Optical device and manufacturing method thereof

Abstract

To provide an optical device capable of achieving an improvement in quality of a well layer.SOLUTION: An optical device includes: a well layer 107 comprising an In-containing III nitride semiconductor; a barrier layer (first barrier layer 109) provided on the well layer 107 and comprising a III nitride semiconductor having band gap energy greater than that of the well layer 107; and a compositionally graded layer 108 provided between the well layer 107 and the barrier layer, contacting the well layer 107 and the barrier layer, comprising an In-containing III nitride semiconductor, and having an In composition in a thickness direction which is set to gradually decrease toward the barrier layer. The In composition on a surface of the compositionally graded layer 108 on the well layer 107 side is less than the In composition of the well layer 107.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an optical device and a method for manufacturing the same. [Background technology]

[0002] In light-emitting devices using group III nitride semiconductors, InGaN with a high In composition is required to achieve long-wavelength light emission.

[0003] In recent years, there has been a demand for higher resolution displays, and micro LED displays, in which each pixel is a tiny LED on the order of 1 to 100 μm, have been attracting attention. To achieve full color, red light emission is required, and to achieve this with group III nitride semiconductors, InGaN with a high In composition, as mentioned above, is required.

[0004] Patent Document 1 describes that strain is alleviated by providing a compositionally graded layer between the top surface of the well layer and the barrier layer, in which the In composition decreases toward the barrier layer. It also describes that the In composition of the compositionally graded layer is made to match the In composition of the well layer on the surface facing the well layer, and to match the In composition of the barrier layer on the surface facing the barrier layer. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Publication No. 2022-15556 Summary of the Invention [Problem to be solved by the invention]

[0006] However, the higher the In composition of InGaN, the greater the lattice mismatch, making it more susceptible to misfit dislocations. This is because a GaN template layer is generally formed on a sapphire substrate, which increases the lattice mismatch with GaN. While it is possible to reduce the lattice mismatch by forming an InGaN template layer, forming an InGaN template layer is difficult and expensive. This problem is particularly pronounced in red light emission, which is required for display applications.

[0007] Furthermore, in the method of Patent Document 1, when the In composition of the well layer is high, the difference between the average lattice constant of the compositionally graded layer and the lattice constant of the barrier layer becomes large, which may result in defects in the well layer.

[0008] The present invention has been made in view of the above background, and aims to provide an optical device that can improve the quality of the well layer, and a method for manufacturing the same. [Means for solving the problem]

[0009] One aspect of the present invention is a well layer made of a group III nitride semiconductor containing In; a barrier layer formed on the well layer and made of a Group III nitride semiconductor having a band gap energy larger than that of the well layer; a composition gradient layer provided between the well layer and the barrier layer, in contact with the well layer and the barrier layer, the composition gradient layer being a Group III nitride semiconductor containing In, the In composition in the thickness direction being set to gradually decrease toward the barrier layer; In the optical device, the In composition at the surface of the compositionally graded layer on the well layer side is smaller than the In composition of the well layer.

[0010] Another aspect of the present invention is a well layer formation step of forming a well layer made of a group III nitride semiconductor containing In by an MOCVD method; a compositionally graded layer forming step of forming, by MOCVD, a compositionally graded layer that is in contact with the well layer and is a group III nitride semiconductor containing In, the In composition in the thickness direction gradually decreasing with increasing distance from the well layer, and the In composition at the surface of the compositionally graded layer facing the well layer being smaller than the In composition of the well layer; a barrier layer formation step of forming, by MOCVD, a barrier layer that is in contact with the composition gradient layer and is made of a Group III nitride semiconductor and has a band gap energy larger than that of the well layer; The composition gradient layer forming step is a method for manufacturing an optical device in which the In composition is controlled by the vapor phase ratio of the In raw material gas. [Effects of the Invention]

[0011] In the above embodiment, a compositionally graded layer is provided on the well layer, and the In composition in the thickness direction gradually decreases with increasing distance from the well layer, and the In composition at the surface of the compositionally graded layer facing the well layer is set to be smaller than the In composition of the well layer, thereby improving the quality of the well layer.

[0012] As described above, according to the above aspects, it is possible to provide an optical device that can improve the quality of the well layer, and a method for manufacturing the same. [Brief explanation of the drawings]

[0013] [Figure 1] 1 is a cross-sectional view showing the configuration of a light-emitting device in Embodiment 1, taken along a plane perpendicular to the main surface of a substrate. [Figure 2] 1 shows the energy of the conduction band minimum and the In composition in the band diagrams of the fourth superlattice layer, well layer, compositionally graded layer, and first barrier layer. [Figure 3] 3 is a flowchart showing each step in a method for manufacturing a light-emitting element according to the first embodiment. [Figure 4] 4A to 4C are diagrams showing the configuration of the light-emitting device at each stage of the manufacturing process of the light-emitting device according to the first embodiment, and correspond to step S1 in the flowchart of FIG. [Figure 5]4A to 4C are diagrams showing the configuration of the light-emitting device at each stage of the manufacturing process of the light-emitting device according to the first embodiment, and correspond to step S2 in the flowchart of FIG. [Figure 6] 4A to 4C are diagrams showing the configuration of the light-emitting device at each stage of the manufacturing process of the light-emitting device according to the first embodiment, and correspond to step S3 in the flowchart of FIG. [Figure 7] FIG. 10 is a cross-sectional view showing the configuration of a light-emitting device according to a second embodiment, taken along a plane perpendicular to the main surface of the substrate. [Figure 8] FIG. 4 is a cross-sectional view showing the configuration of a third active layer, taken along a plane perpendicular to the main surface of the substrate. [Figure 9] 1 is a diagram showing the energy of the conduction band minimum and the In composition in the band diagram of the first strain relaxation layer, well layer, compositionally graded layer, and first barrier layer. DETAILED DESCRIPTION OF THE INVENTION

[0014] The optical device comprises a well layer which is a group III nitride semiconductor containing In; a barrier layer which is provided on the well layer and is a group III nitride semiconductor having a band gap energy larger than that of the well layer; and a compositionally graded layer which is provided between the well layer and the barrier layer and is in contact with the well layer and the barrier layer, the compositionally graded layer being a group III nitride semiconductor containing In, the In composition in the thickness direction gradually decreasing toward the barrier layer, and the In composition at the surface of the compositionally graded layer facing the well layer is smaller than the In composition of the well layer.

[0015] The optical device further includes an n-type layer which is an n-type Group III nitride semiconductor, a first active layer which is provided on the n-type layer and is also a Group III nitride semiconductor and has a predetermined emission wavelength, an intermediate layer which is provided on the first active layer and is also a Group III nitride semiconductor, a second active layer which is provided on the intermediate layer and is also a Group III nitride semiconductor and has a longer emission wavelength than the first active layer, a groove which reaches the intermediate layer from the second active layer side, a first p-type layer which is provided on the second active layer and is also a p-type Group III nitride semiconductor, a second p-type layer which is provided on the intermediate layer exposed at the bottom surface of the groove and is also a p-type Group III nitride semiconductor, a first p-side electrode which is provided on the first p-layer, and a second p-side electrode which is provided on the second p-layer, and the second active layer has a well layer, a barrier layer, and a compositionally graded layer.

[0016] In the optical device, the In composition at the surface of the composition-graded layer facing the well layer may be 30% to 70% of the In composition of the well layer. This reduces the difference between the average lattice constant of the well layer and the lattice constant of the barrier layer, thereby reducing defects in the well layer.

[0017] In the optical device, the In composition of the surface of the composition gradient layer on the barrier layer side may be 90% or more and 110% or less of the In composition of the barrier layer, thereby further alleviating strain between the composition gradient layer and the barrier layer.

[0018] In the optical device, the total thickness of the well layer and the compositionally graded layer may be 3 nm or more and 10 nm or less.

[0019] In the optical device, the In composition of the well layer may be 25% or more and 50% or less.

[0020] In the optical device, the average In composition of the combined layer of the well layer and the compositionally graded layer may be set so that the emission wavelength of the well layer is 500 to 800 nm.

[0021] In the optical device, the barrier layer may include a first barrier layer provided on and in contact with the compositionally graded layer, the first barrier layer being a Group III nitride semiconductor having a larger bandgap energy than the well layer and an In composition of 20% or less; a second barrier layer provided on and in contact with the first barrier layer, the second barrier layer being a Group III nitride semiconductor containing Al; and a third barrier layer provided on and in contact with the second barrier layer, the third barrier layer being a Group III nitride semiconductor having a larger bandgap energy than the well layer.

[0022] An optical device having a first superlattice layer having a superlattice structure in which Group III nitride semiconductor thin films with different compositions are alternately stacked, and a second superlattice layer provided on the first superlattice layer and below a well layer, also having a superlattice structure in which Group III nitride semiconductor thin films with different compositions are alternately stacked, wherein the average In composition of the second superlattice layer may be higher than the average In composition of the first superlattice layer. Strain can be relaxed in a stepwise manner.

[0023] The optical device may further include a strain relaxation layer provided below the well layer, having a quantum well structure, and having a second well layer whose thickness is adjusted so as not to emit light, and the wavelength corresponding to the band edge energy of the second well layer of the strain relaxation layer is set to be shorter than the emission wavelength of the well layer, thereby further relaxing the strain.

[0024] In the optical device, the first active layer may have a third well layer made of a group III nitride semiconductor containing In, and a fourth barrier layer provided on the third well layer and made of a group III nitride semiconductor having a band gap energy larger than that of the third well layer. The configuration can be simplified by not providing a compositionally graded layer for the first active layer, which has a shorter emission wavelength than the second active layer.

[0025] The method for manufacturing an optical device includes a well layer formation step of forming a well layer, which is a Group III nitride semiconductor containing In, by an MOCVD method; a compositionally graded layer formation step of forming, on and in contact with the well layer, a compositionally graded layer, which is a Group III nitride semiconductor containing In, by an MOCVD method, such that the In composition in the thickness direction of the compositionally graded layer gradually decreases with increasing distance from the well layer and the In composition at the surface of the compositionally graded layer facing the well layer is smaller than the In composition of the well layer; and a barrier layer formation step of forming, on and in contact with the compositionally graded layer, a barrier layer, which is a Group III nitride semiconductor having a larger band gap energy than the well layer, by an MOCVD method, wherein the In composition may be controlled by the vapor phase ratio of an In source gas.

[0026] (Embodiment 1) 1. Overview of the light-emitting element configuration Fig. 1 is a cross-sectional view showing the configuration of a light-emitting device according to embodiment 1, taken perpendicular to the main surface of a substrate. As shown in Fig. 1, the light-emitting device according to embodiment 1 is a face-up type, and includes a substrate 100, an n-type layer 101, a first superlattice layer 102, a second superlattice layer 103, a third superlattice layer 104, an n-type intermediate layer 105, a fourth superlattice layer 106, a well layer 107, a compositionally graded layer 108, a first barrier layer 109, a second barrier layer 110, a third barrier layer 111, an electron blocking layer 112, a p-type contact layer 113, a transparent electrode 114, a p-side electrode 115, and an n-side electrode 116.

[0027] Although the light emitting element in the first embodiment is a face-up type, it may be a flip-chip type. When a conductive material such as GaN is used for the substrate 100, it may have a vertical structure.

[0028] 2. Details of each component of the light-emitting element The substrate 100 is a growth substrate for growing a group III nitride semiconductor. For example, it may be sapphire, Si, GaN, ScAlMgO4 (SAM), etc. A GaN template substrate in which GaN is grown on sapphire or the like may also be used.

[0029] The n-type layer 101 is an n-type semiconductor layer provided on the substrate 100 via a low-temperature buffer layer or a high-temperature buffer layer (not shown). However, the buffer layer may be provided as needed, and if the substrate is GaN, the buffer layer may not be provided. The n-type layer 11 is, for example, n-GaN, n-AlGaN, or n-InGaN. The Si concentration is, for example, 1×10 18 ~100×10 18 cm -3 is.

[0030] The first superlattice layer 102 is a semiconductor layer provided on the n-type layer 101. The first superlattice layer 102 has a superlattice structure in which group III nitride semiconductor thin films with different compositions are alternately stacked. For example, GaN and InGaN, or two InGaNs with different In compositions. The average In composition of the first superlattice layer 102 is, for example, 2 to 10%. The number of pairs in the superlattice structure is, for example, 3 to 30. It may be undoped, or may contain Si in an amount of 1×10 17 ~100×10 17 cm -3 The first superlattice layer 102 can relax the strain in the well layer 107.

[0031] The second superlattice layer 103 is a semiconductor layer provided on the first superlattice layer 102. The second superlattice layer 103 has a superlattice structure similar to that of the first superlattice layer 102. However, the In composition is different, and the average In composition of the second superlattice layer 103 is higher than that of the first superlattice layer 102, for example, 2 to 15%. The number of pairs in the superlattice structure is, for example, 3 to 30. It may be undoped or may have a doping concentration of 1×10 Si. 17 ~100×10 17 cm -3 The second superlattice layer 103 can further relax the strain in the well layer 107.

[0032] Instead of the second superlattice layer 103, a quantum well structure may be used in which the wavelength corresponding to the band gap energy of the well layer is in the near ultraviolet to blue range (for example, a wavelength of 380 to 480 nm). As with the superlattice structure, the strain in the well layer 107 can be relaxed. The quantum well structure may be either SQW or MQW, but SQW is preferable because it will be thicker. Note that even with a quantum well structure, the presence of the n-type intermediate layer 105 above the second superlattice layer 103 prevents actual light emission.

[0033] The third superlattice layer 104 is a semiconductor layer provided on the second superlattice layer 103. The third superlattice layer 104 has a superlattice structure similar to that of the first superlattice layer 102. However, the In composition is different, and the average In composition of the third superlattice layer 104 is higher than the average In composition of the second superlattice layer 103, for example, 2 to 20%. The number of pairs in the superlattice structure is, for example, 3 to 30. It may be undoped or may have a doping ratio of 1×10 17 ~100×10 17 cm -3 The third superlattice layer 104 can further relax the strain in the well layer 107.

[0034] Instead of the third superlattice layer 104, a quantum well structure may be used in which the wavelength corresponding to the band gap energy of the well layer is blue-green to yellow-green (for example, a wavelength of 480 to 580 nm). As with the superlattice structure, the strain in the well layer 107 can be relaxed. The quantum well structure may be either SQW or MQW, but SQW is preferable because it will be thicker. Note that even with a quantum well structure, the presence of the n-type intermediate layer 105 on the third superlattice layer 104 will prevent actual light emission.

[0035] As described above, by providing three superlattice structures, namely, the first superlattice layer 102, the second superlattice layer 103, and the third superlattice layer 104, and gradually increasing the In composition, strain is gradually alleviated and defects are prevented from occurring.

[0036] N-type intermediate layer 105 is a semiconductor layer provided on third superlattice layer 104. N-type intermediate layer 105 is a layer made of n-type GaN or InGaN. When InGaN is used, the In composition is, for example, 15% or less. By providing n-type intermediate layer 105, it is possible to prevent first superlattice layer 102, second superlattice layer 103, and third superlattice layer 104 from emitting light.

[0037] The fourth superlattice layer 106 is a layer provided on the n-type intermediate layer 105. The fourth superlattice layer 106 has a superlattice structure similar to that of the first superlattice layer 102. However, the In composition is different, and the average In composition of the third superlattice layer 104 is equal to or greater than the average In composition of the first superlattice layer 102 and equal to or less than the average In composition of the second superlattice layer 103. The average In composition is, for example, 2 to 10%. The number of pairs in the superlattice structure is, for example, 1 to 30. It may be undoped, or may contain Si in an amount of 1×10 17 ~100×10 17 cm -3 The fourth superlattice layer 106 can further relax the strain in the well layer 107. The fourth superlattice layer 106 also functions as an n-side barrier layer for the well layer 107.

[0038] Instead of or in addition to the fourth superlattice layer 106, a first strain relaxation layer 16A and a second strain relaxation layer 16B as described in the second embodiment may be provided. As described below, this allows for efficient strain relaxation. Only one of the first strain relaxation layer 16A and the second strain relaxation layer 16B may be provided.

[0039] The well layer 107 is a layer provided on the fourth superlattice layer 106. The well layer 107 is made of a group III nitride semiconductor containing undoped In, such as InGaN. The emission wavelength of the light emitting device in embodiment 1 is determined by the average In composition of the combined layer of the well layer 107 and the compositionally graded layer 108. The average In composition is set to achieve an emission wavelength in the yellow to near-infrared range. For example, the In composition of the well layer 107 is 25 to 50%. The present invention is particularly effective for red light emission, in which case the In composition is 30 to 45%.

[0040] Although the first embodiment has an SQW structure with one well layer 107, an MQW structure with two or more well layers 107 may also be used. In this case, a compositionally graded layer 108 is provided on each well layer 107. The number of pairs is preferably 1 to 5, and more preferably 1 to 3.

[0041] The compositionally graded layer 108 is a semiconductor layer provided on and in contact with the well layer 107. The compositionally graded layer 108 is made of a group III nitride semiconductor containing undoped In, such as InGaN.

[0042] The In composition of the compositionally graded layer 108 will be described with reference to Fig. 2. Fig. 2 is a diagram showing the energy of the bottom of the conduction band in a band diagram and the In composition for the fourth superlattice layer 106, the well layer 107, the compositionally graded layer 108, and the first barrier layer 109.

[0043] As shown in FIG. 2, the compositionally graded layer 108 is configured so that the In composition in the thickness direction gradually decreases toward the first barrier layer 109 (as it moves away from the well layer 107). Conversely, the energy of the conduction band minimum of the compositionally graded layer 108 is configured so that it gradually increases toward the first barrier layer 109. While FIG. 2 shows a case where the In composition is configured to decrease continuously and linearly, it may also decrease in a stepwise manner. Alternatively, it may decrease continuously and curvedly, for example, exponentially. Furthermore, if the In composition decreases linearly, the slope of the line may be changed in two or more steps.

[0044] Setting the In composition of the compositionally graded layer 108 in this manner makes it possible to alleviate strain between the compositionally graded layer 108 and the first barrier layer 109. Furthermore, as shown in FIG. 2, the energy of the conduction band minimum of the compositionally graded layer 108 decreases toward the well layer 107. Therefore, the compositionally graded layer 108 also contributes to electron confinement in the well layer 107. Therefore, the effective thickness of the well layer 107 can be set to the total thickness of the well layer 107 and the compositionally graded layer 108, and the well layer 107 can be made substantially thick. As a result, carrier confinement is improved.

[0045] The In composition at the surface of the composition graded layer 108 facing the well layer 107 is set to a smaller value than the In composition of the well layer 107. This makes it possible to reduce the difference between the average lattice constant of the composition graded layer 108 and the lattice constant of the first barrier layer 109, compared to when the In composition is the same as that of the well layer 107, and makes it less likely that relaxation accompanied by defects will occur in the well layer 107 or the composition graded layer 108. Preferably, the In composition at the surface of the composition graded layer 108 facing the well layer 107 is 30 to 70% of the In composition of the well layer 107, and more preferably 40 to 60%.

[0046] The In composition of the surface of the compositionally graded layer 108 on the side of the first barrier layer 109 is set to the same value as the In composition of the first barrier layer 109. This makes it possible to effectively alleviate strain between the compositionally graded layer 108 and the first barrier layer 109. Note that the In composition does not need to be exactly the same as that of the first barrier layer 109, but it is preferable that the difference be as small as possible, and it is preferably 90 to 110% of the In composition of the first barrier layer 109. It is more preferably 95 to 105%.

[0047] As a result of setting the In composition of the compositionally graded layer 108 as described above, it is possible to maintain the surface states of the well layer 107 and the compositionally graded layer 108. This also makes it possible to improve the quality of the barrier layers (first barrier layer 109, second barrier layer 110, third barrier layer 111) formed on the compositionally graded layer 108.

[0048] In manufacturing, the gas flow is set so that the In composition becomes steep at the interface between the well layer 107 and the compositionally graded layer 108 and at the interface between the compositionally graded layer 108 and the first barrier layer 109. However, in the actual structure, the In composition may be slightly graded near the interfaces.

[0049] The total thickness of the well layer 107 and the compositionally graded layer 108 is preferably 3 nm or more. If the total thickness of the well layer 107 and the compositionally graded layer 108 is too small, a subband is formed, shifting the wavelength to the shorter wavelength side. The total thickness of the well layer 107 and the compositionally graded layer 108 is preferably 10 nm or less. If the well layer 107 is too thick, strain relaxation occurs due to the high In composition, which may cause defects in the well layer 107 and the compositionally graded layer 108. A thickness of 3 to 5 nm is more preferable. The thickness of the well layer 107 may be equal to or less than the critical thickness at which defects occur.

[0050] The difference between the maximum and minimum values ​​of the In composition in the compositionally graded layer 108 is preferably 10 to 30%. The rate of change of the In composition in the compositionally graded layer 108 is preferably 5 to 17% / nm.

[0051] The effects of the compositionally graded layer 108 can be summarized as follows. First, the strain between the compositionally graded layer 108 and the first barrier layer 109 can be relaxed, allowing the well layer 107 to be substantially thicker. As a result, carrier confinement is improved. Second, the strain due to the difference in lattice constant with the first barrier layer 109 can be reduced, making it difficult for defect-related relaxation to occur in the well layer 107 or the compositionally graded layer 108. Third, the quality of the first barrier layer 109, the second barrier layer 110, and the third barrier layer 111 can be improved. These effects make it possible to form a quantum well structure without degrading the crystalline quality of the well layer 107.

[0052] The first barrier layer 109 is a semiconductor layer provided on and in contact with the compositionally graded layer 108. The first barrier layer 109 functions as a p-side barrier layer for confining carriers in the well layer 107, and also functions as a protective layer for preventing In from evaporating from the well layer 107 during device formation. The first barrier layer 109 is made of undoped GaN or InGaN. The first barrier layer 109 has a larger band gap energy than the well layer 107, and has an In composition of, for example, 15% or less. The In composition is constant in the thickness direction. The thickness of the first barrier layer 109 is, for example, 0.5 to 10 nm.

[0053] The second barrier layer 110 is a semiconductor layer provided on and in contact with the first barrier layer 109. Like the first barrier layer 109, the second barrier layer 110 functions as a p-side barrier layer for confining carriers in the well layer 107, and also functions as a protective layer for preventing In from evaporating from the well layer 107 during device formation. The second barrier layer 110 is made of undoped AlGaN or AlN. By using a material with a larger bandgap energy than the first barrier layer 109, the function as a barrier layer is improved, and carriers can be efficiently confined in the well layer 107 and the compositionally graded layer 108. The thickness of the second barrier layer 110 is, for example, 0.5 to 10 nm.

[0054] The third barrier layer 111 is a semiconductor layer provided on and in contact with the second barrier layer 110. The third barrier layer 111 functions as a p-side barrier layer for confining carriers in the well layer 107. The third barrier layer 111 is made of an undoped Group III nitride semiconductor having a bandgap energy larger than that of the well layer 107, such as InGaN, GaN, AlGaN, or AlGaInN. GaN is particularly preferred. The thickness of the third barrier layer 111 is, for example, 2 to 50 nm.

[0055] The electron blocking layer 112 is a p-type semiconductor layer provided on the third barrier layer 111. The electron blocking layer 112 is a layer that blocks electrons injected from the n-type layer 101 in order to efficiently confine them in the well layer 107. Furthermore, by making the electron blocking layer 112 p-type, holes can be efficiently injected into the well layer 107, and a larger barrier against electrons can be formed, thereby enhancing the electron blocking function. The Mg concentration of the electron blocking layer 112 is 1×10 19 ~100×10 19 cm -3 The electron blocking layer 112 may be undoped.

[0056] The electron blocking layer 112 may be a single layer of GaN or AlGaN, or may have a structure in which two or more of AlGaN, GaN, and InGaN are stacked, or a structure in which layers are stacked with only the composition ratio changed. It may also have a superlattice structure. The superlattice structure can block electrons more efficiently. Examples of the superlattice structure include a structure in which p-AlGaN and p-InGaN are stacked alternately, or a structure in which p-AlGaN and p-GaN are stacked alternately.

[0057] The thickness of the electron blocking layer 112 is, for example, 2 to 50 nm, and preferably 2 to 25 nm.

[0058] The p-type contact layer 113 is a p-type semiconductor layer provided on the electron blocking layer 112. The p-type contact layer 113 is composed of a first layer and a second layer in this order from the electron blocking layer 112 side.

[0059] The first layer is preferably p-GaN or p-InGaN. The thickness of the first layer is preferably 10 to 500 nm, more preferably 10 to 200 nm, and further preferably 10 to 100 nm. The Mg concentration of the first layer is 1×10 19 ~100×10 19 cm -3 The second layer is preferably p-GaN or p-InGaN. The thickness of the second layer is preferably 2 to 50 nm, more preferably 4 to 20 nm, and further preferably 6 to 10 nm. The Mg concentration of the second layer is 1×10 20 ~100×10 20 cm -3 It is best to do so.

[0060] A recess 117 is provided on the surface of the p-type contact layer 113. The recess 117 has a depth that reaches the n-type layer 101. The n-type layer 101 is exposed at the bottom of the recess 117.

[0061] The transparent electrode 114 is an electrode provided on the p-type contact layer 113. The transparent electrode 114 is made of a transparent conductive material, such as ITO, IZO, ICO, or ZnO.

[0062] The p-side electrode 115 is an electrode provided on the transparent electrode 114. For example, it is made of Ni / Au.

[0063] The n-side electrode 116 is an electrode provided on the bottom surface of the recess 117. It is made of, for example, Ti / Al.

[0064] As described above, according to the light-emitting device of embodiment 1, the compositionally graded layer 108 is provided between the well layer 107 and the first barrier layer 109, the In composition of the compositionally graded layer 108 is set to decrease toward the first barrier layer 109, and the In composition of the surface of the compositionally graded layer 108 on the well layer 107 side is set to a smaller value than the In composition of the well layer 107, thereby improving the crystal quality of the well layer 107.

[0065] 3. Light-emitting device manufacturing method Next, a method for manufacturing the light-emitting device according to embodiment 1 will be described with reference to the drawings. Fig. 3 is a flowchart showing the manufacturing process of the light-emitting device according to embodiment 1, and Figs. 4 to 6 are cross-sectional views showing the configuration of the light-emitting device at each stage of the manufacturing process.

[0066] First, the substrate 100 is prepared, and hydrogen, nitrogen, and, if necessary, ammonia are added to perform a heat treatment on the substrate.

[0067] Next, a buffer layer is formed on substrate 100, and then n-type layer 101, first superlattice layer 102, second superlattice layer 103, third superlattice layer 104, n-type intermediate layer 105, fourth superlattice layer 106, well layer 107, compositionally graded layer 108, first barrier layer 109, second barrier layer 110, third barrier layer 111, electron blocking layer 112, and p-type contact layer 113 are formed in this order on the buffer layer (see FIG. 4, step S1 in FIG. 3). Each layer is formed using MOCVD.

[0068] Here, the growth temperature of the well layer 107 and the compositionally graded layer 108 is preferably 500 to 900°C. This can improve the crystal quality and increase the light emission efficiency. Within the above temperature range, the well layer 107 and the compositionally graded layer 108 may be grown at different temperatures. For example, the growth temperature of the well layer 107 may be lower than the growth temperature of the compositionally graded layer 108.

[0069] The In composition of the compositionally graded layer 108 is controlled by changing the vapor phase ratio of the In source gas. This makes it possible to easily achieve the gradient of the In composition as shown in Fig. 2. The vapor phase ratio of the In source gas is the molar fraction of the In source gas relative to the Group III source gas.

[0070] The first barrier layer 109 and the second barrier layer 110 are formed at the same growth temperature as the compositionally graded layer 108. This makes it possible to suppress thermal damage to the well layer 107. Furthermore, the first barrier layer 109 and the second barrier layer 110 can be formed without evaporating In from the well layer 107 or the compositionally graded layer 108. The growth temperatures do not need to be exactly the same, and an error of about -20 to +20°C is permissible.

[0071] The third barrier layer 111 is formed at a higher growth temperature than the first barrier layer 109 and the second barrier layer 110. Increasing the temperature can improve the crystallinity. Furthermore, even if the temperature is increased, the first barrier layer 109 and the second barrier layer 110 function as protective layers, so evaporation of In from the well layer 107 and the compositionally graded layer 108 can be suppressed.

[0072] Next, the transparent electrode 114 is formed in a predetermined region on the p-type contact layer 113 by sputtering or vapor deposition (see FIG. 5, step S2 in FIG. 3).

[0073] Next, predetermined regions on the surface of the p-type contact layer 113 are dry-etched until the n-type layer 101 is exposed, thereby forming recesses 117 (see FIG. 6, step S3 in FIG. 3).

[0074] Next, a p-side electrode 115 is formed on the transparent electrode 114 and an n-side electrode 116 is formed on the bottom surface of the recess 117 by vapor deposition or sputtering (see FIG. 1, step S4 in FIG. 3). In this way, the light emitting device of Embodiment 1 is manufactured.

[0075] (Embodiment 2) 1. Overview of the light-emitting element configuration 7 is a cross-sectional view showing the configuration of a light-emitting element in embodiment 2, taken perpendicular to the main surface of the substrate. The light-emitting element in embodiment 2 is capable of emitting blue, green, and red light. The light-emitting element in embodiment 2 is a flip-chip type that extracts light from the back side of the substrate, and is mounted face-down on a mounting substrate (not shown).

[0076] The light-emitting element in the second embodiment may be a monolithic micro LED display element. That is, the blue, green, and red light-emitting portions in the second embodiment may be one pixel, and the structure of one pixel may be arranged in a matrix on the same substrate. This structure makes it possible to realize a display with one chip. In this case, the size of the subpixel is, for example, 1 to 100 μm.

[0077] Furthermore, the light emitting device in the second embodiment may have a structure in which one pixel is one chip.

[0078] 2. Details of each component of the light-emitting element As shown in FIG. 7, the light-emitting element in this embodiment has a substrate 10, an n-type layer 11, a first active layer 12, a first intermediate layer 13, a second active layer 14, a second intermediate layer 15, a third active layer 16, electron blocking layers 17, 19A to 19C, p-type layers 18, 20A to 20C, an n-side electrode 21, and p-side electrodes 22A to 22C.

[0079] The substrate 10 is a growth substrate for growing a group III nitride semiconductor, such as sapphire, Si, GaN, or ScAlMgO4 (SAM).

[0080] The n-type layer 11 is an n-type semiconductor provided on the substrate 10 via a low-temperature buffer layer or a high-temperature buffer layer (not shown). However, the buffer layer may be provided as needed, and if the substrate is GaN, the buffer layer may not be provided. The n-type layer 11 is, for example, n-GaN, n-AlGaN, or n-InGaN. The Si concentration is, for example, 1×10 18 ~100×10 18 cm -3 is.

[0081] The first active layer 12 is a light emitting layer of SQW or MQW structure provided on the n-type layer 11. The emitted light wavelength is blue, 430 to 480 nm. The first active layer 12 has a structure in which barrier layers made of AlGaN and well layers made of InGaN are alternately stacked in 1 to 9 pairs, more preferably 1 to 7 pairs, and even more preferably 1 to 5 pairs.

[0082] An underlayer may be provided between the n-type layer 11 and the first active layer 12, if necessary. The underlayer is a semiconductor layer with a superlattice structure provided on the n-type layer 11, and serves to alleviate lattice distortion of the semiconductor layer formed on the underlayer. The underlayer is formed by alternately laminating III nitride semiconductor thin films with different compositions (for example, two of GaN, InGaN, and AlGaN), and the number of pairs is, for example, 3 to 30. The underlayer may be undoped or may contain 1×10 Si. 17 ~100×10 17 cm -3 Also, as long as the strain can be relaxed, the superlattice structure is not necessary.

[0083] An ESD layer may be provided between the n-type layer 11 and the underlayer. The ESD layer is a layer provided to improve electrostatic breakdown voltage. The ESD layer may be made of, for example, undoped or lightly Si-doped GaN, InGaN, or AlGaN.

[0084] The first intermediate layer 13 is a semiconductor layer provided on the first active layer 12. The first intermediate layer 13 is a layer provided to enable separate control of light emission from the first active layer 12 and light emission from the second active layer 14. The first intermediate layer 13 also serves to protect the first active layer 12 from etching damage when forming second grooves 31, which will be described later.

[0085] The first intermediate layer 13 has a structure in which an undoped intermediate layer 13A and an n-type intermediate layer 13B are laminated in this order from the first active layer 12 side. The undoped intermediate layer 12A and the n-type intermediate layer 12B may be made of the same material except for impurities. The reason for making the first intermediate layer 13 have such a two-layer structure will be explained later.

[0086] The material of the first intermediate layer 13 is a group III nitride semiconductor containing In, such as InGaN. The surfactant effect of In can suppress roughness on the surface of the first intermediate layer 13, improving surface flatness. It can also alleviate lattice distortion.

[0087] The In composition of the first intermediate layer 13 (the molar ratio of In to the total Group III metals of the Group III nitride semiconductor) may be set to have a band gap that does not absorb light emitted from the first active layer 12 and the second active layer 14. A preferred In composition is 10% or less, more preferably 5% or less, and even more preferably 2% or less. If the In composition is greater than 10%, it will cause the surface of the first intermediate layer 13 to become rough. The In content may be any value greater than 0%, and may be at a doping level (a level that does not form a mixed crystal). For example, if the In concentration is 1×10 14 cm -3 More than 1×10 22 cm -3 The following is GaN.

[0088] The non-doped intermediate layer 13A is non-doped, and the n-type intermediate layer 13B is Si-doped. The Si concentration of the n-type intermediate layer 13B is 1×10 17 ~1000×10 17 cm -3 It is preferable to set it to 10×10 17 ~100×1017 cm -3 , and more preferably 20×10 17 ~80×10 17 cm -3 The n-type intermediate layer 13B may be modulation doped with Si, or a part of the n-type intermediate layer 13B may have an undoped region.

[0089] The thickness of the first intermediate layer 13 is preferably 20 to 150 nm. If it is thicker than 150 nm, the surface of the first intermediate layer 13 may become rough. If it is thinner than 20 nm, it may be difficult to control the depth of the second grooves 31 (described later) so that they lie within the non-doped intermediate layer 13A when they are formed. The thickness is more preferably 30 to 100 nm, and even more preferably 50 to 80 nm.

[0090] The thickness of the non-doped intermediate layer 13A is preferably 10 nm or more in order to control the etching depth and avoid etching damage to the first active layer 12. The thickness of the n-type intermediate layer 13B is preferably 10 nm or more in order to independently control the light-emitting characteristics of each active layer.

[0091] The second active layer 14 is a layer provided on the first intermediate layer 13 and has a quantum well structure of SQW or MQW. The emission wavelength is green and is 510 to 570 nm. The quantum well structure is formed by alternately stacking 1 to 7 pairs of barrier layers made of GaN or AlGaN and well layers made of InGaN.

[0092] A strain relaxation layer may be provided between the first intermediate layer 13 and the second active layer 14. By providing the strain relaxation layer, the strain in the second active layer 14 stacked thereon can be relaxed, improving the crystal quality. The strain relaxation layer has an SQW structure or MQW structure in which a barrier layer and a well layer are stacked in order, and is a quantum well structure in which the thickness of the well layer is adjusted to be thin so as not to emit light. For example, light emission can be prevented by setting the thickness of the well layer to 1 nm or less. The barrier layer is made of AlGaN, and the well layer is made of InGaN. The wavelength corresponding to the band edge energy of the well layer of the strain relaxation layer needs to be shorter than the emission wavelength of the second active layer 14; for example, if the emission wavelength is 500 to 560 nm, it is 400 to 460 nm.

[0093] The second intermediate layer 15 is a semiconductor layer provided on the second active layer 14. The second intermediate layer 15 is provided for the same reason as the first intermediate layer 13, and is a layer provided to enable separate control of the light emission from the second active layer 14 and the light emission from the third active layer 16. The second intermediate layer 15 also serves to protect the second active layer 14 from etching damage when forming the first grooves 30 described below.

[0094] The second intermediate layer 15 has a structure in which an undoped intermediate layer 15A and an n-type intermediate layer 15B are laminated in this order from the second active layer 14 side. The undoped intermediate layer 15A and the n-type intermediate layer 15B have the same structures as the undoped intermediate layer 13A and the n-type intermediate layer 13B. In other words, the undoped intermediate layer 15A and the n-type intermediate layer 15B are made of the same materials as the undoped intermediate layer 13A and the n-type intermediate layer 13B except for the impurities, and the thickness ranges and the like are also the same as the undoped intermediate layer 13A and the n-type intermediate layer 13B. The undoped intermediate layer 15A is undoped, and the n-type intermediate layer 15B is Si-doped.

[0095] The third active layer 16 is a layer provided on the second intermediate layer 15. As shown in Fig. 8, the third active layer 16 has a structure in which, from the second intermediate layer 15 side, a first strain relaxation layer 16A, a second strain relaxation layer 16B, a well layer 16C, a compositionally graded layer 16D, a first barrier layer 16E, a second barrier layer 16F, and a third barrier layer 16G are stacked.

[0096] The first strain relaxation layer 16A is a semiconductor layer provided on the second intermediate layer 15. The first strain relaxation layer 16A has an SQW structure or MQW structure in which a barrier layer and a well layer are stacked in this order, and is a quantum well structure in which the thickness of the well layer is adjusted to be thin so as not to emit light. For example, light emission can be prevented by setting the well layer thickness to 1 nm or less. The barrier layer is AlGaN, and the well layer is InGaN. The wavelength corresponding to the band edge energy of the well layer of the strain relaxation layer needs to be shorter than the emission wavelength of the second active layer 14; for example, if the emission wavelength is 500 to 560 nm, it is 400 to 460 nm. Preferably, it is 40 to 100 nm shorter than the emission wavelength of the second active layer 14. In this case, the growth temperature of the first strain relaxation layer 16A is 700 to 800°C.

[0097] The wavelength corresponding to the band edge energy of the well layer of the first strain relaxation layer 16A may be equal to the emission wavelength of the first active layer 12. In this case, the first strain relaxation layer 16A may be grown at the same growth temperature as the first active layer 12.

[0098] The band edge energy of the well layer of the first strain relaxation layer 16A can be controlled by the thickness of the well layer. Specifically, by making the well layer thickness of the first strain relaxation layer 16A sufficiently thin, the subband energy in the well increases, resulting in a larger band edge energy. This allows the emission wavelength to be shorter than that of the second active layer 14. The growth temperature is optional, but the layer may be grown at the same temperature as that of the second active layer. Furthermore, by making the well layer thickness of the first strain relaxation layer 16A thinner, the subband energy further increases, reducing the energy difference with the barrier layer. That is, the band edge energy approaches that of the barrier layer. As a result, carrier confinement in the well layer of the first strain relaxation layer 16A becomes difficult, making it difficult to emit light. This allows the well layer to function as part of the barrier layer of the third active layer 16 and simultaneously achieve strain relaxation. Thus, by forming a first strain relaxation layer 16A with a well layer that has poorer carrier confinement than the well layer of the third active layer 16, a non-emitting first strain relaxation layer 16A can be formed.

[0099] In short, the material and layer structure of the first strain relaxation layer 16A are set so that the average lattice constant of the entire first strain relaxation layer 16A is between the lattice constant of the second intermediate layer 15 and the lattice constant of the second strain relaxation layer 16B, and the thickness of the well layer is set so that the first strain relaxation layer 16A does not emit light.

[0100] The first strain relaxation layer 16A may have either an SQW structure or an MQW structure, but the SQW structure is preferable because the third active layer 16 would be thicker.

[0101] By providing the first strain relaxation layer 16A as described above, the strain in the well layer 16C thereon can be relaxed, and the crystal quality of the well layer 16C can be improved.

[0102] The second strain relaxation layer 16B is a semiconductor layer provided on the first strain relaxation layer 16A. The wavelength corresponding to the band edge energy of the well layer of the second strain relaxation layer 16B is shorter than the emission wavelength of the well layer 16C and longer than the wavelength corresponding to the band edge energy of the well layer of the first strain relaxation layer 16A. For example, it is 510 to 570 nm. Other than that, it is the same as the first strain relaxation layer 16A.

[0103] It is preferable that the difference between the wavelength corresponding to the band edge energy of the well layer of the first strain relaxation layer 16A and the wavelength corresponding to the band edge energy of the well layer of the second strain relaxation layer 16B, and the difference between the wavelength corresponding to the band edge energy of the well layer of the second strain relaxation layer 16B and the emission wavelength of the well layer 16C are 40 to 100 nm.

[0104] By providing the first strain relaxation layer 16A and the second strain relaxation layer 16B in this manner, strain can be relaxed in stages, and the strain in the well layer 16C stacked thereon can be effectively relaxed, thereby improving the quality of the well layer 16C.

[0105] In the second embodiment, the first strain relaxation layer 16A and the second strain relaxation layer 16B are used to relax the strain in two stages. However, three or more strain relaxation layers may be provided to relax the strain in three or more stages. Alternatively, only one of the first strain relaxation layer 16A and the second strain relaxation layer 16B may be provided. The second active layer 14 may also include multiple strain relaxation layers to relax the strain in stages. Alternatively, the fourth superlattice layer 106 in the first embodiment may be provided instead of or in addition to the first strain relaxation layer 16A and the second strain relaxation layer 16B.

[0106] The well layer 16C is a semiconductor layer provided on the second strain relaxation layer 16B. The well layer 16C has the same structure as the well layer 107 in the first embodiment. The emission wavelength of the third active layer 16 of the light-emitting device in the second embodiment is determined by the average In composition of the combined layer of the well layer 16C and the compositionally graded layer 16D. The average In composition is set to achieve a red emission wavelength (590 to 750 nm).

[0107] The compositionally graded layer 16D is a semiconductor layer provided on and in contact with the well layer 16C. The compositionally graded layer 16D has the same structure as the compositionally graded layer 108 in the first embodiment. The In composition of the compositionally graded layer 16D will be described with reference to Fig. 9. Fig. 9 is a diagram showing the energy of the bottom of the conduction band and the In composition in a band diagram for the second strain relaxation layer 16B, the well layer 16C, the compositionally graded layer 16D, and the first barrier layer 16E.

[0108] 9, the composition gradient layer 16D is configured so that the In composition in the thickness direction gradually decreases toward the first barrier layer 16E. Conversely, the energy of the conduction band minimum of the composition gradient layer 16D is configured so that it gradually increases toward the first barrier layer 16E.

[0109] The In composition of the surface of the composition-graded layer 16D facing the well layer 16C is set to a value smaller than the In composition of the well layer 16C. This reduces strain due to the difference between the average lattice constant of the composition-graded layer 16D and the lattice constant of the first barrier layer 16E, compared to when the In composition is the same as that of the well layer 16C, making it less likely that relaxation accompanied by defects will occur in the well layer 16C or the composition-graded layer 16D. Preferably, the In composition of the surface of the composition-graded layer 16D facing the well layer 16C is 30 to 70% of the In composition of the well layer 16C, and more preferably 40 to 60%.

[0110] The In composition of the surface of the composition gradient layer 16D facing the first barrier layer 16E is set to the same value as the In composition of the first barrier layer 16E. This effectively alleviates strain between the composition gradient layer 16D and the first barrier layer 16E. Note that the In composition does not need to be exactly the same as that of the first barrier layer 16E, but it is preferable that the difference be as small as possible, and it is preferably 80 to 110% of the In composition of the first barrier layer 109. It is more preferably 90 to 105%.

[0111] Setting the In composition of the compositionally graded layer 16D as described above provides the same effects as the compositionally graded layer 108 in the first embodiment. Specifically, the following effects can be achieved. First, the strain between the compositionally graded layer 16D and the first barrier layer 16E can be relaxed, allowing the well layer 16C to be substantially thicker. As a result, carrier confinement is improved. Second, the strain due to the difference in lattice constant with the first barrier layer 16E can be reduced, making it difficult for defect-related relaxation to occur in the well layer 16C and the compositionally graded layer 16D. Third, the quality of the first barrier layer 16E, the second barrier layer 16F, and the third barrier layer 16G can be improved. These effects enable the formation of a quantum well structure without degrading the crystalline quality of the well layer 16C.

[0112] In addition, various modifications and preferred examples of the compositionally graded layer 108 described in the first embodiment can be similarly applied to the compositionally graded layer 16D in the second embodiment.

[0113] The first barrier layer 16E, the second barrier layer 16F, and the third barrier layer 16G have the same structures as the first barrier layer 109, the second barrier layer 110, and the third barrier layer 111 in the first embodiment, respectively, and therefore description thereof will be omitted.

[0114] In the first active layer 12 and the second active layer 14, a barrier layer is provided in contact with the well layer, and no compositionally graded layer is provided, but a compositionally graded layer may be provided as in the third active layer 16. However, in the blue to green wavelength region, sufficient luminous efficiency can be obtained without providing a compositionally graded layer, so it is preferable to simplify the configuration by not providing a compositionally graded layer as in the second embodiment.

[0115] The electron blocking layer 17 is a semiconductor layer provided on the third active layer 16. The electron blocking layer 17 is a layer that blocks electrons injected from the n-type layer 11 in order to confine them efficiently in the third active layer 16. The electron blocking layer 17 not only functions as an electron blocking layer but also as a protective layer that protects the active layer. The electron blocking layer 17 may be made of a material that has a wider band gap than the well layer of the third active layer 16, such as AlGaN, GaN, or InGaN. The thickness of the electron blocking layer 17 is preferably 2 to 50 nm, and more preferably 2 to 25 nm. The electron blocking layer 17 may be doped with an impurity, or may be doped with Mg. In this case, the Mg concentration should be 1×10 18 ~1000×10 18 cm -3 It is best to do so.

[0116] Between the third active layer 16 and the electron blocking layer 17, a strain relaxation layer made of InGaN having a narrower band gap than the electron blocking layer 17 may be provided.

[0117] The p-type layer 18 is a semiconductor layer provided on the electron blocking layer 17. The p-type layer 18 is preferably p-GaN or p-InGaN. The thickness of the p-type layer 18 is preferably 10 to 500 nm, more preferably 10 to 200 nm, and even more preferably 10 to 100 nm. The Mg concentration of the p-type layer 18 is 1×10 19 ~100×10 19 cm -3It is best to do so.

[0118] A portion of the surface of the p-type layer 18 is etched to form grooves, and a first groove 30 extending from the p-type layer 18 to the second intermediate layer 15, a second groove 31 extending to the first intermediate layer 13, and a third groove 32 extending to the n-type layer 11 are provided.

[0119] The first grooves 30 have a depth that reaches the non-doped intermediate layer 15A of the second intermediate layer 15. By removing the n-type intermediate layer 15B of the second intermediate layer 15 below the p-side electrode 22B in this way, an n-type layer is prevented from being positioned above the second active layer 14, allowing the second active layer 14 to emit light.

[0120] The second grooves 31 are deep enough to reach the non-doped intermediate layer 13A of the first intermediate layer 13. This is also for the same reason, and by removing the n-type intermediate layer 13B of the first intermediate layer 13 below the p-side electrode 22C, an n-type layer is prevented from being positioned above the first active layer 12, allowing the first active layer 12 to emit light.

[0121] The electron blocking layers 19A to 19C are semiconductor layers respectively provided on the p-type layer 18, on the undoped intermediate layer 15A exposed at the bottom of the first groove 30, and on the undoped intermediate layer 13A exposed at the bottom of the second groove 31, and are layers that block electrons injected from the n-type layer 11 in order to efficiently confine them in the first active layer 12, the second active layer 14, and the third active layer 16.

[0122] The electron blocking layers 19A to 19C may be single layers of GaN or AlGaN, or may have a structure in which two or more of AlGaN, GaN, and InGaN are stacked, or a structure in which layers are stacked with only the composition ratio changed. They may also have a superlattice structure. The superlattice structure can block electrons more efficiently. Examples of the superlattice structure include a structure in which p-AlGaN and p-InGaN are stacked alternately, or a structure in which p-AlGaN and p-GaN are stacked alternately.

[0123] The thickness of the electron blocking layers 19A to 19C is preferably 2 to 50 nm, and more preferably 2 to 25 nm.

[0124] The electron blocking layers 19A to 19C are p-type doped with Mg. By making them p-type, holes can be efficiently injected into the active layer. Furthermore, a larger barrier against electrons can be formed, enhancing the electron blocking function. The electron blocking layers 19A to 19C may be non-doped, but for the reasons mentioned above, they are preferably p-type by doping with Mg. The Mg concentration of the electron blocking layers 19A to 19C is 1×10 19 ~100×10 19 cm -3 It is best to do so.

[0125] The p-type layers 20A to 20C are semiconductor layers provided on the electron blocking layers 19A to 19C, respectively, and are composed of a first layer and a second layer in that order from the electron blocking layer 19A to 19C side.

[0126] The first layer is preferably p-GaN or p-InGaN. The thickness of the first layer is preferably 10 to 500 nm, more preferably 10 to 200 nm, and further preferably 10 to 100 nm. The Mg concentration of the first layer is 1×10 19 ~100×10 19 cm -3 The second layer is preferably p-GaN or p-InGaN. The thickness of the second layer is preferably 2 to 50 nm, more preferably 4 to 20 nm, and even more preferably 6 to 10 nm. However, when applying the optical interference effect, the thicknesses of the first and second layers are not limited to this. The Mg concentration of the second layer is 1×10 20 ~100×10 20 cm -3 It is best to do so.

[0127] In the embodiment, the electron blocking layers 19A to 19C are formed separately from each other, and the p-type layers 20A to 20C are formed separately from each other, but they may be formed as a continuous film. In this case, layers made of the same material as the electron blocking layers 19A to 19C are also formed on the side surfaces of the first groove 30 and the second groove 31, and layers made of the same material as the p-type layers 20A to 20C are also formed.

[0128] The n-side electrode 21 is an electrode provided on the n-type layer 11 exposed at the bottom surface of the third groove 32. When the substrate 10 is made of a conductive material, the n-side electrode 21 may be provided on the back surface of the substrate 10 without providing the third groove 32. The material of the n-side electrode 21 is, for example, Ti / Al or V / Al.

[0129] The p-side electrodes 22A-22C are electrodes provided on the p-type layers 20A-20C, respectively. The p-side electrodes 22A-22C are preferably made of a material that has high reflectivity for light of the emission wavelength and low contact resistance with the p-type layers 20A-20C. Examples include Ag, Ni / Au, Co / Au, ITO / Ni / Al, Rh, and Ru. Of the red light emitted from the third active layer 16, the light traveling toward the p-type layer 20A is reflected by the p-side electrode 22A and travels toward the substrate 10. Similarly, of the green light emitted from the second active layer 14, the light traveling toward the p-type layer 20B is reflected by the p-side electrode 22B and travels toward the substrate 10. Meanwhile, of the blue light emitted from the first active layer 12, the light traveling toward the p-type layer 20C is reflected by the p-side electrode 22C and travels toward the substrate 10.

[0130] 3. Operation of light-emitting element Next, the operation of the light-emitting device in the embodiment will be described. In the light-emitting device in the embodiment, red light can be emitted from the third active layer 16 by applying a voltage between the p-side electrode 22A and the n-side electrode 21, green light can be emitted from the second active layer 14 by applying a voltage between the p-side electrode 22B and the n-side electrode 21, and blue light can be emitted from the first active layer 12 by applying a voltage between the p-side electrode 22C and the n-side electrode 21. These lights can be controlled individually, and two or more of blue, green, and red can also be emitted simultaneously.

[0131] In this way, in the light-emitting element of the embodiment, the light emission of blue, green, and red can be controlled by selecting the electrode to which the voltage is applied, and it can be used as one pixel in a micro LED display.

[0132] (Modification of the second embodiment) Although the light-emitting element in the second embodiment emits light of three colors, red, green, and blue, the present invention is not limited to this, and any light having two or more colors with different emission wavelengths may be used. For example, the light-emitting element may emit light of four colors, red, yellow, green, and blue.

[0133] (Other variations) Although the first and second embodiments are directed to light-emitting devices, the present invention can be applied to optical devices using semiconductors other than light-emitting devices, such as light-receiving elements and solar cells.

[0134] Furthermore, the light-emitting element and light-receiving element of the present invention can be used as a light source for generating an optical signal and a light-receiving element for receiving an optical signal in a wavelength division multiplexing optical communication device, thereby simplifying the optical communication device and reducing its cost. For example, a transmission signal can be generated and transmitted by wavelength-multiplexing the red light, green light, and blue light emitted by the light-emitting element of embodiment 2. Furthermore, if the received signal is received by the light-receiving element of the present invention, the red light, green light, and blue light can be separately received. [Explanation of symbols]

[0135] 10: Circuit board 11:n-type layer 12: 1st active layer 13: First middle class 14:Second active layer 15: Second middle class 16: 3rd active layer 17, 19A, 19B: Electron blocking layer 18, 20A, 20B: p-type layer 21:n side electrode 22A~22C:p side electrode

Claims

1. a well layer made of a group III nitride semiconductor containing In; a barrier layer formed on the well layer and made of a Group III nitride semiconductor having a band gap energy larger than that of the well layer; a composition gradient layer provided between the well layer and the barrier layer, in contact with the well layer and the barrier layer, the composition gradient layer being a Group III nitride semiconductor containing In, the In composition in the thickness direction being set to gradually decrease toward the barrier layer; an In composition at a surface of the compositionally graded layer facing the well layer is smaller than an In composition in the well layer;

2. an n-type layer that is an n-type Group III nitride semiconductor; a first active layer provided on the n-type layer, the first active layer being a Group III nitride semiconductor and emitting light at a predetermined wavelength; an intermediate layer formed on the first active layer and made of a Group III nitride semiconductor; a second active layer provided on the intermediate layer, the second active layer being made of a Group III nitride semiconductor and having an emission wavelength longer than that of the first active layer; a groove extending from the second active layer side to the intermediate layer; a first p-type layer formed on the second active layer and made of a p-type Group III nitride semiconductor; a second p-type layer formed on the intermediate layer exposed at the bottom surface of the groove and made of a p-type Group III nitride semiconductor; a first p-side electrode provided on the first p-type layer; a second p-side electrode provided on the second p-type layer; and The optical device according to claim 1 , wherein the second active layer includes the well layer, the barrier layer, and the compositionally graded layer.

3. 3. The optical device according to claim 1, wherein the In composition at the surface of the composition-graded layer facing the well layer is 30% to 70% of the In composition of the well layer.

4. 3. The optical device according to claim 1, wherein the In composition of the surface of the composition gradient layer on the barrier layer side is 90% or more and 110% or less of the In composition of the barrier layer.

5. 3. The optical device according to claim 1, wherein the total thickness of said well layer and said compositionally graded layer is 3 nm or more and 10 nm or less.

6. 3. The optical device according to claim 1, wherein the In composition of said well layer is 25% or more and 50% or less.

7. 3. The optical device according to claim 1, wherein an average In composition of the combined layer of said well layer and said compositionally graded layer is set so that the emission wavelength of said well layer is 500 to 800 nm.

8. The barrier layer is a first barrier layer provided on and in contact with the compositionally graded layer, the first barrier layer having a band gap energy larger than that of the well layer and made of a Group III nitride semiconductor having an In composition of 20% or less; a second barrier layer formed on and in contact with the first barrier layer and made of a Group III nitride semiconductor containing Al; 3. The optical device according to claim 1, further comprising: a third barrier layer formed on and in contact with the second barrier layer, the third barrier layer being a group III nitride semiconductor having a bandgap energy greater than that of the well layer.

9. a first superlattice layer having a superlattice structure in which group III nitride semiconductor thin films having different compositions are alternately stacked; a second superlattice layer provided on the first superlattice layer and below the well layer, the second superlattice layer having a superlattice structure in which Group III nitride semiconductor thin films having different compositions are alternately stacked; 3. The optical device according to claim 1, wherein an average In composition of said second superlattice layer is higher than an average In composition of said first superlattice layer.

10. a strain relaxation layer provided under the well layer, having a quantum well structure, and a thickness of a second well layer adjusted so as not to emit light; 3. The optical device according to claim 1, wherein a wavelength corresponding to the band edge energy of said second well layer of said strain relaxation layer is set to be shorter than an emission wavelength of said well layer.

11. The first active layer is a third well layer made of a group III nitride semiconductor containing In; 3. The optical device according to claim 2, further comprising: a fourth barrier layer provided on the third well layer and made of a Group III nitride semiconductor having a bandgap energy larger than that of the third well layer.

12. a well layer forming step of forming a well layer made of an In-containing Group III nitride semiconductor by MOCVD; a compositionally graded layer forming step of forming, by MOCVD, a compositionally graded layer on and in contact with the well layer, the compositional graded layer being a group III nitride semiconductor containing In, the In composition in the thickness direction gradually decreasing with increasing distance from the well layer, and the In composition at the surface on the well layer side being smaller than the In composition in the well layer; a barrier layer formation step of forming, by MOCVD, a barrier layer made of a Group III nitride semiconductor having a band gap energy larger than that of the well layer, in contact with the compositionally graded layer; The method for manufacturing an optical device, wherein the composition gradient layer forming step controls the In composition by adjusting the vapor phase ratio of an In source gas.

Images