Memsm optical switch based on movable waveguide mirror and nxn optical switch array device

Abstract

The application discloses a MEMS optical switch based on a movable waveguide mirror and an N*N optical switch array device. The device comprises a substrate, an optical structure and a mechanical structure arranged on the substrate. The optical structure comprises a fixed waveguide mirror and a movable waveguide mirror. The fixed waveguide mirror is fixed on the substrate, and the movable waveguide mirror is suspended on the substrate. The fixed waveguide mirror and the movable waveguide mirror are connected through a pair of half-folded springs. The fixed waveguide mirror and the movable waveguide mirror are both provided with a sub-wavelength reflecting surface. The application realizes the effect of optical path switching, has the advantages of a large bandwidth range, low insertion loss, low crosstalk, high extinction ratio, simple manufacturing process, low processing cost, low power consumption, strong scalability and the like.

Description

Technical Field

[0001] This invention relates to a MEMS optical switch and optical array device in the field of integrated optoelectronic devices. Specifically, it relates to an integrated waveguide MEMS optical switch and its array device that achieves optical path adjustment by driving the spacing between a pair of waveguide mirrors with subwavelength structure reflective surfaces through an electrostatic comb. Background Technology

[0002] In recent years, the emergence of cloud computing and data-intensive computing has spurred a rapid expansion in the scale of data center networks. The explosive growth in data volume has placed further demands on low-latency, low-power, and reconfigurable networks. Compared with traditional electrical interconnect technologies, optical interconnect technologies have significant advantages in terms of latency and power consumption, making optical interconnects crucial in the construction of large-scale data centers.

[0003] Large-scale optical switch arrays, as core components of reconfigurable optical networks, greatly enhance the flexibility of optical networks. Among various optical switch structures, MEMS-based free-space optical switches have attracted widespread attention due to their low loss, high reliability, and strong scalability. Currently, most optical switches in data center networks are free-space MEMS optical switches. For free-space MEMS optical switches, the optical signal propagates and switches in free space after entering the optical fiber and is then received by the output optical fiber. Its scalability is excellent, and it is currently possible to achieve hundreds of ports while maintaining low insertion loss and low crosstalk. However, free-space MEMS optical switches have relatively long switching times due to the inability of free-space MEMS to achieve higher mechanical action speeds; their switching speeds are generally in the range of a few milliseconds to tens of milliseconds, which significantly limits the reconfiguration speed of large-scale data centers.

[0004] Integrated waveguide MEMS optical switches can solve this problem. The integrated waveguide MEMS structure allows for a smaller mechanical structure size, thereby improving the switching speed. In an integrated waveguide MEMS optical switch, the optical signal is input to the chip from the input fiber and then confined within the integrated waveguide for propagation. The switching of the optical path is achieved through the integrated waveguide structure and mechanical structure.

[0005] Currently, integrated waveguide MEMS optical switches mainly fall into two categories: those based on movable directional couplers and those based on vertical thermally adiabatic couplers. Both achieve low switching losses, fast switching speeds, and large switch array sizes. However, they also have their own drawbacks, limiting further expansion of their switch array size. Specifically, integrated waveguide MEMS optical switches based on movable directional couplers suffer from drawbacks such as high losses, small bandwidth, sensitivity to fabrication errors, and the need for correction of the drive voltage of the switching units due to the limitations of the directional coupler. Integrated waveguide MEMS optical switches based on vertical thermally adiabatic couplers replace the directional coupler with a novel double-layer vertical thermally adiabatic coupler. This results in low switching losses, large bandwidth, and large fabrication tolerance. However, the fabrication process of this double-layer thermally adiabatic coupler is very complex due to the need for two waveguide layers, which is not conducive to cost reduction and scalability. Summary of the Invention

[0006] In view of the above background technology, the purpose of this invention is to provide an integrated waveguide MEMS optical switch and an N×N optical switch array device driven by an electrostatic comb based on a movable subwavelength reflective waveguide mirror. This switch applies a driving voltage to a fixed electrostatic comb, causing it to attract an attraction between itself and a grounded movable electrostatic comb, thus moving the movable electrostatic comb. Furthermore, a transmission rod drives the movable waveguide mirror closer to the fixed waveguide mirror, and the optical path is switched by the coupling of optical signals between the two waveguide mirrors.

[0007] The technical solution adopted in this invention is:

[0008] I. A MEMS optical switch based on a movable waveguide mirror:

[0009] It includes a substrate and optical and mechanical structures placed on the substrate; wherein the optical structure includes a fixed waveguide mirror and a movable waveguide mirror, the fixed waveguide mirror is fixed to the substrate and the movable waveguide mirror is suspended from the substrate; the fixed waveguide mirror and the movable waveguide mirror are connected by a pair of semi-folding springs, and both the fixed waveguide mirror and the movable waveguide mirror are provided with subwavelength reflecting surfaces.

[0010] The mechanical structure includes a transmission rod, two sets of fixed islands, a folding spring, a fixed electrostatic comb, and a movable electrostatic comb. The fixed islands and the fixed electrostatic comb are fixed to the base, while the folding spring, transmission rod, and movable electrostatic comb are suspended from the base. The movable waveguide reflector is fixedly connected to the movable electrostatic comb via the transmission rod, which is connected via the folding spring and the two sets of fixed islands. The comb teeth of the fixed electrostatic comb and the movable electrostatic comb are arranged opposite to each other and are staggered.

[0011] A pair of semi-folded springs are symmetrically distributed on both sides of the fixed waveguide mirror and the movable waveguide mirror, and each side of the fixed waveguide mirror and the movable waveguide mirror is connected by a semi-folded spring.

[0012] The semi-folding spring mainly consists of a spring handle, two beams connected to both ends of the spring handle and perpendicular to the length direction of the spring handle, and a mechanical stopping structure located on one side of the beam away from the spring handle. The ends of the two beams near the spring handle are fixedly connected to both ends of the spring handle, and the ends of the two beams away from the spring handle are connected to the edges of the subwavelength reflecting surfaces of the fixed waveguide reflector and the movable waveguide reflector, respectively.

[0013] The aforementioned mechanical stopping structures are all located on beams connected to the movable waveguide mirror. The mechanical stopping structure is a protrusion on the beam, which contacts the subwavelength reflecting surface of the fixed waveguide mirror when the movable waveguide mirror approaches the fixed waveguide mirror.

[0014] The fixed waveguide reflector and the movable waveguide reflector have the same structure. Both consist of a pair of orthogonal multimode wide waveguides, an adiabatic tapered waveguide connected to the end of the multimode wide waveguide, and a single-mode waveguide connected to the end of the adiabatic tapered waveguide.

[0015] Both the fixed waveguide reflector and the movable waveguide reflector are provided with subwavelength reflecting surfaces. The subwavelength reflecting surfaces are located on the end faces of the intersection of a pair of orthogonal multimode wide waveguides. The subwavelength reflecting surfaces of the fixed waveguide reflector and the movable waveguide reflector are arranged facing each other, and the subwavelength reflecting surfaces are specifically wave-tooth structures that satisfy the subwavelength conditions.

[0016] When the movable waveguide mirror approaches the fixed waveguide mirror, the subwavelength reflecting surfaces of the fixed waveguide mirror and the movable waveguide mirror engage with each other.

[0017] II. An N×N silicon-based MEMS optical switch array:

[0018] The N×N optical switch array includes at least four cascaded integrated waveguide MEMS optical switches.

[0019] The fixed waveguide reflector, movable waveguide reflector, semi-folding spring, transmission rod, two sets of fixed island spring structures and electrostatic comb driver are made of the same material and have the same thickness. The entire optical switch structure or N×N array can be fabricated using monolithic integrated processing.

[0020] The N×N optical switch array should include at least four integrated waveguide MEMS optical switches as described above. Adjacent optical switches are connected by single-mode waveguides and waveguide cross-connections. The connection method between optical switches can adopt, but is not limited to, Benes, Cross-Bar and other topologies.

[0021] The electrostatic comb driver, during operation, controls the displacement of the movable waveguide mirror by adjusting the voltage applied by the external circuit, thereby regulating the distance between the subwavelength reflecting surfaces of the fixed and movable waveguide mirrors. As the voltage increases, the movable waveguide mirror gradually approaches the fixed waveguide mirror, and the wavy toothed structures on the subwavelength reflecting surfaces of both mirrors gradually engage. The ends of the two beams of the semi-folded spring connected to the subwavelength reflecting surfaces move closer together. When the voltage reaches a certain level, the mechanical stopping structure on one side of the movable waveguide mirror contacts the beam on the other side of the fixed waveguide mirror. At this point, the distance between the two waveguide mirrors reaches its minimum, determined by the height of the mechanical stopping structure.

[0022] Using the input port of the fixed waveguide mirror as the input port of the switch, when no voltage is applied to the fixed electrostatic comb by the external circuit, the switch structure is in the OFF state. At this time, the input optical signal is reflected by the subwavelength reflecting surface of the fixed waveguide mirror, and the optical signal is output from the through port with low loss. Since the movable waveguide mirror is far away, the optical signal is almost not coupled into the movable waveguide mirror, and the optical signal strength output from the cross port is very low. When the external circuit applies a sufficiently large voltage to the fixed electrostatic comb to minimize the gap between the two subwavelength reflecting surfaces, the switch structure is in the ON state. The subwavelength reflecting surfaces of the fixed and movable waveguide mirrors interlock, with only a small air gap between them. The two waveguide mirrors approximately form a multimode interference cross waveguide, and almost all the optical signal is coupled into the movable waveguide mirror, outputting from the cross port with low loss, thus achieving the function of switching the optical propagation path. At this time, very little optical signal is reflected by the fixed mirror, and the optical signal strength output from the through port is very low. When the switch is in the ON state, if the voltage applied to the fixed electrostatic comb by the external circuit is removed, making the potential difference between the fixed and movable electrostatic combs zero, the driving force provided by the electrostatic comb driver will decrease to zero. The switch structure will then return to the OFF state under the elastic force of the semi-folding and folding springs. This constitutes a complete switching operation.

[0023] Existing technologies typically address the issue that the facing end faces of fixed and movable waveguide mirrors are planar. In this case, if the two mirrors are brought into contact (the ON state of the switch), they will adhere together due to van der Waals forces and cannot be separated. Even after removing the voltage applied to the fixed electrostatic comb by the external circuit, the switch structure cannot return to the OFF state under the spring force. Furthermore, the tilting of the waveguide sidewalls caused by experimental fabrication in this adhered state results in a wedge-shaped air gap between the end faces of the two mirrors, leading to low optical coupling. If the facing end faces of the fixed and movable waveguide mirrors are not in contact, with a small air gap between them (the ON state of the switch), the adhesion due to van der Waals forces can be avoided. However, the wedge-shaped air gap between the end faces is wider, resulting in even lower optical coupling. A solution with planar end faces for both waveguide mirrors is difficult to achieve experimentally.

[0024] This invention features a subwavelength reflective surface on the end face. The fixed and movable waveguide mirrors face each other without contact, with a small air gap between them serving as the ON state of the switch. This avoids the two waveguide mirrors sticking together due to van der Waals forces and significantly enhances optical coupling. The subwavelength reflective surface structure is designed with the following considerations in mind: First, the structure introduced on the reflective surface must not affect the low-loss reflection of light in the waveguide mirror; therefore, the structure must meet the subwavelength condition. Second, the structural shapes on the reflective surfaces of the two waveguide mirrors need to be complementary so that they can mesh tightly without contact when the switch is in the ON state, leaving a small air gap. Considering the limitations of manufacturing technology, a wavy, toothed structure is ultimately introduced on the reflective surface of the waveguide mirror as the subwavelength reflective surface. Compared to other reflective surfaces, this enhances light coupling, allowing almost all light to be coupled, thus improving light transmission.

[0025] The beneficial effects of this invention are:

[0026] (1) The present invention uses only one type of material, has a simple structure, and has low manufacturing cost;

[0027] (2) Optical switches have low insertion loss, low crosstalk and high extinction ratio over a wide wavelength range;

[0028] (3) The optical switch adopts a capacitor-type driving method, and the optical switch has extremely low power consumption;

[0029] (4) The optical switch has two input ports and two output ports. This 2×2 optical switch can be cascaded into a large-scale optical switch array using various topologies, and has strong scalability.

[0030] In summary, the switching structure in this invention achieves optical path switching by driving a movable waveguide mirror close to a fixed waveguide mirror via an electrostatic comb. It has significant advantages such as a large bandwidth range, low insertion loss, low crosstalk, high extinction ratio, simple manufacturing process, low processing cost, low power consumption, and strong scalability. Attached Figure Description

[0031] Figure 1 This is a top view of the structure of the present invention in its initial state (OFF);

[0032] Figure 2 yes Figure 1 Cross-sectional view of line A-A';

[0033] Figure 3 yes Figure 1 Cross-sectional view of B-B';

[0034] Figure 4 This is an enlarged view of the subwavelength reflective surface structure;

[0035] Figure 5 This is a top view of the structure of the present invention after voltage is applied (ON);

[0036] Figure 6 yes Figure 5 Magnified view of the middle waveguide reflector;

[0037] Figure 7 This is a schematic diagram of an N×N integrated waveguide MEMS optical switch structure based on the Benes topology.

[0038] Figure 8 These are light field transmission diagrams of the optical structure of the present invention in OFF and ON states, respectively;

[0039] In the diagram: 1. Fixed waveguide reflector, 2. Subwavelength reflector, 3. Semi-folded spring, 4. Stop structure, 5. Movable waveguide reflector, 6. Fixed island, 7. Folded spring, 8. Drive rod, 9. Fixed electrostatic comb, 10. Movable electrostatic comb, 11. Substrate, 12. Buried layer, pz, spacing of subwavelength structures, hz, height of subwavelength structures, hs, height of mechanical stop structure. Detailed Implementation

[0040] The present invention will be further described below with reference to the accompanying drawings and embodiments, but this should not be construed as limiting the scope of protection of the present invention.

[0041] like Figure 1 and Figure 5As shown, the present invention mainly includes a substrate 11 and an optical structure and a mechanical structure placed on the substrate 11; wherein the optical structure includes a fixed waveguide mirror 1 and a movable waveguide mirror 5, the fixed waveguide mirror 1 is fixed to the substrate 11 and the movable waveguide mirror 5 is suspended from the substrate 11; the fixed waveguide mirror 1 and the movable waveguide mirror 5 are connected by a pair of semi-folding springs 3, and both the fixed waveguide mirror 1 and the movable waveguide mirror 5 are provided with a subwavelength reflecting surface 2 for enhancing optical coupling.

[0042] The mechanical structure includes a transmission rod 8, two sets of fixed islands 6, a folding spring 7, a fixed electrostatic comb 9, and a movable electrostatic comb 10. The fixed islands 6 and the fixed electrostatic comb 9 are fixed to the base 11, while the folding spring 7, transmission rod 8, and movable electrostatic comb 10 are suspended from the base 11. The movable waveguide reflector 5 is fixedly connected to the movable electrostatic comb 10 via the transmission rod 8. The transmission rod 8 is connected to the two sets of fixed islands 6 via the folding spring 7. The comb teeth of the fixed electrostatic comb 9 and the movable electrostatic comb 10 are arranged opposite each other and are staggered. The movable electrostatic comb 10 drives the movable waveguide reflector 5 to move closer to or away from the fixed waveguide reflector 1.

[0043] like Figure 1 As shown, the transmission rod 8 is a long strip structure. One end of the transmission rod 8 is connected to the intersection of a pair of multimode wide waveguides of the movable waveguide reflector 5, and its length direction is perpendicular to the subwavelength reflecting surface 2. The transmission rod 8 is narrower where it connects to the movable waveguide reflector 5, which reduces the influence of the transmission rod 8 on the optical signal transmission in the movable waveguide reflector 5. The transmission rod 8 is wider away from the movable waveguide reflector 5, which enhances the mechanical stability of the transmission rod 8. Starting from the end connected to the movable waveguide reflector 6, the transmission rod 8 is sequentially connected to the first set of fixed island spring structures, the electrostatic comb driver, and the second set of fixed island spring structures. The first set of fixed island spring structures is connected to the end of the transmission rod closest to the movable waveguide reflector, and the second set of fixed island spring structures is connected to the other end of the transmission rod.

[0044] like Figure 1 As shown, each set of fixed island spring structures consists of a pair of folding springs 7 and four fixed islands 6 symmetrically distributed on both sides of the transmission rod 8. The folding springs 7 are suspended structures, consisting of a spring handle, two short beams connected to both ends of the spring handle and perpendicular to its length, and two long beams connected to the middle of the spring handle and perpendicular to its length. The ends of the two short beams furthest from the spring handle are connected to a fixed island 6, and the ends of the two long beams furthest from the spring handle are connected to the transmission rod 8. The length direction of the spring handle is parallel to the length direction of the transmission rod 8. The four beams—two long beams and two short beams—are parallel and equidistantly distributed, with their length directions perpendicular to the length direction of the transmission rod 8. The two sets of fixed island spring structures are distributed along the length direction of the transmission rod on both sides of the electrostatic comb driver.

[0045] like Figure 2 As shown, the fixed island 6 has a square structure and is connected to the base 11 via the buried layer 12. The folding spring 7 is a suspended structure, and the fixed island 6 supports the folding spring 7. Each folding spring requires support from two fixed islands distributed on the same side of the transmission rod. Therefore, each set of fixed island spring structures contains four fixed islands, which are distributed in pairs on both sides of the transmission rod.

[0046] like Figure 1 As shown, the electrostatic comb driver includes a pair of fixed electrostatic combs 9 and a pair of movable electrostatic combs 10. The fixed electrostatic combs 9 and movable electrostatic combs 10 have similar structures, both consisting of a comb handle and periodically distributed comb teeth on the handle. The length direction of the comb handle is perpendicular to the drive rod 8, and the length direction of the comb teeth is parallel to the drive rod 8. The comb teeth are periodically arranged along the length direction of the comb handle, with one end connected to the comb handle and the other end unconnected. The comb teeth of the fixed electrostatic combs 9 and movable electrostatic combs 10 are arranged in an alternating, opposing pattern.

[0047] like Figure 3 As shown, the fixed electrostatic comb 9 is supported by a buried layer 12 below the comb handle, and is connected to the base 11 through the buried layer 12. A pair of fixed electrostatic combs 9 are symmetrically distributed on both sides of the transmission rod 8 along its length, located near the first set of fixed island spring structures. The side of the fixed electrostatic comb 9 closest to the transmission rod is not connected to the transmission rod, while the side furthest from the transmission rod is connected to the external circuit. The comb teeth of the fixed electrostatic comb 9 are distributed on the side furthest from the first set of fixed island spring structures.

[0048] like Figure 1 As shown, a pair of movable electrostatic combs 10 are symmetrically distributed on both sides of the transmission rod 8 along its length, located near the second set of fixed island spring structures. The side of the movable electrostatic comb 10 closest to the transmission rod 8 is connected to the transmission rod, while the side furthest from the transmission rod 8 is not connected. The comb teeth of the movable electrostatic comb 10 are distributed on the side furthest from the second set of fixed island spring structures.

[0049] like Figure 5As shown, when the electrostatic comb driver is working, it applies a voltage to the fixed electrostatic comb 9 through an external circuit, creating a potential difference between it and the movable electrostatic comb 10. This potential difference generates an attractive force between the fixed electrostatic comb 9 and the movable electrostatic comb 10. Since the fixed electrostatic comb 9 is connected to the substrate 11 through the buried layer material 12 and is immovable, while the movable electrostatic comb 10 is a suspended and movable structure, the movable electrostatic comb 10 will move closer to the fixed electrostatic comb 9. This movement drives the transmission rod 8 to displace parallel to its length direction through the connection between the movable electrostatic comb 10 and the transmission rod 8. Furthermore, the connection between the end of the transmission rod 8 and the movable waveguide reflector 5 drives the movable waveguide reflector 5 to displace parallel to the length direction of the transmission rod 8. At the same time, the semi-folding spring 3 connected to the fixed waveguide reflector 1 and the movable waveguide reflector 5, as well as the two sets of spring structures connected to the transmission rod 8, all undergo a certain degree of elastic deformation.

[0050] like Figure 1 As shown, a pair of semi-folded springs 3 are symmetrically distributed on both sides of the fixed waveguide mirror 1 and the movable waveguide mirror 5, and each side of the fixed waveguide mirror 1 and the movable waveguide mirror 5 is connected by a semi-folded spring 3.

[0051] The semi-folding spring 3 is a suspended structure, mainly composed of a spring handle, two beams connected to both ends of the spring handle and perpendicular to the length direction of the spring handle, and a mechanical stopping structure 4 located on one side of the beam away from the spring handle. The ends of the two beams near the spring handle are fixedly connected to both ends of the spring handle, and the ends of the two beams away from the spring handle are connected to the edges of the subwavelength reflecting surfaces 2 of the fixed waveguide reflector 1 and the movable waveguide reflector 5, respectively. The spring handle, beams, and mechanical stopping structure 4 are all suspended.

[0052] like Figure 6 As shown, the mechanical stopping structure 4 of each semi-folding spring 3 is located on the beam connected to the movable waveguide reflector 5. The mechanical stopping structure 4 is a protrusion on the beam. When the movable waveguide reflector 5 approaches the fixed waveguide reflector 1, the protrusion contacts the subwavelength reflecting surface 2 of the fixed waveguide reflector 1. Its function is to prevent the two waveguide reflecting surfaces from being tightly attached when the movable waveguide reflector 5 approaches the fixed waveguide reflector 1, thus preventing the two waveguide reflectors from being stuck together due to intermolecular forces and unable to separate.

[0053] The fixed waveguide reflector 1 and the movable waveguide reflector 5 have the same structure. Both consist of a pair of orthogonal multimode wide waveguides, an adiabatic tapered waveguide connected to the end of the multimode wide waveguide, and a single-mode waveguide connected to the end of the adiabatic tapered waveguide.

[0054] Both the fixed waveguide reflector 1 and the movable waveguide reflector 5 are suspended structures and are distributed relative to each other. The fixed waveguide reflector 1 is fixed in place, while the movable waveguide reflector 5 can move closer to or away from the fixed waveguide reflector 1 under the action of the mechanical structure.

[0055] Both fixed and movable waveguide mirrors have single-mode waveguides as input and output ports. Adjacent optical switches can be interconnected via single-mode waveguides and waveguide crossovers to form an N×N switch array. Each fixed and movable waveguide mirror has one input port and one output port. If the input port of the fixed waveguide mirror is used as the input port of the switch, then the output port of the fixed waveguide mirror is a straight-through port, and the output port of the movable waveguide mirror is a crossover port. Conversely, if the input port of the movable waveguide mirror is used as the input port of the switch, then the output port of the fixed waveguide mirror is a crossover port, and the output port of the movable waveguide mirror is a straight-through port.

[0056] Both the fixed waveguide mirror 1 and the movable waveguide mirror 5 are provided with a subwavelength reflecting surface 2. The subwavelength reflecting surface 2 is located on the end face of the original plane at the intersection of a pair of orthogonal multimode wide waveguides. The subwavelength reflecting surfaces 2 of the fixed waveguide mirror 1 and the movable waveguide mirror 5 are arranged facing each other.

[0057] like Figure 4 As shown, the cross-section of the subwavelength reflecting surface 2 is a periodic wave-tooth structure that satisfies the subwavelength condition, with a periodic tooth spacing of pz and a tooth height of hz.

[0058] like Figure 6 As shown, the wavy toothed structures of the subwavelength reflecting surfaces 2 of the two waveguide mirrors are staggered. When the movable waveguide mirror 5 approaches the fixed waveguide mirror 1, the subwavelength reflecting surfaces 2 of the fixed waveguide mirror 1 and the movable waveguide mirror 5 interlock with each other.

[0059] The function of the subwavelength reflector is to enhance the coupling of optical signals between the two waveguide reflectors when they are close to each other and the wavy toothed structures of the two waveguide reflectors mesh.

[0060] With subwavelength reflecting surfaces 2 provided on the fixed waveguide mirror 1 and the movable waveguide mirror 5, and connected by a pair of semi-folding springs 3, not only can an elastic connection be provided between the fixed waveguide mirror 1 and the movable waveguide mirror 5, but the subwavelength reflecting surfaces 2 of the fixed waveguide mirror 1 and the movable waveguide mirror 5 can also be better aligned.

[0061] When the electrostatic comb driver is working, the displacement of the movable waveguide mirror 5 can be controlled by adjusting the voltage applied by the external circuit, that is, adjusting the distance between the subwavelength reflecting surface 2 of the fixed waveguide mirror 1 and the subwavelength reflecting surface 2 of the movable waveguide mirror 5. As the voltage increases, the movable waveguide mirror 5 gradually moves closer to the fixed waveguide mirror 1, and the wavy tooth-shaped structures on the subwavelength reflecting surfaces 2 of the two waveguide mirrors gradually mesh. The ends of the two beams of the semi-folded spring 3 connected to the subwavelength reflecting surfaces 2 also move closer to each other.

[0062] like Figure 5 and Figure 6 As shown, when the voltage increases to a certain level, the mechanical stop structure 4 located on one side of the movable waveguide reflector 5 will contact the beam located on one side of the fixed waveguide reflector 1. At this time, the distance between the two waveguide reflectors reaches its minimum, and the elastic deformation of the semi-folding spring 3 and the folding spring 7 also reaches its maximum, and the switch is switched to the ON state.

[0063] In the ON state, the minimum distance between the two waveguide mirrors is determined by the height hs of the mechanical stop structure 4. In the ON state, if the voltage applied to the fixed electrostatic comb 9 by the external circuit is removed, there will be no potential difference between the fixed electrostatic comb 9 and the movable electrostatic comb 10. The electrostatic comb driver can no longer provide driving force, and the elastic potential energy stored in the semi-folding spring 3 and the folding spring 7 will be released. The movable waveguide mirror 5 will return to the OFF position under the action of the elastic force of the semi-folding spring 3 and the folding spring 7. At this point, the switch has completed a complete operation process.

[0064] like Figure 1 As shown, the MEMS optical switch in this invention has two input ports and two output ports, making it a 2×2 optical switch, which is more scalable than a 1×2 optical switch structure. 1×2 optical switches are typically cascaded in a cross-bar topology to form a large-scale switch array; an N×N array formed using this topology requires as many as N optical switches. 2 One, while 2×2 optical switches offer greater flexibility in the selection of topology for large-scale arrays; they can also be cascaded using a Benes topology, such as Figure 7 As shown, forming an N×N switch array requires only N(log2N-0.5) optical switches, which significantly reduces the number of optical switches needed for large-scale switch arrays.

[0065] The following describes specific implementations of the present invention:

[0066] Silicon-on-insulator (SOI) was chosen as the implementation platform, with a 220nm thick silicon top layer, a 2µm thick silicon dioxide buried layer, and a silicon substrate. The considered wavelength range is 1425nm to 1700nm, using TE-polarized optical signals. The multimode wide waveguide of the waveguide reflector has a width of 2.68µm and a length of 8.43µm, while the thermally adiabatic tapered waveguide has a length of 6µm. The end connected to the multimode wide waveguide has the same width as the multimode wide waveguide, and the end connected to the single-mode waveguide has the same width as the single-mode waveguide, 0.4µm. For ease of design and fabrication, the subwavelength reflecting surface in this embodiment uses a periodic wavy tooth structure with a tooth height hz of 300nm and a period of 290nm. The height hs of the mechanical stopping structure is 15nm. The distance between the fixed and movable waveguide reflectors in the OFF state is 1µm.

[0067] The optical performance of the device was simulated and verified using the three-dimensional finite-difference time-domain (3D-FDTD) method. In the OFF state, almost all the incident light field was reflected by the waveguide mirror, such as... Figure 8 As shown in the left figure, low loss and high extinction ratio can be achieved in the wavelength range of 1425nm to 1700nm, with losses between 0.12dB and 0.49dB and extinction ratios between 33.6dB and 44.1dB. When a voltage is applied to the fixed electrostatic comb, the movable electrostatic comb pushes the movable waveguide mirror towards the fixed waveguide mirror via a transmission rod. When the mechanical stopping structure contacts the beam on its opposite side, the switch enters the ON state. The wavy toothed structures on the subwavelength reflecting surfaces of the two waveguide mirrors interlock, and the two waveguide mirrors approximate a multimode interference cross waveguide, so the optical signal is no longer reflected. Figure 8 As shown in the right figure, when the switch is in the ON state, it achieves low loss and high extinction ratio in the 1425nm to 1700nm wavelength band, with losses ranging from 0.1dB to 0.47dB and extinction ratios ranging from 32.5dB to 34.1dB. Therefore, the MEMS optical switch proposed in this invention can achieve ultra-wide bandwidth, ultra-low loss, high extinction ratio, and ultra-low power consumption.

[0068] The above embodiments are used to explain and illustrate the present invention, but not to limit the present invention. Any modifications and changes made to the present invention within the spirit and scope of the claims shall fall within the protection scope of the present invention.

Claims

1. A MEMS optical switch based on a movable waveguide mirror, characterized in that: It includes a substrate (11) and optical and mechanical structures placed on the substrate (11); wherein the optical structure includes a fixed waveguide mirror (1) and a movable waveguide mirror (5), the fixed waveguide mirror (1) is fixed to the substrate (11) and the movable waveguide mirror (5) is suspended above the substrate (11); the fixed waveguide mirror (1) and the movable waveguide mirror (5) are connected by a pair of semi-folding springs (3), and both the fixed waveguide mirror (1) and the movable waveguide mirror (5) are provided with subwavelength reflecting surfaces (2). Both the fixed waveguide reflector (1) and the movable waveguide reflector (5) are provided with subwavelength reflecting surfaces (2). The subwavelength reflecting surfaces (2) are provided on the end faces at the intersection of a pair of orthogonal multimode wide waveguides. The subwavelength reflecting surfaces (2) of the fixed waveguide reflector (1) and the movable waveguide reflector (5) are arranged opposite each other, and the subwavelength reflecting surfaces (2) are specifically wave-tooth structures that meet the subwavelength conditions. When the movable waveguide mirror (5) approaches the fixed waveguide mirror (1), the subwavelength reflecting surfaces (2) of the fixed waveguide mirror (1) and the movable waveguide mirror (5) engage with each other.

2. A MEMS optical switch based on a movable waveguide mirror according to claim 1, characterized in that: The mechanical structure includes a transmission rod (8), two sets of fixed islands (6), a folding spring (7), a fixed electrostatic comb (9), and a movable electrostatic comb (10). The fixed islands (6) and the fixed electrostatic comb (9) are fixed to the base (11), while the folding spring (7), the transmission rod (8), and the movable electrostatic comb (10) are suspended above the base (11). The movable waveguide mirror (5) is fixedly connected to the movable electrostatic comb (10) via the transmission rod (8). The transmission rod (8) is connected via the folding spring (7) and the two sets of fixed islands (6). The comb teeth of the fixed electrostatic comb (9) and the movable electrostatic comb (10) are arranged opposite to each other and are staggered.

3. A MEMS optical switch based on a movable waveguide mirror according to claim 1, characterized in that: A pair of semi-folding springs (3) are symmetrically distributed on both sides of the fixed waveguide mirror (1) and the movable waveguide mirror (5), and each side of the fixed waveguide mirror (1) and the movable waveguide mirror (5) is connected by a semi-folding spring (3).

4. A MEMS optical switch based on a movable waveguide mirror according to claim 1 or 3, characterized in that: The semi-folding spring (3) mainly consists of a spring handle, two beams connected to both ends of the spring handle and perpendicular to the length direction of the spring handle, and a mechanical stopping structure (4) located on one side of the beam away from the spring handle. The ends of the two beams near the spring handle are fixedly connected to both ends of the spring handle, and the ends of the two beams away from the spring handle are connected to the edges of the subwavelength reflecting surfaces (2) of the fixed waveguide reflector (1) and the movable waveguide reflector (5).

5. A MEMS optical switch based on a movable waveguide mirror according to claim 4, characterized in that: The mechanical stopping structure (4) is located on the beam connected to the movable waveguide mirror (5). The mechanical stopping structure (4) is a protrusion on the beam. When the movable waveguide mirror (5) approaches the fixed waveguide mirror (1), the protrusion contacts the subwavelength reflecting surface (2) of the fixed waveguide mirror (1).

6. A MEMS optical switch based on a movable waveguide mirror according to claim 1, characterized in that: The fixed waveguide reflector (1) and the movable waveguide reflector (5) have the same structure. Both consist of a pair of orthogonal multimode wide waveguides, an adiabatic tapered waveguide connected to the end of the multimode wide waveguide, and a single-mode waveguide connected to the end of the adiabatic tapered waveguide.

7. An N×N silicon-based MEMS optical switch array device, characterized in that: The optical switch array device includes at least four cascaded MEMS optical switches as described in any one of claims 1 to 6.

Images