Low-phase-shift reconfigurable phase-change optical switch with ultra-wide free spectral range

Abstract

The invention discloses a low-phase-shift reconfigurable phase-change optical switch with an ultra-wide free spectral range, which comprises a silicon dioxide substrate layer and is characterized in that a silicon waveguide layer is arranged on the substrate layer, a heat conduction layer is arranged on the waveguide layer, and electrodes are symmetrically distributed on two sides of the silicon waveguide layer; the structure of the silicon waveguide layer is a side coupling structure of a bent waveguide and a nano-beam cavity waveguide, the two ends of the bent waveguide are respectively connected with a grating vertical coupler, a phase change material is arranged on the upper layer of a coupling area of the bent waveguide and located on the lower surface of the heat conduction layer, and the phase change material is connected with two electrodes and forms ohmic contact; a round hole array on the nano-beam cavity waveguide is designed to be of an axisymmetric structure and is divided into a gradual change area and a mirror surface area, the distance and the size of round holes in the gradual change area are the smallest in the center and gradually increased on the two sides, the distance and the size of the round holes in the mirror surface area are kept unchanged, and the round holes of the nano-beam cavity waveguide are filled with silicon dioxide. In two states of the switch, the wavelength shift is less than 2nm.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of microelectronics, and in particular to a low phase shift reconfigurable phase change optical switch with ultra-wide free spectral range. BACKGROUND

[0002] In the cutting-edge fields of autonomous driving, medical diagnosis, speech recognition, financial analysis and intelligent recommendation systems, advanced machine learning algorithms represented by artificial neural networks (ANN) are playing an increasingly important role. However, as the complexity of tasks continues to increase, the massive multiplication and accumulation operations required by neural networks pose a severe challenge to existing electronic computing hardware in terms of computation and power consumption. In this context, optical computing, with its low power consumption, high speed, large capacity information storage and inherent parallel processing capability, shows great potential. In particular, matrix vector multiplication based on wavelength division multiplexing technology (WDM-MVMs) provides an important path for realizing high-speed, large-scale and high-parallelism optical neuromorphic hardware.

[0003] Currently, WDM-MVM systems mainly rely on micro ring resonator (MRR) arrays for implementation, however, such devices have significant bottlenecks in terms of integration density, power efficiency and system scalability. On the one hand, limited by waveguide bending loss, the physical size of MRRs is difficult to further reduce, restricting the continuous optimization of chip area and power consumption; on the other hand, the limited free spectral range (FSR) limits the number of available channels, hindering large-scale cascading. Therefore, developing new photonic devices with wider FSR, narrower bandwidth and low phase shift characteristics is an urgent need for realizing complex neural network optical computing.

[0004] Silicon (Si) material has the advantage of low cost and full compatibility with mature CMOS manufacturing processes, which is conducive to large-scale and low-cost production of optoelectronic integrated chips. The silicon-on-insulator (SOI) platform provides strong light field confinement through a buried oxide layer, enabling the manufacture of low-loss, small-size optical waveguides, which is an ideal choice for realizing high-density integration and high-performance photonic devices such as modulators, detectors, optical switches.

[0005] The application of phase change materials in the field of waveguide photonics is increasingly expanding, covering a wide range of fields from on-chip photonic data storage to new computing paradigms. Phase change materials have a huge optical contrast (refractive index and extinction coefficient change) between crystalline and amorphous states, which can achieve extremely high storage contrast. Its state transition speed is fast (nanosecond level), energy consumption is low, and data is non-volatile storage, which does not lose power. These characteristics enable it to achieve ultra-high density, high speed, long life, and integration of computing and storage, which is the core of the next generation of optical discs and optical phase change optical switches. Optical switches, as key components for controlling and modulating optical signals, provide significant flexibility and efficiency for optical computing systems. However, traditional optical switches have small free spectral range, large bandwidth, large phase shift, and high power consumption, which severely restrict the practical application of large-scale photonic integrated circuits and optical neural networks. The silicon-based phase change optical switch currently used for optical computing is mainly a micro-ring series, with a free spectral range of less than 50 nm. The device cascade scale is limited by the free spectral range and bandwidth, and cannot achieve large-scale wavelength division multiplexing. To achieve more complex neural network computing, the device needs to have ultra-wide free spectral range and narrow bandwidth, low phase shift characteristics. SUMMARY

[0006] The technical problem to be solved by the present application is to provide an ultra-wide free spectral range low phase shift reconfigurable phase change optical switch with a free spectral range of 100 nm and a wavelength shift of less than 2 nm in two states of the switch.

[0007] The technical solution adopted by the present application to solve the above technical problem is: an ultra-wide free spectral range low phase shift reconfigurable phase change optical switch, comprising a silicon dioxide substrate layer, a silicon waveguide layer disposed on the substrate layer, a heat conduction layer disposed on the silicon waveguide layer, and electrodes symmetrically distributed on both sides of the silicon waveguide layer and located on the upper surface of the heat conduction layer; the structure of the silicon waveguide layer is a side coupling structure of a curved waveguide and a nanobeam cavity waveguide, the two ends of the curved waveguide are respectively connected with a grating vertical coupler, a phase change material is disposed on the upper layer of the coupling region of the curved waveguide and located on the lower surface of the heat conduction layer, the state of the phase change material is reversibly switched between crystalline and amorphous states under electric induction, and the phase change material connects the two electrodes through the heat conduction layer and forms an ohmic contact; The circular hole array on the nanobeam cavity waveguide is designed as an axisymmetric structure, which is distributed from the center to both sides and divided into a gradual change region and a mirror region, the circular holes in the gradual change region are designed to be smallest in the center and gradually increase on both sides, the spacing between the circular holes in the single-sided gradual change region satisfies where d1 is the minimum spacing between the circular holes, d2 is the spacing between the circular holes in the mirror region, and n is the number of circular holes in the single-sided gradual change region; the radius of the circular hole in the gradual change region satisfies , Duty cycle; the spacing between the mirror area circular holes remains unchanged, and the radius of the mirror area circular holes all satisfy , Duty cycle, the circular hole of the nanobeam cavity waveguide is filled with silicon dioxide.

[0008] Preferably, the width of the curved waveguide is 400-600 nm, and the thickness is 220 nm; the width of the nanobeam cavity waveguide is 400-600 nm, and the thickness is 220 nm, and the coupling distance between the curved waveguide and the nanobeam cavity waveguide is 150-300 nm.

[0009] Preferably, the distance between the electrode and the waveguide layer is 500-1000 nm; the grating vertical coupler is a Bragg grating, the period is 300-1000 nm, the filling factor is 0.5, and the coupling efficiency at a wavelength of 1.5-1.6 um is 1-50%.

[0010] Preferably, the phase change material is GSST, the length is 500 nm, the width is 200 nm, and the thickness is 20 nm.

[0011] Preferably, the phase change material is Sb2Se3, the length is 1000-2000 nm, the width is 150-300 nm, and the thickness is 220 nm.

[0012] Preferably, the electrode is a cuboid structure, the length is 500-2000 nm, the width is 200 nm, and the thickness is 20-50 nm.

[0013] Preferably, the diameter of the circular hole of the nanobeam cavity waveguide is 160-260 nm, and the spacing between the centers of the circular holes is 300-450 nm.

[0014] Preferably, the number of single-sided gradual change area circular holes of the nanobeam cavity waveguide is 6, the number of single-sided mirror area is 10-15, the minimum spacing d1 of the gradual change area is 300 nm, the maximum spacing is 450 nm, the spacing a between the mirror area circular holes is 450 nm, and the diameter of all the circular holes is 0.57 times the spacing.

[0015] Preferably, the heat-conducting layer is indium tin oxide ITO.

[0016] Preferably, the heat-conducting layer is divided into 4-8 segments along the width direction of the silicon waveguide layer, and the interval between adjacent segments is 100-200 nm.

[0017] The working mode of the low phase shift reconfigurable phase change optical switch with ultra-wide free spectral range is as follows: the probe light is coupled into the left side (input end) of the curved waveguide through the grating vertical coupler, and the phase change material is reversibly switched among the crystal state, amorphous state and mixed state by applying different electric pulses. When the phase change material is in the amorphous state, in the TE0 mode, the effective refractive index of the curved waveguide and the nanobeam cavity waveguide in the coupling area meets the Bloch mode phase matching, and has high coupling efficiency, so that most of the light is coupled from the curved waveguide to the nanobeam cavity waveguide, and at this time the switch is in the "off state. When the phase change material is in the crystal state, in the TE mode, the two waveguides are phase mismatched in the coupling area, and have low coupling efficiency, only a small part of the light enters the nanobeam cavity waveguide, and at this time the switch is in the "on" state.

[0018] Compared with the prior art, the advantages of the present application are that: the low phase shift reconfigurable phase change optical switch with ultra-wide free spectral range of the present application combines photonic crystal nanobeam cavity waveguide and phase change material, realizes small resonance wavelength shift before and after phase change switching, and the free spectral range is more than 100 nm. The device is based on the nanobeam cavity waveguide, fully utilizes the small mode volume, enhances the interaction between light and matter in the tiny resonant cavity, uses the bandgap characteristics of one-dimensional photonic crystals, and greatly improves the free spectral range. At the same time, the state of the phase change material is switched by electrode heating, which has the characteristics of low power consumption and multi-level modulation. Through careful design, the resonance wavelength shift is less than 2 nm, which is lower than the working bandwidth of the device, the free spectral range is more than 100 nm, which is much higher than the free spectral range of the traditional micro-ring device, and the present application guarantees the pure operation in the practical spectral range by virtue of the ultra-wide FSR, provides a solution for optoelectronic hybrid AI acceleration, programmable photonic integrated circuit and optical matrix calculation core. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 Fig. 1 is a structural schematic diagram of the low phase shift reconfigurable phase change optical switch covering GSST in embodiment 1, wherein (a) is a perspective view, (b) is a sectional view, and (c) is a top view; Figure 2 Fig. 2 is a band analysis diagram of the one-dimensional photonic crystal with a period of a=350 nm, 400 nm and 450 nm in embodiment 1, (b) is a diagram of the influence of the minimum spacing and duty cycle of the nanobeam cavity waveguide on the resonance wavelength, (c) is a diagram of the influence of the minimum spacing and duty cycle of the nanobeam cavity waveguide on the optical transmittance at the resonance wavelength, and (d) is a diagram of the ultra-wide free spectral range of the side coupling structure under the optimal size d1=300 and η=0.57; Figure 3The physical pattern of the low phase shift reconfigurable phase-change optical switch in Example 1, wherein (a) is an optical microscope image of the low phase shift reconfigurable phase-change optical switch structure, (b) is an SEM image of the device at a magnification of 8000 times, and (c) is an SEM image of the device at a magnification of 40000 times in a local part; Figure 4 In (a), the simulated optical transmission spectrum of the low phase shift reconfigurable phase-change optical switch in Example 1, (b) is the actual test transmission spectrum of the low phase shift reconfigurable phase-change optical switch in the specific embodiment, (c) is the electric field distribution of the optical switch in the on state, and (d) is the electric field distribution of the optical switch in the off state; Figure 5 The structural schematic diagram of the low phase shift reconfigurable phase-change optical switch embedded with Sb2Se3 in Example 2, wherein (a) is a perspective view, (b) is a sectional view, and (c) is a top view; Figure 6 In (a), the transmission spectrum of the low phase shift reconfigurable phase-change optical switch in Example 2 in the TE0 mode, (b) is the electric field distribution of the optical switch in the on and off states, (c) is the transmittance of the optical switch at different operating wavelengths, and (d) is the insertion loss (IL) of the optical switch at different duty cycles; Figure 7 In (a), the electric pulse required for the crystallization process of the phase-change optical switch in Example 2, (b) is the electric pulse required for the amorphization process of the phase-change optical switch, (c) is the device temperature distribution when the phase-change material reaches the crystallization temperature, and (d) is the device temperature distribution when the phase-change material reaches the melting temperature. DETAILED DESCRIPTION

[0020] The application will be further described in detail below with reference to the embodiments of the drawings.

[0021] Example 1: Low phase shift reconfigurable phase-change optical switch based on phase-change material GSST with ultra-wide free spectral range.

[0022] The device is prepared by two-step electron beam-exposure etching and magnetron sputtering process on a Si substrate, that is, the waveguide layer and the grating vertical coupler are first prepared by electron beam-exposure etching, then the phase-change material film is grown at the corresponding position by electron beam exposure and magnetron sputtering, and finally a 2um thick silicon dioxide protective layer is coated on the device.

[0023] Figure 1Figure a shows the three-dimensional structure of the optical switching device in this embodiment. The device includes a substrate layer (silicon dioxide), a waveguide layer (silicon), a thermally conductive layer (indium tin oxide, ITO), a phase change material, and a cladding layer (silicon dioxide, not shown in the figure) enclosing the waveguide and phase change material. The phase change material is GSST, with a length of 500 nm, a width of 200 nm, and a thickness of 20 nm. It is located on top of the coupling region of the curved waveguide, and the circular holes of the nanocavity waveguide are filled with silicon dioxide. The waveguide layer has a side-coupled structure of a curved waveguide and a nanocavity waveguide. The state of the phase change material reversibly switches between crystalline and amorphous states under electrical induction. Figure 1 Figure b is a cross-sectional view of the device of the present invention. Figure 1 C is a top view of the device of the present invention, wherein the nanobeam cavity waveguide is designed as an axisymmetric structure, and its distribution from the center to both sides is divided into a gradient region ( N taper ) and mirror area ( N mirror The number of circular holes in the single-sided gradient zone is 6. Figure 1 (The diagram below is for illustrative purposes only; the number of holes is omitted.) The spacing between the holes satisfies the following conditions: Where d1 is the minimum spacing (optimized to 300nm, referring to the spacing between the centers of the circular holes). The period refers to the spacing between the circular apertures in the mirror area, with a value of 450nm. The number of mirror areas is 10-15. Figure 1 (The diagram in the middle is for illustrative purposes only; the number of holes is omitted.) The spacing between the centers of the holes (period) Keeping the wavelength constant at 450nm, the radius of the circular holes all meet the requirements. (η optimized to 0.57). The probe light is input from one side of the curved waveguide. An electrical pulse induces a phase switch in the phase change material. When the GSST is in the amorphous state, the curved waveguide and the nanocavity waveguide are strongly coupled, and most of the light is reflected back to the input port through the nanocavity waveguide, resulting in reduced optical power at the output port. This state is defined as the off state. When the GSST is in the crystalline state, the curved waveguide and the nanocavity waveguide are weakly coupled, and a small portion of the light is reflected back to the input port through the nanocavity waveguide. This state is defined as the on state. Since only the two on / off states are considered, and intermediate states are not investigated, the ITO covering the top of the phase change material does not require a segmented structure.

[0024] like Figure 2 Figure a shows the photonic band structure analysis diagram of a one-dimensional photonic crystal (nano-beam cavity waveguide), that is, the dispersion relationship between photon frequency and wave vector. Figure 2 The shaded area of ​​'a' represents the light cone. Modes located within the light cone region are extended modes, whose energy radiates into the surrounding environment; while the region outside the light cone is the bandgap, where electromagnetic waves are prohibited from propagating in the photonic crystal within this frequency range. Figure 2Figure 'a' depicts the energy band structure of a one-dimensional photonic crystal with periods a = 350 nm, 400 nm, and 450 nm. It can be observed that for each periodic distribution of the photonic crystal structure, there exists a series of separate conduction bands (frequency bands that allow photon propagation) and band gaps (frequency forbidden zones). Our target frequency is 193.55 THZ (corresponding to a wavelength of 1550 nm). Figure 2 The value 'a' (marked with a dashed line) lies precisely between the first and second conduction bands at a = 450 nm (the middle), representing the center frequency of the bandgap region. This position means that for this period, light of the target frequency is completely confined and cannot propagate, which is crucial for designing photonic crystal bandgap devices. As the period 'a' gradually increases, all conduction bands shift towards lower frequencies, a phenomenon consistent with the scaling theory of photonic crystals: with a constant refractive index configuration, the band structure of a photonic crystal is determined by the ratio of the period to the radius of the circular aperture. As the period increases, the characteristic size of the entire photonic crystal increases, causing a corresponding redshift in its resonant frequency and bandgap position. The nanocrystal waveguide structure is designed with a conical region and a mirror region. The mirror region has a perfectly periodic array of nanopores, making it an ideal one-dimensional photonic crystal with a specific photonic bandgap, theoretically providing very high reflectivity. The gradient region, as an adiabatic transition layer, ensures that the energy is reflected back and forth without radiation and with low loss during the conversion of light from the central cavity mode to the external mirror region mode. This design allows light energy to be highly confined to a very small central region (small mode volume), while the photon reflection time within the cavity is very long (high quality factor). The maximum spacing of the gradient region is set to 450 nm, and the minimum spacing (d1) and duty cycle (η) are used to control the resonant wavelength and cavity loss. As d1 and η change, the resonant wavelength exhibits a certain linear change. The circular aperture array of the nanocrystal waveguide is designed with the smallest spacing and size at the center, gradually increasing towards the sides, according to the formula... , Based on the device's resonant wavelength and extinction ratio as evaluation indicators, the optimal duty cycle (η) and minimum spacing (d1) were selected.

[0025] like Figure 2 As shown in Figure b, the resonant wavelength gradually redshifts (wavelength increases) as d1 increases and blueshifts (wavelength decreases) as η increases. From this, we can find a curve that allows the device to operate around 1550nm. By comparing the changes in d1 and η, the lower the loss within the resonant cavity, the higher the coupling efficiency and the larger the extinction ratio. Figure 2 As shown in Figure c, the optimal coupling efficiency is achieved when d1 = 300 nm and η = 0.57. At this point, the switch is designed to operate at 1550 nm (the later phase change material coating will cause a slight shift in the operating wavelength), with a transmittance of -19.87 dB. Besides switching performance, the switch's free spectral range is also an important indicator, such as...Figure 2 As shown in FIG. 9a, without the nanobeam cavity waveguide side-coupling structure covering the phase change material, a super-wide free spectral range of 100 nm has been achieved, which benefits from the extremely small mode volume of the nanobeam cavity waveguide and the extremely short cavity length of the optical microcavity (FSR is inversely proportional to the cavity length).

[0026] Figure 3 For the optical microscope image of the device of the present application, electron beam lithography (EBL) exposure and ICP-RIE etching waveguide, magnetron sputtering GSST, deposition of silica protective layer, to realize high performance device manufacturing. Figure 3 The microscope image of FIG. 9a, Figure 3 The microscope image of FIG. 9b, and Figure 3 The SEM magnification images of the overall device structure and the coupling region structure under different magnifications (8000 times, 40000 times) are shown in FIG. 9c, wherein the width (Ws) of the phase change material deposited on the device is 197.2 nm and the length (Ls) is 488.8 nm. The results obtained by 3D-FDTD finite element algorithm simulation of the device are shown in Figure 4 As shown in FIG. 9a, the simulated phase change optical switch FSR is greater than 100 nm, and the resonance wavelength shift before and after the phase change is about 1 nm, which is much smaller than the bandwidth. Figure 4 The test results of the device after experimental preparation are shown in FIG. 9b, the laser power is 5 dBm, and the laser input wavelength is 1500-1600 nm. Under this wavelength range, the device has no free spectral range limitation. By means of high temperature annealing, the actual phase state of the phase change material is changed. When the phase change material is in amorphous state, the output power measured by the power meter is-18.1 dB. When the phase change material is in crystalline state, the output power measured by the power meter is-15.3 dB. The resonance wavelength shift under the two phase states is about 0.5 nm, which is consistent with the simulation calculation results. In the actual work of the device, we configure the threshold value-15 dB in the acquisition circuit. When the output power collected is lower than-15 dB, we define the switch state as off. When the output power is higher than-15 dB, we define the switch state as on. The switch ratio of the switch is 3 dB. Figure 4 The microscope image of FIG. 9c, and Figure 4 The electric field distribution of the 3D-FDTD simulation optical switch in on and off states is shown in FIG. 9d.

[0027] Example 2, a low phase shift reconfigurable phase change optical switch with super-wide free spectral range based on phase change material Sb2Se3.

[0028] The same as specific example one, the difference is that the phase change material is selected as Sb2Se3, the length of the phase change material is 1800 nm, the width is 160 nm, and the thickness is 220 nm. The phase change mode adopts simulation and analog to apply electric pulse to induce phase change. Six electrode points are set, respectively adopting 5 V, 295 ns crystallization pulse and 7 V, 136 ns amorphous pulse to induce lengthwise phase change of Sb2Se3.

[0029] Figure 5 Figure a is a three-dimensional structural diagram of the optical switch device in the embodiment, which includes a substrate layer (silicon dioxide), a waveguide layer (silicon), a heat-conducting layer (indium tin oxide ITO), a phase-change material structure, and a cladding layer (silicon dioxide) wrapping the waveguide and the phase-change material. Figure 5 Figure b is a cross-sectional view of the device, Figure 5 Figure c is a top view of the device. The design of the nanobeam cavity waveguide is an axisymmetric structure, and the distribution from the center to both sides is divided into a gradual change region and a mirror region. The number of single-side gradual change region circular holes is 6, the minimum distance (d1) of the gradual change region is 300 nm, and the maximum distance is 450 nm, satisfying the formula . The distance (a) of the mirror region is 450 nm, and the diameter of all the circular holes is 0.57 times the distance (a). wherein η is optimized to be 0.57. The advantage of this design is that when the curved waveguide and the nanobeam cavity waveguide are coupled, the light field is mainly distributed in the central region of the nanobeam cavity waveguide, the circular holes and the distance of the gradual change region gradually increase from the center to both sides, which can make the effective refractive index of the TE0 mode change uniformly, reduce the reflection loss caused by the sudden change of the effective refractive index when the light is reflected back and forth in the resonant cavity, and make the working wavelength of the optical switch controllable, which can be further determined according to the photonic crystal energy band analysis. At the same time, the working wavelength of the optical switch is located in the forbidden band region, has a stronger reflection effect, improves the coupling efficiency when the switch is in the amorphous state, and improves the extinction ratio. The probe light is input from one side of the curved waveguide, and the phase-change material is induced to undergo phase state switching by applying an electric pulse. When Sb2Se3 is in the amorphous state, the curved waveguide and the nanobeam cavity waveguide are strongly coupled, most of the light is reflected back to the input port through the nanobeam cavity waveguide, the output port light power is weakened, and the switch is defined as the off state. When Sb2Se3 is in the crystalline state, the curved waveguide and the nanobeam cavity waveguide are weakly coupled, and a small part of the light is reflected back to the input port through the nanobeam cavity waveguide, and this switch is defined as the on state. Since the switch has a large on-off ratio, we can obtain more intermediate states by segmenting the phase-change material, changing the crystalline length of the phase-change material, and adjusting the coupling efficiency, so as to realize multi-level modulation of the optical switch.

[0030] The transmission spectrum of the switch device is obtained by 3D-FDTD finite element algorithm simulation as Figure 6As shown in Figure a, segmented heating of the electrodes alters the crystal length of the phase change material, essentially changing the coupling length and thus affecting the coupling efficiency. The phase change material states are set as weighted states, with crystalline states mapped to 1 and amorphous states to 0. Light enters from one end of a curved waveguide, with the other end serving as the output port. By changing the phase state of the phase change material, different effective coupling lengths (Leff) are achieved, resulting in different output values. Assuming the input is set to 1, the output power has seven possible values: 0.002, 0.1, 0.2, 0.3, 0.6, 0.9, and 0.98, enabling multi-level modulation of the optical switch. Figure 6 As shown in Figure a, the effective coupling length is modulated between 0 and 1.8 μm. At a working wavelength of 1553 nm, the transmittance of the output light varies from 1 to 0. The conclusion is that the longer the effective coupling length, the higher the coupling efficiency between the curved waveguide and the nano-beam cavity waveguide. This results in more input light entering the nano-beam cavity waveguide and being reflected back to the input waveguide, leading to lower optical power at the output end. This enables multi-level modulation of the optical switch. Figure 6 Figure b shows the electric field distribution in the simulated on and off states of the optical switch. To study the scalability of the optical switch, we simulated devices operating at different wavelengths by changing the duty cycle of the circular aperture in the nanobeam cavity waveguide. Within the probe light range of 1450-1650 nm, the optical switch exhibited no free spectral range limitations, with the widest FSR exceeding 100 nm. The resonant wavelength did not shift before and after the phase change material switching. The on / off ratio and insertion loss are shown below. Figure 6 c in the middle Figure 6 As shown in Figure d, the on / off ratios all exceed 10 dB, and the insertion loss is less than 0.4 dB. The devices operate at different wavelengths without affecting each other, and all can achieve pure operation. This demonstrates that the optical switches we studied can be adjusted according to parameters to obtain optical switches operating at arbitrary wavelengths, and these optical switches exhibit good performance in terms of free spectral range, on / off ratio, and insertion loss.

[0031] Next, we used COMSOL thermal simulation to simulate the heating process of the phase change material and calculate the electrical pulses and switching speed required for the optical switch conversion. For example... Figure 7 As shown in Figure a, during the crystallization process, we employ a multi-electrode independent heating scheme. By applying voltage to different electrode sites, we achieve segmental crystallization of the phase change material from left to right. This process alters the effective coupling length (Leff) of the amorphous region, thereby controlling the coupling efficiency between waveguides. When a 5V, 295ns electrical pulse is applied to unit 1, the temperature in this region exceeds the crystallization temperature (reaching a maximum of 645K, but still below the amorphization temperature), while the adjacent unit 2 remains below the crystallization temperature. The temperature tolerance between the two is 26K, ensuring the local controllability of the crystallization process. Unit 1 and unit 2 are two arbitrary adjacent segments of the heating layer.

[0032] like Figure 7In the amorphization process, the whole bulk phase change material is brought to the melting temperature by applying a 7V, 136ns pulse to all electrodes simultaneously, realizing fast erasing and resetting to a completely amorphous state. The highest temperature in this process is strictly controlled and does not exceed the melting point of other materials in the device (such as silicon, indium tin oxide, silicon dioxide), ensuring the thermal safety of the device.

[0033] Figure 7 In the amorphization process, the whole bulk phase change material is brought to the melting temperature by applying a 7V, 136ns pulse to all electrodes simultaneously, realizing fast erasing and resetting to a completely amorphous state. The highest temperature in this process is strictly controlled and does not exceed the melting point of other materials in the device (such as silicon, indium tin oxide, silicon dioxide), ensuring the thermal safety of the device. Figure 7 In the amorphization process, the whole bulk phase change material is brought to the melting temperature by applying a 7V, 136ns pulse to all electrodes simultaneously, realizing fast erasing and resetting to a completely amorphous state. The highest temperature in this process is strictly controlled and does not exceed the melting point of other materials in the device (such as silicon, indium tin oxide, silicon dioxide), ensuring the thermal safety of the device.

[0034] The above description is not a limitation of the present application, and the present application is not limited to the above examples. As long as the phase change material with refractive index change caused by optical or electrical pulse excitation is combined with the nanobeam cavity to realize the change of optical transmittance, a reconfigurable optical switch without free spectral range can be realized. At the same time, the size of the waveguide can also be changed to adapt to different modes of light source. In addition to the above-mentioned Sb2Se3 and GSST, GST, GSTS, Sb2Se3 can also achieve the same effect. Changes, modifications, additions or substitutions made by those skilled in the art within the essential scope of the present application should also be within the protection scope of the present application.

Claims

1. A low-phase-shift reconfigurable phase-change optical switch with an ultrawide free spectral range, comprising a silicon dioxide substrate layer, characterized in that: A silicon waveguide layer is disposed on the substrate layer, and a thermally conductive layer is disposed on the silicon waveguide layer. Electrodes are symmetrically distributed on both sides of the silicon waveguide layer and on the upper surface of the thermally conductive layer. The structure of the silicon waveguide layer is a side-coupled structure of a curved waveguide and a nano-beam cavity waveguide. The two ends of the curved waveguide are respectively connected to grating vertical couplers. A phase change material is disposed on the upper layer of the coupling region of the curved waveguide and on the lower surface of the thermally conductive layer. The phase change material can reversibly switch between crystalline and amorphous states under electrical induction. The phase change material is connected to the two electrodes through the thermally conductive layer to form an ohmic contact. The circular aperture array on the nanobeam cavity waveguide is designed as an axisymmetric structure, distributed from the center to both sides, divided into a gradient region and a mirror region. The circular apertures in the gradient region are designed to be the smallest at the center and gradually increase in size on both sides. The spacing between the circular apertures in the gradient region on one side satisfies... Where d1 is the minimum spacing between the circular holes. Let n be the spacing between the circular holes in the mirror area, and n be the number of circular holes in the gradient area on one side; the radii of the circular holes in the gradient area all satisfy... , The duty cycle is constant; the spacing between the circular holes in the mirror area remains unchanged, and the radii of the circular holes in the mirror area all satisfy... , For the duty cycle, the circular holes of the nanobeam cavity waveguide are filled with silicon dioxide.

2. The low-phase-shift reconfigurable phase-change optical switch with an ultrawide free spectral range according to claim 1, characterized in that: The curved waveguide has a width of 400nm-600nm and a thickness of 220nm; the nanocrystal waveguide has a width of 400nm-600nm and a thickness of 220nm; and the coupling distance between the curved waveguide and the nanocrystal waveguide is 150nm-300nm.

3. The low-phase-shift reconfigurable phase-change optical switch with an ultrawide free spectral range according to claim 1, characterized in that: The distance between the electrode and the waveguide layer is 500-1000nm; the grating vertical coupler is a Bragg grating with a period of 300nm-1000nm, a fill factor of 0.5, and a coupling efficiency of 1-50% at a wavelength of 1.5-1.6um.

4. A low-phase-shift reconfigurable phase-change optical switch with an ultrawide free spectral range according to any one of claims 1-3, characterized in that: The phase change material is GSST, which has a length of 500 nm, a width of 200 nm, and a thickness of 20 nm.

5. A low-phase-shift reconfigurable phase-change optical switch with an ultrawide free spectral range according to any one of claims 1-3, characterized in that: The phase change material is Sb2Se3, with a length of 1000nm-2000nm, a width of 150nm-300nm, and a thickness of 220nm.

6. The low-phase-shift reconfigurable phase-change optical switch with an ultrawide free spectral range according to claim 5, characterized in that: The electrode has a cuboid structure with a length of 500nm-2000nm, a width of 200nm, and a thickness of 20nm-50nm.

7. The low-phase-shift reconfigurable phase-change optical switch with an ultrawide free spectral range according to claim 5, characterized in that: The diameter of the circular holes in the nanobeam cavity waveguide is 160-260nm, and the spacing between the centers of the circular holes is 300nm-450nm.

8. The low-phase-shift reconfigurable phase-change optical switch with an ultrawide free spectral range according to claim 5, characterized in that: The nanobeam cavity waveguide has 6 circular holes in the single-sided gradient region and 10-15 circular holes in the single-sided mirror region. The minimum spacing d1 of the gradient region is 300nm and the maximum spacing is 450nm. The spacing a between the circular holes in the mirror region is 450nm. The diameter of all the circular holes is 0.57 times the spacing.

9. The low-phase-shift reconfigurable phase-change optical switch with an ultrawide free spectral range according to claim 8, characterized in that: The thermally conductive layer is indium tin oxide (ITO).

10. The low-phase-shift reconfigurable phase-change optical switch with an ultrawide free spectral range according to claim 8, characterized in that: The thermally conductive layer is divided into 4-8 segments along the width direction of the silicon waveguide layer, with an interval of 100-200nm between adjacent segments.

Images