JPH0983059A - Semiconductor laser - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a semiconductor laser in which low loss optical coupling is realized through direct abutting against an optical waveguide having large spot size while reducing the cost and increasing the yield of manufacture. SOLUTION: The semiconductor laser comprises a stripe semiconductor active layer 1 of constant width emitting a light upon injection of current, and a clad layer 2 surrounding the semiconductor active layer 1 where the refractive index of clad layer 2 is lower than that of the semiconductor active layer 1. The confinement coefficient Γ of light in the active layer 1 is set in the range of 0.002-0.05 thus coupling the outgoing spot light optically with an optical waveguide with low loss without requiring any lens or optical coupling device for converting the spot size.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a semiconductor laser having a large emission beam radius.

[0002]

2. Description of the Related Art In a conventional semiconductor laser, the width of an active layer is set to an almost upper limit value (for example, about 1 to 3 μm) so that a single transverse mode is generated during laser oscillation, and the vertical and horizontal directions of a substrate are set. The total optical confinement factor Γ in the active layer considering the direction was set to be somewhat strong (for example, 0.1 to 0.2) so that the low threshold operation is possible. For this reason, the radius of the emitted beam was as small as about 1 μm, and the beam spread rapidly from the emission end.

When such a semiconductor laser is directly butted against an optical waveguide having a large spot size such as an optical fiber to perform optical coupling, coupling loss becomes extremely large. For example, when the optical coupling is performed by directly abutting against the single mode optical fiber, there is a problem that the coupling loss becomes about 10 dB.

Therefore, generally, at least one lens or an optical coupling device for converting the spot size of light is inserted between the semiconductor laser and the optical fiber to reduce the coupling loss.

[0005]

However, when a lens is used, there is a problem in that the coupling loss increases sharply with the misalignment of the optical fiber, and the axis alignment must be performed with extremely high accuracy. In addition, an optical coupling device that converts the spot size of light is integrated with a semiconductor laser to develop an element that can obtain a light beam with a large spot size. However, there is a problem in that the cost increases and the yield decreases.

An object of the present invention is to provide a semiconductor laser which enables low loss optical coupling by direct butting to an optical waveguide having a large spot size such as an optical fiber, and which can be manufactured at a low cost and a high yield. is there.

[0007]

According to the present invention, as shown in FIG. 1, a semiconductor active layer 1 having a uniform width and emitting light by current injection, and a clad formed so as to surround the semiconductor active layer 1. A layer 2 and the cladding layer 2 has a refractive index smaller than that of the semiconductor active layer 1, a light confinement coefficient Γ in the active layer 1 is 0.002 to 0.0
It is characterized in that the spot size is enlarged by setting to about 5 so that the spot size is coupled with low loss to the optical waveguide optically coupled with the emitted light.

FIG. 2 shows the relationship between the coupling loss for the dispersion-shifted fiber and the optical confinement coefficient Γ of the laser obtained by calculation. In the conventional semiconductor laser, Γ is generally 0.1 or more, and the coupling loss is 5 dB or more.
In the present invention, by setting Γ to about 0.002 to 0.05, a coupling loss of 3 dB or less can be realized. Such Γ
In order to achieve the above, the following configuration may be adopted.

Structure (1) When a semiconductor having a uniform composition is used for the active layer 1, Γ is 0.002 to 0.002 by mainly narrowing or narrowing the active layer.
It can be set to about 0.05. It is also effective to select the material so that the refractive index of the cladding layer 2 is close to that of the active layer 1. However, when a material having a refractive index close to that of the carrier is selected, the confinement of carriers is generally weakened and the oscillation threshold and efficiency are significantly lowered.

Structure (2) When an active layer 1 is a multiple quantum well layer or multiple quantum well layers and SCH layers forming separate confirmation heterostructures on both sides thereof, these layers are made narrower or thinner. By doing so, Γ can be set to about 0.002 to 0.05. Further, in this case, a material having a wide gap and a refractive index close to that of the cladding layer is used for the barrier layer or the SCH layer of the multiple quantum well so that the equivalent active layer has a smaller refractive index than that of the uniform composition. Even if the width or thickness of the multiple quantum well layer or the SCH layers on both sides thereof is relatively large, Γ can be set to about 0.002 to 0.05.

By using the configuration (1) or (2) as described above, the spot size of light is uniformly expanded in the laser cavity, and the optical waveguide having a large spot size such as an optical fiber is directly abutted. Optical coupling with low loss becomes possible. Further, as compared with the conventional device in which a laser and an optical coupling device are integrated, a process for integration is not necessary, and the yield improvement due to the process is synergistically achieved, which is extremely effective in reducing the cost of the device. Furthermore, unlike the element in which the conventional laser and the optical coupling device having the tapered waveguide are integrated, there is no occurrence of electrical loss due to the integration, and the integration of the tapered waveguide causes the generation of the coupling loss and the tapered portion. The efficiency is improved without increasing the optical loss due to the generation of the radiated light component due to the mode change of the light in the above.

[0012]

BEST MODE FOR CARRYING OUT THE INVENTION

(First Embodiment) FIG. 3 shows a first embodiment of the present invention, in which 21 is an n-InP substrate and 22 is an n-In.
P clad layer, 23 InGaAsP (λ = 1.3 μm composition) active layer, 24 p-InP clad layer, 25 p ⁺
-InGaAsP cap layer, 26 is a p-InP buried layer, 27 is an n-InP buried layer, and 28 and 29 are an n-electrode and a p-electrode, respectively.

The active layer 23 has a width of 1.0 μm and a thickness of 0.04 μm. In an element with an active layer width of 1.5 μm and a thickness of 0.1 μm of the conventional structure, the light confinement coefficient Γ was 0.1 or more, and the beam radius at the exit end was about 1 μm, but the active layer width was narrowed as shown in FIG. Further, by reducing the thickness, Γ was reduced to about 0.006, and as a result, the beam radius was expanded to about 2.5 μm, and a coupling efficiency of −2.0 dB was obtained by direct butting with the dispersion shift fiber.

Also, the coupling tolerance, which is the tolerance for the axis deviation between the fiber and the laser, is within the coupling efficiency deterioration range of 1 dB, and is ± 2 μ in the X and Y directions in the abutting plane.
A large value of m was obtained. It should be noted that with the conventional structure, only a coupling efficiency of about -10 dB can be obtained, and the coupling tolerance with the fiber using the lens is a coupling efficiency deterioration range of 1 dB, and only ± 0.4 μm in the X and Y directions in the butt plane. I can't get it.

The oscillation spectrum was similar to that of a normal Fabry-Perot laser. The oscillation threshold value was about 9 mA for the conventional structure having a cavity length of 500 μm, but even if the optical confinement was weakened as in the present embodiment, the oscillation threshold value was about 17 mA. Further, as compared with an element in which a passive type optical coupling device is integrated with a laser, a process for manufacturing the optical coupling device section is not necessary, and the yield improvement due to it is synergistically achieved, and the cost can be reduced.

Although the substrate is n-type InP here, it may be p-type InP. In that case, in FIG. 3, 22 is a p-InP cladding layer and 23 is InGaAs.
P (λ = 1.3 μm composition) active layer, 24 n-InP clad layer, 25 n ⁺ -InGaAsP cap layer, 26
Is an n-InP buried layer, 27 is a p-InP buried layer, and 28 and 29 are a p-electrode and an n-electrode, respectively.

Further, here, the active layer is InGaAsP.
(Λ = 1.3 μm composition) is shown.
Other compositions such as nGaAsP (λ = 1.55 μm composition) may be used. InGaA other than InGaAsP / InP system
Similar effects can be obtained with a material system such as 1As / InGaAsP system or AlGaAs / GaAs system, or a material system having a strain therein. Although this example is an example of a Fabry-Perot laser, it goes without saying that the present invention can be applied to other types of lasers such as a distributed feedback (DFB) laser and a distributed reflection (DBR) laser.

(Second Embodiment) FIG. 4 shows a second embodiment of the present invention, in which 31 is an n-InP substrate, 3
2 is an n-InP clad layer, 33 is InGaAsP (λ
= 1.3 μm composition) active layer, 34 p-InP cladding layer, 35 p ⁺ -InGaAsP cap layer, 36 semi-insulating InP buried layer, 37 and 38 n-electrode and p-electrode, respectively.

The active layer 33 has a width of 1.0 μm and a thickness of 0.04 μm. In the conventional structure with an active layer width of 1.5 μm and a thickness of 0.1 μm, the light confinement coefficient Γ was 0.1 or more, and the beam radius at the exit end was about 1 μm, but the active layer width was narrowed as shown in FIG. Further, by reducing the thickness, Γ was reduced to about 0.006, and as a result, the beam radius was expanded to about 2.5 μm, and a coupling efficiency of −2.0 dB was obtained by direct butting with the dispersion shift fiber.

Also, the coupling tolerance, which is the tolerance for the axis deviation between the fiber and the laser, is within the coupling efficiency deterioration range of 1 dB, and is ± 2 μ in the X and Y directions in the abutting plane.
A large value of m was obtained. It should be noted that with the conventional structure, only a coupling efficiency of about -10 dB can be obtained, and the coupling tolerance with the fiber using the lens is a coupling efficiency deterioration range of 1 dB, and only ± 0.4 μm in the X and Y directions in the butt plane. I can't get it.

The oscillation spectrum was similar to that of a normal Fabry-Perot laser. The oscillation threshold value was about 8 mA for the conventional structure having a cavity length of 500 μm, but the oscillation threshold value was about 15 mA even if the optical confinement was weakened as in the present embodiment. Further, as compared with an element in which a passive type optical coupling device is integrated with a laser, a process for manufacturing the optical coupling device section is not necessary, and the yield improvement due to it is synergistically achieved, and the cost can be reduced.

Although the substrate is made of n-type InP here, it may be made of p-type InP. In that case, in FIG. 4, 32 is a p-InP clad layer and 33 is InGaAs.
P (λ = 1.3 μm composition) active layer, 34 n-InP clad layer, 35 n ⁺ -InGaAsP cap layer, 36
Is a semi-insulating InP buried layer, and 37 and 38 are a p-electrode and an n-electrode, respectively.

The active layer is InGaAsP here.
(Λ = 1.3 μm composition) is shown.
Other compositions such as nGaAsP (λ = 1.55 μm composition) may be used. InGaA other than InGaAsP / InP system
Similar effects can be obtained with a material system such as 1As / InGaAsP system or AlGaAs / GaAs system, or a material system having a strain therein. Although this example is an example of a Fabry-Perot laser, it goes without saying that the present invention can be applied to other types of lasers such as a distributed feedback (DFB) laser and a distributed reflection (DBR) laser.

(Third Embodiment) FIG. 5 shows a third embodiment of the present invention. In the drawing, 41 is an n-InP substrate, 4
2 is an n-InP clad layer, 43 is 7 layers of 6 nm In
GaAsP well and 6 layers of 10 nm InGaAsP (λ
= 1.1 μm composition) Multiple quantum well active layer (λ
_PL = 1.3 μm), 44a and 44b are 10 nm InGaA
sP (λ = 1.1 μm composition) separate confinement hetero (SCH) layer, 45 is p-InP clad layer,
46 is p ⁺ -InGaAsP cap layer, 47 is p-I
An nP buried layer, 48 is an n-InP buried layer, and 49 and 50 are an n electrode and a p electrode, respectively.

Multiple quantum well active layer 43 and SCH layer 44
The widths of a and 44b are 0.7 μm. Further, a high reflection film is formed on one end surface thereof. Uniform active layer width of conventional structure
In the element with 1.5 μm and SCH layer thickness of 100 nm, the light confinement coefficient Γ was 0.15 and the beam radius at the exit end was about 1 μm. However, by narrowing the active layer width and thinning the SCH layer, As a result, the beam radius was expanded to about 2 μm, and the coupling efficiency of −3 dB or more was obtained by direct butting of the dispersion shift fiber.

Also, the coupling tolerance, which is the tolerance for the axis deviation between the fiber and the laser, is within the coupling efficiency deterioration range of 1 dB, and is ± 2 μ in the X and Y directions in the butt plane.
A large value of m was obtained. It should be noted that with the conventional structure, only a coupling efficiency of about -10 dB can be obtained, and the coupling tolerance with the fiber using the lens is a coupling efficiency deterioration range of 1 dB, and only ± 0.4 μm in the X and Y directions in the butt plane. I can't get it.

The oscillation threshold is about 6 mA when the cavity length of the conventional structure is 500 μm and the high reflection film is formed on one end face, but it oscillates even if the light confinement is weakened as in this embodiment. The threshold value was about 9 mA. In addition, the efficiency of the device is 0.4 W / A or more, and compared with the device in which the passive optical coupling device is integrated in the laser, there is no increase in electrical loss and optical loss due to integration, so it is 20%. As a result, a highly efficient device was obtained. Further, the efficiency was improved by shortening the cavity length, and at a cavity length of 300 μm, an efficiency of 0.5 W / A or higher was obtained.

The oscillation spectrum was similar to that of a normal Fabry-Perot laser. Further, as compared with an element in which a passive type optical coupling device is integrated with a laser, a process for integration is not required, and the yield improvement due to it is synergistically achieved, and the cost can be reduced.

Although the substrate is n-type InP here, it may be p-type InP. In that case, in FIG. 5, 42 is a p-InP clad layer and 43 is a 7-layer InG.
aAsP well and 6 layers of InGaAsP (λ = 1.1 μm
Multiple quantum well active layer (λ _PL = 1.3μ)
m), 44a and 44b are InGaAsP (λ = 1.1 μm)
Composition) Separate Confinement Hetero (SCH)
Layer, 45 is an n-InP clad layer, 46 is n ⁺ -InG
aAsP cap layer, 47 is n-InP buried layer, 4
8 is a p-InP buried layer, and 49 and 50 are p-type.
It becomes an electrode and an n-electrode.

Further, here, the case where the active layer is an InGaAsP multiple quantum well active layer (λ _PL = 1.3 μm) and the number of wells is 7 is shown, but the number of wells may be about 8 or less. The thickness of the SCH layer and the widths thereof may be adjusted in accordance with the above, and the spot size of light may be enlarged so as to couple with the optical waveguide that is optically coupled with low loss. For example, the number of wells is 4
When the width of the active layer is 1.0 μm, the SCH layer thickness is 4
When it is set to 0 nm, Γ becomes as small as 0.03, and as a result, the spot size of light is 3 dB and the coupling loss with the optical fiber.
The following can be expanded to combine. Also, 7 wells,
When the SCH layer thickness is 70nm, the width of the active layer is 0.4μ
When m is set, Γ is reduced to about 0.03, and as a result, the spot size of light can be expanded so as to be coupled with the optical fiber with a coupling loss of 3 dB or less.

Although the active layer is an InGaAsP multiple quantum well active layer (λ _PL = 1.3 μm) here, other configurations such as InGaAsP (λ _PL = 1.55 μm) and a strained quantum well structure are shown. Alternatively, a strain compensation quantum well structure or the like may be used.

Although the composition of the InGaAsP barrier layer and the SCH layer is λ = 1.1 μm here, λ = 1.0
By setting the thickness to 5 μm, the equivalent refractive index of the multiple quantum well active layer including the SCH layer becomes small, and the coupling occurs at a coupling loss of 2.5 dB or less. It goes without saying that the SCH layer may be a gradient composition layer (GRIN-SCH) from InGaAsP to InP. Also, InGaAsP /
Similar effects can be obtained with a material system other than the InP system, such as InGaAlAs / InGaAsP system, AlGaAs / GaAs system, or the like, or a material system having a strain therein.

Further, although this example is an example of the pn buried structure, it may be a semi-insulating buried structure. Although this example is an example of a Fabry-Perot laser, it goes without saying that the present invention can be applied to other types of lasers such as a distributed feedback (DFB) laser and a distributed reflection (DBR) laser.

[0034]

As described above, according to the present invention, in the waveguide type semiconductor laser in which the width of the active layer is constant,
Mainly by narrowing or thinning the active layer and setting the light confinement coefficient in the active layer to about 0.002 to 0.05,
The spot size of the light is uniformly expanded in the laser cavity, which enables low loss optical coupling by direct abutment with a large spot size optical waveguide such as an optical fiber. In addition, in the quantum well laser including the SCH layer, the SCH layer is narrowed or thinned in addition to the quantum well portion so that the confinement coefficient of light in the active layer is 0.002 to 0.0.
By setting it to about 5, the spot size of the light is uniformly expanded in the laser cavity, which enables low loss optical coupling by direct abutting against an optical waveguide with a large spot size such as an optical fiber. .

Further, as the spot size is enlarged, the tolerance for the axis deviation between the fiber and the laser is greatly improved as compared with the conventional combination using the laser and the lens. Further, as compared with the conventional device in which a laser and an optical coupling device are integrated, a process for integration is not necessary, and the yield improvement due to the process is synergistically achieved, which is extremely effective in reducing the cost of the device. Furthermore, as compared with the conventional element in which a laser and an optical coupling device are integrated, there is no increase in electrical loss and optical loss due to integration, so that efficiency can be improved.

[Brief description of drawings]

FIG. 1 is a perspective view showing the structure of a semiconductor laser of the present invention.

FIG. 2 is a diagram showing a relationship between a coupling loss for a dispersion-shifted fiber and an optical confinement coefficient Γ.

FIG. 3 is a sectional view showing a first embodiment of a semiconductor laser of the present invention.

FIG. 4 is a sectional view showing a semiconductor laser according to a second embodiment of the present invention.

FIG. 5 is a sectional view showing a third embodiment of a semiconductor laser according to the present invention.

[Explanation of symbols]

21, 31, 41 ... n-InP substrate, 22, 32, 42
... n-InP clad layer, 23, 33 ... InGaAsP
(Λ = 1.3 μm composition) Active layer, 24, 34, 45 ... P-
InP clad layer, 25, 35, 46 ... P ⁺ -InGa
AsP cap layer, 26, 47 ... p-InP buried layer, 27, 48 ... n-InP buried layer, 28, 37,
49 ... N-electrode, 29, 38, 50 ... P-electrode, 36 ... Semi-insulating InP buried layer, 43 ... Multiple quantum well active layer (λ
_PL = 1.3 μm), 44a, 44b ... InGaAsP (λ
= 1.1 μm composition) Separate Confinement Hetero (SCH) layer.

Claims (5)

[Claims] 1. A waveguide type semiconductor laser having a constant width of an active layer, wherein a confinement coefficient of light in the active layer is set to about 0.002 to 0.05. 2. The semiconductor laser according to claim 1, wherein an active layer having a uniform composition is used as the active layer. 3. The semiconductor laser according to claim 1, wherein a multiple quantum well layer is used as the active layer. 4. The semiconductor laser according to claim 1, wherein a multiple quantum well layer and a separate confirmation hetero layer formed on both sides of the multiple quantum well layer are used as the active layer. 5. The confinement coefficient of light in the active layer is set to about 0.002 to 0.05 by narrowing or thinning the active layer. Or the described semiconductor laser.