A polarization-insensitive waveguide device design method

Abstract

The application discloses a polarization-insensitive waveguide device design method, which comprises the following steps: setting a photonic integrated material platform; setting a functional waveguide device or system for optical signal processing on the photonic integrated material platform; setting an input end conversion device in front of the functional waveguide device or system; and setting an output end conversion device behind the functional waveguide device or system. The design method reduces the design difficulty, does not introduce extra power consumption and on-chip loss, significantly improves the system integration, and provides a new scheme for the realization of polarization-insensitive waveguide devices and systems.

Description

Technical Field

[0001] This invention relates to the field of semiconductor technology, and more specifically to a design method for a polarization-insensitive waveguide device and system. Background Technology

[0002] Photonic integration, using photons as the carrier of information transmission, features low latency, low power consumption, and high bandwidth, demonstrating enormous development potential in data communication and attracting widespread research from scientists both domestically and internationally. Photonic integration platforms can be categorized based on substrates such as indium phosphide-based, sapphire-based, and silicon-based. To increase the waveguide's confinement of optical modes and improve integration density, a certain refractive index difference must be maintained between the core and cladding materials. The difference in stress in different directions leads to strong birefringence and significant polarization dependence in most platforms. Typical devices such as directional couplers and modulators are often only suitable for one polarization; otherwise, device performance will be severely degraded.

[0003] However, during optical fiber communication, due to the bending of ordinary single-mode optical fibers and changes in the surrounding environment, the transmitted optical signal exhibits random and dynamic changes, rather than a single polarization state. While polarization-maintaining fibers can maintain the stability of a single polarization state, they are incompatible with existing infrastructure for long-distance transmission, significantly increasing operating costs. Therefore, to adapt to the random polarization state of optical signals in optical fibers and receive complete information, polarization-insensitive functional devices or systems are often designed on mainstream photonic integration platforms. These solutions mainly include the following approaches.

[0004] The first approach is to use symmetrical waveguide structures and stress engineering to achieve isotropic waveguides, and then design polarization-insensitive waveguide devices based on this. However, this waveguide structure usually has weak constraint on optical modes and low integration, making it difficult to achieve monolithic integration with microelectronic devices. Furthermore, stress control increases the operational difficulty.

[0005] The second approach is polarization diversity, which utilizes devices with polarization separation or rotation capabilities, such as polarization beam splitters (PBS), polarization rotators (PSRs), and two-dimensional gratings, to orthogonally decompose the random polarization states coupled from the optical fiber into two paths, or further rotate the orthogonal polarization states into two paths of the same polarization. These paths then pass through a functional waveguide device or system, and finally, through the aforementioned photonic device with polarization separation or rotation capabilities placed in opposite directions, the two beams are combined. This approach has a simple technical route, but the addition of two functional waveguide devices or systems at least doubles the chip area. Furthermore, as the scale of the functional waveguide devices or systems increases, the number of electrodes also increases the difficulty of electrical packaging and driver circuit design.

[0006] The third approach involves adding an active polarization control device before the functional waveguide device or system to convert the random polarization state coupled from the optical fiber to the optical chip into a single polarization state. First, the random polarization is orthogonally decomposed into two paths by a polarization separator and rotated to the same polarization. Then, phase and intensity modulation is used to achieve high-efficiency beam combining of the two beams before they enter the functional waveguide device or system. During this process, the intensity and phase of the orthogonal components of the random polarization are random, requiring a feedback algorithm to find the optimal modulation state. Therefore, the active polarization control scheme increases system power consumption and is more complex to operate.

[0007] In summary, how to design simple and compact on-chip polarization-insensitive functional devices or systems has become an urgent problem to be solved. Summary of the Invention

[0008] This invention provides a method for converting orthogonal polarization states in a waveguide into different modes of a single polarization state, thereby realizing a polarization-insensitive functional waveguide device or system. By utilizing a passive conversion device to achieve the requirement of converting polarization independence to mode independence, the design difficulty of integrated devices is effectively alleviated, system integration is improved, no additional power consumption is added, and on-chip losses are not significantly increased. This provides a novel implementation scheme for a polarization-insensitive functional waveguide device or system.

[0009] A polarization-insensitive waveguide device design method includes the following steps:

[0010] 1) Establish a photonic integrated materials platform;

[0011] 2) Set up functional waveguide devices or systems for information processing on the aforementioned photonic integrated material platform;

[0012] 3) Install the functional waveguide device or the input conversion device before the system;

[0013] 4) Set the output conversion device after the functional waveguide device or system.

[0014] The material platform mentioned includes any platform with birefringence effect and polarization dependence, such as silicon-based platforms.

[0015] The input conversion device connects the input waveguide and the functional waveguide device or system front end, and can convert orthogonal polarization modes into different modes of a single polarization state.

[0016] The different modes of a single polarization state are combinations of any two or more modes of a single polarization state. Combinations of any two or more modes of a single polarization state include combinations of two or more higher-order modes or combinations of the fundamental mode and one or more higher-order modes, such as transforming TE0\TM0 into combinations of TE0\TE1, TM1\TM2, TM1\TM0 and TM2, etc. Higher-order modes are modes with orders higher than the fundamental mode. For example, the fundamental mode of TE polarization is TE0, and higher-order modes include TE1, TE2, TE3, etc.

[0017] The output conversion device is connected to the functional waveguide device or the system back end and the output waveguide; the output conversion device is an input conversion device placed in reverse. According to the optical path reversibility, the output conversion device converts different modes of a single polarization state back to the corresponding orthogonal polarization state mode.

[0018] The aforementioned functional waveguide device or system is used for optical signal processing, including but not limited to cross waveguides, optical switches, modulators, etc., and can operate in different modes of the single polarization.

[0019] When the functional waveguide device described in step 2) is an optical switch or modulator, the photonic integrated material platform is set as a thin-film lithium niobate (LNOI) platform; the optical switch or modulator is set as a Mach-Zehnder interferometer (MZI) structure, including a beam splitter / combiner and a phase shifter arm, both of which are suitable for TE0 and TE1 modes; an input conversion device is set at the front end of the optical switch or modulator to convert the TM0 mode to the TE1 mode without changing the TE0 mode; the two modes, TE0 and TE1, are switched or modulated by the waveguide device; an output conversion device is set at the rear end of the optical switch or modulator to convert the TE1 mode back to the TM0 mode without changing TE0, thereby obtaining a polarization-insensitive optical switch or modulator.

[0020] Figure 1 This is a schematic diagram of the method. Its main operations include the following processes:

[0021] 1. After the random polarization state in a single-mode fiber is coupled into the input waveguide through a polarization-insensitive end-face coupler, it can be decomposed into a combination of a pair of orthogonal polarization optical modes, such as A0 and B0, where the A polarization state and the B polarization state are a pair of orthogonal polarization states.

[0022] 2. After passing through the input conversion device, the orthogonal polarization mode is converted into different modes of a single polarization state, such as C. m C n It should be noted that the number of different modes of the single polarization state is unlimited, C m C n These are just two examples; three or more modes are also possible, such as converting A0 to C. mB0 is transformed into C x With C y The mixing mode; here, C polarization can be a component of another orthogonal basis different from A and B polarization, or it can be A or B polarization, such as B0 being transformed into its higher-order orthogonal polarization mode A. n However, its orthogonal polarization mode A0 remains unaffected.

[0023] 3. Different modes of the same polarization state can achieve specific functions such as signal switching, cross transmission, phase and intensity modulation through functional waveguide devices or systems.

[0024] 4. The output conversion device is an input conversion device placed in reverse. Due to the reversibility of the optical path, C0 and C... n The pattern is reverted to A0 and B0.

[0025] Based on this schematic diagram, polarization-insensitive functional waveguide devices or systems such as polarization-insensitive cross waveguides and modulators can be designed.

[0026] Compared with the prior art, the present invention has the following advantages:

[0027] 1. Compared with isotropic waveguides, this design is suitable for photonic integrated material platforms with birefringence effects, such as silicon-based and indium phosphide-based waveguides, and has no special requirements for processing technology, thus reducing costs.

[0028] 2. Compared to active control schemes, this scheme only introduces passive conversion devices and does not introduce additional power consumption;

[0029] It does not significantly increase on-chip loss.

[0030] 3. Compared to polarization diversity schemes, this scheme only requires one polarization and mode conversion at each of the input and output ends of the polarization-insensitive functional system, greatly saving area, significantly improving system integration, and reducing costs. Especially for large-scale polarization-insensitive functional systems, such as when constructing large-scale array optical switches, only one interchange between orthogonal polarization and single polarization modes is needed at the array input and output stages, with intermediate stages involving only one polarization. This scheme is even more beneficial for integration.

[0031] 4. Unlike multiple polarization states, the principal electric field directions of different modes under the same polarization are the same, and they are less restricted by waveguide structures. Therefore, mode insensitivity under the same polarization is easier to achieve. This invention realizes polarization-insensitive devices by utilizing mode insensitivity, reducing design difficulty and alleviating process limitations. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the invention;

[0033] Figure 2 This is a schematic diagram of a polarization-insensitive optical switch unit.

[0034] Among them, [1] is a silicon substrate, [2] is a silicon dioxide buried oxide layer, [3] is a thin film lithium niobate material plate layer, [4]

[14] are input and output single-mode ridge waveguides, [5]

[13] are input and output polarization and mode converters, [6] and

[12] are mode-insensitive 3dB couplers (taking MMI as an example, other types of beam splitters such as DC can also be selected), [7]

[11] are phase shift arms, [8]

[10] are ground electrodes, and [9] are signal electrodes.

[0035] Figure 3 The simulated optical field diagrams of the polarization and mode converter are as follows: (a) Input is TE0 and no conversion is performed; (b) Input is TM0 and converted to TE1.

[0036] Figure 4 The simulated optical field diagrams of the mode-insensitive multimode interference coupler are as follows: (a) Input is TE0; (b) Input is TE1.

[0037] Figure 5 This is a schematic diagram of the electric field line distribution at the cross section of the phase-shifting arm;

[0038] Figure 6 Voltage-optical power simulation curves for TE0 and TM0 mode inputs; Detailed Implementation

[0039] Example:

[0040] As a core device for optical path switching and modification, optical switches can effectively overcome the limitations of electrical domain switching and better adapt to the needs of high-speed, high-capacity, and low-loss optical transmission networks, thus attracting extensive research. Thin-film lithium niobate platforms, as a novel material platform, possess excellent linear electro-optic effects, extremely high modulation speeds, and extremely low transmission losses. The following section will elaborate on this concept using a polarization-insensitive optical switch unit based on an x-cut y-transmission insulator lithium niobate (LNOI) platform.

[0041] Figure 2 This is a three-dimensional schematic diagram of an optical switching unit, which includes a silicon substrate 1, a silicon dioxide buried oxide layer 2, a thin-film lithium niobate material planar layer 3, an input single-mode ridge waveguide 4, an output single-mode ridge waveguide 14, an input polarization and mode converter 5, an output polarization and mode converter 13, mode-insensitive 3dB couplers 6 and 12 (taking MMI as an example, other types of beam splitters such as DC can also be selected), phase shift arms 7 and 11, ground electrodes 8 and 10, and a signal electrode 9.

[0042] This section describes the specific working process using a single-port input scenario. First, an end-face coupler couples the randomly polarized optical signal from the fiber into a single-mode waveguide 4, which can be decomposed into a pair of orthogonally polarized fundamental modes, TE0 and TM0. The input polarization and mode converter 5 converts the TM0 mode to TE1 mode without affecting the TE0 mode. Then, TE0 and TE1 enter the mode-insensitive 2×2 MMI6 and are uniformly split at the two output ports with a phase difference of π / 2. The two beams then enter the phase-shifting arms 7 and 11, which are insensitive to the TE0 and TE1 modes, respectively. When voltage is applied to the GSG electrodes 8, 9, and 10, the two modes, after modulation by the two arms, produce the same phase difference and interfere again in the second MMI12, changing the switching state. The two output beams enter the output-end conversion device 13, where the TE1 mode reverts to the TM0 mode, while the TE0 mode remains unaffected. Finally, the beams are output through the single-mode waveguide or transmitted to subsequent devices. It is worth noting that when this switching unit is used to construct a polarization-insensitive optical switch array, only one polarization and mode conversion is needed at the array's input and output, respectively, thus ensuring that the array operates with only a single polarization. Therefore, the units used in the switch array can be simplified based on the complete unit structure described above. For example, the intermediate-level units of the array do not need conversion devices 4 and 5 and single-mode waveguides 13 and 14; instead, multi-mode waveguides can be used to connect the input and output ports. Simultaneously, the input-level optical switch units no longer require 13 and 14, and the output-level switch units do not require 4 and 5.

[0043] The optical switch unit employs GSG differential electrodes 8, 9, and 10. After applying voltage, the electric fields on the two phase shift arms are equal in magnitude and opposite in direction, with a phase difference twice the phase change of a single phase shift arm. For the MZI structure, the optical switch achieves crossover and through states when the phase difference between the two phase shift arms is 0 and π, respectively.

[0044] 5.13 An adiabatic wedge structure can be adopted. Based on the dispersion curve of the lithium niobate (LN) waveguide structure, TM0 and TE1 hybridization exists within a certain waveguide width range. Since TE0 has no hybridization points with other modes, a suitable waveguide width can be selected to construct a wedge structure, thereby achieving high-efficiency conversion from TM0 to TE1 modes without affecting TE0. Figure 3 The simulation yields the light field diagram of 5, which verifies that (a) shows the case with TE0 input, and the output mode remains TE0; (b) shows the case with TM0 input, and the output mode changes to TE1. Therefore, before entering the first MMI6, the polarization state of the input light is converted to TE polarization. Due to the reversibility of the optical path, when 5 is reversed, i.e., the directions of the input and output are exchanged, the TE1 mode can be converted back to the TM0 mode, while TE0 remains unchanged after transmission.

[0045] Multimode interference couplers [6]

[12] have a self-image effect. Based on the general interference principle, a 2×2 MMI that is insensitive to TE0 and TE1 can be simulated. Figure 4 The image shows the light field obtained from the Lumerical simulation software for photonic devices. Both modes achieve uniform light splitting, and the phase difference between the two output ports is π / 2.

[0046] Lithium niobate is anisotropic and is a negative uniaxial crystal. Its modulation efficiency is highest when the polarization direction is the same as the applied electric field direction. On the x-cut y-transmission LNOI platform, the TE polarization direction is consistent with the applied electric field direction, and the corresponding LN principal axis has a larger electro-optic coefficient than the principal axis corresponding to TM polarization. The modulation efficiency of the modulation arm for TE polarization is approximately three times that for TM polarization. For LN waveguides with a width on the micrometer scale and a height on the submicrometer scale, specific electrode thicknesses and electrode-to-waveguide distances can be selected to approximate a uniform electric field within the waveguide.

[0047] Accordingly, the following Figure 5 The image is a simulation diagram from the optoelectronic simulation software COMSOL, where 201 is the upper cladding, 202 is the lower cladding, 203 is the LN waveguide layer, and 204 and 205 are gold electrodes. The TE polarization direction is the same as the optical axis, and different modes correspond to the same electro-optic coefficient in the material. Therefore, according to the linear electro-optic effect, the effective refractive index changes of different TE polarization modes are basically the same when the same bias voltage is applied, thus obtaining modulation arms with similar modulation effects for TE0 and TE1. Through joint simulation using COMSOL and Lumerical, an LNOI platform with a core thickness of 600 nm and an etching depth of 200 nm was obtained. When the distance between the metal and the waveguide edge is 1.8 μm, and the waveguide width is 2.8 or 3 μm, the change in effective refractive index of TE0 and TE1 modes after applying voltage is the same, 0.000015 / V.

[0048] Figure 6 The simulation curves of voltage-optical power are shown for TE0 and TM0 mode inputs. TE0 and TM0 are input from either of the input ports shown in the figure, with Port1 and Port2 corresponding to the through and cross-output ports, respectively. It can be seen that when TE0 is input, the extinction ratios of the optical switch in the BAR and CROSS states are 29dB and 46dB, respectively, with a loss of 1dB in both states. When TM0 is input, the extinction ratios of the optical switch in the BAR and CROSS states are also 29dB and 46dB, respectively, with a loss of approximately 1.9dB. The driving voltages for the two polarization states, BAR and CROSS, are relatively consistent, and the polarization-dependent loss (PDL) is approximately 0.9dB. The PDL is mainly related to the polarization-dependent loss of the converter and beam splitter, and the absorption loss of the gold electrode pair in both modes. The former can be reduced through design optimization, while the latter can be improved through process optimization.

[0049] In summary, by adding conversion devices before and after the mode-independent optical switch, a novel implementation scheme for polarization-insensitive optical switches is provided. Compared with traditional design methods, it has significant advantages in terms of integration, operational complexity, power consumption, and polarization-related losses.

Claims

1. A method for designing polarization-insensitive waveguide devices, characterized in that, Includes the following steps: 1) Establish a photonic integrated materials platform; The aforementioned photonic integrated material platform is a platform with birefringence effect; 2) Configure functional waveguide devices or systems on the aforementioned photonic integrated material platform; 3) Install the functional waveguide device or the input conversion device before the system; The input conversion device connects the input waveguide and the functional waveguide device or system front end, and can convert orthogonal polarization mode into different modes of single polarization. The different modes of the single polarization state are any combination of two or more modes of the single polarization state. Any combination of two or more modes of the single polarization state includes the combination of two or more higher-order modes or the combination of the fundamental mode and one or more higher-order modes. Higher-order modes are modes with orders higher than the fundamental mode. 4) Set the output conversion device after the functional waveguide device or system.

2. The design method for polarization-insensitive waveguide devices according to claim 1, characterized in that, In step 4), the output conversion device is connected to the functional waveguide device or the system back end and the output waveguide; the output conversion device is an input conversion device placed in reverse. According to the optical path reversibility, the output conversion device converts different modes of a single polarization state back to the corresponding orthogonal polarization state mode.

3. The design method for polarization-insensitive waveguide devices according to claim 1, characterized in that, In step 2), the functional waveguide device or system is used for optical signal processing and can operate normally in different modes of the single polarization state as described in claim 1.

4. The polarization-insensitive waveguide device design method according to claim 1, characterized in that, When the functional waveguide device described in step 2) is an optical switch or modulator, the photonic integrated material platform is set to a thin-film lithium niobate platform; the optical switch or modulator is set to a Mach-Zehnder interferometer structure, including a beam splitter / combiner and a phase shifter arm, both of which are suitable for... , model; An input conversion device is installed at the front end of the optical switch or modulator to convert the input to the input of the optical switch or modulator. Pattern transformation Pattern, without change model; , The two modes are switched or modulated via functional waveguide devices; An output conversion device is installed at the back end of the optical switch or modulator to convert the output of the optical switch or modulator. Mode switch Pattern, unchanged This leads to polarization-insensitive optical switches or modulators.

Images