JP5144306B2 - Optical semiconductor device and manufacturing method thereof - Google Patents

Description

The present invention relates to an optical semiconductor device and a method for manufacturing the same .

  In a WDM (Wavelength Division Multiplexing) system in a metro network, there is a high demand for large capacity. In order to increase the capacity, it is necessary to greatly increase the number of channels. As the number of channels increases, the number of elements such as lasers and modulators used at each node increases accordingly. As the number of elements used increases, the required mounting volume capacity including components for mounting them increases. In addition, power consumption for operating a large number of optical components also increases.

  Therefore, development of a small and low power consumption optical modulator and development of a small module of a laser and an external optical modulator are being carried out.

  For example, it has been reported that an InP MZ (Mach-Zehnder) modulator has a smaller chip size, a lower driving voltage, and a wider wavelength band that can be modulated compared to an LN modulator or an EA modulator (non-patent document). Reference 1).

  In addition, a low-power consumption and compact wavelength tunable module in which an InP MZ optical modulator and a wavelength tunable laser are mounted on a multichip has been developed in recent years.

J. et al. S. Barton, et al. , "A Widely Tunable High-Speed Transmitter Using an Integrated SGDBR Laser-Semiconductor Optical Amplifier and Mach-Zehnder Modulator, JE. Sel. Topics Quantum Electron. , Vol. 9, no. 5, pp. 1113-1117, Sep. / Oct. 2003.

  However, since the spot diameter of the semiconductor InP MZ optical modulator described above is only 1/10 of the spot diameter of the optical fiber, the coupling loss does not become 1.5 dB or less, and the coupling loss is large.

  Further, there is a problem that the loss in the semiconductor InP MZ optical modulator (MZM) is as large as about 4 dB.

Accordingly, the present invention has been made in view of the above problems, and in manufacturing a multichip module of a laser and an optical modulator, the coupling loss between the optical fiber and the optical modulator is compensated, and a semiconductor InP Mach-Zehnder modulator is provided. It is an object of the present invention to provide an optical semiconductor device and a method for manufacturing the same , which compensate for losses in the semiconductor device and increase the output power from the device.

An optical semiconductor device according to a first invention for solving the above-described problem is
On the (100) surface of the same n-type compound semiconductor substrate having a zinc blende structure, a semiconductor optical amplifier having a p-type surface layer and having a npin layer structure and a surface layer having a p-type surface layer An n-type Mach-Zehnder modulator made of a high mesa structure is an integrated optical semiconductor device formed and guided in the forward mesa direction in the forward mesa direction. And
The Mach-Zehnder modulator is
And two wavelength division means for coupling the branch to the light, and two optical waveguides connecting one output of the wavelength division means and the other an input unit of the wavelength division means each of the two optical waveguides Two phase control means provided in each,
A p-type surface layer identical to the p-type surface layer of the semiconductor optical amplifier is formed between the two optical coupling / branching means and the two phase control means , and the p-type surface layer A non-p-type layer is formed between the p-layer in the npin layer structure .

An optical semiconductor device according to a second invention for solving the above-described problem is an optical semiconductor device according to the first invention,
An interface between the semiconductor optical amplifier and the Mach-Zehnder modulator, or an interface between the Mach-Zehnder modulator and the p-type surface layer is formed to extend in a direction inclined with respect to the reverse mesa direction. And

An optical semiconductor device according to a third invention for solving the above-mentioned problem is an optical semiconductor device according to the second invention,
An inclination angle of the interface is larger than 0 degree and smaller than 45 degrees.

An optical semiconductor device according to a fourth invention for solving the above-described problem is an optical semiconductor device according to any one of the first to third inventions,
The Mach-Zehnder modulator is
A first 2 × 2 MMI coupler having two inputs and two outputs;
A second 2 × 2 MMI coupler having two inputs and two outputs;
Two optical waveguides respectively connecting the output part of the first 2 × 2 MMI coupler and the input part of the second 2 × 2 MMI coupler;
Comprising a two phase control means provided in each of the two optical waveguides,
The semiconductor optical amplifier is formed at an input portion of the first 2 × 2 MMI coupler.

An optical semiconductor device according to a fifth invention for solving the above-described problem is the optical semiconductor device according to any one of the first to third inventions,
The Mach-Zehnder modulator is
A first 2 × 2 MMI coupler having two inputs and two outputs;
A second 2 × 2 MMI coupler having two inputs and two outputs;
Two optical waveguides respectively connecting the output part of the first 2 × 2 MMI coupler and the input part of the second 2 × 2 MMI coupler;
Comprising a two phase control means provided in each of the two optical waveguides,
The semiconductor optical amplifier is formed at an input portion of the second 2 × 2 MMI coupler.
Furthermore, an invention of a method of manufacturing an optical semiconductor device according to the sixth invention for solving the above-described problem is as follows.
A method of manufacturing one of the first to fifth optical semiconductor devices,
In forming a non-p-type layer between the p-type surface layer and the p-layer in the npin layer structure,
On the substrate, a p-type surface layer of the semiconductor optical amplifier is laminated partway, and an etch stop layer and an InP cover layer are sequentially formed on the pin layer structure in the npin layer structure from the bottom in a portion other than the semiconductor optical amplifier. And a p-type surface layer identical to the p-type surface layer of the semiconductor optical amplifier is formed on the entire surface.
Thereafter, in a portion excluding the portion of the semiconductor optical amplifier and the portion where the p-type surface layer between the two optical coupling / branching means and the two phase control means is to be formed, the uppermost layer To the etch stop layer,
A new n layer is laminated on the removed portion, so that the etch stop layer and the InP cover are formed as a non-p-type layer between the p-type surface layer and the p-layer in the npin layer structure. Layer and formed state
It is characterized by that.

  According to the optical semiconductor device of the present invention, when a multi-chip module of a laser and an optical modulator is manufactured, the coupling loss between the optical fiber and the optical modulator can be compensated, and the loss in the semiconductor InP Mach-Zehnder modulator is compensated. it can. Furthermore, the output power from the device can be increased.

  The best mode for carrying out an optical semiconductor device according to the present invention will be described in detail based on examples.

A first embodiment of the optical semiconductor device according to the present invention will be specifically described with reference to FIGS.
FIG. 1 is a diagram schematically showing an optical semiconductor device. FIG. 2 is a diagram for explaining a layer structure of an SOA (semiconductor optical amplifier) and MZM included in the optical semiconductor device. FIG. 2A schematically shows a state in which these layers are connected. FIG. 2B shows a cross section of the SOA, and FIG. 2C shows a cross section of the MZM.

  As shown in FIG. 1, an optical semiconductor device 100 according to the present embodiment is provided on one semiconductor substrate 1 and includes first and second 2 × 2 MMI couplers (optical coupling / branching means) 10. , 20 and two phase control regions 30, 40 provided between the first 2 × 2 MMI coupler 10 and the second 2 × 2 MMI coupler 20. These elements (first and second 2 × 2 MMI couplers 10, 20 and two phase control regions (phase modulation regions) 30, 40) are manufactured by an MZM (Mach-Zehnder modulator) 60 having a high mesa structure.

The first 2 × 2 MMI coupler 10 includes two input units 11 and 12 and two output units 13 and 14. The second 2 × 2 MMI coupler 20 also includes two input units 21 and 22 and two output units 23 and 24. One phase control region (second phase control region) 30 includes one output unit 13 of the first 2 × 2 MMI coupler 10 and one input unit 21 of the second 2 × 2 MMI coupler 20. Is provided with a second control electrode 32 for applying a voltage to the second phase control optical waveguide 31. The other phase modulation region (first phase control region) 40 connects the other output unit 14 of the first 2 × 2 MMI coupler 10 and the other input unit 22 of the second 2 × 2 MMI coupler 20. A first control electrode 42 for applying a voltage to the first phase control optical waveguide 41 is provided. That is, the control electrodes 32 and 42 form phase control means. A ridge-structured SOA (Semiconductor Optical Amplifier) 50 is fabricated in the vicinity of the input port 1 a that guides the signal light λ 1 to one input portion 11 of the first 2 × 2 MMI coupler 10. Yes. That is, the SOA 50 is monolithically integrated in the MZM 60.

  Here, as shown in FIG. 2B, the ridge structure is obtained by mesa-processing only the upper layers 55 and 56 of the active layer 53 without mesa-processing the active layer 53. On the other hand, as shown in FIG. 2C, the high mesa structure is a structure in which the active layer 62 is also mesa-processed and the active layer side wall is exposed to the atmosphere. The substrate is (100) InP. The SOA 50 and MZM 60 are fabricated in the forward mesa direction ([011] direction) with respect to the n-InP substrate 1. Since the Pockels effect is large in this direction due to the fabrication in the forward mesa direction ([011] direction), the modulation characteristics of MZM are improved.

The core layer of the MZM 60 is a multiple quantum well (MQW) 62, the quantum well layers constituting the MQW are 20 layers, InGaAsP (layer thickness: 10 nm), the barrier layers are 19 layers, and InP (layer thickness: 10 nm). is there. The composition of MQW 62 is adjusted so that the photoluminescence wavelength of MQW is 1.43 μm. The mesa width W _{1 of the} MZM 60 is 2.0 μm, and its height H ₁ is 3.5 μm.

On the other hand, the active layer of the SOA 50 is a multiple quantum well (MQW) 53, the quantum well layers constituting the MQW are eight layers, InGaAsP (layer thickness: 5 nm), the barrier layers are seven layers, and InGaAsP (photoluminescence wavelength: 1.25 μm, layer thickness: 10 nm). The quantum well composition is adjusted so that the photoluminescence wavelength of MQW is 1.55 μm. Note that i-InGaAsP SCH layers 54 and 52 are formed above and below the MQW 53, and a p-InP layer 55 is formed above the i-InGaAsP SCH layer 54. A p-InGaAsP layer 56 is formed on the p-InP layer 55. The SOA 50 has a ridge width W ₂ of 2.0 μm and a length L ₂ of 600 μm.

  In the optical semiconductor device 100 described above, when the light λ1 is input from the input port 1a, it is amplified by the SOA 50 and is equally distributed to the two phase control optical waveguides 31 and 41 by the first 2 × 2 MMI coupler 10. Is done. Then, the first and second control electrodes (lumped constant type electrodes) 42 and 32 individually apply voltages to the phase control optical waveguides 41 and 31 to change the phase state of the light. The lights that have passed through the two phase control optical waveguides 41 and 31 are combined by the second 2 × 2 MMI coupler 20, and the upper and lower output ports 1d and 1c (upper and lower are shown in FIG. Lights λ3 and λ2 are output from above and below.

Here, FIG. 3 shows extinction characteristics when the voltage applied to the first control electrode 42 is Vb1 and the voltage applied to the second control electrode 32 is Vb2 = 0V. In FIG. 3, Bar port represents the lower output port 1c, and Cross port represents the upper output port 1d.
As shown in FIG. 3, by changing the voltage Vb1 and changing the phase of the light passing through the first phase control optical waveguide 41, the first and second phase control optical waveguides 41 and 31 are changed. The phase of the combined light of the passed light also changes, and the ratio at which the light is output from the Bar port (output port 1c) or the Cross port (output port 1d) changes.

In addition, the voltage applied to the first control electrode 42 is Vb1 = 0 V, the voltage applied to the second control electrode 32 is Vb2 = 0 V, and the injection current to the SOA 50 is 50 mA, 100 mA, 150 mA, FIG. 4 shows the device input power dependence of the output power from the Bar port (output port 1c) when changing to 200 mA. In FIG. 4, the horizontal axis represents the device input power, and the vertical axis represents the device output power.
As shown in FIG. 4, by amplifying light by the SOA 50, when the SOA injection current is 100 mA or more, power that is 8 dB or more larger than the input power is output. This indicates that not only the loss in the MZM 60 is compensated but also the coupling loss between the device of about 8 dB and the optical fiber can be compensated.
Therefore, by monolithically integrating the SOA 50, the output power from the output ports 1c and 1d increases.

When the device input power is −8.6 dBm, the voltage applied to the first control electrode 42 is Vb1 = 0V, and the voltage applied to the second control electrode 32 is Vb2 = 0V, Bar FIG. 5 shows the result of measuring the SOA injection current dependence of the output from the port (output port 1c) over the entire C band. In FIG. 5, the horizontal axis represents the injection current into the SOA, and the vertical axis represents the device output power.
As shown in FIG. 5, it can be seen that the wavelength dependence is small throughout the C band. This is because the photoluminescence wavelength of the MZ modulator is about 1.43 μm, so the C band (1530 nm to 1565 nm) is longer than the band edge, and the wavelength dependence of the absorption coefficient and refractive index change with respect to voltage is small. is there.

Here, FIG. 6 shows an eye pattern of an output light waveform of an NRZ signal (9.9532 Gb / s, PRBS: 2 ³¹ −1) from 1530 nm to 1560 nm. 6A is a diagram when the input light wavelength λ is 1530 nm, FIG. 6B is a diagram when the input light wavelength λ is 1540 nm, and FIG. 6C is a diagram when the input light wavelength λ is FIG. 6D is a diagram in the case of 1550 nm, and FIG. 6D is a diagram in the case where the input light wavelength λ is 1560 nm. The drive voltage is 3 Vpp, the bias voltage to the electrode 42 is −8.5 V, the bias voltage to the electrode 32 is 0 V, the input optical power is −4.6 dBm, and the SOA injection current is 100 mA.
As shown in FIG. 6, it was found that almost the same good waveform quality was obtained in the entire C band, and the extinction ratio was 9.6 dBm.

Further, in FIG. 7, 0 km (back-to-back notation in the figure) of an NRZ signal (10 Gbit / s) from 1530 nm to 1560 nm, a code error rate (bit error rate: BER) indicating a single mode fiber 100 km transmission characteristic, and The relationship is shown.
As shown in FIG. 7, it was found that an error-free operation was observed in the entire C band, and that the power penalty was good when the BER was ^{10-12 and} less than 1.5 dB.

  Here, a method for manufacturing the SOA 50-MZM 60 joint in the optical semiconductor device 100 described above will be described with reference to FIGS. The element shape processing and electrode formation of the SOA 50 and MZM 60 are performed by a normal semiconductor device process.

First, as shown in FIG. 8A, the upper p-InP layer 59 is stacked by MOVPE (Metal-organic vapor phase epitaxy), and then the SiO ₂ film 71 is formed on the SOA 50. Then, the SiO ₂ film 71 is used as a mask to etch (remove) the MZM 60 up to the i-InGaAsP SCH layer 52. That is, in the SOA 50, an n-InP layer 51, an i-InGaAsP SCH layer 52, an i-MQW layer 53, an i-InGaAsP SCH layer 54, and a p-InP layer 59 are formed in this order from the lower layer. An SiO ₂ film 71 is formed on the layer 59. On the other hand, in the MZM 60, only the n-InP layer 61 is left. Note that the layer thickness of the p-InP layer 59 in the SOA 50 is 0.5 μm.

  Subsequently, as shown in FIG. 8B, only the region of MZM 60 is laminated to the core layer by MOVPE (selectively grown). That is, in the MZM 60, an i-MQW layer (layer thickness: 0.4 μm) 62 made of InGaAsP / InP (10 nm / 10 nm) is formed on the n-InP layer 61, and the i-MQW layer 62 is formed with i An InP layer (layer thickness: 0.3 μm) 63 is formed.

  Subsequently, as shown in FIG. 8C, the p layer, the n layer, the etch stop layer, and the cover layer are laminated on the MZM 60 only by MOVPE. That is, in the MZM 60, a material having a band gap larger than that of InP, for example, a p-InGaAlAs layer or a p-InAlAs layer (layer thickness: 50 nm) 64, an n-InGaAsP layer in order from the lower layer side on the i-InP layer 63. (Layer thickness: 50 nm) 65, an n-InP etch stop layer 69, an InGaAsP etch stop layer 301, and an InP cover layer 302 are formed.

Subsequently, as shown in FIG. 9A, the SiO ₂ film 71 of the SOA 50 is removed, and a p layer and a cover layer are laminated on the entire surface (SOA 50 and MZM 60). This is because the MZM 60 is extremely large compared to the SOA 50, and therefore only the SOA 50 cannot be selectively grown. That is, in the SOA 50, a p-InP layer (layer thickness: 1.2 μm) 55, a p-InGaAsP / InGaAs cap layer 56, and an InP cover layer 72 are formed on the p-InP layer 59 in order from the lower layer side. . This p-InP layer 55 becomes a cladding layer by the SOA 50. The p-InP layer produced in this step has the same composition as that of the p-InP layer 59 produced on the i-InGaAsP SCH layer 54. The layer including this layer is denoted by reference numeral 55 in FIG. Indicated. On the other hand, in the MZM 60, a p-InP layer (layer thickness: 1.2 μm) 303, a p-InGaAsP / InGaAs cap layer 304, and an InP cover layer 305 are formed on the InP cover layer 302 in order from the lower layer side.

Subsequently, as shown in FIG. 9B, after the InP cover layer 72 is removed and the SiO ₂ mask 73 is formed in the SOA 50, the p layer stacked on the MZM 60 is all removed by etching. That is, in the MZM 60, the InGaAsP etch stop layer 301, the InP cover layer 302, the p-InP layer 303, the p-InGaAsP / InGaAs cap layer 304, and the InP cover layer 305, which are layers above the n-InP etch stop layer 69, are provided. Removed.

Subsequently, as illustrated in FIG. 9C, the n layer and the cover layer are laminated on the MZM 60. That is, in the MZM 60, an n-InP layer (layer thickness: 1.3 μm) 66, an n-InGaAs (P) cap layer 67, and an InP cover layer 68 are formed on the n-InP etch stop layer 69 in order from the lower layer side. It is formed. The n-InP layer produced in this step has the same composition as that of the n-InP etch stop layer 69, and this layer is indicated by reference numeral 66 in FIG. The MZM 60 is manufactured with an npin layer structure. Thereby, the loss of the electric signal and the optical signal is smaller than that of the pin type and the electric field application efficiency is high, so that it is small and can be operated at a low driving voltage.
Therefore, the above-described optical semiconductor device 100 can be manufactured.

  Therefore, according to the optical semiconductor device 100 according to the present embodiment, the SOA 50 and the MZM 60 are integrated in the direction in which light is guided on the same substrate 1, so that a multichip module of a laser and an optical modulator can be obtained. In manufacturing, the coupling loss between the optical fiber and the optical modulator can be compensated, and the loss in the semiconductor InP Mach-Zehnder modulator can be compensated. Furthermore, the output power from the device can be increased.

  In this embodiment, InP is used for the substrate and InGaAs (P) is used in a stacked manner. However, if the zinc blende structure is the same crystal structure, the same Pockels effect is obtained. Therefore, GaAs, ZnTe, ZnS, or the like can be used for the substrate. Further, InAl (Ga) As, GaInNAs, AlGaAs, InAlGaP, ZnSeTe, or the like can be stacked on the substrate.

  In this embodiment, the wavelength band of 1.55 μm band is targeted, but by changing the composition of InGaAsP used for the quantum well layer, it is possible to cope with the long wavelength band of 1.1-1.65 μm. Further, by stacking AlGaAs, InAlGaP, ZnSeTe or the like on a substrate such as GaAs, ZnTe, or ZnS, it is possible to cope with other wavelength bands.

A second embodiment of the optical semiconductor device according to the present invention will be specifically described with reference to FIGS.
FIG. 10 is a diagram schematically showing the optical semiconductor device, and FIG. 11 is a diagram for explaining the layer structure of SOA and MZM included in the optical semiconductor device. In FIG. 10, λ11 indicates light input to the input port 1a, λ12 indicates light output from the output port 1d, and λ13 indicates light output from the output port 1c.

  The optical semiconductor device according to this embodiment has the same element configuration (first and second 2 × 2 MMI couplers 10 and 20 and phase control regions 30 and 40 as those of the optical semiconductor device according to the first embodiment described above. ), And an SOA is applied to the input part of the second 2 × 2 MMI coupler 20. That is, the SOA having the ridge structure included in the optical semiconductor device according to the present embodiment is inserted between the MZM having the high mesa structure. In this embodiment, the same elements as those in the optical semiconductor device according to the first embodiment described above are denoted by the same reference numerals, and the description thereof is omitted. In addition, the layer structure of each of the SOA and MZM included in the optical semiconductor device according to the present embodiment and the fabrication process of the SOA-MZM coupling portion are the same as those of the optical semiconductor device according to the first embodiment described above. Description is omitted.

  The optical semiconductor device 200 according to the present embodiment is an apparatus including SOAs 250a and 250b provided in the input units 21 and 22 of the second 2 × 2 MMI coupler 20, as shown in FIG. Specifically, as shown in FIG. 11, the apparatus includes SOAs 250a and 250b provided between the MZMs 60.

As described with reference to FIG. 4 in the first embodiment, the output power from the SOA increases as the SOA input power increases and increases as the injection current into the SOA increases, but saturates at a certain level. . From the graph of FIG. 4, even if the SOA injection current and the input power are increased, the maximum device output is predicted to be about 7 to 8 dBm. Here, in the optical semiconductor device configured as shown in FIG. 1 (an optical semiconductor device in which an SOA is fabricated at the input portion of the first 2 × 2 MMI coupler), the maximum output of the device is derived from the saturation output power of the SOA. The value is obtained by subtracting the loss of light in the waveguide.
Therefore, the maximum output power of the device greatly depends on the saturation output power of the SOA.

Here, FIG. 12 shows a level diagram when the pulse light is propagated to the device (optical semiconductor device).
As shown in FIG. 12, when an SOA is manufactured at the input portion of the first 2 × 2 MMI coupler, light input to the device is amplified by the SOA (the gain at this time is 9 dB), but is propagated. Due to the loss, the output from the device is a loss-1 dB less than the SOA output.
Therefore, in order to avoid a decrease in output due to light propagation loss, generally, when optical amplification is performed using SOA, the SOA is installed at a position closest to the output end.

  Therefore, when an SOA is produced not at the input portion of the first 2 × 2 MMI coupler closest to the output end but at the output portion of the second 2 × 2 MMI coupler, that is, the output portions 23 and 24, the light enters the SOA. The light loss at this time is −1.5 dB and becomes smaller than the input value of the device due to the waveguide propagation loss. However, since the gain is larger when the SOA input is smaller due to gain saturation, the SOA output is the same as in the configuration of FIG. A value not much different from the SOA output value is obtained, and the device output is larger than that in the configuration of FIG. 1 because the optical propagation loss after the SOA output is small.

  However, in the actual use state, in the configuration of FIG. 1, the light input to the SOA is CW light. However, when the SOA is produced at the output portion of the second 2 × 2 MMI coupler, the intensity-modulated light is not generated. Will be input.

  For this reason, in the latter case, the pattern effect causes a pattern effect in which the output of one level differs depending on the pattern of the modulation signal, and the waveform deteriorates.

  Thus, FIG. 1 shows a configuration for performing optical amplification before intensity modulation so as not to be affected by the pattern effect, and FIG. 10 shows a configuration for obtaining a larger device output than the configuration of FIG.

  In the case of the configuration shown in FIG. 10, the light incident on the device is equally divided into the phase control optical waveguides 31 and 41 by the first 2 × 2 MMI coupler 10. It is amplified by the SOAs 250a and 250b manufactured in the No. 41.

Then, the lights amplified by the respective SOAs 250a and 250b are multiplexed by the second 2 × 2 MMI coupler 20 and output from the output ports 1c and 1d.
In the case of this configuration, when the output from the SOAs 250a and 250b produced in the both phase controlling optical waveguides 31 and 41 is maximum (saturated output power), the second 2 × 2 MMI coupler 20 uses the both phase controlling optical waveguides. As a result of combining the light propagated through 31 and 41, the device output becomes twice the value of the saturated output power.

FIG. 13 shows the relationship between the device input / output power and the SOA input / output power of the configuration shown in FIG.
As shown in FIG. 13, it can be seen that the maximum device output is about 9 to 10 dBm, which is twice that of FIG.

  Therefore, according to the optical semiconductor device 200 according to the present embodiment, since the SOAs 250a and 250b are produced in the input portions 21 and 22 of the second 2 × 2 MMI coupler, the output power from the device can be increased. .

A third embodiment of the optical semiconductor device according to the present invention will be specifically described with reference to FIGS.
FIG. 14 is a diagram for explaining an insulation state between optical waveguides in a conventional MZM, and FIG. 15 is a diagram for explaining an MZM included in the optical semiconductor device. FIG. 16 is a diagram for explaining a method for manufacturing the optical semiconductor device. FIG. 17 is a diagram schematically showing the optical semiconductor device. FIG. 17A is a plan view thereof, and FIG. 17B is an output section of the first 2 × 2 MMI coupler included in the optical semiconductor device. Shows an enlargement of In FIG. 17, λ31 indicates light input to the input port 1a, λ32 indicates light output from the output port 1c, and λ33 indicates light output from the output port 1d.

  The devices (optical semiconductor devices) 100 and 200 manufactured in the first and second embodiments described above have n-InP between the first phase control optical waveguide 41 and the second phase control optical waveguide 31. Since the layers 66 are electrically connected to each other, it becomes difficult to individually control the phase of light propagating through the phase control optical waveguides 31 and 41 through the control electrodes 32 and 42 with a voltage.

In order to solve this, it is necessary to increase the resistance between both phase control optical waveguides of the MZM.
Conventionally, as shown in FIG. 14, in order to increase the resistance between the first and second phase control optical waveguides, a separation groove 81 is formed in the n-InP layer 66, and the first and second phase control are performed. The optical waveguides for use are insulated by air. That is, as shown in FIG. 14, the MZM region 80 includes an n-InP layer 61, i-MQW (layer thickness: 0.4 μm) 62, i-InP (layer thickness: 0.3 μm) in order from the lower layer side. 63, P-type layer (layer thickness: 50 nm) 64 having a larger band gap than InP, n-InGaAsP layer (layer thickness: 50 nm) 65, n-InP layer (layer thickness: 1.3 μm) 66, n− A separation groove 81 having a depth from the InGaAs (P) cap layer 67 and the InP cover layer 68 to the InP cover layer 68, the n-InGaAs (P) cap layer 67, and the n-InP layer 66 is formed. .

  However, in the conventional method, when the propagating light λ20 passes through the groove 81, the mode spreads in the air. Therefore, when the light λ20 recombines with the n-InP layer 66, it cannot be coupled 100%, resulting in a large optical loss. .

In this embodiment, in order to increase the resistance between the first and second phase control optical waveguides, a p-type layer is provided on part of the n-type surface layer of the first and second phase control optical waveguides. A layer (layer structure 310) is inserted.
As shown in FIG. 15, the MZM 360 included in the optical semiconductor device according to this example is an MZM 80 in which a conventional separation groove 81 is formed, and a layer structure 310 is formed in the separation groove 81. The layer structure 310 includes an InGaAsP etch stop layer 301, an InP cover layer 302, a p-InP layer 303, an InGaAsP / InGaAs cap layer 304, and an SiO ₂ film 306 in order from the lower layer side.

  Hereinafter, a method for manufacturing the optical semiconductor device according to this example will be described with reference to FIGS. The manufacturing method of the optical semiconductor device according to the present example is the same as the manufacturing method of the optical semiconductor device according to the first example described above up to FIG. That is, the fabrication of the optical semiconductor device according to this example is performed after the fabrication process of the optical semiconductor device shown in FIGS. 8A to 8C and FIG. A manufacturing process is performed.

As shown in FIG. 16A, after etching the p layer of MZM 360, after removing the InP cover 305 shown in FIG. 9A, a SiO ₂ mask 306 is formed on a part of the p layer surface, Leave a part without removing it by etching. That is, in a place where the SiO ₂ mask 306 is not formed, the InGaAsP etch stop layer 301, the InP cover layer 302, the p-InP layer 303, the p-InGaAsP / P, which are layers above the n-InP etch stop layer 69, are formed. The InGaAs cap layer 304 is removed.

  Subsequently, as illustrated in FIG. 16B, an n-InP layer, a cap layer, and a cover layer are selectively grown on the MZM 360 other than the region of the layer structure 310. That is, an n-InP layer (layer thickness: 1.3 μm) 66, an n-InGaAs (P) cap layer 67, and an InP cover layer 68 are formed on the n-InP etch stop layer 69 in order from the lower layer side. . The n-InP layer produced by selective growth in this step has the same composition as that of the n-InP etch stop layer 69, and this layer is indicated by reference numeral 66 in FIG.

  Therefore, according to the manufacturing method of the optical semiconductor device described above, the p-type layer of the layer structure 310 can be stacked at the same time when the p-type surface layer of the SOA 50 is stacked. Therefore, the manufacturing process itself is simplified, and the manufacturing cost can be reduced.

  Here, the optical semiconductor device 300 according to this example is an apparatus to which the above-described layer structure 310 is applied. In the element configuration shown in the above-described first example, as shown in FIG. In this device, layer structures 310a, 310b, 310c, and 310d are respectively formed at four locations of the output units 13 and 14 of the 2 × 2 MMI coupler 10 and the input units 21 and 22 of the second 2 × 2 MMI coupler 20. As a result, the p-InP layer 303 is introduced into the n-InP layer 66 between the first and second phase control optical waveguides 41 and 31, and the first phase control optical waveguide 41 and the second phase control optical waveguide 41 and 31 are introduced. Since an npn layer structure is formed between the control optical waveguide 31 and the n-layer of the first phase control optical waveguide 31 and the n layer of the second phase control optical waveguide 31, there is a potential difference. When this occurs, one of the two n / p-n p / n interfaces is always reverse-biased, resulting in a high resistance regardless of the polarity of the potential difference. As a result, light propagation loss can be significantly reduced. In addition, electrolysis tends to concentrate on the corner of the separation groove, which tends to cause device breakdown. However, by introducing the p-InP layer 303 and increasing the resistance, electrolysis concentration does not occur. It is also excellent in terms of reliability.

  In the present embodiment, the optical semiconductor device 300 in which the layer structures 310a, 310b, 310c, and 310d are fabricated in a total of four locations has been described. However, the output units 13 and 14 of the first 2 × 2 MMI coupler 10 are not described. It is also possible to provide an optical semiconductor device in which the layer structure 310 is produced in a total of two places, any one place and any one of the input portions 21 and 22 of the second 2 × 2 MMI coupler 20. Even in such an optical semiconductor device, similarly to the optical semiconductor device 300 described above, there is an effect that the resistance between the first phase control optical waveguide 41 and the second phase control optical waveguide 31 becomes high. can get.

  In the above description, the layer structures 310a, 310b, 310c, and 310d have been described using the optical semiconductor device 300 formed over the waveguide width in the reverse mesa direction. The optical semiconductor device is formed so as to have a high resistance (electrically insulated) between the first phase control optical waveguide 41 and the second phase control optical waveguide 31 formed over the width. Is also possible.

A fourth embodiment of the optical semiconductor device according to the present invention will be specifically described with reference to FIG.
FIG. 18 is a diagram schematically showing an optical semiconductor device. In FIG. 18, λ41 indicates light input to the input port 1a, λ42 indicates light output from the output port 1c, and λ43 indicates light output from the output port 1d.

  An optical semiconductor device 400 according to the present embodiment is an apparatus to which the layer structure 310 included in the optical semiconductor device 300 according to the third embodiment described above is applied, and in the element configuration shown in the second embodiment described above. 18, the input portions 21 and 22 of the second 2 × 2 MMI coupler 20 are increased in resistance by the p-type surface layers of the SOAs 250a and 250b, so that the first 2 × 2 MMI coupler 10 In this device, layer structures 410a and 410b are produced at two locations of the output sections 13 and 14, respectively. As a result, the p-InP layer 303 is introduced into the n-InP layer 66 between the first and second phase control optical waveguides 41 and 31, and the first phase control optical waveguide 41 and the second phase control optical waveguide 41 and 31 are introduced. The resistance between the control optical waveguides 31 is increased (electrically insulated).

  As described above, according to the optical semiconductor device 400 according to the present example, by introducing the p-InP layer 303 and increasing the resistance, the light propagation loss can be significantly reduced.

  Electrolysis tends to concentrate on the corners of the separation groove, which tends to cause device breakdown. However, by introducing the p-InP layer 303 and increasing the resistance, the concentration of electrolysis does not occur and reliability is improved. Is also excellent. The p-type layers of the layer structures 410a and 410b included in the above-described optical semiconductor device 400 can be manufactured by laminating simultaneously with the p-type surface layers of the SOAs 250a and 250b. Therefore, the manufacturing process itself is simplified, and the manufacturing cost can be reduced.

  In the present embodiment, the layer structures 410a and 410b are described using the optical semiconductor device 400 formed over the waveguide width in the reverse mesa direction, but the waveguide is not required to be in the waveguide direction and the reverse mesa direction. The optical semiconductor device is formed so as to have a high resistance (electrically insulated) between the first phase control optical waveguide 41 and the second phase control optical waveguide 31 formed over the width. Is also possible.

A fifth embodiment of the optical semiconductor device according to the present invention will be specifically described with reference to FIG.
FIG. 19 is a plan view schematically showing a connection location (joining location) between the SOA and MZM included in the optical semiconductor device. In FIG. 19, arrow A indicates the forward mesa direction with respect to the substrate, and arrow B indicates the reverse mesa direction with respect to the substrate.
The optical semiconductor device according to this example is obtained by changing the interface where the SOA and the MZM manufactured by the above-described optical semiconductor device are connected.

  In the optical semiconductor device 500 according to the present embodiment, as shown in FIG. 19, the coupling interface (coupling portion) 501 between the SOA 50 and the MZM 60 is from a direction perpendicular to the element direction (forward mesa direction) (reverse mesa direction). Inclined.

  That is, in the optical semiconductor device 500 described above, when the coupling interface 501 between the SOA 50 and the MZM 60 is formed in a direction (reverse mesa direction) perpendicular to the element direction (forward mesa direction), for example, light incident on the coupling interface, A part of the light emitted from the SOA 50 and incident on the MZM 60 is reflected at the coupling interface, and may be (re-) entered on the SOA as return light to promote the deterioration of the SOA. Therefore, it is possible to prevent the return light from entering the SOA 50 by inclining the coupling interface 501 from the reverse mesa direction.

  Here, the effect of suppressing the influence of the return light is obtained regardless of the inclination angle θ of the bonding interface 501, but abnormal growth is suppressed during crystal regrowth when the bonding interface is in the reverse mesa direction. The inclination angle from the reverse mesa direction should be small. On the other hand, a larger tilt angle θ in the reverse mesa direction is more effective in preventing the return light from entering the SOA 50. Therefore, the inclination angle θ from the reverse mesa direction is effective when it is larger than 0 degree and not larger than 45 degrees.

  In the above description, the optical semiconductor device 500 applied to the coupling interface 501 between the SOA 50 and the MZM 60 has been described. However, the optical semiconductor applied to the coupling interface between the layer structures 310a and 310b of the optical semiconductor device 300 and the MZM 360 described above. An optical semiconductor device applied to the coupling interface between the layer structures 410a, 410b, 410c, 410d of the optical semiconductor device 400 and the MZM 60 described above can be used. Even these optical semiconductor devices have the same operational effects as the optical semiconductor device 500 described above.

1 is a diagram schematically showing an optical semiconductor device according to a first embodiment of the present invention. It is a figure for demonstrating the layer structure of SOA and MZM which the optical semiconductor device based on 1st Example of this invention comprises. It is a graph which shows the quenching characteristic in MZM of the optical semiconductor device which concerns on 1st Example of this invention. It is a graph which shows the relationship between the input power and output power in the optical semiconductor device which concerns on 1st Example of this invention. 4 is a graph showing the relationship between the current injected into the SOA and the output power in the optical semiconductor device according to the first example of the present invention. It is a figure which shows the eye pattern of the output waveform of the NRZ signal of each wavelength. It is a graph which shows the transmission characteristic of the NRZ signal of each wavelength. It is a figure for demonstrating the manufacturing method (the first half) of the optical semiconductor device which concerns on 1st Example of this invention. It is a figure for demonstrating the manufacturing method (latter half) of the optical semiconductor device which concerns on 1st Example of this invention. It is the figure which showed typically the optical semiconductor device which concerns on the 2nd Example of this invention. It is a figure for demonstrating the layer structure of SOA and MZM which the optical semiconductor device based on 2nd Example of this invention comprises. It is the figure which showed the level diagram at the time of propagating pulsed light to a device. It is a figure which shows the relationship between the device input / output power and SOA input / output power in the optical semiconductor device based on 2nd Example of this invention. It is a figure for demonstrating the insulation state between the optical waveguides in the conventional MZM. It is a figure for demonstrating MZM which the optical semiconductor device which concerns on the 3rd Example of this invention comprises. It is a figure for demonstrating the manufacturing method of the optical semiconductor device which concerns on the 3rd Example of this invention. It is a figure which shows typically the optical semiconductor device which concerns on the 3rd Example of this invention. It is a figure which shows typically the optical semiconductor device which concerns on the 4th Example of this invention. It is the top view which showed typically the connection location of SOA and MZM which the optical semiconductor device which concerns on the 5th Example of this invention comprises.

Explanation of symbols

10, 20 2 × 2 MMI coupler 30, 40 Phase control region 50 Semiconductor optical amplifier 60 Mach-Zehnder modulator 310 Layer structure 100, 200, 300, 400, 500 Optical semiconductor device

Claims (6)

On the (100) surface of the same n-type compound semiconductor substrate having a zinc blende structure, a semiconductor optical amplifier having a p-type surface layer and having a npin layer structure and a surface layer having a p-type surface layer An n-type Mach-Zehnder modulator made of a high mesa structure is an integrated optical semiconductor device formed and guided in the forward mesa direction in the forward mesa direction. And
The Mach-Zehnder modulator is
And two wavelength division means for coupling the branch to the light, and two optical waveguides connecting one output of the wavelength division means and the other an input unit of the wavelength division means each of the two optical waveguides Two phase control means provided in each,
A p-type surface layer identical to the p-type surface layer of the semiconductor optical amplifier is formed between the two optical coupling / branching means and the two phase control means , and the p-type surface layer A non-p-type layer is formed between the n layer and the p layer in the npin layer structure . An optical semiconductor device according to claim 1,
The interface between the semiconductor optical amplifier and the Mach-Zehnder modulator or the interface between the Mach-Zehnder modulator and the p-type surface layer is formed to extend in a direction inclined with respect to the reverse mesa direction. An optical semiconductor device. An optical semiconductor device according to claim 2,
An optical semiconductor device characterized in that an inclination angle of the interface is larger than 0 degree and smaller than 45 degrees. An optical semiconductor device according to any one of claims 1 to 3,
The Mach-Zehnder modulator is
A first 2 × 2 MMI coupler having two inputs and two outputs;
A second 2 × 2 MMI coupler having two inputs and two outputs;
Two optical waveguides respectively connecting the output part of the first 2 × 2 MMI coupler and the input part of the second 2 × 2 MMI coupler;
Comprising a two phase control means provided in each of the two optical waveguides,
An optical semiconductor device, wherein the semiconductor optical amplifier is formed at an input portion of the first 2 × 2 MMI coupler. An optical semiconductor device according to any one of claims 1 to 3,
The Mach-Zehnder modulator is
A first 2 × 2 MMI coupler having two inputs and two outputs;
A second 2 × 2 MMI coupler having two inputs and two outputs;
Two optical waveguides respectively connecting the output portion of the first 2 × 2 MMI coupler and the input portion of the second 2 × 2 MMI coupler;
Comprising a two phase control means provided in each of the two optical waveguides,
An optical semiconductor device, wherein the semiconductor optical amplifier is formed at an input portion of the second 2 × 2 MMI coupler.   An optical semiconductor device manufacturing method according to any one of claims 1 to 5, comprising:
  In forming a non-p-type layer between the p-type surface layer and the p-layer in the npin layer structure,
  On the substrate, a p-type surface layer of the semiconductor optical amplifier is laminated partway, and an etch stop layer and an InP cover layer are sequentially formed on the pin layer structure in the npin layer structure from the bottom in a portion other than the semiconductor optical amplifier. And a p-type surface layer identical to the p-type surface layer of the semiconductor optical amplifier is formed on the entire surface.
  Thereafter, in a portion excluding the portion of the semiconductor optical amplifier and the portion where the p-type surface layer between the two optical coupling / branching means and the two phase control means is to be formed, the uppermost layer To the etch stop layer,
  A new n layer is laminated on the removed portion, so that the etch stop layer and the InP cover are formed as a non-p-type layer between the p-type surface layer and the p-layer in the npin layer structure. Layer and formed state
An optical semiconductor device manufacturing method.

Images