JPH0882810A - Optical switch - Google Patents

Abstract

PURPOSE: To eliminate the absorption loss of light in optical waveguides and to enable low-voltage operation with a small size by providing an optical switch with an n type region and p type region for impressing reverse biases to these optical waveguides on both sides of the optical waveguide regions. CONSTITUTION: The semiconductor optical waveguide of a branch type consisting of one piece of the main optical waveguide 3 and two pieces of the branched optical waveguides 4, 5 is composed of a multiple quantum well structure. The region held by these branch optical waveguides is formed with either conduction type region 10 and the regions facing this conduction type region 10 via the branch optical waveguides 4, 5 are formed with the opposite conduction type regions 11, 12. A refractive index change arises when a reverse bias voltage is impressed between the one conduction type region 10 of the optical switch and the p type region. Total reflection is generated between the branch optical waveguide 4 on the side impressed with the reverse bias voltage and the main waveguide 3 and incident light hνis introduced toward the branch optical waveguide 5 on the side not impressed with the reverse bias. Then, the change in the refractive index is utilizable to the max. possible extent if light of about 1.54μm is used as the incident light hν to be used. The absorption loss of the light is thus lessened.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an optical switch, and more particularly, to an optical switch which constitutes an optical circuit of an information processing device utilizing light such as an optical computer and optical interconnection.

[0002]

2. Description of the Related Art A basic optical operation circuit that constitutes an optical circuit of an information processing device utilizing light is an optical switch. Conventionally, as this optical switch, a ferroelectric crystal such as LiNbO ₃ crystal is used. The optical switch used is widely used, and an optical switch using a bulk semiconductor is also proposed. In this case, integration with a semiconductor optical active element such as a semiconductor laser or a pin photodiode is also possible. Become.

In any case, in order to construct an optical computer, an element capable of high-speed optical switch operation is required. In order to perform this high-speed operation, the parasitic capacitance of the element itself is reduced and the driving voltage is reduced. Reduction is needed.

However, an optical switch using a ferroelectric crystal has a large element size (on the order of cm) and is difficult to be integrated with a semiconductor optical active element, so that it is used as an optical switch for an optical arithmetic unit. Is not suitable for.

On the other hand, an optical switch using a bulk semiconductor can be integrated with a semiconductor photoactive device and downsized, but its change in refractive index with voltage is smaller than that of an optical switch using a LiNbO ₃ crystal. Therefore, in order to obtain a sufficient change in the refractive index, it is necessary to increase the driving voltage or increase the length of the optical operation element (optical waveguide length).

[0006] However, these are factors that hinder high-speed operation, cause optical loss, and hinder the improvement of the degree of integration. Further, as an optical switch using a semiconductor, a current injection type optical switch has been proposed in addition to the one that utilizes the change in the refractive index due to the application of a voltage. However, when high-speed modulation is applied to this, Since the injected carriers cannot follow the modulation, there is a drawback that the high speed response is inferior to the voltage application type.

Several proposals have been made in order to improve these drawbacks, and the principle of these proposals will be described with reference to FIGS. 7 to 11.

FIG. 7 is a diagram schematically illustrating a semiconductor optical switch (Japanese Patent Laid-Open No. 6-148699) filed by the present inventor, and FIG. 7A is a plan view. 7B is a cross-sectional view taken along the chain line BB ′ of FIG. 7A, and FIG. 7C is cut along the chain line CC ′ of FIG. 7A. FIG.

Referring to FIG. 7, a clad layer 2 provided on a semi-insulating substrate 1 is provided with a Y-shaped branched groove portion composed of one main optical waveguide 3 and two branched optical waveguides 4 and 5, and MQW After depositing the (multiple quantum well) layer 7, the cladding layer 8 and the contact layer 9, an n-type impurity reaching the lower cladding layer 2 is introduced into the branch portion of the optical waveguide to form the n-type region 15. At the same time, p-type impurities are similarly introduced into the central portions of the two branched optical waveguides 4 and 5 to form p-type regions 16 and 17 to form a semiconductor optical switch.

Next, the operating principle of this semiconductor optical switch will be described with reference to FIGS. FIG. 8 is a diagram showing the voltage dependence of the refractive index with respect to the wavelength of incident light. In the case where a reverse bias voltage is applied between the n-type region 15 and the p-type regions 16 and 17 in FIG. , The refractive index of the MQW layer 7 changes from the solid line to the broken line.

This is because the bound state is released due to the electric field to which the exciton level formed in the MQW layer 7 is applied. In the exciton, electrons in the conduction band and holes in the valence band generated by absorption of light in the semiconductor are bound to each other by the Coulomb force to form a level.

In the bulk semiconductor, the binding energy of excitons is low (for example, in the case of GaAs, 4.
Since it is 2 meV), it does not exist stably at room temperature due to the influence of thermal energy, and the voltage dependence of the refractive index change is small, but the MQW structure traps carriers, so the binding energy is high (also in the case of GaAs). , 16 meV), for example, it stably exists even at room temperature, and the voltage dependence of the refractive index change becomes large.

Therefore, as shown in FIG. 9A, when a reverse bias voltage is applied between the n-type region 15 and the p-type region 16, (18 is an n-type electrode, 19 is a p-type electrode). Type electrode), n-type region 15
Almost all the voltage is applied to the i-type region between the p-type region 16 and the p-type region 16, and the uniform refractive index distribution as shown in FIG. As a result of the application, the incident wavelength λ _op in the i-type region as shown in FIG.
Is reduced. In this case, if the length of the i-type region between the n-type region 15 and the p-type region 16 is set to 2 μm, a voltage of about 2 V may be applied to eliminate the excitons.

10A and 10B, when light is incident on the boundary where the refractive index changes shown by the broken line in FIG. 10A, the refractive index change and the optical waveguide are changed so that the incident angle of the incident light hν exceeds the total reflection angle. If the crossing angle of is set, reflection will occur at the boundary on the optical waveguide side to which voltage is applied, and light will be guided through the optical waveguide on the opposite side to which no voltage was applied, and optical switching operation will occur. It has been broken. The left side of FIG. 10B shows the n-type region 15 and the first p-type region.
FIG. 10 shows the path of light when a voltage is applied between the n-type region 15 and the second p-type region 17 on the right side of FIG. This shows the path of the light when it is done.

When a voltage is applied between the n-type region and the p-type region, the excitons disappear and the refractive index is, for example, 3.
540 to 3.520. Since the total reflection angle between the region having the refractive index of 3.540 and the region having the refractive index of 3.520 is about 84 °, it is necessary to guide the incident light to only one branch waveguide. The angle θ formed by the main waveguide and the branch waveguide may be set to 180 ° -84 ° (that is, 96 °) or less, and this angle θ may be set according to the total reflection angle.

FIG. 11 shows another proposal for improvement (Japanese Patent Laid-Open No. 62-296129) in which Schottky electrodes 20 and 21 are provided on an optical waveguide which also uses the MQW structure. is there. In this case, since the step of introducing impurities is not necessary, the optical switch can be formed by a simple step only by forming the electrodes.

[0017]

However, in the case of the semiconductor optical switch shown in FIG. 7, since the n-type region 15 and the p-type regions 16 and 17 are provided in the region where light propagates, the impurity-doped region is formed. N-type region 15 and p-type region 1
In Nos. 6 and 17, absorption loss of light occurred, and the optical switch characteristics were not always sufficient.

Although the semiconductor optical switch shown in FIG. 11 does not cause light absorption loss, when an electric field is applied in parallel to the MQW layer 7 in order to obtain a sufficient change in refractive index, the electrodes are structurally changed. Wider distance between 20, 21 is 2
It must be 5 μm. Then, it is necessary to apply a voltage to the region of this wide interval to apply a large voltage for obtaining a sufficient change in the refractive index, and eventually it becomes difficult to operate at a low voltage and downsize.

Accordingly, it is an object of the present invention to provide an optical switch that is small in size and capable of operating at a low voltage while eliminating absorption loss of light in an optical waveguide.

[0020]

FIG. 1 is a diagram for explaining the principle configuration of the present invention. The means for solving the problems of the present invention will be described with reference to FIG. A branch type semiconductor optical waveguide composed of two main optical waveguides 3 and two branch optical waveguides 4 and 5 is constituted by a multiple quantum well structure, and a region sandwiched between the branch optical waveguides is a one conductivity type region 10. The optical switch is characterized in that the regions facing the one conductivity type region 10 via the branched optical waveguides 4 and 5 are opposite conductivity type regions 11 and 12.

Further, the present invention is characterized in that light having a wavelength longer than the exciton absorption wavelength in the multiple quantum well structure is used as the light guided in the branched semiconductor optical waveguide, that is, the incident light hν.

Further, the present invention is characterized in that a semi-insulating substrate is used as a substrate on which the optical waveguide is provided, and further, in order to form the optical waveguide, a branched groove or a branched groove is formed in the cladding layer. A feature is that a mesa portion is provided.

[0023]

According to the present invention, the branched semiconductor optical waveguide is constructed by the multiple quantum well structure, and the n-type region and the p-type region for applying a reverse bias to the optical waveguide are provided on both sides of the optical waveguide region. Therefore, the light absorption loss is reduced.

Further, since the light having a wavelength longer than the exciton absorption wavelength in the multiple quantum well structure is used as the incident light hν, the change in the refractive index can be utilized to the maximum, and
The light absorption loss can be reduced.

Further, by using a semi-insulating substrate as the substrate, the parasitic capacitance can be reduced, and further, by providing a branched groove portion or a branched mesa portion in the cladding layer, a light confinement structure can be obtained by a simple manufacturing process. That is, a so-called pseudo buried heterojunction structure can be formed.

[0026]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The structure and manufacturing process of an optical switch in which a branched groove is provided in a clad layer, which is a first embodiment of the present invention, will be described with reference to FIGS.

Referring to FIG. 2, first, 5000 Å i-type In is formed on the semi-insulating InP substrate 1.
After depositing the P clad layer 2, a Y-shaped branched groove portion 6 composed of a main optical waveguide 3 and two branched optical waveguides 4 and 5 is formed by a normal photolithography process as shown in FIG. To form. Note that FIG.
It is sectional drawing when it cuts by the dashed-dotted line which connects between BB 'in (a), Moreover, FIG.2 (c) cut | disconnects by the dashed-dotted line which connects between CC' in FIG.2 (a). FIG.

Next, referring to FIG. 3, 100 Å I of I-type InP clad layer 2 is formed.
nP barrier layer and 100Å In _1-x Ga _x As _y P _1-y
A multiple quantum well (MQW) layer 7 composed of well layers (x = 0.438, y = 0.940), a 1000Å i-type InP clad layer 8, and an i-type InGaAs contact layer 9 are sequentially formed by MOVPE (or MOCVD). It is deposited by a crystal growth method such as a method. Thereby, in the step portion, both sides of the flat region of the groove bottom portion are surrounded by the side wall portions composed of the InP cladding layers 2 and 8 having a lower refractive index than the MQW layer 7 to form an optical waveguide. The period of the MQW layer 7 in this case is 20 periods. Note that FIGS. 3A and 3B correspond to FIGS. 2B and 2C, respectively.

Next, referring to FIG. 4, an n-type region 10 is formed by ion-implanting n-type impurities such as Sn and Si into the vicinity of the branch points of the branch optical waveguides 4 and 5 using a focused ion implanter. On the other hand, p-type impurities such as Zn are ion-implanted into a region facing the n-type region 10 via the branched optical waveguides 4 and 5 to p-type region 1.
1, 12 are formed. Note that FIG. 4B is the same as FIG.
6 is a cross-sectional view taken along the alternate long and short dash line connecting AA ′ in FIG. 6B. Impurities are ion-implanted so as to penetrate the MQW layer 7 and deeply reach the i-type InP cladding layer 2.

Next, referring to FIG. 5A, n made of AuGe / Au or the like is formed on the n-type region 10.
A mold electrode 13 is formed, and Ti is formed on the p-type region 11 (12).
The p-type electrode made of / Pt or the like is provided to complete the optical switch.

Next, the operation principle of this optical switch will be described with reference to FIG. See FIG. 5. When a reverse bias voltage is applied between the n-type region 10 and the p-type region 11 of this optical switch, the refractive index changes according to the same principle as in the conventional example shown in FIG. 9, and the reverse bias voltage is applied. The side branch optical waveguide (4 in FIG. 4) and the main optical waveguide (3 in FIG. 4) undergo total reflection, and the incident light hν is the side branch optical waveguide to which the reverse bias is not applied (5 in FIG. 4). Be guided to.

In the case of this embodiment, the exciton absorption wavelength in the MQW layer is about 1.526 μm, and when the exciton is extinguished, the maximum refractive index is around 1.54 μm (λ _{op in} FIG. 8). Therefore, the incident light hν used
When light having a wavelength of about 1.54 μm is used, the change in the refractive index can be utilized to the maximum and the light absorption loss can be reduced.

In the first embodiment described above, the clad layer 2 is provided with a branched groove portion. On the contrary, as shown in FIG. 6, the clad layer 2 is provided with a branched mesa portion. In this case, the optical waveguide portion is the flat portion on the top of the mesa portion, the optical confinement mechanism is formed by the step portion as in the first embodiment, and the other optical switching mechanism is also used. It is similar to the first embodiment. Note that FIG.
It is sectional drawing when it cuts by the dashed-dotted line which connects BB 'in (a), Moreover, FIG.6 (c) cut | disconnects by the dashed-dotted line which connects CC' in FIG.6 (a). FIG.

In the above embodiments, a single optical switch is described, but the present invention is a light receiving element such as a semiconductor laser as a light source and a pin photodiode as a light receiving portion, or a semiconductor optical modulator. Is also monolithically integrated with this optical switch to form a semiconductor optical integrated circuit device (optoelectronic IC), and in this case, there is an advantage that optical axis alignment becomes unnecessary.

Further, the present invention is not limited to the numerical range of the above-mentioned embodiment, but its layer thickness, the number of cycles, or the number of cycles, depending on the intended use, that is, the difference in the wavelength band of the light used. It is sufficient to set the branch angle of the branch optical waveguide, and the one conductivity type region 10 in the vicinity of the branch point has been described as the n type region. However, this region is the p type and the opposite conductivity type regions 11 and 12 are described. May be an n-type, and impurities are not limited to the listed elements.

The manufacturing process in the above embodiment is as follows.
A conventional optical switch shown in FIG. 7 (Japanese Patent Laid-Open No. 6-148699).
Since it is the same as the manufacturing process of (Japanese Patent Laid-Open Publication), it has not been described in detail, but the semiconductor layer deposition method is not limited to MOVPE, and the impurity introduction method may be a diffusion method in some cases. Further, when forming the electrodes, patterning may be performed using a normal photolithography process, or a lift-off method using a photoresist pattern may be used.

In the above embodiment, InGaA is used.
Although the explanation is given using the sP / InP system, GaAs / A
It is also possible to use 1 GaAs, in which case the substrate is semi-insulating GaAs and the cladding layer is i-type AlG.
The aAs layer, the barrier layer are AlGaAs or AlAs layers, the well layers are GaAs or AlGaAs layers, and the contact layers are GaAs layers. Further, the material is not limited to this, and other semiconductor material systems such as InAlAs system and InGaAlAs system used as MQW semiconductor lasers are naturally covered by the present invention.

[0038]

According to the present invention, since the n-type region and the p-type region are provided so as to be opposed to each other via the branched optical waveguide, the absorption loss of light in the optical waveguide is significantly reduced, and the optical switch is downsized. It is possible to increase the sensitivity as well as increase the sensitivity.

[Brief description of drawings]

FIG. 1 is a diagram illustrating a principle configuration of the present invention.

FIG. 2 is a diagram illustrating a substrate structure of a branch type optical switch that is a first embodiment of the present invention.

FIG. 3 is a diagram illustrating a laminated structure of a branch type optical switch that is a first embodiment of the present invention.

FIG. 4 is a diagram illustrating a structure of a reverse bias applying section of the branch type optical switch according to the first embodiment of the present invention.

FIG. 5 is a diagram illustrating a change in the refractive index in the optical waveguide when a reverse bias is applied to the branch type optical switch that is the first example of the present invention.

FIG. 6 is a diagram illustrating another embodiment of the present invention.

FIG. 7 is a diagram illustrating a structure of a conventional branch type optical switch.

FIG. 8 is a diagram illustrating the voltage dependence of refractive index versus incident light wavelength in the MQW layer.

FIG. 9 is a diagram illustrating a change in refractive index in an optical waveguide when a reverse bias is applied to a conventional branch type optical switch.

FIG. 10 is a diagram for explaining a state of optical waveguide when a reverse bias is applied to a conventional branch type optical switch.

FIG. 11 is a diagram illustrating a conventional optical switch using a Schottky electrode.

[Explanation of symbols]

 1 Semi-Insulating Substrate 2 Cladding Layer 3 Main Optical Waveguide 4 First Branching Optical Waveguide 5 Second Branching Optical Waveguide 6 Groove 7 Multiple Quantum Well (MQW) Layer 8 Cladding Layer 9 Contact Layer 10 n-type Region 11 First p-type region 12 second p-type region 13 n-type electrode 14 p-type electrode 15 n-type region 16 first p-type region 17 second p-type region 18 n-type electrode 19 p-type electrode 20 first Schottky Electrode 21 Second Schottky electrode

Claims (7)

[Claims] 1. A branch type semiconductor optical waveguide comprising one main optical waveguide and two branch optical waveguides is constituted by a multi-quantum well structure comprising a semi-insulating semiconductor layer, and said two branches are provided. An optical switch, wherein a region sandwiched by optical waveguides is a one-conductivity type region, and two regions facing the one-conductivity type region through the two branched optical waveguides are opposite conductivity type regions. 2. The optical waveguide is switched from the main optical waveguide to the branched optical waveguide by applying a reverse bias between the one conductivity type region and one of the opposite conductivity type regions. The semiconductor optical switch according to claim 1, wherein: 3. The optical switch according to claim 1, wherein the wavelength of light guided through the branched semiconductor optical waveguide is longer than the exciton absorption wavelength of the multiple quantum well structure. . 4. The optical switch according to claim 1, wherein a semi-insulating semiconductor substrate is used as a substrate on which the branch type semiconductor optical waveguide is provided. 5. A branch type groove provided in a cladding layer corresponding to the branch type semiconductor optical waveguide is used, and the multiple quantum well structure of a portion corresponding to the bottom surface of the groove is used as an optical waveguide. The optical switch according to any one of claims 1 to 4, which is characterized. 6. A branch type mesa portion provided in a clad layer corresponding to the branch type semiconductor optical waveguide is used, and the multiple quantum well structure of the portion corresponding to the top of the mesa portion is used as an optical waveguide. The optical switch according to any one of claims 1 to 4, wherein: 7. The monolithically integrated at least one of the branched semiconductor optical waveguide and a semiconductor laser for injecting light into the main optical waveguide or a semiconductor photodetector for detecting light from the branched waveguide. The optical switch according to any one of claims 1 to 6, which is characterized.