8-channel structure and manufacturing method based on polarization beam splitter and photodetector
By introducing polarization beam splitter and 8-channel structure into the photodetector, the sensitivity of the photodetector in terms of polarization state is solved, the response speed and bandwidth are improved, the demand for large-scale data transmission is met, and the polarization is simplified. Test of isolation.
Patent Information
- Application Number
- CN202110359483.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-04-02
- Publication Date
- 2025-02-18
- Estimated Expiration
- 2041-04-02
AI Technical Summary
Existing photodetectors are extremely sensitive in polarization state, resulting in reduced measurement accuracy, long response time, and insufficient bandwidth to meet the needs of large-scale integrated chip data transmission.
Design an 8-channel structure based on polarization beam splitter and photodetector. By forming a polarization beam splitter region and photodetector region on the SOI substrate, combining 16 photovoltaic tubes and Y branch waveguides, the polarization sensitivity is reduced and the polarization sensitivity is improved. bandwidth.
It effectively reduces the polarization sensitivity of the system, improves the response speed and bandwidth of the photodetector, meets the needs of large-scale data transmission, and simplifies the polarization isolation test process.
Smart Images

Figure CN115188776B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of semiconductor silicon photonics technology, and in particular to an 8-channel structure based on a polarization beam splitter and a photoelectric detector and a manufacturing method thereof. Background Art
[0002] In the past few decades, silicon photonics technology on silicon-on-insulator (SOI) platforms has made great progress due to its compact footprint, high integration, low power consumption, and compatibility with complementary metal oxide semiconductor (CMOS) processes. Silicon photonics technology is a low-cost, high-speed optical communication technology based on silicon photonics. Based on silicon-based substrate materials, using CMOS processes, combined with integrated circuits represented by microelectronics and photonic technologies, laser beams are used instead of electronic information to transmit data. Through silicon photonic integration, light is used instead of electricity for information transmission, which will greatly reduce the cost of integrated circuits. At the optical path receiving end of the photonic device, photodetectors are common devices. Photodetectors can convert optical signals into electrical signals, thereby realizing photoelectric conversion at the optical path receiving end.
[0003] Polarization sensitivity is a serious problem in the practical application of photodetectors. In SOI waveguides, the refractive index difference between Si and SiO2 is large, which makes the effective refractive index in the two polarization modes, TE and TM, very different, causing devices such as photodetectors to be extremely sensitive to the polarization state. Although the polarization state of light itself is also very useful information, it is a kind of interference information in most cases. Unnecessary polarization will cause errors, reduce measurement accuracy, and affect the results of photodetectors. At the same time, it is also necessary to find ways to reduce the response time of photodetectors to reduce relaxation, and ensure that the photogenerated carriers in the detector have a very short diffusion time to enable the circuit to be quickly turned on. In addition, the bandwidth of existing devices cannot meet the needs of large-scale integrated chip data transmission.
[0004] Based on the above defects in the prior art, it is necessary to propose an 8-channel structure and a manufacturing method based on a polarization beam splitter and a photodetector. Summary of the invention
[0005] In order to achieve the above technical requirements, the purpose of the present invention is to provide an 8-channel structure and manufacturing method based on a polarization beam splitter and a photodetector, which is used to reduce the polarization sensitivity of the system and reduce the response time of the photodetector, while increasing the bandwidth to meet the needs of large-scale data transmission.
[0006] The present invention provides an 8-channel structure based on a polarization beam splitter and a photodetector, wherein the 8-channel structure at least comprises:
[0007] An SOI substrate, wherein the SOI substrate comprises a silicon substrate, a dielectric layer and a top silicon layer stacked in sequence, and the SOI substrate comprises a polarization beam splitter region and a photodetector region;
[0008] The polarization beam splitter region includes 8 waveguides arranged at intervals obtained by patterning the top silicon, the waveguides include a single-mode input waveguide, a double-etched waveguide and a Y-branch waveguide connected in sequence, and the Y-branch waveguide includes a first branch waveguide and a second branch waveguide;
[0009] The photodetector area includes 16 phototubes, and the 16 phototubes include 8 first phototubes arranged corresponding to the first branch waveguides and 8 second phototubes arranged corresponding to the second branch waveguides;
[0010] A silicon dioxide cladding layer covers the polarization beam splitter region and the photodetector region to form the 8-channel structure.
[0011] Optionally, the photoelectric tube includes a silicon layer, a photosensitive layer, an N+ type doped region and a metal plug;
[0012] The silicon layer includes a P-type doped region, two P+ type doped regions and an undoped silicon layer, the two P+ type doped regions are in contact with the dielectric layer and are located on both sides of the P-type doped region, the photosensitive layer is located on the P-type doped region, the undoped silicon layer is located on the upper part of the P-type doped region and on both sides of the photosensitive layer, the N+ type doped region is formed on the top of the photosensitive layer, and the P-type doped region is connected to the Y-branch waveguide;
[0013] The metal plugs include a first metal plug, a second metal plug and a third metal plug, and the metal plugs are exposed in the silicon dioxide cladding, the first metal plug is electrically connected to the N+ type doped region, and the second metal plug and the third metal plug are electrically connected to the two P+ type doped regions respectively.
[0014] Optionally, the photosensitive layer includes a metal Ge layer, and the photosensitive layer is in a columnar or truncated cone shape.
[0015] Optionally, the ion doping concentration of the P+ type doping region is 1e18-1e20 / cm 3 The ion doping concentration of the P-type doping region is 1e14-1e18 / cm 3 The ion doping concentration of the N+ doping region is 1e18-1e20 / cm 3 .
[0016] Optionally, the width of the 8-channel structure is 300 μm to 400 μm, and the spacing distance between the single-mode input waveguides in the 8 spaced-apart waveguides is 180 μm to 200 μm.
[0017] The present invention also provides a method for manufacturing an 8-channel structure based on a polarization beam splitter and a photodetector, the manufacturing method comprising the following steps:
[0018] Providing an SOI substrate, wherein the SOI substrate comprises a silicon substrate, a dielectric layer, and a top silicon layer stacked in sequence, and the SOI substrate comprises a polarization beam splitter region and a photodetector region;
[0019] Patterning the top silicon to form eight waveguides arranged at intervals in the polarization beam splitter region, wherein the waveguides include a single-mode input waveguide, a double-etched waveguide, and a Y-branch waveguide connected in sequence, wherein the Y-branch waveguide includes a first branch waveguide and a second branch waveguide;
[0020] 16 photoelectric tubes are formed in the photoelectric detector area, and the 16 photoelectric tubes include 8 first photoelectric tubes arranged corresponding to the first branch waveguides and 8 second photoelectric tubes arranged corresponding to the second branch waveguides;
[0021] A silicon dioxide cladding is formed, the silicon dioxide cladding covering the polarization beam splitter region and the photodetector region.
[0022] Optionally, the method for manufacturing the photoelectric tube comprises the following steps:
[0023] Patterning the top silicon in the photodetector region and forming a silicon layer through an ion implantation process, wherein the silicon layer includes: a P-type doped region, two P+ type doped regions and an undoped silicon layer, wherein the two P+ type doped regions are in contact with the dielectric layer and are located on both sides of the P-type doped region, the undoped silicon layer is located on the upper surface of the P-type doped region, and the P-type doped region is connected to the Y-branch waveguide;
[0024] Patterning the undoped silicon layer and depositing a photosensitive layer, wherein the photosensitive layer is connected to the P-type doped region;
[0025] Forming an N+ type doped region on the top of the photosensitive layer by ion implantation;
[0026] A metal plug is formed in the photodetector region, the metal plug is exposed in the silicon dioxide cladding, and the metal plug includes a first metal plug electrically connected to the photosensitive layer and a second metal plug and a third metal plug electrically connected to the two P+ type doped regions respectively.
[0027] Optionally, the photosensitive layer includes a metal Ge layer, and the photosensitive layer is in a columnar or truncated cone shape.
[0028] Optionally, the ion doping concentration of the P+ type doping region is 1e18-1e20 / cm 3 The ion doping concentration of the P-type doping region is 1e14-1e18 / cm3 The ion doping concentration of the N+ doping region is 1e18-1e20 / cm 3 .
[0029] Optionally, the width of the 8-channel structure is 300 μm to 400 μm, and the spacing distance between the single-mode input waveguides in the 8 spaced-apart waveguides is 180 μm to 200 μm.
[0030] As described above, the 8-channel structure based on the polarization beam splitter and the photodetector of the present invention has the following beneficial effects: the present invention effectively combines the photodetector and the polarization beam splitter to form an 8-channel structure, which reduces the polarization sensitivity of the system while maintaining low losses. The 8-channel design effectively improves the bandwidth of the system and meets the needs of large-scale data transmission. The present invention also improves the structure of the photodetector, and designs the photosensitive layer of the photodetector into a truncated cone shape. The truncated cone-shaped photosensitive layer balances the influencing factors of current transmission and current expansion, ensuring that the photogenerated carriers diffuse in a very short time to enable the circuit to be quickly turned on. In addition, the present invention also provides a method for testing its own polarization isolation through the 8-channel structure, and the polarization isolation of the system is obtained by the photocurrent value of the photodetector, which simplifies the testing process and avoids errors such as loss caused by testing optical power. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] Figure 1 Shown is a schematic diagram of a top view of an 8-channel structure in the present invention.
[0032] Figure 2 Display as Figure 1 Schematic diagram of the enlarged structure of the middle H region.
[0033] Figure 3 It is shown that the 8-channel structure of the present invention is Figure 2 Schematic diagram of the cross-sectional structure at the AB position.
[0034] Figure 4-9 The steps of the method for making an 8-channel structure in the present invention are shown in Figure 2 Schematic diagram of the cross-sectional structure at the CD position.
[0035] Fig.10 Shown is a schematic diagram of the device connection structure of the 8-channel structure in the present invention.
[0036] Fig.11 Shown is the bandwidth test result of the 8-channel structure in the present invention.
[0037] Fig.12 Shown is the IV test curve of the 8-channel structure in the present invention.
[0038] Component number description
[0039] 1 Photodetector area
[0040] 2 Polarization beam splitter area
[0041] 11. First Photocell
[0042] 12 Second photocell
[0043] 21 Waveguide
[0044] 10 Silicon substrate
[0045] 20 Dielectric layer
[0046] 30 Top Silicon
[0047] 100 Undoped silicon layer
[0048] 101 P+ doping region
[0049] 102 P-type doping region
[0050] 103 Photosensitive layer
[0051] 104 N+ doping region
[0052] 105 Silica cladding
[0053] 106 Metal plug
[0054] 201 Single-mode input waveguide
[0055] 202 Double Etched Waveguide
[0056] 203 Y-branch waveguide
[0057] 2000 Polarization Beam Splitter
[0058] 1001 First Photodetector
[0059] 1002 Second photodetector
[0060] 2031 First Branch Waveguide
[0061] 2032 Second branch waveguide
[0062] 1061 First Metal Plug
[0063] 1062 Second metal plug
[0064] 1063 Third Metal Plug
[0065] L1 Width
[0066] L2 Spacing DETAILED DESCRIPTION
[0067] The following describes the embodiments of the present invention through specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention.
[0068] It should be noted that the illustrations provided in this embodiment only illustrate the basic concept of the present invention in a schematic manner. Although the illustrations only show the components related to the present invention rather than being drawn according to the number, shape and size of the components in actual implementation, the form, quantity, positional relationship and proportion of each component in actual implementation can be changed at will under the premise of realizing the technical solution of this party, and the component layout form may also be more complex. Therefore, it can be foreseen that the shape in the schematic diagram may change, for example, due to manufacturing technology and / or tolerance. Therefore, the exemplary embodiment should not be considered to be limited to the specific shape of the area shown in the figure, but may also include shape deviations caused by, for example, the manufacturing process.
[0069] Embodiment 1
[0070] This embodiment provides an 8-channel structure based on a polarization beam splitter and a photodetector, such as Figure 1-3 and Fig. 9 As shown, the 8-channel structure includes:
[0071] An SOI substrate, wherein the SOI substrate comprises a silicon substrate 10, a dielectric layer 20 and a top silicon layer 30 stacked in sequence, and the SOI substrate comprises a polarization beam splitter region 2 and a photodetector region 1;
[0072] The polarization beam splitter region 2 includes eight waveguides 21 arranged at intervals obtained by patterning the top silicon 30, wherein the waveguide 21 includes a single-mode input waveguide 201, a double-etched waveguide 202, and a Y-branch waveguide 203 connected in sequence, and the Y-branch waveguide 203 includes a first branch waveguide 2031 and a second branch waveguide 2032;
[0073] The photodetector area 1 includes 16 phototubes, and the 16 phototubes include 8 first phototubes 11 arranged corresponding to the first branch waveguides 2031 and 8 second phototubes 12 arranged corresponding to the second branch waveguides 2032;
[0074] The silicon dioxide cladding 105 covers the polarization beam splitter region 2 and the photodetector region 1 to form the 8-channel structure.
[0075] As an example, the photoelectric tube includes a silicon layer obtained by patterning the top silicon 30, a photosensitive layer 103, an N+ type doped region 104 and a metal plug 106; wherein,
[0076] The silicon layer includes a P-type doping region 102, two P+ type doping regions 101 and an undoped silicon layer 100, the two P+ type doping regions 101 are in contact with the dielectric layer 20 and are located on both sides of the P-type doping region 102, the photosensitive layer 103 is located on the P-type doping region 102, the undoped silicon layer 100 is located on the upper part of the P-type doping region 102 and on both sides of the photosensitive layer 103, the N+ type doping region 104 is formed on the top of the photosensitive layer 103, and the P-type doping region 102 of the photoelectric tube is connected to the Y-branch waveguide 203;
[0077] The metal plug 106 includes a first metal plug 1061, a second metal plug 1062 and a third metal plug 1063, and the metal plug 106 is exposed in the silicon dioxide cladding 105, the first metal plug 1061 is electrically connected to the N+ type doped region 104, and the second metal plug 1062 and the third metal plug 1063 are electrically connected to the two P+ type doped regions 101 respectively.
[0078] As an example, the width L1 of the 8-channel structure is 300 μm to 400 μm, preferably 345 μm; the spacing distance L2 between the single-mode input waveguides 201 in the 8 waveguides 21 arranged at intervals is 180 μm to 200 μm, preferably 190 μm. Providing intervals between the 8 waveguides 21 can effectively avoid crosstalk.
[0079] As an example, the photosensitive layer 103 includes a metal Ge layer, and the shape of the photosensitive layer 103 may also be a columnar shape. Considering the influencing factors of current transmission and current expansion, in this embodiment, the photosensitive layer 103 is preferably in a truncated cone shape.
[0080] Specifically, the photosensitive layer 103 generates carriers and emits photocurrent after receiving light. In order to make the speed of photocurrent transmission as large as possible and thus reduce the response time, the cross-sectional area of the photosensitive layer 103 perpendicular to the direction of photocurrent transmission needs to be as small as possible to reduce current expansion; however, for CMOS circuits, it is necessary to achieve rapid conduction of the source and drain, so the P-type doped region 102 in the present invention needs to expand the carriers generated by the photosensitive layer 103 as much as possible to enable the source and drain to conduct quickly. Therefore, considering the two factors of current transmission and current expansion, the truncated cone-shaped photosensitive layer 103 is the best choice. Among them, the truncated cone-shaped or columnar photosensitive layer 103 can be obtained by etching an inverted truncated cone-shaped hole or a columnar hole through a hard mask and then depositing a photosensitive material.
[0081] As an example, the stacking thickness of the undoped silicon layer 100 and the P-type doping region 102 is the same as the thickness of the P+ type doping region 101, including 100-400nm, such as 100nm, 200nm, 400nm, etc. Further, the P+ type doping region 101 is a heavily doped region, which is intentionally doped, and its ion doping concentration is 1e18-1e20 / cm 3 In this embodiment, the preferred doping concentration is 1e19 / cm 3 The P-type doping region 102 is a lightly doped region, which is intentionally doped, and its ion doping concentration is 1e14-1e18 / cm 3 In this embodiment, the preferred doping concentration is 1e16 / cm 3 ; The undoped silicon layer 100 is not doped.
[0082] Specifically, the P-type doping region 102 is designed as a lightly doped region. The main reason is to form a Schottky contact at the interface between the P-type doping region 102 and the photosensitive layer 103. Too high a doping concentration will form an ohmic contact at the interface. If there is no potential barrier at the interface, bidirectional conductivity will occur, and the unidirectional conductivity of the pn junction will not be present. Therefore, in order to ensure that a Schottky contact is formed at the interface between the P-type doping region 102 and the photosensitive layer 103, the P-type doping region 102 cannot have an excessively high doping concentration. In this embodiment, the preferred doping concentration of the P-type doping region 102 is 1e16 / cm 3 .
[0083] As an example, the ion doping concentration of the N+ doping region 104 is 1e18-1e20 / cm 3 In this embodiment, the preferred doping concentration is 1e19 / cm 3 .
[0084] Specifically, the silicon substrate 10, the dielectric layer 20, the waveguide 21 and the silicon dioxide cladding 105 constitute a polarization beam splitter 2000, and the silicon substrate 10, the dielectric layer 20, the photoelectric tube and the silicon dioxide cladding 105 constitute a photodetector, and the photodetector includes a first photodetector 1001 and a second photodetector 1002. It should be noted that, Figure 3 Only the second branch waveguide 2032 of the polarization beam splitter 2000 is shown, and the first photodetector 1001 and Figure 3 The structure of the second photodetector 1002 shown in FIG. 1 is similar to that of the second photodetector 1002 and will not be described in detail here.
[0085] Specifically, the working process of the 8-channel structure is as follows: light is incident from the left side of the single-mode input waveguide 201 of the polarization beam splitter 2000, undergoes mode conversion in the double-etched waveguide 202, and finally passes through the first branch waveguide 2031 and the second branch waveguide 2032 of the Y-branch waveguide 203 and reaches the first photodetector 1001 and the second photodetector 1002 respectively. The single-mode light entering the single-mode input waveguide 201 can be a TM0 mode (polarization direction is perpendicular to the propagation direction) and a TE0 mode (polarization direction is parallel to the propagation direction). By utilizing the mode conversion function of the polarization beam splitter 2000, when the TE0 mode is incident on the single-mode input waveguide 201, after the mode conversion of the polarization beam splitter 2000, the TE0 mode remains unchanged and is output from the first branch waveguide 2031 of the Y branch waveguide 203 and then reaches the first photodetector 1001; when the TM0 mode is incident on the single-mode input waveguide 201, after the mode conversion of the polarization beam splitter 2000, the TM0 mode is converted into the TE0 mode and is output from the second branch waveguide 2032 of the Y branch waveguide 203 and then reaches the second photodetector 1002. Thus, the light of two different polarization states, the TE mode and the TM mode, are separated according to different propagation directions. In this process, light is incident from the Y-branch waveguide 203 to the P-type doped region 102, and then reaches the bottom surface of the photosensitive layer 103. Thanks to the truncated cone structure of the photosensitive layer 103, the bottom surface of the photosensitive layer 103 can have more contact area with the P-type doped region 102, thereby increasing the light receiving area to further reduce the response time.
[0086] like Fig.11 The figure shows the bandwidth test result of the 8-channel structure in the present invention. From the test result, when the loss is -3dB, the bandwidth of the 8-channel structure is 20GB, which can meet the needs of large-scale data transmission.
[0087] like Fig.12 The IV test curve of the 8-channel structure in the present invention is shown. From the test results, the IV curve of the 8-channel structure conforms to the characteristics of the semiconductor device and exhibits a small reverse leakage current, indicating that the 8-channel structure has good electrical performance.
[0088] This embodiment also provides a method for manufacturing an 8-channel structure based on a polarization beam splitter and a photodetector, the manufacturing method comprising the following steps:
[0089] An SOI substrate is provided, wherein the SOI substrate comprises a silicon substrate 10, a dielectric layer 20, and a top silicon layer 30 stacked in sequence, and the SOI substrate comprises a polarization beam splitter region 2 and a photodetector region 1, such as Figure 1 and Figure 4 shown.
[0090] The top silicon 30 is patterned to form eight waveguides 21 arranged at intervals in the polarization beam splitter region. The waveguide 21 includes a single-mode input waveguide 201, a double-etched waveguide 202, and a Y-branch waveguide 203 connected in sequence. The Y-branch waveguide 203 includes a first branch waveguide 2031 and a second branch waveguide 2032. Figure 1 As shown;
[0091] 16 photoelectric tubes are formed in the photoelectric detector area 1, and the 16 photoelectric tubes include 8 first photoelectric tubes 11 arranged corresponding to the first branch waveguides 2031 and 8 second photoelectric tubes 12 arranged corresponding to the second branch waveguides 2032;
[0092] A silicon dioxide cladding 105 is formed, and the silicon dioxide cladding 105 covers the polarization beam splitter region 2 and the photodetector region 1. Figure 3 shown.
[0093] As an example, the method for manufacturing the photoelectric tube includes the following steps:
[0094] The top silicon 30 is patterned in the photodetector region 1 and a silicon layer is formed by an ion implantation process, wherein the silicon layer includes: a P-type doped region 102, two P+ type doped regions 101 and an undoped silicon layer 100, wherein the two P+ type doped regions 101 are in contact with the dielectric layer 20 and are located on both sides of the P-type doped region 102, and the undoped silicon layer 100 is located on the upper surface of the P-type doped region 102, wherein the P-type doped region 102 is connected to the Y-branch waveguide 203, as shown in FIG. Figure 3 and Figure 5 As shown;
[0095] The undoped silicon layer 100 is patterned and a photosensitive layer 103 is deposited. The photosensitive layer 103 is connected to the P-type doped region 102. Figure 6 and Figure 7 As shown;
[0096] An N+ doped region 104 is formed on the top of the photosensitive layer 103 by ion implantation. Figure 8 As shown;
[0097] A metal plug 106 is formed in the photodetector region 1, the metal plug 106 is exposed in the silicon dioxide cladding layer 105, and the metal plug 106 includes a first metal plug 1061 electrically connected to the photosensitive layer 103, and a second metal plug 1062 and a third metal plug 1063 electrically connected to the two P+ doped regions 101, respectively. Fig. 9 shown.
[0098] Specifically, the width L1 of the 8-channel structure is 300 μm to 400 μm, preferably 345 μm; the spacing distance L2 between the single-mode input waveguides 201 of the 8 spaced waveguides 21 is 180 μm to 200 μm, preferably 190 μm. Providing spacing between the 8 waveguides 21 can effectively avoid crosstalk.
[0099] As an example, the photosensitive layer 103 may include a metal Ge layer, but is not limited to this. The shape of the photosensitive layer 103 may also be cylindrical, preferably a truncated cone shape as shown in this embodiment, wherein the truncated cone shape of the photosensitive layer 103 is a choice made by comprehensively considering the influencing factors of current transmission and current expansion.
[0100] Specifically, the photosensitive layer 103 generates carriers and emits photocurrent after receiving light. In order to make the speed of photocurrent transmission as large as possible and thus reduce the response time, the cross-sectional area of the photosensitive layer 103 perpendicular to the direction of photocurrent transmission needs to be as small as possible to reduce current expansion; however, for CMOS circuits, it is necessary to achieve rapid conduction of the source and drain, so the P-type doped region 102 in the present invention needs to expand the carriers generated by the photosensitive layer 103 as much as possible to enable the source and drain to conduct quickly. Therefore, considering the two factors of current transmission and current expansion, the truncated cone-shaped photosensitive layer 103 is the best choice. Among them, the truncated cone-shaped or columnar photosensitive layer 103 can be obtained by etching an inverted truncated cone-shaped hole or a columnar hole through a hard mask and then depositing a photosensitive material, and no excessive restrictions are made here.
[0101] As an example, the stacking thickness of the undoped silicon layer 100 and the P-type doping region 102 is the same as the thickness of the P+ type doping region 101, which is 100-400nm. The P+ type doping region 101 is a heavily doped region, which is intentionally doped, and its ion doping concentration is 1e18-1e20 / cm 3 In this embodiment, the preferred doping concentration is 1e19 / cm 3 The P-type doping region 102 is a lightly doped region, which is intentionally doped, and its ion doping concentration is 1e14-1e18 / cm 3 In this embodiment, the preferred doping concentration is 1e16 / cm 3 ; The undoped silicon layer 100 is not doped.
[0102] Specifically, the P-type doping region 102 is designed as a lightly doped region. The main reason is to form a Schottky contact at the interface between the P-type doping region 102 and the photosensitive layer 103. Too high a doping concentration will form an ohmic contact at the interface. If there is no potential barrier at the interface, bidirectional conductivity will occur, and the unidirectional conductivity of the pn junction will not be present. Therefore, in order to ensure that a Schottky contact is formed at the interface between the P-type doping region 102 and the photosensitive layer 103, the P-type doping region 102 cannot have an excessively high doping concentration. In this embodiment, the preferred doping concentration of the P-type doping region 102 is 1e16 / cm 3 .
[0103] As an example, the ion doping concentration of the N+ doping region 104 is 1e18-1e20 / cm 3 In this embodiment, the preferred doping concentration is 1e19 / cm 3 .
[0104] Embodiment 2
[0105] Based on the first embodiment, this embodiment provides a method for testing the on-chip loss of the polarization beam splitter 2000 using the above 8-channel structure. Fig.10 As shown, the specific steps include:
[0106] T1: When the light is directly coupled into the photodetector by the grating without passing through the polarization beam splitter 2000, the photocurrent I1, input light power E1, and coupling loss C1 of the grating of the photodetector are measured, and then the responsivity R of the photodetector is obtained by using formula (a);
[0107] The responsivity R is the ratio of the photoelectric detector output photocurrent to the input light power, which represents the photoelectric conversion efficiency of the photoelectric detector. Specifically, in this embodiment, the responsivity R of the photoelectric detector is measured to be 0.75A / W.
[0108] T2: When the light passes through the polarization beam splitter 2000 and the grating coupling and enters the photodetector, the photocurrent I2 of the photodetector, the input optical power E2, and the coupling loss C2 of the grating are measured, and then the on-chip loss W of the polarization beam splitter 2000 is obtained using formula (a);
[0109]
[0110] The calculation principle of the above formula (a) is as follows: a numerical relationship is established between the results of the two tests (i.e., whether or not a polarization beam splitter is added) through the responsivity R, and the on-chip loss W of the polarization beam splitter 2000 is calculated.
[0111] Specifically, in this embodiment, the on-chip loss W of the polarization beam splitter 2000 is measured to be 1.19 dB using the above-mentioned method for testing on-chip loss, indicating that the 8-channel structure has a relatively low loss.
[0112] This embodiment also provides a method for testing the polarization isolation of the 8-channel structure. Polarization isolation refers to the degree of isolation between different polarized lights. The greater the isolation, the better the effect of converting the TM0 mode into the TE0 mode, which means that the distinction between TE0 and TM0 is greater. In layman's terms, a large isolation can effectively prevent crosstalk between signals and cause distortion of the transmitted signal. Fig.10 As shown, the above method for testing polarization isolation specifically includes the following steps:
[0113] D1: measuring the responsivity R of the photodetector;
[0114] Specifically, the responsivity R is the ratio of the output photocurrent of the photodetector to the input optical power, which represents the photoelectric conversion efficiency of the photodetector. Specifically, in this embodiment, the responsivity R of the photodetector is measured to be 0.75 A / W.
[0115] D2: Measure the polarization isolation in TE mode. Specifically, the light in TE0 mode is incident on the single-mode input waveguide 201, and the photocurrents A1 and A2 of the first photodetector 1001 and the second photodetector 1002 are measured respectively. The responsivity R measured in step D1 and the photocurrents A1 and A2 measured in this step are substituted into the following formula (b) to calculate P1 and P2, wherein the values of P1 and P2 are the optical powers input to the first photodetector 1001 and the second photodetector 1002 respectively, and finally the polarization isolation in TE mode is calculated to be P1-P2;
[0116]
[0117] D3: Use steps similar to those in D2 to measure the polarization isolation in TM mode. Specifically, light in TM0 mode is incident on the single-mode input waveguide 201, and the photocurrents A1 and A2 of the first photodetector 1001 and the second photodetector 1002 are measured respectively. The responsivity R measured in step D1 and the photocurrents A1 and A2 measured in this step are substituted into the above formula (b) to calculate P1 and P2, wherein the values of P1 and P2 are the optical powers input to the first photodetector 1001 and the second photodetector 1002 respectively, and finally the polarization isolation in TM mode is calculated to be P2-P1;
[0118] The calculation principle of the above formula (b) is: a numerical relationship is established between the photocurrent and the optical power through the response R, the optical power is obtained using the photocurrent, and then the polarization isolation is calculated.
[0119] It should be noted that, as described above, when the TE0 mode light is incident on the single-mode input waveguide 201, after the mode conversion of the polarization beam splitter 2000, the TE0 mode remains unchanged and is output from one branch of the Y-branch waveguide 203 and then reaches the first photodetector 1001; when the TM0 mode is incident on the single-mode input waveguide 201, after the mode conversion of the polarization beam splitter 2000, the TM0 mode is converted to the TE0 mode and is output from another branch of the Y-branch waveguide 203 and then reaches the second photodetector 1002. Therefore, in the TE mode, the P1 obtained by the test is greater than P2; while in the TM mode, the P1 obtained by the test is less than P2;
[0120] Specifically, in this embodiment, the polarization isolation of the 8-channel structure in the TE mode and the TM mode is calculated and tested using the above method for testing polarization isolation, and the polarization isolation is 14.59 dB and 5.141 dB respectively. It can be seen that the 8-channel structure has good polarization isolation, thereby improving the measurement accuracy of the photodetector.
[0121] The present embodiment provides a method for testing its own polarization isolation through the 8-channel structure, and obtains the polarization isolation of the system through the photocurrent of the photodetector. Originally, it was necessary to obtain the polarization isolation by testing the optical power received by the first photodetector and the second photodetector. After adopting the above method, the polarization isolation can be obtained directly by testing the photocurrent of the first photodetector and the second photodetector, which simplifies the test process and the instruments used, and avoids errors such as loss caused by testing the optical power, making the test result of the polarization isolation more accurate.
[0122] In summary, the present invention effectively combines the photodetector and the polarization beam splitter to form an 8-channel structure, which reduces the polarization sensitivity of the system while maintaining low losses. The 8-channel design effectively improves the bandwidth of the system and meets the needs of large-scale data transmission. The present invention also improves the structure of the photodetector, and designs the photosensitive layer of the photodetector into a truncated cone shape. The truncated cone-shaped photosensitive layer balances the influencing factors of current transmission and current expansion, ensuring that the photogenerated carriers diffuse in a very short time to allow the circuit to be quickly turned on. In addition, the present invention also provides a method for testing its own polarization isolation through the 8-channel structure, and the polarization isolation of the system is obtained by the photocurrent value of the photodetector, which simplifies the test process and avoids errors such as loss caused by testing optical power.
[0123] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.
Claims
1. An 8-channel structure based on a polarization beam splitter and a photodetector, characterized in that: The 8-channel structure includes at least: An SOI substrate, wherein the SOI substrate comprises a silicon substrate, a dielectric layer and a top silicon layer stacked in sequence, and the SOI substrate comprises a polarization beam splitter region and a photodetector region; The polarization beam splitter region includes 8 waveguides arranged at intervals obtained by patterning the top silicon, the waveguides include a single-mode input waveguide, a double-etched waveguide and a Y-branch waveguide connected in sequence, and the Y-branch waveguide includes a first branch waveguide and a second branch waveguide; The photodetector area includes 16 phototubes, and the 16 phototubes include 8 first phototubes corresponding to the first branch waveguides and 8 second phototubes corresponding to the second branch waveguides; A silicon dioxide cladding layer covers the polarization beam splitter region and the photodetector region to form the 8-channel structure.
2. The 8-channel structure according to claim 1, characterized in that: The photoelectric tube includes a silicon layer, a photosensitive layer, an N+ type doped region and a metal plug; wherein, The silicon layer includes a P-type doped region, two P+ type doped regions and an undoped silicon layer, the two P+ type doped regions are in contact with the dielectric layer and are located on both sides of the P-type doped region, the photosensitive layer is located on the P-type doped region, the undoped silicon layer is located on the upper part of the P-type doped region and on both sides of the photosensitive layer, the N+ type doped region is formed on the top of the photosensitive layer, and the P-type doped region is connected to the Y-branch waveguide; The metal plugs include a first metal plug, a second metal plug and a third metal plug, and the metal plugs are exposed in the silicon dioxide cladding, the first metal plug is electrically connected to the N+ type doped region, and the second metal plug and the third metal plug are electrically connected to the two P+ type doped regions respectively.
3. The 8-channel structure according to claim 2, characterized in that: The photosensitive layer includes a metal Ge layer, and the shape of the photosensitive layer is a column or a truncated cone.
4. The 8-channel structure according to claim 2, characterized in that: The ion doping concentration of the P+ type doping region is 1e18-1e20 / cm 3 The ion doping concentration of the P-type doping region is 1e14-1e18 / cm 3 The ion doping concentration of the N+ doping region is 1e18-1e20 / cm 3 .
5. The 8-channel structure according to claim 1, characterized in that: The width of the 8-channel structure is 300 μm to 400 μm, and the spacing between the 8 single-mode input waveguides is 180 μm to 200 μm.
6. A method for manufacturing an 8-channel structure based on a polarization beam splitter and a photodetector, characterized in that: The production method comprises the following steps: Providing an SOI substrate, wherein the SOI substrate comprises a silicon substrate, a dielectric layer, and a top silicon layer stacked in sequence, and the SOI substrate comprises a polarization beam splitter region and a photodetector region; Patterning the top silicon to form eight waveguides arranged at intervals in the polarization beam splitter region, wherein the waveguides include a single-mode input waveguide, a double-etched waveguide, and a Y-branch waveguide connected in sequence, wherein the Y-branch waveguide includes a first branch waveguide and a second branch waveguide; 16 photoelectric tubes are formed in the photoelectric detector area, and the 16 photoelectric tubes include 8 first photoelectric tubes arranged corresponding to the first branch waveguides and 8 second photoelectric tubes arranged corresponding to the second branch waveguides; A silicon dioxide cladding is formed, the silicon dioxide cladding covering the polarization beam splitter region and the photodetector region.
7. The manufacturing method according to claim 6, characterized in that: The manufacturing method of the photoelectric tube comprises the following steps: Patterning the top silicon in the photodetector region and forming a silicon layer through an ion implantation process, wherein the silicon layer includes: a P-type doped region, two P+ type doped regions and an undoped silicon layer, wherein the two P+ type doped regions are in contact with the dielectric layer and are located on both sides of the P-type doped region, the undoped silicon layer is located on the upper surface of the P-type doped region, and the P-type doped region is connected to the Y-branch waveguide; Patterning the undoped silicon layer and depositing a photosensitive layer, wherein the photosensitive layer is connected to the P-type doped region; forming an N+ type doped region on top of the photosensitive layer by ion implantation; A metal plug is formed in the photodetector region, the metal plug is exposed in the silicon dioxide cladding, and the metal plug includes a first metal plug electrically connected to the photosensitive layer and a second metal plug and a third metal plug electrically connected to the two P+ type doped regions respectively.
8. The method according to claim 7, characterized in that: The photosensitive layer includes a metal Ge layer, and the shape of the photosensitive layer is a column or a truncated cone.
9. The manufacturing method according to claim 7, characterized in that: The ion doping concentration of the P+ type doping region is 1e18-1e20 / cm 3 The ion doping concentration of the P-type doping region is 1e14-1e18 / cm 3 The ion doping concentration of the N+ doping region is 1e18-1e20 / cm 3 .
10. The manufacturing method according to claim 6, characterized in that: The width of the 8-channel structure is 300 μm to 400 μm, and the spacing distance between the single-mode input waveguides in the 8 waveguides arranged at intervals is 180 μm to 200 μm.
Citation Information
Patent Citations
Eight-channel structure based on polarization beam splitter and photoelectric detector
CN215266304U