KR20170121182A - III-Nitride of Reduced Cracks Selecting and Growing Seeds for Bulk Crystals - Google Patents

Abstract

In one example, the present invention is directed to a method of measuring the X-ray oscillation curve of a seed crystal, comprising: (a) measuring an X-ray oscillation curve of the seed crystal at one or more points, (b) quantifying the peak width of the measured X- A method of growing a bulk crystal of a Group III nitride using an evaluated seed crystal. The present invention also includes a method for selecting seed crystals for growing bulk crystals of a Group III nitride. Bulk crystals of a Group III nitride can be grown in a supercritical ammonia or Group III metal melt using at least one seed selected by the method described above.

Description

Selecting and growing seeds for Group-III nitride bulk crystallization of reduced cracks

Cross reference of related application

This application claims priority from U.S. Patent Application No. 62 / 106,709, entitled " METHOD FOR SELECTION AND GROWTH METHOD FOR GROUP DELIVERY OF CIRCLE III-NITRIDE BULK CRYSTALS "filed on Jan. 22, 2015 by the inventors Tadao Hashimoto, And is attorney docket number SIXPOI-024USPRV1, which also relates to:

PCT Patent Application No. US2005 / 024239 filed on Jul. 08, 2005, inventors Kenji Fujito, Tadao Hashimoto and Shuji Nakamura entitled " Method for Growing Group III-Nitride Crystals in Supercritical Ammonia Using Autoclaves, Document No. 30794.0129-WO-Ol (2005-339-1);

U.S. Patent Application No. 11 / 784,339, filed on Apr. 06, 2007, inventors Tadao Hashimoto, Makoto Saito and Shuji Nakamura entitled " Method for growing large surface area gallium nitride crystals in supercritical ammonia and large surface area gallium nitride crystals " , Attorney Docket No. 30794.179-US-Ul (2006-204), filed on Apr. 04, 2006 by Tadao Hashimoto, Makoto Saito and Shuji Nakamura, Growth method and large surface area gallium nitride crystals ", U.S. Provisional Patent Application No. 60 / 790,310, Attorney Docket No. 30794.179-US-Pl (2006-204);

U.S. Patent Application No. 60 / 973,602 filed on Sep. 19, 2007, inventors Tadao Hashimoto and Shuji Nakamura entitled " Gallium Nitride Bulk Crystals and Their Growth Methods ", Attorney Docket No. 30794.244-US- 809-1);

U.S. Patent Application No. 11 / 977,661, filed on October 25, 2007, inventor Tadao Hashimoto, entitled " Method for Growing Group III-Nitride Crystals in a Mixture of Supercritical Ammonia and Nitrogen and the Group III- , Current US registration number 7,803,344, attorney docket number 30794.253-US-Ul (2007-774-2);

U.S. Patent Application No. 61 / 067,117, filed February 25, 2008, inventor Tadao Hashimoto, Edward Litz and Masanori Ikari, "Method of Manufacturing Group III-Nitride Wafer and Group III-Nitride Wafer" 8,728,234, Attorney Docket No. 62158-30002.00 or SIXPOI-003;

U.S. Patent Application No. 61 / 058,900 filed on Jun. 4, 2008, Inventor Edward Litz, Tadao Hashimoto and Ikari Masanori, "Group III-nitride with improved crystallinity from the initial Group III- Method of Manufacturing Crystalline ", current U.S. Registration No. 8,728,234, Attorney Docket No. 62158-30004.00 or SIXPOI-002;

U.S. Patent Application No. 61 / 058,910, filed Jun. 04, 2008, inventors Tadao Hashimoto, Edward Litz and Masanori Ikari, entitled "Group III nitride crystals grown in high pressure vessels and high- Growth methods and Group-III nitride determinations ", current U.S. Patent Nos. 8,236,267 and 8,420,041, Attorney Docket No. 62158-30005.00 or SIXPOI-005;

U.S. Patent Application No. 61 / 131,917, June 12, 2008, inventors Tadao Hashimoto, Masanori Ikari and Edward Litz, "III-nitride wafers with test method and test data for III-nitride wafers" No. 8,357,243, 8,577,043 and 8,585,822, Attorney Docket No. 62158-30006.00 or SIXPOI-001;

U.S. Patent Application No. 61 / 106,110, Oct. 16, 2008, inventor, Tadao Hashimoto, Masanori Ikari and Edward Litz, "Design of Reactor for Group III Group Nitride Crystal Growth and Growth of Group III-Nitride Crystals" Document No. SIXPOI-004;

U.S. Patent Application No. 61 / 705,540, filed on Sep. 25, 2012, by Tadao Hashimoto, Edward Litz and Sierra Hoff, entitled "Method for Growing Group III Nitride Crystals," U.S. Patent No. 9,202,872, No. SIXPOI-014;

All of which are incorporated herein by reference in their entirety, as if fully set forth below.

Field of invention

The present invention relates to bulk crystals of semiconductor materials used to produce semiconductor wafers for a variety of devices including optoelectronic devices such as light emitting diodes (LEDs) and laser diodes (LDs) and electronic devices such as transistors. More specifically, the present invention provides a bulk crystal of a Group III nitride such as gallium nitride. The present invention also provides a method of selecting seed crystals for growth of Group-III nitride bulk crystals.

Explanation of existing technology

The present application refers to named publications and patents as indicated by numbers in square brackets, e.g., [x]. The following is a list of these publications and patents:

[1] R. Dwilinski, R. Doradzinski, J. Garczynski, L. Sierzputowski, Y. Kanbara, U.S. Pat. Patent No. 6,656,615.

[2] R. Dwilinski, R. Doradzinski, J. Garczynski, L. Sierzputowski, Y. Kanbara, U.S. Pat. Patent No. 7,132,730.

[3] R. Dwilinski, R. Doradzinski, J. Garczynski, L. Sierzputowski, Y. Kanbara, U.S. Pat. Patent No. 7,160,388.

[4] K. Fujito, T. Hashimoto, S. Nakamura, International Patent Application No. PCT / US2005 / 024239, W007008198.

[5] T. Hashimoto, M. Saito, S. Nakamura, International Patent Application No. PCT / US2007 / 008743, W0071l7689. See also US20070234946, U.S. Pat. Application Serial Number 11 / 784,339 filed April 6, 2007.

[6] D 'Evelyn, U.S. Patent No. 7,078,731.

[7] Wang et al., Journal of Crystal Growth volume 318 (2011) pl 030.

Each of the above-listed references cited herein are incorporated by reference in their entirety, particularly as expressed herein, in connection with the description of how to make and use a Group III nitride substrate.

Gallium nitride (GaN) and its associated Ill-nitride alloys are key materials for a variety of optoelectronic and electronic devices such as LEDs, LDs, microwave power transistors and solar-blind photodetectors. Currently, LEDs are widely used in displays, indicators, general lighting, and LDs are used in data storage disk drives. However, most of these devices are epitaxially grown on heterogeneous substrates such as sapphire and silicon carbide because GaN substrates are extremely expensive compared to these heteroepitaxial substrates. Heteroepitaxial growth of Group-III nitrides results in highly defective or even cracked films, which hinders the realization of high-end optical and electronic devices such as high brightness LEDs or high power microwave transistors for general illumination .

In order to solve the fundamental problem caused by heteroepitaxy, it is essential to use a crystalline Group-III nitride wafer sliced from a bulk Group-III nitride crystal ingot. In most devices, a crystalline GaN wafer is preferred because it is relatively easy to control the conductivity of the wafer and the GaN wafer will provide the least lattice / thermal mismatch with the device layers. However, due to the high nitrogen vapor pressure at high melting point and elevated temperature, it was difficult to grow the GaN crystal ingot. Currently, most of the commercially available GaN substrates are produced by hydride vapor phase epitaxy (HVPE). HVPE is one of the vapor-phase methods, and it is difficult to reduce the dislocation density to less than 10 ⁵ cm ^-2 .

Ammonothermal growth, flux growth, and high temperature melting growth have been developed to obtain high quality GaN substrates with dislocation density less than 10 ⁵ cm ^-2 . Amorphous growth of group III nitride crystals in supercritical ammonia [1-6]. The flux method and high-temperature melting growth utilize the dissolution of group III metals.

Recently, a high-quality GaN substrate having a dislocation density of less than 10 < ^{5 &} gt ^; cm < ^-2 > Since thermal annealing can produce pure bulk crystals, one or more thick crystals can grow and can be sliced to produce GaN wafers. We have developed bulk crystallization of GaN by amorphization. However, we have found that it is difficult to avoid cracking of bulk crystals when the total thickness of bulk crystals exceeds 1 mm. We believe that the cracking problem of bulk Group III nitride is a universal problem of any bulk growth method, including cancer thermodynamics. Therefore, the present invention aims to obtain crack-free bulk Group-III nitride crystals using any bulk growth method such as growth in supercritical ammonia or melt of a Group III metal.

In one example, the present invention is directed to a method of determining the peak amplitude of a measured X-ray oscillation curve, comprising: (a) measuring an X-ray rocking curve of a seed crystal at one or more points, Lt; RTI ID = 0.0 > III-nitride < / RTI > The present invention also includes a method of selecting seed crystals for growing bulk crystals of a Group III nitride.

Reference is now made to the drawings, wherein like reference numerals throughout the following figures represent corresponding parts.
1 is an example of a process flow of the present invention.
FIG. 2 is a graph showing the half width (FWHM) (square point) of the 201 X ray oscillation curve in the seed crystal, the half width (diamond point) of the 201 X ray oscillation curve in the bulk GaN crystal using the corresponding seed, Fig. (a) is for a seed of a distributed distribution of FWHM, and (b) is for a seed of a less distributed distribution of FWHM. The zero point is approximately at the center of the seed plane along the longest line on the m plane. Exemplary XRD data was collected at several points along this line across the face of the seed crystal.

summary

The bulk crystals of the present invention are sliced to produce a Group III nitride wafer suitable for manufacturing a variety of optoelectronic and electronic devices such as LEDs, LDs, transistors and photodetectors by generally known techniques. Many optoelectronic and electronic devices are fabricated using thin films of Group III nitride alloys (i.e., alloys of GaN, AiN, and InN). Group-III nitride alloys are generally represented by Ga _x Al _y In _x N (0 _? X? 1, 0? _Y? 1). Since Group III metal elements (i.e., Al, Ga, In) exhibit similar chemical properties, the nitrides of these Group III elements make alloys or solid solutions. In addition, the crystal growth characteristics of these Group-III nitride are very similar.

Because of the limited solubility and high cost of Group-III nitride single crystal substrates, these devices are being fabricated on so-called heteroepitaxial substrates such as sapphire and silicon carbide. Because the heteroepitaxial substrate is chemically and physically different from the Group III nitride, the device generally has a high dislocation density (10 < ⁸ > -10 < ^{10 &} gt ^; cm <" ² >) at the interface between the heteroepitaxial substrate and the device layer. Since such dislocations deteriorate the performance and reliability of the device, a substrate composed of a crystalline Group-III nitride such as GaN and AlN is preferable.

Currently, most commercially available GaN substrates are produced with HVPE which is difficult to reduce the dislocation density to below 10 < ^{5 &} gt ^; cm <" 2 >. The dislocation density of the HVPE-GaN substrate is several times lower than that of the GaN film on the heteroepitaxial substrate, but dislocation density is still several times higher than typical silicon devices of electronic devices. In order to improve the performance of the device, the dislocation density should be lower.

Ammonothermal growth using supercritical ammonia was developed to make the dislocation density less than 10 ⁵ cm ^-2 . Calorimetry can produce GaN substrates with a dislocation density less than 10 < ^{5 &} gt ^; cm <" ² >. One of the advantages of amalgamation is the ability to grow bulk crystals with thicknesses greater than 1 mm. Calorimetry can be used to grow crystals with various dopants such as donors (i.e., electrons), acceptors (i.e., holes) or magnetic dopants. However, it is difficult to obtain a bulk crystal exceeding 1 mm in thickness without cracking. For example, Wang et al. [7] have initiated a procedure to evaluate the FWHM of an X-ray oscillation curve to select a good seed. Nevertheless, we experienced cracking problems in the selection process. Causes and mechanisms of crack formation are not well known, but possible causes may be stress scales within the crystal due to some discrepancy in thermal expansion coefficient between seed crystals and growth crystals or other physical properties. It is necessary to obtain a bulk crystal of a Group-III nitride free from cracks in order to produce a crack-free Group-III nitride substrate.

Description of the invention

In an effort to reduce or eliminate cracks within the bulk crystal of a Group III nitride having a thickness greater than 1 mm, the present invention contemplates that (a) the seed crystal measures the X-ray oscillation curve of the seed crystal at one or more points, b) quantifying the peak width of the measured x-ray oscillation curve, and c) evaluating the distribution of the quantified peak width. Figure 1 shows the process flow of the present invention.

First, a seed crystal for growing Group-III nitride crystal such as GaN is prepared. The seed crystal is preferably a Group III nitride single crystal such as GaN. The orientation of the seed crystal may be c-plane, a-plane, m-plane or other semipolar planes, although c-plane crystals are preferred. The monocrystalline seeds are formed by hydride vapor-phase epitaxy (HVPE), molecular beam epitaxy (MBE), metal organic vapor-phase epitaxy (MOVPE) Solution growth or other methods.

The seed crystal is then measured with an X-ray diffractometer to obtain the oscillation curve at one or more points of the seed crystal. An example of selecting a measurement location is a straight line along one crystallographic orientation such as the m-direction or the a-direction. Another example is to select a point within or at the intersection of a square grid drawn on the face of the seed. Another example is to obtain a statistically significant number of random measurements of the seed crystal structure on the seed surface.

If Group-III nitride crystals such as c-plane GaN are used, non-axis rotations such as 201 and 102 reflections are preferably used. This is because the non-shrinkage reflections are known to be more sensitive to the quality of seed crystals for bulk crystal growth. As a result, a direction (e.g., c-plane, m-plane, a-plane) that is more sensitive to the crystal structure of the seed crystal is first determined for the specific seed used, and then the crystal structure at various points It is helpful to use that direction to measure quality.

Although FWHM is commonly used to quantify the peak width of the X-ray oscillation curve, other methods of quantifying the peak width are also used. As generally known, the peak width of the X-ray oscillation curve indicates the quality of the microstructure of the crystal. Peak width is generally measured in arcsec, arcmin, or degree units.

In order to evaluate the distribution of the peak width, statistical values such as standard deviation can be used. Alternatively, the peak width data can be plotted and the distribution of the data visually identified. The size of the data distribution can be evaluated as an absolute value using units of arcsec, arcmin, or degree. Alternatively, the magnitude of the data variance can be evaluated against a representative value such as the average value of the entire data.

If the standard deviation and the average value are used to select a good seed crystal, the standard deviation is preferably less than 30% of the average value, more preferably less than 20% of the average value, or more preferably less than 10% of the average value.

The selected seed crystals are used to grow Group III nitride bulk crystals such as bulk GaN. The most preferred seed orientation and polarity can be selected depending on the growth method of the bulk crystal. For example, when growing bulk crystals in supercritical ammonia, nitrogen-polarized c-plane GaN is preferably used.

Example 1

Single crystal GaN seed crystals with a c-plane basal plane were prepared with HVPE. The thickness of the GaN seed was about 430 microns. The X-ray shake curve at 201 reflection was recorded at multiple points on the nitrogen polar side of the seed crystal. Measurements were made at point intervals of 0.5 mm along the m-direction. The peak width was quantified using FWHM in units of arcsec. The square points in FIG. 2 (a) show the FWHM at each measurement point. As shown in FIG. 2 (a), the FWHM value has a large variance. The average value of FWHM was 78 seconds (arcsec), and the standard deviation was 29 seconds (arcsec), which was 37% of the average value. The variance of the data appears throughout the scan line.

Subsequently, bulk crystals of GaN were grown using supercritical ammonia in a high pressure reactor. The chamber in the high-pressure reactor was divided into lower and upper portions using a baffle plate. Approximately 15 g of polycrystalline GaN was used as a nutrient and 3.1 g of sodium was used as a mineralizer. The mineralizer and seed crystals are located at the bottom of the high pressure reactor and the nutrients are located at the top of the high pressure reactor. The autoclave is then sealed, pumped under vacuum and filled with anhydrous liquid ammonia. The volumetric ammonia charge coefficient was about 53%. The autoclave was heated to about 510-520 ° C to allow crystal growth of GaN on the seed. After sufficient time, the ammonia was released and the autoclave was cooled. The resulting bulk GaN crystal has a thickness of about 5 mm.

The X-ray vibration curves at 201 reflection were measured at multiple points on the surface of the grown bulk GaN crystal, as described above and in Example 2. FWHM is shown in Figure 2 (a) as a diamond point. As shown, the FWHM in the grown bulk crystals also showed large dispersion. The mean value of FWHM was 89 seconds (arcsec), the standard deviation was 38 seconds (arcsec), and the mean value was 43%. The bulk crystal is then sliced on the wafer with a multi-wire saw. 2 (a) is a view of a sliced wafer. The wafer has a large number of cracks.

Example 2

As in Example 1, the c-plane GaN seed crystals were prepared as HVPE. The thickness of the GaN seed was about 430 microns. The X-ray oscillation curve at 201 reflection was recorded at multiple points on the nitrogen polar side of the seed crystal. Measurements were made at 0.5 mm intervals along the m-direction. Peak width was quantified as FWHM in units of arcsec. The square points in FIG. 2 (b) represent the FWHM of each measurement site. As shown in FIG. 2 (b), the FWHM value has a small variance. The mean value of FWHM was 41 seconds (arcsec), and the standard deviation was 7 seconds (arcsec). The mean value was 17%.

The bulk GaN was then grown on the seed crystal in a similar manner as in Example 1. The FWHM of the 201 X ray oscillation curve at multiple points on the grown bulk crystal is shown in FIG. 2 (b) as a diamond point and shows a small data variance. The mean value of FWHM was 48 seconds (arcsec), the standard deviation was 18 seconds (arcsec), and the mean value was 38%. As can be seen between the +12 and +16 mm positions in FIG. 2 (b), the large standard deviation is due to the edge effect of the measurement and is the cause of the lower reliability of the XRD data. The center portion of the wafer has less data dispersion. As shown in Fig. 2 (b), the wafer sliced in this bulk GaN crystal was greatly reduced in the cracks. The crack density was less than 1 cm ^-2 . Comparing Example 1 and Example 2, we found that there is a strong correlation between the dispersion of the peak width data of the X-ray oscillation curve of the seed crystal and the crack density of the bulk crystal using the seed.

The evaluation of the data variance can be performed in combination with standard deviations, visual judgments, and other criteria. For example, in this example (Fig. 2 (b)), if the central portion of the data is used from the seed decision, then the standard deviation may be less than 10% of the mean value. This can eliminate the edge effect of the measurement. A crack-free bulk crystal can be obtained in consideration of the correlation between the data scattering and the crack density of the peak width of the oscillation curve.

Benefits and Improvements

The bulk GaN crystal obtained by the method disclosed in the present invention contains little or no amount of cracks. The crack free bulk GaN crystal that can be obtained is sliced into wafers. Such wafers are used in optical devices such as LEDs and laser diodes, or electronic devices such as high power transistors. Since cracks greatly degrade the performance and reliability of such devices, the present invention can improve device performance and reliability.

Possible variants

Although the preferred embodiment describes bulk crystallization of GaN, similar advantages of the present invention may be expected for Group III nitride alloys of various compositions such as AlN, AlGaN, InNm, InGaN or GaAlInN.

Although the preferred embodiment describes GaN seed crystals having a thickness of about 430 microns, similar advantages of the present invention can be expected for other thicknesses of 100 microns to 2000 microns.

While the preferred embodiment describes the development of the grit heat, similar advantages of the present invention may be expected in other bulk growth processes such as flux processes or high pressure, hot solution growth. In the flux process, fluxes such as Group III metals and sodium are melted together and nitrogen is subsequently dissolved in the melt. One of the flux methods is disclosed in U.S. Patent No. 5,868,837. One suitable high pressure, high temperature solution growth method is disclosed in U.S. Patent No. 627,3948 B1. Each of these patents is incorporated herein.

Although the preferred embodiment describes seed crystals having a size of about 50 mm, a similar advantage of the present invention can be expected for smaller or larger seeds such as 1 ", 2 ", 4 ", 6 ".

The bulk crystals described, manufactured or used in one of the above descriptions may have a thickness of, for example, 1 mm, 2 mm, 3 mm, 4 mm, 5 mm, 6 mm, 7 mm, 8 mm, 9 mm, 10 mm or more.

Claims (28)

A method of growing a Group III nitride bulk crystal having a composition of Ga _{x 1} Al _{y 1} In _{1-x 1} -y ₁ N (0? X _{1? 1} , 0? X 1 + _{y 1? 1} )
(a) measuring an X-ray oscillation curve of the seed crystal at one or more points;
(b) quantifying a peak width of the measured X-ray oscillation curve;
(c) comparing the indicated value of the quantized peak width distribution with an acceptable value; And
(d) growing monocrystalline Ga _{x 1} Al _{y 1} In _{1-x 1-y 1} N on the surface of the seed crystal having an acceptable value of the quantified peak width distribution to form a bulk crystal of the Group III nitride Lt; RTI ID = 0.0 > III-nitride < / RTI > The method according to claim 1,
The method of quantifying the peak width comprises calculating a half width of the peak of the X-ray oscillation curve. The method according to claim 1 or 2,
Wherein the quantized peak width distribution is determined as a standard deviation. The method of claim 3,
Wherein the standard deviation is less than 30% of the average value of the quantized peak width. The method of claim 3,
Wherein the standard deviation is less than 20% of the average value of the quantized peak width. The method of claim 3,
Wherein the standard deviation is less than 10% of the average value of the quantized peak width. The method according to any one of claims 1 to 6,
Wherein the seed crystal is predominantly c-plane oriented, and wherein the X-ray swing curve is measured on one or more off-axis planes. The method of claim 7,
Wherein the X-ray swing curve is measured in the m-direction. The method of claim 7,
Lt; RTI ID = 0.0 > 201 < / RTI > reflex. The method of claim 7,
Lt; RTI ID = 0.0 > 111 < / RTI > reflex. The method according to any one of claims 1 to 10,
Wherein the seed crystal is gallium nitride. The method according to any one of claims 1 to 11,
Wherein the Group III nitride is GaN. The method according to any one of claims 1 to 12,
Wherein the Group III nitride is grown in supercritical ammonia. The method according to any one of claims 1 to 13,
Wherein the Group III nitride bulk crystal has a crack density of less than 1 cm <" ^{2 &} gt ;. A method for selecting a seed crystal for growing a Group III nitride bulk crystal having a composition of Ga _{x 1} Al _{y 1} In _{1-x 1} -y ₁ N (0? X _{1? 1} , 0? X 1 + _{y 1? 1} )
(a) measuring an X-ray oscillation curve of the seed crystal at one or more points;
(b) quantifying a peak width of the measured X-ray oscillation curve;
(c) comparing the indicated value of the quantized peak width distribution with an acceptable value; And
(d) indicating acceptably or incapable of the seed crystal based on the indicated value of the quantized peak width distribution. 16. The method of claim 15,
The method of quantifying the peak width includes calculating a half-width of a peak of the X-ray oscillation curve. A method of selecting a seed crystal for growing Group-III nitride bulk crystal. The method according to claim 15 or 16,
Wherein a distribution of the quantized peak width is determined by a standard deviation. 18. The method of claim 17,
Wherein the standard deviation is less than 30% of the average value of the quantized peak width. 18. The method of claim 17,
Wherein the standard deviation is less than 20% of an average value of the quantized peak width. 18. The method of claim 17,
Wherein the standard deviation is less than 10% of the average value of the quantized peak width. The method according to any one of claims 15 to 20,
Wherein the seed crystal is predominantly c-plane oriented, and wherein the X-ray swing curve is measured on at least one off-axis plane to grow a Group III nitride bulk crystal. 23. The method of claim 21,
Wherein the X-ray swing curve is a measure of seed crystal selection for growing Group-III nitride bulk crystals measured in the m-direction. 23. The method of claim 21,
Wherein the pit plane is 201 reflex. ≪ RTI ID = 0.0 > 11. < / RTI > 23. The method of claim 21,
Wherein the pit plane is a 102 reflections group III nitride bulk crystal. The method according to any one of claims 15 to 24,
Wherein the seed crystal is a gallium nitride, and the Group III nitride bulk crystal is grown. The method of any one of claims 15 to 25,
Wherein the Group-III nitride is GaN. A bulk Ill-nitride grown by the method of any one of claims 1 to 26 A Group III nitride wafer formed by the method of any one of claims 1 to 26.

Images