JPWO2015052861A1 - Semiconductor light emitting device - Google Patents

Abstract

In a semiconductor light emitting device having a ridge, an object is to achieve both a lateral light confinement effect and a waveguide loss suppression effect. The semiconductor light emitting device has a first guide layer (313) between the first clad layer (12) and the active layer (13), the second clad layer (14) has a ridge (14A), When the film thickness of the first cladding layer (12) is Tnc and the film thickness of the first guide layer (313) is Tng, Tng [μm] ≦ −0.98 Tnc [μm] +0.76 [μm ] Tng [μm] ≧ −0.98 Tnc [μm] +0.56 [μm] Two conditions were satisfied, and the emission wavelength was 500 to 560 nm.

Description

  The present disclosure relates to a semiconductor light emitting device used as a light source necessary for an illumination light source, an in-vehicle light source, a display light source, other electronic devices, information processing devices, and the like.

  Conventionally, as shown in FIG. 17, this type of semiconductor light emitting device includes a substrate 101 made of GaN, a first clad layer 102 made of n-type InAlGaN provided above the substrate 101, and the first clad layer 102. A second clad layer 103 made of n-type GaN provided above the clad layer 102 and a third clad layer 104 made of n-type InGaN provided above the second clad layer 103 are provided. Further, an active layer 105 containing InGaN provided above the third cladding layer 104, a fourth cladding layer 106 made of InGaN provided above the active layer 105, and the fourth cladding layer 106 And a fifth clad layer 107 made of p-type InAlGaN provided above.

  As prior art document information relating to this application, for example, Patent Document 1 is known.

JP 2012-248575 A

  In such a conventional semiconductor light emitting device, it is difficult to achieve both a lateral light confinement effect and a waveguide loss suppression effect in a laser having an oscillation wavelength of 500 nm to 560 nm. The reason for this will be described first.

  FIG. 2 shows a general basic structure of a semiconductor laser. In this structure, an n-type cladding layer 2, an active layer 3, a p-type cladding layer 4, and a p-type contact layer 6 are sequentially formed on an n-type substrate 1. A ridge type waveguide is formed in the p-type cladding layer 4. On the ridge side wall, a current blocking layer 5 having a refractive index lower than that of the p-type cladding layer 4 and a band gap energy larger than that of laser oscillation light is formed. Here, the direction parallel to the substrate surface in the resonator plane of the semiconductor laser is called the horizontal direction, and the normal direction of the substrate is called the vertical direction.

  In order to cause laser oscillation in a semiconductor laser, it is necessary to confine light in the active layer 3 and confine electrons and holes injected from the n-type and p-type cladding layers. By doing so, as a result of the inversion distribution state of carriers in the active layer, stimulated emission occurs, and the light confined in the active layer is amplified, and laser oscillation can be generated. In this case, the more light that is trapped in the active layer, the greater the amplification effect, and the smaller the injection current amount (oscillation threshold current value) required for laser oscillation. The ratio of the light confined in the active layer to the entire light distribution is called a light confinement coefficient. The larger this coefficient, the smaller the oscillation threshold current value can be made.

  In order to increase the optical confinement factor, it is effective to increase the refractive index difference between the cladding layer and the active layer. In the nitride laser, in order to obtain laser oscillations in the 405 nm band (blue purple), 445 nm band (blue), and 530 nm band (green), the InGaN used for the active layer (well layer in the case of a quantum well active layer) The In composition must be about 0.06, 0.15, and 0.3, respectively. In InGaN, when the In composition is increased, the band gap energy is decreased, and the oscillation wavelength can be changed according to the In composition. Therefore, in order to increase the optical confinement coefficient in a laser having a wavelength of 530 nm, a cladding layer material having a large refractive index difference from InGaN having an In composition of 0.3 is required.

Conventionally, AlGaN has been widely used as a material for the cladding layer. This is because the AlGaN material is suitable for the cladding layer material because the crystal growth is relatively easy and the refractive index is lower than that of the InGaN material. However, in the AlGaN material, when the Al composition is increased, the difference between the thermal expansion coefficient and the lattice constant with the GaN substrate is increased, and lattice defects and cracks are likely to occur. Since the efficiency decreases, the resistance of the p-type AlGaN layer increases. In order to avoid this problem, the Al composition of the AlGaN layer used for the cladding layer must be about 0.1 or less. In order to solve this problem, the In _0.03 Al _0.14 Ga _0.83 N layer was used as the cladding layer in the conventional example shown in FIG. An example is shown. By using the In _0.03 Al _0.14 Ga _0.83 N layer, the generation of the lattice defects can be suppressed.

Consider the difference in refractive index between an AlGaN layer having an Al composition of 0.1, an In _0.03 Al _0.14 Ga _0.83 N layer, and an InGaN layer that is an active layer. Here, the refractive index difference between the AlGaN layer having the Al composition of 0.1 and the InGaN layer is denoted by ΔNA, and the refractive index difference between the In _0.03 Al _0.14 Ga _0.83 N layer and the InGaN layer is denoted by ΔNI. When the wavelength changes, the refractive index of the material also changes accordingly. Therefore, the refractive index difference between the cladding layer and the active layer needs to be estimated for wavelengths of 405 nm, 445 nm, and 530 nm, respectively. The In composition for InGaN for obtaining the above wavelengths needs to have values of about 0.06, 0.15, and 0.3, respectively.

In the 405 nm band, the refractive index of InGaN with an In composition of 0.06 is 2.67, the refractive index of AlGaN with an Al composition of 0.1 is 2.48, and the refractive index of In _0.03 Al _0.14 Ga _0.83 N is In this case, ΔNA is 0.19 and ΔNI is 0.2.

In the 445 nm band, the refractive index of InGaN with an In composition of 0.15 is 2.55, the refractive index of AlGaN with an Al composition of 0.1 is 2.41, and the refractive index of In _0.03 Al _0.14 Ga _0.83 N is In this case, ΔNA is 0.14, and ΔNI is 0.14.

In the 530 nm band, the refractive index of InGaN with an In composition of 0.30 is 2.49, the refractive index of AlGaN with an Al composition of 0.1 is 2.36, and the refractive index of In _0.03 Al _0.14 Ga _0.83 N is In this case, ΔNA is 0.13, and ΔNI is 0.15.

  As described above, it can be seen that when the oscillation wavelength is increased from the 405 nm band to the 445 nm band and the 530 nm band, the refractive index difference between the active layer and the cladding layer is smaller than that in the 405 nm band. Furthermore, since the lattice constant difference with the GaN substrate increases as the In composition increases, lattice defects are likely to occur, so the thickness of the active layer (in the case of a quantum well active layer, the total thickness of the well layers) is increased. It becomes difficult. In order to increase the oscillation wavelength, it is necessary to increase the In composition of the active layer. As a result, when the oscillation wavelength is increased from the 405 nm band to the 445 nm band and the 530 nm band, it is difficult to increase the thickness of the active layer, and the confinement factor is decreased. The decrease in the optical confinement factor leads to an increase in the operating carrier density in the active layer in the oscillation threshold state, so that the overflow of injected carriers from the active layer to the cladding layer during high-temperature operation increases, resulting in a high-temperature operation. This leads to a decrease in the thermal saturation level in the light output. From this, it can be seen that the structure according to the conventional example in the oscillation wavelength band of 530 nm tends to cause a decrease in the thermal saturation level due to a decrease in the refractive index difference between the cladding layer and the active layer.

In order to improve the thermal saturation level, it is effective to reduce the waveguide loss of the light distribution propagating through the waveguide and improve the differential efficiency (slope efficiency) in the current-light output characteristics. Most of the waveguide loss is absorption loss caused by free carrier absorption by impurities doped in the cladding layer. In a nitride material, the doping concentration of impurities in the n-type cladding layer is usually about 5 × 10 ¹⁷ cm ⁻³ , whereas in the p-type cladding layer, it is 10 times the impurity concentration of the n-type cladding layer. A low-resistance p-type layer cannot be obtained unless doping is performed at about 5 × 10 ¹⁸ cm ⁻³ . In a normal nitride semiconductor laser, a p-type contact layer is formed in which the impurity concentration is further increased to about 10 times or more that of the p-type cladding layer in order to reduce the contact resistance with the electrode.

  Free carrier absorption increases as the impurity concentration increases. For this reason, the main component of the waveguide loss is a free carrier absorption component that is subjected to light distribution in the p-type cladding layer and the p-type contact layer. Therefore, in order to reduce the waveguide loss, it is effective to reduce the ratio of the light distribution distributed in the p-type layer.

Incidentally, the vertical light distribution in the region outside the ridge shown in FIG. 2 is slightly biased toward the substrate 1 due to the current blocking layer 5 having a low refractive index as compared with the vertical light distribution in the central portion of the ridge. As a result, since the shape of the light distribution in the vertical direction is different between inside and outside the ridge, an effective refractive index difference (ΔN) is formed in the light distribution inside and outside the ridge with respect to the horizontal direction. ΔN is important for confining the light distribution in the ridge region in the horizontal direction. The refractive index of the active layer varies depending on the injected carriers. For this reason, if ΔN is made too small, the light distribution becomes unstable due to the influence of the refractive index change due to the change in the operating carrier density in the active layer, and nonlinear bending (kink) occurs in the current-light output characteristics. At this time, it becomes difficult not only to lower the thermal saturation level but also to obtain a stable light output, which causes a serious problem in practical use. In order to prevent such instability of operation, ΔN of 3 × 10 ⁻³ or more is required to form a stable horizontal light confinement mechanism even if the refractive index of the active layer changes.

  Here, when the vertical light distribution is closer to the n-type cladding layer 2, the light distribution shape in the vertical direction inside and outside the ridge approaches the same shape, so that ΔN decreases. Conversely, when the vertical light distribution is closer to the p-type cladding layer 4, the vertical light distribution shape inside and outside the ridge becomes different due to the influence of the current blocking layer 5, so ΔN increases. Will do. Therefore, in order to increase ΔN, the light distribution should be close to the p-type cladding layer 4. However, if the light distribution is closer to the p-type cladding layer 4, the light distribution is affected by the p-type contact layer having the highest impurity concentration, and the waveguide loss increases. Furthermore, as described above, in the wavelength 530 nm band, the vertical light distribution tends to spread in the vertical direction, and the waveguide loss is likely to increase as compared with the laser in the wavelength 405 nm band. In order to reduce the waveguide loss, the thickness of the InGaN active layer having a high refractive index may be increased (in the case of a quantum well active layer, the total thickness of the InGaN well layer). Since the difference in lattice constant with the high InGaN layer is large, it cannot be made too thick. In a laser with a wavelength of 530 nm, the In composition of the InGaN layer used for the active layer must be set to about 0.3 as described above, and the total thickness of the InGaN layer having a high In composition of about 0.3 is maximum. However, if the thickness is not about 10 nm, the luminous efficiency is reduced due to the generation of lattice defects due to lattice irregularities. As a result, in the 530 nm band semiconductor laser, the optical confinement coefficient becomes a small value of about 1% or less.

For these reasons, in the nitride laser structure of the conventional structure in the wavelength 530 nm band, even if the confinement factor is small, the magnitude of ΔN of 3 × 10 ⁻³ or more for obtaining a stable high temperature and high output operation is obtained. While obtaining, it was difficult to achieve both low waveguide loss (10 cm ⁻¹ or less).

  Therefore, the present disclosure aims to achieve both a lateral light confinement effect and a waveguide loss suppression effect in a semiconductor light emitting device having a ridge.

In order to achieve this object, a semiconductor light emitting device of the present disclosure includes a substrate, a first refractive index correction layer having a first conductivity provided above the substrate, and a first refractive index correction layer. A first cladding layer provided above the active layer, an active layer provided above the first cladding layer, a second cladding layer having second conductivity provided above the active layer, Is provided. A first guide layer is provided between the first cladding layer and the active layer, and the second cladding layer has a ridge. The substrate is a GaN layer, the first refractive index correction layer includes an InAlN layer, the first cladding layer is a GaN layer, the first guide layer is an InGaN layer, and the active layer is an InGaN layer. When the second cladding layer is an AlGaN layer, the thickness of the first cladding layer is Tnc, and the thickness of the first guide layer is Tng,
Tng [μm] ≦ −0.98 Tnc [μm] +0.76 [μm]
Tng [μm] ≧ −0.98 Tnc [μm] +0.56 [μm]
These two formulas were satisfied and the emission wavelength was 500 to 560 nm.

  By adopting such a configuration, since the refractive index of the first refractive index correction layer is lower than that of the conventional AlGaN layer having an Al composition of 0.1, the active layer is reduced in free carrier absorption loss due to impurity doping. The vertical optical confinement can be increased, and the waveguide loss can be reduced. Furthermore, it is possible to precisely control the vertical spread of the light distribution to the second cladding layer by the film thickness of the first guide layer, and to obtain a desired ΔN. As a result, it is possible to increase the effective refractive index difference between the inside and outside of the ridge while concentrating the light distribution in the low loss region, so that both the lateral optical confinement effect and the waveguide loss suppression effect can be achieved. .

FIG. 1 is a cross-sectional view of the semiconductor light emitting device according to the first embodiment of the present disclosure. FIG. 2 is a diagram showing the relationship between the cross section of a ridge type semiconductor laser and the effective refractive index difference (ΔN) inside and outside the ridge. FIG. 3 shows the film thickness dependency of the first guide layer and the film thickness dependency of the first cladding layer with respect to the waveguide loss, optical confinement coefficient, and effective refractive index difference of the semiconductor light emitting device according to the first embodiment of the present disclosure. FIG. FIG. 4 shows the first guide layer thickness dependency and the first cladding layer thickness dependency on the waveguide loss, optical confinement coefficient, and effective refractive index difference of the semiconductor light emitting device according to the first embodiment of the present disclosure. FIG. FIG. 5 shows the first guide layer thickness dependency and the first cladding layer thickness dependency on the waveguide loss, optical confinement coefficient, and effective refractive index difference of the semiconductor light emitting device according to the first embodiment of the present disclosure. FIG. FIG. 6 shows the thickness dependence of the first guide layer and the thickness dependence of the first cladding layer with respect to the waveguide loss, optical confinement coefficient, and effective refractive index difference of the semiconductor light emitting device in the first embodiment of the present disclosure. FIG. FIG. 7 shows the film thickness dependence of the first guide layer and the film thickness dependence of the first cladding layer with respect to the waveguide loss, optical confinement coefficient, and effective refractive index difference of the semiconductor light emitting device in the first embodiment of the present disclosure. FIG. FIG. 8 shows the thickness dependence of the first guide layer and the thickness dependence of the first cladding layer with respect to the waveguide loss, optical confinement coefficient, and effective refractive index difference of the semiconductor light emitting device according to the first embodiment of the present disclosure. FIG. FIG. 9 is a diagram illustrating the oscillation wavelength dependence of the effective refractive index difference of the semiconductor light emitting device according to the first embodiment of the present disclosure. FIG. 10 is a diagram illustrating the oscillation wavelength dependence of the waveguide loss of the semiconductor light emitting device according to the first embodiment of the present disclosure. FIG. 11 is a diagram illustrating the oscillation wavelength dependence of the optical confinement coefficient of the semiconductor light emitting device according to the first embodiment of the present disclosure. FIG. 12 shows the first guide layer thickness dependency and the first cladding layer thickness dependency on the waveguide loss, optical confinement coefficient, and effective refractive index difference of the semiconductor light emitting device in the first embodiment of the present disclosure. FIG. FIG. 13 shows the film thickness dependence of the first guide layer and the film thickness dependence of the first cladding layer with respect to the waveguide loss, optical confinement coefficient, and effective refractive index difference of the semiconductor light emitting device in the first embodiment of the present disclosure. FIG. FIG. 14 shows the thickness dependence of the first guide layer and the thickness dependence of the first cladding layer with respect to the waveguide loss, optical confinement coefficient, and effective refractive index difference of the semiconductor light emitting device according to the first embodiment of the present disclosure. FIG. FIG. 15 is a diagram illustrating the film thickness dependency and Al composition dependency of the second guide layer with respect to the waveguide loss and the effective refractive index difference of the semiconductor light emitting device according to the first embodiment of the present disclosure. FIG. 16 is a cross-sectional view illustrating another example of the semiconductor light emitting device according to the first embodiment of the present disclosure. FIG. 17 is a cross-sectional view of a conventional semiconductor light emitting device.

(Embodiment 1)
In the present disclosure, the expression that “A” is provided “above” B includes the case where A is provided on B via another member and the case where A is provided on and in contact with B. Including both. The same applies to the expression that A is provided “above” B.

  In the present disclosure, the first conductivity and the second conductivity are different from each other. When the first conductivity is n-type, the second conductivity is p-type, and the first conductivity is p-type. In this case, the second conductivity is n-type.

  Hereinafter, the semiconductor light emitting device in the first embodiment will be described with reference to the drawings. As shown in FIG. 1, the semiconductor light emitting device in the first embodiment includes a substrate 11, a first refractive index correction layer 15 having a first conductivity provided above the substrate 11, and a first refractive index. The first clad layer 12 provided above the correction layer 15, the active layer 13 provided above the first clad layer 12, and the second conductivity provided above the active layer 13 A second cladding layer 14.

  A first guide layer 313 is provided between the first cladding layer 12 and the active layer 13, and the second cladding layer 14 has a ridge 14A.

  The substrate 11 is a GaN layer, the first refractive index correction layer 15 includes an InAlN layer, the first cladding layer 12 is a GaN layer, the first guide layer 313 is an InGaN layer, an active layer Reference numeral 13 denotes an InGaN layer, and the second cladding layer 14 is an AlGaN layer.

  When the film thickness of the first cladding layer 12 is Tnc and the film thickness of the first guide layer 313 is Tng, the following (Expression 1) and (Expression 2) are satisfied.

  The emission wavelength is 500 nm or more and 560 nm or less.

  By adopting such a configuration, it is possible to increase the effective refractive index difference between the inside and outside of the ridge 14A while concentrating the light distribution in a region where the waveguide loss is small. Both suppression effects can be achieved. The reason will be described below.

  3 shows a waveguide loss, an optical confinement coefficient, and the like when the oscillation wavelength of the light emitting device is 530 nm, the In composition of the first guide layer 313 is 0.02, and the second guide layer 315 in FIG. 1 is not provided. The film thickness dependence of the first guide layer 313 and the film thickness dependence of the first cladding layer 12 with respect to the effective refractive index difference are shown.

  Here, the solid line indicates the contour line of the waveguide loss, the alternate long and short dash line indicates the contour line of the optical confinement coefficient, and the broken line indicates the contour line of the effective refractive index difference. The shaded area in FIG. 3 indicates a region that satisfies (Expression 1) and (Expression 2) at the same time.

  4 shows a case where the oscillation wavelength of the light emitting device is 530 nm, the In composition of the first guide layer 313 is 0.04, and the second guide layer 315 in FIG. The film thickness dependence of the first guide layer 313 and the film thickness dependence of the first cladding layer 12 with respect to the refractive index difference are shown.

  Here, the solid line indicates the contour line of the waveguide loss, the alternate long and short dash line indicates the contour line of the optical confinement coefficient, and the broken line indicates the contour line of the effective refractive index difference. The shaded area in FIG. 4 indicates a region that satisfies (Expression 1) and (Expression 2) at the same time.

From FIGS. 3 and 4, the smaller the film thickness Tnc of the first cladding layer 12 and the film thickness Tng of the first guide layer 313, the larger the effective refractive index difference indicated by the broken line. Recognize. And if it is in the range which satisfy | fills said (Formula 1), it turns out that the effective refractive index difference will be ^3x10 < ^-3 > or more in the most range. If the effective refractive index difference is 3 × 10 ⁻³ or more, the transverse mode light can be stably confined in the ridge 14A including the higher order mode. Therefore, by satisfying the above (Equation 1), the effect of lateral light confinement can be improved.

3 and 4, the waveguide loss indicated by the solid line increases as both the film thickness Tnc of the first cladding layer 12 and the film thickness Tng of the first guide layer 313 decrease. I understand. If it is within the range satisfying the above (Equation 2), the generation of the waveguide loss can be suppressed to 10 cm ⁻¹ or less in many of the ranges, and the generation of the waveguide loss in all the ranges is 15 cm. ^-1 or less.

  Therefore, by satisfying the above (Equation 1) and (Equation 2), it is possible to achieve both a lateral light confinement effect and a waveguide loss suppression effect.

  Further, from FIGS. 3 and 4, as both the film thickness Tnc of the first cladding layer 12 and the film thickness Tng of the first guide layer 313 become smaller, the optical confinement coefficient indicated by the alternate long and short dash line increases. I understand that. And if it is in the range which satisfy | fills said (Formula 1), it turns out that a light confinement coefficient will be 1.0% or more in the whole range. Therefore, by satisfying the above (Equation 1), not only the above-described lateral light confinement effect but also the vertical light confinement effect can be obtained more reliably.

  Subsequently, as shown in FIG. 1, a second guide layer 315 having second conductivity is provided between the second cladding layer 14 and the active layer 13, and the second guide layer 315 is formed from the InGaN layer. The structure which becomes will be described.

  FIG. 5 shows a waveguide when the oscillation wavelength of the light emitting device is 530 nm, the In composition of the first guide layer 313 is 0.02, and the thickness of the second guide layer 315 in FIG. 1 is 0.05 μm. The film thickness dependency of the first guide layer 313 and the film thickness dependency of the first cladding layer 12 with respect to loss, optical confinement coefficient, and effective refractive index difference are shown.

  Here, the solid line indicates the contour line of the waveguide loss, the alternate long and short dash line indicates the contour line of the optical confinement coefficient, and the broken line indicates the contour line of the effective refractive index difference. The hatched portion in FIG. 5 indicates a region that satisfies (Expression 3) and (Expression 4) described later simultaneously.

  6 shows the waveguide loss when the oscillation wavelength of the light emitting device is 530 nm, the In composition of the first guide layer 313 is 0.02, and the thickness of the second guide layer 315 in FIG. 1 is 0.1 μm. 4 shows the film thickness dependence of the first guide layer 313 and the film thickness dependence of the first cladding layer 12 with respect to the optical confinement coefficient and the effective refractive index difference.

  Here, the solid line indicates the contour line of the waveguide loss, the alternate long and short dash line indicates the contour line of the optical confinement coefficient, and the broken line indicates the contour line of the effective refractive index difference. The shaded area in FIG. 6 indicates a region that satisfies (Expression 3) and (Expression 4) described later simultaneously.

  7 shows a waveguide when the oscillation wavelength of the light emitting device is 530 nm, the In composition of the first guide layer 313 is 0.04, and the thickness of the second guide layer 315 in FIG. 1 is 0.05 μm. The film thickness dependency of the first guide layer 313 and the film thickness dependency of the first cladding layer 12 with respect to loss, optical confinement coefficient, and effective refractive index difference are shown.

  Here, the solid line indicates the contour line of the waveguide loss, the alternate long and short dash line indicates the contour line of the optical confinement coefficient, and the broken line indicates the contour line of the effective refractive index difference. The shaded area in FIG. 7 indicates a region that satisfies (Expression 3) and (Expression 4) described later simultaneously.

  8 shows a waveguide loss when the oscillation wavelength of the light emitting device is 530 nm, the In composition of the first guide layer 313 is 0.04, and the thickness of the second guide layer 315 in FIG. 1 is 0.1 μm. 4 shows the film thickness dependence of the first guide layer 313 and the film thickness dependence of the first cladding layer 12 with respect to the optical confinement coefficient and the effective refractive index difference.

  Here, the solid line indicates the contour line of the waveguide loss, the alternate long and short dash line indicates the contour line of the optical confinement coefficient, and the broken line indicates the contour line of the effective refractive index difference. The shaded area in FIG. 8 indicates a region that satisfies (Expression 3) and (Expression 4) at the same time.

  From FIGS. 5, 6, 7, and 8, the effective refractive index indicated by the broken line becomes smaller as both the film thickness Tnc of the first cladding layer 12 and the film thickness Tng of the first guide layer 313 become smaller. It can be seen that the difference increases.

Here, it is understood that the effective refractive index difference is 3 × 10 ⁻³ or more in most of the range as long as the range satisfies the following (formula 3).

If the effective refractive index difference is 3 × 10 ⁻³ or more, it can be stably confined in the ridge 14A including higher order modes.

  Therefore, in the configuration having the second guide layer 315, the effect of optical confinement in the lateral direction can be improved by satisfying the above (Equation 3).

Further, from FIGS. 5, 6, 7, and 8, the conductive line indicated by the solid line becomes smaller as both the film thickness Tnc of the first cladding layer 12 and the film thickness Tng of the first guide layer 313 become smaller. It can be seen that the waveguide loss increases. And if it is in the range satisfying the following (Equation 4), the generation of the waveguide loss can be suppressed to 10 cm ⁻¹ or less in many of the ranges, and the generation of the waveguide loss in all the ranges is 15 cm. ^-1 or less.

  Therefore, in the configuration having the second guide layer 315, satisfying the above (Equation 3) and (Equation 4) makes it possible to achieve both the lateral optical confinement effect and the waveguide loss suppression effect. .

  Furthermore, from FIG. 5, FIG. 6, FIG. 7, and FIG. 8, as the film thickness Tnc of the first clad layer 12 and the film thickness Tng of the first guide layer 313 are both smaller, it is indicated by a one-dot chain line. It can be seen that the optical confinement factor increases. If the above (Equation 3) is satisfied, it can be seen that the optical confinement coefficient is at least 1.0%. Therefore, by satisfying the above (Equation 3), the light confinement effect in the vertical direction can be obtained more reliably.

  Further, the hatched portions in FIG. 5, FIG. 6, FIG. 7, and FIG. 8 indicate regions that simultaneously satisfy (Expression 3) and (Expression 4).

  Note that (Equation 1), (Equation 2), (Equation 3), and (Equation 4) are calculated based on simulation results when the light emitting device oscillation wavelength is 530 nm. Even if it is extended to a range of 560 nm or less, it is considered that it can be applied to achieve both improvement of the light confinement effect in the vertical direction and suppression of distortion generation. The reason will be described below.

FIG. 9 shows that the thickness of the first guide layer 313 is 0.17 μm, the thickness of the first cladding layer 12 is 0.5 μm, the Al composition of the second cladding layer 14 is 0.07, and the first refraction. It shows the oscillation wavelength dependence of the effective refractive index difference when the film thickness of the rate correction layer 15 is 0.4 μm. From FIG. 9, the effective refractive index difference at the oscillation wavelength of 530 nm is 4.2 × 10 ⁻³ , whereas the effective refractive index difference at 500 nm is 3.95 × 10 ⁻³ , and the effective refractive index at 560 nm. Since the difference is 4.45 × 10 ⁻³ , it can be seen that the deviation is only about 6% based on the effective refractive index difference at the oscillation wavelength of 530 nm. Therefore, even if the oscillation wavelength range is extended to a range of 500 nm to 560 nm, it is considered that the light confinement effect in the lateral direction can be sufficiently obtained.

FIG. 10 shows that the thickness of the first guide layer 313 is 0.17 μm, the thickness of the first cladding layer 12 is 0.5 μm, the Al composition of the second cladding layer 14 is 0.07, and the first refraction. It shows the oscillation wavelength dependence of the waveguide loss when the thickness of the rate correction layer 15 is 0.4 μm. From FIG. 10, with respect to the waveguide loss at an oscillation wavelength 530nm in the range of 8.2 cm ^-1, the waveguide loss is ^{6.9 cm} -1 at 500 nm, the waveguide loss at 560nm is 9.5cm ^-1 Therefore, it can be seen that the deviation is only about 1.3 cm ⁻¹ . Therefore, even if the oscillation wavelength range is extended to a range of 500 nm to 560 nm, it is considered that the effect of suppressing the waveguide loss can be sufficiently obtained.

  Therefore, the above (Equation 1), (Equation 2), (Equation 3), and (Equation 4) are calculated based on the simulation result when the light emitting device oscillation wavelength is 530 nm, which is 500 nm to 560 nm. It can be said that even if it is expanded to the above range, it can be applied to achieve both improvement of the light confinement effect in the vertical direction and suppression of distortion generation.

  In FIG. 11, the thickness of the first guide layer 313 is 0.17 μm, the thickness of the first cladding layer 12 is 0.5 μm, the Al composition of the second cladding layer 14 is 0.07, This shows the oscillation wavelength dependence of the optical confinement coefficient when the film thickness of the refractive index correction layer 15 is 0.4 μm. From FIG. 11, the optical confinement factor at the oscillation wavelength of 530 nm is 1.23 (%), whereas the optical confinement factor at 500 nm is 1.33 (%), and the optical confinement factor at 560 nm is 1.13. (%), It can be seen that the deviation is only about 0.1%. Therefore, even if the oscillation wavelength range is extended to a range of 500 nm to 560 nm, it is considered that the light confinement effect in the vertical direction can be sufficiently obtained.

  Hereinafter, more specific embodiments including arbitrary configurations that are not essential configurations will be described.

  In the embodiment shown in FIG. 1, the upper portion of the second cladding layer 14 forms a ridge 14A.

  The width of the ridge 14A affects the highest order of the horizontal transverse mode that can guide the ridge 14A. When the width of the ridge 14A is large, the highest order of the horizontal transverse mode that can be guided becomes large, and multi transverse mode oscillation is likely to occur. Conversely, when the width of the ridge 14A is small, the maximum order is also reduced, and when all the high-order transverse modes are cut off, only the fundamental (0th-order) transverse mode oscillates.

  In order to prevent kinks from occurring in the current-light output characteristics, the horizontal higher-order transverse mode may be cut off. Kink occurs because horizontal transverse modes of different orders are combined to deform the light distribution, and the mode gain generated in the light distribution by current injection and the efficiency of using the injected current are reduced. Therefore, in general, in order to suppress the occurrence of kinks, the horizontal high-order transverse mode is cut off, and a waveguide having a narrow stripe width with a ridge width of 2 μm or less is formed so as to operate only in the fundamental transverse mode. To do.

  Here, considering a wide stripe structure in which the horizontal high-order transverse mode having a width of 3 μm or more in which the ridge 14A has a width is not cut off, in order not to generate a kink having a large non-linearity in current-light output characteristics, the ridges 14A oscillate simultaneously. It is better to have as many horizontal and horizontal modes of different orders as possible. This is because when horizontal transverse modes of different orders are coupled, the light distribution is deformed and kinks occur. However, if the number of transverse modes oscillating at the same time is large, the influence of the coupling becomes relatively small. On the contrary, it is because it becomes difficult to produce a large nonlinear kink. If there are at least three horizontal transverse modes that oscillate at the same time, the effect of deformation of the light distribution when the transverse modes are coupled can be reduced, and the current has good linearity with no practical problems. -Light output characteristics can be realized.

  Therefore, in order to obtain a current-light output characteristic with excellent linearity, it is preferable that the width of the ridge 14A is large, and the width of the ridge 14A is at least a basic (0th order) transverse mode, a primary transverse mode, and a secondary order. The transverse mode must be large enough to allow laser oscillation at the same time. Here, the highest-order horizontal transverse mode that can be guided has a smaller effective refractive index than other horizontal transverse modes, and therefore, the spread of the light distribution in the horizontal direction outside the ridge 14A is large. For this reason, the highest-order horizontal transverse mode that can be guided has a larger oscillation threshold for laser oscillation in that mode than other horizontal transverse modes, and the laser is relatively in comparison with other modes. Oscillation is difficult to occur. For this reason, in order to perform laser oscillation in at least three types of horizontal and transverse modes simultaneously, the width of the ridge 14A is required so that the highest order that can be guided is the third order or more.

  Also, if the width of the ridge 14A is large, the density of the current injected into the active layer 13 below the ridge 14A becomes small. Therefore, in the high-temperature operation, carriers thermally excited from the active layer 13 are the second cladding layer. The carrier overflow leaking to 14 can be suppressed. However, since the area of the active layer 13 into which current is injected under the ridge 14A is increased, the oscillation threshold current value necessary for laser oscillation is increased, so that the operating current value is increased and the power consumption during operation is increased. .

  For the above reason, in order not to cause a large kink in the current-light output characteristics, the width of the ridge 14A needs to be large enough to cause laser oscillation in the third-order transverse mode or more, and on the other hand, power consumption is not increased too much. Therefore, it is preferable that the width of the ridge 14A is not too large. Specifically, the width of the ridge 14A is preferably 5 μm or more and 16 μm or less.

  In the second cladding layer 14, the effective refractive index difference between the inside and outside of the ridge 14A can be controlled by the film thickness dp between the lower surface of the current blocking layer 318 and the upper surface of the second guide layer 315. is there. When the film thickness dp is small, the current blocking layer 318 having a low refractive index approaches the active layer 13, so that the effective refractive index difference increases. Conversely, when the film thickness dp is large, the effective refractive index difference decreases. Further, in a nitride laser, since the wet etching process using an acidic or alkaline solution is very difficult to form the ridge 14A, it is very difficult to accurately control the size of the film thickness dp. Need to use process. When the width of the ridge 14A is 4 μm or more, the light distribution in the vertical direction is hardly affected by the film thickness dp, and is determined by the laminated structure of the waveguide structure of the ridge 14A. Only changes. Therefore, even if the film thickness dp is changed, the light distribution shape change in the ridge 14A is very small, and the waveguide loss and the light confinement factor in the vertical direction to the active layer 13 do not change. Therefore, it is more effective to suppress the kink by reducing the film thickness dp as much as possible, increasing the effective refractive index difference, and increasing the order of the horizontal transverse mode that can be guided. However, if the film thickness dp is too small, there is a concern that a region that is etched to the depth of the active layer 13 may be generated due to the variation in the etching depth in the wafer surface when the ridge is formed by dry etching. For this reason, the film thickness dp is preferably as small as possible, taking into account the in-wafer variation of the dry etching depth, and is preferably 50 nm or less. In this embodiment, it is 30 nm. Even if dp changes within a range of 30 nm ± 20 nm, the optical loss coefficient in the vertical direction to the active layer 13 of the waveguide loss and the light distribution is almost constant, and the effective refractive index difference is accurately controlled to a desired value. Is possible.

  As shown in FIG. 16, the second refractive index correction layer 16 having the second conductivity is provided above the second cladding layer 14, and the second cladding layer 14 and the second cladding layer 14 The ridge 14 </ b> A may be configured by the refractive index correction layer 16. The second refractive index correction layer 16 is composed of an InAlN layer. With such a configuration, the refractive index of the second refractive index correction layer 16 is smaller than the refractive index of the second cladding layer 14 made of an AlGaN layer, so that the vertical light distribution is changed to the active layer 13. This is because it becomes possible to increase the light confinement effect.

A current blocking layer 318 is provided on the side surface of the ridge 14A and the upper surface of the second cladding layer 14 shown in FIG. As the current blocking layer 318, SiO ₂ or the like can be used. When SiO ₂ is used, the refractive index is 1.5 or less. By providing such a current blocking layer 318, an effective refractive index difference inside and outside the ridge 14A can be formed. For example, when the Al composition of the second cladding layer 14 is 0.1, the refractive index is 2.4, and the refractive index of the current blocking layer 318 is lower than the refractive index of the second cladding layer 14. The effective refractive index difference between the inside and outside of the ridge 14A can be formed.

  In the configuration shown in FIG. 16 as well, an effective refractive index difference inside and outside the ridge 14A is formed by providing a current blocking layer 318 made of SiO2 or the like on the side surface of the ridge 14A and the upper surface of the second cladding layer 14. can do.

  The substrate 11 may be made of GaN, but a sapphire substrate may be prepared as the first layer, and a multi-layer structure in which a GaN layer is formed thereon may be used. Alternatively, instead of the sapphire substrate, a silicon substrate or the like may be prepared as the first layer, and a multi-layer structure in which a GaN layer is formed thereon may be used.

  The In composition of the first guide layer 313 shown in FIGS. 1 and 16 is 0.04 or less. By setting it as such a structure, it can suppress that the active layer 13 produces a defect. This is because the difference between the lattice constant of GaN used for the substrate 11 and the lattice constant of the first guide layer 313 can be reduced.

  Similarly, the In composition of the second guide layer 315 is desirably 0.04 or less.

  When the film thickness of the first guide layer 313 shown in FIGS. 1 and 16 is Tng [μm] and the film thickness of the second guide layer 315 is Tpg [μm], the total film thickness is 0.3 [ [mu] m] or less. By adopting such a configuration, it is possible to suppress an increase in resistance, and as a result, it is possible to suppress upward of the operating voltage. Furthermore, it is possible to suppress the occurrence of defects in the active layer 13 by forming the first guide layer 313 and the second guide layer 315 containing In thin. Reduce the difference between the lattice constant of GaN used for the substrate 11 and the lattice constant of the first guide layer 313, and the difference between the lattice constant of GaN used for the substrate 11 and the lattice constant of the second guide layer 315. It is because it can do.

  Note that the Al composition of InAlN in the first refractive index correction layer 15 shown in FIGS. 1 and 16 is preferably 0.76 or more and 0.9 or less. By setting the composition range, the deviation between the lattice constant of GaN used for the substrate 11 and the lattice constant of the first refractive index correction layer 15 can be reduced, so that the critical film thickness can be 10 nm or more.

  Similarly, in the second refractive index correction layer 16, it is desirable that the Al composition of the InAlN layer is 0.76 or more and 0.9 or less. This is because the deviation between the lattice constant of the GaN layer used for the substrate 11 and the lattice constant of the second refractive index correction layer 16 can be reduced, and the critical film thickness can be 10 nm or more.

  As shown in FIGS. 1 and 16, a first electrode 321 is provided on the lower surface of the substrate 11.

  As shown in FIGS. 1 and 16, a contact layer 317 is provided above the ridge 14 </ b> A, and a second electrode 320 is provided above the contact layer 317 and on the side and upper surfaces of the current blocking layer 318.

  1 and FIG. 16, the first refractive index correction layer 15 is composed of an InAlN layer. However, the first refractive index correction layer 15 is composed of an InAlN layer and a GaN layer. You may comprise by a superlattice. In that case, the total film thickness of the InAlN layer is desirably 20% or more of the film thickness of the first refractive index correction layer 15. This is because the superlattice having such a structure has a refractive index lower than that of AlGaN having an Al composition of 0.1, and thus can sufficiently function as a refractive index correction layer.

  In addition, since the first refractive index correction layer 15 includes the InAlN layer, it is possible to achieve both improvement of the vertical light confinement effect and suppression of distortion generation. The reason will be described below.

  First, a first refractive index correction layer 15 including an InAlN layer having a refractive index smaller than that of GaN is disposed between a substrate 11 made of a GaN layer and a first cladding layer 12 made of a GaN layer. Directional light confinement effect can be obtained.

  In addition, since the lattice constant of InN is large and the lattice constant of AlN is small compared to the lattice constant of GaN, by adjusting the composition ratio of In and Al in the first refractive index correction layer 15, The lattice constant deviation with respect to the substrate 11 can be reduced.

  As a result, it is possible to achieve both improvement in the vertical light confinement effect and suppression of distortion generation.

  Further, similarly, the second refractive index correction layer 16 may be composed of only an InAlN layer, or may be composed of a superlattice of an InAlN layer and a GaN layer. When the InAlN layer and the GaN layer are used as a superlattice, the total thickness of the InAlN layer is preferably 20% or more of the thickness of the second refractive index correction layer 16. This is because the superlattice having such a structure has a refractive index lower than that of AlGaN having an Al composition of 0.1, and thus can sufficiently function as a refractive index correction layer.

  The film thickness of the first refractive index correction layer 15 is desirably 0.4 μm or more. By setting the thickness to 0.4 μm or more, it is possible to realize a configuration in which the waveguide loss is small and both the effective refractive index difference and the optical confinement coefficient are large. If the film thickness of the first refractive index correction layer 15 is 0.4 μm or more, changes in the values of the waveguide loss, the effective refractive index difference, and the optical confinement coefficient are small enough to be ignored. The reason will be described below.

  In FIG. 12, the thickness of the first refractive index correction layer 15 is 0.2 μm, the oscillation wavelength of the light emitting device is 530 nm, the In composition of the first guide layer 313 is 0.02, and the second guide layer 315 The film thickness dependence of the first guide layer 313 and the first cladding layer 12 with respect to the waveguide loss, optical confinement coefficient, and effective refractive index difference when the film thickness is 0.1 μm is shown.

  Here, the solid line indicates the contour line of the waveguide loss, the alternate long and short dash line indicates the contour line of the optical confinement coefficient, and the broken line indicates the contour line of the effective refractive index difference.

  FIG. 13 shows the waveguide loss, optical confinement coefficient, and effective refractive index difference when the thickness of the first refractive index correction layer 15 is 0.4 μm and other conditions are the same as those in FIG. The film thickness dependence of the first guide layer 313 and the first cladding layer 12 is shown.

  Here, the solid line indicates the contour line of the waveguide loss, the alternate long and short dash line indicates the contour line of the optical confinement coefficient, and the broken line indicates the contour line of the effective refractive index difference.

  FIG. 14 shows the waveguide loss, optical confinement coefficient, and effective refractive index difference when the film thickness of the first refractive index correction layer 15 is 0.6 μm and other conditions are the same as those in FIG. The film thickness dependence of the first guide layer 313 and the first cladding layer 12 is shown.

  Here, the solid line indicates the contour line of the waveguide loss, the alternate long and short dash line indicates the contour line of the optical confinement coefficient, and the broken line indicates the contour line of the effective refractive index difference.

  First, comparing FIG. 12 with FIG. 13, the waveguide loss contour shown in FIG. 13 shows the thickness of the first guide layer 313 and the first cladding layer 12 rather than the waveguide loss contour shown in FIG. It can be seen that the region of low waveguide loss as a whole is widened. On the other hand, when FIG. 13 and FIG. 14 are compared, it can be confirmed that there is almost no difference between the two in terms of the arrangement relationship of the contour lines of the waveguide loss.

  Further, when FIG. 12 is compared with FIG. 13, the effective refractive index difference contour shown in FIG. 13 is more effective than the effective refractive index difference contour shown in FIG. It can be seen that the cladding layer 12 is positioned in the direction of increasing film thickness, and as a whole, a region having a high effective refractive index difference is widened. On the other hand, when FIG. 13 and FIG. 14 are compared, it can be confirmed that there is almost no difference between the contour lines of the effective refractive index difference in the positional relationship.

  From the comparison of FIGS. 12 to 14, by setting the film thickness of the first refractive index correction layer 15 to 0.4 μm or more, a configuration in which the waveguide loss is small and the lateral light confinement effect is large is realized. be able to. Further, it can be confirmed that when the film thickness of the first refractive index correction layer 15 is 0.4 μm or more, changes in the values of the waveguide loss and the effective refractive index difference are small enough to be ignored.

  Further, comparing FIGS. 12 and 13, the contour lines of the optical confinement coefficient shown in FIG. 13 are larger than the contour lines of the optical confinement coefficient shown in FIG. It can be seen that the region where the optical confinement coefficient is large spreads as a whole. On the other hand, when FIG. 13 and FIG. 14 are compared, it can be confirmed that there is almost no difference between the two in terms of the arrangement relationship of the contour lines of the optical confinement coefficient. Therefore, by setting the film thickness of the first refractive index correction layer 15 to 0.4 μm or more, it is possible to realize a configuration in which vertical light confinement is large. Further, it can be confirmed that if the film thickness of the first refractive index correction layer 15 is 0.4 μm or more, the change in the value of the optical confinement factor is small enough to be ignored.

  Based on these results, the above-described simulations of FIGS. 3 to 8 are calculated assuming that the film thickness of the first refractive index correction layer 15 is 0.4 μm or more.

  The film thickness of the second cladding layer 14 is desirably 0.5 μm or more. By setting the thickness to 0.5 μm or more, a configuration in which the waveguide loss is small and the effective refractive index difference is large can be realized more stably. If the film thickness of the second cladding layer 14 is 0.5 μm or more, changes in the values of the waveguide loss and the effective refractive index difference are negligibly small. The reason will be described below.

  FIG. 15 shows that the thickness of the first refractive index correction layer 15 is 0.4 μm or more, the oscillation wavelength of the light emitting device is 530 nm, the In composition of the first guide layer 313 is 0.02, and the second guide layer 315 The graph shows the Al composition dependency of the second cladding layer 14 and the film thickness dependency of the second cladding layer 14 with respect to the waveguide loss and effective refractive index difference when the film thickness is 0.1 μm.

  Here, the solid line indicates the contour line of the waveguide loss, and the broken line indicates the contour line of the effective refractive index difference.

  As shown in FIG. 15, the waveguide loss decreases or increases as the thickness of the second cladding layer 14 increases, and saturates when the thickness of the second cladding layer 14 becomes 0.5 μm. .

  Further, as shown in FIG. 15, the effective refractive index difference increases as the thickness of the second cladding layer 14 increases, and when the thickness of the second cladding layer 14 becomes 0.5 μm. Saturates.

  For these reasons, by setting the thickness of the second cladding layer 14 to 0.5 μm or more, it is possible to more stably realize a configuration in which the waveguide loss is small and the effective refractive index difference is large. In addition, if the thickness of the second cladding layer 14 is 0.5 μm or more, it can be confirmed that changes in the values of the waveguide loss and the effective refractive index difference are negligibly small.

Further, in FIG. 15, focusing on the range in which the film thickness of the second cladding layer 14 is 0.5 μm or more, the effective refractive index difference is increased if the Al composition of the second cladding layer 14 is 0.08 or less. 3 × can be ^{10 -3} or more, the waveguide loss can be reduced to 15cm ^-1 or less as long as the Al composition of the second cladding layer 14 is 0.05 or more. Therefore, by setting the Al composition of the second cladding layer 14 to 0.05 or more and 0.08 or less, it is possible to realize a configuration in which the effective refractive index difference is large and the waveguide loss is small.

  Based on these results, in the simulations of FIGS. 3 to 8 described above, the thickness of the second cladding layer 14 is 0.5 μm or more, and the Al composition of the second cladding layer 14 is 0.07. Calculated assuming a range.

As described above with reference to FIG. 15, if the film thickness of the second cladding layer 14 is in the range of 0.5 μm or more, the waveguide loss with respect to the film thickness of the second cladding layer 14 and the effective refraction It can be seen that the change in the rate difference shows a saturation tendency. Further, when the Al composition of the second cladding layer 14 is in the range of 0.05 to 0.08 and the film thickness of the second cladding layer 14 is in the range of 0.5 μm to 0.8 μm. The waveguide loss is controlled in the range of 8 cm ⁻¹ to 12 cm ⁻¹ , and the effective refractive index difference is controlled in the range of 3.5 × 10 ⁻³ to 5.1 × 10 ⁻³ . That is, it can be seen that a low loss of 12 cm ⁻¹ or less and an effective refractive index difference of 3 × 10 ⁻³ or more can be stably obtained. Specifically, when the film thickness of the second cladding layer 14 is 0.5 μm and the Al composition of the second cladding layer 14 is 0.05, the waveguide loss is 12 cm ⁻¹ , the effective refractive index. It can be seen that the difference is 4.2 × 10 ⁻³ . On the other hand, when the thickness of the second cladding layer 14 is 0.5 μm and the Al composition of the second cladding layer 14 is 0.08, the waveguide loss is 8.7 cm ⁻¹ , the effective refractive index difference. Is 3.2 × 10 ⁻³ . Therefore, if the Al composition of the second cladding layer 14 is in the range of 0.05 to 0.08, the influence of the Al composition on the waveguide loss and the effective refractive index difference is considered to be very small. It is done. Therefore, it can be said that the simulation results of FIGS. 3 to 8 described above are effective when the Al composition of the second cladding layer 14 is in the range of 0.05 to 0.08.

  The semiconductor light-emitting device of the present disclosure is a semiconductor light-emitting device having a substrate made of GaN, and can achieve both improvement of the light confinement effect and suppression of distortion generation. It is useful in other electronic devices and information processing devices.

1 substrate 2 n-type cladding layer 3 active layer 4 p-type cladding layer 5 current blocking layer 6 p-type contact layer 11 substrate 12 first cladding layer 13 active layer 14 second cladding layer 14A ridge 15 first refractive index correction Layer 16 Second refractive index correction layer 313 First guide layer 315 Second guide layer 317 Contact layer 318 Current blocking layer 320 Second electrode 321 First electrode

Claims (18)

A substrate,
A first refractive index correction layer having a first conductivity provided above the substrate;
A first cladding layer provided above the first refractive index correction layer;
An active layer provided above the first cladding layer;
A second clad layer having second conductivity provided above the active layer;
With
Having a first guide layer between the first cladding layer and the active layer;
The second cladding layer has a ridge;
The substrate is a GaN layer;
The first refractive index correction layer includes an InAlN layer;
The first cladding layer is a GaN layer;
The first guide layer is an InGaN layer;
The active layer is an InGaN layer;
The second cladding layer is an AlGaN layer;
The film thickness of the first cladding layer is Tnc,
When the film thickness of the first guide layer is Tng,
Tng [μm] ≦ −0.98 Tnc [μm] +0.76 [μm]
Tng [μm] ≧ −0.98 Tnc [μm] +0.56 [μm]
Satisfies the following two formulas,
A semiconductor light emitting device having an emission wavelength of 500 nm or more and 560 nm or less. The first refractive index correction layer comprises a superlattice of an InAlN layer and a GaN layer,
The semiconductor light emitting device according to claim 1, wherein a total film thickness of the InAlN layer is 20% or more of a film thickness of the first refractive index correction layer. The semiconductor light emitting device according to claim 1, wherein an In composition of the first guide layer is 0.04 or less. Having a second guide layer between the second cladding layer and the active layer;
The semiconductor light emitting device according to claim 1, wherein the second guide layer is an InGaN layer. The film thickness of the first cladding layer is Tnc,
The film thickness of the first guide layer is Tng,
When the film thickness of the second guide layer is Tpg,
The semiconductor light-emitting device according to claim 4, satisfying a relational expression of Tng [μm] ≦ −0.98 Tnc [μm] −1.2 Tpg [μm] +0.76 [μm]. The semiconductor light emitting device according to claim 5, satisfying a relational expression of Tng [μm] ≧ −0.98 Tnc [μm] −1.2 Tpg [μm] +0.56 [μm]. Tng [μm] + Tpg [μm] ≦ 0.3 [μm]
The semiconductor light emitting device according to claim 6, satisfying the relational expression: The semiconductor light-emitting device according to claim 4, wherein the In composition of the first guide layer is 0.04 or less. The semiconductor light emitting device according to claim 4, wherein the In composition of the second guide layer is 0.04 or less. A second refractive index correction layer having a second conductivity above the second cladding layer;
The semiconductor light emitting device according to claim 1, wherein the second refractive index correction layer includes an InAlN layer. The said 2nd refractive index correction layer consists of a superlattice of an InAlN layer and a GaN layer, and the total film thickness of the said InAlN is 20% or more of the film thickness of the said 2nd refractive index correction layer. Semiconductor light emitting device. The substrate is a sapphire substrate;
The semiconductor light-emitting device according to claim 1, comprising a GaN layer provided above the sapphire substrate. The substrate is a silicon substrate;
The semiconductor light-emitting device according to claim 1, comprising a GaN layer provided above the silicon substrate. 2. The semiconductor light emitting device according to claim 1, wherein an Al composition of InAlN in the first refractive index correction layer is 0.76 or more and 0.9 or less. The semiconductor light emitting device according to claim 10, wherein an Al composition of InAlN in the second refractive index correction layer is 0.76 or more and 0.9 or less. 2. The semiconductor light emitting device according to claim 1, wherein an Al composition of the second cladding layer is 0.05 or more and 0.08 or less. The semiconductor light emitting device according to claim 1, wherein the film thickness of the second cladding layer is 0.5 μm or more. The semiconductor light-emitting device according to claim 1, wherein the film thickness of the first refractive index correction layer is 0.4 μm or more.

Images