Group iii nitride semiconductor and manufacturing method for the same
The method forms an AlGaN or AlN crystalline nucleus layer on a sapphire substrate to facilitate substrate separation from an AlN semiconductor layer by laser-induced void formation, addressing the high bond energy challenge and reducing laser output requirements.
Patent Information
- Application Number
- JP2024009500
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-01-25
- Publication Date
- 2025-08-06
AI Technical Summary
Laser lift-off technology is challenging for separating a sapphire substrate from an AlN semiconductor layer due to the high bond energy of Al-N bonds, requiring an extremely high laser output that is difficult to achieve.
A method involving the formation of an AlGaN or AlN crystalline nucleus layer on a sapphire substrate, followed by laser irradiation to generate heat and form voids, allowing substrate separation at the interface.
Reduces the laser output required for substrate separation by forming voids in the sapphire substrate without decomposing the crystal nucleus layer, enabling efficient separation of the substrate from the semiconductor layer.
Smart Images

Figure 2025115135000001_ABST
Abstract
Description
[Technical Field]
[0001] The present invention relates to a Group III nitride semiconductor and a method for producing the same. [Background technology]
[0002] Laser lift-off (LLO) is a known technique for separating a substrate from a semiconductor layer formed on it. In this method, after a semiconductor layer is formed on a substrate, laser light is irradiated from the backside of the substrate to decompose the semiconductor layer at the interface between the substrate and the semiconductor layer, thereby separating the substrate from the semiconductor layer.
[0003] Patent Document 1 describes a light-emitting device in which an n-type layer, a light-emitting layer, and a p-type layer are formed on a growth substrate made of sapphire via a buffer layer made of AlN, and the buffer layer is decomposed by irradiating a laser from the back side of the growth substrate, thereby separating the growth substrate. [Prior art documents] [Patent documents]
[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2018-61049 Summary of the Invention [Problem to be solved by the invention]
[0005] Laser lift-off technology is established and commercialized when the substrate is sapphire and the semiconductor layer is GaN. However, AlN has a higher bond energy than GaN, so Al-N bonds cannot be easily broken. For this reason, laser lift-off technology has not been established when the substrate is sapphire and the semiconductor layer is AlN. In Patent Document 1, the growth substrate is separated by thermally decomposing the AlN buffer layer, but thermally decomposing AlN requires an extremely high laser output, which is difficult to achieve.
[0006] The present invention has been made in view of the above background, and aims to provide a method for manufacturing a Group III nitride semiconductor that enables separation of a substrate from a stack of a sapphire substrate and a semiconductor layer made of a Group III nitride semiconductor. [Means for solving the problem]
[0007] One aspect of the present invention is a crystalline nucleation layer formation step of generating nuclei of AlGaN or AlN on a substrate made of sapphire to form a crystalline nucleation layer; a semiconductor layer forming step of forming a semiconductor layer made of a Group III nitride semiconductor on the crystalline nucleus layer; a void forming step of irradiating a laser beam from the back side of the substrate, transmitting the laser beam through the substrate and causing the crystal nucleus layer to absorb the laser beam, thereby generating heat in the crystal nucleus layer, and conducting the heat of the crystal nucleus layer to the substrate to decompose a region of the substrate near the interface with the crystal nucleus layer, thereby forming voids; and a substrate separation step of separating the substrate from the crystal nucleus layer at the position of the void.
[0008] Another aspect of the present invention is a crystal nucleation layer, which is a layer formed by generating and growing nuclei of AlGaN or AlN; a semiconductor layer formed on one surface of the crystal nucleus layer and made of a Group III nitride semiconductor; a plurality of protrusions made of sapphire formed on the other surface of the crystal nucleus layer; The surface of the crystal nucleus layer is exposed between the adjacent protrusions of the Group III nitride semiconductor. [Effects of the Invention]
[0009] In the above embodiment, voids are formed in the sapphire substrate by the laser beam, and there is no need to form voids in the crystal nucleus layer, which allows the output of the laser beam to be reduced.
[0010] As described above, according to the above aspect, it is possible to provide a method for manufacturing a Group III nitride semiconductor that allows the substrate to be separated from a stack of a substrate made of sapphire and a semiconductor layer made of a Group III nitride semiconductor. [Brief explanation of the drawings]
[0011] [Figure 1] 1 is a cross-sectional view showing the configuration of a light-emitting device according to Embodiment 1, the cross-section being perpendicular to the main surface of the substrate. [Figure 2] 3A to 3C are diagrams showing the structure of each layer at each growth stage up to the formation of a two-dimensional growth layer. [Figure 3] Schematic diagram showing substrate warpage at each growth stage. [Figure 4] Graph showing the change in temperature and pressure over time during AlN layer formation. [Figure 5] 2A to 2C are diagrams illustrating a manufacturing process of the light-emitting element according to the first embodiment. [Figure 6] 2A to 2C are diagrams illustrating a manufacturing process of the light-emitting element according to the first embodiment. [Figure 7] 2A to 2C are diagrams illustrating a manufacturing process of the light-emitting element according to the first embodiment. [Figure 8] 2A to 2C are diagrams illustrating a manufacturing process of the light-emitting element according to the first embodiment. [Figure 9] 2A to 2C are diagrams illustrating a manufacturing process of the light-emitting element according to the first embodiment. [Figure 10] FIG. 1 shows a photograph of a sample viewed from the substrate side. [Figure 11] A cross-sectional SEM image of the sample. [Figure 12] A cross-sectional SEM image of the sample. [Figure 13] Elemental mapping of Al, Ga, O, and N. [Figure 14] FIG. 1 shows the relationship between the Al composition of the crystal nucleation layer, the energy density of the laser beam, and the separation of the substrate. [Figure 15] FIG. 10 is a cross-sectional view showing the configuration of a light-emitting device in a modified form of the first embodiment, the cross-section being perpendicular to the main surface of the substrate. DETAILED DESCRIPTION OF THE INVENTION
[0012] The method for manufacturing a Group III nitride semiconductor includes a crystalline nucleus layer formation step of generating AlGaN or AlN nuclei on a sapphire substrate to form a crystalline nucleus layer; a semiconductor layer formation step of forming a semiconductor layer made of a Group III nitride semiconductor on the crystalline nucleus layer; a void formation step of irradiating the backside of the substrate with laser light, which passes through the substrate and is absorbed by the crystalline nucleus layer, thereby generating heat in the crystalline nucleus layer and conducting the heat of the crystalline nucleus layer to the substrate, thereby decomposing a region of the substrate near the interface with the crystalline nucleus layer and forming voids; and a substrate separation step of separating the substrate from the crystalline nucleus layer at the location of the void.
[0013] In the method for producing a Group III nitride semiconductor, in the substrate separation step, a protrusion may be formed on the surface of the crystalline nucleus layer facing the substrate, leaving a part of the substrate.
[0014] In the method for producing a Group III nitride semiconductor, the Al composition of the crystal nucleus layer is set to x, and the energy density of the laser light (J / cm 2 ) is defined as y, the energy density y of the laser light may satisfy y≧3x−1.4.
[0015] In the method for producing a Group III nitride semiconductor, the energy density of the laser light is 1.6 J / cm 2 It may be more than that.
[0016] In the method for producing a Group III nitride semiconductor, the energy density of the laser light is 5 J / cm 2 It may be the following:
[0017] In the method for producing a Group III nitride semiconductor described above, the nuclei of the crystal nucleus layer may be AlGaN or AlN having an Al composition of 50% or more.
[0018] In the method for manufacturing a Group III nitride semiconductor described above, the semiconductor layer may have a low-temperature three-dimensional growth layer formed on the crystal nucleus layer and a high-temperature three-dimensional growth layer formed on the low-temperature three-dimensional growth layer, and the semiconductor layer formation step may have a low-temperature three-dimensional growth layer formation step of growing AlGaN or AlN from the nuclei at a temperature lower than that of the crystal nucleus layer formation step to coalesce crystals from adjacent nuclei to form the low-temperature three-dimensional growth layer, and a high-temperature three-dimensional growth layer formation step of growing AlGaN or AlN from the low-temperature three-dimensional growth layer at a temperature higher than that of the low-temperature three-dimensional growth layer formation step but not higher than the temperature of the crystal nucleus layer formation step to form the high-temperature three-dimensional growth layer.
[0019] In the method for producing a Group III nitride semiconductor, the temperature in the crystalline nucleus layer formation step may be 1100°C or higher and 1200°C or lower, the temperature in the low-temperature three-dimensional growth layer formation step may be 900°C or higher and 1100°C or lower, and the temperature in the high-temperature three-dimensional growth layer formation step may be 1050°C or higher and 1200°C or lower.
[0020] In the method for manufacturing a Group III nitride semiconductor, the semiconductor layer may have a low-temperature three-dimensional growth layer formed on the crystal nucleus layer and a high-temperature three-dimensional growth layer formed on the low-temperature three-dimensional growth layer, and the semiconductor layer formation step may have a low-temperature three-dimensional growth layer formation step of growing AlGaN or AlN from the nuclei at a growth rate slower than that of the crystal nucleus layer to coalesce crystals from adjacent nuclei to form the low-temperature three-dimensional growth layer, and a high-temperature three-dimensional growth layer formation step of growing AlGaN or AlN from the low-temperature three-dimensional growth layer at a growth rate faster than that of the low-temperature three-dimensional growth layer but equal to or lower than that of the crystal nucleus layer to form the high-temperature three-dimensional growth layer.
[0021] In the method for producing a Group III nitride semiconductor, the growth rate of the crystal nucleus layer may be 5 nm / min or more and 100 nm / min or less, the growth rate of the low-temperature three-dimensional growth layer may be 2 nm / min or more and 20 nm / min or less, and the growth rate of the high-temperature three-dimensional growth layer may be 5 nm / min or more and 50 nm / min or less.
[0022] The method for producing a Group III nitride semiconductor may further include, after the substrate separation step, a roughness forming step of etching the crystal nucleus layer side exposed by the separation of the substrate to form roughness having a depth extending from the crystal nucleus layer to the semiconductor layer.
[0023] The method for producing a Group III nitride semiconductor may further include a thermal cleaning step, prior to the crystalline nucleus layer forming step, of heat-treating the substrate in a hydrogen-dominated atmosphere at a temperature higher than that in the crystalline nucleus layer forming step.
[0024] The Group III nitride semiconductor has a crystalline nucleus layer, which is a layer formed by generating and growing AlGaN or AlN nuclei, a semiconductor layer made of a Group III nitride semiconductor and formed on one surface of the crystalline nucleus layer, and a plurality of protrusions made of sapphire and formed on the other surface of the crystalline nucleus layer, with the surface of the crystalline nucleus layer exposed between adjacent protrusions.
[0025] In the above-mentioned Group III nitride semiconductor, the height of the protrusions may be 0.01 to 1 μm, the width of the protrusions may be 0.01 to 1 μm, and the center-to-center distance between adjacent protrusions may be 0.02 to 5 μm.
[0026] In the above-mentioned Group III nitride semiconductor, the semiconductor layer may have a three-dimensional growth layer formed on the crystal nucleus layer, in which AlGaN or AlN is grown from the nuclei and crystals from adjacent nuclei are united, and a two-dimensional growth layer formed on the three-dimensional growth layer, in which AlGaN or Ga-doped AlN is grown from the three-dimensional growth layer.
[0027] (Embodiment 1) 1. Structure of light-emitting element Fig. 1 is a cross-sectional view showing the configuration of a light-emitting device according to embodiment 1, showing a cross section perpendicular to the main surface of the substrate. As shown in Fig. 1, the light-emitting device according to embodiment 1 includes a crystal nucleus layer 11, a three-dimensionally grown layer 12, a two-dimensionally grown layer 13, an n-type layer 14, an active layer 15, an electron blocking layer 16, a p-type layer 17, a p-side electrode 18, an n-side electrode 19, and a protrusion 20, and the p-side electrode 18 and the n-side electrode 19 are bonded to a support substrate 21. The light-emitting device according to embodiment 1 emits light in the UVC band.
[0028] The crystalline nucleus layer 11 is located on the substrate 10. The crystalline nucleus layer 11 is a layer formed by generating nuclei 11A made of AlN on the surface of the substrate 10 and three-dimensionally growing them.
[0029] The thickness of the crystalline nucleus layer 11 (height of the nuclei 11A) is preferably 1 to 200 nm. If the thickness of the crystalline nucleus layer 11 is in this range, the nuclei 11A are sufficiently large, and the quality of the crystal layer formed after the nuclei 11A can be improved. The thickness of the crystalline nucleus layer 11 is more preferably 2 to 40 nm, and even more preferably 3 to 30 nm. Since the nuclei 11A are sufficiently large, the density of the nuclei 11A is low. For example, when the thickness of the crystalline nucleus layer 11 is in the range of 5 to 100 nm, the density of the nuclei 11A is 3×10 11 / cm -2 It is preferable to do the following:
[0030] The size of the nuclei 11A (average diameter in plan view) is preferably 20 to 50 nm. Here, the diameter is the diameter when the nuclei 11A are converted into a circle of the same area. Also, this is the size of the nuclei 11A when the thickness of the crystal nucleus layer 11 is 5 to 100 nm. By setting the size of the nuclei 11A within this range, tensile stress can be sufficiently reduced. Furthermore, the variation in size of the nuclei 11A (the difference between the maximum diameter and the average diameter, and the difference between the average diameter and the minimum diameter) is preferably 10 nm or less.
[0031] It is important that the nuclei 11A are large. When the crystal nuclei coalesce to form a flat film, tensile stress occurs on the surface. Therefore, if the nuclei size is small and the density is high, the stress generated on the surface when they coalesce to form a flat film will also be large. Therefore, it is necessary to reduce the nuclei density. To achieve this, it is effective to increase the nuclei size.
[0032] A low nucleus density has another advantage. Threading dislocations are formed at the interface where the nuclei 11A coalesce. A low nucleus density also reduces the number of interface where the nuclei 11A coalesce. Therefore, by reducing each density, it is possible to reduce the formation of threading dislocations, and to form a high-quality crystal film.
[0033] The three-dimensional growth layer 12 is located on the crystal nucleus layer 11. The three-dimensional growth layer 12 is made of AlN. The three-dimensional growth layer 12 is a layer formed by three-dimensionally growing the nuclei 11A, merging adjacent nuclei 11A, and then three-dimensionally growing the combined AlN. The three-dimensional growth layer 12 is a laminate of a low-temperature three-dimensional growth layer 12A and a high-temperature three-dimensional growth layer 12B. The high-temperature three-dimensional growth layer 12B is a layer grown at a higher temperature than the low-temperature three-dimensional growth layer 12A. By three-dimensionally merging the nuclei 11A rather than growing them all at once into a two-dimensional flat film, stress generated on the surface can be alleviated.
[0034] On the surface of the low-temperature three-dimensionally grown layer 12A, the proportion of flat crystal planes parallel to the principal surface of the substrate 10 is preferably smaller than the proportion of obliquely inclined crystal planes. The obliquely inclined planes may be formed by low-order or high-order facet planes, such as {10-11} planes or {11-22} planes.
[0035] The thickness of the low-temperature three-dimensionally grown layer 12A is preferably 100 to 500 nm. Within this range, the nuclei 11A can be sufficiently combined to reduce the dislocation density. The surface roughness RMS of the low-temperature three-dimensionally grown layer 12A is, for example, 10 to 100 nm. The thickness of the low-temperature three-dimensionally grown layer 12A is more preferably 250 to 400 nm, and even more preferably 250 to 350 nm.
[0036] On the surface of the high-temperature three-dimensionally grown layer 12B, the proportion of flat crystal planes parallel to the main surface of the substrate 10 is preferably larger than the proportion of obliquely inclined crystal planes. In other words, the high-temperature three-dimensionally grown layer 12B is a flatter film than the low-temperature three-dimensionally grown layer 12A.
[0037] The thickness of the high-temperature three-dimensionally grown layer 12B is 750 to 2000 nm. With such a thickness, the nuclei 11A and the low-temperature three-dimensionally grown layer 12A can be combined with each other without any gaps, thereby reducing the dislocation density. The thickness of the high-temperature three-dimensionally grown layer 12B is more preferably 1000 to 1750 nm, and even more preferably 1250 to 1500 nm.
[0038] Voids may be present in the crystals of the crystal nucleus layer 11 and the three-dimensional growth layer 12. The voids are regions that were not completely filled by the growth of the nuclei 11A, and are spaces in the crystals filled with carrier gas. The voids can relieve residual stress in the film and suppress the occurrence of cracks.
[0039] The two-dimensional growth layer 13 is located on the three-dimensional growth layer 12. The two-dimensional growth layer 13 is a layer in which AlN is grown while Ga is supplied. The two-dimensional growth layer 13 is made of AlN or AlGaN containing Ga. Ga is supplied as a surfactant that promotes lateral growth during AlN film formation. In order to exert the surfactant effect, the molar ratio of Ga to Al in the formed AlN film must be greater than 0. Due to the surfactant effect, the surface of the two-dimensional growth layer 13 is a two-dimensionally flat surface without unevenness caused by pits or crystal facets.
[0040] The AlN containing Ga is AlN in a state where Ga is dissolved in solid solution, and is AlN containing Ga to an extent that it does not form a mixed crystal of GaN and AlN. For example, the molar ratio of Ga to Al in the two-dimensionally grown layer 13 is greater than 0 and 0.5% or less.
[0041] The surfactant effect of Ga is also exerted when the Ga content of AlN increases to a level at which an AlN-GaN mixed crystal (AlGaN) is formed. The mixed crystal is formed when the molar ratio of Ga to Al is greater than 0.5%. On the other hand, if the Ga composition of AlGaN becomes too high, it is more likely to absorb ultraviolet light from the active layer 15. Therefore, when the two-dimensionally grown layer 13 is made of AlGaN, it is preferable to set the Ga composition to greater than 0.5% and not more than 10%.
[0042] The Ga supply amount in the two-dimensional growth layer 13 may vary in the thickness direction, or may increase continuously or stepwise with increasing distance from the substrate 10. The back side region of the two-dimensional growth layer 13 (the three-dimensional growth layer 12 side) is in a state where Ga is solid-dissolved in AlN, and the front side is in a mixed crystal state of AlN and GaN (i.e., Al 1-x Ga x N) and Al. 1-x Ga x The Ga composition x of N also changes in the thickness direction, and the Ga composition x increases continuously or stepwise from 0 to a predetermined value with increasing distance from the substrate 10. The maximum value of the Ga composition x (i.e., the Ga composition x near the surface of the two-dimensionally grown layer 13) is 0.01 to 0.1.
[0043] Thus, the surface vicinity of the two-dimensional growth layer 13 is not AlN but a mixed crystal of AlN and GaN, but the Ga composition x is low enough to allow it to function equivalent to AlN as an underlayer for deep-ultraviolet light-emitting devices. Furthermore, the two-dimensional growth layer 13 is flat, has few cracks, and has a low dislocation density. By continuously increasing the Ga ratio, the lattice mismatch with the n-type layer 14 (n-type AlGaN layer) stacked on top of it also becomes smaller, contributing to improving the crystal quality of the n-type layer 14.
[0044] The thickness of the two-dimensional growth layer 13 is 750 to 2000 nm. With such a thickness, the surface of the two-dimensional growth layer 13 is sufficiently flattened. For example, the surface roughness RMS of the two-dimensional growth layer 13 can be set to 0.5 to 5 nm. Step bunching may be present on the surface of the two-dimensional growth layer 13. The thickness of the two-dimensional growth layer 13 is more preferably 1000 to 1750 nm, and even more preferably 1250 to 1500 nm.
[0045] The threading dislocation density of the high-temperature three-dimensional growth layer 12B and the two-dimensional growth layer 13 is lower than that of the crystal nucleus layer 11 and the low-temperature three-dimensional growth layer 12A. The threading dislocation density of the high-temperature three-dimensional growth layer 12B and the two-dimensional growth layer 13 is, for example, 5×10 9 cm -2 The threading dislocation density of the crystal nucleus layer 11 and the low-temperature three-dimensionally grown layer 12A is, for example, 1×10 10 cm -2 That's all.
[0046] Furthermore, the dislocation density of dislocations with edge dislocation components in the two-dimensionally grown layer 13 is higher than that of dislocations with screw dislocation components. For example, the dislocation density of dislocations with edge dislocation components is 10 times or more higher.
[0047] In the X-ray diffraction pattern of the two-dimensionally grown layer 13, the full width at half maximum (FWHM) of the (002) diffraction line is, for example, 20 to 200 arcsec, and the full width at half maximum of the (102) diffraction line is, for example, 200 to 800 arcsec.
[0048] The n-type layer 14 is located on the two-dimensionally grown layer 13. The n-type layer 14 is made of n-AlGaN and has a Ga composition higher than that of the two-dimensionally grown layer 13. The n-type impurity is Si, and the Si concentration is 5×10 18 ~5×10 19 / cm 3 The n-type layer 14 may be made up of multiple layers.
[0049] The active layer 15 is located on the n-type layer 14. The active layer 15 has an MQW structure in which well layers and barrier layers are alternately stacked. The number of repetitions is, for example, 2 to 5. The well layers are made of AlGaN, and the Al composition is set according to the desired emission wavelength. For example, it is set so that the emission wavelength is a predetermined value in the range of 200 to 280 nm. The barrier layers are AlGaN with a higher Al composition than the well layers. They may also be AlGaInN with a larger band gap energy than the well layers. The active layer 15 may also have an SQW structure.
[0050] The electron blocking layer 16 is located on the active layer 15. The electron blocking layer 16 is made of AlGaN or AlN with a higher Al composition ratio than the barrier layer of the active layer 15. The electron blocking layer 16 may be doped with impurities to provide p-type conductivity. The doping method may be constant or modulated, or may be combined with an undoped layer. The electron blocking layer 16 prevents electrons injected from the n-side electrode 19 from diffusing beyond the active layer 15 toward the p-type layer 17.
[0051] The p-type layer 17 is located on the electron blocking layer 16. The p-type layer 17 is made of p-AlGaN. In the light emitting device of the first embodiment, all the semiconductor layers from the n-type layer 14 to the p-type layer 17 are made of AlGaN, which suppresses the absorption of ultraviolet light emitted from the active layer 15 by the semiconductor layers. The p-type impurity is Mg. The Mg concentration is 1×10 18 / cm 3 That's all. The p-type layer 17 may be made up of multiple layers with different Al compositions or Mg concentrations. In that case, the layer in contact with the p-side electrode 18 may be made of p-GaN to reduce contact resistance.
[0052] The lowest Al composition in the p-type layer 17 is preferably an Al composition with a band gap that does not absorb the wavelength of the light emitted from the active layer 15. For example, if the Al composition of the well layer of the active layer 15 is 40%, the lowest Al composition in the p-type layer 17 is preferably 40% or more. However, since a high Al composition increases the contact resistance with the p-side electrode 18, a thin layer with a composition of 0 to 40% may be laminated as a contact layer with a thickness of 1 to 50 nm. The layer with a composition of 0 to 40% absorbs the ultraviolet light emitted from the active layer 15, but is thin enough to transmit it to some extent. This prevents a significant decrease in the external quantum efficiency of the LED.
[0053] Furthermore, layers may be combined with an Al composition in the range of 0 to 90%. In this case, it is preferable that the Al composition is set to decrease continuously or stepwise from the active layer 15 side. A superlattice structure may be formed, and the average Al composition may decrease stepwise.
[0054] A groove is provided in a partial area of the surface of the p-type layer 17, the groove having a depth reaching the n-type layer 14. This groove is for exposing the n-type layer 14 so that an n-side electrode 19 can be provided therein.
[0055] The p-side electrode 18 is provided on the p-type layer 17. The p-side electrode 18 is a reflective electrode that reflects ultraviolet light emitted from the active layer 15 toward the substrate 10 to improve light extraction efficiency. The p-side electrode 18 is made of a material such as Ru, Rh, Ni / Au, or Ni / Al.
[0056] The n-side electrode 19 is provided on the n-type layer 14 exposed at the bottom of the groove. The material of the n-side electrode 19 is Ti / Al, V / Al, or the like.
[0057] The protrusions 20 are located on the back surface of the crystalline nucleus layer 11 (the surface on the low-temperature three-dimensional growth layer 12A side). The protrusions 20 are made of sapphire. As described below, these protrusions 20 are formed by forming voids 22 in a substrate 10 made of sapphire and separating the substrate 10 from the crystalline nucleus layer 11, leaving a portion of the substrate 10 on the back surface of the crystalline nucleus layer 11. A plurality of protrusions 20 are present at intervals. A flat back surface of the crystalline nucleus layer 11 may be exposed between the protrusions 20. This flat crystalline nucleus layer 11 corresponds to the position of the voids 22 described below. By providing the protrusions 20 in this manner, ultraviolet light emitted from the active layer 15 is scattered, thereby improving the light extraction efficiency.
[0058] The height of the protrusions 20 is, for example, 0.01 to 1 μm, the width of the protrusions 20 is, for example, 0.01 to 1 μm, and the center-to-center distance between adjacent protrusions 20 is, for example, 0.02 to 5 μm.
[0059] The support substrate 21 is a substrate bonded to the p-side electrode 18 and the n-side electrode 19. An electrode pattern (not shown) is provided on the support substrate 21. This support substrate 21 is used to support the wafer during laser lift-off, which will be described later. The support substrate 21 is preferably made of a material with high thermal conductivity, such as AlN.
[0060] 2. Light-emitting device manufacturing method Next, a description will be given of a manufacturing method of the light-emitting device according to embodiment 1. The group III nitride semiconductor is formed by MOCVD, and the source gases used are, for example, ammonia as the nitrogen source gas, TMGa (trimethylgallium) or TEGa (triethylgallium) as the Ga source gas, and TMAl (trimethylaluminum) as the Al source gas, and hydrogen or nitrogen as the carrier gas.
[0061] 2 to 9. Fig. 2 is a diagram schematically illustrating the structure of each layer at each growth stage up to the formation of the two-dimensionally grown layer 13.
[0062] FIG. 3 is a conceptual diagram showing wafer warpage at each growth stage up to the formation of the two-dimensional growth layer 13. Wafer warpage can also occur in bulk substrates on which nothing is grown. Because the surface of the substrate 10 is cooled by a carrier gas such as hydrogen, its temperature is strictly different from that of the backside of the substrate 10. This temperature difference can cause warpage due to differences in thermal expansion. Therefore, the depiction of warpage in FIG. 3 conceptually illustrates the behavior caused by differences in lattice mismatch and thermal expansion coefficient when different materials are stacked, and may differ slightly from the actual state of warpage. Therefore, FIG. 3 conceptually illustrates the stress generated in the film due to the stacking of layers and the resulting warpage.
[0063] FIG. 4 is a graph showing the time changes in the growth temperature and growth pressure until the two-dimensionally grown layer 13 is formed.
[0064] 5 to 9 show the steps after the two-dimensional growth layer 13 is formed.
[0065] 2-1. Thermal cleaning First, a substrate 10 made of c-plane sapphire is prepared. The plane orientation of the main surface of the sapphire may be the a-plane. The substrate has an off-angle of 0.1 to 2 degrees in the m-axis or a-axis direction.
[0066] Next, the substrate 10 is thermally cleaned in a hydrogen atmosphere at a temperature of 1150 to 1250°C for 1 second to 15 minutes. This removes impurities from the surface of the substrate 10 and flattens it (see Figures 2(a), 3(a), and 4). In Figure 4, the growth temperature is 1190°C and the pressure is 4 kPa, as an example. Also in Figure 4, the period between this thermal cleaning and the nitriding treatment in the next step is shown as period T1. It is believed that this thermal cleaning removes oxygen from the surface of the substrate 10, resulting in an Al-rich surface.
[0067] The thermal cleaning temperature should be higher than that of the crystalline nucleus layer formation step described below. A more preferred range for the thermal cleaning temperature is 1170 to 1230°C, and even more preferably 1180 to 1210°C. The atmosphere may be dominated by hydrogen, for example, a mixed gas containing 80% or more by volume of hydrogen. The mixed gas may be, for example, a mixed gas of hydrogen and nitrogen. The pressure may be normal pressure, but reduced pressure is preferred, for example, 1 to 80 kPa, preferably 1 to 50 kPa, and even more preferably 1 to 20 kPa. The flow rate of the hydrogen gas or mixed gas of hydrogen and nitrogen above the substrate 10 may be 5 to 500 m / min (meters per minute), preferably 10 to 300 m / min, and even more preferably 15 to 150 m / min.
[0068] 2-2. Nitriding Next, the temperature is set lower than that of thermal cleaning, and ammonia is supplied to nitride the Al-rich surface of the substrate 10. The pressure is the same as that of thermal cleaning. The flow rate of the carrier gas (hydrogen or a mixed gas of hydrogen and nitrogen) should also be the same as or higher than the optimal range for thermal cleaning. The flow rate of ammonia is the same as or lower than that of the next process, and the ammonia partial pressure in the carrier gas, which is hydrogen or a mixed gas of hydrogen and nitrogen, is preferably in the range of 0.001 to 0.1 atm, more preferably 0.01 to 0.05 atm. The nitriding time is 0.5 seconds to 10 minutes, more preferably 60 to 300 seconds. An AlN layer (not shown) of one to several monolayers is formed on the surface of the substrate 10, which serves as the starting point for nucleation.
[0069] 2-3. Crystal nucleus layer 11 formation process Next, nitrogen source gas and Al source gas are supplied at the same temperature as the nitriding process in the previous step or at a lower temperature than the nitriding process in the previous step, generating AlN nuclei 11A on the substrate 10, and the nuclei 11A are grown three-dimensionally to form the crystal nucleus layer 11 (see Figures 2(b) and 4). In Figure 4, the period of this step is shown as period T2, and a temperature of 1155°C is shown as an example. The pressure is the same as in the previous step. Conventionally, the nuclei 11A are generated at about 1000°C and grown three-dimensionally, but in embodiment 1, the temperature is higher, which increases the surface migration of the source atoms and allows the nuclei 11A to be formed larger than conventional ones. This allows the nuclei 11A made of AlN to be of high quality.
[0070] The lattice mismatch between the nuclei 11A and the substrate 10 made of sapphire causes compressive stress in the crystal nucleus layer 11, but because the lattice mismatch between the nuclei 11A and the substrate 10 is large, the stress is relieved within a few nanometers, and the stress at the interface between the substrate 10 and the nuclei 11A is small.
[0071] The crystal nucleus layer 11 is preferably grown three-dimensionally until it has a thickness of 1 to 200 nm, which allows the nuclei 11A to be sufficiently large. It is also preferable that the size and density of the nuclei 11A are within the above-mentioned ranges.
[0072] The preferred growth temperature for the crystal nucleus layer 11 is 1100 to 1200°C. By setting the temperature in this range, the nuclei 11A can be made sufficiently large. This is because the surface migration of the raw material atoms can be enhanced. A more preferred temperature is 1125 to 1190°C, and even more preferred is 1150 to 1180°C. The same temperature as that used in the previous nitriding process may be used. By making the nuclei 11A larger, the nucleus density decreases. Therefore, the interfaces between the nuclei 11A also decrease. Tensile stress may occur in the film when the nuclei 11A combine and become flat, but by reducing the nucleus density as described above, the stress that occurs after the nuclei 11A combine can be reduced.
[0073] The growth temperature of the crystalline nucleus layer 11 may be changed stepwise or continuously. In this case, the average growth temperature should be 1100 to 1200°C.
[0074] The growth rate is preferably 5 to 100 nm / min. By increasing the growth rate, the diameter of the nuclei 11A can be increased. The growth rate is more preferably 10 to 70 nm / min, and even more preferably 20 to 50 nm / min.
[0075] The V / III ratio is preferably 5 to 500. By setting the V / III ratio in this range, the growth rate can be controlled within the above range. It is more preferably 5 to 400, and even more preferably 5 to 300.
[0076] Although -c-plane AlN may be formed on the nitrided substrate 10, it can be made into a +c-plane by controlling growth conditions such as a low V / III ratio of the nuclei 11A and a high growth rate. Normally, a +c-plane crystal layer is preferable, and even if +c and -c are mixed in the crystal nucleus layer 11, the proportion of -c is small, and +c becomes dominant during the growth process, ultimately resulting in a crystal layer having only a +c-plane surface.
[0077] 2-4. Low temperature three-dimensional growth layer 12A formation process Once the crystalline nucleus layer 11 has grown to a predetermined thickness, the growth temperature is lowered to a temperature lower than that of the formation process of the crystalline nucleus layer 11 (period T2). This causes the nuclei 11A of the crystalline nucleus layer 11 to grow three-dimensionally, and adjacent nuclei 11A merge to form a low-temperature three-dimensionally grown layer 12A (see Figures 2(c) and 4). Figure 4 shows the period of this process as period T3, with a temperature of 975°C as an example. The pressure is the same as in the previous process. By merging the nuclei 11A, threading dislocations can be reduced, resulting in the formation of a high-quality low-temperature three-dimensionally grown layer 12A.
[0078] Here, because the temperature is lower than the temperature at which the nuclei 11A are formed, the stress acting on the layers up to the low-temperature three-dimensionally grown layer 12A becomes compressive stress due to the difference in linear expansion coefficient with the substrate 10 (see FIG. 3(b)). Generally, cracks occur in the grown layer when tensile stress is applied. However, compressive strain occurs in the layers up to the low-temperature three-dimensionally grown layer 12A, so cracks can be suppressed.
[0079] Furthermore, by lowering the temperature, the low-temperature three-dimensionally grown layer 12A can be grown slowly, and the coalescence of the low-temperature three-dimensionally grown layer 12A growing from the nuclei 11A can also proceed slowly. Tensile stress may be generated when the nuclei 11A coalesce, but by growing the low-temperature three-dimensionally grown layer 12A slowly as described above, cracks due to tensile stress that occur when the low-temperature three-dimensionally grown layer 12A on the nuclei 11A coalesce to form a continuous film can also be suppressed.
[0080] The low-temperature three-dimensionally grown layer 12A is preferably grown thicker than the crystal nucleus layer 11, for example, to a thickness of 200 to 500 nm. By sufficiently coalescing the nuclei 11A starting from the nuclei to change from a nucleus shape to a film shape, the threading dislocation density can be reduced. Furthermore, the thicker the low-temperature three-dimensionally grown layer 12A, the more uneven the three-dimensional surface can be. This reduces the number of threading dislocations that bend and propagate to the surface of the grown layer due to the lateral growth promoted in subsequent processes. A more preferred thickness is 250 to 400 nm, and even more preferably 250 to 350 nm.
[0081] The growth temperature of the low-temperature three-dimensionally grown layer 12A is preferably 900 to 1100°C. By setting the temperature to 900°C or higher, impurities are less likely to enter the crystal, and light absorption can be reduced. The temperature is more preferably 950 to 1050°C, and even more preferably 975 to 1025°C.
[0082] Furthermore, the growth rate of the low-temperature three-dimensional growth layer 12A is set to be slower than that of the crystal nucleus layer 11. By slowing the growth rate and allowing the nuclei 11A to slowly coalesce starting from the nuclei 11A, dislocations can be efficiently reduced. In addition, the formation of tensile stress that occurs when the nuclei 11A coalesce to form a flat surface can be alleviated. The growth rate is preferably 2 to 20 nm / min, more preferably 2 to 15 nm / min, and even more preferably 2 to 10 nm / min.
[0083] The V / III ratio of the low-temperature three-dimensionally grown layer 12A is preferably set to 500 to 2000, which is larger than that of the crystal nucleus layer 11. This allows the growth rate to be controlled within the above range. The V / III ratio is more preferably set to 750 to 1750, and even more preferably set to 1000 to 1500.
[0084] 2-5. High temperature three-dimensional growth layer 12B formation process When the low-temperature three-dimensionally grown layer 12A, which has grown from the nuclei 11A as starting points, has sufficiently coalesced and grown to a predetermined thickness, the growth temperature is increased from that of the low-temperature three-dimensionally grown layer 12A formation step (period T3) to a temperature equal to or lower than that of the crystal nucleus layer 11 formation step, and further growth is performed on the low-temperature three-dimensionally grown layer 12A to form the high-temperature three-dimensionally grown layer 12B (see FIG. 2(d)), FIG. 4). In FIG. 4, the period of this step is shown as period T4, and a temperature of 1155°C is shown as an example. The pressure is the same as in the previous step.
[0085] The high-temperature three-dimensional growth layer 12B is a layer grown to smoothly transition from the three-dimensional growth of the low-temperature three-dimensional growth layer 12A to the two-dimensional growth of the two-dimensional growth layer 13. The high-temperature three-dimensional growth layer 12B grows laterally faster than the low-temperature three-dimensional growth layer 12A, and because the growth temperature of the high-temperature three-dimensional growth layer 12B is higher than that of the low-temperature three-dimensional growth layer 12A, the lateral growth is faster than that of the low-temperature three-dimensional growth layer 12A, which allows for more rapid crystal coalescence. As a result, the threading dislocations propagating toward the surface can be reduced by bending the threading dislocations laterally, resulting in a crystalline film with a low threading dislocation density.
[0086] Here, the temperature at which the high-temperature three-dimensional growth layer 12B is formed is higher than the temperature at which the low-temperature three-dimensional growth layer 12A is formed, and tensile stress occurs in the crystal layer up to the high-temperature three-dimensional growth layer 12B during growth of the low-temperature three-dimensional growth layer 12A due to the difference in linear expansion coefficient with the substrate 10 (see FIG. 3(c)). Therefore, it is preferable to grow the high-temperature three-dimensional growth layer 12B not as a completely flat film, but as a three-dimensional surface partially having pits or facets, thereby easing the tensile stress occurring on the surface.
[0087] The high-temperature three-dimensionally grown layer 12B is preferably grown thicker than the low-temperature three-dimensionally grown layer 12A, for example, to a thickness of 750 to 2000 nm. This range allows for more crystal coalescence and further reduces the threading dislocation density. A more preferred thickness is 1000 to 1750 nm, and even more preferably 1250 to 1500 nm.
[0088] The growth temperature of the high-temperature three-dimensionally grown layer 12B is preferably 1050 to 1200°C. By setting the temperature in this range, the growth mode can be efficiently converted from three-dimensional growth to two-dimensional growth. The temperature is more preferably 1075 to 1175°C, and even more preferably 1100 to 1150°C.
[0089] Furthermore, the growth rate of the high-temperature three-dimensionally grown layer 12B is preferably set to 5 to 50 nm / min, which is faster than the growth rate of the low-temperature three-dimensionally grown layer 12A but lower than the growth rate of the crystal nucleus layer 11. By setting the growth rate within this range, the crystals coalesce slowly, and the change in tensile stress generated on the crystal layer surface is gradual, thereby suppressing the occurrence of cracks during crystal coalescence. A growth rate of 10 to 40 nm / min is more preferable, and a rate of 15 to 30 nm / min is even more preferable.
[0090] The V / III ratio of the high-temperature three-dimensionally grown layer 12B is smaller than that of the low-temperature three-dimensionally grown layer 12A and equal to or greater than that of the crystal nucleus layer 11, and is preferably 100 to 1000. This allows the growth rate to be controlled within the above range. It is more preferably 150 to 700, and even more preferably 200 to 500.
[0091] 2-6. Two-dimensional growth layer 13 formation process Once the high-temperature three-dimensional growth layer 12B has grown to a predetermined thickness, Ga source gas is supplied in addition to the nitrogen source gas and Al source gas at the same temperature as or higher than the temperature in the high-temperature three-dimensional growth layer 12B formation step (period T4) to form a two-dimensional growth layer 13 made of AlN or AlGaN containing Ga (see Figures 2(e) and 4). In Figure 4, the period for this step is shown as period T5. The pressure is the same as in the previous step.
[0092] When Ga is supplied during AlN growth, Ga atoms, which have a higher surface migration rate than Al atoms, promote lateral growth of the AlN, resulting in two-dimensional growth. This allows the crystal to be flattened by the two-dimensional growth layer 13. Furthermore, the amount of Ga supplied is increased continuously and stepwise over time. This allows the lateral growth rate to be increased slowly, easing the change in tensile stress that occurs when the crystals coalesce, thereby suppressing the occurrence of cracks in the two-dimensional growth layer 13.
[0093] The Ga supply amount is controlled by the molar ratio of the Ga source gas to the Al source gas, which is preferably 0.05 to 0.5, more preferably 0.08 to 0.4, and even more preferably 0.1 to 0.3.
[0094] Although the ratio of Ga source gas is very high, Ga is not actually incorporated into AlN at the above ratio. This is because Ga atoms evaporate more easily than Al atoms on the AlN surface, so the Ga atoms actually incorporated into the AlN crystal are only a few percent even when the molar ratio of the source gas is 0.3. The higher the growth temperature, the more Ga atoms evaporate preferentially, so the number of Ga atoms incorporated into the AlN crystal becomes even smaller. Although it depends on conditions such as the growth temperature, the amount of Al grown within the above molar ratio range is 1-x Ga x The Ga solid phase ratio of N is x = about 0.01 to 0.1%.
[0095] By applying a continuous and stepwise gradient within the above molar ratio range, the Ga supply rate is increased over time. The two-dimensionally grown layer 13 changes from AlN to a mixed crystal of AlN and GaN (Al 1-x Ga x Here, the Ga composition is preferably 0.01 to 0.1.
[0096] The two-dimensionally grown layer 13 is preferably grown to a thickness equal to or greater than that of the high-temperature three-dimensionally grown layer 12B, for example, to a thickness of 750 to 2000 nm. Within this range, the surface of the two-dimensionally grown layer 13 can be sufficiently flattened. For example, the surface roughness RMS can be set to 0.5 to 5 nm. A more preferred thickness is 1000 to 1750 nm, and even more preferably 1250 to 1500 nm.
[0097] The growth temperature of the two-dimensionally grown layer 13 is preferably 1100 to 1200°C. If the temperature is within this range, the two-dimensionally grown layer 13 can be sufficiently flattened. The temperature is more preferably 1120 to 1190°C, and even more preferably 1140 to 1180°C. The temperature may be the same as that in the step of forming the crystal nucleus layer 11 and the step of forming the high-temperature three-dimensionally grown layer 12B.
[0098] The growth rate of the two-dimensionally grown layer 13 is preferably 5 to 50 nm / min, which is equal to or higher than the growth rate of the high-temperature three-dimensionally grown layer 12B and lower than the growth rate of the crystal nucleus layer 11. If the growth rate is within this range, the growth rate is sufficiently slow, which can moderate the change in tensile stress due to crystal flattening and suppress the occurrence of cracks. A more preferred range is 10 to 40 nm / min, and even more preferred is 15 to 30 nm / min.
[0099] The V / III ratio of the two-dimensionally grown layer 13 is set to be equal to or lower than the V / III ratio of the high-temperature three-dimensionally grown layer 12B and equal to or higher than the V / III ratio of the crystal nucleus layer 11, and is preferably set to 50 to 500. This allows the growth rate to be controlled within the above range. It is more preferably set to 100 to 400, and even more preferably set to 150 to 300.
[0100] The curvature of the wafer at the end of the growth of the two-dimensional growth layer 13 is 50 km. -1 More than 300km -1 It is preferable that the curvature is 100 km or less. Here, the concave curvature is a positive value, and the convex curvature is a negative value. If the curvature of the wafer is within this range, cracks in the crystal layer can be further suppressed. It is more preferable that the curvature is 100 km or less. -1 Over 200km -1 The following is the result.
[0101] The above manufacturing process makes it possible to obtain a high-quality two-dimensionally grown layer 13 that is flat, has suppressed crack generation, and has a low dislocation density. In particular, a high-quality two-dimensionally grown layer 13 can be obtained at a temperature of 1250° C. or less. Therefore, a general crystal growth device using quartz components can be used, and the device costs and running costs can be reduced.
[0102] After the two-dimensionally grown layer 13 is formed, when the temperature is lowered to room temperature, compressive stress is applied to the crystal layers up to the two-dimensionally grown layer 13 due to the difference in linear expansion coefficient with the substrate 10, causing the wafer to warp in a convex shape (see FIG. 3(d)).
[0103] When switching between periods T1 to T5, it is advisable to temporarily stop the supply of the source gas before changing the temperature, and then restart the supply of the source gas after changing the temperature to a predetermined temperature.
[0104] The crystal nucleus layer 11, the three-dimensional growth layer 12, and the two-dimensional growth layer 13 are grown under reduced pressure. This is because TMAl is highly reactive, and reducing the pressure reduces the reactivity, allowing for the production of high-quality crystals. In the first embodiment, the pressure is constant, but the pressure may be reduced as the growth temperature increases. This allows for the production of higher-quality crystals.
[0105] 2-7. Device structure formation process Next, an n-type layer 14, an active layer 15, an electron blocking layer 16, and a p-type layer 17 are laminated in this order on the two-dimensionally grown layer (see FIG. 5).
[0106] Next, a portion of the p-type layer 17 is dry-etched to form a groove reaching the n-type layer 14 (see FIG. 6). Then, a p-side electrode 18 is formed on the p-type layer 17, and an n-side electrode 19 is formed on the n-type layer 14 exposed at the bottom of the groove (see FIG. 7).
[0107] 2-8.Void formation process Next, the p-side electrode 18 and the n-side electrode 19 are bonded to the support substrate 21. Then, laser light is irradiated from the rear surface side of the substrate 10 (see FIG. 8).
[0108] The wavelength of the laser light is set to a wavelength that is transmitted through the substrate 10 and absorbed by the crystal nucleus layer 11. Absorption occurs in sapphire at approximately 150 nm or less, and in AlN at approximately 220 nm or less. Therefore, it is recommended to set the wavelength to longer than 150 nm and 220 nm or less. For example, an ArF excimer laser with a wavelength of 193 nm can be used.
[0109] When laser light having such a wavelength and a predetermined energy density is irradiated from the back side of the substrate 10, multiple voids 22 are formed on the substrate 10 side near the interface between the substrate 10 and the crystal nucleus layer 11 (see FIG. 9). When the voids 22 are spatially continuous with the outside of the device, the voids 22 are filled with air. When the voids 22 are not spatially continuous with the outside of the device, the voids 22 are considered to be filled with oxygen. Multiple voids 22 may be formed discretely, or adjacent voids 22 may overlap and be continuous. Increasing the overlap of the voids 22 makes it easier to separate the substrate 10. The shape, width, height, and spacing of the voids 22 can be controlled by the energy density of the laser light. The cross-sectional shape of the voids 22 may be elliptical, circular, rectangular, or the like. For example, the width of the voids 22 is 0.02 to 1 μm, the height of the voids 22 is 0.01 to 1 μm, and the spacing between the voids is 1 μm or less.
[0110] The energy density of the laser light may be any value as long as it is high enough to heat the crystal nucleus layer 11 sufficiently and decompose the substrate 10 with the heat.
[0111] In conventional laser lift-off, the substrate is separated by decomposing the buffer layer near the interface between the substrate 10 and the buffer layer. Therefore, when the buffer layer is made of AlN, the Al-N bond is strong, so the energy density of the laser light needs to be high.
[0112] In contrast, in the first embodiment, the substrate 10 is decomposed instead of the buffer layer (crystal nucleus layer 11), so there is no need to break the Al—N bond, and the substrate 10 can be separated at a lower energy density than conventionally.
[0113] For example, the Al composition of the crystal nucleus layer 11 is x, and the energy density of the laser light (J / cm 2 ) is defined as y, and if y≧3x−1.4 is satisfied, voids 22 can be easily formed in the substrate 10, and the substrate 10 can be easily separated. In particular, when the energy density is 1.6 J / cm 2 If the above conditions are met, the voids 22 can be formed regardless of the Al composition.
[0114] The energy density of the laser light is 5J / cm 2 It is preferable that the energy density of the laser beam is 3 J / cm or less. This is because, in general, the higher the energy density of the laser beam, the more difficult it is to achieve and the more damage it can cause to the crystal. Since AlN has a higher thermal conductivity than sapphire, it is thought that if the energy density is high, heat is more likely to be transferred to the active layer 15, p-side electrode 18, and n-side electrode 19, causing damage. It is more preferable that the energy density is 3 J / cm or less. 2 The following is the result.
[0115] The thickness of the substrate 10 is preferably 1000 μm or less, more preferably 500 μm or less, and most preferably 250 μm or less. While sapphire theoretically absorbs light at wavelengths below 150 nm, actual industrial sapphire contains impurities and oxygen vacancies, causing light absorption. Therefore, the intensity of the laser light gradually attenuates as it travels through the sapphire. Therefore, a thinner substrate 10 allows the energy of the incident laser light to be directly absorbed by the crystal nucleus layer 11.
[0116] The reason why voids 22 occur in the substrate 10 but not in the crystalline nucleus layer 11 is presumed to be as follows: The irradiated laser light passes through the sapphire substrate 10 without being absorbed and is absorbed by the AlN crystalline nucleus layer 11. Therefore, the crystalline nucleus layer 11 becomes a heat source. The region where the laser light is first absorbed, i.e., the interface between the substrate 10 and the crystalline nucleus layer 11, generates the most heat. Here, sapphire has a lower melting point than AlN. Therefore, it is thought that the substrate 10, which has a lower melting point at the interface between the substrate 10 and the crystalline nucleus layer 11, decomposes first. In addition, sapphire has a lower thermal conductivity than AlN and dissipates heat more easily. This difference in heat dissipation is thought to be one of the reasons why voids 22 occur in the substrate 10. It is thought that the above mechanism prevents voids 22 from occurring in the crystalline nucleus layer 11.
[0117] The crystalline nucleus layer 11 may be made of AlGaN, as described below. However, the higher the Ga composition of the crystalline nucleus layer 11, the lower the laser light transmittance. In other words, the laser energy is absorbed with a thinner film thickness. Therefore, the higher the Ga composition, the greater the amount of heat generated at the interface between the substrate 10 and the crystalline nucleus layer 11, which is thought to make the substrate 10 more likely to decompose.
[0118] 2-9.Substrate separation process Next, a physical force is applied to crack the substrate 10 at the position where the void 22 was formed, and the substrate 10 is separated and removed from the crystal nucleus layer 11. Depending on the size of the void 22, the substrate 10 may be separated without applying a physical force. At this time, protrusions 20 are formed on the back surface of the crystal nucleus layer 11 as remaining portions of the substrate 10. The planar pattern of the protrusions 20 is approximately the same as the inverted planar pattern of the voids 22.
[0119] The substrate 10 may be separated by allowing a solution capable of wet etching the crystalline nucleus layer 11 to penetrate through the voids 22 and wet-etch the crystalline nucleus layer 11 near the interface between the substrate 10 and the crystalline nucleus layer 11. For example, a solution such as TMAH or phosphoric acid may be used.
[0120] In this manner, the light emitting device according to the first embodiment shown in FIG. 1 is manufactured.
[0121] (Modification 1 of Embodiment 1) The cores 11A are not limited to AlN, but may be any III-nitride semiconductor containing Al. For example, AlGaN may be used. In particular, AlGaN with an Al composition of 50% or more is preferable. If the Al composition is less than 50%, there is a possibility that voids 22 may be generated on the crystal core layer 11 side rather than the substrate 10 side during laser irradiation in laser lift-off. However, if the Al composition is 50% or more, this can be suppressed. An Al composition of 70% or more is more preferable.
[0122] If AlGaN is used as the nucleus 11A, the difference in lattice constant with the sapphire substrate 10 becomes larger, and strain relaxation at the interface between the substrate 10 and the nucleus 11A increases. In addition, if the nucleus 11A is AlGaN, the lattice constant of the low-temperature three-dimensional growth layer 12A formed on the nucleus 11A has a higher Al composition than the nucleus 11A, so the low-temperature three-dimensional growth layer 12A grown on the nucleus 11A is subjected to tensile strain. This is because the growing crystal forms in conformity with the lattice constant of the substrate. Therefore, it is possible to relax the compressive strain that occurs in the low-temperature three-dimensional growth layer 12A and beyond due to thermal stress between the substrate and the crystal layer that occurs when growth is completed and the temperature returns to room temperature.
[0123] Furthermore, during strain relaxation at the interface between the substrate 10 and the nuclei 11A, the nuclei 11A are formed discretely rather than in a film shape, thereby suppressing the occurrence of cracks. Furthermore, the low-temperature three-dimensionally grown layer 12A formed on the crystal nuclei layer 11 is subjected to tensile stress from the crystal nuclei layer 11 due to the difference in lattice constants, but the occurrence of cracks due to the tensile stress is suppressed due to the three-dimensional growth. As a result of the above, the threading dislocation density of the two-dimensionally grown layer 13 can be reduced.
[0124] Furthermore, when the nuclei 11A are AlGaN, the surfactant effect of Ga allows the nuclei 11A to be larger, thereby reducing the nucleus density. As mentioned above, a low nucleus density reduces the tensile stress that occurs on the surface when the nuclei 11A coalesce to form a flat film. Furthermore, the number of meeting surfaces where the nuclei 11A coalesce is reduced, reducing the formation of threading dislocations and enabling the formation of a high-quality crystalline film.
[0125] When nuclei 11A are made of AlGaN, the size of nuclei 11A is preferably 20 to 100 nm. By setting the size of nuclei 11A in this range, tensile stress can be sufficiently reduced. Furthermore, the size variation of nuclei 11A (the difference between the maximum diameter and the average diameter, and the difference between the average diameter and the minimum diameter) is preferably 20 nm or less.
[0126] Furthermore, when the nuclei 11A are AlGaN, the thickness of the crystal nucleus layer 11 is preferably 5 to 100 nm. Since the nuclei 11A are larger than when the nuclei 11A are AlN, the crystal nucleus layer 11 is also thicker. As a result of the larger nuclei 11A, the quality of the crystal layers formed after the nuclei 11A can be further improved. A more preferable thickness of the crystal nucleus layer 11 is 5 to 50 nm. When the nuclei are AlGaN and the Ga composition is higher, the crystals become more mobile by annealing, and it becomes easier to enlarge the nuclei by solid-phase growth. Annealing can also increase the size of the nuclei immediately before growing a three-dimensional growth layer, thereby reducing the density.
[0127] In addition, when the nuclei 11A are made of AlGaN, the density of the nuclei 11A is 3×10 11 / cm -2 It is preferable that the density of the nuclei 11A is 1.5×10 or less. Since the nuclei 11A are larger than in the case of AlN, the density of the nuclei 11A is also smaller than in the case of AlN. It is more preferable that the density of the nuclei 11A is 1.5×10 or less. 11 / cm -2 or less, more preferably 1 × 10 11 / cm -2 In addition, when the nuclei 11A are AlGaN, the nuclei 11A can be enlarged by annealing. That is, the nuclei density can be increased to 1×10 11 / cm-2 It can be as follows:
[0128] Although the low-temperature three-dimensional growth layer 12A is made of AlN in the first embodiment, it may be made of AlGaN instead of AlN. However, from the viewpoint of strain relaxation and crystallinity, it is preferable that the difference between the Al composition of the low-temperature three-dimensional growth layer 12A and the Al composition of the crystal nucleus layer 11 is 40% or less.
[0129] Although the high-temperature three-dimensional growth layer 12B is made of AlN in the first embodiment, it may be made of AlGaN instead of AlN. However, from the viewpoint of strain relaxation and crystallinity, it is preferable that the difference in Al composition between the high-temperature three-dimensional growth layer 12B and the low-temperature three-dimensional growth layer 12A be 30% or less.
[0130] In addition, although the two-dimensional growth layer 13 is made of AlN in the first embodiment, it may be made of AlGaN instead of AlN. However, from the viewpoint of strain relaxation and crystallinity, it is preferable that the difference between the Al composition of the two-dimensional growth layer 13 and the Al composition of the high-temperature three-dimensional growth layer 12B be 20% or less.
[0131] In addition, if the device formed on the two-dimensional growth layer 13 is a light-emitting device, the crystal layers below the two-dimensional growth layer 13 must have an Al composition that does not absorb light. If the device to be formed on the two-dimensional growth layer 13 is an ultraviolet-emitting LED, the Al composition of the two-dimensional growth layer 13 is preferably the same as or greater than the Al composition of the n-type layer of the LED.
[0132] For these reasons, most of the device structure is often formed with an Al composition lower than that of the two-dimensional growth layer 13. To alleviate the strain in the device structure, it is preferable to gradually change the lattice mismatch. It is preferable that the Al composition of the three-dimensional growth layer 12 and subsequent layers, excluding the crystal nucleus layer 11, gradually decrease. This makes it possible to alleviate the strain on the device structure.
[0133] As the Al composition of the nuclei 11A decreases, the lattice constant changes from that of AlN to that of GaN. That is, the in-plane lattice constant increases. Therefore, the three-dimensional growth layer 12 formed on the nuclei 11A, which has a larger Al composition than that of the nuclei 11A, experiences tensile strain from the nuclei 11A. As the Al composition of the nuclei 11A decreases, the in-plane lattice constant of the three-dimensional growth layer 12 tends to increase. The three-dimensional growth layer 12 experiences tensile strain during growth, increasing the likelihood of crack formation. Therefore, as described above, there is an optimal value for the difference in Al composition between the nuclei 11A and the three-dimensional growth layer 12. The tensile strain formed in the three-dimensional growth layer 12 and subsequent layers during growth becomes compressive strain when growth is completed and the temperature is raised to room temperature due to the difference in thermal expansion coefficients with the sapphire substrate 10. However, as during growth, the larger the lattice constant of the nuclei 11A, the more the compressive strain occurring in the three-dimensional growth layer 12 and subsequent layers at room temperature is alleviated.
[0134] Furthermore, when the nuclei 11A are AlGaN, the thickness of the two-dimensionally grown layer 13 is preferably 0.5 to 5 μm. The surface of the two-dimensionally grown layer 13 can be sufficiently flattened. For example, the surface roughness RMS can be set to 0.5 to 5 nm. The thickness of the two-dimensionally grown layer 13 is more preferably 1 to 3 μm.
[0135] Even when the nuclei 11A are AlGaN, the threading dislocation density of the two-dimensionally grown layer 13 is 5×10 11 cm -2 or less, thereby obtaining high-quality crystals. Furthermore, in rocking curve measurement of X-ray diffraction of the two-dimensionally grown layer 13, the full width at half maximum (FWHM) of the (002) diffraction line can be set to, for example, 100 to 300 arcsec, and the full width at half maximum of the (102) diffraction line can be set to, for example, 300 to 600 arcsec. This is because the full width at half maximum can be sufficiently reduced by reducing the threading dislocation density.
[0136] (Another variation of embodiment 1) In the first embodiment, the n-side electrode 19 is provided on the same side as the p-side electrode 18, but instead, a light-emitting device with a vertical structure (a structure that provides electrical conduction in a direction perpendicular to the principal surface of the semiconductor layer) can be realized by etching the crystalline nucleus layer 11 side after laser lift-off to expose the n-type layer 14 and providing the n-side electrode 19 on the exposed n-type layer. Furthermore, if the crystalline nucleus layer 11, the three-dimensional growth layer 12, and the two-dimensional growth layer 13 are n-type, a light-emitting device with a vertical structure can be realized by forming the n-side electrode 19 on the back surface of the crystalline nucleus layer 11.
[0137] Although the first embodiment is directed to a light-emitting device, the present invention is not limited to light-emitting devices and can be applied to any semiconductor device. Furthermore, the growth of the two-dimensional growth layer 13 can be completed and the device can be used as a template substrate.
[0138] 15, the surface exposed by the separation of the substrate 10 may be etched to form large irregularities extending from the crystal nucleus layer 11 to the low-temperature three-dimensional growth layer 12A and the high-temperature three-dimensional growth layer 12B. This can further enhance the light scattering effect. In this case, the protrusions 20 may remain or may be completely removed. These irregularities may be random as shown in FIG. 15, or may have a periodic irregular structure. Furthermore, it is preferable that the period of these irregularities is greater than the wavelength of the ultraviolet light emitted from the active layer 15.
[0139] The formation of voids 22 and the separation of substrate 10 may be performed not only after the device is divided, but also in the wafer state. Here, the wafer may be in the state shown in FIG. 5 after the crystal growth process is completed, or in the state after electrode formation and before device separation.
[0140] When bonded to the support substrate 21 as shown in Figure 1, gaps are formed between the p-side electrode 18, the n-side electrode 19, and the support substrate 21. Because these gaps are not supported, they have low mechanical strength. To improve the strength of these areas, an underfill such as resin (liquid curable resin) may be inserted into these gaps.
[0141] The light-emitting element in embodiment 1 may be sealed with a fluororesin film or a quartz lens. In this case, a fluorocarbon compound may be filled between the nucleation layer 11 side of the light-emitting element and the fluororesin film, or between the nucleation layer 11 side and the quartz lens. By reducing the refractive index difference, the light extraction efficiency can be improved.
[0142] The thicknesses of the electron blocking layer 16 and the p-type layer 17 are preferably set so that ultraviolet light traveling from the active layer 15 toward the substrate 10 and light traveling from the active layer 15 toward the p-side electrode 18, reflected by the p-side electrode 18, and traveling toward the substrate 10 are mutually strengthened by interference.
[0143] 3. Experimental Results Next, various experimental results relating to the first embodiment will be described.
[0144] Experiment 1 After fabricating a 1 mm square element structure by the light-emitting device manufacturing method of embodiment 1, laser light was irradiated from the substrate 10 side. The crystal nucleus layer 11 had an Al composition of 79% and a thickness of 50 nm. A ceramic substrate made of AlN was used as the support substrate 21. An ArF excimer laser was used as the laser light, and the energy density was 1.5 J / cm. 2 It was decided.
[0145] The area where voids can be formed by laser irradiation tends to increase as the laser energy density increases, but at 1.5 J / cm 2 With this, it is possible to form voids in an area of approximately 0.5 mm x 0.5 mm. Therefore, in order to form voids in the entire 1 mm square, irradiation was performed on each of the four divided areas.
[0146] This four-part irradiation resulted in overlapping areas of the laser light near the center. The laser energy becomes weaker toward the periphery of the spot, so the size of the voids formed tends to become smaller. Therefore, it is expected that the irradiation energy around the periphery of the element is insufficient.
[0147] Fig. 10 is a photograph of the sample viewed from the side of the substrate 10. The sample was cut along the dotted line in Fig. 10, and cross-sectional SEM images of the region near the interface between the substrate 10 and the crystal nucleus layer 11 were obtained at positions A to H in Fig. 10.
[0148] FIG. 11 shows the cross-sectional SEM image obtained. As shown in FIG. 11, at positions A to G, voids 22 were formed in the sapphire at the interface between the sapphire substrate 10 and the crystal nucleus layer 11. Furthermore, the voids were large near the center of the sample, i.e., in the areas where the laser irradiation overlapped (D, E), and near the center of the laser (C, F), and the voids 22 became smaller as they approached the periphery (B, G). This shows that the stronger the irradiation intensity of the laser light, the larger the voids 22 become, and that it is possible to control the size of the voids 22. In the areas (A, H) corresponding to the periphery of the laser spot, the voids were extremely small or could not be confirmed.
[0149] For these reasons, when actually peeling off the substrate 10, the laser irradiation energy, number of times, and method are adjusted to create uniform voids throughout the entire device. Note that for the sample in Figure 10, conditions were selected to prevent the substrate 10 from peeling off in order to evaluate the cross section.
[0150] Furthermore, EDS analysis was performed on the region of the SEM image shown in FIG. 12, and elemental analysis of Al, Ga, O, and N was performed. FIG. 13 shows elemental mapping. The results in FIG. 13 show that Ga is distributed in a thin layer. Since only the crystalline nucleus layer 11 contains Ga in the SEM image region, it was confirmed that the crystalline nucleus layer 11 remains between the substrate 10 (sapphire) and the low-temperature three-dimensional growth layer 12A (AlN). It was also found that voids 22 were formed on the substrate 10 side, but not in the crystalline nucleus layer 11. It was also found that no damage due to laser light irradiation was observed in the low-temperature three-dimensional growth layer 12A.
[0151] Experiment 2 Three types of samples were fabricated in the same manner as in Experiment 1, with the crystalline nucleus layer 11 being AlN, AlGaN with an Al composition of 89%, and AlGaN with an Al composition of 79%. The energy density of the laser light was 1.5 J / cm when the crystalline nucleus layer 11 was AlN or AlGaN with an Al composition of 89%. 2 , 1.7J / cm 2 In the case of AlGaN with an Al composition of 79%, the 2 , 1.5J / cm 2 Then, a predetermined force was applied to the substrate 10 of each sample, and it was confirmed whether or not the substrate 10 was separated.
[0152] FIG. 14 shows the relationship between the Al composition of the crystal nucleus layer 11 and the energy density of the laser beam (J / cm 2 14 is a graph showing the relationship between the Al content and whether or not the substrate 10 peeled off. From FIG. 14, it was found that the lower the Al content, the smaller the energy density of the laser beam required to peel off the substrate 10. This is presumably because the lower the Al content, the closer the physical properties become to GaN, making it easier to absorb laser beams. This increases the amount of heat generated at the interface between the substrate 10 and the crystalline nucleus layer 11, making it easier for voids 22 to form. From this, it is also presumed that if the Al content is too low, the absorption of laser beams in the crystalline nucleus layer 11 becomes too great, causing the crystalline nucleus layer 11 to reach its decomposition temperature, resulting in the formation of voids 22 in the crystalline nucleus layer 11. To prevent this, it is presumed that the Al content of the crystalline nucleus layer 11 needs to be 50% or more.
[0153] 14, the Al composition of the crystal nucleus layer 11 is set to x, and the energy density of the laser beam (J / cm 2 ) is defined as y, and it is estimated that the substrate 10 can be separated if y≧3x−1.4 is satisfied. [Explanation of symbols]
[0154] 10: Circuit board 11: Crystal nucleation layer 12: Three-dimensional growth layer 12A: Low temperature three-dimensional growth layer 12B: High temperature three-dimensional growth layer 13: Two-dimensional growth layer 14:N-type layer 15:Active layer 16: Electron blocking layer 17:p-type layer 18:p side electrode 19:n side electrode 20: Convex 21: Support substrate 22: Void
Claims
1. a crystalline nucleus layer forming step of generating nuclei of AlGaN or AlN on a substrate made of sapphire to form a crystalline nucleus layer; a semiconductor layer forming step of forming a semiconductor layer made of a Group III nitride semiconductor on the crystalline nucleus layer; a void forming step of irradiating a laser beam from the back side of the substrate, transmitting the laser beam through the substrate and causing the crystal nucleus layer to absorb the laser beam, thereby generating heat in the crystal nucleus layer, and conducting the heat of the crystal nucleus layer to the substrate to decompose a region of the substrate near the interface with the crystal nucleus layer, thereby forming voids; and a substrate separation step of separating the substrate from the crystal nucleus layer at the location of the void.
2. The method for producing a Group III nitride semiconductor according to claim 1 , wherein in the substrate separation step, a convex portion is formed on the surface of the crystal nucleus layer facing the substrate, leaving a part of the substrate.
3. The Al composition of the crystal nucleus layer is defined as x, and the energy density of the laser light (J / cm 2 2. The method for producing a Group III nitride semiconductor according to claim 1, wherein the energy density y of the laser light satisfies y≧3x−1.1, where y is the energy density of the laser light.
4. The energy density of the laser light is 1.6 J / cm 2 The method for producing a Group III nitride semiconductor according to claim 1 .
5. The energy density of the laser light is 5 J / cm 2 5. The method for producing a Group III nitride semiconductor according to claim 3, wherein:
6. 2. The method for producing a Group III nitride semiconductor according to claim 1, wherein the nuclei of the crystal nucleation layer are AlGaN or AlN having an Al composition of 50% or more.
7. the semiconductor layer has a low-temperature three-dimensional growth layer formed on the crystal nucleus layer and a high-temperature three-dimensional growth layer formed on the low-temperature three-dimensional growth layer, The semiconductor layer forming step includes: a low-temperature three-dimensional growth layer formation step of growing AlGaN or AlN from the nuclei at a temperature lower than that of the crystal nucleus layer formation step, thereby combining crystals from adjacent nuclei to form the low-temperature three-dimensional growth layer; and a high-temperature three-dimensional growth layer formation step of growing AlGaN or AlN from the low-temperature three-dimensional growth layer at a temperature higher than that of the low-temperature three-dimensional growth layer formation step but equal to or lower than that of the crystal nucleus layer formation step to form the high-temperature three-dimensional growth layer.
8. the temperature of the crystal nucleus layer forming step is 1100°C or higher and 1200°C or lower, the temperature of the low-temperature three-dimensional growth layer formation step is 900°C or higher and 1100°C or lower; The method for producing a Group III nitride semiconductor according to claim 7 , wherein the temperature in the high-temperature three-dimensional growth layer formation step is 1050° C. or higher and 1200° C. or lower.
9. the semiconductor layer has a low-temperature three-dimensional growth layer formed on the crystal nucleus layer and a high-temperature three-dimensional growth layer formed on the low-temperature three-dimensional growth layer, The semiconductor layer forming step includes: a low-temperature three-dimensional growth layer formation step of growing AlGaN or AlN from the nuclei at a growth rate slower than that of the crystal nucleus layer, and merging crystals from adjacent nuclei to form the low-temperature three-dimensional growth layer; and a high-temperature three-dimensional growth layer formation step of growing AlGaN or AlN from the low-temperature three-dimensional growth layer at a growth rate that is faster than that of the low-temperature three-dimensional growth layer but not greater than that of the crystal nucleus layer to form the high-temperature three-dimensional growth layer.
10. the growth rate of the crystal nucleus layer is 5 nm / min or more and 100 nm / min or less; the growth rate of the low-temperature three-dimensionally grown layer is 2 nm / min or more and 20 nm / min or less; The method for producing a Group III nitride semiconductor according to claim 9 , wherein the growth rate of the high-temperature three-dimensionally grown layer is 5 nm / min or more and 50 nm / min or less.
11. 3. The method for producing a Group III nitride semiconductor according to claim 1, further comprising, after the substrate separation step, a roughness forming step of etching the crystal nucleus layer side exposed by the separation of the substrate to form roughness having a depth extending from the crystal nucleus layer to the semiconductor layer.
12. 2. The method for producing a Group III nitride semiconductor according to claim 1, further comprising, before said crystalline nucleus layer forming step, a thermal cleaning step of heat-treating said substrate in a hydrogen-dominated atmosphere at a temperature higher than that in said crystalline nucleus layer forming step.
13. a crystal nucleation layer which is a layer in which nuclei of AlGaN or AlN are generated and grown; a semiconductor layer formed on one surface of the crystal nucleus layer and made of a Group III nitride semiconductor; a plurality of protrusions made of sapphire formed on the other surface of the crystal nucleus layer; The surface of the crystal nucleus layer is exposed between the adjacent protrusions.
14. 14. The Group III nitride semiconductor according to claim 13, wherein the height of said protrusions is 0.01 to 1 μm, the width of said protrusions is 0.01 to 1 μm, and the center-to-center distance between adjacent said protrusions is 0.02 to 5 μm.
15. The semiconductor layer is a three-dimensional growth layer formed on the crystal nucleus layer, which is a layer in which AlGaN or AlN is grown from the nuclei and crystals from adjacent nuclei are united; 15. The Group III nitride semiconductor according to claim 13 or 14, further comprising: a two-dimensional growth layer formed on the three-dimensional growth layer and being a layer of AlGaN or Ga-doped AlN grown on the three-dimensional growth layer.
Citation Information
Patent Citations
Nitride semiconductor light-emitting device and method for manufacturing the same
JP2018061049A