Hybrid optical device and hybrid optical device production method

Abstract

A hybrid optical device according to the present disclosure comprises, in order from the outermost layer thereof, a p-InP layer (7), a light-absorbing layer (9), an n-InP layer (6), a dielectric thin film (10), an optical waveguide (8), a BOX layer (11), and a core substrate (12). The light-absorbing layer (9) has a second taper (2), and the light-absorbing layer (9) is exposed in the second taper (2).

Description

Hybrid optical element and method for producing the same

[0001] The disclosed technology relates to hybrid optical devices.

[0002] Hybrid optical elements are known that integrate highly efficient optical devices on a silicon waveguide. The term "hybrid" in hybrid optical elements is synonymous with that used in "hybrid IC" and "hybrid integrated circuit," and is used as the opposite of "monolithic." Hybrid optical elements are used, for example, as devices for optical Ethernet communications. More specifically, hybrid optical elements are used, for example, as communication devices for transmission methods such as NRZ (Non-Return-to-Zero) or PAM (Pulse Amplitude Modulation).

[0003] High-efficiency optical devices are realized, for example, using compound semiconductors. A silicon waveguide and a compound semiconductor waveguide are connected by adiabatic transition using a taper. The compound semiconductor has an n-type semiconductor on the bottom and a p-type semiconductor on the top, forming a p-n junction. The taper can be either where the n-type semiconductor is exposed on the surface (hereinafter referred to as the "first taper") or where the p-type semiconductor is exposed on the surface (hereinafter referred to as the "second taper").

[0004] For example, Patent Document 1 discloses a hybrid optical device having a structure that improves the electrical resistance in a light source device.

[0005] Japanese Patent Application Laid-Open No. 2019-140319

[0006] Hybrid optical elements are applied, for example, to electroabsorption modulators (also called EA modulators). When applied to EA modulators, the tip of the second taper of the hybrid optical element is likely to couple to the TM mode, which reflects the vertically elongated waveguide shape, and a transition of (1) TE [Si waveguide] ⇒ (2) TM [taper tip] ⇒ (3) TE [taper base] ⇒ TE [EA modulator] is often obtained. As a result, in order to achieve sufficient coupling, the taper becomes long, resulting in a large device. Note that the above TE and TM represent the propagation modes of electromagnetic waves, respectively.

[0007] When hybrid optical devices are applied to high-speed modulation devices such as EA modulators, the area of ​​the modulator must be minimized to reduce parasitic capacitance (also called floating capacitance or stray capacitance). An increase in parasitic capacitance degrades the high-frequency characteristics of the device. On the other hand, reducing the area of ​​the modulator shortens the taper length, making it impossible to achieve the above-mentioned coupling. Furthermore, in EA modulators, the modulation bias is also applied to the taper section, causing unintended optical absorption in the taper section and resulting in insertion loss. Therefore, the taper length must be designed with consideration given to the trade-off between sufficient coupling and reduced parasitic capacitance.

[0008] The disclosed technology aims to solve the above trade-off problem, and shortens the taper while ensuring sufficient coupling in a hybrid optical element, thereby reducing parasitic capacitance.

[0009] The hybrid optical element according to the disclosed technology comprises, in order from the surface, p-InP 7, a light absorbing layer 9, n-InP 6, a dielectric thin film 10, an optical waveguide 8, a BOX layer 11, and a core substrate 12, and the light absorbing layer 9 has a second taper 2 at which the light absorbing layer 9 is exposed.

[0010] The hybrid optical element according to the disclosed technique has the above-described configuration, and therefore the taper is shortened while ensuring sufficient coupling, thereby realizing a reduction in parasitic capacitance.

[0011] FIG. 1 is a structural diagram of a hybrid optical element according to a first embodiment. FIG. 2 is a first explanatory diagram illustrating a bonding process, which is part of a production process of a hybrid optical element according to the first embodiment. FIG. 3 is a second explanatory diagram illustrating a bonding process, which is part of a production process of a hybrid optical element according to the first embodiment. FIG. 4 is a graph illustrating a propagation mode transition occurring in the hybrid optical element according to the first embodiment. FIG. 5 is a structural diagram when a conventional hybrid optical element is applied to an EA modulator. FIG. 6 is a graph illustrating a propagation mode transition occurring in the conventional hybrid optical element. FIG. 7 is a graph illustrating the coupling efficiency of a hybrid optical element according to the disclosed technology compared with the conventional technology. FIG. 8 is a structural diagram of a hybrid optical element according to a second embodiment. FIG. 9 is a graph illustrating a propagation mode transition occurring in the hybrid optical element according to the second embodiment. FIG. 10 is a structural diagram of a hybrid optical element according to a third embodiment.

[0012] The hybrid optical element according to the disclosed technique is applied as an EA modulator.

[0013] First Embodiment Fig. 1 is a structural diagram (perspective view) of a hybrid optical device according to a first embodiment.

[0014] 1, the hybrid optical device according to the first embodiment includes an n-electrode 4, a p-electrode 5, an n-InP layer 6s, a p-InP layer 7s, an optical waveguide 8, a light absorption layer 9, a dielectric thin film 10, a BOX layer 11, and a core substrate 12. The n-InP 6 has a first taper 1 formed therein. The light absorption layer 9 has a second taper 2 formed therein.

[0015] The optical waveguide 8 utilizes a difference in refractive index. The optical waveguide 8 utilizes the property that light is localized and propagates in a region with a high refractive index, such as silicon. The optical waveguide 8 may be a Si waveguide or a SiN waveguide. In the hybrid optical device according to the first embodiment, the optical waveguide 8 is a Si waveguide. An example in which the optical waveguide 8 is a SiN waveguide is shown in the fifth embodiment.

[0016] <<Production Process of Hybrid Optical Device>> The raw materials of the hybrid optical device include a silicon photonics substrate 14 and an InP epitaxial substrate 13. The silicon photonics substrate 14 has a dielectric thin film 10 as the outermost layer and has a waveguide structure. The dielectric in the dielectric thin film 10 is SiO ２ , or Al ２ O ３ In this specification, the dielectric thin film 10 is generally made of SiO ２ The InP epitaxial substrate 13 is produced by epitaxially growing a layer structure including the light absorption layer 9 on an n-InP substrate.

[0017] The manufacturing process of the hybrid optical device includes a bonding process between the silicon photonics substrate 14 and the InP epitaxial substrate 13. The bonding between the silicon photonics substrate 14 and the InP epitaxial substrate 13 is realized by wafer bonding using surface activated bonding technology. In principle, the bonding surface between the silicon photonics substrate 14 and the InP epitaxial substrate 13 is formed by a dielectric thin film 10 (SiO ２ Alternatively, the n-InP 6 side may be provided with SiO ２ By depositing the dielectric thin film 10 (SiO ２ thin film) and SiO deposited on n-InP6 ２ and SiO ２ The substrate obtained by bonding the silicon photonics substrate 14 and the InP epitaxial substrate 13 is called a hybrid substrate.

[0018] 2 is a first explanatory diagram showing a bonding process, which is part of the production process of the hybrid optical device according to the first embodiment. As shown in FIG. 2, the bonding between the silicon photonics substrate 14 and the InP epitaxial substrate 13 may be a bonding between circular substrates.

[0019] 3 is a second explanatory diagram illustrating a bonding process, which is part of the production process of the hybrid optical element according to the first embodiment. As shown in FIG. 3 , the bonding between the silicon photonics substrate 14 and the InP epitaxial substrate 13 may be performed by bonding InP epitaxial substrate pieces 13p, which are fragments of the InP epitaxial substrate 13, onto the silicon photonics substrate 14.

[0020] The dielectric thin film 10 in the silicon photonics substrate 14 preferably has a thickness of 300 nm or less. By making the thickness 300 nm or less, optical coupling between the optical waveguide 8 and the light absorption layer 9 can be easily achieved.

[0021] The InP epitaxial substrate 13 is produced by epitaxially growing an n-InP buffer layer, a light absorption layer 9, and a p-InP layer 7s on an n-InP substrate. The epitaxial growth used to produce the InP epitaxial substrate 13 produces a layer structure on the order of several μm. The thickness of the n-InP substrate is approximately 300 μm to 1 mm, depending on the diameter of the substrate.

[0022] In either the n-InP layer 6s or the p-InP layer 7s formed by epitaxial growth, alloy crystal materials such as GaInAs, AlInAs, InGaAlAs, or GaInAsP may be inserted. The insertion of alloy crystal materials is performed for purposes such as trapping carriers generated by photocurrent, reducing contact resistance with metal electrodes, and forming an etch stop layer required in the processing process. In the semiconductor field, GaInAs is often referred to as indium gallium arsenide (InGaAs). GaInAs is a ternary alloy (compound) of indium arsenide (InAs) and gallium arsenide (GaAs). Indium arsenide (InAs) and gallium arsenide (GaAs) both have a 3-5 （ローマ数字） It is a group compound semiconductor.

[0023] The light absorption layer 9 has a quantum well structure. A quantum well refers to a state in which the direction of movement of particles such as electrons is restricted. The light absorption layer 9 is designed to absorb light in the 1.3 μm wavelength band or the 1.55 μm wavelength band due to the quantum Stark effect when an electric field is applied. The quantum well is made of a mixed crystal material such as InGaAlAs or GaInAsP.

[0024] The manufacturing process for hybrid optical devices includes a semiconductor process for forming a device structure on a hybrid substrate. First, the n-InP layer 6s, which is the outermost layer of the hybrid substrate, is mostly removed by etching, thinning it to a thickness of several hundred μm to several μm. Then, lithography and dry etching techniques are used to form the tapered structure shown in Figure 1.

[0025] Regarding the thickness of the layers, for example, the thickness of the n-InP 6 is 200 nm, the thickness of the light absorption layer 9 is 500 nm, and the thickness of the p-InP 7 is 2 μm. The thickness of the n-InP 6 needs to be thin so that optical coupling can be obtained between the optical waveguide 8 and the light absorption layer 9. If the thickness of the n-InP 6 exceeds approximately 300 nm, the optical coupling weakens and the taper length required for the transition of the propagation mode becomes significantly long. It is important that the p-InP 7 has a sufficient thickness so that the propagation mode does not reach the p-electrode 5 and cause optical loss.

[0026] The first taper 1 is processed so that the n-InP 6 is exposed on the surface. The second taper 2 is processed so that the light absorbing layer 9 is exposed on the surface. The light absorbing layer 9 is made of a ternary or quaternary mixed crystal with a higher refractive index than the n-InP 6. Since it is undesirable for the n-InP layer 6s and the light absorbing layer 9 to be exposed to the atmosphere during the tapering process, they may be covered with a very thin InP protective layer of 100 nm or less.

[0027] The hybrid optical element formed by the above process has a small effective refractive index difference between the optical waveguide 8 and the tip of the second taper 2. Therefore, the first taper 1 can achieve adiabatic transition over a very short distance. In the hybrid optical element according to the disclosed technology, the taper length of the first taper 1 can be set to 10 μm.

[0028] The longer the length of the modulation section 15 in the p-InP layer 7s, the better the extinction ratio; however, if it is too long, the parasitic capacitance increases and the operating bandwidth decreases. Regarding the length of the modulation section 15, there is also a trade-off between the extinction ratio and the parasitic capacitance. If the hybrid optical element is used as a device for 53 GBaud PAM4, the currently mainstream optical Ethernet communication method, the length of the modulation section 15 can be approximately 200 μm. If the hybrid optical element is used as a device for 106 GBaud PAM4, the next-generation optical Ethernet communication method, the operating bandwidth is doubled, and the length of the modulation section 15 is approximately 100 μm. PAM4 stands for 4-level Pulse-Amplitude Modulation, which means four-level pulse amplitude modulation. More specifically, PAM4 is a method of transmitting a four-valued pulse signal instead of the conventional binary bit string consisting of "0" and "1".

[0029] <<Operation or Behavior of Light in Hybrid Optical Element>> Input light to the hybrid optical element propagates through the optical waveguide 8 in the TE mode. TE stands for Transverse Electric. The TE mode is a term that describes the state when there is no electric field in the propagation direction. In this specification, the mode in which input light propagates through the optical waveguide 8 is referred to as the "TE fundamental mode." In addition, in this specification, the TE mode for the light absorption layer 9 is referred to as the "TE fundamental mode," and the two are distinguished from each other. The propagating light reaches the modulation section 15 via the first taper 1 and the second taper 2.

[0030] A reverse bias is applied between the p-electrode 5 and the n-electrode 4. When a reverse bias is applied, light absorption occurs in the light absorption layer 9 due to the quantum confined Stark effect (also simply referred to as the "quantum Stark effect"). The Stark effect is a phenomenon in which the spectrum changes when a uniform external electric field is applied to atoms or molecules. By changing the magnitude of the reverse bias, the amount of light absorption in the modulation section 15 can be changed. By utilizing this property, the hybrid optical element according to the disclosed technology is capable of light intensity modulation such as NRZ or PAM.

[0031] It is believed that light absorption occurs only in the region where the p-InP 7 exists. The light that has passed through the modulation section 15 passes through the tapered structure again and returns to the optical waveguide 8.

[0032] FIG. 4 is a graph showing propagation mode transitions occurring in the hybrid optical element according to the first embodiment. In the graph shown in FIG. 4, the horizontal axis represents position, and the vertical axis represents effective refractive index. The effective refractive index is also sometimes referred to as the effective refractive index. The graph shown in FIG. 4 is the result of a mode analysis performed on the cross section of the waveguide shown in FIG. 1. The graph in FIG. 4 shows the top five modes with the highest effective refractive index values. The mode with the highest effective refractive index, indicated by a thick solid curve, is the dominant mode. As shown in FIG. 4, the mode analysis results show that there is only one mode crossing occurring in the second taper 2 of the hybrid optical element according to the first embodiment, at a location marked "mode crossing 1."

[0033] The light propagating through the optical waveguide 8 in the TE fundamental mode intersects with the TE fundamental mode of the light absorption layer 9 in the second taper 2 and undergoes mode conversion. This mode conversion is an adiabatic behavior.

[0034] The technical feature of the hybrid optical element according to the first embodiment is that the optical absorption layer 9 is exposed in the second taper 2. FIG. 5 is a structural diagram (perspective view) of a conventional hybrid optical element applied to an EA modulator. By comparing FIG. 1 (first embodiment) with FIG. 5 (conventional), the technical feature of the hybrid optical element according to the first embodiment becomes clear. In other words, the technical feature of the hybrid optical element according to the first embodiment is that p-InP is not present in the second taper 2.

[0035] 6 is a graph showing the propagation mode transition that occurs in a conventional hybrid optical element. As shown in FIG. 6 , mode analysis revealed that mode crossings that occur in the tapered portion (corresponding to second taper 2) of the conventional hybrid optical element are located at two locations: a location labeled "mode crossing 1" and a location labeled "mode crossing 2." In other words, in the conventional hybrid optical element, two mode conversions occur via the TM mode. The reasons for the two mode conversions that occur in the conventional hybrid optical element are believed to be that the electric field distribution of the TM mode is easily affected in the thickness direction, the TM fundamental mode of the optical absorption layer is increased near the tapered tip due to the influence of p-InP, and the effective refractive index is higher than that of the TE fundamental mode of the optical absorption layer.

[0036] Due to the above technical features, the hybrid optical element according to the first embodiment keeps the effective refractive index of the TM fundamental mode low near the tapered tip. Due to this effect, no mode crossing occurs between the propagating mode and the TM fundamental mode in the hybrid optical element according to the first embodiment. The technical features of the disclosed technology compared to the conventional technology are as shown in the table below.

[0037] FIG. 7 is a graph showing the coupling efficiency (simulation results) of the hybrid optical element according to the present disclosure compared to the conventional technology. In the graph shown in FIG. 7, the horizontal axis represents the taper length of the second taper 2, and the vertical axis represents the coupling efficiency. In the graph shown in FIG. 7, the simulation results of the hybrid optical element according to the first embodiment are shown by the solid curve, and the simulation results of the hybrid optical element according to the conventional structure are shown by the dashed curve. As shown in FIG. 7, the taper length required to achieve sufficient coupling in the hybrid optical element according to the present disclosure is shorter than that of the conventional technology. More specifically, in the simulation shown in FIG. 7, the taper length required to achieve a coupling efficiency of 80% is approximately one-third shorter than that of the conventional technology.

[0038] The hybrid optical element according to the first embodiment functions as an EA modulator, and transitions the mode from the optical waveguide 8 to the optical absorption layer 9 by an adiabatic transition type optical coupler having a tapered shape.

[0039] As described above, the hybrid optical element according to the first embodiment has the above technical features, and therefore has the effect of shortening the taper and reducing the parasitic capacitance while ensuring sufficient coupling.

[0040] Second Embodiment A hybrid optical element according to a second embodiment is a modified example of the hybrid optical element according to the disclosed technology. Unless otherwise specified, the same reference numerals as those used in the first embodiment are used in the second embodiment. Furthermore, in the second embodiment, descriptions that overlap with those in the first embodiment are omitted as appropriate.

[0041] 8 is a structural diagram (perspective view) of a hybrid optical device according to embodiment 2. As shown in Fig. 8, the hybrid optical device according to embodiment 2 includes a very short third taper 3 between the second taper 2 and the modulation section 15 in the p-InP layer 7s.

[0042] Fig. 9 is a graph showing the propagation mode transition that occurs in the hybrid optical element according to the second embodiment. As can be seen from Fig. 9 (Embodiment 2) in comparison with Fig. 4 (Embodiment 1), in the hybrid optical element according to the second embodiment, the discontinuous, sudden change in the effective refractive index that occurred at the boundary between the second taper 2 and the modulation section 15 has been changed to a continuous, gradual change. The characteristic shown in Fig. 9 is the result of the provision of a very short third taper 3 between the second taper 2 and the modulation section 15, which prevents the presence or absence of p-InP from occurring suddenly along the optical path.

[0043] The unique effect of the hybrid optical element according to the second embodiment is that the modes of the second taper 2 and the modulation section 15 can be converted adiabatically, thereby reducing the optical coupling loss that may occur in this section.

[0044] Incidentally, the effective refractive index difference between the end portion of the second taper 2 and the modulation section is low, at about 0.1, so a taper length of about 10 μm provides good coupling. Care must be taken to design the taper length of the third taper 3 so that it is not too long, in order to fully obtain the effects described in the first embodiment. To keep the increase in parasitic capacitance resulting from the addition of the third taper 3 to, for example, 10% or less, the taper length of the third taper 3 relative to the length of the modulation section 15 should be 5% or less (because the third taper 3 is located on both sides of the modulation section 15).

[0045] Embodiment 3. A hybrid optical element according to embodiment 3 is a modified example of the hybrid optical element according to the disclosed technology. Unless otherwise specified, in embodiment 3, the same symbols as those used in the previous embodiments are used. Furthermore, in embodiment 3, descriptions that overlap with those in the previous embodiments are omitted as appropriate.

[0046] 10 is a structural diagram of a hybrid optical element according to the third embodiment. As shown in FIG. 10, the first taper 1 is not an essential component in the hybrid optical element according to the disclosed technique. According to simulation results, the effective refractive index difference between the optical waveguide 8 and the tip of the second taper 2 is low, about 0.02. Therefore, even if a structure without the first taper 1 (see FIG. 10) is adopted, no significant increase in loss occurs.

[0047] The structure without the first taper 1 (see FIG. 10) requires fewer tapers that require high-precision processing in the production process, which allows, for example, lower alignment precision and, as a result, lower production costs. In addition, the variation in characteristics caused by misalignment that can occur in the first taper 1 is eliminated, which is expected to improve yield.

[0048] Fourth Embodiment A hybrid optical element according to a fourth embodiment is a modified example of the hybrid optical element according to the disclosed technology. Unless otherwise specified, the same reference numerals as those used in the previous embodiments are used in the fourth embodiment. Furthermore, in the fourth embodiment, descriptions that overlap with those in the previous embodiments are omitted as appropriate.

[0049] In the first embodiment, the silicon photonics substrate 14 and the InP epitaxial substrate 13 are bonded together by wafer bonding using a surface activated bonding technique, but the disclosed technology is not limited to this. The silicon photonics substrate 14 and the InP epitaxial substrate 13 may also be bonded together using a resin material such as BCB.

[0050] Bonding using a resin material such as BCB can prevent bonding defects that can occur with surface activated bonding. Resin materials are relatively soft, and even if minute foreign matter is mixed into the substrate surface, good bonding can be achieved without voids.

[0051] Fifth Embodiment A hybrid optical element according to a fifth embodiment is a modified example of the hybrid optical element according to the disclosed technology. Unless otherwise specified, the same reference numerals as those used in the preceding embodiments are used in the fifth embodiment. Furthermore, in the fifth embodiment, descriptions that overlap with those in the preceding embodiments are omitted as appropriate.

[0052] As described in the first embodiment, the optical waveguide 8 provided on the silicon photonics substrate may be a SiN waveguide. SiN waveguides have the property of being less susceptible to nonlinear optical effects than Si waveguides, and therefore can be used for applications requiring higher optical densities. The refractive index of SiN is lower than that of Si. Therefore, SiN waveguides can increase the propagation mode and achieve high coupling efficiency with optical fibers (light absorption layer 9 in the figure). SiN does not have a particularly high electro-optic effect compared to Si. Therefore, it is difficult to fabricate an optical modulator using SiN itself, but a SiN waveguide modulator can be produced using the production process according to the disclosed technology.

[0053] Sixth Embodiment A hybrid optical element according to a sixth embodiment is a modified example of the hybrid optical element according to the disclosed technology. Unless otherwise specified, the same reference numerals as those used in the preceding embodiments are used in the sixth embodiment. Furthermore, in the sixth embodiment, descriptions that overlap with those in the preceding embodiments are omitted as appropriate.

[0054] In the above-described embodiments, the hybrid optical element is described as an EA modulator, but the disclosed technology is not limited to this. The hybrid optical element according to the disclosed technology may be a device in which a directly modulated laser diode is formed on an optical waveguide 8. A device in which a directly modulated laser diode according to the disclosed technology is formed has reduced parasitic capacitance and is capable of high-speed response.

[0055] The hybrid optical element according to the disclosed technique can be applied as an optical Ethernet communication device and has industrial applicability.

[0056] 1 first taper, 2 second taper, 3 third taper, 4 n electrode, 5 p electrode, 6 n-InP (6s n-InP layer), 7 p-InP (7s p-InP layer), 8 optical waveguide, 9 light absorption layer, 10 dielectric thin film, 11 BOX layer, 12 core substrate, 13 InP epitaxial substrate, 14 silicon photonics substrate, 15 modulation section.

Claims

1. The structure is made up of, in order from the surface layer, p-InP, an optical absorption layer, n-InP, a dielectric thin film, an optical waveguide, a BOX layer, and a core substrate. the light absorbing layer has a second taper; the light absorbing layer is exposed at the second taper, The p-InP includes a p-electrode and a modulation section, The n-InP has an n-electrode, The amount of light absorption in the modulation section can be changed by changing the magnitude of the reverse bias applied between the p-electrode and the n-electrode. Hybrid optical element.

2. a third taper is provided between the second taper and the modulation section; The hybrid optical element according to claim 1 .

3. The optical waveguide is a Si waveguide or a SiN waveguide. The hybrid optical element according to claim 1 .

4. A high-speed modulation element that functions as an EA modulator, a mode transition is performed from the optical waveguide to the optical absorption layer by an adiabatic transition type optical coupler having the second taper; The hybrid optical element according to claim 1 .

5. A high-speed modulation element that functions as an EA modulator or a directly modulated laser diode. The hybrid optical element according to claim 1 .

6. A method for producing a hybrid optical element capable of changing the amount of light absorption in a modulation section by changing the magnitude of a reverse bias applied between a p-electrode and an n-electrode, comprising: A silicon photonics substrate and an InP epitaxial substrate are bonded to form a hybrid substrate. forming a device structure on the hybrid substrate; The light absorbing layer is formed to have a second taper; The light absorbing layer is formed so as to be exposed at the second taper. A method for producing a hybrid optical element.

7. forming a third taper between the second taper and the modulation section; A method for producing the hybrid optical device of claim 6.

8. The bonding between the silicon photonics substrate and the InP epitaxial substrate is This is carried out by wafer bonding using surface activated bonding techniques or by using resin materials. A method for producing the hybrid optical device of claim 6.