Optical Semiconductor Device and Method of Manufacturing the Same

Abstract

The optical semiconductor device (100) of the present disclosure comprises a striped ridge structure (130) on a substrate (101) consisting of a lower cladding layer (102), an active layer (111), and a first upper cladding layer (103); a filling layer (120) that fills both sides of the ridge structure (130); a second upper cladding layer (104) formed on the upper surface of the ridge structure (130) and the upper surface of the filling layer (120); and a striped contact layer (141) formed on the upper surface of the second upper cladding layer (104), wherein in at least a portion of the region along the stripe direction, the contact layer (141) and the two The contact layer width Wc is the distance between one end and the other end of the upper cladding layer (103) that are in contact. is active Smaller than the sex layer width Wa Furthermore, the active layer width Wa is greater than the full width at half maximum Wf of the horizontal near-field pattern of light guiding the active layer (111). It is characterized by the following:

Description

Technical Field

[0001] The present disclosure relates to an optical semiconductor device and a method for manufacturing the optical semiconductor device.

Background Art

[0002] In a semiconductor laser, a semiconductor optical amplifier, etc., light is generated and amplified by injecting current into an active layer. A current that is not injected into the active layer and does not contribute to the generation and amplification of light, that is, a leakage current, not only reduces the amount of current injected into the active layer, but also raises the device temperature. When the device temperature rises, there occurs a problem that the optical output decreases due to carrier overflow. Therefore, improving the current injection efficiency into the active layer is an effective means for improving the optical output in a light-emitting device.

[0003] For example, in the semiconductor laser disclosed in Patent Document 1, it is said that by realizing a narrow and reproducible distance of the n-type current blocking layer from the active layer, the leakage current is suppressed and uniform laser characteristics are obtained.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, when the width of the active layer is wide with respect to the optical intensity distribution of the light guided in the active layer, the current injected into both ends of the active layer where the optical intensity is relatively low does not contribute to light emission and is wasted as a leakage current. In the current blocking layer structure as shown in Patent Document 1, although the leakage current flowing outside the active layer can be improved, there is room for improvement with respect to a structure in which the injection current required inside the active layer has a distribution.

[0006] The present disclosure has been made to solve the above-described problems, and an object thereof is to provide an optical semiconductor device and a method for manufacturing the optical semiconductor device that can achieve high photoelectric conversion efficiency even when the active layer width is relatively wide with respect to the light intensity distribution of light guided in the active layer.

Means for Solving the Problems

[0007] The optical semiconductor device according to the present disclosure includes: a stripe-shaped ridge structure including a lower cladding layer, an active layer that guides light, and a first upper cladding layer sequentially laminated on the substrate; and an embedding layer formed so as to embed both side surfaces of the ridge structure; a second upper cladding layer formed on the upper surface of the ridge structure and the upper surface of the embedding layer; a stripe-shaped contact layer formed on the upper surface of the second upper cladding layer, and in at least a part of a region along the stripe direction, a contact layer width Wc which is the distance between one end and the other end where the contact layer and the second upper cladding layer are in contact; by is smaller than an active layer width Wa which is the distance between one end and the other end where the active layer and the embedding layer are in contact in the active layer located directly below the contact layer; doing so the active layer width Wa; to is larger than a full width at half maximum Wf of a horizontal near-field pattern of light guiding the active layer. control so as to become It is characterized by this.

[0008] The method for manufacturing an optical semiconductor device according to the present disclosure includes: a step of sequentially laminating each semiconductor layer including a lower cladding layer, an active layer that guides light, and a first upper cladding layer on a substrate by epitaxial crystal growth; Forming a stripe-shaped ridge structure including at least each of the semiconductor layers; Epitaxially growing an embedding layer so as to embed both side surfaces of the ridge structure; Epitaxially growing a second upper cladding layer and a contact layer on the upper surface of the ridge structure and the upper surface of the embedding layer; In at least a part of the region along the stripe direction, the contact layer width Wc, which is the distance between one end and the other end where the contact layer contacts the first upper cladding layer, is smaller than the active layer width Wa, which is the distance between one end and the other end where the active layer contacts the embedding layer in the active layer located directly below the contact layer. such that the active layer width Wa is larger than the full width at half maximum Wf of the horizontal near-field pattern of the light guiding the active layer Processing, and comprising.

Advantages of the Invention

[0009] According to the optical semiconductor device according to the present disclosure, it is possible to provide an optical semiconductor device capable of realizing high photoelectric conversion efficiency even in a structure where the active layer width is relatively wide with respect to the intensity distribution of light guiding the active layer.

[0010] According to the method for manufacturing an optical semiconductor device according to the present disclosure, it is possible to easily manufacture an optical semiconductor device capable of realizing high photoelectric conversion efficiency even in a structure where the active layer width is relatively wide with respect to the intensity distribution of light guiding the active layer.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6A

Figure 6B

Figure 7A

Figure 7B

Figure 8A

Figure 8B

Figure 8C

Figure 9

Figure 10A

Figure 10B

Figure 10C

Figure 11A

Figure 11B

Figure 11C

Figure 12A

Figure 12B

Figure 12C

Figure 12D

Figure 12E

Figure 13A

Figure 13B

Figure 13C

Figure 14A

Figure 14B

Figure 14C

Figure 15A

Figure 15B

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21A

Figure 21B

Figure 22A

Figure 22B

Figure 23A

Figure 23B

Figure 24A

Figure 24B

Figure 24C

Figure 25A

Figure 25B

Figure 25C

Figure 26A

Figure 26B

Figure 26C

Figure 27A

Figure 27B

Figure 27C

Figure 28A

Figure 28B

Figure 28C

Figure 29A

Figure 29B

Figure 29C

Figure 30A

Figure 30B

Figure 30C

Figure 31A

Figure 31B

Embodiments for Carrying Out the Invention

[0012] Hereinafter, embodiments of the present disclosure will be described. Unless otherwise specified, the same or corresponding parts in the following drawings are denoted by the same reference numerals, and the description thereof will not be repeated.

[0013] Embodiment 1. <Configuration of the optical semiconductor device according to Embodiment 1> FIG. 1 is a top view of an optical semiconductor device 100 according to Embodiment 1. FIG. 2 is a cross-sectional view of the optical semiconductor device 100 according to Embodiment 1 taken along line I-I of FIG. 1. FIG. 3 is a cross-sectional view of the optical semiconductor device 100 according to Embodiment 1 taken along line II-II of FIG. 1.

[0014] As the optical semiconductor device 100 shown in FIGS. 1 to 3, for example, a semiconductor laser and a semiconductor optical amplifier (SOA) are each given as an example.

[0015] The optical semiconductor device 100 according to Embodiment 1 has, as shown in the top view of FIG. 1, a second electrode 162 whose width changes along the light propagation direction, and a surface protective film 151 that covers portions other than the second electrode 162. In FIG. 1, the dotted line represents the positions of both ends of the active layer 111 provided below the second electrode 162, and the dashed-dotted line represents the positions of both ends of the contact layer 141 also provided below the second electrode 162.

[0016] In FIG. 1, the width of the active layer 111 (hereinafter referred to as the active layer width) is denoted as Wa, and the width of the region where the contact layer 141 and the second electrode 162 are in contact, that is, the width of the contact layer 141 (hereinafter referred to as the contact layer width) is denoted as Wc.

[0017] As an example, the optical semiconductor device 100 according to Embodiment 1 is composed of four regions, i.e., a first region 184, a second region 185, a third region 186, and a fourth region 187, along the light propagation direction as shown in the top view of FIG. 1.

[0018] In the first region 184, the active layer width Wa and the contact layer width Wc each have a predetermined constant width. Further, there is a relationship that Wa < Wc, that is, the active layer width Wa is smaller than the contact layer width Wc. Note that the active layer width Wa and the contact layer width Wc are in a relationship such that the width of the active layer 111 located directly below the contact layer 141, which is the contact layer width Wc, corresponds to the active layer width Wa.

[0019] In the second region 185 connected to the first region 184, the contact layer width Wc changes along the light propagation direction, specifically, monotonically decreases. That is, in the second region 185, the contact layer 141 narrows in a tapered shape. On the other hand, the active layer width Wa is a constant width up to near the center of the second region 185, but changes in the direction further beyond the center, specifically, monotonically increases. That is, in the second region 185, the active layer 111 has a constant width up to the central portion, but spreads in a tapered shape in the direction beyond the central portion. In the second region 185, there is a relationship that Wa ≦ Wc, that is, the active layer width Wa is less than or equal to the contact layer width Wc.

[0020] In the third region 186 connected to the second region 185, the contact layer width Wc changes along the light propagation direction, specifically, monotonically increases. That is, in the third region 186, the contact layer 141 spreads in a tapered shape. On the other hand, the active layer width Wa also changes along the light propagation direction, specifically, monotonically increases. That is, in the third region 186, the contact layer 141 also spreads in a tapered shape. In the third region 186, there is a relationship that Wa ≧ Wc, that is, the active layer width Wa is greater than or equal to the contact layer width Wc.

[0021] In the fourth region 187, the active layer width Wa and the contact layer width Wc each have a predetermined constant width. Further, there is a relationship such that Wa > Wc, that is, the active layer width Wa is larger than the contact layer width Wc.

[0022] As shown in the cross-sectional views of FIGS. 2 and 3 respectively, the optical semiconductor device 100 according to Embodiment 1 includes at least a substrate 101, a lower cladding layer 102 sequentially formed on the substrate 101, an active layer 111 that guides light, a stripe-shaped ridge structure 130 composed of a first upper cladding layer 103, an embedding layer 120 formed so as to embed both side surfaces of the ridge structure 130, a second upper cladding layer 104 formed on the upper surface of the ridge structure 130 and the upper surface of the embedding layer 120, a stripe-shaped contact layer 141 formed on the upper surface of the second upper cladding layer 104, a surface protection film 151 composed of an insulating film covering the upper surface of the second upper cladding layer 104 other than the contact layer 141, a second electrode 162 formed on the upper surface of the contact layer 141 and on the surface protection film 151 and electrically connected to the contact layer 141 through the opening of the surface protection film 151, and a first electrode 161 formed on the back surface side of the substrate 101. Note that the term "stripe-shaped" means a state in which the target shape extends in the light guiding direction.

[0023] FIG. 2 is a cross-sectional view along line segment I-I of FIG. 1 in the optical semiconductor device 100 according to Embodiment 1, that is, a cross-sectional view of the third region 186, and the active layer width Wa is larger than the contact layer width Wc. FIG. 3 is a cross-sectional view along line segment II-II of FIG. 1 in the optical semiconductor device 100 according to Embodiment 1, that is, a cross-sectional view of the second region 185, and the active layer width Wa is smaller than the contact layer width Wc.

[0024] The embedded layer 120 does not have to be composed of only a single material and may be composed of a plurality of materials. Also, a contact layer 141 may be provided between the second upper cladding layer 104 and the second electrode 162, but it does not have to be provided. When the contact layer 141 is not provided, the contact layer width Wc is the width of the portion where the upper surface of the second upper cladding layer 104 and the second electrode 162 are in contact through the opening of the surface protection film 151 made of an insulating film.

[0025] A surface protection film 151 made of an insulating film is formed between the semiconductor layer composed of the second upper cladding layer 104 or the contact layer 141 and the second electrode 162 except for the portion where they are in contact.

[0026] The active layer 111 functions as an optical waveguide for guiding light. The lower cladding layer 102, the embedded layer 120, the first upper cladding layer 103, and the second upper cladding layer 104 function as cladding layers covering the active layer 111 as an optical waveguide.

[0027] The lower cladding layer 102 is connected on the substrate 101. The lower cladding layer 102 is connected to the lower surface of the active layer 111. The upper surface of the active layer 111 is connected to the first upper cladding layer 103.

[0028] The ridge structure 130 may or may not include the substrate 101. The ridge structure 130 includes the lower cladding layer 102, the active layer 111, and the first upper cladding layer 103. Both side surfaces of the ridge structure 130 are embedded by the embedded layer 120. The bottom surface of the embedded layer 120 is connected to the substrate 101 or the lower cladding layer 102. The upper surface of the embedded layer 120 and the upper surface of the first upper cladding layer 103 are connected to the second upper cladding layer 104.

[0029] When the optical semiconductor device 100 according to Embodiment 1 has the contact layer 141, the contact layer 141 is connected to the second upper cladding layer 104 through the opening of the surface protective film 151. The width of the opening of the surface protective film 151 is equal to the contact layer width Wc. The second upper cladding layer 104 is connected to the surface protective film 151 in a region other than the opening of the surface protective film 151. In this case, a part of the contact layer 141 and the surface protective film 151 may be connected.

[0030] When the optical semiconductor device 100 according to Embodiment 1 has the contact layer 141, the second electrode 162 is electrically connected to the contact layer 141. On the other hand, when the optical semiconductor device 100 does not have the contact layer 141, the second electrode 162 is electrically connected to the second upper cladding layer 104 through the opening of the surface protective film 151. In this case, a part of the second electrode 162 may be connected to the surface protective film 151.

[0031] As shown in FIG. 1, the ridge structure 130 is formed in a stripe shape with respect to the light propagation direction. The light propagation direction, that is, the light waveguide direction, is the direction from the first end face 191 to the second end face 192, or from the second end face 192 to the first end face 191. The ridge structure 130 includes the active layer 111. As shown in FIG. 1, the active layer width Wa may not be constant with respect to the light propagation direction. That is, the active layer width Wa may change along the light propagation direction. However, for example, the active layer width Wa may be constant along the light propagation direction as in the case where the optical semiconductor device 100 is composed of only the first region 184.

[0032] In an example of the optical semiconductor device 100 according to Embodiment 1 shown in FIG. 1, a first region 184, a second region 185, a third region 186, and a fourth region 187 are shown, but the optical semiconductor device 100 does not necessarily have all of these four regions. The optical semiconductor device 100 only needs to have at least one of the third region 186 and the fourth region 187 where the active layer width Wa and the contact layer width Wc satisfy Wa≧Wc. Further, if the optical semiconductor device 100 has at least one of the third region 186 and the fourth region 187, it may or may not have the first region 184 or the second region 185.

[0033] In the first region 184 and the second region 185 where the active layer width Wa and the contact layer width Wc satisfy Wa≦Wc, the relationship between the horizontal direction of light, that is, the full width at half maximum (Full Width at Half Maximum: FWHM, hereinafter referred to as the half-width FWHM Wf) of the near field pattern (Near Field Pattern: NFP) in the horizontal direction of light on the surface of the substrate 101 and the active layer width Wa is Wa≦Wf. Further, in the third region 186 and the fourth region 187 where the active layer width Wa and the contact layer width Wc satisfy Wa>Wc, the relationship between the active layer width Wa and the half-width FWHM Wf of the NFP in the horizontal direction of light is Wa>Wf.

[0034] In a semiconductor laser, light is amplified by stimulated emission that emits the same light as the incident light. In stimulated emission, light is emitted when electrons with high-level energy transition to a low-level energy level. Therefore, efficient stimulated emission can be caused by forming an inverted distribution in which many electrons exist in the high-level energy band. Since electrons as carriers are required to form an inverted distribution, current injection into the active layer 111 is necessary.

[0035] In the optical semiconductor device 100 according to Embodiment 1, as shown in FIGS. 1 and 3, in the first region 184 and the second region 185 where the full width at half maximum Wf of the NFP in the horizontal direction of light and the active layer width Wa satisfy the relationship Wf≧Wa, light with a light intensity of 1 / 2 or more is incident on the entire active layer 111 in the horizontal direction. Therefore, by efficiently injecting current into the entire active layer 111, an improvement in the photoelectric conversion efficiency can be obtained. Incidentally, in the semiconductor laser disclosed in Patent Document 1, the photoelectric conversion efficiency is improved by suppressing the leakage current flowing outside the active layer and thereby improving the current injection into the active layer. In this regard, the principle of improving the photoelectric conversion efficiency is different from that of the optical semiconductor device of the present disclosure.

[0036] In the optical semiconductor device 100 according to Embodiment 1, as shown in FIGS. 1 and 2, in the third region 186 and the fourth region 187 where the full width at half maximum Wf of the NFP in the horizontal direction of light and the active layer width Wa satisfy the relationship Wa>Wf, a region where the light incident on the active layer 111 has a light intensity of 1 / 2 or less is generated. As described above, since light is generated by stimulated emission, a large number of electrons in the high-energy level are used in a region where the light intensity of the incident light is high, but not so many electrons in the high-energy level are required in a region where the light intensity of the incident light is low. Therefore, when current is uniformly injected into the active layer 111, the injected current is not converted into light at both ends of the active layer 111 where the light intensity is low, but is consumed as heat. For the above reasons, in the conventional optical semiconductor device, there is room for improvement when the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy the relationship Wa>Wf.

[0037] Here, the feature of the optical semiconductor device 100 according to Embodiment 1 is that, as shown in FIGS. 1 to 3, when the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa>Wf, the active layer width Wa and the contact layer width Wc satisfy Wa>Wc.

[0038] As shown in Patent Document 1 and FIG. 3 of the present disclosure, when the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa ≦ Wf, in order to minimize heat generation due to the resistance from the second electrode 162 to the active layer 111, it is desirable to increase the contact area between the semiconductor layer and the second electrode 162, that is, the contact layer width Wc.

[0039] However, as shown in FIGS. 1 and 2, when the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa > > Wf, it is effective to suppress current injection into both ends of the active layer 111 where only light with an intensity of 1 / 2 or less is incident. That is, by making the contact layer width Wc smaller than the active layer width Wa, that is, setting Wa > Wc, it becomes possible to efficiently cause stimulated emission.

[0040] The optimal contact layer width Wc for improving the photoelectric conversion efficiency changes depending on the magnitude relationship between the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light. Therefore, as shown in FIG. 1, it is preferable to change the contact layer width Wc according to the active layer width Wa. Also, as shown in the second region 185 and the third region 186 of FIG. 1, the difference value |Wa - Wc| between the active layer width Wa and the contact layer width Wc does not have to be constant.

[0041] In a region where the active layer width Wa is sufficiently smaller than the full width at half maximum Wf of the NFP in the horizontal direction of light, the contact layer width Wc may be increased to reduce the element resistance. Also, in a region where the active layer width Wa is sufficiently larger than the full width at half maximum Wf of the NFP in the horizontal direction of light, in order to concentrate and inject current into the central part of the active layer 111 with a large light intensity, the contact layer width Wc may be decreased. Also, although not shown in FIG. 1, in the transition region from the second region 185 to the third region 186, due to manufacturing tolerances, a region where Wa ≦ Wf and Wa > Wc may locally occur.

[0042] <Manufacturing Method of Optical Semiconductor Device According to Embodiment 1> FIG. 4 is a flowchart for explaining the manufacturing method of the optical semiconductor device 100 according to Embodiment 1.

[0043] As shown in FIG. 4, in the manufacturing method of the optical semiconductor device 100 according to the first embodiment, first, a step S1a of preparing a substrate 101 is performed. The substrate 101 may be, for example, an InP (indium phosphide) substrate 101. The conductivity type of the substrate 101 may be p-type or n-type. Further, the substrate 101 may be a semi-insulating substrate.

[0044] Next, a step S2a of forming the lower cladding layer 102 is performed. In step S2a, the lower cladding layer 102 is formed on the substrate 101. The lower cladding layer 102 is formed by a crystal growth method such as Metal Organic Chemical Vapor Deposition (MOCVD).

[0045] The material constituting the lower cladding layer 102 is a material having a refractive index lower than that of the active layer 111 as described later. The material constituting the lower cladding layer 102 may be a material lattice-matched with the active layer 111, for example, InP. The conductivity type of InP may be p-type. Zinc (Zn) may be used as the p-type dopant. The conductivity type of InP may be n-type. Sulfur (S) may be used as the n-type dopant. Further, InP may not be doped.

[0046] Next, a step S3a of forming the active layer 111 is performed. In step S3b, the active layer 111 is formed on the lower cladding layer 102. Similar to the method of forming the lower cladding layer 102, the active layer 111 may be formed using the MOCVD method. The material constituting the active layer 111 may be, for example, indium gallium arsenide phosphide (InGaAsP), aluminum gallium indium arsenide (AlGaInAs), indium gallium arsenide (InGaAs), or aluminum indium arsenide (AlInAs). The lower cladding layer 102 and the active layer 111 may be formed simultaneously by the MOCVD method.

[0047] Next, a step S4a of forming the first upper cladding layer 103 is performed. In step S4b, the first upper cladding layer 103 is formed on the active layer 111. Similar to the method of forming the active layer 111, the first upper cladding layer 103 may be formed using the MOCVD method. The material constituting the first upper cladding layer 103 may be, for example, InP. The conductivity type of InP may be p-type. As the p-type dopant, Zn may be used. The conductivity type of InP may be n-type. As the n-type dopant, S may be used. Further, InP may not be doped. The lower cladding layer 102, the active layer 111, and the first upper cladding layer may be formed simultaneously by the MOCVD method. A cross-sectional view of the wafer after each step from step S1a to step S4a is shown in FIG. 5.

[0048] Next, a step S5a of etching to form the ridge structure 130 is performed. In step S5a, as shown in FIGS. 6A and 6B, a first processing mask 171 is formed on the first upper cladding layer 103. As the material of the first processing mask 171, for example, an insulating film of either a SiO2 film or a SiN film is used. As a method of forming the SiO2 film and the SiN film, for example, Chemical Vapor Deposition (CVD) can be mentioned.

[0049] The first processing mask 171 is patterned into a desired shape using photolithography technology and etching technology. As shown in FIG. 6A, in plan view, the shape of the first processing mask 171 does not have to be of equal width in the light propagation direction. Also, in plan view, the shape of the first processing mask 171 may be of equal width in the light propagation direction.

[0050] The shape of the first processing mask 171 may have, for example, a first active layer region 181 with a constant width, a third active layer region 183 that is wider than the first active layer region 181 and has a constant width, and a second active layer region 182 whose width changes in a tapered shape so as to connect the first active layer region 181 and the third active layer region 183. Further, the shape of the first processing mask 171 may be composed of, for example, only the second active layer region 182 that does not have a region with a constant width.

[0051] The shape of the first processing mask 171 is such that, for example, the second active layer region 182 has a shape in which the width increases from the first active layer region 181 side toward the third active layer region 183 side, but it may also have a shape in which the width increases from the third active layer region 183 side toward the first active layer region 181 side. The shape of the first processing mask 171 may be composed of straight lines, or may be composed of curves. Further, the shape of the first processing mask 171 may be composed of a combination of straight lines and curves.

[0052] At least a part of the active layer width Wa formed in step S5a is wider than the full width at half maximum Wf of the NFP in the horizontal direction of light. In this case, in all regions along the light propagation direction, the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light may have a relationship of Wa > Wf. In Embodiment 1, in the first region 184 and the second region 185, it is treated as Wa ≤ Wf, and in the third region 186 and the fourth region 187, it is treated as Wa > Wf.

[0053] Next, using the above-described first processing mask 171 as a mask, the first upper cladding layer 103, the active layer 111, and the lower cladding layer 102 are dry-etched. Specifically, as shown in FIG. 7A, in the region where the first processing mask 171 is disposed in a plan view, the first upper cladding layer 103, the active layer 111, and the lower cladding layer 102 are not etched. On the other hand, in the region where the first processing mask 171 is not disposed in a plan view, the first upper cladding layer 103, the active layer 111, and the lower cladding layer 102 are etched.

[0054] As a result of etching, as shown in FIG. 7B, a part of the lower cladding layer 102 is etched such that a part of the first upper cladding layer 103, the active layer 111, and the lower cladding layer 102 are exposed. At this time, a part of the substrate 101 may be etched through the lower cladding layer 102. As the etching method, wet etching may be used in addition to the above-described dry etching.

[0055] As described above, a ridge structure 130 having a desired shape is formed from the first upper cladding layer 103, the active layer 111, and the lower cladding layer 102. The ridge structure 130 has a stripe shape in the light propagation direction. However, the width of the ridge structure 130 varies along the light propagation direction.

[0056] Next, a step S6a of forming the embedded layer 120 is performed. In step S6a, as shown in FIG. 8A, the first processing mask 171 is used as a selective growth mask to form the embedded layer 120 so as to embed the side surfaces of the ridge structure 130. Note that the bottom surface of the embedded layer 120 is in contact with the portion that remained without being etched by the lower cladding layer 102 when the ridge structure 130 was formed. Note that the bottom surface of the embedded layer 120 may be in contact with the substrate 101.

[0057] The embedded layer 120 may be formed using the MOCVD method. The material for forming the embedded layer 120 may be, for example, InP. The conductivity type of InP may be p-type or n-type by adding a dopant. As the dopant for p-type conversion, Zn may be used. As the dopant for n-type conversion, S may be used. Alternatively, dopants other than these may be used. Further, semi-insulating InP may be used. As the dopant for insulating InP, Fe may be used.

[0058] The embedded layer 120 may be formed by combining a plurality of layers of InP with different dopants. For example, as shown in FIG. 8B, the embedded layer 120 may be configured by combining a first embedded layer 121 and a second embedded layer 122. The material constituting the first embedded layer 121 may be semi-insulating InP. The material constituting the second embedded layer 122 may be n-type InP.

[0059] For example, as shown in FIG. 8C, the embedded layer 120 may be configured by combining three layers: a first embedded layer 121, a second embedded layer 122, and a third embedded layer 123. The material constituting the first embedded layer 121 may be p-type InP. The material constituting the second embedded layer 122 may be semi-insulating InP. The material constituting the third embedded layer 123 may be n-type InP.

[0060] The embedded layer 120 may have a configuration other than those shown in FIGS. 8B and 8C. Even when the embedded layer 120 is composed of a plurality of layers, it may be formed simultaneously by the MOCVD method. It is desirable that the embedded layer 120 has a current constriction function.

[0061] Next, a step S7a of forming the second upper cladding layer is performed. In this step S7a, first, the first processing mask 171 is removed. Hydrofluoric acid may be used to remove the first processing mask 171.

[0062] Next, the second upper cladding layer 104 is formed on the upper surface of the first upper cladding layer 103, that is, the upper surface of the ridge structure 130, and the upper surface of the embedded layer 120. At this time, the second upper cladding layer 104 may be formed using the MOCVD method. The material constituting the second upper cladding layer 104 may be, for example, InP. The conductivity type of InP may be p-type. Zn may be used as the p-type dopant. The conductivity type of InP may be n-type. S may be used as the dopant. Furthermore, InP may not be doped.

[0063] Next, a step S8a of forming the contact layer 141 is performed. In step S8a, the contact layer 141 is formed on the second upper cladding layer 104. Similar to the method of forming the second upper cladding layer 104, the contact layer 141 may be formed using the MOCVD method. The material constituting the contact layer 141 may be, for example, InGaAs, InGaAsP, AlInAs, AlGaInAs, or InP. The conductivity type of these materials may be p-type. As a dopant, Zn may be used. The conductivity type of these materials may be n-type. As a dopant, S may be used. Further, these materials may not be doped. The second upper cladding layer 104 and the contact layer 141 may be formed simultaneously by the MOCVD method. A cross-sectional view of the wafer after each step from step S7a to step S8a is shown in FIG. 9.

[0064] Next, a step S9a of etching the contact layer 141 is performed. In step S9a, as shown in FIGS. 10A, 10B, and 10C, a second processing mask 172 is formed on the contact layer 141. As the material of the second processing mask 172, for example, an insulating film such as a SiO2 film or a SiN film is used. As a method for forming the SiO2 film and the SiN film, for example, CVD can be mentioned.

[0065] The second processing mask 172 is patterned into a desired shape using photolithography technology and etching technology. As shown in FIG. 10A, in a plan view, the shape of the second processing mask 172 may not be of equal width in the light propagation direction. In a plan view, the shape of the second processing mask 172 may be of equal width in the light propagation direction.

[0066] For example, it may have a first region 184 of equal width, a fourth region 187 that is narrower than the first region 184 and of equal width, a second region 185 that changes to be tapered and narrower in width with respect to the first region 184, and a third region 186 that changes to be tapered and wider in width with respect to the second region 185 and is connected to the fourth region 187.

[0067] For example, it may be composed of only at least one of the second region 185 and the third region 186 that do not have regions of equal width. For example, an equal-width first region 184 and a fourth region 187 that are equal in width but narrower than the first region 184 may be combined. For example, an equal-width region may be provided between the second region 185 and the third region 186 whose widths change in a tapered shape.

[0068] Also, in FIG. 10A, the shape of the second processing mask 172 is defined by four regions, but four or more regions may be combined. The shape of the second processing mask 172 may be composed of straight lines, or may be composed of curves. The shape of the second processing mask 172 may be composed by combining straight lines and curves. Although not shown, the second processing mask 172 may be interrupted midway between the first region 184 and the fourth region 187.

[0069] At least a part of the contact layer width Wc of the second processing mask formed in step S9a is narrower than the active layer width Wa. At this time, the active layer width Wa and the half-value full width Wf of the NFP in the horizontal direction of light have a relationship of Wa>Wf. In order to obtain better current injection efficiency, in the first region 184 and the second region 185 where Wa≦Wf with respect to the active layer width Wa, the active layer width Wa and the contact layer width Wc satisfy the relationship of Wa≦Wc, and further, in the third region 186 and the fourth region 187 where Wa>Wf, it is desirable that the relationship be Wa>Wc.

[0070] However, for example, in the transition region from the second region 185 to the third region 186, due to manufacturing tolerances such as mask alignment, a region where locally Wa≦Wf and Wa>Wc may occur. Also, for the purpose of isolating between the first region 184 and the fourth region 187, the second processing mask 172 may be interrupted midway, and at that time, a region where Wa>Wc may occur.

[0071] Next, using the above-described second processing mask 172 as a mask, the contact layer 241 is dry-etched. Specifically, as shown in FIG. 11A, the contact layer 241 in the region where the second processing mask 172 is disposed in a plan view is not etched. On the other hand, the contact layer 241 in the region where the second processing mask 172 is not disposed in a plan view is etched. As a result, as shown in FIGS. 11B and 11C, a part of the contact layer 241 is etched so that a part of the second upper cladding layer 104 is exposed. At this time, a part of the second upper cladding layer 104 may be etched through the contact layer 141.

[0072] As the etching method, wet etching may be used in addition to the above-described dry etching. As the etching solution, tartaric acid, hydrochloric acid, a mixed solution of hydrochloric acid and phosphoric acid, or the like may be used. After etching the contact layer 141, the second processing mask 172 is removed by wet etching.

[0073] Next, a step S10a of forming a surface protective film 151 made of an insulating film is performed. In step S10a, first, the surface protective film 151 is formed on the second upper cladding layer 104 and the contact layer 141. The surface protective film 151 is, for example, either a SiO2 film or a SiN film. The film formation method of the SiO2 film and the SiN film is, for example, CVD.

[0074] Next, as shown in FIGS. 12A, 12B, and 12C, a third processing mask 173 is formed on the surface protective film 151. As the material of the third processing mask 173, for example, photoresist can be mentioned. Other than photoresist, any material that can be selectively etched with respect to the surface protective film 151 may be used. For example, when the surface protective film 151 is made of a SiO2 film, the third processing mask 173 is made of a SiN film.

[0075] The third processing mask 173 is patterned into a desired shape using photolithography technology and etching technology. As shown in FIG. 12A, in a plan view, the shape of the third processing mask 173 does not have to be of equal width along the light propagation direction. In a plan view, the shape of the third processing mask 173 may be of equal width along the light propagation direction.

[0076] For example, in the region shown in FIG. 12A, it may have a first region 184 of equal width, a fourth region 187 that is wider than the first region 184 and of equal width, a second region 185 that changes to be wider in an asymmetric taper shape with respect to the first region 184, and a third region 186 that changes to be narrower in an asymmetric taper shape with respect to the second region 185 and is connected to the fourth region 187.

[0077] For example, the third processing mask 173 may be composed of only at least one of the second region 185 and the third region 186 that do not have a region of equal width. For example, the third processing mask 173 may be a combination of a first region 184 of equal width and a fourth region 187 of equal width and narrower than the first region 184. For example, a region of equal width may be provided between the second region 185 and the third region 186, each of which changes in width in an asymmetric taper shape.

[0078] Also, in FIG. 12A, the shape of the third processing mask 173 is defined by four regions, but more than four regions may be combined. The shape of the third processing mask 173 may be composed of straight lines, or may be composed of curves. Furthermore, the shape of the third processing mask 173 may be composed of a combination of straight lines and curves. As shown in FIGS. 12D and 12E, for the purpose of isolating between the first region 184 and the fourth region 187, the third processing mask 173 may connect the vertical patterns in FIGS. 12D and 12E in the middle between the first region 184 and the fourth region 187.

[0079] Next, a step S11a of etching the surface protective film 151 is performed. In step S11a, the surface protective film 151 is dry-etched using the above-described third processing mask 173 as a mask. Specifically, the surface protective film 151 in the region where the third processing mask 173 is disposed in a plan view is not etched. On the other hand, the surface protective film 151 in the region where the third processing mask 173 is not disposed in a plan view is etched. As a result, a part of the surface protective film 151 is etched so that a part of the contact layer 141 is exposed.

[0080] As an etching method, wet etching may be used in addition to the above-described dry etching. As an etching solution, a mixed solution of tartaric acid, hydrochloric acid, hydrochloric acid, and phosphoric acid, etc. may be used. After etching the surface protective film 151, as shown in FIGS. 13A, 13B, and 13C, the third processing mask 173 is removed by wet etching.

[0081] Next, a step S12a of forming the first electrode 161 and the second electrode 162 is performed. In step S12a, as shown in FIGS. 14A, 14B, and 14C, the first electrode 161 is formed on the back surface side of the substrate 101, and the second electrode 162 is formed on the upper surface of the contact layer 141 on the front surface side of the substrate 101 and the surface of a part of the surface protective film 151. The first electrode 161 and the second electrode 162 are formed by an electrolytic plating method, a vacuum evaporation method, a sputtering method, etc. Examples of the materials constituting the first electrode 161 and the second electrode 162 include Au, Ti, and Pt.

[0082] <Operation of the optical semiconductor device according to Embodiment 1> Next, the operation of the optical semiconductor device 100 according to Embodiment 1 will be described. Examples of the optical semiconductor device 100 according to Embodiment 1 include a semiconductor laser and a semiconductor optical amplifier. As an example of the optical semiconductor device 100, a semiconductor laser that is a light source connected from the first region 184 side and a semiconductor optical amplifier will be described below, and its operation will be described.

[0083] Light propagates from a semiconductor laser to a semiconductor optical amplifier. The light is incident on the active layer 111 of the semiconductor optical amplifier. The incident light propagates through the active layer 111 while being amplified and travels from the first region 184 toward the fourth region 187. Since the active layer width Wa expands toward the fourth region 187, the light propagating through the active layer 111 also spreads horizontally. At the interface between the second region 185 and the third region 186, the full width at half maximum Wf of the NFP of the light in the horizontal direction becomes equal to the active layer width Wa.

[0084] After the third region 186, the full width at half maximum Wf of the NFP of the light in the horizontal direction becomes smaller than the active layer width Wa. Along with this, as the active layer width Wa expands, (active layer width Wa - contact layer width Wc) increases. As a result, current is injected into the region with high light intensity, and high photoelectric conversion efficiency can be obtained. s

[0085] <Operation of the optical semiconductor device according to Embodiment 1> The optical semiconductor device 100 according to the present disclosure includes an active layer 111 having an active layer width Wa and a contact layer having a contact layer width Wc. In the active layer 111 of the optical semiconductor device 100, in at least a part of the region along the light propagation direction, the active layer width Wa is wider than the full width at half maximum Wf of the NFP of the light in the horizontal direction, that is, Wa > Wf. In at least a part of the region where Wa > Wf, the active layer width Wa and the contact layer width Wc satisfy Wa > Wc.

[0086] With the above configuration, in the region where Wa > Wf, since the contact layer width Wc is smaller than the active layer width Wa, current is not injected into the region where the light intensity in the active layer 111 becomes 1 / 2 or less. Therefore, it is possible to selectively inject current into the portion with high light intensity.

[0087] In the optical semiconductor device 100 according to Embodiment 1, in the first region 184 and the second region 185, the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa ≤ Wf, and the active layer width Wa and the contact layer width Wc satisfy Wa ≤ Wc. In the third region 186 and the fourth region 187, the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa > Wf, and the active layer width Wa and the contact layer width Wc satisfy Wa > Wc. With the above configuration, it becomes possible to always inject current only into the portion with high light intensity within the active layer 111. As a result, the photoelectric conversion efficiency of the optical semiconductor device is improved.

[0088] FIGS. 15A and 15B are diagrams showing simulation results of current-light output characteristics when the contact layer width Wc is changed in the case where the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa = 7 μm > Wf. As shown in FIG. 15A, it can be seen that the maximum light output value increases when the contact layer width Wc is smaller than Wa = Wc = 7 μm. Also, as shown in FIG. 15B, in the region represented by the following formula (1), the improvement effect of the light output is particularly large.

[0089]

Equation

[0090] <Operation and Effect of the Optical Semiconductor Device According to Embodiment 1> As described above, according to the optical semiconductor device according to Embodiment 1, since a region where the active layer width Wa is larger than the contact layer width Wc is provided in a part of the region along the light propagation direction, as a result, it becomes possible to always inject current only into the portion with high light intensity, and the photoelectric conversion efficiency of the optical semiconductor device is improved.

[0091] Furthermore, according to the optical semiconductor device according to Embodiment 1, in a partial region along the light propagation direction, a region is provided in which the active layer width Wa is larger than the contact layer width Wc and the active layer width Wa is larger than the full width at half maximum Wf of the NFP in the horizontal direction of light. As a result, it becomes possible to always inject current only into the portion with high light intensity, and thus the photoelectric conversion efficiency of the optical semiconductor device is further improved.

[0092] Embodiment 2. <Configuration of the optical semiconductor device according to Embodiment 2> FIG. 16 is a top view of an optical semiconductor device 200 according to Embodiment 2. FIG. 17 is a cross-sectional view of the optical semiconductor device 200 according to Embodiment 2 taken along line I-I in FIG. 16. FIG. 18 is a cross-sectional view of the optical semiconductor device 200 according to Embodiment 2 taken along line II-II in FIG. 16.

[0093] In FIGS. 16, 17, and 18, the active layer width of the active layer 211 is denoted as Wa, and the width of the region where the bottom surface of the opening of the current blocking layer 222 is connected to the second upper cladding layer 204 (hereinafter referred to as the opening width) is denoted as Wb. The active layer width Wa and the opening width Wb are in a relationship such that the width of the active layer 111 located directly below the opening of the current blocking layer 222, which is the opening width Wb, corresponds to the active layer width Wa.

[0094] Also, a region where the active layer width Wa and the opening width Wb satisfy Wa ≦ Wb is defined as the first region 288, a region where the active layer width Wa and the opening width Wb satisfy Wa > Wb and both the active layer width Wa and the opening width Wb change is defined as the second region 289, and a region where the active layer width Wa and the opening width Wb satisfy Wa > Wb and both the active layer width Wa and the opening width Wb are each a preset constant width is defined as the third region 290.

[0095] Examples of the optical semiconductor device 200 shown in FIGS. 16 to 18 include a semiconductor laser and a semiconductor optical amplifier.

[0096] As shown in the top view of Fig. 16, the optical semiconductor device 200 according to Embodiment 2 has a second electrode 262 covering the surface. In Fig. 16, the dotted lines represent the positions of both ends of the active layer 211 provided below the second electrode 262, and the dashed-dotted lines represent the positions of both ends of the opening of the current blocking layer 222 also provided below the second electrode 262.

[0097] As shown respectively in the cross-sectional views of Figs. 17 and 18, the optical semiconductor device 200 according to Embodiment 2 includes a substrate 201, a lower cladding layer 202 sequentially formed on the substrate 201, an active layer 211 that guides light, a stripe-shaped ridge structure 230 composed of a first upper cladding layer 203, an embedded layer 221 formed so as to embed both side surfaces of the ridge structure 230, a current blocking layer 222 formed on the upper surface of the ridge structure 230 and the upper surface of the embedded layer 221 and having an opening with an opening width Wb at the bottom surface made of the first upper cladding layer 203, a second upper cladding layer 204 formed on the upper surface of the first upper cladding layer 203 exposed at the bottom surface of the opening of the current blocking layer 222 and on the upper surface of the current blocking layer 222, a contact layer 241 formed on the upper surface of the second upper cladding layer 204, a second electrode 262 formed on the upper surface of the contact layer 241 and electrically connected to the contact layer 241, and a first electrode 261 formed on the back surface side of the substrate 201.

[0098] The embedded layer 221 does not necessarily have to be composed of only a single material and may be composed of a plurality of materials. Also, a contact layer 241 may be provided between the second upper cladding layer 204 and the second electrode 262, but it may not be provided.

[0099] Although not shown, a surface protection film made of an insulating film may be formed between the semiconductor layer composed of the second upper cladding layer 204 or the contact layer 241 and the second electrode 262 in a region other than the region where they are electrically connected.

[0100] The active layer 211 functions as an optical waveguide that guides light. The lower cladding layer 202, the embedding layer 221, the first upper cladding layer 203, and the second upper cladding layer 204 function as cladding layers that cover the active layer 211 as an optical waveguide.

[0101] The lower cladding layer 202 is connected on the substrate 201. The lower cladding layer 202 is connected to the lower surface of the active layer 211. The upper surface of the active layer 211 is connected to the first upper cladding layer 203.

[0102] The ridge structure 230 may or may not include the substrate 201. The ridge structure 230 includes the lower cladding layer 202, the active layer 211, and the first upper cladding layer 203. Both side surfaces of the ridge structure 230 are embedded by the embedding layer 221. The bottom surface of the embedding layer 221 is connected to the substrate 201 or the lower cladding layer 202. The upper surface of the embedding layer 221 and the upper surface of the first upper cladding layer 203 are connected to the second upper cladding layer 204.

[0103] When the optical semiconductor device 200 according to Embodiment 2 has the contact layer 241, the contact layer 241 is connected to the second upper cladding layer 204. Although not shown, the second upper cladding layer 204 may be connected to the surface protection film in a region where the contact layer 241 is not formed. At this time, a part of the contact layer 241 and the surface protection film may be connected.

[0104] When the optical semiconductor device 200 according to Embodiment 2 does not have the contact layer 241, the second upper cladding layer 204 and the second electrode 262 are electrically connected. A part of the second electrode 262 may be connected to the surface protection film.

[0105] As shown in FIG. 16, the ridge structure 230 is formed in a stripe shape along the light propagation direction. The light propagation direction is the direction from the first end face 291 to the second end face 292, or from the second end face 292 to the first end face 291. The ridge structure 230 includes the active layer 211. As shown in FIG. 16, the active layer width Wa may not be constant along the light propagation direction. That is, the active layer width Wa may change along the light propagation direction. Also, the active layer width Wa may be constant along the light propagation direction.

[0106] In an example of the optical semiconductor device 200 according to Embodiment 2 shown in FIG. 16, three regions, a first region 288, a second region 289, and a third region 290, are shown, but the optical semiconductor device 200 does not necessarily have all of these three regions. It is sufficient to have at least one of the second region 289 and the third region 290 where the active layer width Wa and the opening width Wb satisfy Wa > Wb. If it has at least one of the second region 289 and the third region 290, it may or may not have the first region 288.

[0107] Also, the first region 288 is composed of a region where the active layer width Wa and the opening width Wb are both of equal width, and a region where the active layer width Wa and the opening width Wb increase in the direction from the first end face 291 to the second end face 292, but it does not necessarily have such a configuration. That is, any region where the relationship Wa ≦ Wb between the active layer width Wa and the opening width Wb is maintained is acceptable.

[0108] The second region 289 is composed of a region where the active layer width Wa and the opening width Wb monotonically increase along the direction from the first end face 291 to the second end face 292, but it does not necessarily have such a configuration. Similarly, the third region 290 is composed of a region where the active layer width Wa and the opening width Wb are of equal width, but it does not necessarily have such a configuration. The second region 289 and the third region 290 may be any regions where the relationship Wa > Wb regarding the active layer width Wa and the opening width Wb is maintained.

[0109] In the first region 288 where the active layer width Wa and the opening width Wb satisfy Wa ≦ Wb, the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa ≦ Wf. Further, in the second region 289 and the third region 290 where the active layer width Wa and the opening width Wb satisfy Wa > Wb, the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa > Wf.

[0110] The problems in the conventional optical semiconductor device are as described in Embodiment 1. As shown in FIGS. 16 and 18, in the first region 288 where the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa ≦ Wf, light with a light intensity of 1 / 2 or more in the horizontal direction is incident on the entire active layer 211.

[0111] On the other hand, as shown in FIGS. 16 and 17, in the second region 289 and the third region 290 where the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa > Wf, a region where light with a light intensity of 1 / 2 or less is incident is generated in the active layer 211. In such an optical semiconductor device having such a region, there is room for improving the photoelectric conversion efficiency.

[0112] The feature of the optical semiconductor device 200 according to Embodiment 2 is that, as shown in FIGS. 16 to 18, when the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa > Wf, the active layer width Wa and the opening width Wb satisfy Wa > Wb. As shown in Patent Document 1 or FIG. 18, when the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa ≦ Wf, it is desirable that the opening width Wb be equal to the active layer width Wa in order to minimize the heat generation due to the resistance from the second electrode 262 to the active layer 211.

[0113] However, as shown in FIGS. 16 and 17, when the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa > Wf, it is effective to suppress the current injection into both ends of the active layer 211 where only light with a light intensity of 1 / 2 or less is incident. That is, by making the opening width Wb smaller than the active layer width Wa, it becomes possible to efficiently cause stimulated emission.

[0114] The optimum aperture width Wb for improving the photoelectric conversion efficiency changes depending on the magnitude relationship between the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light. Therefore, as shown in FIG. 16, it is desirable to change the aperture width Wb according to the active layer width Wa.

[0115] Also, as shown in the first region 288, the second region 289, and the third region 290 of FIG. 16, the difference value |Wa - Wb| between the active layer width Wa and the aperture width Wb does not have to be constant.

[0116] In a region where the active layer width Wa is sufficiently smaller than the full width at half maximum Wf of the NFP in the horizontal direction of light, the aperture width Wb may be widened to reduce the element resistance. Also, in a region where the active layer width Wa is sufficiently larger than the full width at half maximum Wf of the NFP in the horizontal direction of light, the aperture width Wb may be narrowed to concentrate and inject current into the central portion of the active layer 211 where the light intensity is high.

[0117] Also, although not shown in FIG. 16, in the transition region from the first region 288 to the second region 289, a region where Wa ≤ Wf and Wa > Wb may locally occur due to manufacturing tolerances.

[0118] <Method for manufacturing an optical semiconductor device according to Embodiment 2> FIG. 19 is a flowchart for explaining a method for manufacturing an optical semiconductor device 200 according to Embodiment 2. From step S1b to step S5b shown in the flowchart of FIG. 19, they are manufactured by the same manufacturing method as in Embodiment 1, as shown in FIGS. 20, 21A, 21B, 22A, and 22B, respectively. Note that the first processing mask 271 shown in FIG. 21A is composed of three regions: the first active layer region 281, the second active layer region 282, and the third active layer region 283.

[0119] After step S5b, step S6b of forming the embedded layer 221 is performed. In step S6b, using the first processing mask 271 shown in FIG. 23A as a selective growth mask, the embedded layer 221 is formed so as to embed the ridge structure 230. Note that the bottom surface of the embedded layer 221 is in contact with the portion that remained without being etched by the lower cladding layer 202 when forming the ridge structure 230. Note that the bottom surface of the embedded layer 221 may be in contact with the substrate 201.

[0120] The embedded layer 221 may be formed using the MOCVD method. The material for forming the embedded layer 221 may be, for example, InP. The conductivity type of InP may be p-type or n-type by adding a dopant. As a dopant for p-type conversion, Zn may be used. As a dopant for n-type conversion, S may be used. Alternatively, dopants other than these may be used. Further, semi-insulating InP may be used. As a dopant for insulating InP, Fe may be used.

[0121] The embedded layer 221 may be formed by combining a plurality of InPs having different dopants. For example, as shown in FIG. 8C of Embodiment 1, the first embedded layer 121 and the second embedded layer 122 may be combined. The material constituting the first embedded layer 121 may be p-type InP. The material constituting the second embedded layer 122 may be semi-insulating InP.

[0122] Here, the case where the embedded layer 221 is used for embedding will be described. The embedded layer 221 may be formed by the MOCVD method. It is desirable that the embedded layer 221 has a current confinement function.

[0123] Next, step S7b of etching the first processing mask 271 is performed. In step S7b, as shown in FIGS. 24A, 24B, and 24C, the second processing mask 272 is partially formed on the entire first processing mask 271 and on the embedded layer 221. As a material for the second processing mask 272, for example, photoresist can be mentioned.

[0124] The second processing mask 272 is patterned into a desired shape using photolithography technology and etching technology. As shown in FIG. 24A, in plan view, the shape of the second processing mask 272 does not have to be of equal width in the light propagation direction. In plan view, the shape of the second processing mask 272 may be of equal width in the light propagation direction.

[0125] For example, it may have a first region 284 of equal width and a fourth region 287 that is wider than the first region 284 and of equal width, a second region 285 that changes to a tapered shape with a narrower width from the first region 284 toward the third region 286, and a third region 286 that changes to a tapered shape with a wider width from the second region 285 toward the fourth region 287. For example, it may be composed only of the second region 285 and the third region 286 that do not have regions of equal width.

[0126] For example, the second region 285 has a shape with a narrower width from the first region 284 side toward the third region 286 side, but may have a shape with a narrower width from the third region 286 side toward the first region 284 side. The shape of the second processing mask 272 may be composed of straight lines, or may be composed of curves. The shape of the second processing mask 272 may be composed of a combination of straight lines and curves.

[0127] At least a part of the active layer width Wa formed in step S7b is wider than the full width at half maximum Wf of the NFP in the horizontal direction of light. In this case, in all regions, the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light may be such that Wa > Wf. In Embodiment 2, in the first region 284 and the second region 285, it is treated as Wa ≤ Wf, and in the third region 286 and the fourth region 287, it is treated as Wa > Wf.

[0128] The mask width Wp of the second processing mask 272 satisfies Wa ≤ Wp for the active layer width Wa and the mask width Wp in the first region 284 and the second region 285, and Wa > Wp in the third region 286 and the fourth region 287. In FIG. 24A, in the first region 284 and the fourth region 287, both the active layer width Wa and the mask width Wp are of equal width. In the second region 285 and the third region 286, the active layer width Wa and the mask width Wp are changing. In the first region 284 and the second region 285, Wa - Wp ≤ 0.

[0129] In the third region 286 and the fourth region 287, Wa - Wp > 0. That is, when the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa ≤ Wf, Wa - Wp ≤ 0. On the other hand, when Wa > Wf, Wa - Wp > 0. However, for example, in the transition region from the second region 285 to the third region 286, due to manufacturing tolerances such as mask alignment, a region where Wa ≤ Wf and Wa > Wp may locally occur.

[0130] Next, using the above-described second processing mask 272 as a mask, the first processing mask 271 is etched. Specifically, as shown in FIG. 24A, the first processing mask 271 in the region where the second processing mask 272 is disposed in a plan view is not etched. On the other hand, the first processing mask 271 in the region where the second processing mask 272 is not disposed in a plan view is etched. As a result, as shown in FIGS. 25A, 25B, and 25C, a part of the first processing mask 271 is etched so that a part of the first upper cladding layer 203 is exposed.

[0131] As an etching method, wet etching may be used in addition to the above-described dry etching. As shown in FIGS. 26A, 26B, and 26C, after etching the first processing mask 271, the second processing mask 272 is removed by wet etching. As a result, as shown in FIG. 26A, the opening width Wb of the etched first processing mask 271 satisfies Wa ≤ Wb in the first region 284 and the second region 285.

[0132] In the third region 286 and the fourth region 287, the active layer width Wa and the aperture width Wb satisfy Wa > Wb. That is, when the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa ≤ Wf, Wa - Wb ≤ 0. When Wa > Wf, Wa - Wb > 0. However, for example, in the transition region from the second region 285 to the third region 286, due to manufacturing tolerances such as mask alignment, a region where Wa ≤ Wf and Wa > Wb may locally occur. In FIG. 26A, the first region 284 and the second region 285 are illustrated such that Wa = Wb, but Wa ≤ Wb may also be possible.

[0133] Next, a step S8b of forming the current blocking layer 222 is performed. In step S8b, as shown in FIGS. 27A, 27B, and 27C, the first processing mask 271 is used as a selective growth mask to form the current blocking layer 222 on a part of the embedded layer 221 and the first upper cladding layer 203. At this time, the current blocking layer 222 may be formed using the MOCVD method.

[0134] The material constituting the current blocking layer 222 may be, for example, InP. The conductivity type of InP may be p-type or n-type by adding a dopant. As a dopant for p-type conversion, Zn may be used. As a dopant for n-type conversion, S may be used. Dopants other than these may also be used. Furthermore, semi-insulating InP may be used. As a dopant for semi-insulating InP, Fe may be used.

[0135] The current blocking layer 222 may be formed by combining a plurality of InPs having different dopants. It is desirable that the embedded layer 221 and the current blocking layer 222 have a current confinement function in combination. As shown in FIGS. 28A, 28B, and 28C, after forming the current blocking layer 222, the first processing mask 271 is removed by wet etching.

[0136] Next, a step S9b of forming the second upper cladding layer 204 and a step S10b of forming the contact layer 241 are performed. The step S9b and the step S10b are the same as the step S7a and the step S8a in the first embodiment as shown in FIGS. 29A, 29B, and 29C.

[0137] Next, a step S11b of etching the contact layer 241 is performed. In the step S11b, although not shown, the contact layer 241 is etched in any part such as forming isolation between devices. The etching method is the same as the etching step S9a in the first embodiment. However, no restrictions are imposed on the shape of the contact layer 241 and the shape of the processing mask.

[0138] Next, a step S12b of forming a surface protection film made of an insulating film, a step S13b of etching the surface protection film, and a step S14b of forming the first electrode 261 and the second electrode 262 are performed. In the step S12b and the step S13b, although not shown, the surface protection film is etched in any part such as forming isolation between devices. The etching method and the formation location of the surface protection film are the same as the etching step S9a in the first embodiment. However, no restrictions are imposed on the shape of the surface protection film and the shape of the processing mask.

[0139] Also, as shown in FIGS. 30A, 30B, and 30C, the first electrode 261 and the second electrode 262 are formed by the same manufacturing method as the step S12a of forming the electrodes in the first embodiment.

[0140] <Operation of the optical semiconductor device according to the second embodiment> Next, the operation of the optical semiconductor device 200 according to the second embodiment will be described. Examples of the optical semiconductor device 200 according to the second embodiment include a semiconductor laser and a semiconductor optical amplifier. As an example of the optical semiconductor device 200, a semiconductor optical amplifier connected to a semiconductor laser serving as a light source from the first region 288 side will be described below, and its operation will be described.

[0141] Light propagates from a semiconductor laser to a semiconductor optical amplifier. The light is incident on the active layer 211 of the semiconductor optical amplifier. The incident light propagates while being amplified in the active layer 211 and travels from the first region 288 toward the third region 290. Since the active layer width Wa widens toward the third region 290, the light propagating within the active layer 211 also spreads horizontally. At the interface between the first region 288 and the second region 289, the full width at half maximum Wf of the NFP of the light in the horizontal direction becomes equal to the active layer width Wa.

[0142] After the second region 289, the full width at half maximum Wf of the NFP of the light in the horizontal direction becomes smaller than the active layer width Wa. In the third region 290 where the active layer width Wa and the full width at half maximum Wf of the NFP of the light in the horizontal direction satisfy Wa < Wf, the aperture width Wb becomes narrower with respect to the active layer width Wa. As a result, current is injected into the region with high light intensity, and high photoelectric conversion efficiency can be obtained.

[0143] <Operation of the optical semiconductor device according to Embodiment 2> The optical semiconductor device 200 according to Embodiment 2 includes an active layer 211 having an active layer width Wa and a current blocking layer 222 having an aperture width Wb. The active layer 211 in the optical semiconductor device 200 has a region where the active layer width Wa is wider than the full width at half maximum Wf of the NFP of the light in the horizontal direction, that is, a region where Wa > Wf, at least in part along the light propagation direction. In the region where the active layer width Wa and the full width at half maximum Wf of the NFP of the light in the horizontal direction satisfy Wa > Wf, at least in part along the light propagation direction, there is a region where the active layer width Wa and the aperture width Wb satisfy Wa > Wb.

[0144] With the configuration as described above, in the region where the active layer width Wa and the full width at half maximum Wf of the NFP of the light in the horizontal direction satisfy Wa > Wf, since the aperture width Wb is narrower than the active layer width Wa, current is not injected into the region where the light intensity in the active layer 211 becomes 1 / 2 or less, and current can be selectively injected into the portion with high light intensity.

[0145] In the above-described optical semiconductor device 200, in the first region 284 and the second region 285, the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light satisfy Wa ≤ Wf, and the active layer width Wa and the aperture width Wb satisfy Wa ≤ Wb. In the third region 286 and the fourth region 287, the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light are such that Wa > Wf, and the active layer width Wa and the aperture width Wb are such that Wa > Wb. With the above configuration, it becomes possible to always inject current only into the portion where the light intensity is high within the active layer 211. As a result, the photoelectric conversion efficiency of the optical semiconductor device is improved.

[0146] Figs. 31A and 31B are diagrams showing the simulation results of the current-light output characteristics when the aperture width Wb is changed in the case where the active layer width Wa and the full width at half maximum Wf of the NFP in the horizontal direction of light are such that Wa = 7 μm > Wf. As shown in Fig. 31A, it can be seen that the maximum light output value increases when the aperture width Wb is smaller than Wa = Wb = 7 μm. As shown in Fig. 31B, in the region represented by the following formula (2), the improvement effect of the light output is particularly large.

[0147] [Number] More preferably, in the region represented by the following formula (3), the improvement effect of the light output is even greater.

[0148] [Number] Therefore, in order to obtain a more effective photoelectric conversion efficiency, it can be said that it is preferable to set the active layer width Wa and the aperture width Wb such that they satisfy formula (3).

[0149] <Operation and Effect of the Optical Semiconductor Device According to Embodiment 2> As described above, according to the optical semiconductor device according to Embodiment 2, since a region where the active layer width Wa is larger than the aperture width Wb is provided in a part of the region along the light propagation direction, it becomes possible to always inject current only into the portion where the light intensity is high. As a result, the photoelectric conversion efficiency of the optical semiconductor device is improved.

[0150] Furthermore, according to the optical semiconductor device according to Embodiment 2, in a part of the region along the light propagation direction, a region is provided where the active layer width Wa is larger than the opening width Wb and the active layer width Wa is larger than the full width at half maximum Wf of the NFP in the horizontal direction of light. As a result, it becomes possible to always inject current only into the portion with high light intensity, and thus the photoelectric conversion efficiency of the optical semiconductor device is further improved.

[0151] Although various exemplary embodiments and examples are described in the present disclosure, the various features, aspects, and functions described in one or more of the embodiments are not limited to the application of a specific embodiment, but are applicable to the embodiments alone or in various combinations.

[0152] Therefore, countless variations not illustrated are envisioned within the scope of the technology of the present disclosure. For example, it includes cases where at least one component is modified, added, or omitted, and further cases where at least one component is extracted and combined with components of other embodiments.

Description of Reference Numerals

[0153] 100, 200 optical semiconductor devices, 101, 201 substrates, 102, 202 lower cladding layers, 103, 203 first upper cladding layers, 104, 204 second upper cladding layers, 111, 211 active layers, 120, 221 embedded layers, 121 first embedded layer, 122 second embedded layer, 123 third embedded layer, 130, 230 ridge structures, 141, 241 contact layers, 151 surface protective films, 161, 261 first electrodes, 162, 262 second electrodes, 171, 271 first processing masks, 172, 272 second processing masks, 173 third processing mask, 181, 281 first active layer regions, 182, 282 second active layer regions, 183, 283 third active layer regions, 184, 284, 288 first regions, 185, 285, 289 second regions, 186, 286, 290 third regions, 187, 287 fourth regions, 191, 291 first end faces, 192, 292 second end faces, 222 current blocking layer, Wa active layer width, Wb opening width, Wc contact layer width, Wf half-value full width of NFP in the horizontal direction of light, Wp mask width

Claims

1. circuit board and A stripe-shaped ridge structure consisting of a lower cladding layer, an active layer that guides light, and a first upper cladding layer, which are sequentially stacked on the substrate, An embedding layer formed to embed both sides of the ridge structure, A second upper cladding layer formed on the upper surface of the ridge structure and the upper surface of the embedding layer, The second upper cladding layer comprises a striped contact layer formed on the upper surface of the second upper cladding layer, In at least a portion of the region along the stripe direction, the contact layer width Wc, which is the distance between one end and the other end where the contact layer and the second upper cladding layer are in contact, is smaller than the active layer width Wa, which is the distance between one end and the other end where the active layer and the embedding layer are in contact in the active layer located directly beneath the contact layer. The optical semiconductor device is characterized in that the active layer width Wa is greater than the full width at half maximum Wf of the horizontal near-field pattern of light guiding the active layer.

2. circuit board and A stripe-shaped ridge structure consisting of a lower cladding layer, an active layer that guides light, and a first upper cladding layer, which are sequentially stacked on the substrate, An embedding layer formed to embed both sides of the ridge structure, A second upper cladding layer formed on the upper surface of the ridge structure and the upper surface of the embedding layer, A surface protective film formed on the second upper cladding layer and having openings, The system comprises a striped contact layer formed on the upper surface of the second upper cladding layer through an opening in the surface protective film, In at least a portion of the region along the stripe direction, the contact layer width Wc, which is the distance between the two ends where the contact layer and the opening of the surface protective film are in contact, is smaller than the active layer width Wa, which is the distance between one end of the active layer located directly beneath the contact layer and the other end where the active layer and the embedding layer are in contact. The optical semiconductor device is characterized in that the active layer width Wa is greater than the full width at half maximum Wf of the horizontal near-field pattern of light guiding the active layer.

3. The optical semiconductor device according to claim 1 or 2, characterized in that the active layer width Wa has at least a portion of the region in which it changes along the stripe direction.

4. The optical semiconductor device according to claim 1 or 2, characterized in that the contact layer width Wc is smaller than the active layer width Wa over the entire stripe direction.

5. The optical semiconductor device according to claim 1 or 2, characterized in that in a region where the active layer width Wa along the stripe direction is greater than the full width at half maximum Wf of the horizontal near-field pattern of light guiding the active layer, the contact layer width Wc is smaller than the active layer width Wa.

6. The optical semiconductor device according to claim 5, characterized in that in a region where the full width at half maximum Wf along the stripe direction is equal to or greater than the active layer width Wa, the contact layer width Wc is equal to or greater than the active layer width Wa.

7. circuit board and A stripe-shaped ridge structure consisting of a lower cladding layer, an active layer that guides light, and a first upper cladding layer, which are sequentially stacked on the substrate, An embedding layer formed to embed both sides of the ridge structure, A current block layer is formed on the upper surface of the ridge structure and the upper surface of the embedded layer, and its bottom surface is made of the first upper cladding layer, and is provided with an opening width Wb which is the distance between one end and the other end that are in contact with the first upper cladding layer on the bottom surface, The system comprises the upper surface of the first upper cladding layer exposed on the bottom surface of the opening and a second upper cladding layer formed on the upper surface of the current block layer, In at least a portion of the region along the stripe direction, the opening width Wb is smaller than the active layer width Wa, which is the distance between one end of the active layer and the embedding layer that are in contact with each other in the active layer located directly below the opening. The optical semiconductor device is characterized in that the active layer width Wa is greater than the full width at half maximum Wf of the horizontal near-field pattern of light guiding the active layer.

8. The optical semiconductor device according to claim 7, characterized in that the active layer width Wa has at least a portion of a region in which it changes along the stripe direction.

9. The optical semiconductor device according to claim 7, characterized in that the aperture width Wb is smaller than the active layer width Wa over the entire stripe direction.

10. The optical semiconductor device according to claim 7 or 8, characterized in that in a region where the active layer width Wa along the stripe direction is greater than the full width at half maximum Wf of the horizontal near-field pattern of light guiding the active layer, the aperture width Wb is smaller than the active layer width Wa.

11. The optical semiconductor device according to claim 10, characterized in that in a region where the full width at half maximum Wf along the stripe direction is equal to or greater than the active layer width Wa, the aperture width Wb is equal to or greater than the active layer width Wa.

12. The optical semiconductor device according to claim 1 or 2, characterized in that the value obtained by subtracting the contact layer width Wc from the active layer width Wa and dividing it by the active layer width Wa is within the range of 0.3 or more and 0.5 or less.

13. The optical semiconductor device according to claim 7, characterized in that the value obtained by subtracting the aperture width Wb from the active layer width Wa and dividing it by the active layer width Wa is within the range of 0.1 to 0.

6.

14. A substrate and, A stripe-shaped ridge structure consisting of a lower cladding layer, an active layer that guides light, and a first upper cladding layer, which are sequentially stacked on the substrate, An embedding layer formed to embed both sides of the ridge structure, A second upper cladding layer formed on the upper surface of the ridge structure and the upper surface of the embedding layer, The second upper cladding layer comprises a striped contact layer formed on the upper surface of the second upper cladding layer, In at least a portion of the region along the stripe direction, the contact layer width Wc, which is the distance between one end and the other end where the contact layer and the second upper cladding layer are in contact, is smaller than the active layer width Wa, which is the distance between one end and the other end where the active layer and the embedding layer are in contact in the active layer located directly beneath the contact layer. An optoelectronic semiconductor device characterized in that, in at least some other regions along the stripe direction, there is a region in which the contact layer width Wc is greater than the active layer width Wa.

15. A substrate and, A stripe-shaped ridge structure consisting of a lower cladding layer, an active layer that guides light, and a first upper cladding layer, which are sequentially stacked on the substrate, An embedding layer formed to embed both sides of the ridge structure, A current block layer is formed on the upper surface of the ridge structure and the upper surface of the embedded layer, and its bottom surface is made of the first upper cladding layer, and is provided with an opening width Wb which is the distance between one end and the other end that are in contact with the first upper cladding layer on the bottom surface, The system comprises the upper surface of the first upper cladding layer exposed on the bottom surface of the opening and a second upper cladding layer formed on the upper surface of the current block layer, In at least a portion of the region along the stripe direction, the opening width Wb is smaller than the active layer width Wa, which is the distance between one end of the active layer and the embedding layer that are in contact with each other in the active layer located directly below the opening, An optoelectronic semiconductor device characterized in that, in at least some other regions along the stripe direction, there is a region in which the aperture width Wb is greater than the active layer width Wa.

16. A process of sequentially stacking semiconductor layers, each consisting of a lower cladding layer, a light-guiding active layer, and a first upper cladding layer, on a substrate by epitaxial crystal growth, A step of forming a stripe-shaped ridge structure including at least each of the aforementioned semiconductor layers, A step of epitaxially growing a filling layer so as to fill both sides of the ridge structure, The process involves epitaxially growing a second upper cladding layer and a contact layer on the upper surface of the ridge structure and the upper surface of the embedding layer. A step of processing such that, in at least a portion of the region along the stripe direction, the contact layer width Wc, which is the distance between one end and the other end where the contact layer and the second upper cladding layer are in contact, is smaller than the active layer width Wa, which is the distance between one end and the other end where the active layer and the embedding layer are in contact in the active layer located directly beneath the contact layer, so that the active layer width Wa is larger than the full width at half maximum Wf of the horizontal near-field pattern of light guiding the active layer; A method for manufacturing an optoelectronic semiconductor device.

17. A process of sequentially stacking semiconductor layers, each consisting of a lower cladding layer, a light-guiding active layer, and a first upper cladding layer, on a substrate by epitaxial crystal growth, A step of forming a stripe-shaped ridge structure including at least each of the aforementioned semiconductor layers, A step of epitaxially growing a filling layer so as to fill both sides of the ridge structure, The process involves epitaxially growing a current-blocking layer on the upper surface of the ridge structure and the upper surface of the embedded layer. The process of forming an opening in the current block layer, the bottom surface of which is made of the first upper cladding layer and which has an opening width Wb that is the distance between one end and the other end that are in contact with the first upper cladding layer on the bottom surface, The system comprises the upper surface of the first upper cladding layer exposed on the bottom surface of the opening and a second upper cladding layer formed on the upper surface of the current block layer, A method for manufacturing an optical semiconductor device, characterized in that, in at least a portion of a region along the stripe direction, the aperture width Wb is smaller than the active layer width Wa of the active layer, which is the distance between one end of the active layer and the embedding layer that are in contact with each other in the active layer located directly below the aperture, and the active layer width Wa is larger than the full width at half maximum Wf of the horizontal near-field pattern of light guiding the active layer.

Images