High-speed polarization controller

Abstract

The disclosed technology herein relates to high-speed performance polarization control systems. In one or more cases, the polarization controller includes an optical circuit configured to split light into a first beam of light and a second beam of light and provide the first beam and second beam of light to a Mach-Zehnder Interferometer (MZI). The MZI is configured to transmit with controlled phase delayed between the first beam and the second beam via waveguides with embedded PN junctions forming a phase modulator. In one or more cases, the polarization controller includes an inverse-designed polarization beam splitter-rotator (PBSR) configured to combine the first beam and second beam and provide the combined first beam and second beam to an output waveguide and waveguide taper.

Description

HIGH-SPEED POLARIZATION CONTROLLERCROSS-REFERENCE TO RELATED APPLICATONS

[0001] The present disclosure claims priority to U.S. Application No. 63 / 574,651 filed April 4, 2024, the contents of which is incorporated by reference in its entirety herein.BACKGROUND

[0002] A polarization controller is a system capable of generating a single output with any combination of polarization states in a waveguide. A conventional polarization controller operates on two principles. The first of which relates to the phase between two coherent sources being modified to control interference between the sources (i.e., the phase shifter). The second of which includes a device that converts two sources into |p) and |s) polarization states at a single output depending on the phase between the relative phase between its two inputs (i.e., the combiner).

[0003] Recently, the proliferation of inverse design techniques in photonic integrated circuits (PICs) has expanded the number of applications to use the adjoint method, particularly in developing ultra-compact and efficient on-chip integrated components. The adjoint method is employed in on-chip polarization control. For instance, polarization-diverse systems, such as polarization splitters and filters, are built by employing multi-level waveguide etching. These systems are typically developed using shape optimization and topology-optimized structures. However, to enable polarization rotation operations, a vertical asymmetry in the geometry of the system needs to be implemented. One possible method to implement vertical asymmetry includes leaving the waveguide system uncladded, thereby causing an asymmetry between the top and bottom boundaries. This method allows mode conversion but at the cost of low performance and increased scattering.

[0004] An alternative method includes using multiple levels of silicon heights, partial etching, or using a trapezoidal waveguide. However, typical approaches for creating multi-layer inverse- designed structures require complex manufacturing processes. For example, in some approaches, topology-optimized polarization rotators converge to solutions in which the top layer is not always contained within the bottom layer and cannot be manufactured in a single silicon (Si) layer with multiple etching steps. While these multi-layer inverse-designed structures can bemanufactured in foundry processes that involve the sequential deposition of secondary optical layers, such as poly-silicon, amorphous Si, or Silicon Nitride, these secondary layers are generally subjected to scattering losses and process complexity. Other approaches may create multi-layer inverse-designed structures by defining geometric constraints in the optimization process. By defining the geometric constraints, the upper layers of Si is contained within the boundaries of the layers below, allowing solutions that match manufacturing restrictions. As such, a multi-layer inverse-designed structure can be fabricated into polarization switches with a single Si layer in a two etch step process. However, the performance of these switches is generally limited to low-speed operation.

[0005] In short, the conventional multi-layer inverse-designed structures are generally large in size, and yet are still subject to high losses, performance degradation, and / or limited to low- speed operation. Thus, there is a need in the art for high-speed polarization control systems having an inverse design structure. The present invention satisfies this need.SUMMARY

[0006] The present disclosure relates generally to integrated circuits, and more particularly, to high-speed performance polarization control systems.

[0007] The disclosed technology herein relates to high-speed performance polarization control systems. In one or more cases, the polarization controller includes an optical circuit configured to split light into a first beam of light and a second beam of light and provide the first beam and second beam of light to a Mach-Zehnder Interferometer (MZI). The MZI is configured to transmit with controlled phase delayed between the first beam and the second beam via waveguides with embedded PN junctions forming a phase modulator. In one or more cases, the polarization controller includes an inverse-designed polarization beam splitter-rotator (PBSR) configured to combine the first beam and second beam and provide the combined first beam and second beam to an output waveguide and waveguide taper.

[0008] A variety of additional aspects will be set forth in the description that follows. The aspects can relate to individual features and to combination of features. It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the broad inventive concepts upon which the embodiments disclosed herein are based.BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The following drawings are illustrative of particular embodiments of the present disclosure and therefore do not limit the scope of the present disclosure. The drawings are not to scale and are intended for use in conjunction with the explanations in the following detailed description.

[0010] FIG. 1 illustrates a block diagram of an example polarization control system.

[0011] FIG. 2A illustrates a graphical representation of the example polarization control system of FIG. 1.

[0012] FIG. 3A is a schematic view of the system 100. FIG. 3B is a side view of a portion of the system 100.

[0013] FIG. 4A illustrates an example Scanning Electron Microscopy (SEM) micrograph of the example inverse designed integrated polarization beam splitter-rotator (PBRS). FIG. 4B illustrates an Atomic Force Microscopy (AFM) of the example PBRS.

[0014] FIGs. 5A-5D illustrate scanning electron micrographs of the example PBRS.

[0015] FIG. 6 illustrates a 3D representation of PBRS of the example polarization control system of FIG. 1.

[0016] FIG. 7 illustrates beam propagation simulations of the PBRS.

[0017] FIG. 8A is a graph illustrating the calculated insertion loss from transmittance measurements.

[0018] FIG. 8B illustrates a layout schematic of the series of PBRS to calculate insertion loss from transmittance measurements.

[0019] FIG. 9A illustrates polarization modulation using a thermos-optic control. FIG. 9B illustrates a Stoke’s sphere corresponding to the polarization modulation of FIG. 9A.

[0020] FIG. 10 illustrates an example measurement arrangement for orthogonal polarizations tests.

[0021] FIG. 11 is a graph illustrating power transmittance of polarization states and a fixed transverse electric (TE) input to the polarization controller.

[0022] FIGs. 12A and 12B are graphs illustrating peak power interference.

[0023] FIG. 13A illustrates an example transmission spectra.

[0024] FIG. 13B illustrates an optical microscopy image.

[0025] FIG. 14 illustrates another example transmission spectra.

[0026] FIG. 15 illustrates an example measurement arrangement for high-speed characterization.

[0027] FIGs. 16A and 16B illustrates eye diagrams for polarization outputs.

[0028] FIG. 17 illustrates eye diagrams for polarization at multiple frequencies.

[0029] FIG. 18 is a table illustrating a high-speed polarization modulation signal response performance.

[0030] FIG. 19 is a table illustrating a high-speed polarization modulation signal response performance.

[0031] FIGs. 20A and 20B are graphs illustrating example bandwidth signals.

[0032] FIG. 21 illustrates Bit Error Rate Test at standard modulation speeds for communications.

[0033] FIG. 22 illustrates a block diagram of an example full stokes polarization controller.DETAILED DESCRIPTION

[0034] The following discussion omits or only briefly describes conventional features of polarization control systems that are apparent to those skilled in the art. It is noted that various embodiments are described in detail with reference to the drawings, in which like reference numerals represent like parts and assemblies throughout the several views. Reference to various embodiments does not limit the scope of the claims attached hereto. Additionally, any examples set forth in this specification are intended to be non-limiting and merely set forth some of the many possible embodiments for the appended claims. Further, particular features described herein can be used in combination with other described features in each of the various possible combinations and permutations.

[0035] Unless otherwise specifically defined herein, all terms are to be given their broadest reasonable interpretation including meanings implied from the specification as well as meanings understood by those skilled in the art and / or as defined in dictionaries, treatises, etc. It is noted that, as used in the specification and the appended claims, the singular forms “a,” “an” and “the” include plural referents unless otherwise specified, and that the terms “includes” and / or “including,” when used in this specification, specify the presence of stated features, elements, and / or components, but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and / or groups thereof.

[0036] Relative terms such as “horizontal,” “vertical,” “up,” “down,” “top,” and “bottom” as well as derivatives thereof (e.g., “horizontally,” “downwardly,” “upwardly,” etc.) should beconstrued to refer to the orientation as then described or as shown in the drawing figure under discussion. These relative terms are for convenience of description and normally are not intended to require a particular orientation. Terms including “inwardly” versus “outwardly,” “longitudinal” versus “lateral” and the like are to be interpreted relative to one another or relative to an axis of elongation, or an axis or center of rotation, as appropriate. Terms concerning attachments, coupling and the like, such as “connected” and “interconnected,” refer to a relationship wherein structures are secured or attached to one another either directly or indirectly through intervening structures, as well as both movable or rigid attachments or relationships, unless expressly described otherwise. The term “operatively or operably connected” is such an attachment, coupling or connection that allows the pertinent structures to operate as intended by virtue of that relationship.

[0037] Reference throughout the specification to “one embodiment”, “an embodiment” or “some embodiments” means that a particular feature, structure, or characteristic described in connection with an embodiment is included in at least one embodiment of the subject matter disclosed. Thus, the appearance of the phrases “in one embodiment”, “in an embodiment” or “in some embodiments” in various places throughout the specification is not necessarily referring to the same embodiment. Further, the particular features, structures or characteristics of “one embodiment”, “an embodiment” or “some embodiments” may be combined in any suitable manner with each other to form additional embodiments of such combinations. It is intended that embodiments of the disclosed subject matter cover modifications and variations thereof. Terms such as “first,” “second,” “third,” etc., merely identify one of a number of portions, components, steps, operations, functions, and / or points of reference as disclosed herein, and likewise do not necessarily limit embodiments of the present disclosure to any particular configuration or orientation.

[0038] Moreover, throughout this disclosure, various aspects of the invention can be presented in a range format. It should be understood that the description in range format is merely for convenience and brevity and should not be construed as an inflexible limitation on the scope of the invention. Accordingly, the description of a range should be considered to have specifically disclosed all the possible subranges as well as individual numerical values within that range. For example, description of a range such as from 1 to 6 should be considered to have specifically disclosed subranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3to 6, etc., as well as individual numbers within that range, for example, 1 , 2, 2.7, 3, 4, 5, 5.3, 6, and any whole and partial increments there between. This applies regardless of the breadth of the range. As used herein, the term “about” in reference to a measurable value, such as an amount, a temporal duration, and the like, is meant to encompass variations of plus or minus 20%, plus or minus 10%, plus or minus 5%, plus or minus 1%, and plus or minus 0.1% of the specified value, as such variations are appropriate.

[0039] The terms “proximal,” “distal,” “anterior,” “posterior,” “medial,” “lateral,” “superior,” and “inferior” are defined by their standard usage indicating a directional term of reference. For example, “proximal” refers to a position that is situated nearer to the center of a body or point of attachment, while “distal” refers to a position that is situated away from the center of the body or point of attachment. In another example, “anterior” refers to the front of a body or structure, while “posterior” refers to the rear of a body or structure. In another example, “medial” refers to the direction towards the midline of a body or structure, and “lateral” refers to the direction away from the midline of a body or structure. In some examples, “lateral” or “laterally” may refer to any sideways direction. In another example, “superior” refers to the top of a body or structure, while “inferior” refers to the bottom of a body or structure. It should be understood, however, that the directional term of reference may be interpreted within the context of a specific body or structure, such that a directional term referring to a location in the context of the reference body or structure may remain consistent as the orientation of the body or structure changes.

[0040] In free space propagation, polarization can be controlled using waveplates and birefringent elements with precise thickness. An electromagnetic wave propagates through a waveplate in both its principal and ordinary axes at different phase velocities, causing the combined output to have a different polarization state. However, in ordinary waveguides, the two polarization states propagate over modes with a large birefringence and, consequently, significantly different phase velocities. As such, when a composite state exists in a waveguide (e.g., TE (|s» + TM ( |p))), the combined polarization changes between diagonal and circular every few microns.

[0041] In early developments, electro-optic effects were used to create waveguide-based polarization controllers with low birefringence. In such systems, waveguides are designed at their degenerate state (i.e., where the effective indices nrE and nrM are equal), and a hybrid is forced by applying a voltage across the semiconductor. Typical modulators using these systemsmay be designed in InGaAs and AlGaAs to reach modulations bandwidths exceeding 40GHz. However, these designs generally have large dimensions (e.g., >15 mm) and modulate at large voltages (50V). Further, for the cases in which modulators use silicon, the small Pockels coefficient (e.g., < 0.3 pm / V) requires different and more complex design approaches.

[0042] Alternatively, passive mode conversion may be used to achieve active polarization rotation. For instance, passive mode conversion may occur between the transverse magnetic TM mode and the first odd TE mode (TE1 or |1)), then splitting between in-phase and out-of-phase waveguides, followed by an active switching between the TE1 and TEO modes (|0)) modes using controlled phase shifting between the separate waveguides. To allow the initial transfer of energy between the TM and TE1 modes, the hybrid polarization mode may be generated using vertical asymmetries and waveguide dimensions chosen to ensure mode conversion in a small size while minimizing reflections. Furthermore, the |0) <— > |1) hybrid mode conversion (TEO <— > TE1) may be carried out with high efficiency using an inverse-designed multimode y-combiner controller and used for active mode switching. Conversion may be achieved in the y-combiner design by employing a phase shift of it between back-to-back connected junction inputs.

[0043] The combined mode conversion and splitting acts as a Polarization Beam Splitter and Rotator (PBSR). This approach may typically be used in silicon-integrated devices to build a passive polarization PBSR, in which the phase delay is induced using a waveguide delay line. Further, this approach may be used to achieve active polarization control using compact silicon structures with a PBSR device size reaching 90 im. The base component for this controller, the PBSR, uses, for instance, directional couplers and is enhanced by bent waveguides and subwavelength gratings. Although not used for active control, similar PBSR structures use inverse design techniques to reach smaller dimensions, such as 17 [im via multilevel silicon etching and 10 / rm for air-cladded devices.

[0044] Typically, for proposed hybrid mode schemes, mode conversion is carried out using two separate components. First, the TM mode is converted into TE1, and then the wave is evenly split between the two outputs where it propagates as in-phase TEo.o and out-of-phase TEo, n across two outputs. However, for the embodiments discussed herein, the inverse design eliminates the need to enforce a unique behavior in separate devices and allows both processes to occur simultaneously. Accordingly, the embodiments discussed herein provide a compact device that continuously implements the transformation. Thus, the embodiments provided herein areless susceptible to mode mismatches, resulting in lower reflections and scattering losses. When the PBSR is used as a combiner, the polarization state at its output is a function of the phase between each (|0) input. A combination of thermal-optic and plasma-dispersion phase shifting allow for low speed fine tuning and high-speed switching, respectively of the polarization conversion process.

[0045] Thus, as photonic circuits become more densely packed, it is essential to design smaller- sized devices without sacrificing their performance. The embodiments herein provide a novel integrated optical system for high-speed polarization control. In one or more cases, the integrated optical system for high-speed polarization control uses devices designed by the adjoint method for shape optimization. As such, the embodiments discussed herein relate to the integration of an inverse-designed ultra-compact polarization beam splitter-rotator (PBSR) with a high-speed PN junction phase shifter within a silicon photonics platform. The combination forms a plasma dispersion-based modulator integrated into a Mach-Zehnder interferometer (MZI), which facilitates rapid changes in output polarization states. That is, the plasma dispersion-based modulator integrated into the MZI layout enables rapid adjustments to the output polarization at high speeds. In one or more cases, the high-speed polarization modulator employs inverse- designed active polarization-sensitive y-junctions integrated at the input and output of the MZI. The y-junctions may perform passive mode conversion of the TE (|s)) and TM (|p)) modes into phase-symmetric TEo.o and asymmetric TEO. TT modes at its two outputs, respectively. The reverse process may occur when the y-junction is excited by symmetric and asymmetric modes in its output, thus enabling its use in polarization control. Further, the embodiments discussed herein demonstrate the performance of the polarization control optical circuit at 32 Gbit / s at 1550 nm wavelength, illustrating the ability for advanced high-speed polarization control applications.Further, the embodiments discussed herein demonstrate an ultra-compact device (e.g., sized at 17 im) with an insertion loss of < 0.5 dB and an extinction ratio > 20 dB. Embodiments of the highspeed performance polarization control systems are described below with reference to the Figures.

[0046] FIG. 1 illustrates a block diagram of the example polarization control system 100 (hereinafter “system 100”). FIG. 2 illustrates a graphical representation of the system 100. In one or more cases, the system 100 includes one or more of an optical circuit 101, a modulator 110, the PBRS 116, and an output waveguide taper 118.

[0047] The optical circuit 101 may be, for example, a test optical circuit. The optical circuit 101 may include an input waveguide taper 102, a beam splitter 104, and an interferometer 103. The input waveguide taper 102 may be operably connected to the beam splitter 104. The input waveguide taper 102 may be configured for input TE polarization |s). The beam splitter 104 may be, for example, a waveguide 50:50 beam splitter. The beam splitter 104 may form branches that define the interferometer 103. One branch of the interferometer 103 may include a thermal phase modulator 106. The thermal phase modulator 106 may be a thermal heater configured to thermally tune, for example, a phase of the signal. Another branch of the interferometer 103 may include a matching waveguide section 108. For example, the matching waveguide section 108 may be, for example, a dispersive matching waveguide.

[0048] In one or more cases, the optical circuit 101 of the polarization switch includes an input waveguide taper operably connected to the modulator 110 forming branches of the interferometer 103. The modulator 110 may be a 50:50 y-splitter. In one or more cases, the modulator 110 may be, for example, but not limited to, a travelling wave plasma dispersion modulator. One branch may include a thermal heater, allowing for the thermal tuning of the controller’s response. The modulator 110 may convert the output of each of the branches of the interferometer 103 into a rib waveguide doped to form a PN junction at the center of the waveguide’s core. For instance, the modulator 110 may include a positive bias active phase shifter 112 and a negative bias active phase shifter 114. The active phase shifter 112 may be a PN-doped layer configured to receive output from the thermal phase modulator 106. The active phase shifter 114 may be a NP-doped layer configured to receive output from the matching waveguide section 108. The PN junction may be extended for 4 mm, or about 4 mm, after which the two waveguides are transformed back into a ridge cross-section. The inverse-designed PBRS 116 may act as a combiner to reconnect the two branches and provide the output to the output waveguide taper 118. The output waveguide taper 118 may be configured for output TE or TM polarization |s) or |p). The output waveguide taper 118 may be configured to transmit the signal off-chip. The system 100 may be configured to function as either a mode controller, such as, a switch or a mode controller configured to provide precise control over the output mode excitation when the input mode is fixed. Precise mode control may be used in advanced communication modulation techniques, along with photonic computing systems in which information operates in an optical mode distribution. The waveguides may be designed tosupport fundamental mode transmission. Tn some cases, the waveguides may be designed to support fundamental mode transmission, except for specific regions of the devices where multiple modes get excited during the conversion processes. In one or more cases, the polarization controller may be tested for the |s) input (TE).

[0049] As such, the system 100 may be a polarization controller based on an inverse-designed polarization rotator and beam splitter with connected junction outputs for TE and TM polarizations. The y-junctions of the system 100 may be integrated at the inputs and outputs of the modulator 110, which may be a high-speed MZI based plasma dispersion modulator, to achieve ultrafast polarization conversion. The PBRS 116 may perform passive mode conversion of the TE (|s)) and TM (|p)) modes into phase-symmetric TEo.o and asymmetric TEo, modes at its two outputs, as illustrated in FIG. 6. The reverse process may occur when the y-junction is excited by symmetric and asymmetric modes in its output, thus enabling its use in polarization control. Using a high-speed phase shifter, efficient mixing and conversion between the polarization modes are achieved at the output of the MZI to any arbitrary level.

[0050] An intermediate phase state projected on the TEo.o and TEo,^ basis may be transformed into a corresponding mixed TE / TM state. As described herein, to achieve polarization mode conversion and rotation at the output of the MZI, the relative phase between the MZI’s arms may be adjusted using a combination of thermal-optic and plasma-dispersion phase shifting which enables low and high-speed tuning. That is, the MZI is configured to control the relative phase between the two intermediate waveguide branches. The MZI may, for example, be integrated in a silicon photonics platform and use high-speed PN-junction phase shifter in a travelling wave modulation configuration, to form plasma a dispersion-based high-speed controller.

[0051] In some cases, the system 100 is fabricated with a CMOS-compatible foundry process. The system 100 may operate in the C-band. In some cases, the PBRS 116 is compact (e.g., 17 Um) and designed using adjoint-based optimization. In one or more examples, the system 100 may include a 3-dB bandwidth at 10.2 GHz and have a 20 Gbit / s transmission operation. In one or more cases, the waveguides of the system 100 may be configured to support fundamental mode transmission. In some cases, the waveguides of the system 100 may not support fundamental mode transmission in regions where multiple modes are excited during the conversion processes. In some cases, the system 100 may be configured to use the TE (|s)) state as input.

[0052] FIG. 3 A is a schematic view of the system 100. FIG. 3B is a side view of a portion of the system 100. In one or more cases, the system 100 employs a 220 nm high Silicon on Insulator (SOI) platform 306 as illustrated in FIG. 3B with a 90 nm rib. The edges of both the core and ribs may be symmetric across the direction of propagation. The edges of both the core and ribs may be constrained such that the rib edge position is at or further away than the core edge (i.e., from the center of the waveguide). The input waveguide of the PBRS 116 may be a 650 nm wide ridge and the output waveguide of the system 100 may include two 500 nm waveguides with a 350 nm gap between them. A spline with 19 control points may be positioned on the edges of each layer in the PBSR 116. In one or more cases, the first and last control points of the spline may be fixed to the input and output outer widths, defining a total of 17 optimizable parameters per layer. Each parameter may correspond to the y-coordinate of the spline control point, and the x-coordinates may be homogeneously distributed to have a constant 1 zm step size between them to ensure smooth transitions along the device.

[0053] The system 100 and PBRS 116 were manufactured using the Advanced Microsystem Foundry (AMF). The technology uses 193 nm UV lithography on a 220 nm thick SOI platform 306. The platform 306 may be disposed on a silicon oxide layer 304 and a Silicon substrate layer 302. The multiple levels of silicon may be etched from a single crystalline layer, such that a shape of the PBRS 116 is inverse designed for each layer as a multi-objective optimization process. Multiple etch levels of the SOI platform 306 allow for structure with heights of 90, 150, and 220 nm. The technology includes N and P doping layers for manufacturing active elements manufacturing, Germanium deposition, as well as two layers of metallization for connections and an extra one for heaters manufacturing.

[0054] FIG. 4A illustrates an example Scanning Electron Microscopy (SEM) micrograph of the example PBRS 116. FIG. 4B illustrates AFM of the PBRS 116. FIGs. 5A-5D illustrate scanning electron micrographs of the PBRS 116.

[0055] The AFM resolution of the side ribs has a width of 210 nm at the widest point of the side ribs. The AFM resolution of the side ribs may be limited by the tip radius of 20 nm and the sidewall angle of 18 degrees during AFM scanning. The wider region of the PBRS 116 (i.e., region A highlighted in FIGs. 4A and 5A) may be used to benchmark the fabrication, as it is subject to a higher manufacturing variation when a homogeneous change affects the photonic structures. The design is 1715 nm wide at its widest part, but the fabricated PBSR 116 is 1687nm, resulting in a 28 nm error. The measured structure of the PBSR 1 16 at the waist is 1275 nm, below the resolution of SEM-based measurements, while the design is 1272 nm wide. At the port connection, the two waveguides are separated by 350 nm in the design and are separated by 360 nm in the fabricated structure. This difference is typically associated with proximity effects between the two waveguides, which is not a concern for the stem of the PBRS 116. The device fabrication variations are generally minor. However, the variations exceed the tolerances used during design in some regions, set as random homogenous variations in the range of 0-20 nm. Specifically, region D highlighted in FIGs. 4A and 5D may contain 150 wide ribs that are not resolved in the 193nm-UV fabricated structure. Despite the significant variation of the rib dimensions, during optimization, the system has fabrication thresholds (i.e., set as random variations in the geometry at each iteration), indicating that the obtained structure can withstand imprecisions in the manufacturing processes.

[0056] To optimize the PBRS 116, the following figure of the merit function is used and given by:FOM = fTE+ fTM(1) where fTEand fTMare the independent figures of merit of the |s) and p) polarizations. Each of these figures of merits may include coupling to both the symmetric and antisymmetric TE outputs in the output. This gives an optimization function in the form of:where gtare normalization constants and a^ndenotes the complex mode coupling coefficient of the mthmode at the nthwaveguide port in the forward (+) direction (transmission) or backward (-) direction (reflection), for the |0) input,for the |1). The coefficientsand (3*n nare calculated from the power overlap of the electric field profile and the mode distribution at the port, integrated over the design bandwidth. This optimization includes two forward simulations and four adjoint simulations to extract the permittivity gradients in each step fully. Given the device symmetries, this problem can be simplified. First, using the symmetry ofthe output waveguides, a super mode at the output may be described as including both waveguides together. In this configuration, for the first output mode TE0 0, light is coupled into both waveguides with identical phase, and for the second output mode, TEO n, light is coupled to both waveguides as well, but with a phase of it between them, as represented in the onsets 602 and 604 of FIG. 6. Onsets 602 and 604 correspond to input |00) and |07t) modes, respectively. As such, two forward solutions are solved in each optimization step (i.e., |s) and \p)) and two adjoints per forward solution (i.e., one for each of the outputs). By describing the design as a maximization of the power coupling between the modes of interest, the other components of the eigenspace may be neglected to simplify the problem further. The resulting simplified FOM is given by:where A* and B are the calculated output mode overlaps for the |s) input (first forward solution) and the \p) input (second forward solution), respectively, measured for the nthsuper mode at the output port. Here A* is equivalent to the power transmission from an input |s) to the output |00), andthe transmission from an input \p) to the output |Ojr). In this expression, the negative contribution of crosstalk to the figure of merit is not included, and yet it still converges correctly to the solution of interest. This occurs because the optimization involves a maximization problem in a system where the total energy is fixed (i.e., increasing the power in one mode necessarily decreases it in the others). Such reduction may only be implemented when slow transitions guarantee no coupling to scattering modes and when the modes used span the space of possible supported photonic modes. In this final form, the problem needs the numerical solution of only two forward and two adjoint problems (i.e., one per forward solution).

[0057] FIG. 7 illustrates beam propagation simulations of the optimized PBRS. FIG. 7 illustrates the principal components of the electric field intensity for the |s) (TE) and \p} ™ inputs at X = 1550 nm. The simulations illustrate an effective mode conversion. As illustrated, the PBRS generates a mode conversion. The PBRS initiates polarization rotation first, and subsequently, a wider multimode interference (MMI) section then appropriately splits the two output beams. It is noted that other conversion modes to split the output beams, such as, but not limited to, MMI, adiabatic transformations, mode matching, coupling, and the like.

[0058] The low-speed performance (i.e., non-modulated (DC) performance) of the device was analyzed. In particular, insertion losses and mode conversion of the device were analyzed.

[0059] To measure the device’s insertion losses, a series of PBRS 800 were measured for different numbers of devices being connected one after another. FIG. 8A is a graph illustrating the calculated insertion loss from transmittance measurements, and FIG. 8B illustrates a layout schematic of the series of PBRS 800 used to measure the device’s insertion loss. The overall loss is determined as the slope of the loss versus the number of devices connected together, subtracted by 3 dB (i.e., where only one branch is connected). This measurement assumes that a balanced splitting is achieved, which is expected from the axial symmetry of the device. FIG. 8A illustrates the resulting insertion loss (IL). Overall, the IL remains below 1 dB across the C- and L- bands and below 0.5 dB across the C Band ([1535 nm - 1565 nm]). During testing, the input polarization state was scrambled using a polarization controller.

[0060] To analyze mode conversion, an input heater (e.g., such as the thermos-optic heater in the chip) was used to adjust the phase shift between the branches of the MZI connecting the PBRS. To adjust the phase shift, a voltage was applied to the input heater using a Source Meter Unit (SMU - Keysight B2900A). FIG. 9A illustrates polarization modulation using a thermo-optic control. FIG. 9B illustrates a Stoke’s sphere corresponding to the polarization modulation of FIG. 9A. As such, FIGs. 9A and 9B illustrate the modulation of the Stoke’s parameters at multiple applied heater voltages. A quasi-orthogonal polarization state is reached at a power of Pn= 66.8mW, measured at the minimum of the SI parameter. The results demonstrate that the device can effectively control the polarization state of the output at slow speeds, limited by the thermal dissipation dynamics of the substrate.

[0061] As observed, the output polarization efficiently rotates between orthogonal states. After a full rotation, at a drive power of P = 139mV, there is a mismatch in the polarization of 0.314rad, equivalent to a polarization crosstalk of 0.7 dB (approximated by XT — -10 • log Q(cos(jt -(p ), with <p being the angle between the two polarization states. This mismatch may arise from crosstalk in the device. The crosstalk may be related to the parasitic coupling of a small TM input, coupling of scattered light, or the inherent crosstalk of the device. If a precise orthogonal state is needed, the output completes a full transformation between |s) and \p) at P = 203mVF, with an error of only 23 x 10-3rad, which is equivalent to a polarization cross talk of 1.5 mdB. However, for polarization control applications and switch applications, it is important to have theability to convert the polarization between any of the states along the path, specifically between |s) and \p) *, that is, the quasi -orthogonal state. Upon detection, the light is typically separated into two orthogonal components using a polarizing beam splitter (PBS). Any error in the rotation may result in an optical power crosstalk in the opposite output with no further impact.

[0062] Given that the optical circuit is configured as an imbalanced MZI, the transmission spectra for each polarization state may follow a sinusoidal pattern. By projecting the output onto orthogonal states using a free space PBS, the optical power transmission is measured to determine the polarization extinction ratio (PER). The PER is equivalent to the interference peaks’ extinction ratio. After the PBS, since the optical circuit is configured as an imbalanced MZI, the power transmission for each polarization state follows a sinusoidal pattern with an FSR of AZFSR = A2 / (ncAL), with AL the difference in path lengths between the two MZI branches an no the group index of the |0) mode in the waveguide. As illustrated in FIG. 10, to measure each polarization state, a continuous wave (CW) laser at 1550 nm with its polarization set to TE was coupled to the device under test (DUT) using lensed fibers (LF). The light is guided out of the fibers using a collimator lens, and transmitted through an Iceland spar (calcite) crystal in a Gian- Taylor configuration (Thorlabs GT10). The crystal acts as a high-quality polarizing beam splitter (PBS) with a nominal extinction ratio of 50 dB at 1550 nm. The PBS crystal is mounted on a rotation mount to select the filtered polarization, and pairs of mirrors are used to efficiently align the beam in and out of the collimators. A manual fiber polarization controller is used to correct the fiber polarization dispersion, and at the input, a polarization synthesizer maintains the polarization state constant across the spectrum. A tunable laser source (Keysight 8164B) is swept across the bands of interest to measure the spectral response.

[0063] As was observed in the Stoke’s plots, the polarization can be fully rotated between opposing states. This appears as a shift of the interference peaks in the transmission spectrum, as shown in Fig. 11. The two functions are complementary. Most importantly, in this graph, it can be observed that several interference peaks reach an extinction ratio (ER) beyond 40 dB (averaging >25 dB across the C-Band). Across the spectrum, orthogonal polarizations may exhibit very high extinction ratios, approaching the detection limit of 50dB (ER of the PBS). The polarization dispersion within the optical fibers of the test setup generates components that need to be corrected every 10 nm in bandwidth. To assess the maximum extinction ratio of the device across the complete C- and L- bands, fiber dispersion is compensated for at each interferencepeak. Doing so for both polarization outputs results in the spectra plotted in FIGs. 12A and 12B, in which FIG. 12A illustrates peak power interference for the |s) state across multiple measurements and FIG. 12B illustrates peak power interference for the \p) state across multiple measurements. The nominal extinction ratio, taken as the peak ER measured, is, in general, above 30 dB for the \p) polarization state and above 40 dB for the |s) state. At 1550 nm, the PER may be set close to 50 dB in either mode by tuning the projector’s polarization controller.

[0064] In wavelength, the actuation of the thermal tuning of the device appears as a shift in the interference peaks, as illustrated in FIG. 13. FIG. 13 illustrates a transmission spectra for multiple DC powers applied to the waveguide thermal actuator. That is, FIG. 13 illustrates a recording the spectral response of the system using a tunable laser source and applying a varying voltage to the waveguide heater.

[0065] The high-speed performance of the device was analyzed. In particular, a DC modulator / gate response, high-speed modulation, bandwidth, and bit error rate (BER) performance of the device were analyzed.

[0066] With respect to DC modulator / gate response, the high-speed polarization switch uses a traveling wave Mach-Zehnder phase modulator (TWMZM). Each of the branches of the MZI induces a phase shift in the propagating wave. When driven over a Vn range, the phase shift between the inputs of the PBRS enables a complete rotation of the polarization state. The phase shift that enables a complete rotation of the polarization state may be similar to the effect presented with the heater controller, but uses free carrier dispersion as the governing phenomena. The TWMZM configurations allow the microwave and optical signal to have a matched group velocity for high-bandwidth operation while the voltage changes the index of refraction in the PN junction. This electro-optic setup is identical to configurations demonstrated for amplitude modulation in the same technology node using adiabatic combiners or y-splitters at the interferometer’s output and has been reported with bandwidths exceeding 20 GHz. FIG. 13B illustrates an optical microscopy image of the device. To test the optoelectronic response, the setup described in FIG. 10 is used, and additionally, the SMU is connected to the TWMZM driver. FIG. 14 illustrates a recorded high-resolution transmission spectrum, tracking the shift of the interference peaks around 1550 nm. To generate the transmission spectrum, the DC voltage applied to the TWMZM was varied between 0V and 5V.

[0067] The measurement was limited to 5V to ensure that the DC current would not exceed 100 mA, thereby avoiding damage to the junction. The wavelength shifts linearly for voltages above 0.4 V. The linear shift may be equivalent to a linear phase change. Extrapolating using the recorded data, the voltage at which the device shows a phase shift of = n is found to be Vn= 7.687. For RF (radio frequency) operation, the junction may be biased at any point along the resulting linear region. The wavelength is selected to generate good optical amplitude modulation for both polarization components. The heater tuner may be used to shift the spectral response when a fixed wavelength laser is used. The measurements carried use a tunable laser source to more accurately control the operation point. Correctly biasing the modulator may generate an optical modulation depth of approximately 15 dB for a 1 Vpp RF small signal.

[0068] With respect to high-speed modulation, to evaluate the polarization controller’s highspeed optoelectronic response, a high-speed arbitrary wave-function generator (AWG) is used to test the system. FIG. 15 illustrates the measurement setup for High-Speed characterization. As illustrated in FIG. 15, the TWMZM modulator is driven by the AWG signal, which is amplified and connected to an external bias-T before being connected to the modulator. The bias-T is driven using a Source Measurement Unit (SMU) to apply a DC bias to the PN junction, while the AWG generates only the RF signal. All components are impedance-matched and have a 3dB bandwidth over 12 GHz.

[0069] The circuit was tested for multiple modulation speeds using non-return to zero (NRZ) amplitude modulation of a 1024-bit long-bit stream from a pseudo-random binary sequence generator (PRBS). The signal was fitted as a raised cosine function at the AWG. Eye diagrams for both states of polarization were recorded and are illustrated in FIGs. 16A and 16B. Eye diagrams for the |s) state at multiple frequencies are illustrated in FIG. 17.

[0070] As illustrated in FIG. 18, the system displays a signal-to-noise ratio (SNR) surpassing 8dB at 12 GHz. This level is suitable for fast communication and polarization-based gates. Considering the high extinction ratio observed for each polarization state, the system may be further improved.

[0071] With respect to bandwidth, the PN semiconductor layer has an intrinsic bandwidth governed by the junction capacitance and series resistance of the p and n layers, which has been reported as 30 GHz. To assess the performance of the fabricated TWMZM, its frequency response was measured using a Vector Network Analyzer (Keysight PNA). To perform theassessment, the same setup as described FIG. 15 was used, and the analyzer’s ports replaced the AWG and OSC. The recorded spectrum is illustrated in FIGs. 20A and 20B, which is measured for both output polarization states. The 3dB bandwidth of the systems is 10.2 GHz and 10.5 GHz for the |s) and \p) states, respectively. The measured bandwidth performance may still be improved to reach the reported foundry values by further optimizing the PN junction design, primarily by extending its length.

[0072] With respect to BER performance, experiments were run at a single bias voltage for multiple frequencies and for the |s) output polarization state. To analyze BER performance, the same setup as described FIG. 15 was used, except for replacing the AWG for a sequence generator from a BER testing system (Keysight M8040A) and the OSC for the test unit’s input. The clock and triggers are recycled from the unit itself and reported BER values were measured over 30 minutes each or 1014bits, whichever was shorter. FIG. 21 illustrates BER Test at standard modulation speeds for communications. When testing NRZ signals at data rates below 25 GBaud, the gate generated no detectable errors in sequences over 1014long. At 41.25 GBaud (40Gb Ethernet base), the system has a BER of 1.8 X 10-9, while at 53.12 GBaud (400GB ASE), it has a BER of 1.24 x 10-5. Specifically for optical communications, at 28.05 GBaud and 28.9 GBaud (base for 32G and 64G fiber optic communications), the measured NRZ BER was 3.32 x 10-12and 5.77 x 10-12, respectively. PAM modulation at the same speed show errors in the order of 1 x 10-3. Overall the performance confirms the ability to utilize the polarization controller in applications at 30GHz for long-range communications. Moreover, the lower BER rate attained at 40GBaud in the range of 1 x 10-9is still usable for short-distance communications.

[0073] The system was further tested for PAM4 modulation at 12.25 GBaud (Base for STM-64) and 14.025 GBaud (Base for 16G Fibre). The measured BER showed error rates of 1 x 10-13and 2.33 x 10-12, respectively. Consequently, the device shows the capability for multi-symbol transformations at this reduced channel capacity, offering the ability to use the polarization switch in spaces with four symbols for more complex encoding techniques.

[0074] In one or more cases, the embodiments discussed herein provide a compact (17 pm) 1550 nm wavelength inverse-designed hybrid mode y-junction polarization beam splitter rotator on the SOI platform. The device’s minimum critical dimension is fully compatible with 193 nm UV lithography. Measurements demonstrate low insertion loss (IL) (< 0.5dB) and high extinctionratio (ER) (> 20dB). The designed splitter is further employed to construct a high-speed polarization switch. Operational experiments successfully modulate the polarization state of a continuous wave laser, revealing open-eye diagrams at 32 Gbit / s across the C-Band. The speed of the controller being, however, constrained by the PN-junction bandwidth. Additionally, experiments showed excellent performance with low BER up to 40 Gbit / s, confirming the versatility of the device for polarization-diverse high-speed communications based on the silicon photonics platform. The device high polarization extinction ratio and high-speed operation offer a promising platform for on-the-fly operations on information encoded in the polarization state of light, which in combination with its <1 dB losses across the C-Band makes it a promising platform for applications like polarization encryption and single-photon quantum key distribution.

[0075] FIG. 22 illustrates a block diagram of a full stokes polarization controller. In one or more cases, when an input splitter is replaced by the polarization rotator splitter (PRS), the system may be used as a high-speed polarization converter. That is, the system may be configured to convert any input polarization state into any real polarization state at the output. In one or more cases, by adding a polarization splitter and a phase delay in any one of the separated waveguides and recombining the waveguides using a second polarization splitter constructs a full stokes polarization controller. The second polarization splitter may be used as a combiner. The full stokes polarization controller may be configured to generate any complex polarization state at its output, either in phase (e g., linear or diagonal polarization phases) or circular states.

[0076] Moreover, this application incorporates, by reference and in its entirety, J. E. Villegas and M. Rasras, "A Compact Mode Controller (Switch) Using an Inverse-Designed 2-Mode ¥- Junction," in Journal of Lightwave Technology., vol. 42, no. 6, pp. 2083-2089, March 15, 2024, doi: 10.1109 / JLT.2023.3327120; and J. E. Villegas and M. Rasras, "A High-Speed Compact Polarization Controller in Silicon Photonics,?2024 Optical Fiber Communications Conference and Exhibition, March 25, 2024.

[0077] The various embodiments described above are provided by way of illustration only and should not be construed to limit the claims attached hereto. Those skilled in the art will readily recognize various modifications and changes that may be made without following the example embodiments and applications illustrated and described herein, and without departing from the spirit and scope of the following claims.

Claims

CLAIMSWhat is claimed is:

1. A polarization controller, comprising an optical circuit configured to split light into a first beam of light and a second beam of light and provide the first beam and second beam of light to a Mach-Zehnder Interferometer (MZI); the MZI configured to transmit with controlled phase delayed between the first beam and the second beam via waveguides with embedded PN junctions forming a phase modulator; and an inverse-designed polarization beam splitter-rotator (PBSR) configured to combine the first beam and second beam and provide the combined first beam and second beam to an output waveguide and waveguide taper.

2. The polarization controller of claim 1, wherein the optical circuit comprises an input waveguide taper operably connected to a beam splitter, wherein the beam splitter is configured to provide the first beam and the second beam to the MZI via a respective first branch and second branch.

3. The polarization controller of claim 2, wherein the first branch comprises a thermal phase modulator configured to tune a phase of the first beam, and the second branch comprises a dispersive matching waveguide.

4. The polarization controller of claim 2, wherein the first branch and the second branch each comprise waveguides of equal length with a P-doped section and a N-doped section forming a PN semiconductor junction used as a phase shifter.

5. The polarization controller of claim 2, wherein the input waveguide taper is configured to receive a transverse electric polarized light as an input.

6. The polarization controller of claim 1, wherein the MZI comprises a travelling wave plasma dispersion modulator.

7. The polarization controller of claim 1, wherein the output waveguide taper is configured to transmit a signal from the polarization controller.

8. The polarization controller of claim 1, wherein the PBSR comprises a polarization converter and splitter using inverse-design and adjoint method optimization.

9. The polarization controller of claim 8, wherein the PBRS comprises connected junction outputs for transverse electric polarized light and transverse magnetic polarized light.

10. The polarization controller of claim 1 is configured to operate in a C-band.

11. The polarization controller of claim 9, wherein the PBSR comprises multiple levels of Silicon etched from a single crystalline layer.

12. The polarization controller of claim 11, wherein a shape of the PBSR is inverse designed for each layer as a multi-objective optimization process.

Images