base is used as the cleaning solution, it is preferred that the concentration of the organic base is at least 0.5% by weight relative to the entire cleaning solution. By defining the concentration of the cleaning solution as above, the number of particles _on the surface of the _{AlxGayIn1 -xyN} substrate tends to be more stably controlled as above.

Preferably, the immersion time _of the _{AlxGayIn1 -xyN} substrate in the cleaning solution is at least 30 seconds. In this case, since the _{AlxGayIn1 -xyN} substrate is sufficiently immersed in the cleaning solution, the number of particles on the surface of the _{AlxGayIn1 -xyN substrate tends} to be more stably controlled as above . Here, the immersion time of the Al _x Ga _y In _1-xy N substrate is calculated from the time point when the cleaning liquid is ultrasonicated.

Furthermore, the present invention relates to an Al _x Ga _y In _1-xy N substrate, wherein the photoelectron energy at the surface of the Al _x Ga _y In _1-xy N substrate obtained by X-ray photoelectron spectroscopy at a detection angle of 10° In the spectrum, the ratio between the peak area of C _1s electrons and the peak area of N _1s electrons (C _1s electron peak area/N _1s electron peak area) is at most 3. This is based on the inventor's finding that when the ratio between the peak area of C _1s electrons and that of N _1s electrons is controlled as above, a high-quality epitaxial film without any tarnishing can be grown. Here, the peak area _of C _1s electrons and the peak area of N _1s electrons in the photoelectron spectroscopy of _{AlxGayIn1 -xyN} substrate surface obtained by X-ray photoelectron spectroscopy (XPS) method at a detection angle of 10° The ratio between indicates _the amount of organic matter on the surface _of the _{AlxGayIn1 -xyN} substrate relative to the amount of nitrogen close to the surface of the _{AlxGayIn1 -xyN} substrate. By controlling its ratio as above, a high-quality epitaxial film without any tarnishing can be grown.

Here, C _1s electrons refer to electrons on C (carbon) 1s orbitals, and N _1s electrons refer to electrons on N (nitrogen) 1s orbitals. As shown in FIG. 1 , C _1s electrons and N _1s electrons on the surface of Al _x Ga _y In _1-xy N substrate 51 are released in the form of photoelectrons 71 by irradiation of X-rays 61 . Thereafter, photoelectrons 71 released in a _direction 10° from the surface of _{AlxGayIn1 -} xyN substrate 51 were detected by monitor 81 (detection angle 10°), thereby obtaining a photoelectron spectrum. Determine the ratio between the peak area of C _1s electrons and the peak area of N _1s electrons of this photoelectron spectrum.

Furthermore, the present invention relates to a cleaning method in which an _{AlxGayIn1 -} xyN substrate is _soaked in an acid solution so that _Alx obtained by X-ray photoelectron spectroscopy at a detection angle of 10° In the photoelectron spectroscopy of the Ga _y In _1-xy N substrate surface, the ratio between the peak area of C _1s electrons and the peak area of N _1s electrons (peak area of C _1s electrons/peak area of N _1s electrons) is at most 3.

Here, it is preferable that the acid solution is a solution of at least one acid selected from hydrofluoric acid, hydrochloric acid, and sulfuric acid. More preferably the acid solution is a mixture of at least one acid selected from hydrofluoric acid, hydrochloric acid and sulfuric acid and a hydrogen peroxide solution. In this case, the ratio between the peak area of C _1s electrons and the peak area of N _1s electrons in the above photoelectron spectrum tends to be more stably controlled to be at most 3.

In addition, when the acid solution is a solution made of at least one acid selected from hydrofluoric acid, hydrochloric acid and sulfuric acid, it is preferable that the total concentration of hydrofluoric acid, hydrochloric acid and sulfuric acid in the acid solution is at least 0.5 weight percent. Furthermore, when the acid solution is a solution made of a mixture of at least one acid selected from hydrofluoric acid, hydrochloric acid and sulfuric acid and a hydrogen peroxide solution, it is preferred that the total concentration of hydrofluoric acid, hydrochloric acid and sulfuric acid in the acid solution is at least 0.1% by weight, and the concentration of the hydrogen peroxide solution is at least 0.1% by weight. In this case, the ratio between the peak area of C _1s electrons and the peak area of N _1s electrons in the above photoelectron spectrum tends to be more stably controlled to be at most 3.

It is also preferred that the immersion _time of the _{AlxGayIn1 -xyN} substrate in the acid solution is at least 30 seconds. In this case, since the _{AlxGayIn1 -xyN substrate} is sufficiently soaked in the acid solution, the ratio between the peak area of C _1s electrons and that of N _1s electrons as above tends to be more stably Control as above.

Furthermore, the present invention relates to an AlN substrate in which the peak area of Al _2s electrons and the peak area of N _1s electrons in the photoelectron spectrum of the AlN substrate surface obtained by the X-ray photoelectron spectroscopy (XPS) method at a detection angle of 10° The ratio between the peak areas (Al _2s electron peak area/N _1s electron peak area) is at most 0.65. This is based on the inventor's finding that when the ratio between the Al _2s electron peak area and the N _1s electron peak area is controlled as above, an epitaxial film with a low haze level can be grown.

When growing an epitaxial film on an AlN substrate using the MOVPE method (Metal Organic Vapor Phase Epitaxy) or the MBE method (Molecular Beam Epitaxy), the surface of the AlN substrate is heated before growing the film. Since the volatilization of N (nitrogen) on the surface of the AlN substrate exceeds that of Al (aluminum) by heating the surface of the AlN substrate, the chemical composition of the surface of the AlN substrate contains a large amount of Al when growing an epitaxial film. Therefore, by making the AlN substrate surface contain a large amount of N in its chemical composition in advance, an epitaxial film with a low haze level can be grown. This is based on the inventors' discovery that an AlN substrate can be used in which Al The ratio between the peak area of _{the 2s} electrons and that of the N _1s electrons is at most 0.65, as a criterion for the AlN substrate surface.

Here, Al _2s electrons refer to electrons on Al (aluminum) 2s orbitals, and N _1s electrons refer to electrons on N 1s orbitals. As shown in FIG. 2 , through the irradiation of X-rays 62 , Al _2s electrons and N _1s electrons on the surface of AlN substrate 52 are released in the form of photoelectrons 72 . Thereafter, photoelectrons 72 released in a direction 10° from the surface of the AlN substrate 52 are detected by a monitor 82 (detection angle 10°), thereby obtaining a photoelectron spectrum. Determine the ratio between the peak area of Al _2s electrons and the peak area of N _1s electrons of this photoelectron spectrum.

Furthermore, the present invention relates to a cleaning method in which an AlN substrate is immersed in an acid solution so that in the photoelectron spectrum of the AlN substrate surface obtained by the X-ray photoelectron spectroscopy method at a detection angle of 10°, Al _2s electrons The ratio between the peak area of Al and the peak area of N _1s electrons (Al _2s electron peak area/N _1s electron peak area) is at most 0.65.

Here, it is preferable that the acid solution is a solution made of at least one acid selected from hydrofluoric acid, hydrochloric acid, and sulfuric acid, or a mixed solution made of at least two of these acids. In this case, the ratio between the peak area of Al _2s electrons and the peak area of N _1s electrons as above tends to be more stably controlled to be at most 0.65.

In addition, it is preferred that the concentration of the acid solution is at least 0.5% by weight relative to the entire acid solution. In this case, the ratio between the peak area of Al _2s electrons and the peak area of N _1s electrons in the photoelectron spectroscopy as above tends to be more stably controlled to be at most 0.65.

Preferably, the immersion time of the AlN substrate in the acid solution is at least 40 seconds. In this case, since the AlN substrate is sufficiently soaked in the acid solution, the ratio between the peak area of Al _2s electrons and the peak area of N _1s electrons tends to be more stably controlled to be at most 0.65 as above.

Detailed ways

implementation plan

Example 1

First, 50 AlN substrates with a diameter of 2 inches were prepared, each of which was obtained by mirror-polishing AlN crystals grown by the HVPE method, and then removing the damaged layer due to the mirror-polishing. Here, the thickness of each of 50 AlN substrates was 400 µm, and the AlN substrate surface was a plane deviated from the orientation (0001) by 2°.

Next, 50 AlN substrates were cleaned with different immersion times by using the cleaning apparatus shown in the schematic cross-sectional view of FIG. 3 . Here, the cleaning bath 13 shown in FIG. 3 is filled with tetramethylammonium hydroxide aqueous solutions of various concentrations as the cleaning solution 23 . Ultrasonic waves 33 having a frequency of 900 kHz were applied to the cleaning solution 23 in which the AlN substrates 43 were soaked, and the ultrasonic conditions were the same for each of the 50 AlN substrates 43 .

Then, for each cleaned AlN substrate, the particles with a grain size of at least 0.2 μm on the surface of the AlN substrate were counted with the substrate surface inspection instrument of the light scattering scheme.

Thereafter, under the same conditions, an epitaxial film formed of AlN crystal was grown to a thickness of 1 µm on each of the substrate surfaces of 50 AlN substrates by the MOVPE method (Metal Organic Vapor Phase Epitaxy). Then, the number of defects in the epitaxial film was counted with the same substrate surface inspection instrument as the above instrument.

Figure 4 shows the results of this experiment. In Fig. 4, the abscissa indicates the number of particles with a grain size of at least 0.2 μm on the cleaned AlN substrate surface counted in the above-mentioned manner, while the ordinate indicates the counted AlN substrate surface corresponding to the number of particles on the abscissa The number of defects in epitaxial films grown on

As can be seen from FIG. 4, when the number of particles with a grain size of at least 0.2 μm on the surface of an AlN substrate with a diameter of 2 inches is at most 20, the number of defects in the epitaxial film grown on the surface is less than 50. Therefore, compared with the case where the number of particles exceeds 20, a high-quality epitaxial film with fewer defects is obtained.

an AlN substrate having a particle size of at least 0.2 μm on the surface and a population of at most 20 particles is cleaned with a cleaning solution having a concentration of tetramethylammonium hydroxide of at least 0.5 percent by weight relative to the entire cleaning solution, and Soak time is at least 30 seconds.

Although an AlN substrate was used in Embodiment 1 above, it is considered that similar results can be obtained using an _{AlxGayIn1 -xyN substrate} other than an AlN substrate. In addition, as described above, there are no restrictions on the thickness and plane orientation of the AlN substrate, and results similar to those of Embodiment 1 can be obtained at any values thereof.

Example 2

First, similarly to Example 1, 50 AlN substrates with a diameter of 2 inches were prepared, each of which was obtained by mirror-polishing an AlN crystal and then removing a damaged layer due to the mirror-polishing. Here, the thickness of each of 50 AlN substrates was 400 µm, and the AlN substrate surface was a plane deviated from the orientation (0001) by 2°.

Next, 50 AlN substrates were cleaned with different immersion times by using the cleaning apparatus shown in the schematic cross-sectional view of FIG. 5 . Here, the cleaning bath 15 shown in FIG. 5 is filled with hydrochloric acid of various concentrations as the acid solution 25 to soak the respective AlN substrates 45 .

Then, for each cleaned AlN substrate, use an X-ray photoelectron spectrometer, using Kα rays of Mg as the X-ray source, measure the photoelectron spectrum of the AlN substrate surface at a detection angle of 10°, and calculate the C _1s electron The ratio between the peak area and the peak area of N _1s electrons (C _1s electron peak area/N _1s electron peak area).

Thereafter, under the same conditions, an epitaxial film formed of AlN crystal was grown to a thickness of 1 µm on each of the substrate surfaces of 50 AlN substrates by the MOVPE method. Then, for each of the epitaxial films grown in this way, the presence or absence of tarnish was visually evaluated based on the following criteria. For each category shown in Table 1 (C _1s electron peak area/N _1s electron peak area), the number of epitaxial films with tarnishing was counted, and the results are shown in Table 1.

Table 1 (C _1s electron peak area/N _1s electron peak area) up to 3 greater than 3 and at most 5 greater than 5 Number of AlN substrates with tarnish/total number of AlN substrates 0/15 5/27 7/8

Evaluation criteria for the presence of tarnish

Tarnish exists - a part of the epitaxial film is not polished to become a mirror surface

There is no tarnish - the epitaxial film is all polished to a mirror surface

As can be seen in Table 1, there is a tendency that the smaller the ratio between the peak area of C _1s electrons and the peak area of N _1s electrons (peak area of C _1s electrons/peak area of N _1s electrons), the less tarnish. Especially when the ratio is at most 3, there is no tarnish on the epitaxial film, and a high-quality epitaxial film is grown.

AlN substrates having a ratio between the peak area of C _1s electrons and the peak area of N _1s electrons (peak area of C _1s electrons/peak area of N _1s electrons) of at most 3 are cleaned with a cleaning solution in which The concentration of hydrochloric acid is at least 0.5% by weight relative to the entire cleaning solution, and the soaking time is at least 30 seconds.

Although an AlN substrate was used in Embodiment 2 above, it is considered that similar results can be obtained using an _{AlxGayIn1 -xyN substrate} other than an AlN substrate. In addition, as described above, there are no restrictions on the thickness and plane orientation of the AlN substrate, and results similar to those of Embodiment 2 can be obtained at any values thereof.

Example 3

First, 50 AlN substrates with a diameter of 2 inches were prepared, each of which was obtained by mirror-polishing an AlN crystal grown by the HVPE method, and then removing a damaged layer due to the mirror-polishing. Here, the thickness of each of 50 AlN substrates was 400 µm, and the AlN substrate surface was a plane deviated from the orientation (0001) by 2°.

Next, each of the 50 AlN substrates was cleaned using the cleaning apparatus shown in the schematic sectional view of FIG. 6 . Here, the cleaning bath 16 shown in FIG. 6 is filled with mixed solutions of nitric acid, phosphoric acid and acetic acid of various concentrations as the cleaning solution 26, and the AlN substrate 46 is cleaned with different cleaning times.

Then, for each cleaned AlN substrate, use an X-ray photoelectron spectrometer to measure the photoelectron spectrum of the AlN substrate surface at a detection angle of 10° using the Kα ray of Mg as the X-ray source, and calculate the Al _2s electron The ratio between the peak area and the peak area of N _1s electrons (Al _2s electron peak area/N _1s electron peak area).

Thereafter, under the same conditions, an epitaxial film formed of AlN crystal was grown to a thickness of 1 µm on each of the substrate surfaces of 50 AlN substrates by the MOVPE method. Each epitaxial film grown in this way was then evaluated for haze levels using the Light Scattering Protocol's Substrate Surface Inspection instrument. The results are shown in Figure 7. In Fig _. 7, the abscissa represents the ratio (Al _2s electron peak area/N _1s electron peak area/N _1s Electronic peak area), while the vertical axis represents the haze level of the epitaxial film grown on each AlN substrate with the value shown on the abscissa (Al _2s electronic peak area/N _1s electronic peak area).

As can be seen from FIG. 7 , there is a tendency that the smaller the ratio between the peak area of Al _2s electrons and the peak area of N _1s electrons (the peak area of Al _2s electrons/the peak area of N _1s electrons), the lower the turbidity level of the epitaxial film. Especially when the ratio is at most 0.65, the haze level of the epitaxial film is less than 2 ppm, and the grown epitaxial film is excellent.

The AlN substrate whose ratio between the Al _2s electron peak area and the N _1s electron peak area (Al _2s electron peak area/N _1s electron peak area) is at most 0.65 is cleaned with an acid solution in which nitric acid, The concentration of phosphoric acid and acetic acid is at least 0.5% by weight relative to the whole cleaning solution, and the soaking time is at least 40 seconds.

Although a mixed solution of nitric acid, phosphoric acid, and acetic acid was used in the above-mentioned Example 3, it is considered that a similar solution can be obtained using one of nitric acid, phosphoric acid, and acetic acid, or a mixed solution made of two of these acids. the result of.

In addition, as described above, there are no restrictions on the thickness and plane orientation of the AlN substrate, and results similar to those of Embodiment 3 can be obtained at any values thereof.

The present invention can be applied to the manufacture of semiconductor devices using _{AlxGayIn1 -xyN substrates} . Furthermore, the present invention is also applicable to the manufacture of semiconductor devices using AlN substrates.

Although the present invention has been described and illustrated in detail, it should be clearly recognized that the above illustrations and embodiments are only illustrations and examples, not limiting, and the spirit and scope of the present invention are only limited by the terms of the appended claims.

Claims (15)

1, a kind of Al _xGa _yIn _1-x-yThe N substrate, wherein

As described Al _xGa _yIn _1-x-yWhen the diameter of N substrate is 2 inches, described Al _xGa _yIn _1-x-yThe population that grain size is at least 0.2 μ m on surface of N substrate is at most 20.

2, a kind of Al _xGa _yIn _1-x-yThe cleaning method of N substrate (43), wherein

With described Al _xGa _yIn _1-x-yN substrate (43) is immersed in a kind of cleaning solution of being made by the solution that is selected from ammoniacal liquor, ammonium hydroxide/hydrogen peroxide mixture and the organic base aqueous solution, carries out ultrasonic (33) simultaneously, thereby as this Al _xGa _yIn _1-x-yWhen the diameter of N substrate (43) is two inches, make Al _xGa _yIn _1-x-yThe population that grain size is at least 0.2 μ m on (43) surfaces of N substrate is at most 20.

3, the Al of claim 2 _xGa _yIn _1-x-yThe cleaning method of N substrate (43), wherein

Use ammonia concentration to be at least the ammoniacal liquor of 0.5 percentage by weight, hydrogenperoxide steam generator concentration is at least the ammonium hydroxide/hydrogen peroxide mixture that 0.1 percentage by weight and ammonia concentration are at least 0.1 percentage by weight, and organic base concentration is at least the organic base aqueous solution of 0.5 percentage by weight as described cleaning solution (23).

4, the Al of claim 2 _xGa _yIn _1-x-yThe cleaning method of N substrate (43), wherein

The described organic base aqueous solution is the organic base that is dissolved in the water, and described organic base is in tetramethylammonium hydroxide and the hydroxide 2-ethoxy trimethylammonium.

5, the Al of claim 2 _xGa _yIn _1-x-yThe cleaning method of N substrate (43), wherein

Described Al _xGa _yIn _1-x-yThe soak time of N substrate (43) was at least 30 seconds.

6, a kind of Al _xGa _yIn _1-x-yN substrate (51), wherein

Detecting the described Al that the angle obtains by the X-ray photoelectron spectroscopy method with 10 ° _xGa _yIn _1-x-yIn the photoelectron spectroscopy on N substrate (51) surface, C _1sThe peak area of electronics and N _1sRatio (C between the peak area of electronics _1sElectronics peak area/N _1sThe electronics peak area) is at most 3.

7, a kind of Al _xGa _yIn _1-x-yThe cleaning method of N substrate (45), wherein

With described Al _xGa _yIn _1-x-yN substrate (45) is immersed in the acid solution (25), thereby is detecting the Al that the angle obtains by the X-ray photoelectron spectroscopy method with 10 ° _xGa _yIn _1-x-yIn N substrate (45) the surface light electron spectrum, make C _1sThe peak area of electronics and N _1sRatio (C between the peak area of electronics _1sElectronics peak area/N _1sThe electronics peak area) is at most 3.

8, the Al of claim 7 _xGa _yIn _1-x-yThe cleaning method of N substrate (45), wherein

Described acid solution (25) is to be made by at least a acid that is selected from hydrofluoric acid, hydrochloric acid and the sulfuric acid, or make by the mixture that is selected from least a acid in hydrofluoric acid, hydrochloric acid and the sulfuric acid and hydrogenperoxide steam generator.

9, the Al of claim 8 _xGa _yIn _1-x-yThe cleaning method of N substrate (45), wherein

When described acid solution (25) be by be selected from that at least a acid in hydrofluoric acid, hydrochloric acid and the sulfuric acid makes the time, the total concentration of hydrofluoric acid, hydrochloric acid and sulfuric acid is at least 0.5 percentage by weight in the described acid solution (25), and

When described acid solution (25) be by the mixture that is selected from least a acid in hydrofluoric acid, hydrochloric acid and the sulfuric acid and hydrogenperoxide steam generator make the time, the total concentration of hydrofluoric acid, hydrochloric acid and sulfuric acid is at least 0.1 percentage by weight in the acid solution (25), and the concentration of hydrogenperoxide steam generator is at least 0.1 percentage by weight simultaneously.

10, the Al of claim 7 _xGa _yIn _1-x-yThe cleaning method of N substrate (45), wherein

Described Al _xGa _yIn _1-x-yThe soak time of N substrate (45) was at least 30 seconds.

11, a kind of AlN substrate (52), wherein

In the photoelectron spectroscopy on AlN substrate (52) surface of detecting the angle acquisition by the X-ray photoelectron spectroscopy method with 10 °, Al _2sThe peak area of electronics and N _1sRatio (Al between the peak area of electronics _2sElectronics peak area/N _1sThe electronics peak area) is at most 0.65.

12, the cleaning method of a kind of AlN substrate (46), wherein

Described AlN substrate (46) is immersed in the acid solution (26), thereby in the photoelectron spectroscopy on AlN substrate (46) surface of detecting the angle acquisition by the X-ray photoelectron spectroscopy method with 10 °, makes Al _2sThe peak area of electronics and N _1sRatio (Al between the peak area of electronics _2sElectronics peak area/N _1sThe electronics peak area) is at most 0.65.

13, the cleaning method of the AlN substrate (46) of claim 12, wherein

Described acid solution (26) is to be made by at least a acid that is selected from nitric acid, phosphoric acid and the acetic acid.

14, the cleaning method of the AlN substrate (46) of claim 12, wherein

The concentration of described acid solution (26) is at least 0.5 percentage by weight.

15, the cleaning method of the AlN substrate (46) of claim 12, wherein

The soak time of described AlN substrate (46) was at least 40 seconds.

Images