Optical semiconductor device

Abstract

The present disclosure obtains an optical semiconductor device capable of preventing a decrease in a frequency band. A first pad (8) connected to a first electrode (6) of an optical modulator (2) and a second pad (9) connected to a second electrode (7) of the optical modulator (2) are disposed on a first terrace (11) via an insulating film (22). In the first terrace (11), a first conductive type layer (13, 14) is removed between the first pad (8) and the second pad (9).

Description

Technical Field

[0001] This disclosure relates to optical semiconductor devices. Background Technology

[0002] An optical semiconductor device that monolithically integrates a laser unit and an optical modulator is proposed (see, for example, Patent Document 1). The optical modulator operates differentially by applying a differential voltage between the anode pad and the cathode pad. By arranging the anode pad and the cathode pad of the optical modulator on the same side of the platform relative to the waveguide, the lengths of the leads connected to the two pads can be made to be the same.

[0003] Patent Document 1: Japanese Patent No. 5891920

[0004] Because leakage current flows between the two pads of the optical modulator during differential operation, the voltage applied to the absorber layer of the optical modulator decreases. Since the leakage current flows through the capacitance beneath the electrodes of the optical modulator, the leakage current increases and the extinction ratio decreases at higher frequencies. As a result, there is a reduction in the frequency band in which the optical modulator can operate normally. Summary of the Invention

[0005] This disclosure is made to solve the aforementioned problems, and its purpose is to obtain an optical semiconductor device that can prevent bandwidth degradation.

[0006] The optical semiconductor device disclosed herein is characterized by comprising: a substrate; an optical modulator having a semiconductor layer comprising a first conductivity layer, an absorption layer, and a second conductivity layer sequentially formed on the substrate, a first electrode connected to the first conductivity layer, and a second electrode connected to the second conductivity layer; a first pad connected to the first electrode; and a second pad connected to the second electrode, wherein the semiconductor layer has a waveguide and a first platform and a second platform disposed opposite to each other relative to the waveguide, the first pad and the second pad being disposed on the first platform via an insulating film, wherein the first conductivity layer is removed between the first pad and the second pad on the first platform.

[0007] In this disclosure, in the first platform, the first conductivity layer is removed between the first pad and the second pad. This breaks the leakage current path between the first pad and the second pad, reducing leakage current and thus improving the response, particularly in the high-frequency region. Consequently, bandwidth degradation can be prevented. Attached Figure Description

[0008] Figure 1 This is a top view of the optical semiconductor device according to Embodiment 1.

[0009] Figure 2 It is along Figure 1 A cross-sectional view of the laser section cut by A-A'.

[0010] Figure 3 It is along Figure 1 A cross-sectional view of an optical modulator cut by B-B'.

[0011] Figure 4 It is along Figure 1 A cross-sectional view of a C-C' cut optical modulator.

[0012] Figure 5 It is along Figure 1 A cross-sectional view of a waveguide cut by a D-D' section.

[0013] Figure 6 It is along Figure 1 A cross-sectional view of the first platform cut by E-E'.

[0014] Figure 7 This is a top view of the optical semiconductor device involved in the comparative example.

[0015] Figure 8 It is along Figure 7 A cross-sectional view of a waveguide cut by A-A'.

[0016] Figure 9 It is along Figure 7 A cross-sectional view of the first platform cut by B-B'.

[0017] Figure 10 This is a graph showing the frequency response characteristics of Embodiment 1 and the comparative example.

[0018] Figure 11 This is a top view of the optical semiconductor device involved in Embodiment 2.

[0019] Figure 12 It is along Figure 11 A cross-sectional view of an optical modulator cut by A-A'.

[0020] Figure 13 It is along Figure 11 A cross-sectional view of an optical modulator cut by B-B'.

[0021] Explanation of reference numerals in the attached figures

[0022] 1...Laser unit; 2...Optical modulator; 3...Semi-insulating InP substrate (substrate); 4...Cathode electrode (electrode); 6...Cathode electrode (first electrode); 7...Anode electrode (second electrode); 8...Cathode pad (first pad); 9...Anode pad (second pad); 10...Waveguide; 11...First platform; 12...Second platform; 13...n-InGaAs contact layer (first conductivity layer); 14...n-InP cladding layer (first conductivity layer); 16...p-InP cladding layer (second conductivity layer); 20...First trench; 21...Second trench; 22...Insulating film; 24...Semiconductor layer; 25...Absorbing layer. Detailed Implementation

[0023] The optical semiconductor device according to the embodiments will be described with reference to the accompanying drawings. The same or corresponding components are labeled with the same reference numerals, and repeated descriptions are sometimes omitted.

[0024] Implementation Method 1

[0025] Figure 1 This is a top view showing the optical semiconductor device according to Embodiment 1. This optical semiconductor device is a modulator-integrated laser diode in which a laser unit 1 and an optical modulator 2 are monolithically integrated on a semi-insulating InP substrate 3. The laser unit 1 is a distributed-feedback laser diode (DFB-LD). The optical modulator 2 is an electroabsorption modulator.

[0026] The laser unit 1 has a cathode electrode 4 and an anode electrode 5. The optical modulator 2 has a cathode electrode 6 and an anode electrode 7. A cathode pad 8 is connected to the cathode electrode 6. An anode pad 9 is connected to the anode electrode 7. The optical modulator 2 operates differentially by applying a differential voltage between the cathode pad 8 and the anode pad 9.

[0027] A first terrace 11 and a second terrace 12 are arranged on opposite sides of the waveguide 10. A cathode pad 8 and an anode pad 9 are disposed on the first terrace 11. This allows the lead connected to the anode pad 9 and the lead connected to the cathode pad 8 to be of the same length.

[0028] In the figure, region V is the area between cathode pad 8 and anode pad 9. Region W is the area between laser unit 1 and optical modulator 2. Region X is the area where cathode pad 8 is formed. Region Y is the area where anode pad 9 is formed. Region Z is the area where the second platform 12 of optical modulator 2 exists. The area enclosed by the dashed line in the figure is the area where the n-type cladding and n-type contact layer, described later, exist.

[0029] Figure 2 It is along Figure 1 A cross-sectional view of the laser section taken along line A-A'. An n-InGaAs contact layer 13, an n-InP cladding layer 14, an active layer 15, a p-InP cladding layer 16, and a p-InGaAs contact layer 17 are sequentially stacked on a semi-insulating InP substrate 3. The active layer 15 is an InGaAsP multi-quantum well (MQW) structure.

[0030] The active layer 15 is patterned into stripes when viewed from above, and is embedded on both sides by Fe-InP blocking layers 18. A diffraction grating 19 is formed in the p-InP cladding layer 16.

[0031] On both sides of the active layer 15, and on the p-InGaAs contact layer 17, p-InP cladding layer 16, Fe-InP barrier layer 18, and n-InP cladding layer 14, a first trench 20 and a second trench 21 are formed. The second trench 21 extends below the n-InGaAs contact layer 13. The upper surface of the p-InGaAs contact layer 17 and the inner surfaces of the first trench 20 and the second trench 21 are covered by an insulating film 22. An opening is formed in the upper part of the mesa structure between the first trench 20 and the second trench 21, through which the anode electrode 5 is connected to the p-InGaAs contact layer 17. An opening is formed in the bottom surface of the first trench 20, through which the cathode electrode 4 is connected to the p-InGaAs contact layer 17. An n-electrode 23 is formed on the lower surface of the semi-insulating InP substrate 3.

[0032] In order to connect the cathode electrode 4 of the laser section 1 to the active layer 15 of the laser section 1 with low resistance, a low-resistance n-InP cladding layer 14 and an n-InGaAs contact layer 13 are required at the connection point between the active layer 15 of the waveguide 10 of the laser section 1 and the cathode electrode 4 of the laser section 1. In particular, the resistance of the n-InGaAs contact layer 13 is very low.

[0033] Figure 3 It is along Figure 1A cross-sectional view of a B-B' cut optical modulator. On a semi-insulating InP substrate 3, as a semiconductor layer 24, an n-InGaAs contact layer 13, an n-InP cladding layer 14, a p-InP cladding layer 16, and a p-InGaAs contact layer 17 are sequentially stacked. A first trench 20 and a second trench 21 are formed between the p-InGaAs contact layer 17, the p-InP cladding layer 16, and the n-InP cladding layer 14 in a mutually separated manner. The first trench 20 and the second trench 21 limit the lateral width of the absorption layer 25, enabling it to function as a waveguide 10. Within the mesa structure between the first trench 20 and the second trench 21, an absorption layer 25 is formed between the n-InP cladding layer 14 and the p-InP cladding layer 16. The absorption layer 25 is a multilayer quantum well structure of InGaAsP.

[0034] Semiconductor layer 24 has waveguide 10 and a first platform 11 and a second platform 12 disposed opposite to each other relative to waveguide 10. In the first platform 11 and the second platform 12, an Fe-InP blocking layer 18 is formed between the n-InP cladding layer 14 and the p-InP cladding layer 16, replacing the absorption layer 25. An opening is formed in the insulating film 22 at the bottom surface of the first trench 20, through which the cathode electrode 6 is connected to the n-InGaAs contact layer 13. The cathode pad 8 is disposed on the first platform 11 via the insulating film 22 and connected to the cathode electrode 6. To electrically decouple the electrostatic capacitor C2 from the lower part of the absorption layer 25 of the optical modulator 2, the depth of the second trench 21 is deeper than that of the n-InGaAs contact layer 13. On the other hand, in order to connect the cathode electrode 6 of the optical modulator 2 and the n-InGaAs contact layer 13, the depth of the first trench 20 at the connection portion between the cathode electrode 6 and the n-InGaAs contact layer 13 extends to the top of the n-InGaAs contact layer 13. These trenches of different depths need to be formed in a separate process.

[0035] In order to connect the cathode electrode 6 of the optical modulator 2 to the absorption layer 25 of the optical modulator 2 with low resistance, a low-resistance n-InP cladding layer 14 and an n-InGaAs contact layer 13 are required at the connection point between the absorption layer 25 of the waveguide 10 of the optical modulator 2 and the cathode electrode 6 of the optical modulator 2. In particular, the resistance of the n-InGaAs contact layer 13 is very low.

[0036] Figure 4 It is along Figure 1 A cross-sectional view of the C-C' cut optical modulator. The anode pad 9, like the cathode pad 8, is disposed on the first platform 11 via the insulating film 22. The insulating film 22 has an opening at the upper part of the waveguide 10. The anode electrode 7 is connected to the p-InGaAs contact layer 17 at the upper part of the waveguide 10 via the opening in the insulating film 22, and is connected to the anode pad 9 via the side and bottom surfaces of the first groove 20.

[0037] An electrostatic capacitance C1 is generated between the n-InGaAs contact layer 13 of the semi-insulating InP substrate 3 and the n-electrode 23, sandwiched below the anode pad 9 of the first platform 11. An electrostatic capacitance C2 is generated between the n-InGaAs contact layer 13 of the semi-insulating InP substrate 3 and the n-electrode 23, sandwiched in the second platform 12. In order to electrically decouple the electrostatic capacitances C1 and C2 from the lower part of the absorption layer 25 of the optical modulator 2, the depth of the first trench 20 at the portion through which the anode electrode 7 passes and the depth of the second trench 21 are made deeper than the n-InGaAs contact layer 13.

[0038] Figure 5 It is along Figure 1 A cross-sectional view of the waveguide cut by a D-D' axis. The active layer 15 and the absorption layer 25 are connected by a transparent waveguide layer 26. A transparent waveguide layer 26 is also formed between the absorption layer 25 and the output end face. The transparent waveguide layer 26 is composed of a single layer of InGaAsP. In the region W of the waveguide 10 between the optical modulator 2 and the laser section 1, the low-resistance n-InP cladding layer 14 and the n-InGaAs contact layer 13 are removed and buried by an Fe-InP blocking layer 18. As a result, the n-InGaAs contact layer 13 and the n-InP cladding layer 14 of the laser section 1 are electrically insulated from the n-InGaAs contact layer 13 and the n-InP cladding layer 14 of the optical modulator 2.

[0039] Figure 6 It is along Figure 1 A cross-sectional view of the first platform cut by E-E'. Cathode pad 8 and anode pad 9 are respectively connected to differential power supply 27. Differential power supply 27 performs push-off operation by alternately applying voltage at one end positive and the other negative. Not limited to this, the anode-side voltage can also be varied to make the cathode-side potential 0. A load resistor R is connected in parallel with respect to the pn junction 28 of the optical modulator 2. The load resistor R is located externally to the device for impedance matching.

[0040] In the first platform 11, in the region V between the cathode pad 8 and the anode pad 9, the low-resistivity n-InP cladding layer 14 and the n-InGaAs contact layer 13 are removed and buried by the Fe-InP blocking layer 18. Additionally, in the region W between the cathode pad 8 of the optical modulator 2 and the cathode electrode 4 of the laser section 1, the low-resistivity n-InP cladding layer 14 and the n-InGaAs contact layer 13 are removed and buried by the Fe-InP blocking layer 18.

[0041] Next, the effects of this embodiment will be compared with those of a comparative example for explanation. Figure 7 This is a top view of the optical semiconductor device involved in the comparative example. Figure 8 It is along Figure 7A cross-sectional view of the waveguide cut along line A-A'. In the comparative example, the n-InGaAs contact layer 13 and the n-InP cladding layer 14 are formed on the entire surface of the chip and are not removed. Between the optical modulator 2 and the laser section 1, the low-resistance n-InP cladding layer 14 and the n-InGaAs contact layer 13 exist in the waveguide 10, thus causing leakage current to flow between the optical modulator 2 and the laser section 1. Therefore, the voltage amplitude applied to the optical modulator 2 from the cathode side of the differential power supply 27 decreases, and the load impedance observed from the anode side and the cathode side of the differential power supply 27 is different. Furthermore, a portion of the leakage current flows to the active layer 15 of the laser section 1, thereby changing the amount of current flowing to the active layer 15, and thus changing the light output of the laser section 1. When the light output of the laser section 1 changes, the light output modulated by the optical modulator 2 also changes accordingly. In addition, a very large wavelength chirp (wavelength variation) is generated. Therefore, suppressing leakage current is important to improve communication quality.

[0042] Figure 9 It is along Figure 7 A cross-sectional view of the first platform cut by B-B'. The cathode pad 8 and the anode pad 9 are connected at high frequency via an insulating film 22, a p-InP cladding layer 16, an Fe-InP barrier layer 18, an n-InP cladding layer 14, and an n-InGaAs contact layer 13. Therefore, a resistance R1 exists in the semiconductor layer 24 between the cathode pad 8 and the anode pad 9 as a leakage current path.

[0043] Furthermore, the cathode pad 8 of the optical modulator 2 is also connected to the cathode electrode 4 of the laser section 1 at a high frequency. Therefore, a resistor R2 exists in the semiconductor layer 24 between the anode pad 9 and the cathode electrode 4, serving as a leakage current path. Additionally, the positions of the cathode pad 8 and the anode pad 9 of the optical modulator 2 can be reversed. In this case, a potential difference is generated between the anode electrode 7 of the optical modulator 2 and the cathode electrode 4 of the laser section 1, resulting in leakage current.

[0044] The current caused by the anode voltage of the differential power supply 27 flows through the paths of the load resistors R and R2, and through the paths of the capacitor C, resistor R1, and resistor R2 of the insulating film 22. The current caused by the cathode voltage flows through resistor R2 to the cathode electrode 4 of the laser section 1.

[0045] The higher the frequency, the lower the impedance of capacitor C, and the greater the leakage current flowing in resistor R1. This increased leakage current leads to a decrease in the current flowing through the load resistor R, thus reducing the voltage applied to the optical modulator 2. Therefore, the higher the frequency, the greater the leakage current and the lower the extinction ratio, resulting in a reduced bandwidth.

[0046] In contrast, in this embodiment, by removing the n-InP cladding layer 14 and the n-InGaAs contact layer 13 in the region V between the cathode pad 8 and the anode pad 9, the leakage current path is broken, the leakage current is reduced, and therefore the response, especially in the high-frequency region, is improved. As a result, bandwidth reduction can be prevented. Furthermore, although it is only necessary to remove one of the n-InP cladding layer 14 and the n-InGaAs contact layer 13, it is preferable to remove both. In addition, even in a semiconductor device without the optical modulator unit of the laser section 1, the same effect can be obtained by removing the n-InP cladding layer 14 and the n-InGaAs contact layer 13 between the pads.

[0047] Furthermore, since resistors R1 and R2 are present in the comparative example, the current path caused by the cathode voltage is different from that caused by the anode voltage, and their impedances are also different. In contrast, in this embodiment, by removing the n-InP cladding layer 14 and the n-InGaAs contact layer 13 in regions V and W, the current path flowing in resistors R1 and R2 is interrupted, so the current caused by the two voltages flows only to the load resistor R. As a result, the impedances on the anode side and the cathode side become equal. Consequently, the frequency response characteristics of the anode modulation and the cathode modulation become approximately the same, thus enabling ideal differential operation. In addition, by making the impedances on the anode side and the cathode side equal, the amplitude and phase of the noise on the anode side and the cathode side can be made consistent, maximizing the noise reduction effect caused by differential operation.

[0048] Figure 10 This is a graph showing the frequency response characteristics of Embodiment 1 and the comparative example. It can be seen that in this embodiment, both the frequency response to anode modulation and the frequency response to cathode modulation are improved compared to the comparative example.

[0049] Implementation Method 2

[0050] Figure 11 This is a top view of the optical semiconductor device involved in Embodiment 2. Figure 12 It is along Figure 11 A cross-sectional view of an optical modulator cut by A-A'. Figure 13 It is along Figure 11 A cross-sectional view of the optical modulator cut by B-B'. In this embodiment, the n-InP cladding layer 14 and the n-InGaAs contact layer 13 are removed from the entire area of ​​the first platform 11 and the entire area of ​​the second platform 12 of the optical modulator 2. The n-InP cladding layer 14 and the n-InGaAs contact layer 13 are present in other areas. Alternatively, in this embodiment, the n-InP cladding layer 14 and the n-InGaAs contact layer 13 can be removed from region W in the same manner as in Embodiment 1. Other structures are the same as in Embodiment 1.

[0051] In the first platform 11, an electrostatic capacitance C1 is generated between the p-InP cladding 16, which sandwiches the semi-insulating InP substrate 3 and the Fe-InP barrier layer 18, and the n-electrode 23. However, in the first platform 11, the n-InP cladding 14 and the n-InGaAs contact layer 13 are removed. Therefore, the resistance between the n-InP cladding 14 and the n-InGaAs contact layer 13 of the waveguide 10 of the optical modulator 2 and the electrostatic capacitance C1 increases significantly. Therefore, it is not necessary to make... Figure 13 The bottom surface of the first trench 20 shown is deeper than the n-InGaAs contact layer 13. Similarly, in the second platform 12, an electrostatic capacitance C2 is generated between the p-InP cladding 16, which sandwiches the semi-insulating InP substrate 3 and the Fe-InP barrier layer 18, and the n-electrode 23. However, in the second platform 12, the n-InP cladding 14 and the n-InGaAs contact layer 13 are removed. Therefore, the resistance between the n-InP cladding 14 and the n-InGaAs contact layer 13 of the waveguide 10 of the optical modulator 2 and the electrostatic capacitance C2 increases significantly. Therefore, it is not necessary for the bottom surface of the second trench 21 to be deeper than the n-InGaAs contact layer 13. Thus, by removing the n-InP cladding 14 and the n-InGaAs contact layer 13 in the first platform 11 and the second platform 12 of the optical modulator 2, even if the depths of the first trench 20 and the second trench 21 are the same, there is no problem of increased electrostatic capacitance.

[0052] Additionally, in implementation method 1, such as Figure 3 As shown, the depth of the first groove 20 at the connection portion between the cathode electrode 6 and the n-InGaAs contact layer 13 is different from the depth of the second groove 21. Therefore, the distribution of guided light in the optical modulator 2 is asymmetrical. Consequently, scattering loss occurs when the laser with a symmetrical light distribution is guided into the optical modulator 2 or during waveguided ...

Claims

1. An optical semiconductor device, characterized in that, The optical semiconductor device includes: substrate; An optical modulator has a semiconductor layer comprising a first conductivity layer, an absorption layer, and a second conductivity layer sequentially formed on the substrate, a first electrode connected to the first conductivity layer, and a second electrode connected to the second conductivity layer; The first pad is connected to the first electrode; and The second pad is connected to the second electrode. The semiconductor layer has a waveguide and a first platform and a second platform disposed opposite to each other relative to the waveguide. The first pad and the second pad are disposed on the first platform via an insulating film. In the first platform, the first conductive layer is removed between the first pad and the second pad.

2. The optical semiconductor device according to claim 1, characterized in that, A laser unit, monolithically integrated with the optical modulator, is also provided on the substrate. The first conductivity layer is removed between the optical modulator and the laser unit.

3. The optical semiconductor device according to claim 1 or 2, characterized in that, In the semiconductor layer of the optical modulator, a first trench is formed between the waveguide and the first plateau, and a second trench is formed between the waveguide and the second plateau. The first electrode is connected to the first conductive layer at the bottom of the first trench. The second electrode is connected to the second conductivity layer at the upper part of the waveguide, and is connected to the second pad through the first trench. In the first and second platforms of the optical modulator, the first conductivity layer is removed. The depth of the first trench at the connection portion between the first electrode and the first conductive layer, the depth of the first trench at the portion through which the second electrode passes, and the depth of the second trench are all the same.

Images