Polarization-insensitive large process-tolerant wavelength division multiplexer

Abstract

This invention discloses a polarization-insensitive wavelength division multiplexer with high process tolerance. It achieves stable power distribution across a wide spectral range by cascading a polarization-insensitive power divider and a group delay converter. The power divider employs three directional couplers and two phase-shifting units with multi-segment waveguide structures to achieve stable power distribution. The polarization-insensitive group delay converter achieves group delay consistency between TE and TM modes through phase-shifting units with multi-segment waveguide structures. The multi-segment waveguide structure is composed of at least four different widths of waveguide segments connected in series, excluding those with a specific width for stable modes. This enables the device to achieve phase line coincidence of dual polarization states and alignment of the free spectral range while exhibiting first-order robustness to waveguide etching deviations. It achieves polarization independence and high process tolerance without requiring additional polarization compensation devices, significantly improving manufacturing yield and providing wavelength division multiplexing with flat-top filtering characteristics.

Description

Technical Field

[0001] This invention belongs to the field of integrated photonics technology, and more specifically, relates to a wavelength division multiplexer that is polarization insensitive and has a large process tolerance. Background Technology

[0002] With the explosive growth in bandwidth demands for modern high-speed optical communication systems and data center interconnects, wavelength division multiplexing (WDM) technology has become a core means to improve fiber optic transmission capacity and frequency utilization. In the field of integrated photonics, achieving wavelength multiplexing and demultiplexing on a single integrated photonic chip can significantly reduce system power consumption and module size. However, the development of high-performance WDM multiplexers faces two major technological challenges: polarization sensitivity and manufacturing process tolerance.

[0003] First, silicon-based photonic waveguides exhibit strong birefringence. Due to the strong field confinement caused by the high refractive index difference, the effective refractive indices of the transverse electric (TE) mode and the transverse magnetic (TM) mode of the waveguide differ significantly. In practical fiber optic transmission links, the polarization state of the input optical signal often exhibits random uncertainty due to factors such as environmental stress, fiber bending, and polarization mode dispersion. If traditional wavelength division multiplexers cannot achieve polarization-insensitive design, the TE and TM modes will face different phase delays and group delays, leading to drift or interleaving of the filter envelope, resulting in significant polarization-dependent losses and polarization-dependent wavelength drift, severely restricting the demodulation quality and bit error rate performance of the signal.

[0004] Secondly, submicron-scale silicon waveguides are extremely sensitive to manufacturing process variations. In the mass production of large-scale integrated photonic chips, factors such as lithography precision errors, etching bias, and uneven material deposition can cause critical waveguide dimensions to deviate from design values. For silicon-based platforms with high refractive index contrast (such as SOI and deposited silicon), nanometer-level fluctuations in waveguide width (e.g., ±10nm) can lead to a significant shift in the effective refractive index, resulting in a drift in the center wavelength of the wavelength division multiplexer and a change in the free spectral range. This low process tolerance not only significantly reduces chip yield but also increases reliance on subsequent active thermal compensation circuitry, increasing system power consumption and control complexity.

[0005] Currently, existing solutions to these problems still have significant limitations. While common polarization-independent circuits can circumvent polarization sensitivity issues, they require the introduction of additional polarization beam splitters and dual wavelength division multiplexing (WDM) structures, significantly increasing device size and link loss. Passive polarization-independent designs achieved through optimized waveguide geometry often lack robustness to process variations, resulting in poor performance in large-scale production, especially in advanced processes involving complex thermodynamic processes such as III-V heterogeneous integration, where stability and performance are difficult to balance.

[0006] Therefore, developing a novel wavelength division multiplexer that can simultaneously achieve polarization insensitivity and large process tolerance is of urgent practical need and significant economic value for realizing high-performance, low-cost, and high-reliability integrated optical communication systems. Summary of the Invention

[0007] In view of the above-mentioned defects or improvement needs of the existing technology, the present invention provides a wavelength division multiplexer that is polarization insensitive and has a large process tolerance, thereby solving the technical problem of how to realize a wavelength division multiplexing scheme that does not require additional polarization compensation devices, has high process tolerance, and can provide flat-top filtering characteristics.

[0008] To achieve the above objectives, according to a first aspect of the present invention, a wavelength division multiplexer with polarization insensitivity and large process tolerance is provided, comprising: Multiple cascaded filters; The filter includes a power divider and a group delay converter; the power divider includes three directional couplers and two phase shift units; the group delay converter includes a phase shift unit; the directional couplers in the power divider satisfy polarization insensitivity by finding the polarization-insensitive point of coupling efficiency; the group delay converter achieves polarization insensitivity through refractive index dispersion compensation; the phase shift units are configured according to the requirements of a Butterworth flat-top filter, and each phase shift unit includes a first waveguide group and a second waveguide group; wherein, both the first waveguide group and the second waveguide group include a multi-segment waveguide structure, the multi-segment waveguide structure consisting of at least four different widths of waveguide segments excluding a specific width used for stable modes. The waveguides are connected in series, and the widths of the waveguide segments in the first waveguide group and the second waveguide group are the same. In the phase shift unit of the power divider, the width sequence of each waveguide segment in the multi-segment waveguide structure and the corresponding logical length sequence enable the phase shift of the corresponding phase shift unit in TE mode and TM mode to be aligned and ensure the stability of the phase shift relative to the waveguide width. In the phase shift unit of the group delay converter, the width sequence of each waveguide segment in the multi-segment waveguide structure and the corresponding logical length sequence enable the group delay of the corresponding phase shift unit in TE mode and TM mode to be aligned and ensure the stability of the effective refractive index relative to the waveguide width.

[0009] According to any of the above-mentioned polarization-insensitive, high-process-tolerance wavelength division multiplexers, in the phase shift unit of any power divider, the width sequence of each waveguide segment in the first waveguide group and the second waveguide group, and the logical length sequence corresponding to the width sequence are configured as follows: Based on the preset coupling efficiency of any power divider and the coupling efficiency selected for the directional coupler in any power divider, determine the target phase shift of the first phase shift unit and the target phase shift of the second phase shift unit in any power divider. Based on the target phase shift of the first phase shift unit, the width sequence of each waveguide segment in the first waveguide group and the second waveguide group in the first phase shift unit and the logical length sequence corresponding to the width sequence are selected so that the selected width sequence and logical length sequence can ensure that the phase shift generated by the first phase shift unit in TE mode and TM mode are equal, and the rate of change of the phase shift with the width of the waveguide segment in TE mode and TM mode is zero. Based on the target phase shift of the second phase shift unit, the width sequence of each waveguide segment in the first waveguide group and the second waveguide group in the second phase shift unit, as well as the logical length sequence corresponding to the width sequence, are selected so that the selected width sequence and logical length sequence can ensure that the phase shift generated by the second phase shift unit in TE mode and TM mode are equal, and the rate of change of the phase shift with respect to the width of the waveguide segment in both TE mode and TM mode is zero.

[0010] According to any of the above-mentioned polarization-insensitive wavelength division multiplexers with large process tolerance, the directional coupler satisfies the following constraints: the partial derivative of the coupling efficiency with respect to wavelength is less than a preset threshold, and the difference between the odd and even modes of the directional coupler in TE mode is equal to the difference between the odd and even modes in TM mode.

[0011] Based on any of the above-mentioned polarization-insensitive wavelength division multiplexers with large process tolerance, the width sequence of each waveguide segment in the first waveguide group and the corresponding logical length sequence of the width sequence are selected based on the following set of constraint equations:

[0012] in, and These represent the phase shift amounts generated by the first phase shift unit or the second phase shift unit in TE mode and TM mode, respectively. For the target wavelength, This refers to the number of waveguide segments with different widths in the first waveguide group and the second waveguide group. Representing different widths. For the i-th width, for The corresponding logical length, and Let be the effective refractive index of the TE mode and the effective refractive index of the TM mode for the waveguide segment with the i-th width, respectively. The target phase shift amount is the first phase shift unit or the second phase shift unit.

[0013] According to any of the above-mentioned polarization-insensitive, high-process-tolerance wavelength division multiplexers, in any group of time-delay converters, the width sequence and the logical length sequence corresponding to each waveguide segment in the first waveguide group and the second waveguide group are configured as follows: Determine the target filter free spectral range of the filter to which any group of time delay converters belongs; Based on the free spectral range of the target filter, the width sequence of each waveguide segment in the first and second waveguide groups of the phase shift unit of any group of time delay converters and the logic length sequence corresponding to the width sequence are selected, so that the group delay of any group of time delay converters in TE mode and TM mode both reach the target delay amount, and the effective refractive index of TE mode and TM mode changes with the waveguide segment width at a rate of zero.

[0014] Based on any of the above-mentioned polarization-insensitive wavelength division multiplexers with large process tolerance, the width sequence of each waveguide segment in the first and second waveguide groups of the phase shift units of any group of time delay converters and the corresponding logic length sequence of the width sequence are selected based on the following set of constraint equations:

[0015] in, and Let be the group delay of any group delay converter in TE mode and TM mode, respectively, and let c be the speed of light. This refers to the number of waveguide segments with different widths in the first waveguide group and the second waveguide group. Representing different widths. For the i-th width, for The corresponding logical length, For the target wavelength, and Let be the refractive indices of the TE mode group and the TM mode group, respectively, for the waveguide segment with the i-th width. For the target latency, Determined by the free spectral range of the target filter. and These represent the phase shift amounts generated by the phase shift units of any group of time delay converters in TE mode and TM mode, respectively. and denoted as the effective refractive index of the TE mode and the effective refractive index of the TM mode for the i-th width waveguide segment, respectively.

[0016] Based on any of the aforementioned polarization-insensitive, high-process-tolerance wavelength division multiplexers, the physical length of each width of waveguide segment in the first and second waveguide groups of any phase-shifting unit is determined as follows: Based on the logical length sequence corresponding to the width sequence of each waveguide segment in the first waveguide group and the second waveguide group, the logical length corresponding to each width is determined. For each width, if the logic length corresponding to the width is positive, then the physical length of the waveguide segment corresponding to the width in the first waveguide group is determined to be equal to the sum of 1 / 2 of the logic length corresponding to that width and the base length, and the physical length of the waveguide segment corresponding to the width in the second waveguide group is equal to the base length; if the logic length corresponding to the width is negative, then the physical length of the waveguide segment corresponding to the width in the first waveguide group is determined to be equal to the base length, and the physical length of the waveguide segment corresponding to the width in the second waveguide group is equal to the difference between the base length and 1 / 2 of the logic length corresponding to that width.

[0017] According to any of the above-mentioned polarization-insensitive wavelength division multiplexers with large process tolerance, the first waveguide group and the second waveguide group of any phase shift unit include a mode stabilization section, and the length of the mode stabilization section is the basic length.

[0018] According to any of the above-mentioned polarization-insensitive wavelength division multiplexers with large process tolerance, two adjacent waveguide segments of different widths in the multi-segment waveguide structure are connected by an adiabatic tapered waveguide.

[0019] According to any of the above-mentioned wavelength division multiplexers that are polarization insensitive and have large process tolerance, the wavelength division multiplexer uses a silicon deposition platform as the basic fabrication platform.

[0020] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects: A cascaded polarization-insensitive, high-process-tolerance power divider and group delay converter, wherein the power divider employs three directional couplers and two phase-shifting units containing multi-segment waveguide structures, achieves stable power distribution across a wide spectrum, independent of polarization. The polarization-insensitive, high-process-tolerance group delay converter achieves group delay consistency between TE and TM modes through phase-shifting units containing multi-segment waveguide structures. The multi-segment waveguide structure is composed of at least four different widths of waveguide segments connected in series, excluding the specific width used for stable modes. Furthermore, in the phase-shifting units of the power divider, the width sequence of each waveguide segment and its corresponding logical length sequence ensure that the corresponding phase-shifting unit... By aligning the phase shifts in TE and TM modes and ensuring the stability of the phase shifts relative to the waveguide width, in the phase shift unit of the group delay converter, the width sequence of each waveguide segment and its corresponding logic length sequence in the multi-segment waveguide structure enable the corresponding phase shift unit to align the group delays in TE and TM modes and ensure the stability of the effective refractive index relative to the waveguide width. This allows for the realization of phase line coincidence in dual polarization states and alignment of the free spectral range, while also enabling the device performance to have first-order robustness to waveguide etching deviations. It also exhibits polarization independence and high process tolerance without the need for additional polarization compensation devices, significantly improving manufacturing yield and providing wavelength division multiplexing with flat-top filtering characteristics. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the structure of a wavelength division multiplexer with polarization insensitivity and large process tolerance provided in an embodiment of the present invention.

[0022] Figure 2 This is a schematic diagram of the power divider in a wavelength division multiplexer with high process tolerance and insensitive polarization, provided in an embodiment of the present invention.

[0023] Figure 3 This is a schematic diagram of the group delay converter in a wavelength division multiplexer with high process tolerance and polarization insensitivity provided in an embodiment of the present invention.

[0024] Figure 4 This is a schematic diagram of the directional coupler in the power divider of a wavelength division multiplexer with high process tolerance and insensitive polarization, provided in an embodiment of the present invention.

[0025] Figure 5 This is a schematic diagram of the phase shifting unit in the power divider of a wavelength division multiplexer with polarization insensitivity and large process tolerance provided in an embodiment of the present invention.

[0026] Figure 6 This is a 3D structural schematic diagram of the power divider in a wavelength division multiplexer with high process tolerance and insensitive polarization, provided in an embodiment of the present invention.

[0027] Figure 7This is a cross-sectional schematic diagram of the directional coupler in the power divider of a wavelength division multiplexer with high process tolerance and insensitive polarization, provided in an embodiment of the present invention.

[0028] Figure 8 This is a schematic diagram of the phase shifting unit of a wavelength division multiplexer with polarization insensitivity and large process tolerance provided in an embodiment of the present invention.

[0029] Figure 9 The graph shows the variation of effective refractive index with waveguide width, as provided in the embodiments of the present invention.

[0030] Figure 10 The graph shows the variation of the effective refractive index with width as provided in the embodiments of the present invention.

[0031] Figure 11 The graph shows the variation of group refractive index with waveguide width, as provided in an embodiment of the present invention.

[0032] Figure 12 The diagram shows the variation of the group refractive index with width as provided in the embodiments of the present invention.

[0033] Figure 13 The diagram illustrates the effective refractive index and group refractive index under two polarization states provided in the embodiments of the present invention.

[0034] Figure 14 This is a cross-sectional schematic diagram of a silicon deposition platform provided for an embodiment of the present invention. Detailed Implementation

[0035] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0036] This invention provides a wavelength division multiplexer (WDM) with high process tolerance and polarization insensitivity, aiming to solve problems such as high polarization correlation loss introduced by different polarization states in WDM systems and the large number of expensive components designed specifically for polarization correlation. Furthermore, this design concept can be applied to various wavelength bands. The WDM includes multiple cascaded filters. Each filter is an independent functional unit, internally integrating a power divider for power allocation and a group delay converter for phase modulation. Through the collaborative operation of these two device modules and the physical cascading and parameter coordination of multiple filters, the device can achieve efficient WDM or demultiplexing processing for specific optical signals.

[0037] Taking three cascaded filters as an example, such as Figure 1As shown, the wavelength division multiplexer is composed of cascaded filters 11, 12, and 13, which are polarization-insensitive and have high process tolerance. The two output ports of filter 11 are connected to one input port of each of filters 12 and 13, respectively. Each filter has the same structure, including a power divider and a group delay converter. In some embodiments, such as... Figure 2 and Figure 3 As shown, filter 11 is composed of cascaded power dividers 21, 22, 23, group delay converter 31, and 32; filter 12 is composed of cascaded power dividers 21, 22, 23, group delay converter 33, and 34; and filter 13 is composed of cascaded power dividers 21, 22, 23, group delay converter 35, and 36. It should be noted that the structures and parameter settings of components with the same reference numerals in the embodiments of the present invention are the same. For example, the structure and parameter settings of power divider 21 in different filters are the same, the structure and parameter settings of power divider 22 are the same, and the structure and parameter settings of power divider 23 are also the same, which will not be described again later.

[0038] Furthermore, each power divider includes three directional couplers and two phase shift units, and each group delay converter includes one phase shift unit. For example... Figure 4 , 5 As shown in Figure 6, power divider 21 consists of three directional couplers 211, phase shift unit 213, and phase shift unit 214; power divider 22 consists of three directional couplers 211, phase shift unit 215, and phase shift unit 216; and power divider 23 consists of three directional couplers 212, phase shift unit 217, and phase shift unit 218. Here, because directional couplers exhibit large fluctuations in coupling efficiency at different wavelengths, this characteristic limits their performance in wavelength division multiplexing (WDM) systems. Therefore, to achieve a wider wavelength division multiplexing effect, this embodiment of the invention combines cascaded directional couplers and phase shift units into a power divider.

[0039] In each power divider, the directional coupler satisfies polarization insensitivity by finding a polarization-insensitive point in the coupling efficiency. In some embodiments, the directional coupler is configured such that the coupling efficiency's partial derivative with respect to wavelength is less than a preset threshold, and the difference between the odd and even modes in TE mode is equal to the difference between the odd and even modes in TM mode, thereby achieving approximately equal coupling efficiency for both TE and TM modes, i.e., satisfying:

[0040] in, and These represent the odd and even modes of the directional coupler in TE mode. and These represent the odd and even modes of the directional coupler in TM mode, respectively.

[0041] By selecting a coupling efficiency that satisfies the above constraints, the directional coupler achieves extremely stable coupling efficiency across a wide spectral range, effectively suppressing power fluctuations caused by wavelength shift and polarization state switching, thus laying the foundation for polarization-insensitive, high-process-tolerance operation. Therefore, regardless of the polarization state of the input optical signal, the power divider can stably distribute optical power to the two output ports according to a preset ratio, providing a reliable foundation for subsequent optical signal processing. In some embodiments, the coupling efficiency of the directional coupler 211... It can be set to The coupling efficiency of directional coupler 212 It can be set to Then, the corresponding coupling efficiency can be achieved by optimizing the main parameters of the directional coupler, including waveguide width, thickness, and coupling spacing. The cross-section of the directional coupler is as follows: Figure 7 As shown.

[0042] Each phase shift unit needs to be configured according to the requirements of a Butterworth flat-top filter, and each phase shift unit includes a first waveguide group (upper arm) and a second waveguide group (lower arm). Both the first and second waveguide groups include multi-segment waveguide structures, which consist of at least four different widths, excluding the specific width used for stable modes. , … The waveguides are connected in series, and the widths of the waveguide segments in the first and second waveguide groups are the same. Specifically, in the phase-shifting unit of the power divider, the width sequence of each waveguide segment and the corresponding logical length sequence in the multi-segment waveguide structure ensure that the phase shift of the corresponding phase-shifting unit is aligned in both TE and TM modes, guaranteeing the stability of the phase shift relative to the waveguide width. In the phase-shifting unit of the group delay converter, the width sequence of each waveguide segment and the corresponding logical length sequence in the multi-segment waveguide structure ensure that the group delay of the corresponding phase-shifting unit is aligned in both TE and TM modes, guaranteeing the stability of the effective refractive index relative to the waveguide width. Furthermore, in the multi-segment waveguide structures of the first and second waveguide groups, all connecting segments, 90° bends, and 180° bends maintain high symmetry in the upper and lower arms, and their phase contributions cancel each other out, thus ensuring that the final phase shift of the phase-shifting unit is entirely determined by the aforementioned width. , … The waveguide segment is generated.

[0043] like Figure 8As shown, taking four different widths as examples, both the first and second waveguide groups include waveguide segments with widths of W1, W2, W3, and W4. It should be noted that the waveguide segments described in this embodiment of the invention are all waveguide segments with widths not equal to the specific width used for the stable mode. Traditional phase shifters typically only focus on the phase difference generated for a specific wavelength, ignoring the waveguide width deviation caused by etching bias during manufacturing. Therefore, the core principle of this embodiment of the invention lies in utilizing the different effective refractive index sensitivities of waveguides of different widths to TE and TM modes, so that the phase deviations generated by waveguide segments of different widths can physically cancel each other out. Even if the waveguide width shifts due to uneven etching during manufacturing, the overall performance remains robust. In some embodiments, to ensure that the loss of optical signals is minimized when transmitting between waveguide segments of different widths, adjacent waveguide segments of different widths in the multi-segment waveguide structure are connected by a linearly varying adiabatic tapered waveguide to achieve a smooth transition of the optical field mode. The length of the tapered waveguide is the basic length required to maintain mode stability. .

[0044] In other embodiments, in the phase shift unit of any power divider, the width sequence and the logical length sequence corresponding to each waveguide segment in the first waveguide group and the second waveguide group are configured as follows: Based on the preset coupling efficiency of the power divider and the coupling efficiency selected for the directional coupler in the power divider, the target phase shift of the first phase shift unit and the target phase shift of the second phase shift unit in the power divider are determined. Based on the target phase shift of the first phase shift unit, the width sequence of each waveguide segment in the first waveguide group and the second waveguide group in the first phase shift unit and the corresponding logical length sequence are selected so that the selected width sequence and logical length sequence can ensure that the phase shift generated by the first phase shift unit in TE mode and TM mode are equal, and the rate of change of the phase shift with respect to the width of the waveguide segment in TE mode and TM mode is zero. Similarly, based on the target phase shift of the second phase shift unit, the width sequence of each waveguide segment in the first and second waveguide groups of the second phase shift unit and the corresponding logical length sequence are selected so that the selected width sequence and logical length sequence can ensure that the phase shift generated by the second phase shift unit in TE mode and TM mode are equal, and the rate of change of the phase shift with respect to the width of the waveguide segment in both TE mode and TM mode is zero.

[0045] Specifically, such as Figure 6 As shown, taking power divider 21 as an example, the input light has a coupling efficiency of... After the directional coupler 211, the phase shift is After phase shift unit 213, the coupling efficiency is... The directional coupler 211, after a phase shift of After phase shift unit 214, the final coupling efficiency is... After the directional coupler 211, the output is from two splitter ports. The coupling efficiency of the power divider 21 can be obtained from the Jones matrix, which shows that it satisfies the following condition:

[0046]

[0047] in, , For the coupling efficiency of power divider 21, , This represents the coupling efficiency of the directional coupler 211.

[0048] The effects of polarization can be reduced if the following conditions are met simultaneously:

[0049] in, The preset coupling efficiency of power divider 21, λ is the wavelength.

[0050] Therefore, once the preset coupling efficiency of the power divider 21 and the coupling efficiency of the directional coupler 211 are determined, the target phase shift of the phase shift unit 213 can be obtained by solving the above set of equations. and the target phase shift amount of phase shift unit 214 .

[0051] In some embodiments, the preset coupling efficiency of power divider 21, power divider 22, and power divider 23 , and They can be set separately Preferably, In the power divider 21, set , , ; In the power divider 22, set , , ; The power divider 23 is set up. , , .

[0052] like Figure 9 and Figure 10 As shown, waveguide segments of different widths Different TE mode effective refractive indices TM mode effective refractive index TE mode effective refractive index differential error Effective refractive index differential error in TM mode The phase shift amounts of the TE mode light and TM mode light in a single phase shift unit are as follows:

[0053]

[0054] The phase error caused by manufacturing process deviations can be derived as the derivative of the phase with respect to the waveguide width:

[0055]

[0056] To ensure the realization of a phase-shifting unit with large process tolerance and insensitivity to polarization, the above formula needs to be constrained, namely:

[0057] in, and These represent the phase shift amounts generated by a single phase shift unit in TE mode and TM mode, respectively. For the target wavelength, This refers to the number of waveguide segments with different widths in the first and second waveguide groups. Representing different widths. For the i-th width, for The corresponding logical length, and Let be the effective refractive index of the TE mode and the effective refractive index of the TM mode for the waveguide segment with the i-th width, respectively. This represents the target phase shift amount for this phase shift unit.

[0058] In the above set of constraint equations, the first two equations ensure that, under standard design dimensions, the TE and TM modes can produce the preset phase shift. The first equation achieves polarization independence; the second equation eliminates the dependence of phase on width, making the width fluctuations caused by the single phase shift on the sidewall etching tilt angle and uneven photoresist development first-order immune, thus achieving large process tolerance.

[0059] The structure of the phase shift unit in a group delay converter is the same as that in a power divider, but the group delay converter achieves polarization insensitivity through refractive index dispersion compensation. In some embodiments, precise quantization of the group delay of the group delay converter aims to preset and control the free spectral range (FSR) of the photonic device. In interferometric devices, FSR is defined as the wavelength interval between adjacent transmission peaks in the spectral response, and its value is related to the group delay difference between the two arms of the interference device. They exhibit an inverse correlation. Therefore, this embodiment of the invention uses a multi-segment width optimization method to constrain the group delay of TE and TM modes, essentially providing a unified and accurate FSR characteristic for the device under different polarization states. To achieve precise control of signal transmission time and eliminate polarization-dependent delay, this embodiment of the invention solves for the group refractive index and group delay sensitivity of the phase shift unit in the group delay converter simultaneously, achieving stability in the time domain performance. For example, Figure 11 and Figure 12 As shown, waveguide segments of different widths have different TE mode group refractive indices. TM mode group refractive index To ensure that group delay compensation with large process tolerance and insensitivity to polarization can be achieved, constraints on the group delay conditions are required.

[0060] Specifically, in any group of time-delay converters, the width sequence and the corresponding logical length sequence of each waveguide segment in the first and second waveguide groups are configured as follows: Determine the target filter free spectral range of the filter to which this group of time delay converters belongs; Based on the free spectral range of the target filter, the width sequence of each waveguide segment in the first and second waveguide groups of the phase shift unit of the group delay converter, as well as the logic length sequence corresponding to the width sequence, are selected so that the group delay of the group delay converter in both TE and TM modes reaches the target delay amount, and the effective refractive index of both TE and TM modes changes with the waveguide segment width at a rate of zero.

[0061] In other embodiments, the width sequence of each waveguide segment in the first and second waveguide groups of the phase shift unit of the group delay converter, and the corresponding logic length sequence of the width sequence are selected based on the following set of constraint equations:

[0062] in, and , where are the group delays of the group delay converter in TE mode and TM mode, respectively, and c is the speed of light. This refers to the number of waveguide segments with different widths in the first and second waveguide groups. Representing different widths. For the i-th width, for The corresponding logical length, For the target wavelength, and Let be the refractive indices of the TE mode group and the TM mode group, respectively, for the waveguide segment with the i-th width. For the target latency, Determined by the free spectral range of the target filter. and These represent the phase shift amounts generated by the phase shift units of this group of time-delay converters in TE and TM modes, respectively. and denoted as the effective refractive index of the TE mode and the effective refractive index of the TM mode for the i-th width waveguide segment, respectively.

[0063] The minimum waveguide width number n=4 is achieved by solving the above linear equations. The width sequence of each waveguide segment and its corresponding logical length sequence can be obtained. This logical length sequence ensures that: (1) the group delay increments generated by the TE mode and TM mode after passing through the phase shift unit are equal, eliminating polarization mode dispersion; (2) since the partial derivative of the phase with respect to the waveguide width is zero, in the actual processing, the influence of the waveguide width deviation caused by process fluctuations such as photolithography and etching on the final phase enters the second-order small range, which greatly improves the yield and timing robustness of the device. A polarization-insensitive flat-top passband can be achieved through this method.

[0064] like Figure 13 As shown, for most waveguide widths, the group refractive index of the TM mode is greater than that of the TE mode, but the refractive index of the TM mode is smaller than that of the TE mode. In the constraint equations of the group delay converter given in the above embodiment, the group delay of the TE mode is equal to that of the TM mode. In this case, the phase of the TE mode is greater than that of the TM mode. To achieve passband alignment between the TE and TM modes, the phase difference between the TE and TM modes must be an even multiple of the wavelength, i.e.:

[0065] in It is an integer. Where λ is the wavelength. Under certain FSR conditions, the above equations may have no solution.

[0066] In some embodiments, a waveguide segment of a specific width can be added to both the first and second waveguide groups of the phase shift unit of the group delay converter. The width of this newly added waveguide segment... With the thickness of the waveguide section The equality ensures that the cross-section of the newly added waveguide segment is a regular square. This square waveguide segment possesses inherent polarization independence characteristics, namely:

[0067] in, and These are the effective refractive indices of the TE mode and TM mode for the square waveguide segment, respectively. and These are the refractive indices of the TE mode group and the TM mode group, respectively, for this square waveguide segment.

[0068] Therefore, the introduction of this square waveguide segment adds phase to both the TE and TM modes. Group latency They are the same. At this point, the overall group delay is... To ensure that the group delay remains unchanged, all waveguide segments except the square waveguide segment need to be shortened proportionally. The phase difference at this point can be written as:

[0069] Simplify to obtain

[0070] in, The logic length of the square waveguide segment.

[0071] This algorithm in the embodiment achieves the goal of meeting the demand. to length sequence The transformation of the solution enables the length calculation under different waveguide width combinations.

[0072] Therefore, when the equations given in the above embodiments cannot be solved, the following constraint equations can be used to solve for the width (the width of the square waveguide segment does not need to be solved) and logic length of all waveguide segments, including the square waveguide segment.

[0073]

[0074] In some embodiments, in order to physically eliminate negative solutions of the logical length that may occur in mathematical calculations and to ensure the symmetry of the optical path, the logical length can be... This is mapped to the length difference between two waveguide segments of the same width in the first and second waveguide groups. Specifically, the physical lengths of waveguide segments of each width in the first and second waveguide groups of any phase-shifting unit (phase-shifting unit in a power divider or group delay converter) are determined based on the following: Based on the logical length sequence corresponding to the width sequence of each waveguide segment in the first waveguide group and the second waveguide group, determine the logical length corresponding to each width. For each width, if the logic length corresponding to the width is positive, then the physical length of the waveguide segment corresponding to the width in the first waveguide group is determined to be equal to the sum of 1 / 2 of the logic length corresponding to that width and the base length, and the physical length of the waveguide segment corresponding to the width in the second waveguide group is equal to the base length; if the logic length corresponding to the width is negative, then the physical length of the waveguide segment corresponding to the width in the first waveguide group is determined to be equal to the base length, and the physical length of the waveguide segment corresponding to the width in the second waveguide group is equal to the difference between the base length and 1 / 2 of the logic length corresponding to that width.

[0075] That is, the first waveguide group and the second waveguide group have the same width. The physical lengths of the two waveguide segments and for:

[0076]

[0077] in, for The corresponding logical length, Based on the length.

[0078] In some embodiments, both the first waveguide group and the second waveguide group of any phase-shifting unit include a mode-stabilized section, the length of which is the aforementioned basic length. Preferably, It can ensure that the light field is redistributed and stabilized in the main mode after passing through the width transformation or bending section, suppressing the excitation of higher-order modes, thereby improving mode purity and reducing loss.

[0079] This differential compensation mechanism ensures that the physical fabrication length of the waveguide segment is always positive, regardless of the calculation result, thus avoiding manufacturability in the design.

[0080] In other embodiments, the polarization-insensitive, high-process-tolerance wavelength division multiplexer of the present invention uses a silicon deposition platform as the basic fabrication platform. For example... Figure 14 As shown, the thickness of the silicon substrate 41 is 500-700 mm. The thickness of the silica buried layer 42 is 2. The thickness of the deposited silicon layer 43 is 300-500 mm. Preferably, 400 A silicon layer is deposited to etch the waveguide for device fabrication. Finally, the silicon layer is encased in a silicon dioxide cladding 44 to form a protective structure with high refractive index contrast.

[0081] In other embodiments, in III-V heterogeneous integration platforms, a thick silicon device layer with a thickness greater than 400 nm (e.g., 500 nm or more) is typically used as the bottom waveguide. This thick silicon structure aims to enhance the mode field coupling efficiency between the silicon waveguide and the heterobonded III-V material above it. However, thick silicon waveguides exhibit extremely high sensitivity to processing dimensions, particularly etching width, and thermal stress during heterobonding can cause minute distortions in the waveguide shape. Therefore, in the wavelength division multiplexing (WDM) design of this III-V heterogeneous integration platform, for the thick silicon device layer with a thickness greater than 400 nm, a multi-segment waveguide structure with four segments having different waveguide widths is employed. The aforementioned constraint equations are used to simultaneously solve for the group refractive index and phase constant unique to this thick silicon waveguide. During the design process, the effective refractive index distribution of the thick silicon waveguide in the heterobonding environment is obtained through pre-simulation, and first-order derivative constraints on the width deviation are introduced to determine the physical length of each waveguide segment. Meanwhile, by utilizing group delay constraint logic, the free spectral range of the wavelength division multiplexer is matched with the gain longitudinal mode spacing of the upper III-V heterogeneous integrated laser array, ensuring the synchronization of each wavelength channel in the time and frequency domains.

[0082] In summary, the wavelength division multiplexer provided in this embodiment of the invention achieves stable power distribution across a wide spectrum by cascading a polarization-insensitive, high-process-tolerance power divider and a group delay converter. The power divider employs three directional couplers and two phase-shifting units containing multi-segment waveguide structures. The polarization-insensitive, high-process-tolerance group delay converter achieves group delay consistency between TE and TM modes through phase-shifting units containing multi-segment waveguide structures. The multi-segment waveguide structure is composed of at least four different widths of waveguide segments connected in series, excluding the specific width used for stable modes. Furthermore, in the phase-shifting units of the power divider, the width sequence and corresponding logical length sequence of each waveguide segment in the multi-segment waveguide structure are displayed. This technology enables the corresponding phase shift units to align their phase shifts in TE and TM modes and ensures the stability of the phase shifts relative to the waveguide width. In the phase shift units of the group delay converter, the width sequence of each waveguide segment and its corresponding logic length sequence in the multi-segment waveguide structure enable the corresponding phase shift units to align their group delays in TE and TM modes and ensure the stability of the effective refractive index relative to the waveguide width. It can achieve phase line coincidence of dual polarization states and alignment of free spectral ranges, while also making the device performance robust to waveguide etching deviations. It has polarization independence and high process tolerance without the need for additional polarization compensation devices, significantly improving manufacturing yield and providing wavelength division multiplexing with flat-top filtering characteristics.

[0083] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A wavelength division multiplexer that is polarization insensitive and has high process tolerance, characterized in that, include: Multiple cascaded filters; The filter includes a power divider and a group delay converter; the power divider includes three directional couplers and two phase shift units. The group delay converter includes a phase shift unit; the directional coupler in the power divider satisfies polarization insensitivity by finding a polarization-insensitive point of coupling efficiency; the group delay converter achieves polarization insensitivity through refractive index dispersion compensation; the phase shift unit is configured according to the requirements of a Butterworth flat-top filter, and each phase shift unit includes a first waveguide group and a second waveguide group; wherein, both the first waveguide group and the second waveguide group include a multi-segment waveguide structure, the multi-segment waveguide structure being composed of at least four different widths of waveguide segments connected in series, excluding a specific width used for stable modes, and the multi-segment waveguide structure of the first waveguide group... The width of the waveguide segments in the multi-segment waveguide structure of the second waveguide group is the same; in the phase shift unit of the power divider, the width sequence of each waveguide segment in the multi-segment waveguide structure and the logical length sequence corresponding to the width sequence enable the corresponding phase shift unit to align the phase shift amount in TE mode and TM mode and ensure the stability of the phase shift amount relative to the waveguide width; in the phase shift unit of the group delay converter, the width sequence of each waveguide segment in the multi-segment waveguide structure and the logical length sequence corresponding to the width sequence enable the corresponding phase shift unit to align the group delay in TE mode and TM mode and ensure the stability of the effective refractive index relative to the waveguide width.

2. The wavelength division multiplexer with polarization insensitivity and large process tolerance as described in claim 1, characterized in that, In the phase shift unit of any power divider, the width sequence and the logical length sequence corresponding to each waveguide segment in the first waveguide group and the second waveguide group are configured as follows: Based on the preset coupling efficiency of any power divider and the coupling efficiency selected for the directional coupler in any power divider, determine the target phase shift of the first phase shift unit and the target phase shift of the second phase shift unit in any power divider. Based on the target phase shift of the first phase shift unit, the width sequence of each waveguide segment in the first waveguide group and the second waveguide group in the first phase shift unit and the logical length sequence corresponding to the width sequence are selected so that the selected width sequence and logical length sequence can ensure that the phase shift generated by the first phase shift unit in TE mode and TM mode are equal, and the rate of change of the phase shift with the width of the waveguide segment in TE mode and TM mode is zero. Based on the target phase shift of the second phase shift unit, the width sequence of each waveguide segment in the first waveguide group and the second waveguide group in the second phase shift unit, as well as the logical length sequence corresponding to the width sequence, are selected so that the selected width sequence and logical length sequence can ensure that the phase shift generated by the second phase shift unit in TE mode and TM mode are equal, and the rate of change of the phase shift with respect to the width of the waveguide segment in both TE mode and TM mode is zero.

3. The wavelength division multiplexer with polarization insensitivity and large process tolerance as described in claim 2, characterized in that, The directional coupler satisfies the following constraints: the partial derivative of the coupling efficiency with respect to wavelength is less than a preset threshold, and the difference between the odd and even modes of the directional coupler in TE mode is equal to the difference between the odd and even modes in TM mode.

4. The wavelength division multiplexer with polarization insensitivity and large process tolerance as described in claim 2, characterized in that, The width sequence and the corresponding logical length sequence of each waveguide segment in the first waveguide group and the second waveguide group of the first phase shift unit or the second phase shift unit are selected based on the following set of constraint equations: in, and These represent the phase shift amounts generated by the first phase shift unit or the second phase shift unit in TE mode and TM mode, respectively. For the target wavelength, This refers to the number of waveguide segments with different widths in the first waveguide group and the second waveguide group. Representing different widths. For the i-th width, for The corresponding logical length, and Let be the effective refractive index of the TE mode and the effective refractive index of the TM mode for the waveguide segment with the i-th width, respectively. The target phase shift amount is the first phase shift unit or the second phase shift unit.

5. The wavelength division multiplexer with polarization insensitivity and large process tolerance as described in claim 1, characterized in that, In any group of time-delay converters, the width sequence and the corresponding logical length sequence of each waveguide segment in the first waveguide group and the second waveguide group are configured as follows: Determine the target filter free spectral range of the filter to which any group of time delay converters belongs; Based on the free spectral range of the target filter, the width sequence of each waveguide segment in the first and second waveguide groups of the phase shift unit of any group of time delay converters and the logic length sequence corresponding to the width sequence are selected, so that the group delay of any group of time delay converters in TE mode and TM mode both reach the target delay amount, and the effective refractive index of TE mode and TM mode changes with the waveguide segment width at a rate of zero.

6. The wavelength division multiplexer with polarization insensitivity and large process tolerance as described in claim 5, characterized in that, The width sequence of each waveguide segment in the first and second waveguide groups of the phase shift units of any given group of time delay converters, as well as the logic length sequence corresponding to the width sequence, are selected based on the following set of constraint equations: in, and Let be the group delay of any group delay converter in TE mode and TM mode, respectively, and let c be the speed of light under vacuum conditions. This refers to the number of waveguide segments with different widths in the first waveguide group and the second waveguide group. Representing different widths. For the i-th width, for The corresponding logical length, For the target wavelength, and Let be the refractive indices of the TE mode group and the TM mode group, respectively, for the waveguide segment with the i-th width. For the target latency, Determined by the free spectral range of the target filter. and These represent the phase shift amounts generated by the phase shift units of any group of time delay converters in TE mode and TM mode, respectively. and denoted as the effective refractive index of the TE mode and the effective refractive index of the TM mode for the waveguide segment with the i-th width, respectively.

7. The wavelength division multiplexer with polarization insensitivity and large process tolerance as described in any one of claims 2 to 6, characterized in that, The physical length of each width of waveguide segment in the first and second waveguide groups of any phase-shifting unit is determined based on the following method: Based on the logical length sequence corresponding to the width sequence of each waveguide segment in the first waveguide group and the second waveguide group, the logical length corresponding to each width is determined. For each width, if the logic length corresponding to the width is positive, then the physical length of the waveguide segment corresponding to the width in the first waveguide group is determined to be equal to the sum of 1 / 2 of the logic length corresponding to that width and the base length, and the physical length of the waveguide segment corresponding to the width in the second waveguide group is equal to the base length; if the logic length corresponding to the width is negative, then the physical length of the waveguide segment corresponding to the width in the first waveguide group is determined to be equal to the base length, and the physical length of the waveguide segment corresponding to the width in the second waveguide group is equal to the difference between the base length and 1 / 2 of the logic length corresponding to that width.

8. The wavelength division multiplexer with polarization insensitivity and large process tolerance as described in claim 7, characterized in that, Each phase-shifting unit includes a mode-stabilized segment in both the first and second waveguide groups, and the length of the mode-stabilized segment is the base length.

9. The wavelength division multiplexer with polarization insensitivity and large process tolerance as described in claim 1, characterized in that, In the multi-segment waveguide structure, two adjacent waveguide segments of different widths are connected by an insulated tapered waveguide.

10. The wavelength division multiplexer with polarization insensitivity and large process tolerance as described in claim 1, characterized in that, The wavelength division multiplexer uses a silicon deposition platform as its basic fabrication platform.

Images