JP2010080757A - Semiconductor light emitting element - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a semiconductor light emitting element wherein the occurrence of crystal defects is prevented and TM mode oscillation within a range free from degradation in characteristics is possible. <P>SOLUTION: An n-type AlGaAs clad layer 2, an n-type AlGaAs beam diffusion layer 3, an n-type AlGaAs light guide layer 4, an MQW active layer 5, a p-type AlGaAs light guide layer 6, a p-type AlGaAs beam diffusion layer 7, a p-type AlGaAs first clad layer 8, an InGaP etching stop layer 9, a p-type AlGaAs second clad layer 10, and a p-type AlGaAs contact layer 11 are formed on a GaAs substrate 1. The MQW active layer 5 includes a multiple quantum well structure constituted of barrier layers and well layers having a tensile strain, and a layer thickness of each well layer is larger than that of each barrier layer, and the layer thickness of at least one well layer is made different from those of the other well layers. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a semiconductor light emitting device including an active layer having a well layer having tensile strain.

 Among semiconductor light emitting devices, semiconductor lasers are used as light sources for DVDs and laser printers, for example. For example, in a semiconductor laser having a wavelength of 780 nm in the infrared wavelength region, AlGaAs has been conventionally used as an active layer material, and polarized light oscillated in a TE mode. In TE mode oscillation, since an electric field exists in a direction perpendicular to the propagation direction, the radiation window cannot be reduced in order to narrow down the emitted light of the semiconductor laser. If the radiation window is made small, the electric field component is blocked by the radiation window, and the laser light is not sufficiently propagated.

In order to sufficiently propagate the electric field component of the laser light even if the radiation window is made small, laser light by TM mode oscillation is generated and the electric field component is given in the light propagation direction. In order to oscillate in the TM mode, it is necessary to use a well layer having a tensile strain in the active layer. For this purpose, the well layer and the barrier layer (barrier layer) are adjusted to apply a tensile strain to the well layer. (For example, refer to Patent Document 1).
JP-A-7-31465

 The greater the tensile strain applied to the well layer in the active layer, the greater the TM mode oscillation. Therefore, in order to increase the amount of strain, materials having greatly different lattice constants may be used for the well layer and the barrier layer. However, if too much strain is applied, the lattice cannot be maintained and relaxed (lattice relaxation). At this time, crystal defects occur, and the characteristics deteriorate. Crystal defects are more likely to occur as the amount of strain increases and the film thickness increases. It is necessary to prevent the occurrence of crystal defects and to enable TM mode oscillation under conditions where the characteristics do not deteriorate.

  The present invention was devised to solve the above-described problems, and provides a semiconductor light emitting device capable of preventing occurrence of crystal defects and performing TM mode oscillation within a range in which characteristics are not deteriorated. It is aimed.

  In order to achieve the above object, an invention according to claim 1 is provided with an active layer having a multiple quantum well structure composed of a barrier layer and a well layer having tensile strain, and each film thickness of the well layer is a barrier layer. The semiconductor light emitting device is characterized in that the thickness of at least one of the well layers is different from those of the other well layers.

  The invention according to claim 2 is the semiconductor light emitting device according to claim 1, wherein each film thickness of the well layer is larger than 6 nm.

  According to a third aspect of the present invention, the active layer forms part of a stacked body stacked on a semiconductor substrate, and the lattice constant of the well layer is smaller than the lattice constant of the semiconductor substrate. The semiconductor light-emitting device according to claim 1 or 2.

  The invention according to claim 4 is the semiconductor light emitting device according to claim 3, wherein a difference in lattice constant between the well layer and the semiconductor substrate is 0.18% or more.

  The invention according to claim 5 is the semiconductor light-emitting element according to claim 3 or 4, wherein the electrode formed on the laminate is configured in a stripe shape.

  A sixth aspect of the present invention is the semiconductor light emitting element according to the third or fourth aspect, wherein a part of the laminate is formed in a ridge structure.

  The invention according to claim 7 is characterized in that a part of the cladding layer formed on the etching stop layer above the active layer is included in the ridge structure. It is a semiconductor light emitting device.

  According to an eighth aspect of the present invention, the cladding layer is divided into a first impurity concentration region and a second impurity concentration region from the side close to the etching stop layer toward the stacking direction. 8. The semiconductor light emitting element according to claim 7, wherein the impurity concentration of the concentration region is lower than that of the second impurity concentration region.

  According to a ninth aspect of the present invention, the bottom surface of the ridge structure is formed on the first impurity concentration region side, and the first impurity concentration region side from the boundary between the first impurity concentration region and the second impurity concentration region. 9. The semiconductor light emitting device according to claim 8, wherein the semiconductor light emitting device is formed in a range up to a depth of 100 nm.

According to a tenth aspect of the present invention, the impurity concentration of the first impurity concentration region ranges from 5 × 10 ¹⁷ cm ⁻³ to 2 × 10 ¹⁸ cm ⁻³ , or the second impurity concentration region and the first impurity. 10. The semiconductor light-emitting element according to claim 8, wherein a concentration difference d with respect to the concentration region satisfies any of 0 <d ≦ 1 × 10 ² (cm ⁻³ ).

  The semiconductor light emitting device of the present invention includes an active layer having a quantum well structure, and this active layer is composed of a barrier layer and a well layer having tensile strain. Further, the well layer thickness is larger than the barrier layer thickness, and at least one of the plurality of well layers is configured to be different from other well layers. As a result, the threshold current of the TM mode wave becomes lower than the threshold current of the TE mode wave, so that a semiconductor light emitting device having a large TM mode oscillation can be obtained.

 Hereinafter, an embodiment of the present invention will be described with reference to the drawings. A configuration example of a semiconductor laser as a semiconductor light emitting device of the present invention is shown in FIG. On a GaAs substrate 1, an n-type AlGaAs cladding layer 2, an n-type AlGaAs beam diffusion layer 3, an n-type AlGaAs light guide layer 4, an MQW active layer 5, a p-type AlGaAs light guide layer 6, a p-type AlGaAs beam diffusion layer 7, A p-type AlGaAs first cladding layer 8, an InGaP etching stop layer 9, a p-type AlGaAs second cladding layer 10, a p-type GaAs contact layer 11, and a p-electrode 12 are laminated, and an n-electrode 13 is formed on the back side of the GaAs substrate 1. Has been. The GaAs substrate 1 is an inclined n-type GaAs substrate. Further, the p electrode 12 is made of a Ti / Au multilayer metal film or the like, and the n electrode 1 is made of an Au / Ge / Ni alloy layer, Ti / Au multilayer metal film or the like.

 As shown in FIG. 3, the MQW active layer 5 has a multi quantum well structure in which GaAsP well layers 5a and AlGaAs barrier layers 5b are alternately stacked. On the n-type AlGaAs light guide layer 4, a GaAsP well layer 5a is first formed, and then an AlGaAs barrier layer 5b is formed, and the well layers and the barrier layers are alternately stacked. Finally, a GaAsP well layer 5a is formed, and a p-type AlGaAs light guide layer 6 is formed thereon. The GaAsP well layer 5a of the MQW active layer 5 is formed so as to generate tensile strain in the lateral direction (a direction perpendicular to the stacking direction) as indicated by arrows in the figure.

 The thickness of each GaAsP well layer 5a in the MQW active layer 5 is larger than the thickness of any AlGaAs barrier layer 5b. Furthermore, the GaAsP well layers 5a are not all configured to have the same thickness, but the thickness of at least one GaAsP well layer is configured to be different from the thickness of other GaAsP well layers. Yes.

 The semiconductor laser of FIG. 1 has a ridge structure in which a p-type AlGaAs second cladding layer 10 and a p-type GaAs contact layer 11 constitute a striped ridge portion. The current flows through the striped ridge portion and the lower portion of the ridge portion.

  The manufacturing method of the semiconductor laser shown in FIG. 1 will be briefly described as follows by a known MOCVD method, a photolithography technique or the like. An n-type AlGaAs cladding layer 2, an n-type AlGaAs beam diffusion layer 3, an n-type AlGaAs light guide layer 4 and an MQW active layer are formed on the GaAs substrate 1 by crystal growth using MOCVD (metal organic chemical vapor deposition). 5, p-type AlGaAs light guide layer 6, p-type AlGaAs beam diffusion layer 7, p-type AlGaAs first cladding layer 8, InGaP etching stop layer 9, p-type AlGaAs second cladding layer 10, and p-type GaAs contact layer 11 in this order. Laminate. As described above, the MQW active layer 5 is formed by alternately laminating the GaAsP well layers 5a and the AlGaAs barrier layers 5b. In this embodiment, three GaAsP well layers 5a and two AlGaAs barrier layers 5b are stacked.

Next, using the striped SiO ₂ as a mask, the p-type GaAs contact layer 11 and the p-type AlGaAs second cladding layer 10 are etched by dry etching to form a ridge portion. Next, wet etching is performed with hydrochloric acid or dilute sulfuric acid and hydrogen peroxide solution, and etching is performed until the InGaP etching stop layer 9 is reached. The ridge etching is automatically stopped by the etching stop layer 9, and the ridge can be formed with good control.

Thereafter, the SiO ₂ mask is removed by HF treatment. Finally, the wafer is thinned to a predetermined thickness by lapping and polishing, and the p electrode 12 and the n electrode 13 are formed by vapor deposition or sputtering.

  Although FIG. 1 shows an example of a semiconductor laser having a ridge structure, as shown in FIG. 2, a semiconductor laser having a p-electrode in a stripe shape can be used without using a stripe ridge structure. The same reference numerals as those in FIG. 1 denote the same configurations as those in FIG. An n-type AlGaAs cladding layer 2, an n-type AlGaAs beam diffusion layer 3, an n-type AlGaAs light guide layer 4, an MQW active layer 5, a p-type AlGaAs light guide layer 6 and a p-type AlGaAs beam diffusion layer 7 are formed on a GaAs substrate 1. The process up to the point of lamination is the same as in FIG. The difference from FIG. 1 is that there is no etching stop layer, and a p-type AlGaAs cladding layer 18 and a p-type GaAs contact layer 19 are sequentially laminated on the p-type AlGaAs beam diffusion layer 7. Further, the p-electrode 20 is formed in a stripe shape without forming the ridge structure of FIG. 1 so that a current flows through the central portion of the element.

 Here, the GaAsP well layer 5a of the MQW active layer 5 and the AlGaAs barrier layer 5b have different lattice constants of the constituent materials as shown in FIG. 8 to be described later. Therefore, tensile strain is generated in the well layer 5a. However, TM mode oscillation cannot be sufficiently obtained simply by generating tensile strain. This is because even if tensile strain is applied to the well layer, both TM mode and TE mode waves are mixed in the laser light. In order to make the TM mode oscillation superior, it is necessary to configure the TM mode wave so that the threshold current is sufficiently low.

  Therefore, the conditions such that the TM mode oscillation becomes dominant by changing the composition of the well layer so as to satisfy the TM mode oscillation condition while keeping the composition of the active layer constant, and the like were examined by simulation. For the simulation, a semiconductor laser having the structure of FIGS. 1 and 3 was used, and specifically, as follows.

The n-type AlGaAs cladding layer 2 has an n-type Al _0.53 GaAs thickness of 3.5 μm, the n-type AlGaAs beam diffusion layer 3 has an n-type Al _0.53 GaAs thickness of 0.05 μm, and an n-type AlGaAs light guide layer 4. Is 0.08 μm thick n-type Al _0.6 GaAs, p-type AlGaAs optical guide layer 6 is 0.08 μm thick p-type Al _0.6 GaAs, and p-type AlGaAs beam diffusion layer 7 is 0.05 μm thick. P-type Al _0.53 GaAs, p-type AlGaAs first cladding layer 8 is 0.03 μm thick p-type Al _0.53 GaAs, and InGaP etching stop layer 9 is 0.015 μm thick InGaP, p-type AlGaAs first. The 2 clad layer 10 was made of p-type Al _0.53 GaAs having a thickness of 1.6 μm, and the p-type GaAs contact layer 11 was made of p-type GaAs having a thickness of 0.4 μm. In any semiconductor layer, Si was used for n-type impurities and Mg for p-type impurities. The ridge width of the ridge structure (the width of the bottom surface of the ridge shape) was 1.5 μm.

The MQW active layer 5 has a multiple quantum well structure in which three GaAsP well layers 5a and two AlGaAs barrier layers 5b are alternately stacked. The AlGaAs barrier layer 5b was made of undoped Al _0.3 GaAs having a thickness of 5 nm. On the other hand, the GaAsP well layer 5a is composed of undoped GaAs _1-X P _X (0 <X <1). In this embodiment, GaAs _0.9 P _{0.1 with} X = _0.1 is used, and the film thickness varies. The state of TM mode oscillation was examined.

  FIG. 4 shows simulation results when the thicknesses of the three GaAsP well layers 5a are changed. The horizontal axis of the figure indicates the film thickness configuration (nm) of the three well layers, and the vertical axis indicates the relative threshold current (arbitrary unit). Here, the relative threshold current indicates a numerical value calculated by (TM mode oscillation threshold current) / (TE mode oscillation threshold current). Even when tensile strain is applied to the well layer, both the TM mode and the TE mode waves are mixed in the laser light, and therefore the relative threshold current is indicated. That is, it can be seen that TM mode oscillation becomes more remarkable as the relative threshold current is smaller.

 For example, when 8/9/10 nm is written on the horizontal axis of the figure, the numeral 8 nm on the left side indicates the film thickness of the GaAsP well layer 5a closest to the p-type AlGaAs light guide layer 6, and then in order. This corresponds to the thickness of the well layer 5a close to the n-type layer side. Therefore, the well layer 5a closest to the n-type AlGaAs optical guide layer 4 has a thickness of 10 nm.

 In the prior art, since the thicknesses of the three well layers are all the same, the case of 9/9/9 nm in FIG. 4 is considered as a reference. When the film thickness is changed even in one of the three well layers, the relative threshold current changes. In the film thickness configurations of 8/8/11 nm, 8/11/8 nm, 7/9/11 nm, and 11/9/7 nm, the relative threshold current is larger than that in the case of 9/9/9 nm. In this case, the difference in film thickness of the well layers is 2 nm or 3 nm. On the other hand, in the film thickness configuration of 8/9/10 nm and 10/9/8 nm, the relative threshold current is smaller than that in the case of 9/9/9 nm. In this case, the difference in film thickness of the well layers is 1 nm. Further, in the film thickness configuration of 8.5 / 9 / 9.5 nm and 9.5 / 9 / 8.5 nm, the relative threshold current is considerably smaller than that in the case of 9/9/9 nm. In this case, the difference in film thickness of the well layers is 0.5 nm.

 From the above, when the thickness of at least one of the well layers constituting the active layer is different from the other well layers, the difference in thickness between the well layers having different thicknesses and the other well layers is less than 2 nm. If so, it can be seen that the relative threshold current is smaller than when all three well layers have the same thickness. More specifically, the relative threshold current is 9/9/9 nm in the structure in which the width of change between the two well layers arranged at the closest positions is 2 nm or 3 nm. growing. On the other hand, when the width of change between the two well layers arranged at the closest position is less than 2 nm, the relative threshold current is smaller than when the thicknesses of the three well layers are all the same. I understand.

   In FIG. 4, the width of the change in film thickness between the two well layers arranged at the closest position is equal. FIG. 5 shows the result including an example in which these are irregular. As shown in the region surrounded by the frame of the figure, 8.5 / 8.8 / 9.7 nm, 9.7 / 8.8 / 8.5 nm, 8.3 / 9.2 / 9.5 nm, 9 In the film thickness configuration of .5 / 9.2 / 8.3 nm, the relative threshold current is smaller than that in the case of 9/9/9 nm. The film thickness differences between the two well layers arranged at the closest positions are 0.3 / 0.9 nm, 0.9 / 0.3 nm, 0.9 / 0.3 nm, and 0.3 / 0.9 nm, respectively. It has become. That is, it can be seen that the TM mode oscillation becomes remarkable when the change width of the film thickness between the two well layers formed at the closest positions is less than 2 nm and under severe conditions is 1 nm or less. Similar conclusions can be drawn.

Next, the active layer 5 was configured in the same manner as described above, and the thicknesses of the three GaAsP well layers 5a were changed while keeping the same thickness. That is, the AlGaAs barrier layer 5b is made of undoped Al _0.3 GaAs having a thickness of 5 nm, and the GaAsP well layer 5a is made of undoped GaAs _0.9 P _0.1 (X = 0.1). The relationship between the thickness of the well layer and the threshold current is shown in FIG. The black rhombus (♦) data shown in the figure indicates the TE mode, and the black square indicates the TM mode. FIG. 7 shows the above-described relative threshold current by taking the ratio of the threshold currents related to the TM mode and the TE mode of FIG.

 As can be seen from FIG. 7, when the thickness of the well layer is 6 nm or less, the relative threshold current becomes large, and there is a possibility that a mixed wave of the TE mode and the TM mode is generated. On the other hand, when the film thickness of the well layer exceeds 6 nm, as shown in FIG. 6, the TE mode threshold current becomes very high, and the TE mode oscillation is prevented. As a result, as shown in FIG. 7, it can be seen that the relative threshold current becomes small and TM mode oscillation becomes remarkable.

Further, the lattice constant difference was calculated when the composition Y was changed between GaAs _{YP 1-Y} and Al _1-Y Ga _Y As. The GaAsP well layer 5a was formed of GaAs _Y P _1-Y having a thickness of 9 nm, and the composition Y was changed using the GaAs substrate 1 as the Al _{1 -Y} Ga _Y As substrate. FIG. 8A shows the change in the lattice constant when the composition Y is changed in GaAs _{YP 1-Y} by L1 (solid line), and the change in the composition Y of Al _1-Y Ga _Y As is shown. The lattice constant is represented by L2 (dotted line). Here, the lattice constant of GaAs is 5.65325, the lattice constant of AlAs is 5.6611, and the lattice constant of GaP is 5.4505. In the meantime, it was calculated as a linear change.

As the composition Y increases, the lattice constant of GaAs _{YP 1-Y} increases linearly, but there is almost no change in the lattice constant of Al _1-Y Ga _Y As. FIG. 8B shows the relationship between the value of 100 × (L1−L2) / L1 (lattice constant difference) and the composition X of GaAs _1−X P _X at this time. The vertical axis represents the lattice constant difference (%), and the horizontal axis represents the composition X of GaAs _1-X P _X. Here, the composition X and the composition Y have a relationship of X = Y−1. As the composition X increases, the lattice constant difference from the GaAs substrate or AlGaAs semiconductor layer increases. An increase in the lattice constant difference means that the amount of tensile strain increases.

Here, as shown in FIG. 9, when looking at the relationship between tensile strain, compression strain, and threshold current density, it is found that tensile strain has a smaller operating range at a lower threshold current density than compression strain. Recognize. This is because, if too much strain is applied, crystal defects are generated and the characteristics deteriorate. Therefore, it is necessary to examine how the threshold current of the TM mode changes depending on the composition X of the P (phosphorus) component of GaAs _1-X P _X that is greatly related to the amount of tensile strain. Figure 10 shows the relationship between the _{GaAs 1-X P X} of P composition X and the threshold current. As in FIG. 6, the black diamond (♦) data indicates the TE mode, and the black square indicates the TM mode. FIG. 11 shows the above-described relative threshold current by taking the ratio of the threshold currents related to the TM mode and the TE mode of FIG. Note that the data indicated by X1 is difficult to discriminate from the figure, but the TE mode and TM mode threshold currents are the same and indicate overlapping data.

 As can be seen from FIG. 10, when the P composition X is 0.04 or less, the TE mode and the TM mode have almost the same threshold current, and the TE mode wave and the TM mode wave are mixed. However, when the P composition X exceeds 0.04, the threshold current in the TM mode becomes smaller, and the relative threshold current becomes smaller than 1 as can be seen from FIG. From the data shown in FIGS. 10 and 11, the TM mode becomes prominent particularly when the P composition X is 0.05 or more. Here, when compared with FIG. 8B, the lattice constant difference when the P composition X is 0.05 is 0.18%. On the other hand, the amount of tensile strain has a critical point until lattice relaxation, and the critical lattice constant difference at that time corresponds to about 0.4, so that the size of P composition X is limited to about 13%. it is conceivable that. Note that if the P composition X is different, the tensile strain also changes and the critical film thickness also changes.

 Next, conditions for achieving TM mode oscillation in the ridge structure of FIG. 1 will be considered from a viewpoint other than the film thickness, lattice constant, and the like. In FIG. 12, the configuration of the semiconductor layer is almost the same as that in FIG. The same reference numerals as those in FIG. 1 denote the same configurations.

 On a GaAs substrate 1, an n-type AlGaAs cladding layer 2, an n-type AlGaAs beam diffusion layer 3, an n-type AlGaAs light guide layer 4, an MQW active layer 5, a p-type AlGaAs light guide layer 6, a p-type AlGaAs beam diffusion layer 7, A p-type AlGaAs first cladding layer 8, an InGaP etching stop layer 9, a p-type AlGaAs second cladding layer 100, a p-type GaAs contact layer 11, and a p-electrode 12 are stacked, and an n-electrode 13 is formed on the back side of the GaAs substrate 1. Has been. The semiconductor laser of FIG. 12 forms a ridge structure in which a part of the p-type AlGaAs second cladding layer 100 and the p-type GaAs contact layer 11 form a striped ridge portion. The difference from FIG. 1 is that the p-type AlGaAs second cladding layer 100 is composed of a striped ridge-shaped portion 100A and a flat plate-shaped portion 100B. D2 indicated by a broken line in the figure indicates the boundary between the ridge-shaped portion 100A and the flat plate-shaped portion 100B, that is, the bottom surface of the ridge structure.

This can be configured by stopping the etching stop position in the middle of the p-type AlGaAs second cladding layer 100 closer to the p-type GaAs contact layer 11 than the InGaP etching stop layer 9. Thereby, the influence of the distortion from the ridge on the active layer and the influence of light leakage can be reduced. Here, the MQW active layer 5 has a multi-quantum well structure in which three GaAsP well layers 5a and two AlGaAs barrier layers 5b are alternately stacked as in the above example. The AlGaAs barrier layer 5b was made of undoped Al _0.3 GaAs having a thickness of 5 nm. On the other hand, the GaAsP well layer 5a is made of undoped GaAs _0.9 P _0.1 (X = 0.1). By using GaAs _0.9 P _0.1 for the well layer 5a, tensile strain is generated in the well layer 5a. Further, the material composition and film thickness of the other semiconductor layers are the same as those described in FIG. As shown in FIG. 12, in order to stop the etching at the position D2, reactive ion etching (RIE) or the like may be used.

 The p-type AlGaAs second cladding layer 100 is formed of two regions, a low-concentration region (first impurity concentration region) and a high-concentration region (second impurity concentration region) having different p-type impurity concentrations. N shown in FIG. 12 represents a boundary between a low concentration region with high impurity concentration and a high concentration region. The upper side (contact layer 11 side) of the impurity concentration boundary N is the high concentration region, and the lower side of the boundary N (MQW active layer 5 side) is the low concentration region. Further, D2 which is the bottom surface of the ridge structure is formed on the lower concentration region side than the boundary N of the impurity concentration.

The interface position between the InGaP etching stop layer 9 and the p-type AlGaAs second cladding layer 100 is D1, and the interface position between the p-type AlGaAs second cladding layer 100 and the p-type GaAs contact layer 11 is D3. The p-type impurity concentration in the high concentration region (second impurity concentration region) is 4 × 10 ¹⁹ cm ^−3, and the p-type impurity concentration in the low concentration region (first impurity concentration region) is 1 × 10 ¹⁸ cm ^−3. It was. The ridge width (width at the D2 position) in FIG. 12 was 1.5 μm. FIG. 13 shows a simulation of the relationship between the boundary N and the boundary D2 by fixing the position of the boundary N at a predetermined position and changing the etching depth.

 The vertical axis in FIG. 13 represents the threshold current (mA), and the horizontal axis represents the distance (μm) from the active layer. The distance from the active layer represents the distance from the interface between the MQW active layer 5 and the p-type AlGaAs light guide layer 6. As shown in the figure, the boundary N is formed at about 0.7 μm from the active layer. Here, when the boundary D2 is changed by etching, the threshold current in the TE mode and the TM mode decreases as the boundary D2 approaches the MQW active layer 5.

 However, since there is no difference between the threshold currents of the TE mode and the TM mode, the possibility that a mixed wave of the TE mode wave and the TM mode wave will be increased. On the other hand, from the figure, within 100 nm from the boundary N to the active layer side, the TM mode threshold current is sufficiently smaller than the TE mode threshold current. Therefore, the bottom surface position D2 of the ridge structure is formed on the low concentration region side rather than the boundary N of the impurity concentration, and the distance difference in the depth direction between N and D1 is set within 100 nm. TE mode oscillation can be made remarkable.

Next, FIG. 14 shows a simulation result in the case where the boundary D2 is set to the same position as the boundary N and the impurity concentration in the low concentration region is changed from 1 × 10 ¹⁹ cm ⁻³ . The high concentration region was fixed at 4 × 10 ¹⁹ cm ⁻³ described above. The vertical axis in FIG. 14 represents the threshold current (mA), and the horizontal axis represents the concentration (cm ⁻³ ) in the low concentration region of the p-type AlGaAs second cladding layer 100. As shown in FIG. 14, when the impurity concentration in the low-concentration region exceeds 2 × 10 ¹⁸ cm ⁻³ , the difference in threshold current between the TE mode and the TM mode becomes small. Therefore, in order to make the TM mode wave more prominent, it is desirable to set it to 2 × 10 ¹⁸ cm ⁻³ or less. The lower limit is 5 × 10 ¹⁷ cm ⁻³ from the figure. From the above, the preferable range of the low concentration region is 5 × 10 ¹⁷ cm ⁻³ or more and 2 × 10 ¹⁸ cm ⁻³ or less.

On the other hand, the range of the high concentration region needs to be higher than that of the low concentration region, and therefore the upper limit is considered to be 5 × 10 ¹⁹ cm ⁻³ at a concentration exceeding at least 2 × 10 ¹⁸ cm ⁻³ . From the above, considering the density difference between the high density region and the low density region, the allowable range of the density difference d is 0 <d ≦ 1 × 10 ² (cm ⁻³ ). Therefore, for the low concentration region, the impurity concentration is 5 × 10 ¹⁷ cm ⁻³ to 2 × 10 ¹⁸ cm ⁻³ , or the concentration difference d is 0 <d ≦ 1 × 10 ² (cm ⁻³ ). It is preferable to satisfy any of the following.

It is a figure which shows an example of the cross-sectional structure of the semiconductor light-emitting device of this invention. It is a figure which shows an example of the cross-sectional structure of the semiconductor light-emitting device of this invention. It is a figure which shows the structural example of the active layer of the semiconductor light-emitting device of this invention. It is a figure which shows the relationship between the film thickness structure of the well layer in an active layer, and a relative threshold current. It is a figure which shows the relationship between the film thickness structure of the well layer in an active layer, and a relative threshold current. It is a figure which shows the relationship between a well layer film thickness and a threshold current. It is a figure which shows the relationship between a well layer film thickness and a relative threshold current. It is a figure which shows the relationship between a lattice constant and a composition of AlGaAs and GaAsP. It is a figure which shows the relationship between tensile strain and compression strain, and a threshold current density. It is a figure which shows the relationship between P composition X of GaAsP and a threshold current. It is a figure which shows the relationship between P composition X of GaAsP and a relative threshold current. It is a figure which shows the other example of the cross-sectional structure of the semiconductor light-emitting device which has a ridge structure. In the configuration of FIG. 12, when there is a change in impurity concentration in the second cladding layer, it is a diagram showing the relationship between the distance from the active layer at the impurity concentration change position and the threshold current. It is a figure which shows the relationship between the impurity concentration and threshold current in the low concentration area | region of a 2nd clad layer.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 GaAs substrate 2 n-type AlGaAs cladding layer 3 n-type AlGaAs beam diffusion layer 4 n-type AlGaAs light guide layer 5 MQW active layer 5a GaAsP well layer 5b AlGaAs barrier layer 6 p-type AlGaAs light guide layer 7 p-type AlGaAs beam diffusion layer 8 p-type AlGaAs first cladding layer 9 InGaP etching stop layer 10 p-type AlGaAs second cladding layer 11 p-type GaAs contact layer 12 p-electrode 13 n-electrode 18 p-type AlGaAs cladding layer 19 p-type GaAs contact layer 20 p-electrode 21 n-electrode 100 p-type AlGaAs second cladding layer

Claims (10)

  An active layer having a multiple quantum well structure composed of a barrier layer and a well layer having a tensile strain, wherein each film thickness of the well layer is larger than each film thickness of the barrier layer, and at least one of the well layers A semiconductor light-emitting element, wherein the thickness of the layer is different from that of other well layers.   2. The semiconductor light emitting device according to claim 1, wherein each thickness of the well layer is larger than 6 nm.   The active layer forms part of a stacked body stacked on a semiconductor substrate, and the lattice constant of the well layer is smaller than the lattice constant of the semiconductor substrate. The semiconductor light emitting element as described.   4. The semiconductor light emitting element according to claim 3, wherein a difference in lattice constant between the well layer and the semiconductor substrate is 0.18% or more.   5. The semiconductor light emitting element according to claim 3, wherein the electrode formed on the laminate is configured in a stripe shape. 6.   5. The semiconductor light emitting element according to claim 3, wherein a part of the stacked body is formed in a ridge structure.  The semiconductor light emitting device according to claim 6, wherein a part of the cladding layer formed on the etching stop layer above the active layer is included in the ridge structure.   The cladding layer is divided into a first impurity concentration region and a second impurity concentration region from the side close to the etching stop layer toward the stacking direction, and the impurity concentration of the first impurity concentration region is the second impurity concentration. The semiconductor light emitting device according to claim 7, wherein the semiconductor light emitting device is lower than the region.   The bottom surface of the ridge structure is formed on the first impurity concentration region side and in a range from the boundary between the first impurity concentration region and the second impurity concentration region to a depth of 100 nm on the first impurity concentration region side. The semiconductor light emitting device according to claim 8, wherein The impurity concentration of the first impurity concentration region is in the range of 5 × 10 ¹⁷ cm ⁻³ to 2 × 10 ¹⁸ cm ⁻³ , or the concentration difference d between the second impurity concentration region and the first impurity concentration region is 0 < The semiconductor light emitting device according to claim 8, wherein any one of d ≦ 1 × 10 ² (cm ⁻³ ) is satisfied.

Images