JP4962367B2 - Optical semiconductor device - Google Patents

Description

  The present invention relates to an optical semiconductor device formed by optically connecting a plurality of semiconductor optical waveguides.

  FIG. 1 is a perspective view of a semiconductor laser 6 with an optical waveguide in which a semiconductor laser 2 and a semiconductor passive waveguide 4 are integrated. Here, FIG. 1A is an overall view of the semiconductor laser 6 with an optical waveguide. On the other hand, FIG. 1B is a perspective view of the semiconductor laser 6 with an optical waveguide as seen through the semiconductor laser 2 through the semiconductor buried layer 8.

  A semiconductor optical waveguide 12 (hereinafter referred to as a first optical waveguide) having a first core layer 10 as an active layer that generates or amplifies light, such as a semiconductor laser 6 with an optical waveguide (see FIG. 1B). ) And a semiconductor optical waveguide 16 (hereinafter referred to as a second optical waveguide) having a second core layer 14 formed of a semiconductor layer having a shorter band gap wavelength than the first core layer 10. The connected structure is the basic structure of the integrated optical semiconductor device.

  As shown in FIG. 1B, the side surface of the first optical waveguide 12 having the active layer is usually covered with a semiconductor buried layer 8. This is to prevent long-term stability of the optical semiconductor device by preventing the active layer (core layer 10) whose side surfaces are exposed from being deteriorated by oxidation or the like.

  Two methods have been reported as a method of forming a structure in which semiconductor optical waveguides are optically connected, such as a semiconductor laser with an optical waveguide.

  In the first method, first, a first slab waveguide including a first core layer and a second slab waveguide including a second core layer are butt-joint on a semiconductor substrate. To be formed.

  Next, the first slab waveguide is processed into a straight line to form the first optical waveguide 12, and then the side surface of the first optical waveguide is embedded by the semiconductor embedded layer. Further, the second slab waveguide is processed so as to be linear and connected to the first optical waveguide to form the second optical waveguide 16 (Patent Document 1).

  In order to improve the optical connection between the first and second optical waveguides, it is necessary to match the central axes of the first and second optical waveguides. However, in the first method, the optical exposure apparatus must be operated for a semiconductor optical waveguide having a width of only 1 to 2 μm. For this reason, the first method has a problem that the center axis of the semiconductor optical waveguide is likely to be shifted (Patent Document 2).

  When such misalignment of the central axis occurs, the optical coupling rate between the first and second optical waveguides decreases, and optical loss occurs. For this reason, the integrated optical semiconductor device formed by the first method has a large loss and does not have satisfactory characteristics.

  Therefore, a second method for simultaneously forming the first and second optical waveguides has been developed (Patent Document 2).

  FIG. 2 is a perspective view for explaining the second method.

  First, the first slab waveguide including the first core layer 10 and the second slab waveguide including the second core layer 14 are arranged on the n-type semiconductor substrate 17 at one interface (butt joint). It is formed so as to be in contact with the surface 22).

  Next, a linear first mask 18 is formed on the first and second slab waveguides across the butt joint surface 22. Using the first mask 18 as an etching mask, the first and second slab waveguides are etched.

  By this etching, as shown in FIG. 2A, the first and second optical waveguides 12 and 16 whose center axes 19 and 19 'are completely coincident are formed in the region protected by the etching mask 18. The

Next, as shown in FIG. 2B, a second mask 20 is formed which covers the second optical waveguide 16 whose top is protected by the first mask 18 and further extends on both sides thereof. Is done. Here, the first and second masks 18 and 20 are formed of, for example, SiO ₂ and Al ₂ O ₃ films, respectively.

  Next, as shown in FIG. 3A, a semiconductor buried layer 8 having a p-type semiconductor layer and an n-type semiconductor layer on a semiconductor substrate not covered with the first and second masks 18 and 20. Is selectively grown. Thereafter, the first and second masks 18 and 20 are removed.

  According to the second method, since the first and second optical waveguides 12 and 16 are formed using a straight single mask as an etching mask, the centers of the first and second optical waveguides 10 and 16 are formed. The axes are completely coincident.

Accordingly, the optical coupling rate between the first and second optical waveguides is increased. Therefore, the characteristics of the integrated optical semiconductor device formed by the second method are good.
JP2002-232069 JP-A-8-162706

  According to the second method, the two semiconductor optical waveguides can be optically connected in a state where the central axes substantially coincide with each other.

  However, this alone is not a perfect optical connection between semiconductor optical waveguides.

  As shown in FIG. 3A, both sides of the first optical waveguide 12 are embedded with the semiconductor embedded layer 8. That is, the first core layer 10 is covered with the semiconductor not only in the vertical direction but also in the horizontal direction parallel to the semiconductor substrate. For this reason, the change in the refractive index in the lateral direction of the first core layer 10 is negligible.

  For this reason, the propagating light oozes out from the first core layer 10 into the semiconductor buried layer 8. Therefore, the lateral beam diameter of the light propagating through the first optical waveguide 12 is wider than the width 24 of the first optical waveguide 12.

  On the other hand, both sides of the second optical waveguide layer 16 are in an empty state, that is, in a state where the refractive index is approximately 1.0. Therefore, the change in the refractive index in the lateral direction with respect to the second core layer 16 formed of a semiconductor (refractive index: 3.0 to 4.0) is as large as 2.0 to 3.0. For this reason, the light propagating through the second core layer 16 is almost completely confined in the second optical waveguide 16.

  Therefore, when the widths 24 and 24 'of the first and second optical waveguides 12 and 16 are equal, the electric field intensity distributions of the light propagating through the respective optical waveguides are not matched. For this reason, when light propagates from the first optical waveguide 12 to the second optical waveguide 16, an optical loss occurs (that is, the optical coupling rate becomes small). Note that what is referred to as “light electric field intensity distribution” here is an electric field intensity distribution in a plane perpendicular to the light traveling direction.

  In order to eliminate such a mismatch in the electric field strength distribution, it is effective to form the second optical waveguide 16 wider than the first optical waveguide 12.

  FIG. 4 is a plan view of such first and second optical waveguides 12 and 16. As shown in the figure, the width 24 ′ of the second optical waveguide 16 is formed wider than the width 24 of the first optical waveguide 12. In this way, the electric field strength distribution 26 of the first optical waveguide 12 and the electric field strength distribution 26 ′ of the second optical waveguide 16 can be substantially matched. Accordingly, the optical coupling rate between the first and second optical waveguides 12 and 16 is increased.

  In this case, the widths of the first and second optical waveguides are designed so as to increase the optical coupling rate under the assumption that the butt joint surface 22 and the end 28 of the semiconductor buried layer 8 coincide.

  The refractive index of the semiconductor material (for example, InP) forming the semiconductor optical waveguide exceeds 3.0. Therefore, the wavelength of light propagating through the semiconductor optical waveguide (for example, 1.55 μm) is well below 1 μm in the optical waveguide. Therefore, it is considered necessary to match the butt joint surface 22 and the end 28 of the semiconductor buried layer with an accuracy of 1 μm or less.

  In order to obtain such a match, the end of the second mask 20 used for the selective growth of the semiconductor buried layer 8 and the butt joint surface 22 must be matched with an accuracy of 1 μm or less (see FIG. 2B). ).

Therefore, in order to form the second mask 20, an insulator film (for example, a film deposited on the semiconductor substrate 17 using a high-precision exposure apparatus such as a stepper (reduction projection exposure apparatus) or an electron beam drawing apparatus (for example, Al ₂ O ₃ etc.) must be processed.

  Such a high-precision exposure apparatus is also required when forming an electrode on the first optical waveguide having an active layer.

  FIG. 5 is a plan view for explaining the state of the first optical waveguide 12 and the second optical waveguide 16 on which the electrodes 30 are formed. Here, the first optical waveguide 12 is drawn through the electrode 30 (the same applies hereinafter).

  FIG. 5 illustrates an ideal state in which the end 32 of the electrode 30 coincides with the end 28 of the semiconductor buried layer 8. However, unless a high-precision exposure apparatus such as a stepper or an electron beam drawing apparatus is used, it is practically difficult to match the ends of the electrode 30 and the semiconductor buried layer 8.

  FIG. 6 is a plan view for explaining a state in which the electrode 30 is detached from the end 28 of the semiconductor buried layer 8.

  FIG. 6A is a plan view for explaining a state in which the end 32 of the electrode 30 is formed inside the semiconductor buried layer 8. On the other hand, FIG. 6B is a plan view illustrating a state in which the end 32 of the electrode 30 is formed outside the semiconductor buried layer 8.

  When the electrode end 32 is formed inside the semiconductor buried layer 8 as shown in FIG. 6A, the electrode 30 does not exist in the region 34 between the butt joint surface 22 and the electrode end 32. . For this reason, the active layer present in this region 34 absorbs light emitted or amplified by the active layer present in the region 36 covered with the electrode 30. As a result, the output of the optical semiconductor device is reduced.

  On the other hand, when the end 32 of the electrode is formed outside the semiconductor buried layer 8 as shown in FIG. 6B, the electrode 30 not only extends on the second optical waveguide 16 but also one of them. The portion 38 protrudes from the second optical waveguide 16. The electrode 30 protruding from the second optical waveguide 16 contacts the substrate 17 exposed by the etching for forming the second optical waveguide 16. For this reason, part of the current to be injected into the active layer through the electrode 30 leaks to the substrate.

  As described above, if the end 32 of the electrode does not coincide with the end 28 of the semiconductor buried layer, the completed optical semiconductor device may deteriorate in characteristics (such as a decrease in optical output) or malfunction (leakage of injected current). In order to avoid such characteristic deterioration and malfunction, the electrode 30 also needs to be formed using a high-precision exposure apparatus such as a stepper or an electron beam drawing apparatus.

  As described above, in order to optically connect a first optical waveguide having an active layer (that is, an optical waveguide having a light emission or optical amplification function) and a second optical waveguide (for example, a passive waveguide). A high-precision exposure apparatus is necessary. However, a high-precision exposure apparatus is extremely expensive.

  Therefore, an object of the present invention is to provide a first optical waveguide (for example, a semiconductor laser or a semiconductor optical amplifier) having a light emission or optical amplification function that can be manufactured by an inexpensive exposure apparatus such as a contact exposure apparatus or a projection exposure apparatus. An optical semiconductor device in which a second optical waveguide (for example, a passive waveguide) is optically connected is provided. The optical semiconductor device includes an integrated semiconductor device.

  In order to achieve the above object, one aspect of the present invention includes a first semiconductor optical waveguide having a first core layer formed of a first semiconductor layer, and a side surface of the first semiconductor optical waveguide. A second core formed by a second semiconductor layer to be embedded, an electrode formed on the first semiconductor optical waveguide, and a third semiconductor layer having a band gap wavelength shorter than that of the first semiconductor layer A second semiconductor optical waveguide having a layer and optically connected to the first semiconductor optical waveguide, wherein the second semiconductor layer includes the first semiconductor optical waveguide and the second semiconductor optical waveguide. An end of the second semiconductor layer is embedded in a part of the side surface of the second semiconductor optical waveguide beyond the boundary of the semiconductor optical waveguide, and the electrode embeds the boundary and the side surface of the second semiconductor optical waveguide. It extends between.

  In the optical semiconductor device according to the above aspect, the second semiconductor layer embeds a part of the side surface of the second semiconductor optical waveguide beyond the boundary between the first semiconductor optical waveguide and the second semiconductor optical waveguide. Therefore, it can be easily manufactured using a contact exposure apparatus (or projection exposure apparatus).

  Further, since the second semiconductor layer embeds a part of the side surface of the second semiconductor optical waveguide beyond the boundary between the first semiconductor optical waveguide and the second semiconductor optical waveguide, The electrode can be easily formed between the end of the two semiconductor optical waveguides and the boundary.

  Therefore, the optical semiconductor device according to the above aspect does not cause malfunction (leakage of injected current) or characteristic deterioration (decrease in output, etc.). Furthermore, in the optical semiconductor device according to the above aspect, the first and second optical waveguides are coupled with high efficiency.

  According to the present invention, a first optical waveguide (for example, a semiconductor laser or semiconductor optical amplifier) having a light emission or optical amplification function and a second optical waveguide (for example, a semiconductor passive waveguide) are optically connected with high efficiency. Manufacturing optical semiconductor devices that do not cause malfunctions (injection current leakage) or characteristic deterioration (decrease in output, etc.) using inexpensive optical exposure devices such as contact exposure devices and projection exposure devices can do.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, the technical scope of the present invention is not limited to these embodiments, but extends to the matters described in the claims and equivalents thereof.

(Embodiment 1)
This embodiment relates to a semiconductor laser 6 with an optical waveguide that can be manufactured by a contact exposure apparatus or a projection exposure apparatus.

(1) Device Configuration FIG. 7 is a plan view of such a semiconductor laser 40 with an optical waveguide. FIG. 8 is a cross-sectional view of the semiconductor laser 40 with an optical waveguide taken along line AA ′ of FIG. FIG. 9 is an enlarged plan view of the region B shown in FIG. 7 where the second optical waveguide 16 (semiconductor passive waveguide) is connected to the first optical waveguide 12 (semiconductor laser). .

  As shown in FIGS. 7 and 8, the semiconductor laser 40 with an optical waveguide according to the present embodiment includes a first core layer 10 formed by a first semiconductor layer (that is, an active layer 46) having an optical amplification function. And a second semiconductor layer (semiconductor embedded layer 8) that embeds the side surface of the first semiconductor optical waveguide 10. Here, the first optical waveguide 12 has a diffraction grating (not shown) extending along the active layer.

  The semiconductor laser 40 with an optical waveguide has an electrode 30 formed on the first semiconductor optical waveguide 12 for injecting current into the first core layer 10.

  The semiconductor laser 40 with an optical waveguide has a second core layer 14 formed of a third semiconductor layer having a shorter bandgap wavelength than the first semiconductor layer (active layer 46), and the first semiconductor A second semiconductor optical waveguide 16 optically connected to the optical waveguide 12 is provided.

  In the semiconductor laser with an optical waveguide 40 according to the present embodiment, as shown in FIG. 9, the second semiconductor layer (semiconductor buried layer 8) includes the first semiconductor optical waveguide 12 and the second semiconductor optical waveguide. The second semiconductor optical waveguide 16 extends from the boundary (butt joining surface 22) of 16 to 1 to 3 times (for example, 5 μm to 15 μm) of the alignment accuracy (for example, 5 μm) of the contact exposure apparatus. The side of is embedded. In FIG. 9, the first and second optical waveguides 12 and 16 are drawn through the electrode 30.

  Furthermore, in the semiconductor laser 40 with an optical waveguide according to the present embodiment, the electrode 30 extends between the boundary (butt joint surface 22) and the end 28 of the second semiconductor layer (semiconductor buried layer 8). ing.

  Here, the second semiconductor layer (semiconductor buried layer 8) is 1 to 3 times the alignment accuracy (for example, 3 μm) of the projection exposure apparatus, not the contact exposure apparatus, from the boundary (butt joint surface 22). (For example, 3 μm or more and 9 μm or less) may be extended.

(2) Manufacturing Method Next, the reason why the semiconductor laser 40 with an optical waveguide can be manufactured by a contact exposure apparatus or a projection exposure apparatus will be described according to the manufacturing method.

  FIG. 10 is a flowchart for explaining an example of the manufacturing procedure of the semiconductor laser 40 with an optical waveguide according to the present embodiment. 11 to 13 are plan process diagrams for explaining an example of the method for manufacturing the semiconductor laser 40 with an optical waveguide according to the present embodiment.

(Step S2)
First, a lower clad layer 42 made of n-type InP and a lower portion made of InGaAsP are formed on a semiconductor substrate 17 formed of n-type InP using metal organic vapor phase epitaxy (MOVPE). An SCH (Separate Confinement Heterostrucure) layer 44, an InGaAsP active layer 46 having a bandgap wavelength λg of 1.55 μm, an upper SCH layer 48, a p-type InP upper cladding layer 50, and an InGaAsP contact layer 52 Are sequentially stacked to form a first semiconductor stacked structure 54 (see FIG. 8). The active layer 46 may be an InGaAsP multi-quantum well (MQW). A diffraction grating (not shown) is formed in the vicinity of the active layer 46.

  Here, the lower SCH layer 44, the active layer 46, and the upper SCH layer 48 form the first core layer 10.

  Next, a part of the first semiconductor multilayer structure 54 is removed to expose the semiconductor substrate 17. On this exposed semiconductor substrate 17, an n-type InP lower cladding layer 42 ′ and a band gap wavelength λg of 1.3 μm are formed using metal organic vapor phase epitaxy (MOVPE). The second core layer 14 made of non-doped InGaAsP and the upper cladding layer 50 'made of p-type InP are sequentially laminated to form a second semiconductor laminated structure 56 (see FIGS. 8 and 11A). .

  That is, by this step, a butt structure 58 in which the first and second semiconductor multilayer structures 54 and 56 are abutted on the butt joint surface 22 is formed (see FIG. 11A).

(Step S4)
Next, for example, a SiO ₂ film is deposited on the butt structure 58.

Next, this SiO ₂ film is processed using a photolithography technique, and as shown in FIG. 11B, the first semiconductor multilayer structure 54 has a width of 2.0 μm and the second semiconductor multilayer structure. On the side of 56, a linear SiO ₂ etching mask 60 having a width of 2.5 μm is formed. 11 to 13 are enlarged views of the vicinity of the butt joint surface 22.

  The etching mask 60 is processed according to the following procedure.

  FIG. 14 is a plan view of a photomask used for manufacturing a semiconductor laser with an optical waveguide according to the present embodiment.

First, a photoresist is applied on the SiO ₂ film.

Next, a photomask (see FIG. 14A) having an opaque region 62 having the same shape as the SiO ₂ etching mask 60 is prepared, and a photoresist applied on the SiO ₂ film is applied to the photomask through the photomask. Irradiate ultraviolet rays. A contact exposure apparatus or a projection exposure apparatus (hereinafter simply referred to as a light exposure apparatus) is used for ultraviolet irradiation. At this time, the light exposure apparatus is operated so that the discontinuous line 64 in which the width of the opaque region 62 changes coincides with the butt joint surface 22 (of the butt structure 58).

Thereafter, the photoresist is developed, and a photoresist etching mask (not shown) having the same shape as the opaque region 62 is formed. Using the photoresist etching mask as a mask, the SiO ₂ film is etched to form an SiO ₂ etching mask 60 (see FIG. 11B).

(Step S6)
Next, using the SiO ₂ etching mask 60 as a mask, the butt structure 58 is etched by ICP (Inductively Coupled Plasma) reactive ion etching. By this etching, the first and second optical waveguides 12 and 16 are formed (see FIG. 12A).

  The first semiconductor optical waveguide having the first core layer 10 formed of the first semiconductor layer (active layer) among the configurations described in the above “apparatus configuration” by the procedure so far (steps S2 to S6). A second optical core 12 having a second core layer 14 formed of a waveguide 12 and a semiconductor layer having a shorter bandgap wavelength than the first semiconductor (active layer) and optically connected to the first semiconductor optical waveguide 12 The semiconductor optical waveguide 16 is formed using an optical exposure apparatus. The active layer is a semiconductor layer having an optical amplification function.

(Step S8)
Next, for example, an Al ₂ O ₃ film is deposited on the semiconductor substrate 17 on which the first and second optical waveguides 12 and 16 are formed.

Next, the Al ₂ O ₃ film is processed using a photolithographic technique, and butt-joining is performed so as to cover the second optical waveguide 16 whose top is covered with the SiO ₂ etching mask 60 and the vicinity thereof. A formation mask 66 is formed (see FIG. 12B).

The processing of the Al ₂ O ₃ film is performed according to substantially the same procedure as that for the SiO ₂ etching mask 60 described in step S4. However, the photoresist etching mask is formed using a photomask having a rectangular opaque region 62 'as shown in FIG.

  In addition, the light exposure apparatus is operated so that the end 68 of the opaque region 62 ′ is located at a position that enters the second optical waveguide 16 side by a distance twice the accuracy of the light exposure apparatus from the butt joint surface 22. The photoresist is irradiated with ultraviolet rays.

  Here, the accuracy of the photoexposure device means that a skilled worker operates the photoexposure device to expose the photoresist film, and further develops the photoresist film to form a desired shape (resist pattern). And an index indicating how much the shape deviates from the target position.

  FIG. 15 is a diagram for explaining the distribution of the deviation δ of the resist pattern with respect to the target position. The deviation δ of the exposure pattern is considered to follow a normal distribution as shown in FIG. The horizontal axis in FIG. 15 is a value obtained by normalizing the deviation δ by the standard deviation σ. On the other hand, the vertical axis represents the frequency (density function) at which the deviation δ occurs.

  As is clear from the nature of the normal distribution, the pattern deviation δ occurs within a range of approximately 100% (more precisely, 99.7%) and −3σ <δ <+ 3σ (see FIG. 15). It is this 3σ that is called the accuracy of the light exposure apparatus.

  As is well known, the accuracy of the contact exposure apparatus is 5 μm. Accordingly, when the standard deviation of the positional deviation distribution of the resist pattern formed by using the contact exposure apparatus is tripled, it becomes 5 μm. On the other hand, the accuracy of the projection exposure apparatus is 3 μm. Accordingly, when the standard deviation of the positional deviation distribution of the resist pattern formed by using the projection exposure apparatus is tripled, it becomes 3 μm.

FIG. 16 is a diagram for explaining a deviation distribution of a resist pattern used for forming the butt-bonding formation mask 66 made of Al ₂ O ₃ . In the upper left part of FIG. 16, a scale 74 with the alignment accuracy of the light exposure apparatus as a unit is shown.

  As described above, in this step, light is emitted so that the end 68 of the opaque region 62 ′ is located at a position that enters the second optical waveguide 16 side from the butt joint surface 22 by a distance twice the accuracy of the light exposure apparatus. The exposure apparatus is operated. Therefore, the end 70 of the photoresist pattern to which the opaque region 62 'has been transferred is at least one time the alignment accuracy of the light exposure apparatus from the butt joint surface 22 to the second optical waveguide 16 as shown in FIG. It is distributed within a range 72 that extends three times or less.

  Specifically, when the contact exposure apparatus is used, the end 70 of the photoresist pattern (for forming the butt-bonding formation mask 66) is 5 μm (== 5 μm from the butt-bonding surface 22 to the second optical waveguide 16 side. It is distributed in the range of 10 μm−5 μm, that is, 5 μm × 1) to 15 μm (= 10 μm + 5 μm, that is, 5 μm × 3). On the other hand, when the projection exposure apparatus is used, the edge 70 of the photoresist pattern is 3 μm (= 6 μm−3 μm, that is, 3 μm × 1) to 9 μm (= 6 μm + 3 μm, that is, from the butt joint surface 22 to the second optical waveguide 16 side. It is distributed in the range of 3 μm × 3).

(Step S10)
Next, by using the MOVPE method, as shown in FIG. 13A, a semiconductor buried layer 8 made of, for example, semi-insulating InP is formed in a region not covered with the butt-bonding formation mask 66 and the SiO ₂ etching mask 60. Will grow. Instead of semi-insulating InP, a PN buried structure in which p-type InP, n-type InP, and p-type InP are stacked may be grown.

  That is, in steps S8 and S10, the semiconductor buried layer 8 that embeds the side surface of the first semiconductor optical waveguide 12 is the boundary between the first semiconductor optical waveguide 12 and the second semiconductor optical waveguide layer 16 (that is, the butt joint surface 22). ) To a distance that is twice the alignment accuracy of the optical exposure apparatus.

As a result, the contact exposure apparatus (or projection exposure apparatus) from the boundary (butt joint surface 22) between the first semiconductor optical waveguide 12 and the second semiconductor optical waveguide 16 in the configuration described in the “apparatus configuration” above. A semiconductor buried layer 8 is formed to extend from one to three times the alignment accuracy and bury the side surface of the second semiconductor optical waveguide 16.
(Step S12)
Next, the butt joint formation mask 66 and the SiO ₂ etching mask 60 are removed.

  Next, the electrode 30 is formed by, for example, a lift-off method using an optical exposure apparatus.

  At this time, in order to form a lift-off resist film, a photomask having a transparent region 76 corresponding to the electrode 30 as shown in FIG. 14C is used.

  As shown in FIG. 13B, the light exposure apparatus is operated so that the end 78 on one side of the transparent region 76 comes to the end 80 of the semiconductor buried layer 8 and the center 80 of the butt joint surface 22 to lift off. A resist film is formed.

  Next, an electrode material for p-type contact is vapor-deposited on the surface on which the lift-off resist film is formed, and then the excess electrode material is removed together with the lift-off resist film.

  Next, an electrode material for n-type contact is deposited on the back surface of the semiconductor substrate 17.

  Finally, heat treatment is performed so that the contact layer and the semiconductor substrate react with the electrode material for p-type and n-type contacts, respectively.

  That is, in this step, the current of the current to the first core layer is extended using the light exposure apparatus so as to extend from the butt joint surface 22 to the end 28 of the semiconductor buried layer 8 and the center 80 of the butt joint surface 22. An electrode 30 for performing the implantation is formed (see FIG. 13B).

  As is apparent from the above description, due to the alignment error of the light exposure apparatus, the distance between the end 28 of the semiconductor buried layer and the butt joint surface 22 becomes 1 to 3 times the accuracy of the light exposure apparatus. As described above, the accuracy of the optical exposure apparatus is three times the standard deviation σ of the positional deviation distribution of the formed resist pattern. Therefore, even when the distance between the end 28 of the semiconductor buried layer and the butt joint surface 22 is the narrowest (that is, when the distance is one time the alignment accuracy), the end 28 of the semiconductor buried layer is not connected to the end 32 of the electrode. The standard deviation σ is 1.5 times (= 3σ × 0.5) away from the planned formation position (the end of the semiconductor buried layer 28 and the center of the butt joint surface 22). Similarly, the butt joint surface 22 is also 1.5 times (= 3σ × 0.5) the standard deviation σ from the position where the electrode end 32 is to be formed.

  FIG. 17 is a cumulative distribution of positional deviations obtained by integrating the frequency of occurrence of positional deviations shown in FIG. 15 (vertical axis in FIG. 15) between −δ and + δ. The horizontal axis is a value obtained by normalizing the positional deviation amount δ with the standard deviation σ. The vertical axis represents the cumulative distribution of positional deviations, that is, the probability that the positional deviation is between -δ and δ.

  As is apparent from this figure, the end 28 of the electrode is formed between the end 28 of the semiconductor buried layer and the butt joint surface 22 with a probability of about 90%. That is, according to the manufacturing method according to the present embodiment, the end 32 of the electrode is extended between the butt joint surface 22 and the end 28 of the semiconductor buried layer with a probability of about 90% even if the optical exposure apparatus is used. be able to.

  Therefore, this step is formed on the first semiconductor optical waveguide 12 in the configuration described in the above “apparatus configuration”, and extends between the butt joint surface 22 and the end 28 of the semiconductor buried layer. An electrode is formed.

  As is apparent from the above description, according to the manufacturing method according to the present embodiment, the semiconductor laser 40 with an optical waveguide described in the above “apparatus configuration” can be manufactured using an inexpensive optical exposure apparatus. .

(3) Operation and Characteristics As described above, in the semiconductor laser with an optical waveguide 40 according to the present embodiment, as shown in FIG. 13B, the electrode 30 is formed between the butt joint surface 22 and the end 28 of the semiconductor buried layer. Extending in between.

  Therefore, the top of the first optical waveguide 12 is completely covered with the electrode 30. For this reason, since current is injected into the entire active layer 46, there is no region that acts as an absorber for propagating light in the active layer. Therefore, in the semiconductor laser with an optical waveguide 40 according to the present embodiment, an operation failure that causes a decrease in the output of the semiconductor laser with an optical waveguide 40 does not occur.

  Further, since the electrode 30 extends between the butt joint surface 22 and the end 28 of the semiconductor buried layer, the electrode 30 does not protrude outside the semiconductor buried layer 8. Accordingly, a part of the electrode 30 contacts the substrate 17 exposed by etching for forming the first and second optical waveguides 12 and 16, and a part of the current to be injected into the active layer leaks to the substrate. There is nothing.

  By the way, in the semiconductor laser 40 with an optical waveguide according to the present embodiment, a position 82 where the widths of the optical waveguides 12 and 16 are switched is formed inside the semiconductor buried layer 8 (see FIG. 7). In this case, the electric field distribution of the light propagating through the first and second optical waveguides 12 and 16 does not coincide with each other at the junction surface, and there is a possibility that optical coupling loss occurs.

  The structure in which the width of the second optical waveguide 16 is made wider than the first optical waveguide 12 as shown in FIG. 7 is that the end 28 of the semiconductor buried layer coincides with the optical waveguide width switching position 82. Is designed to minimize the coupling loss between the first and second optical waveguides. Therefore, if both do not match, the coupling loss increases.

  FIG. 18 is a diagram illustrating the relationship between the distance (Z-axis offset ΔZ) where the optical waveguide width switching position 82 enters the semiconductor buried layer 8 and the coupling loss. The horizontal axis in FIG. 18 represents the Z-axis offset. On the other hand, the vertical axis represents the coupling loss between the first and second optical waveguides.

  FIG. 18 shows a result of obtaining a coupling loss by numerical calculation for the first and second optical waveguides 12 and 16 (waveguide widths of 2.0 μm and 2.5 μm). FIG. 18 shows the coupling loss for each of the TE mode and the TM mode.

  FIG. 19 shows a state in which the first optical waveguide 12 and the second optical waveguide 16 are joined while being displaced. In this figure, a Z-axis offset ΔZ and an X-axis offset described later are shown.

  As shown in FIG. 18, the excess loss (increased coupling loss with respect to ΔZ = 0) is at most −0.25 dB even when ΔZ = 10 μm. Further, even when ΔZ exceeds 10 μm, the increase in coupling loss is slight (for example, the excess loss at ΔZ = 15 μm is 0.15 dB).

  Therefore, even if the optical waveguide width switching position 82 is formed inside the semiconductor buried layer 8, there is almost no deterioration in characteristics (such as a decrease in output) of the semiconductor laser with an optical waveguide due to an increase in coupling loss.

  By the way, as described in step S4 above, the optical exposure apparatus is operated so that the widths of the optical waveguides 12 and 16 are changed at the butt joint surface 22. However, the optical waveguide width switching position 82 is displaced from the butt joint surface 22 within the accuracy range of the optical exposure apparatus. However, even if such a deviation occurs, the first and second optical waveguides 12 and 16 are formed of semiconductors having similar compositions, and thus the calculation result does not change much.

  Further, considering the position of the butt joint surface 22, it is clear that the optical waveguide width switching position 82 does not protrude from the semiconductor buried layer 8.

  For comparison, FIG. 20 shows the coupling loss when the center axes of the first and second optical waveguides are displaced. The horizontal axis in FIG. 20 represents the deviation of the central axes of the first and second optical waveguides (X-axis offset; see FIG. 19). On the other hand, the vertical axis represents the coupling loss between the first and second optical waveguides. As shown in FIG. 20, even if the central axis is slightly shifted, the coupling loss is remarkably increased.

  As described above, in the present embodiment, the second semiconductor layer is one of the contact exposure apparatus and the projection exposure apparatus from the boundary between the first semiconductor optical waveguide and the second semiconductor optical waveguide. It extends 1 to 3 times the alignment accuracy, embeds the side surface of the second semiconductor optical waveguide, and the electrode extends between the boundary and the end of the second semiconductor layer. ing.

  That is, in the optical semiconductor device according to the present embodiment, the second semiconductor layer is either a contact exposure apparatus or a projection exposure apparatus from the boundary between the first semiconductor optical waveguide and the second semiconductor optical waveguide. 1 to 3 times the alignment accuracy. Therefore, it can be easily manufactured using either the contact exposure apparatus or the projection exposure apparatus.

  Furthermore, the distance between the edge of the second semiconductor layer and the boundary extends from 1 to 3 times the alignment accuracy of either the contact exposure apparatus or the projection exposure apparatus, so that the electrode can be securely connected. It can be formed during the above interval.

  Therefore, the optical semiconductor device according to the present embodiment does not cause malfunction (injection current leakage) or characteristic deterioration (output decrease, etc.). Furthermore, in the optical semiconductor device according to the present embodiment, the first and second optical waveguides are coupled with high efficiency.

  Therefore, even if the semiconductor laser with an optical waveguide according to the present embodiment is manufactured using an inexpensive optical exposure apparatus, malfunction (injection current leakage) or characteristic deterioration (decrease in output) due to displacement of the electrode formation position. Etc.).

(Embodiment 2)
The present embodiment relates to a semiconductor laser with an optical waveguide that can be manufactured by a contact exposure apparatus or a projection exposure apparatus, and relates to a semiconductor laser with an optical waveguide that changes in a tapered shape.

  FIG. 21 is a plan view of a semiconductor laser 82 with an optical waveguide according to the present embodiment. The configuration of the semiconductor laser 82 with an optical waveguide according to the present embodiment is a tapered region 86 in contact with the first optical waveguide 12 (semiconductor laser), and the width of the second optical waveguide 16 is directed toward the first optical waveguide 12. And has a taper shape.

  As described in the first embodiment, it is effective to make the width of the second optical waveguide wider than that of the first optical waveguide in order to reduce the coupling loss. However, if the waveguide width becomes too wide, the second optical waveguide becomes multimode. When such a semiconductor laser with an optical waveguide is mounted, the characteristics of the integrated optical device deteriorate.

  In such a case, as shown in FIG. 21, in the straight region 88 away from the first optical waveguide 12, the width of the second optical waveguide 16 is narrowed to satisfy the single mode condition, and the first In the tapered region 86 in contact with the optical waveguide 12, the width of the second optical waveguide 16 may be gradually increased in a tapered shape (that is, the waveguide width may be changed in an adiabatic manner).

  In such an optical waveguide structure, the light generated or amplified in the first optical waveguide 12 does not excite higher-order modes even when entering the tapered region 86 and reaches the linear region 88 while maintaining the fundamental mode.

  Therefore, according to the present embodiment, it is possible to reduce the optical coupling loss while the second optical waveguide 16 maintains the single mode.

  If it is desired to make the width of the second optical waveguide 16 wider than the width that minimizes the coupling loss with the first optical waveguide, the waveguide width is reduced in the tapered region 86, contrary to the example shown in FIG. What is necessary is just to make it narrow toward the 1st optical waveguide 12 little by little.

(Embodiment 3)
The present embodiment relates to an optical gate switch in which semiconductor optical amplifiers with optical waveguides are integrated in an array.

  FIG. 22 is a plan view of optical gate switch 90 according to the present embodiment. The optical gate switch 90 according to the present embodiment has a plurality of semiconductor optical amplifiers 92 with optical waveguides, in which passive optical waveguides are connected to a semiconductor optical amplifier (SOA), and the semiconductor optical amplifiers 92 with optical waveguides as inputs. A multimode interference (MMI) coupler 94 connected in parallel to the side, and a semiconductor optical amplifier 92 ′ with an optical waveguide connected to the output side of the multimode interference coupler 94.

  The configuration of the semiconductor optical amplifier 92 with an optical waveguide is substantially the same as that of the semiconductor laser 40 with an optical waveguide according to the first embodiment, except that no diffraction grating is provided in the vicinity of the active layer.

  The signal light incident on each channel formed by the semiconductor optical amplifier 92 with the optical waveguide on the input side is selected by the switching function of the semiconductor optical amplifier, and the semiconductor optical amplifier 92 ′ with the optical waveguide on the output side is selected by the multimode interference coupler 94. Led to. Then, the semiconductor optical amplifier 92 ′ with an optical waveguide on the output side amplifies and outputs the selected optical signal. That is, the optical gate switch 90 according to the present embodiment can select and output light incident on a desired channel.

  In optical gate switch 90 according to the present embodiment, an electrode for current injection is formed on the entire top of the semiconductor optical amplifier. Accordingly, the amplification factor of each semiconductor optical amplifier forming the optical gate switch 90 is high. In addition, since the coupling ratio between the semiconductor optical amplifier and the optical waveguide is high, high-quality signal processing becomes possible.

  A flat field coupler (FFC) may be used instead of MMI. The details of the flat field coupler have been reported by M. Bouda et al. (M. Bouda, S. Takabayashi, K. Morito and Y. Kotaki, Proceedings of CPT (Contemporary Photonics Technology) International Symposium, “Novel power splitter with significantly reduced device length, ”E-16, pp.91-92, Tokyo, Japan, Jan. 2002.).

(Embodiment 4)
The present embodiment relates to a monolithically integrated tunable light source 96 in which semiconductor lasers with optical waveguides are integrated in an array.

  FIG. 23 is a plan view of monolithically integrated variable wavelength light source 96 according to the present embodiment. The monolithic integrated wavelength tunable light source 96 according to the present embodiment includes a plurality of semiconductor lasers with optical waveguides according to the first embodiment, and a multimode interference coupler 94 in which the plurality of semiconductor lasers with optical waveguides 40 are connected to the input side. And a semiconductor optical amplifier 92 with an optical waveguide connected to the output side of the multimode interference coupler 94.

  However, the period of the diffraction grating is adjusted so that each of the plurality of semiconductor lasers 40 with an optical waveguide emits light having a slightly different wavelength.

  Laser light generated by each semiconductor laser (LD) is guided to a semiconductor optical amplifier (SOA) by a multimode interference coupler 94, amplified by the SOA, and output. Therefore, the wavelength of the output light can be controlled by switching the semiconductor laser to be operated.

  In the monolithic integrated wavelength tunable light source 96 according to the present embodiment, a current injection electrode is formed on the entire top of the semiconductor laser (LD) and the semiconductor optical amplifier (SOA). Therefore, the outputs of the semiconductor laser (LD) and the semiconductor optical amplifier (SOA) are large. Further, since the coupling ratio of the optical waveguide connected to the semiconductor laser (LD) and the semiconductor optical amplifier is high, the loss of light is small.

  Therefore, according to the monolithic integrated variable wavelength light source 96 according to the present embodiment, a large output power can be obtained.

It is a perspective view of a semiconductor laser with an optical waveguide. It is process drawing explaining the 2nd formation method of the structure in which two semiconductor optical waveguides were optically connected (the 1). It is process drawing explaining the 2nd formation method of the structure in which two semiconductor optical waveguides were optically connected (the 2). It is a top view of the 2nd optical waveguide formed wider than the 1st optical waveguide and the 1st optical waveguide. It is a top view explaining the state in which the electrode was formed on the 1st optical waveguide. It is a top view explaining the state which the electrode removed from the edge of the semiconductor embedding layer. 1 is a plan view of a semiconductor laser with an optical waveguide according to a first embodiment. It is sectional drawing of the semiconductor laser with an optical waveguide along the AA 'line of FIG. It is the top view to which the area | region B shown by FIG. 7 where the semiconductor passive waveguide and the semiconductor laser were connected was expanded. FIG. 6 is a flowchart illustrating a manufacturing procedure of the semiconductor laser with an optical waveguide according to the first embodiment. FIG. 6 is a plan process diagram for explaining the method for manufacturing the semiconductor laser with an optical waveguide according to the first embodiment (part 1); FIG. 6 is a planar process diagram for explaining the method of manufacturing the semiconductor laser with an optical waveguide according to the first embodiment (No. 2). FIG. 6 is a plan process diagram for explaining the method for manufacturing the semiconductor laser with an optical waveguide according to the first embodiment (part 3); It is a figure explaining the structure of the photomask used for manufacture of the semiconductor laser with an optical waveguide which concerns on Embodiment 1. FIG. It is a figure explaining the distribution of the shift | offset | difference of the shape (pattern) formed using the optical exposure apparatus. It is a figure explaining the shift | offset | difference of the formation position of the resist mask used for forming the butt joint formation mask. It is a figure explaining the cumulative distribution of the shift | offset | difference of the shape (pattern) formed using the optical exposure apparatus. It is a figure explaining the relationship of the coupling | bonding loss between the distance (DELTA) Z where the junction surface of the optical waveguide entered the inside of the semiconductor embedding layer, and the 1st and 2nd optical waveguide. It is a figure explaining the state where the 1st optical waveguide and the 2nd optical waveguide were shifted and joined. It is a figure explaining the coupling loss when the center axis of the 1st and 2nd optical waveguide has shifted. 6 is a plan view of a semiconductor laser with an optical waveguide according to a second embodiment. FIG. 6 is a plan view of an optical gate switch according to a third embodiment. FIG. FIG. 10 is a plan view of a monolithic integrated wavelength tunable light source according to the fourth embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 2 ... Semiconductor laser 4 ... Semiconductor passive waveguide 6 ... Semiconductor laser with an optical waveguide 8 ... Semiconductor embedding layer 10 ... 1st core layer 12 ... 1st optical waveguide
DESCRIPTION OF SYMBOLS 14 ... 2nd core layer 16 ... 2nd optical waveguide 17 ... Semiconductor substrate 18 ... 1st mask 19, 19 '... Central axis 20 ... 2nd mask 22 ... Butt joint surfaces 24, 24 '... Waveguide widths 26, 26' ... Electric field strength distribution 28 ... End of semiconductor buried layer 30 ... Electrode 32 ... End of electrode 34 ... -Area 36 between butt joint surface and electrode end-Area covered with electrode 38-Extruded electrode 40-Semiconductor laser with optical waveguide (Embodiment 1)
42 ... Lower cladding layer 44 ... Lower SCH layer 46 ... Active layer 48 ... Upper SCH layer 50 ... Upper cladding layer 52 ... Contact layer 54 ... First semiconductor multilayer structure 56 ... Second semiconductor laminated structure 58 ... Butt structure
60 ... SiO ₂ etching mask 60 ... opaque region 64 ... discontinuous line 66 ... butt joint formation mask 68 ... end of opaque region 70 ... photo with opaque region transferred Resist pattern edge 72 ... Photoresist pattern edge distribution range 74 ... Scale with alignment accuracy as unit ... Transparent area 78 ... Transparent area edge 80 ... Semiconductor buried layer End and center of butt joint surface 82 ... Switching position of waveguide width
84... Semiconductor laser with optical waveguide (Embodiment 2)
86 ... Tapered region 88 ... Linear region 90 ... Optical gate switch 92,92 '... Semiconductor optical amplifier 94 with optical waveguide ... MMI 96 ... Monolithically integrated tunable light source

Claims (6)

A first semiconductor optical waveguide having a first core layer formed of a first semiconductor layer;
A second semiconductor layer that embeds a side surface of the first semiconductor optical waveguide;
An electrode formed on the first semiconductor optical waveguide;
A second semiconductor optical waveguide having a second core layer formed by a third semiconductor layer having a shorter bandgap wavelength than the first semiconductor layer and optically connected to the first semiconductor optical waveguide; Having a waveguide;
The second semiconductor layer embeds a part of a side surface of the second semiconductor optical waveguide beyond a boundary between the first semiconductor optical waveguide and the second semiconductor optical waveguide;
And, and extending between the ends of the second semiconductor layer in which the electrodes are embedded side of the said boundary second semiconductor optical waveguide, an end of the electrode from the end and the boundary of the second semiconductor layer An optical semiconductor device in which parts are separated . The optical semiconductor device according to claim 1,
The optical semiconductor device, wherein the second semiconductor layer extends from the boundary to a range of 5 μm to 15 μm. The optical semiconductor device according to claim 1 or 2,
The optical semiconductor device, wherein the second semiconductor layer is one of a semi-insulating semiconductor layer and a pn buried layer. The optical semiconductor device according to any one of claims 1 to 3,
An optical semiconductor device, wherein a width of the second semiconductor optical waveguide gradually increases toward the boundary. The optical semiconductor device according to claim 1,
The optical semiconductor device, wherein the first optical semiconductor waveguide is an optical semiconductor element selected from the group consisting of a semiconductor optical amplifier, a semiconductor laser device, an optical modulator, and a phase modulator. A first semiconductor optical waveguide having a first core layer formed by the first semiconductor layer;
A second semiconductor optical waveguide having a second core layer formed by a second semiconductor layer having a shorter bandgap wavelength than the first semiconductor layer and optically connected to the first semiconductor optical waveguide; Forming a waveguide,
Said third semiconductor layer to embed the side surfaces of the first semiconductor optical waveguide, the first semiconductor optical waveguide and the second semiconductor optical waveguides semiconductor optical waveguide path side of the second beyond the boundaries of the Form to embed a part,
And extending between said third end of the semiconductor layer to embed a portion of the boundary with the previous SL side surface of the second semiconductor optical waveguide path, the electrode from the end and the boundary of the third semiconductor layer in so that you spacing the ends, a method of manufacturing an optical semiconductor light device for forming an electrode for performing the injection of current into the first core layer.

Images