An asymmetric directional coupler and mode-controllable generator, optical circulator

Abstract

The application provides an asymmetric directional coupler, comprising a substrate, wherein a top layer of silicon is arranged on the substrate, the top layer of silicon comprises a first port, a coupling area, a first output waveguide and a second output waveguide, one end of the coupling area is connected with the first port, and the other end is connected with the first output waveguide and the second output waveguide, a phase change material layer is arranged on the coupling area, the coupling area is divided into N*M square units, by adjusting the state of the square units, a non-periodic first perforation array satisfying a predetermined first output target is formed, and the first output target refers to the sum of the transmittances of the first output waveguide and the second output waveguide; the asymmetric directional coupler provided by the application is small in size, easy to process, stable in performance and capable of realizing regulation and control of an optical propagation path.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of micro-nano optoelectronic devices and components, and particularly relates to an asymmetric directional coupler and a mode-controllable mode generator and optical circulator. BACKGROUND

[0002] Due to the compatibility with existing complementary metal-oxide semiconductor (CMOS) technology and dense integration technology, silicon photonics is a very promising optical interconnection platform. Photonics integrated circuits (PICs) have also made great progress in the past decade.

[0003] Nowadays, the most advanced photonic integrated circuits can integrate hundreds of optical devices. Among them, optical switches are essential components, which are used for the dynamic routing of light in different paths. On-chip optical controllable devices are usually based on electro-optic or thermo-optic effects, both of which use electricity or heat to act on the waveguide to slightly change the refractive index of the waveguide to achieve the switching of the light transmission channel, i.e. the switching function. However, this optical switch mechanism is volatile, and continuous energy is needed to maintain the state of the switch, and the optical switch designed by this mechanism is complex in process and large in size, which is not suitable for large-scale integration.

[0004] In order to expand the data capacity, mode division multiplexing (MDM) technology has been widely concerned. Among them, the mode generator is an essential component, which can excite and switch different modes of the carrier in the optical waveguide. In addition, the optical isolator and the optical circulator are a kind of non-reciprocal device that can make light propagate unidirectionally. Among them, the optical isolator is a two-port device, a one-way street for light, which plays a crucial role in preventing unnecessary back reflections and light interactions. Similarly, the optical circulator is a multi-port device, a detour for light, which routes light from each input port to an output port in a non-reciprocal manner. They are indispensable components in optical networks. Although the mode generator and the optical circulator have been reported, they mostly have the disadvantages of being uncontrollable, large in size and complex in design method.

[0005] In recent years, Ge-Sb-Te compound optical phase change materials, such as Ge2Sb2Te5 (GST) and Ge2Sb2Se4Te1 (GSST), integrated on silicon waveguides, have shown the feasibility of designing tunable devices. Unlike electro-optic or thermo-optic effects that use electricity or heat to act on waveguides to slightly change the refractive index, Ge-Sb-Te compounds not only can greatly change the refractive index of waveguides to facilitate the design of small-sized optical tuning devices, but also the phase change is non-volatile, and does not require continuous external energy to maintain its state. Therefore, it can be seen that the tunable device based on phase change material is a very potential solution.

[0006] In view of this, designing a tunable mode generator and circulator based on phase change material has important significance for solving the problems of existing devices, such as non-tunable, large size and complex design method. SUMMARY

[0007] The technical problem to be solved by the present application is to overcome the deficiencies of the prior art optical switch, such as complex process, volatility, large size, etc., and to provide a silicon-based asymmetric directional coupler based on phase change material optimized by using a direct binary search (Direct-binary-search, DBS) algorithm.

[0008] In order to achieve the above-mentioned purpose, the technical scheme of the present application is as follows: an asymmetric directional coupler, comprising a substrate, wherein the substrate is provided with a top layer of silicon, the top layer of silicon comprises a first port, a coupling region, a first output waveguide and a second output waveguide, one end of the coupling region is connected with the first port, and the other end is connected with the first output waveguide and the second output waveguide, a phase change material layer is covered on the coupling region, the coupling region is divided into N×M rectangular units, by adjusting the state of the rectangular units, a non-periodic first perforated array satisfying a predetermined first output target is formed, and the first output target refers to the sum of the transmittances of the first output waveguide and the second output waveguide.

[0009] Preferably, the coupling region comprises an upper waveguide, a waveguide gap and a lower waveguide, the first port, the upper waveguide and the first output waveguide are connected in sequence, the lower waveguide is connected with the second output waveguide, the phase change material layer covers the top surface of the lower waveguide, and the waveguide gap is arranged between the upper waveguide and the lower waveguide.

[0010] Preferably, the phase change material layer is a GSST material.

[0011] Preferably, at least one of the first output waveguide and the second output waveguide is a curved waveguide.

[0012] Preferably, the substrate thickness is 3 μm, the top layer silicon thickness is 220 nm, the coupling region length is 10 μm, the upper waveguide width is 380 nm, the upper waveguide is divided into 4×100 first rectangular units, each first rectangular unit size is 95 nm×100 nm, the depth is 220 nm, and the initial state is no punch; the lower waveguide and the phase change material layer width are both 350 nm, the phase change material layer thickness is 40 nm, the phase change material layer is divided into 4×100 second rectangular units, each second rectangular unit size is 87.5 nm×100 nm, the depth is 40 nm, and the initial state is no punch; the waveguide spacing is divided into 2×100 third rectangular units, each third rectangular unit size is 100 nm×100 nm, the depth is 220 nm, and the initial state is punch.

[0013] The application also provides a controllable mode generator, comprising the above asymmetric directional coupler, the mode division multiplexer optimization area and a third output waveguide, one end of the mode division multiplexer optimization area is connected with the first output waveguide and the second output waveguide respectively, and the other end is connected with the third output waveguide, the mode division multiplexer optimization area is divided into X×Y fourth rectangular units, by adjusting the state of the fourth rectangular units, a non-periodic second punch array satisfying a predetermined second output target is formed, and the second output target refers to the sum of the transmittances of two output different modes in the third output waveguide.

[0014] Preferably, the third output waveguide width is 900 nm, and the two different modes are TE0 mode from 1540 nm to 1560 nm waveband and TE1 mode from 1540 nm to 1560 nm waveband.

[0015] Preferably, the optimization area size is 2400 nm×3000 nm, the mode division multiplexer optimization area is divided into 24×30 fourth rectangular units of 100 nm×100 nm, and the first output waveguide and the second output waveguide and the mode division multiplexer optimization area connection end spacing is 1635 nm.

[0016] The application also provides an optical circulator, comprising the above asymmetric directional coupler, a plurality of asymmetric directional couplers are arranged in a ring array, and the first output waveguide of one asymmetric directional coupler is connected with the second output waveguide of the adjacent asymmetric directional coupler.

[0017] Preferably, the ring optical mode is TE0 mode from 1540 nm to 1560 nm waveband.

[0018] The application has the advantages of small product size, easy processing, stable performance, low loss transmission, and solving the problems of uncontrolled, large size and complex design method of the existing optical switch. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 Structure diagram of asymmetric directional coupler provided by one embodiment of the present application.

[0020] Figure 2 Initial structure diagram of the embodiment shown in FIG. 1. Figure 1

[0021] Figure 3 Parameter scanning result of the embodiment shown in FIG. 1. Figure 1

[0022] Two-dimensional plane structure diagram of the embodiment shown in FIG. 1. Figure 4 Figure 1 Transmission spectrum of the embodiment shown in FIG. 1.

[0023] Figure 5 Figure 1 Structure diagram of controllable mode generator provided by one embodiment of the present application.

[0024] Figure 6 Two-dimensional plane structure diagram of the embodiment shown in FIG. 2.

[0025] Figure 7 Initial structure diagram of mode division multiplexer optimization area of the embodiment shown in FIG. 2. Figure 6

[0026] Figure 8 Transmission spectrum of the embodiment shown in FIG. 2. Figure 6

[0027] Structure diagram of optical circulator provided by one embodiment of the present application. Figure 9 Figure 6 Two-dimensional plane structure diagram of the embodiment shown in FIG. 3.

[0028] Figure 10 Transmission spectrum of the embodiment shown in FIG. 3.

[0029] Figure 11 Figure 10 Two-dimensional plane structure diagram of the embodiment shown in FIG. 4.

[0030] Figure 12 Transmission spectrum of the embodiment shown in FIG. 4. Figure 10

[0031] ​​​​​​​In the figure, 1, first port; 2, coupling area; 21, upper waveguide; 22, lower waveguide; 221, second rectangular unit; 23, waveguide gap; 231, third rectangular unit; 24, phase change material layer; 3, first output waveguide; 4, second output waveguide; 5, mode division multiplexer optimization area; 51, fourth rectangular unit; 6, third output waveguide; 7, second port; 8, third port; 9, fourth port; 10, fifth port; 11, first asymmetric directional coupler; 12, second asymmetric directional coupler; 13, third asymmetric directional coupler; 14, fourth asymmetric directional coupler; 15, first curved waveguide; 16, second curved waveguide; 17, third curved waveguide; 18, fourth curved waveguide. DETAILED DESCRIPTION

[0032] The technical solutions of the present application will be further specifically described below in combination with the drawings and specific embodiments:

[0033] Embodiment one

[0034] Please see Figures 1-5 The asymmetric directional coupler provided by the present application comprises a substrate 100, the substrate 100 is provided with a top layer of silicon, the top layer of silicon comprises a first port 1, a coupling area 2, a first output waveguide 3 and a second output waveguide 4, one end of the coupling area 2 is connected with the first port 1, and the other end is connected with the first output waveguide 3 and the second output waveguide 4, the coupling area 2 is covered with a phase change material layer 24, the coupling area 2 is divided into N×M rectangular units, by adjusting the state of the rectangular units, a non-periodic first perforated array satisfying a predetermined first output target is formed, and the first output target refers to the sum of the transmittances of the first output waveguide 3 and the second output waveguide 4.

[0035] More specifically, the coupling area 2 comprises an upper waveguide 21, a lower waveguide 22 and a waveguide gap 23, the first port 1, the upper waveguide 21 and the first output waveguide 3 are connected in sequence and have the same width, the lower waveguide 22 is connected with the second output waveguide 4 and has the same width, the waveguide gap 23 is between the upper waveguide 21 and the lower waveguide 22, and the phase change material layer 24 covers the top surface of the lower waveguide 22.

[0036] More specifically, the phase change material layer 24 is a GSST material.

[0037] More specifically, at least one of the first output waveguide 3 and the second output waveguide 4 is a curved waveguide, which is used to decouple and thus reduce crosstalk.

[0038] More specifically, the asymmetric directional coupler is designed on a silicon on insulator (SOI) platform, the top silicon thickness is 220 nm, the substrate 100 silicon dioxide thickness is 3 pm, the upper cladding layer is air, the length W4 of the coupling region 2 is 10 pm, the width W1 of the upper waveguide 21 is 380 nm, the upper waveguide 21 is divided into 4x100 first rectangular units, each first rectangular unit has a size of 95 nm x 100 nm, a depth of 220 nm, and an initial state of no punching; the width W2 of the lower waveguide 22 and the phase change material layer 24 is 350 nm, the thickness of the phase change material layer 24 is 40 nm, the phase change material layer 24 is divided into 4x100 second rectangular units 221, each second rectangular unit 221 has a size of 87.5 nm x 100 nm, a depth of 40 nm, and an initial state of no punching; the width W3 of the waveguide gap 23 is 200 nm, and the waveguide gap 23 is divided into 2x100 third rectangular units 231, each third rectangular unit 231 has a size of 100 nm x 100 nm, a depth of 220 nm, and an initial state of punching.

[0039] Using intelligent algorithms to design optical devices relative to traditional methods can optimize less rely on physical models under high degrees of freedom, and is more conducive to designing small size and complex structure devices. Before using the direct binary search algorithm for optimization, the coupling region 2 is divided into a plurality of rectangular units, each rectangular unit has two material states, namely punching and no punching. Punching means that the rectangular unit is filled with air, and no punching means that the rectangular unit is filled with silicon. The state of each rectangular unit is determined by the algorithm to satisfy the objective function, specifically: in the direct binary search algorithm, the objective function reflecting the performance of the device is set, then the algorithm is used to calculate the objective function value of each rectangular unit in two states, and then the state when the objective function value is improved is retained.

[0040] When using the direct binary search algorithm to calculate the rectangular units, the row-by-row calculation and column-by-column calculation are alternately used to calculate all the units. One iteration is called once all the units are traversed, and after a plurality of iterations, the objective function values after the last iteration are compared. If the change value of the two objective function values is less than 0.1%, the objective function converges, and the algorithm stops. The row-by-row calculation means that it is calculated from left to right in the horizontal direction and from top to bottom in the vertical direction; the column-by-column calculation means that it is calculated from top to bottom in the vertical direction and from left to right in the horizontal direction.

[0041] The present application controls the propagation path of light by making the phase change material layer 24 on the asymmetric directional coupler reversibly change phase quickly by light or electricity, affecting the refractive index of the waveguide. Therefore, before completing the device of the present application, the asymmetric directional coupler needs to be designed, and when optimizing the asymmetric directional coupler, the objective function reflecting the performance of the device is set as the sum of the transmittances of the two output waveguides.

[0042] Since the direct binary search algorithm is a search algorithm, it is easy to converge locally prematurely, resulting in an unsatisfactory optimization result. In order to overcome this problem, when optimizing the asymmetric directional coupler, the initial structure is artificially set as Figure 2 , that is, the initial state of all first rectangular units on the upper waveguide 21 is not perforated, the lower waveguide 22 is not perforated, the initial state of all third rectangular units 231 on the waveguide gap 23 is perforated, and the initial state of all second rectangular units 221 on the phase change material layer 24 is not perforated. In order to obtain excellent initial structure parameters, the parameters of the initial structure need to be optimized.

[0043] First, some parameters are constrained, the width of the waveguide gap 23 is set to W3=200nm, the coupling length W4 is 10μm, the width of the phase change material layer 24 covering the surface of the lower waveguide 22 is consistent with the width of the lower waveguide 22, and the thickness of the phase change material layer 24 is 40nm. Under the condition of determining these parameters, the width parameters W1 of the upper waveguide 21 and the width parameters W2 of the lower waveguide 22 are scanned to optimize the initial structure with better performance. The parameter scanning result is shown in Figure 3 , it can be found that when the width W1 of the upper waveguide 21 is about 380nm and the width W2 of the lower waveguide 22 is about 350nm, the GSST of the phase change material layer 24 is adjusted to be amorphous, and due to phase matching, light is coupled into the lower waveguide 22 with good efficiency; the GSST of the phase change material layer 24 is adjusted to be crystalline, and due to phase mismatch, the coupling efficiency is poor, and the light still propagates in the upper waveguide 21. After weighing, the width W1 of the upper waveguide 21 is finally selected to be 380nm, the width W2 of the lower waveguide 22 is 350nm, and the width of the phase change material layer 24 on the surface of the lower waveguide 22 is also 350nm.

[0044] In addition, since the upper and lower waveguides will have a coupling effect, the distance between the two waveguides needs to be expanded at the output position to suppress the coupling effect and not affect the propagation of light in the waveguide. Therefore, at least one of the output portions connecting the upper and lower waveguides needs to be designed as a curved waveguide, that is, at least one of the first output waveguide 3 and the second output waveguide 4 is a curved waveguide, and in the present application, two curved waveguides are used.

[0045] After obtaining the initial structure designed artificially, the coupling region 2 needs to be divided into N×M rectangular units, N and M are integers, and the length of the rectangular edge also needs to be considered for processability. For this purpose, the coupling region 2 is divided into three parts, a total of 10×100 rectangular units, including 4×100 first rectangular units of the upper waveguide 21 part, 4×100 second rectangular units 221 of the phase change material layer 24 part, and 2×100 third rectangular units 231 of the waveguide gap 23 part. Each rectangular unit has two material states, which are punched and not punched respectively. Punched means that the rectangular unit is filled with air, and not punched means that the rectangular unit is filled with silicon.

[0046] The upper waveguide 21 is divided into 4×100 rectangular units, each first rectangular unit has a size of 95 nm×100 nm, a depth of 220 nm, and an initial state of not punched; because the present application is to change the effective refractive index by the 40 nm thick phase change material layer 24 to realize the adjusting effect, therefore the lower waveguide 22 is not divided into rectangular units and not punched; the phase change material layer 24 is divided into 4×100 second rectangular units 221, each second rectangular unit 221 has a size of 87.5 nm×100 nm, a depth of 40 nm, and only penetrates the phase change material layer 24, and the initial state is not punched; the waveguide gap 23 is divided into 2×100 third rectangular units 231, each third rectangular unit 231 has a size of 100 nm×100 nm, a depth of 220 nm, and an initial state of punched.

[0047] The direct binary search algorithm optimizes a total of 10×100 rectangular units on the initial structure. First, the algorithm selects the rectangular unit in the first row and the first column, and calculates the performance of the two states of punched and not punched by using the finite difference time domain method (FDTD). The performance is determined by the objective function set in the algorithm, which is called the figure of merit (FOM), defined as:

[0048] FOM=T A-GSST +T C-GSST

[0049] In the formula, T A-GSST is the transmittance of the second output waveguide 4 when the GSST material is in the amorphous state at the wavelength of 1540 nm to 1560 nm; T C-GSST is the transmittance of the first output waveguide 3 when the GSST material is in the crystalline state at the wavelength of 1540 nm to 1560 nm.

[0050] Then scan to the last rectangular unit by row, which is called an iteration. The next iteration scans to the last rectangular unit by column in the same way. Scan alternately by row and by column, and optimize, iterate multiple times, until the FOM value changes within 0.1% after two iterations, the algorithm converges, and the device performance is stable. The row calculation refers to sequentially from left to right in the horizontal direction and from top to bottom in the vertical direction; the column calculation refers to from top to bottom in the vertical direction and from left to right in the horizontal direction.

[0051] Figure 4 A two-dimensional planar structure diagram of an asymmetric directional coupler optimized using a direct binary search algorithm, Figure 5 The asymmetric directional coupler has a transmission spectrum, and when the GSST is in a non-crystalline state, the TE0 mode has a crosstalk lower than -16.4 dB in the first output waveguide 3 and an insertion loss less than 0.6 dB in the second output waveguide 4 in a bandwidth range of 1540 nm to 1560 nm; when the GSST is in a crystalline state, the TE0 mode has an insertion loss less than 1.0 dB in the first output waveguide 3 and a crosstalk lower than -16.0 dB in the second output waveguide 4.

[0052] Specifically, the working principle is as follows: TE0 mode light with a bandwidth of 1540 nm to 1560 nm is injected into the first port 1, and the phase change of the GSST is controlled by light or electricity; when the GSST is in a non-crystalline state, the input light is efficiently transmitted from the second output waveguide 4, and the crosstalk in the first output waveguide 3 is very low; when the GSST is in a crystalline state, the input light is efficiently transmitted from the first output waveguide 3, and the crosstalk in the second output waveguide 4 is very low. The simulation transmission spectrum of the asymmetric directional coupler provided in the embodiment also illustrates its excellent performance. Compared with other methods for designing an optical switch, the asymmetric directional coupler has the advantages of good performance, small size, easy processing, and the like.

[0053] Embodiment two

[0054] Please refer to Figures 6-9 The embodiment provides a mode generator, which comprises the asymmetric directional coupler, a mode division multiplexer optimization area 5, and a third output waveguide 6. One end of the mode division multiplexer optimization area 5 is connected with the first output waveguide 3 and the second output waveguide 4 respectively, and the other end is connected with the third output waveguide 6. The mode division multiplexer optimization area 5 is divided into X*Y fourth rectangular units 51. By adjusting the state of the fourth rectangular unit 51, a non-periodic second puncturing array satisfying a predetermined second output target is formed. The second output target refers to the sum of the transmission rates of two output different modes in the third output waveguide 6.

[0055] More specifically, the two different modes are TE0 mode from 1540nm to 1560nm waveband and TE1 mode from 1540nm to 1560nm waveband, and the width of the third output waveguide 6 is 900nm.

[0056] More specifically, the size of the mode division multiplexer optimization area 5 is 2400nm x 3000nm, and the mode division multiplexer optimization area 5 is divided into 24 x 30 fourth rectangular units 51 of 100nm x 100nm, and the spacing W5 between the connection end of the mode division multiplexer optimization area 5 and the first output waveguide 3 and the second output waveguide 4 is 1635nm.

[0057] In the embodiment, the first output waveguide 3 and the second output waveguide 4 are both curved waveguides.

[0058] The mode division multiplexer optimization area 5 is also designed by using the direct binary search algorithm, such as Figure 8 The initial structure of the mode division multiplexer optimization area 5 is that the size of the mode division multiplexer optimization area 5 is 2400nm x 3000nm, and the mode division multiplexer optimization area 5 is divided into 24 x 30 fourth rectangular units 51 of 100nm x 100nm, each fourth rectangular unit 51 has two material states, which are punching and non-punching, and the punching means that the fourth rectangular unit 51 is filled with air, and the non-punching means that the fourth rectangular unit 51 is filled with silicon. The direct binary search algorithm optimizes a total of 24 x 30 rectangular units on the initial structure. The performance is determined by the objective function set in the algorithm, which is defined as:

[0059]

[0060] In the formula, is the transmittance of TE0 mode from the first output waveguide 3 to the third output waveguide 6 in the waveband of 1540nm to 1560nm, and the TE0 mode is maintained; is the transmittance of TE0 mode from the second output waveguide 4 to the third output waveguide 6 in the waveband of 1540nm to 1560nm, and the TE0 mode is converted into TE1 mode. The width of the output waveguide is 900nm, which can support the propagation of TE0 and TE1 modes without loss.

[0061] The mode generator works by adjusting the GSST phase change on the asymmetric directional coupler to generate different mode light. Specifically, when TE0 mode light source is injected from the first port 1, the GSST is adjusted to be crystalline, the TE0 mode passes through the asymmetric directional coupler, enters the mode division multiplexer optimization area 5 from the first output waveguide 3, and maintains the TE0 mode output from the third output waveguide 6; when the TE0 mode light source is injected from the first port 1, the GSST is adjusted to be amorphous, the TE0 mode passes through the asymmetric directional coupler, enters the mode division multiplexer optimization area 5 from the second output waveguide 4, and the TE0 mode is converted into the TE1 mode output from the third output waveguide 6.

[0062] Figure 9 The transmission spectrum of the mode generator provided in the embodiment is that when the GSST is crystalline, the mode generator outputs the TE0 mode, the insertion loss is less than 1.5 dB, and the crosstalk is lower than-13.9 dB; when the GSST is amorphous, the mode generator outputs the TE1 mode, the insertion loss is less than 2.2 dB, and the crosstalk is lower than-14.8 dB. The mode generator provided in the embodiment can solve the problems of the existing mode generator, such as uncontrolled, large size, and complex design method.

[0063] Embodiment three

[0064] Please refer to Figures 10-12 The application also provides an optical circulator, which comprises the asymmetric directional coupler described above, and a plurality of asymmetric directional couplers are arranged in a ring array, and the first output waveguide 3 of one asymmetric directional coupler is connected with the second output waveguide 4 of an adjacent asymmetric directional coupler.

[0065] In the embodiment, the first output waveguide 3 and the second output waveguide 4 are connected to form a curved waveguide, and the optical circulator is formed by four asymmetric directional couplers 11-14 connected in series through a first curved waveguide 15, a second curved waveguide 16, a third curved waveguide 17, and a fourth curved waveguide 18, respectively, and the widths of the two ends of each curved waveguide are the same as the width of the connecting part.

[0066] It can be understood that according to different use requirements, the optical circulator can be formed by three, four, five or more asymmetric directional couplers connected in series, and the number of curved waveguides can be increased or decreased accordingly.

[0067] The working principle of the optical circulator is that the TE0 mode light source is outputted from the adjacent port clockwise or counterclockwise by adjusting the GSST phase change on the asymmetric directional coupler. Specifically, when the TE0 mode light source is injected from the second port 7, the GSST on the first asymmetric directional coupler 11 is crystalline, and the GSST on the second asymmetric directional coupler 12, the third asymmetric directional coupler 13 and the fourth asymmetric directional coupler 14 is amorphous, the TE0 mode is outputted from the third port 8 in the clockwise direction, and the like, the TE0 mode light source can be propagated clockwise in the direction of the port second port 7-third port 8-fourth port 9-fifth port 10-second port 7; when the TE0 mode light source is injected from the second port 7, the GSST on the first asymmetric directional coupler 11 is amorphous, and the GSST on the second asymmetric directional coupler 12, the third asymmetric directional coupler 13 and the fourth asymmetric directional coupler 14 is crystalline, the TE0 mode is outputted from the fifth port 10 in the counterclockwise direction, and the like, the TE0 mode light source can be propagated counterclockwise in the direction of the port second port 7-fifth port 10-fourth port 9-third port 8-second port 7.

[0068] Figure 12 The transmission spectrum of the optical circulator provided in the embodiment is that when the TE0 mode is injected into the second port 7 and the GSST layer on the first asymmetric directional coupler 11 is crystalline, the GSST layer on the other ports is amorphous, the TE0 mode light propagates clockwise in the direction of the port second port 7-third port 8-fourth port 9-fifth port 10-second port 7, the insertion loss is less than 1.2 dB, and the crosstalk is lower than -20.0 dB; when the TE0 mode is injected into the second port 7 and the GSST layer on the first asymmetric directional coupler 11 is amorphous, the GSST layer on the other ports is crystalline, the TE0 mode light propagates counterclockwise in the direction of the port second port 7-fifth port 10-fourth port 9-third port 8-second port 7, the insertion loss is less than 1.2 dB, and the crosstalk is lower than -29.4 dB. Compared with the devices designed by other methods, the optical circulator provided in the embodiment has the advantages of controllability, small size, intelligent design and the like.

[0069] The above embodiments are only used to illustrate the technical solutions of the present application, rather than limit the present application; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A programmable pattern generator, characterized by: The asymmetric directional coupler, the mode division multiplexer optimization area and the third output waveguide, one end of the mode division multiplexer optimization area is connected with the first output waveguide and the second output waveguide respectively, and the other end is connected with the third output waveguide, the mode division multiplexer optimization area is divided into X×Y fourth square units, by adjusting the state of the fourth square unit, a non-periodic second punch array satisfying a predetermined second output target is formed, the second output target refers to the sum of the transmittance of two output different modes in the third output waveguide; The asymmetric directional coupler, the mode division multiplexer optimization area and the third output waveguide, one end of the mode division multiplexer optimization area is connected with the first output waveguide and the second output waveguide respectively, and the other end is connected with the third output waveguide, the mode division multiplexer optimization area is divided into X×Y fourth square units, by adjusting the state of the fourth square unit, a non-periodic second punch array satisfying a predetermined second output target is formed, the second output target refers to the sum of the transmittance of two output different modes in the third output waveguide; The coupling area includes an upper waveguide, a waveguide gap and a lower waveguide, the first port, the upper waveguide and the first output waveguide are connected in sequence, the lower waveguide is connected with the second output waveguide, the phase change material layer covers the top surface of the lower waveguide, and the waveguide gap is between the upper waveguide and the lower waveguide; The thickness of the substrate is 3μm, the thickness of the top layer of silicon is 220nm, the length of the coupling area is 10μm, the width of the upper waveguide is 380nm, the upper waveguide is divided into 4×100 first square units, each first square unit has a size of 95nm×100nm, a depth of 220nm and an initial state of no punch, the width of the lower waveguide and the phase change material layer is 350nm, the thickness of the phase change material layer is 40nm, the phase change material layer is divided into 4×100 second square units, each second square unit has a size of 87.5nm×100nm, a depth of 40nm and an initial state of no punch, and the waveguide gap is divided into 2×100 third square units, each third square unit has a size of 100nm×100nm, a depth of 220nm and an initial state of punch.

2. The controllable mode generator of claim 1, wherein: The phase change material layer is a GSST material.

3. The controllable mode generator of claim 1, wherein: The two different modes are TE0 mode from 1540nm to 1560nm waveband and TE1 mode from 1540nm to 1560nm waveband, and the width of the third output waveguide is 900nm.

4. The controllable mode generator of claim 1, wherein: The size of the mode division multiplexer optimization area is 2400nm×3000nm, the mode division multiplexer optimization area is divided into 24×30 fourth square units with a size of 100nm×100nm, and the distance between the connection end of the mode division multiplexer optimization area and the first output waveguide and the second output waveguide is 1635nm.

Images