Micro-ring modulators and systems including phase change materials for athermal operation

Abstract

Examples of systems and methods described herein include waveguide systems. A waveguide system may include a phase-change material and an oxide cladding. The phase-change material, the oxide cladding, or both may have a negative thermos-optic coefficient. Examples of waveguide systems described herein may accordingly provide athermal, or predetermined, thermal behavior. Examples of waveguide systems described herein may be controlled and / or programmed using a diode included in the waveguide system.

Description

Docket No. 0077145-08201MICRO-RING MODULATORS AND SYSTEMS INCLUDING PHASE CHANGE MATERIALS FOR ATHERMAL OPERATION CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit under 35 U.S.C. § 119 of the earlier filing date of U.S. Provisional Application Serial No. 63 / 778,273, filed March 26, 2025, the entire contents of which is hereby incorporated by reference in its entirely for any purpose.STATEMENT REGARDING RESEARCH AND DEVELOPMENT

[0002] This invention was made with government support under Grant No. ECCS 2142996, awarded by the National Science Foundation. The government has certain rights in the invention.TECHNICAL FIELD

[0003] The present disclosure generally relates to the field of integrated photonics. Examples of micro-ring modulators are described.BACKGROUND

[0004] The increasing prevalence of compute-intensive, artificial intelligence (Al) driven workloads has created a need for hyperscale datacenter architectures capable of supporting both scale-out and scale-up deployment models. To meet these performance demands, processor platforms may evolve toward wafer-scale or panel-scale heterogeneously integrated (HI) systems that incorporate multiple reticle-sized GPU or accelerator chiplets in close proximity' to multi-teraby te-class high-bandwidth memory' (HBM) resources. Such integrated processing assemblies may operate at aggregate power levels on the order of kilowatts, with individual chips or packages functioning across elevated operating-temperature ranges that may approach or exceed approximately 100 °C.SUMMARY

[0005] In some embodiments, an apparatus comprises an input port configured to receive optical input energy, a modulation input configured to receive modulation data, an optical device coupled to the input port and the modulation input and comprising one or more waveguides, a phase-change material (PCM) coating at least a portion of the waveguide, and an oxide cladding, wherein the PCM or the cladding or both exhibit a negative thermo-optic coefficient, and wherein the optical device is configured to utilize the waveguide to modulate the input optical energy in accordance with the modulation data to generate modulated outputDocket No. 0077145-08201energy, and an output port coupled to the optical device and configured to provide the modulated output energy.

[0006] In some examples, the optical device comprises a micro-ring modulator. In some examples, control of the PCM material is used to configure the resonant wavelength of the optical device. In some examples, the control is applied through the modulation input. In some examples, the control is applied through separate input from the modulation input.

[0007] In some examples, the waveguide comprises silicon. In some examples, an effective thermo-optic coefficient of an optical mode in the waveguide is reduced due to the phasechange material, the cladding, or both. In some examples, the phase-change material has a thickness, based in part on a width of the waveguide, to cause the effective thermo-optic coefficient of an optical mode in the waveguide to be zero. In some examples, the oxide cladding comprises TiCh. In some examples, the oxide cladding comprises one or more layers. In some examples, the oxide cladding comprises a first material and a second material, the first material is an AI2O3, HfCh, SiN, SiCh, or combinations thereof.

[0008] In some examples, the micro-ring resonator further comprises a diode, the diode configured to program the phase-change material in a forward bias mode and to perform the modulation in a reverse bias mode. In some examples, the micro-ring resonator further comprises a first diode configured to program the phase-change material; and a second diode configured to perform the modulation. In some examples, the diode comprises a P-N junction or a P-i-N.

[0009] In some examples, the waveguide, the PCM layer, and oxide cladding are configured such that the TOC of the waveguide is substantially zero over a target temperature range. In some examples, the thicknesses and TOCs of the core material, the PCM layer, and the cladding layer are configured such that the effective thermo-optic coefficient of the optical waveguide structure is substantially zero over a target temperature range. In some examples, the apparatus further comprising a control interface configured to transition the PCM coating between at least two states to tune the phase or resonant wavelength of the optical waveguide structure.

[0010] In some examples, the optical device comprises a micro-ring resonator, and wherein the control interface is configured to shift the resonant peak of the micro-ring resonator by altering the state of the PCM layer without increasing the thermal sensitivity of the resonator. In some examples, the micro-ring modulator comprises a Mach-Zehnder Interferometer (MZI) having at least two arms, wherein at least one arm includes the PCM layer and a cladding with a negative TOC, enabling athermal phase-shifting for switching or modulation. In someDocket No. 0077145-08201examples, the one or more waveguides are arranged in a series configuration to form a Wavelength Division Multiplexing (WDM) multiplexer, wherein each structure is independently tunable via its respective PCM layer. In some examples, the one or more waveguides are arranged in a parallel configuration to form a Wavelength Division Multiplexing (WDM) demultiplexer, wherein each structure is independently tunable via its respective PCM layer.

[0011] In some embodiments, a method comprises applying a control signal to a PCM material included in an optical device to place the PCM material into a particular state, wherein either the PCM or a cladding of the optical device, or both have a negative thermooptic coefficient; and utilizing the optical device to modulate input optical energy in accordance with data while the PCM material remains persistently in the particular state. In some examples, the particular state remains persistently defined by the state of the PCM without continuous application of the control signal. In some examples, the optical device comprises a micro-ring modulator.BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Embodiments will be readily understood by the following detailed description in conjunction with the accompanying drawings. Embodiments are illustrated by way of example and not by way of limitation in the figures of the accompanying drawings.

[0013] FIG. 1 is a schematic illustration of a micro-ring modulator system arranged in accordance with examples described herein.

[0014] FIG. 2 is a graphical representation of an example ring tuning scheme arranged in accordance with example described herein.

[0015] FIG. 3A is a cross section of an example waveguide arranged in accordance with examples described herein.

[0016] FIG. 3B is a cross sections of an example waveguide arranged in accordance with examples described herein.

[0017] FIG. 4A is a cross-sectional schematic illustration of an example integrated waveguide in accordance with examples described herein.

[0018] FIG. 4B is a cross-sectional schematic illustration of an example integrated waveguide in accordance with examples described herein.

[0019] FIG. 4C is a cross-sectional schematic illustration of an example integrated waveguide in accordance with examples described herein.Docket No. 0077145-08201

[0020] FIG. 4D is a cross-sectional schematic illustration of an example integrated waveguide in accordance with examples described herein.

[0021] FIG. 5A is a schematic cross-section and top-down illustration of example waveguide geometries arranged in accordance with examples described herein.

[0022] FIG. 5B is a schematic cross-section and top-down illustration of example waveguide geometries arranged in accordance with examples described herein.

[0023] FIG. 5C is a schematic cross-section and top-down illustration of example waveguide geometries arranged in accordance with examples described herein.

[0024] FIG. 6A is a schematic illustration of a device including a micro-ring modulator arranged in accordance with examples described herein.

[0025] FIG. 6B is a schematic illustration of a device including a micro-ring modulator arranged in accordance with examples described herein.

[0026] FIG. 7 is a graph of a temperature dependent resonance wavelength shift to a waveguide width in accordance with examples herein.

[0027] FIG. 8 is a graph of a temperature dependent resonance wavelength shift to waveguide width in accordance with examples described herein.

[0028] FIG. 9A is a schematic illustrations of a waveguide device arranged in accordance with examples described herein.

[0029] FIG. 9B is a schematic illustrations of a waveguide device arranged in accordance with examples described herein.

[0030] FIG. 10 is a schematic illustration of a closed-loop tuning system arranged in accordance with examples described herein.

[0031] FIG. 11 is a schematic illustration of circuitry7to perform programming and modulation for a shared PN junction device in accordance with examples herein.DETAILED DESCRIPTION

[0032] In the following detailed description, reference is made to the accompanying drawings, which form a part hereof, and in which are shown by way of illustration embodiments that may be practiced. It is to be understood that other embodiments may be utilized and structural or logical changes may be made without departing from the scope. Therefore, the following detailed description is not to be taken in a limiting sense, and the scope of embodiments is defined by the appended claims and their equivalents.

[0033] Examples of systems and methods described herein include waveguide systems. A waveguide system may include a phase-change material and an oxide cladding. The phase-Docket No. 0077145-08201change material, the oxide cladding, or both may have a negative thermos-optic coefficient. Examples of waveguide systems described herein may accordingly provide athermal, or predetermined, thermal behavior. Examples of waveguide systems described herein may be controlled and / or programmed using a diode included in the waveguide system.

[0034] While the industry has been making progressing in building hyperscale datacenter platforms capable of supporting both scale-out and scale-up architectures, there exists a desire to achieve ultra-low-power and / or low-latency interconnects capable of delivering hundreds of terabits per second (Tb / s) for both in-package inter-die as well as off-package I / O bandwidths.

[0035] Emerging workflows (e.g., applications) such as generative artificial intelligence (Al) has created a demand for dense and ultra-low power inter / intra-rack optical communications (e.g., sub-100-meter reach) and for low-latency (e.g., +10 Tb / s) chip-to-chip interconnects. Similar needs may be desired for edge-to-cloud connectivity and distributed radar, phased-array, and / or 5G / 6G MIMO systems.

[0036] Silicon photonic (SiPh) transceivers (TRx) have been used to co-package optical TRx with high performance GPU, field-programmable gate array (FGPA), and / or system-on-chip (SoC) in a same package (e.g., co-packaged optics (CPO)).

[0037] Micro-ring modulators (MRM) have also been used to improve the energy-efficiency and shoreline (e.g., edge) bandwidth (BW) densities in terms of terabits per second per millimeter (Tb / s / mm) relative to conventional Mach-Zehnder interferometer (MZI) modulators and vertical-cavity surface-emitting lasers (VCSELs).

[0038] Presently reported results indicate support for approximately 100 Gb / s PAM-4 modulation per wavelengths, with energy efficiencies above 5 pj / b and BW densities less than 0.5 Tb / s / mm. These values are typically 5-10 times lower than the performance projected for future CPO applications.

[0039] Increasing baud rates may be associated with increased energy consumption. For example, there may be a large energy penalty in scaling up baud-rates beyond 50 GBaud. Therefore, vast parallelization degrees should be utilized to address multi-TB / s aggregate off-package data-rate demands. Wavelength division multiplexing (WDM) have been proposed and demonstrated to do so.

[0040] A dense wavelength-division multiplexing (DWDM) system may include a plurality of optical transmitters, each configured to generate optical signals at distinct wavelengths within a defined spectral band. The optical signals may be combined onto a common optical waveguide or fiber using a multiplexing element, thereby enabling multiple data channels toDocket No. 0077145-08201be conveyed simultaneously over a shared optical medium. DWDM configuration may enable high aggregate bandwidth, low-latency transmission, and scalable optical input / output architectures suitable for inter-rack, intra-rack, chip-to-chip, or other optical interconnect applications.

[0041] Despite promising advancements in CPO technologies, thermal sensitivity of photonics remains a hurdle in practical implementations of CPO. Delivering beyond lOOTb / s interconnects may necessitate parallel optical I / O enabled by DWDM (e.g., channel spacing of Inm for lasers). However, scaling to 32 or 64 wavelengths introduces thermal management and sensitivity challenges across photonics, packaging, and laser source design. For example, resonant devices such as micro-ring resonators (MRR) and MRM may be sensitive to temperature fluctuations. However, MRR and / or MRM may not be designed to be purely athermal because the initial resonance of MRRs or MRMs may be able to adjust to align with a laser wavelength due to initial process variations.

[0042] Generally , the thermal sensitivity of DWDM devices may be addressed by rely ing on thermo-optical effects and using closed-loop control loop to constantly adjust the local temperature of each device. However, this approach may introduce energy overhead (e.g., poor energy-efficiency), additional control-loop area, constraints on tuning range, and / or further timing or latency overhead in the link. Typically, DWDM optical I / O relies on compact MRM and MRR filters that may be followed by a photodiode (PD) on the drop-port. MRMs may improve the energy-efficiency compared with traditional bulky MZI modulators by an order of magnitude. Although CPO architectures for next-generation Al computing may benefit from tens of Tb / s aggregate bandwidth at sub-pJ / b energy efficiency and low-latency, their deployment may be constrained by the temperature sensitivity of resonant devices such as MRMs and MRRs, which can experience resonance-wavelength shifts on the order of 100 pm / K.

[0043] It may be desirable to exploit the physics of phase-change materials (PCMs). Exploiting the physics of PCM may allow devices to be athermal. In some examples, a device may be athermal when a resonant wavelength of the device becomes independent of temperature (e.g., a weak function of temperature) while staying tunable.

[0044] A PCM may refer to a material capable of reversibly transitioning between two or more structural phases, such as amorphous and crystalline states, in response to the application of thermal, electrical, or optical energy. Each phase may exhibit distinct optical, electrical, or thermal properties, including differences in refractive index, optical absorption, or electrical conductivity. A PCM layer or region may be integrated with a photonic deviceDocket No. 0077145-08201such that transitions between its phases modify the effective refractive index or absorption experienced by an optical mode propagating within an adjacent waveguide or resonant structure. In some embodiments, the PCM may retain its phase state without continuous power, thereby enabling non-volatile tuning or configuration of the associated photonic device. The thickness, composition, and placement of the PCM may be selected to achieve desired tuning ranges, particular thermal behavior (e.g., athermal behavior), and / or reconfigurable optical responses.

[0045] Accordingly, systems and methods described herein may receive input optical energy (e.g., data), modulate the input optical energy in accordance with the data while the PCM remains persistently in a particular state. Examples of systems and methods may be used to output energy modulated in accordance with the data.

[0046] In this way, examples of systems and methods described herein may provide a comprehensive solution for a system to operate in an athermal manner. Furthermore, nonvolatile PCM-based tuning described herein provides ability to tune MRMs at zero power overhead. This may allow DWDM optical I / O with thousands of links with sub-pJ / b energy¬ efficiency.

[0047] Certain details are set forth herein to provide an understanding of described embodiments. However, other examples may be practiced. In some instances, well-known circuits, control signals, optical components, timing protocols, materials, and / or software operations have not been shown in detail in order to avoid unnecessarily obscuring the described embodiments. Other embodiments may be utilized, and other changes may be made, without departing from the spirit or scope of the subject matter and / or claims presented here.

[0048] Turning now to FIG. 1, FIG. 1 is a schematic illustration of a micro-ring modulator system arranged in accordance with examples described herein.

[0049] System 100 may include a waveguide 102 and micro-ring modulator 108. Waveguide 102 may have input port 104 and output port 106. Micro-ring modulator 108 may include a P-N junction 110. Input port 104 may receive a modulation input 112. Output port 106 may provide a modulated output energy 114.

[0050] FIG. 1 is exemplary. Additional, fewer, and / or different components may be used in other examples. For example, while a micro-ring modulator system is shown in FIG. 1 including micro-ring modulator 108, in other examples, systems described herein may utilize an optical device other than a micro-ring modulator and / or micro-ring resonator. Other modulator and / or resonator devices may additionally or instead be used.Docket No. 0077145-08201

[0051] In some embodiments, system 100 may include a waveguide 102. A waveguide 102 may be a medium or structure that guides electromagnetic waves by using differences in refractive index or conductive boundaries to keep the wave confined. In some examples, waveguide 102 may be an optical waveguide. A waveguide 102 may have a core (not shown), which may be a longitudinally extended high-index optical medium. The core of a waveguide 102 may be made from silicon. In some examples, the core of a waveguide 102 may be silicon-based material such as silica. In some embodiments, a PCM may coat the waveguide 102. In some examples, a PCM may coat a portion of the waveguide 102. In some examples, a PCM may coat a circumference of the waveguide 102. In some embodiments, the PCM may have a negative thermo-optic coefficient (TOC). In some embodiments, a PCM may have a positive TOC. A thickness of the PCM coating may vary. In some examples, a thickness of the PCM may be provided (e.g., selected) to cause the effective TOC of an optical mode in the waveguide 102 to be zero. In some examples, when the TOC of the optical mode in the waveguide is zero it may be actually zero or it may be numerically near-zero such that the refractive index does not change substantially with temperature change across an operational range. The response of the waveguide may be considered constant with respect to temperature for the purposes of making or using the waveguide device.

[0052] In some examples, the thickness of the PCM may be provided based on a width of the waveguide 102.

[0053] A waveguide 102 may have a cladding, which may be a low-index media that traverses the core. In some examples, the cladding may be an oxide cladding, such as TiCh cladding. In some embodiments, a cladding may coat the waveguide 102 and / or the PCM. In some examples, the cladding may coat a portion of the waveguide 102 and / or the PCM. In some embodiments, the cladding may have a negative TOC. In some embodiments, the PCM and cladding may both have a negative TOC.

[0054] In some embodiments, a waveguide 102 may include an input port 104. Input port 104 may receive a modulation input 112 (e.g.. optical input energy). In some examples, a modulation input 112 may be data to be modulated by a device. The data may be, in some examples, digital (e.g., Is and 0s). A TOC of an optical mode in the waveguide may be reduced due to PCM, cladding, or a combination thereof. PCM material may be controlled to set and / or adjust a resonant wavelength of the optical device (e.g., of the MRM). In some embodiments, a PCM control may be applied through the input port 104. In some examples, a PCM control input may be similar to the modulation input 112. In some embodiments, the PCM control may be applied through a different input port from the input port 104. In some examples, the different input port may be an additional input port for the PCM control.Docket No. 0077145-08201

[0055] In some embodiments, system 100 may include a diode (e.g., laser diode). Generally, any kind or type of diode may be used. In some examples, diodes may be electrically pumped semiconductor lasers in which a gain is generated by an electric current flowing through a P-N junction or a P-i-N junction. In some embodiments, the diode may be coupled to the PCM. In some embodiments, the diode may be used to program (e.g., heat) the PCM. In some embodiments, the diode may be used to perform modulation. In some examples, the diode may be biased in a forward-bias mode for programming the PCM. In some examples, the diode may be biased in a reverse bias mode to perform modulation. In some embodiments, multiple diodes may be used. For example, one diode may be used to program the PCM and another diode may be used to perform the modulation. In some embodiments, the more the one diodes may be coupled to the PCM via different input ports. For example, one diode may be coupled to the PCM via input port 104, and the other diode may be coupled to the PCM via another input port. Generally, once a PCM state is selected (e.g., programmed), no further energy may be needed and / or required to maintain the PCM in the programmed state. In some examples, a control signal, such as a signal from a diode, to a PCM may place the PCM into a particular state.

[0056] In some embodiments, a waveguide 102 may include an output port 106. In some examples, the output port 106 may provide the modulated output energy 114 (e.g., modulated output laser light). In some examples, the output port 106 may be coupled to the optical device (e.g., the MRM 108). In this manner, an interconnect may be formed, which may receive data for modulation and provide modulated output energy. In some embodiments, the MRM may be used to modulate input optical energy in accordance with data while the PCM remains persistently in the particular state.

[0057] In some embodiments, system 100 may include a micro-ring modulator 108. A MRM 108 described herein may be utilized to form optical interconnects. In some examples, an optical interconnect may be a DWDM optical interconnect. While micro-ring modulators are described, other optical devices may be used in other examples. The optical device may include a waveguide and PCM material as described herein.

[0058] A MRM may be a type of optical modulator that uses a MRR to manipulate light signals. MRM and / or MRR may be structured as a ring-shaped structure. The modulator may modify the phase of the light signal by changing the resonance condition of the micro-ring, allowing for the creation of a variety of modulation formats, including amplitude, phase, and polarization modulation.Docket No. 0077145-08201

[0059] In some embodiments, a MRM 108 may include a P-N junction 110 formed within a portion of a waveguide that defines the optical path of the ring. The P-N junction 110 may be created by introducing p-type dopants and n-type dopants into respective regions of the waveguide or an adjacent semiconductor layer such that the doped regions meet to form a depletion region. The P-N junction 110 may be arranged such that an optical mode circulating within the ring interacts with changes in refractive index that occur when an electrical bias is applied across the junction.

[0060] In some embodiments, a MRM 108 may include a P-i-N junction. The P-i-N junction may comprise a p-doped region, an intrinsic (undoped or lightly doped) region, and an n-doped region, arranged such that the intrinsic region is positioned between the p-type and n-type regions. The junction may be oriented laterally across the waveguide, vertically relative to the waveguide core, or in another suitable geometry that permits electrical control of the refractive index within the optical mode region.

[0061] FIG. 1 further includes a graph of thru-port transmission showing variable temperature behavior. In some examples, thermal fluctuations may occur to the system 100. For example, ambient thermal fluctuations from the package in which the system 100 is wholly or partially packaged may occur. Other thermal inputs may be received, in some examples, where a temperature of an environment may change, which may change a temperature of parts or all of the system 100. Absent techniques used to address these thermal fluctuations, the thermal fluctuations may affect performance of the system 100. For example, the graph shown in FIG. 1 depicts energy transmitted from the input port to the output port of the system 100 on the y-axis, and a wavelength on the x-axis. The through-put transmission varies across wavelength and has a particular change at a wavelength indicative of a “0” bit or a “1” bit, as shown. For example, a particular wavelength (A R) may have a minimum or other predetermined thru-port transmission level indicative of a “0”. Another particular wavelength may have a minimum or other predetermine thru-port transmission level indicative of a “1.” These wavelengths may shift with temperature. For example, the wavelength having a transmission minimum for bit “0” is at a first wavelength at temperature T, but at a different, shifted wavelength at temperature Ti. Examples of systems and devices described herein may be used to reduce this temperature-dependent behavior. While temperature-dependent behavior is depicted in the graph, it is to be understood that process variations in the fabrication of system 100 may also cause performance variations.

[0062] Turning now to FIG. 2, FIG. 2 is a graphical representation of an example ring tuning scheme arranged in accordance with example described herein. The ring tuning scheme mayDocket No. 0077145-08201be used to tune systems including waveguides described herein, such as the system 100 of FIG. 1.

[0063] Generally, MRM-based optical interconnects may utilize thermal tuning to wholly and / or partially compensate for thermal fluctuations and / or process variations. In some examples, less precise and robust thermal tuning may be used when non-switching networks use a laser source is stabilized via thermoelectric cooling (TEC) at a constant temperature and power level. In some examples, a more robust thermal tuning method (e.g.. robust to optical power fluctuations) for MRM-based optical transceivers to optimize and lock MRM’s resonance wavelength may be used when co-packaged I / O with WDM switching networks are used for chip-to-chip communication for supercomputers.

[0064] Generally, MRM-based optical interconnects may utilize precise and robust thermal-tuning to compensate for thermal fluctuations and / or fabrication-related resonance offsets. In some embodiments, less stringent tuning techniques may be sufficient when non-switching optical networks employ laser sources that are stabilized at substantially constant temperature and power levels. In some examples, laser sources that are stabilized at substantially constant temperature and power levels may be through thermoelectric cooling (TEC). In some examples, a more robust thermal tuning method (e.g., robust to optical power fluctuations) for MRM-based optical transceivers to optimize and lock MRM’s resonance wavelength may be needed when co-packaged I / O with WDM switching networks are used for chip-to-chip communication for supercomputers.

[0065] In some embodiments, a MRM configured to undergo an initial non-volatile tuning procedure may exhibit reduced sensitivity to temperature variations. In some examples, an initial non-volatile tuning procedure may use a closed-loop control system. In some examples, a reduced sensitivity to temperature may be accomplished based on utilizing the TOC properties of PCM.

[0066] Accordingly, a first graph 202 in FIG. 2 depicts thru-port transmission against wavelength for an example system. As described herein, the transmission may have a minimum or other predetermined feature at a particular wavelength - for example the transmission in graph 202 has a minimum at MR. Examples of systems described herein may program the phase-change material included in the system. For example, the PCM in FIG. 1 may be programmed. Programming the PCM may refer to placing the PCM in a particular state (e.g., crystalline, amorphous, intermediate). Systems described herein may be tuned to a particular resonant frequency with the PCM in a particular state.Docket No. 0077145-08201

[0067] A next graph 204 in FIG. 2 illustrates that the wavelength of the laser and / or resonant wavelength or other predetermined wavelength ( MR) may be adjusted through the tuning. Once tuned, the behavior of the thru-port transmission may be constant across temperature, as shown in graph 206.

[0068] Turning to FIGs. 3A-3B, FIGs. 3A-3B are cross sections of example waveguides arranged in accordance with examples described herein. The waveguides shown in FIGS. 3A- 3B may be used in systems described herein, such as in system 100 of FIG. 1. FIGS. 3A-3B are exemplary. Additional, fewer, and / or different components may be used in other examples. FIGS. 3A-3B are schematic. For example, material layers are depicted using rightangle comers and distinct boundaries. It is to be understood that example systems as manufactured may have rounded or other shaped corners and / or boundaries which may be less distinct and / or straight between material layers.

[0069] In some embodiments, a MRM may have a separate heater section using P-N or P-i-N junctions, which may facilitate the programming of PCM for tuning the MRM resonance. In some examples, a MRM may have a waveguide with a P-N heater for PCM. In some examples, a MRM may have a waveguide with a P-i-N heater for PCM. The separate heater section configuration decouples the P-N junctions of PCM programming from a high-speed modulation. In some examples, the sections may be optimized independently. In some examples, the sections may have separate driver circuitry.

[0070] In some embodiments, a MRM may utilize a shared P-N junction where the same P-N junction can be used for programming PCM as well as high-speed modulation. A P-N diode may incorporate a P-N junction. In some examples, operating a P-N diode in forward bias regions programs the PCM. In some examples, operating the P-N diode in depletion (e.g., reverse bias mode) performs the high-speed modulation.

[0071] A MRM may operate in an athermal manner by configuring the waveguide to include materials having offsetting TOCs. In some embodiments, a MRM may combine silicon with an optimized layer of low-loss PCM. In some examples, the silicon may have a positive TOC (dn / dT). In some examples, the positive dn / dT may be approximately 1x1 O'4. In some examples, the optimized layer of low-loss PCM may be antimony sulfide (SbS). The layering of PCM may be occur after a commercial foundry process.

[0072] While examples of athermal operation are described herein, and examples of material combinations resulting in a negative TOC are described, in other examples systems may achieve operation having a predetermined relationship with temperature, and varying TOC may be used. For example, some systems may have a PCM material and / or resonantDocket No. 0077145-08201wavelength selected to achieve a predetermined relationship with temperature. A known relationship with temperature may be used, together with one or more temperature measurements, for example, to decode output of an optical system and / or optical device described herein.

[0073] FIG. 3A illustrates a cross-section of a device using negative TOC PCM. The example device of FIG. 3A includes waveguide 302, P-N junction 304, and PCM 306. The waveguide 302 may be provided, for example, as a bulk layer and / or a layer on a substrate. The P-N junction 304 may be provided as two materials adjacent one another and deposited, grown, or otherwise provided on the waveguide 302. Conductive connections may be formed to the two materials forming the P-N junction. The PCM 306 may be provided as a layer deposited, grown, and / or otherwise provided above the materials forming the P-N junction 304. As shown in FIG. 3A, the PCM material may cover the P-N junction 304 in a conformal manner, however, other coating patterns may be used in other examples. In some embodiments, a MRM may combine silicon with an optimized layer of low-loss PCM 306. In some embodiments, the optimized layer of low-loss PCM 306 may have a negative dn / dT. In some examples, a negative dn / dT may be of approximately -2xl0’4. The combination silicon with an optimized layer of low-loss PCM 306 may achieve zero dn / dT. In some examples, zero dn / dT may mean that refractive index becomes independent of temperature, which allows resonance to remain constant across all temperature. In some examples, when the zero dn / dT may be actually zero or it may be numerically near-zero such that the refractive index does not change substantially with temperature change across an operational range.

[0074] FIG. 3B illustrates a cross-section of a device using PCM capped with negative thermo-optic oxide. A device may include a waveguide 308, P-N junction 310, PCM 312, and oxide 314. In some embodiments, the optimized layer of low-loss PCM 312 may have a positive dn / dT. In some embodiments, an oxide 314 layer may have a negative dn / dT.

[0075] In some embodiments, an oxide 314 layer may have one or more layers. In some embodiments, a first oxide 314 layer may be a passivation layer, and a second oxide 314 layer may be a layer with a negative TOC. In some embodiments, the first oxide 314 layer may be immediately deposited on a PCM 312. In some embodiments, the second oxide 314 layer may be immediately deposited on the first oxide 314 layer. In some embodiments, the second oxide 314 layer may be thicker than the first oxide layer. In some examples, a first oxide 314 layer may be AI2O3. In some examples, a first oxide 314 layer may be HfO2. In some examples, a first oxide 314 layer may be SiN. In some examples, a first oxide 314 layer may be SiO2. In some examples, the second oxide 314 layer may be TiCh. In some examples, an oxide layer may have more than two layers. In some examples, an oxide layer may have one or moreDocket No. 0077145-08201passivation layers. In some examples, each of the one or more passivation layers may have different materials. In some examples, an oxide layer may have one or more negative TOC oxide layers. In some examples, an oxide layer may have one or more passivation layers, negative TOC oxide layers, or a combination thereof.

[0076] In some embodiments, various materials may be used as a PCM. In some examples, Ge2Sb2Tes (GST) may be used as a PCM. In some examples, Ge2Sb2 Se4Tei (GSST) may be used as a PCM. In some examples, Sb2Ss (SbS) may be used as a PCM. In some examples, Sb2Ses (SbSe) may be used as a PCM. In some examples. GeTe may be used as a PCM. In some examples, SbSe and / or SbS may be typical choices to lower optical loss. This may be beneficial in cry ogenic applications for non-volatile resonance tuning. PCM has two stable states with distinct optical properties, and the stable states can be reversibly switched using tailored heat pulses. One stable state is amorphous: the other is crystalline. Once a PCM switches its state, the PCM remains in its state in a non-volatile manner. In some examples, the non-volatile manner may be without static power dissipation. In some examples, a PCM may demonstrate a resonance shift of approximately 0.5 nm (with minor Q-factor reduction) for an MRM with a free spectral range (FSR) of 4.5 nm at 4K temperature.

[0077] In some embodiments, various thickness of GST may be layered on the MRM. In some examples, a GST with 12.5 nm thickness may be deposited on a 8 pm long section of the MRM. In some embodiments, the GST layer may vary in thickness depending on the desired tuning efficiency or optical response. In some examples, thickness values include approximately 5nm, 7 nm, 10 nm, 15 nm, 20 nm, 25 nm, 30 nm, 40 nm, or 50 nm, although other thicknesses may also be employed.

[0078] Cryogenic non-volatile photonics may be where a thin film of GST is electrically switched at sub-4K temperatures. Tuning may be done using P-i-N or P-N microheater stricture. Tuning may be continuously tuned using a sequence of tailored pulses with various voltages and pulse-duration. In some embodiments, the tuning may utilize switching pulses. In some examples, the switching pulses may be 10V for hundreds of nanoseconds (e.g., short pulse) for amorphization (e.g., RESET) and around 5V for lOus for crystallization (e.g., SET).

[0079] Turning now to FIGs. 4A-4D, FIGs. 4A-4D are cross-sectional schematic illustrations of examples of integrated waveguides in accordance with examples described herein. FIGs. 4A-4B is an illustration of a waveguide with layers of PCM and cladding utilizing a front-side integration.

[0080] The waveguides shown in FIGS. 4A-4D may be used in systems described herein, such as in system 100 of FIG. 1. FIGS. 4A-4D are exemplary. Additional, fewer, and / orDocket No. 0077145-08201different components may be used in other examples. FIGS. 4A-4D are schematic. For example, material layers are depicted using right-angle corners and distinct boundaries. It is to be understood that example systems as manufactured may have rounded or other shaped corners and / or boundaries which may be less distinct and / or straight between material layers.

[0081] FIG. 4A includes a waveguide 402, a PCM 404, a negative TOC oxide cladding 406, an insulating layer 408, and an oxide cladding 410.

[0082] PCM and cladding may be integrated (e.g., added) to the waveguide or MRM after a standard foundry process In some examples. Methodologies for integrating PCM and negative TOC oxide cladding may involve front-side integration. In some examples, front-side integration may by Back-End-of-Line (BEOL) integration. Front-side integration may integrate the PCM and negative TOC oxide from the top of the chip.

[0083] In some embodiments, oxide openings are created over specific waveguide sections using various techniques to expose the waveguide layer. In some examples, a technique may be HF vapor etching. In some examples, a technique may be reactive-ion etching (RIE). In some examples, a technique may be inductively coupled plasma (ICP) etching. In some embodiments, a thickness of PCM may be deposited onto the exposed waveguide layer. In some examples, an encapsulation layer may be deposited onto the PCM and / or the exposed waveguide layer. In some examples, a negative TOC cladding may later be deposited.

[0084] In some embodiments, waveguide 402 may sit above an insulating layer 408 (e.g., buried oxide later) with the remaining sides exposed. In some examples, the remaining sides are not buried (e.g., covered) under the oxide cladding 410. In some examples, PCM 404 may be deposited on all sides of an exposed waveguide 402. In some examples, negative TOC oxide cladding 406 may be deposited on all exposed sides of the deposited PCM 404.

[0085] In some embodiments, a negative TOC oxide cladding 406 may have multiple layers. In some examples, the negative TOC oxide cladding 406 may have a passivation layer and an oxide cladding layer with a negative TOC. In some examples, the oxide cladding layer may be thicker than the passivation layer. In some embodiments, the passivation layer may be AI2O3, HfO2, SiN, or SiO2. In some embodiments, the oxide cladding 410 may have the same material as the passivation layer. In some embodiments, the negative TOC oxide cladding 406 may have one or more passivation layers, oxide cladding layers with a negative TOC, or a combination thereof.

[0086] FIG. 4B includes a waveguide 412, a PCM 414, a negative TOC oxide cladding 416, insulating layer 418, and oxide cladding 420.Docket No. 0077145-08201

[0087] In some embodiments, waveguide 412 may be provided above an insulating layer 418. In some examples, a portion of the waveguide 412 may be exposed due to the etching, while a remaining portion of the waveguide 412 may be provided under an oxide cladding 420. In some examples, a PCM 414 may be deposited, grown, and / or otherwise provided on the exposed sides of the waveguide 412. In some examples, negative TOC oxide cladding 416 may be deposited, grown, or otherwise provided on some or all exposed sides of the deposited PCM 414.

[0088] In some embodiments, a negative TOC oxide cladding 416 may have multiple layers. In some examples, the negative TOC oxide cladding 416 may have a passivation layer and an oxide cladding layer with a negative TOC. In some examples, the oxide cladding layer may be thicker than the passivation layer. In some embodiments, the passivation layer may be AI2O3, HfO2, SiN, or SiO2. In some embodiments, the oxide cladding 420 may have the same material as the passivation layer. In some embodiments, the negative TOC oxide cladding 416 may have one or more passivation layers, oxide cladding layers with a negative TOC, or a combination thereof.

[0089] FIGs. 4C-4D is an illustration of a waveguide with layers of PCM and cladding utilizing a back-side integration from the insultation layer (e.g., buried oxide (BOX) layer).

[0090] FIG. 4C includes a waveguide 422, a PCM 424, a negative TOC oxide cladding 426, an insulating layer 428, and an oxide cladding 430.

[0091] Methodologies for integrating PCM and negative TOC oxide cladding may involve back-side integration. In back-side integration, a waveguide layer is accessed from a rear of a chip. In some embodiments, a silicon handling layer in target regions may be removed. In some examples, an etching technique to remove the silicon handling layer may be XeF2 vapor etching. In some examples, an etching technique to remove the silicon handling layer may be deep reactive-ion etching (DRIE).

[0092] In some embodiments, oxide openings may be created by etching through a buried oxide layer to reach a waveguide layer followed by the deposition of the PCM, encapsulation layer, and negative TOC cladding on the waveguide section.

[0093] In some embodiments, waveguide 422 may be provided above a 428 and below an oxide cladding 430. An oxide opening may be etched to expose the waveguide 422. In some examples, PCM 424 may be deposited, grown, or otherwise provided on some or all of the exposed sides of the waveguide 422. In some examples, negative TOC oxide cladding 426 may be deposited, grown, or otherwise provided on some or all of the exposed sides of the waveguide PCM 424.Docket No. 0077145-08201

[0094] In some embodiments, a negative TOC oxide cladding 426 may have multiple layers. In some examples, the negative TOC oxide cladding 426 may have a passivation layer and an oxide cladding layer with a negative TOC. In some examples, the oxide cladding layer may be thicker than the passivation layer. In some embodiments, the passivation layer may be AI2O3, HfO2, SiN. or SiO2. In some embodiments, the oxide cladding 430 may have the same material as the passivation layer. In some embodiments, the negative TOC oxide cladding 426 may have one or more passivation layers, oxide cladding layers with a negative TOC, or a combination thereof.

[0095] FIG. 4D includes a waveguide 432, a PCM 434, negative TOC oxide cladding 436, an insulating layer 438, and an oxide cladding 440.

[0096] In some embodiments, waveguide 432 may be provided above an insulating layer 438. In some examples, a portion of the waveguide 432 may be exposed due to the etching, while a remaining portion of the waveguide 432 may be buried under an oxide cladding 440. In some examples, a PCM 434 may be deposited, grown, or otherwise provided on some or all the exposed sides of the waveguide 432. In some examples, negative TOC oxide cladding 436 may be deposited, grown, or otherwise provided on some or all exposed sides of the deposited PCM 434.

[0097] In some embodiments, a negative TOC oxide cladding 436 may have multiple layers. In some examples, the negative TOC oxide cladding 436 may have a passivation layer and an oxide cladding layer with a negative TOC. In some examples, the oxide cladding layer may be thicker than the passivation layer. In some embodiments, the passivation layer may be AI2O3, HfO2, SiN, or SiO2. In some embodiments, the oxide cladding 440 may have the same material as the passivation layer. In some embodiments, the negative TOC oxide cladding 436 may have one or more passivation layers, oxide cladding layers with a negative TOC, or a combination thereof.

[0098] The overall device design of the modulator and the athermal waveguide geometries can differ depending on the integration method. Even with the added complexity of the backside integration, there may be examples where back-side integration may be preferred over front-side integration.

[0099] Turning to FIGs. 5A-5C, FIGs. 5A-5C are schematic cross-section and top-down illustrations of waveguide geometries in accordance with examples described herein. FIG. 5A illustrates a cross-section of a rib-shape waveguide geometry.

[0100] The waveguides shown in FIGS. 5A-5C may be used in systems described herein, such as in system 100 of FIG. 1. FIGS. 5A-5C are exemplary. Additional, fewer, and / orDocket No. 0077145-08201different components may be used in other examples. FIGS. 5A-5C are schematic. For example, material layers are depicted using right-angle corners and distinct boundaries. It is to be understood that example systems as manufactured may have rounded or other shaped corners and / or boundaries which may be less distinct and / or straight between material layers.

[0101] FIG. 5 A illustrates a device 502 that includes a waveguide 504, insulating layer 506, oxide cladding 508, PCM 510, negative TOC oxide cladding 512, and rib 514.

[0102] A rib-shaped waveguide (e.g., raised-ridge waveguide) may be a waveguide structure formed by partially etching a semiconductor or dielectric waveguide layer so that a raised region (e.g., the rib) remains above a thinner slab region. In such a structure, the rib defines a higher-index guiding region, while the surrounding slab supports the remainder of the optical mode and enables lateral confinement.

[0103] In some embodiments, a device 502 may be etched such that waveguide 504 that is provided above insulating layer 506 has a rib 514. In some examples, a rib 514 may be exposed while the remaining waveguide 504 is covered by oxide cladding 508. In some examples, a PCM 510 may be deposited, grown, or otherwise provided over rib 514. In some examples, negative TOC oxide cladding 512 may be deposited, grown, or otherwise provided over TOC 510.

[0104] In some embodiments, a negative TOC oxide cladding 512 may have multiple layers. In some examples, the negative TOC oxide cladding 512 may have a passivation layer and an oxide cladding layer with a negative TOC. In some examples, the oxide cladding layer may be thicker than the passivation layer. In some embodiments, the passivation layer may be AI2O3, HfO2, SiN, or SiO2. In some embodiments, the oxide cladding 508 may have the same material as the passivation layer. In some embodiments, the negative TOC oxide cladding 512 may have one or more passivation layers, oxide cladding layers with a negative TOC, or a combination thereof.

[0105] In some embodiments, temperature-dependent wavelength shift may be controlled by power confinement in a positive TOC waveguide 504 and a negative TOC oxide cladding 512. Power confinement in a waveguide 504 may be engineered by optimizing waveguide 504 geometry and / or dimension to achieve the predetermined power.

[0106] In some embodiments, light (e.g., energy) may be guided in the cladding of a slot waveguide 504. In some embodiments, light may be guided through a low refractive index central core of a hollow core waveguide 504.

[0107] In some embodiments, light may be guided through holes (e.g., defects) filled with a cladding material of a photonic crystal waveguide 504. In some embodiments, light may beDocket No. 0077145-08201guided through holes (e.g., defects) filled with a cladding material of a sub-wavelength grating waveguide 504.

[0108] In some embodiments, a waveguide 504 may have a standard rib-shape geometry. In some embodiments, a waveguide 504 may have a modified L-shaped rib-shape geometry.

[0109] Typically, reducing the height and / or width of a guiding (e.g., rib) layer may lead to an increase in cladding confinement. Similarly, reducing the height of a thinner slab layer may lead to an increase in cladding confinement.

[0110] In some embodiments, a height of a layer may be fixed. In some embodiments, a width of a layer may be modified. In some examples, temperature invariant effective index may be obtained for a width of 310 nm for rib-waveguide geometry for a 45 nm commercial silicon photonics process. A waveguide of an MRM may be implemented with any suitable width, such as about 200 nm, 250 nm, 300 nm, 350 nm, 400 nm, 450 nm, 500 nm, 600 nm, 700 nm, 800 nm, or greater, with the specific width selected to achieve desired optical-mode confinement, modulation efficiency, bending performance, and / or thermal characteristics. A silicon photonics or semiconductor process may be characterized by a technology node having nominal feature sizes such as 180nm, 130 nm, 90 nm, 65 nm, 28 nm, 22 nm, 14 nm, or smaller, while additional photonics-specific layers such as silicon, silicon-nitride, oxide, or phase-change-material layers may have thicknesses ranging from a few nanometers to several hundred nanometers depending on device requirements.

[0111] In some embodiments, a shape (e.g., configuration) of a layer may be modified. A waveguide may be implemented using any suitable cross-sectional geometry. The geometry¬ may be selected to achieve desired optical confinement, bending performance, polarization characteristics, or compatibility with modulation structures such as P-N junction, P-i-N junctions, and / or PCM. In some examples, a cross-sectional geometry may be strip-shaped. In some examples, a cross-sectional geometry- may be rib-shaped. In some examples, a cross-sectional geometry may be half-rib-shaped. Other shapes such as a ridge, slot, buried, suspended, tapered, or multi-material may be used.

[0112] A negative effective TOC may be obtained for a net athermal phase shift in photonic devices to account for thermal expansion of a waveguide.

[0113] In some embodiments, a zero or negative effective TOC while maintaining a well-confined guided optical mode may be challenging for foundry processes that utilize relatively large rib and slab layer heights (e.g.. AMF GP silicon photonic platform) to achieve. In some examples, an L-shaped waveguide may be used. An L-shaped structure may modify a rib waveguide by reducing the slab thickness to zero on one of the sides, reducingDocket No. 0077145-08201confinement in the waveguide medium. In some examples, a waveguide at a width of 206 nm with an L-shaped geometry may have a temperature-dependent wavelength shift of 0 pm / K for a-SbS. In some examples, a waveguide at a width of 206 nm with an L-shaped geometry may have a temperature-dependent wavelength shift of 3 pm / K for c-SbS.

[0114] FIG. 5B illustrates a cross-section of a strip-shape waveguide geometry.

[0115] In some embodiments, device 516 includes waveguide 518, insulating layer 520, oxide cladding 522, PCM 524, and negative TOC oxide cladding 526.

[0116] A strip-shaped waveguide (e.g., fully-etched waveguide) is a waveguide geometry' in which the entire thickness of the waveguide layer surrounding the guiding region is etched. In some examples, after etching, a defined core with no residual slab remains. In some examples, the strip-shaped waveguide may7have a rectangular cross-section. In some examples, the strip-shaped waveguide may have a trapezoidal cross-section.

[0117] In some embodiments, device 516 may be etched such that all of waveguide 518 is exposed and no portions portion of the waveguide 518 is covered by oxide cladding 522. In some examples, PCM 524 may be deposited over waveguide 518. In some examples, negative TOC oxide cladding 526 may be deposited over PCM 524.

[0118] In some embodiments, a negative TOC oxide cladding 522 may have multiple layers. In some examples, the negative TOC oxide cladding 522 may have a passivation layer and an oxide cladding layer with a negative TOC. In some examples, the oxide cladding layer may be thicker than the passivation layer. In some embodiments, the passivation layer may be AI2O3, HfO2, SiN, or SiO2. In some embodiments, the oxide cladding 526 may have the same material as the passivation layer. In some embodiments, the negative TOC oxide cladding 522 may have one or more passivation layers, oxide cladding layers with a negative TOC, or a combination thereof.

[0119] FIG. 5C illustrates a cross-section of a half-rib-shape waveguide geometry'.

[0120] In some embodiments, device 528 includes waveguide 530, insulating layer 532, oxide cladding 534, PCM 536, negative TOC oxide cladding 538, and rib 540.

[0121] A half-rib waveguide may refer to a waveguide geometry in which only a portion of the waveguide layer is etched to form a rib, while the opposite side or remaining lateral region retains the full waveguide thickness. In some examples, the rib extends from one side of the waveguide layer, rather than symmetrically on both sides as in a conventional rib waveguide.

[0122] In some embodiments, a device 528 may be etched such that waveguide 530 that sits above insulating layer 532 has a rib 514. In some examples, a rib 514 may be exposed whileDocket No. 0077145-08201the remaining waveguide 504 is covered by oxide cladding 534. In some examples, rib 540 extends to one end of the waveguide 530 that is covered by oxide cladding 534.

[0123] In some examples, a PCM 536 may be deposited over rib 540. In some examples, negative TOC oxide cladding 538 may be deposited over PCM 536.

[0124] In some embodiments, a negative TOC oxide cladding 536 may have multiple layers. In some examples, the negative TOC oxide cladding 536 may have a passivation layer and an oxide cladding layer with a negative TOC. In some examples, the oxide cladding layer may be thicker than the passivation layer. In some embodiments, the passivation layer may be AI2O3, HfO2, SiN, or SiO2. In some embodiments, the oxide cladding 534 may have the same material as the passivation layer. In some embodiments, the negative TOC oxide cladding 536 may have one or more passivation layers, oxide cladding layers with a negative TOC, or a combination thereof.

[0125] Turning to FIGs. 6A-6B, FIG. 6A is a schematic illustration of a device including a micro-ring modulator arranged in accordance with examples described herein. Examples of interconnects or devices described herein may include one or more optical modulators.

[0126] The devices shown in FIGS. 6A-6B may be used to implement and / or may be implemented by systems described herein, such as system 100 of FIG. 1. The devices shown in FIGS. 6A-6B may include waveguide and material stacks as shown and described with reference to FIGS. 3A-3B, FIGS. 4A-4D, and / or FIGS. 5A-5C. FIGS. 6A-6B are exemplary. Additional, fewer, and / or different components may be used in other examples.

[0127] FIG. 6A illustrates a PCM-integrated MRM with a P-N junction for high-speed modulation separated from a P-N junction for PCM heater.

[0128] The device includes a MRM with an input port 602. modulation P-N junction 604, a PCM heater P-N junction 606, and an output port 608.

[0129] In some embodiments, the input port 602 of the device may receive optical input energy. In some examples, optical input energy may be a laser light. In some embodiments, the output port 608 may provide the modulated output energy (e.g., modulated output laser light).

[0130] High-speed MRM devices may utilize on carrier-plasma effect for high-speed modulation using P-N junctions in depletion mode (e.g., reverse bias). In some embodiments, this P-N junction can be completely separated (e.g., independent) from the P-N heater and / or P-i-N heater junction for tuning the PCM. In some examples, this separated design of the P-N junction may be preferred. In some examples, each P-N junction may have a complementary metal-oxide semiconductor (CMOS) circuitry. In some examples, the circuitry may beDocket No. 0077145-08201fabricated using a 28nm process. In some examples, the circuitry may be fabricated using a 180 nm, 130 nm, 90 nm, 65 nm, 45 nm, or 22 nm process.

[0131] CMOS circuitry’ may refer to an arrangement of electronic circuit components formed using complementary p-type and n-type metal-oxide-semiconductor field-effect transistors (MOSFETs). The circuitry may be fabricated on a semiconductor substrate and incorporates insulating oxide layers, doped semiconductor regions, and metallization layers that provide electrical interconnects. In some embodiments, p-channel and n-channel MOSFETs are configured in complementary pairs such that one device conducts when the other is non-conducting, thereby enabling low static-power operation and high switching efficiency.

[0132] CMOS circuitry may be used to implement logic functions, control circuits, driver structures, biasing networks, signal-processing functions, or other electronic operations. The circuitry can be integrated monolithically with photonic components or other semiconductor structures, allowing electrical control signals to be generated, conditioned, or applied to associated devices such as modulators, heaters, detectors, or phase-change-material tuning elements. In some examples, each P-N junction that has a CMOS circuitry may be dedicated to programming and / or modulation. In some embodiments, the independent P-N junction design may create limitations due to a fragmented ring circumference. In some examples, the independent P-N junction design may limit the modulation efficiency (e.g., resonance shift). In some examples, the independent P-N junction design may limit the range of resonance tuning range.

[0133] FIG. 6B illustrates a PCM-integrated MRM with a P-N junction for high-speed modulation and PCM heater.

[0134] The device includes a MRM with an input port 610, modulation and PCM heater P-N junction 612, and output port 614.

[0135] In some embodiments, the input port 610 of the device may receive optical input energy. In some examples, optical input energy7may be a laser light. In some embodiments, the output port 614 may provide the modulated output energy (e.g., modulated output laser light).

[0136] The device in FIG. 6B may have a shared P-N junction. The shared P-N junction may have be used for both modulation and programming. In some embodiments, the shared P-N junction design may allow the reuse of high-speed P-N junction of the MRM for programing the PCM.

[0137] In some embodiments, the shared P-N junction design may allow7the PCM to be coated on the entire ring. In some examples, a ring may be approximately 7.5 urn in radius.Docket No. 0077145-08201Having the PCM coated on the entire ring may provide a better tuning range. In some examples, a tuning range of an entirely coated ring may be up to 5nm. In some examples, shared P-N junction design may incorporate co-designing a modulation and programming circuitry to drive a single P-N diode without affecting the performance.

[0138] Turning to FIG. 7, FIG. 7 is a graph of a temperature dependent resonance wavelength shift to a waveguide width in accordance with examples herein. The graph may describe behavior of example waveguides and / or devices or systems including waveguides described herein, such as system 100 of FIG. 1, and / or the devices shown and described with reference to FIGS. 6A-6B. The performance shown in FIG. 7 is exemplary - different performance may occur in other examples.

[0139] MRM may operate in an athermal matter by configuring the waveguide to include materials having offsetting TOCs. Athermal operation or minimal response to thermal fluctuations can be achieved if the effective TOC of an optical mode in the photonic waveguide is lowered (e.g., reduced to zero or brought as close to zero).

[0140] In some embodiments, a MRM may combine silicon with an optimized layer of low-loss PCM. In some embodiments, the optimized layer of low-loss PCM may have a negative dn / dT.

[0141] Silicon has a positive TOC (e.g., 1.8 * 104) and can be coated with a material that has a negative TOC so that the effective TOC of an optical mode becomes zero or closer to zero. Typically, a material with a high negative TOC is favorable for achieving athermal operation even if there exists a low confinement factor of the mode in cladding. This allows for larger core dimensions that minimize loss. In some examples, a thick layer of PCM with negative TOC can be deposited (e.g., coated) directly on the waveguide through magnetron sputtering. In some examples, a PCM may be SbS. In some examples, SbS may have a negative TOC of -3xl0'4in its amorphous state (a-SbS). In some examples, SbS may have a negative TOC of -1.5x1 O’4in its crystalline state (c-SbS).

[0142] In some embodiments, a geometry' of a waveguide may determine the appropriate PCM thickness to be applied in order to obtain a near-zero effective TOC in a PCM amorphous state and / or a PCM crystalline state. In some examples, a 265 nm core width and 100 nm thick SbS layer deposited on the waveguide provides a temperature-dependent wavelength shift of the ring resonator with 0 pm / K for a-SbS. In some examples, a 265 nm core width and 100 nm thick SbS layer deposited on the waveguide provides a temperature-dependent wavelength shift of the ring resonator with 14 pm / K for c-SbS. In some examples, when near-zero TOCDocket No. 0077145-08201may be numerically near-zero or actually zero such that the refractive index does not change substantially with temperature change across an operational range.

[0143] FIG. 8 is a graph of a temperature dependent resonance wavelength shift to waveguide width in accordance with examples described herein. The graph may describe behavior of example waveguides and / or devices or systems including waveguides described herein, such as system 100 of FIG. 1, and / or the devices shown and described with reference to FIGS. 6A-6B. The performance shown in FIG. 8 is exemplary - different performance may occur in other examples.

[0144] In some embodiments, precise switching of PCM may be facilitated by applying (e.g., combining) a layer of PCM with an oxide cladding that has a negative TOC. In some examples, the layer of PCM may be a thin layer. In some examples, an oxide cladding may be TiO2, which has a high negative TOC of approximately -2x1 O’4. In some examples, TiO2 may be used because it may be compatible with the CMOS process. In some examples, TiO2 may be deposited over a PCM through magnetron sputtering. In some examples, TiO2 may be deposited over a PCM through atomic layer deposition. In some examples, an oxide cladding may be VO2, which has an ultra-high negative refractive index -3xl0’3at 1550 nm.

[0145] In some embodiments, an ultra-high negative refractive index oxide may the performance of precise switching. In some examples, using an ultra-high negative refractive index oxide may allow a higher waveguide core widths, which can facilitate a smaller ring dimension for dense integration. In some examples, using an ultra-high negative refractive index oxide may allow a large free spectral range (FSR) for a large number of dense wavelength division multiplexing (DWDM) channels

[0146] Similarly, in some embodiments, a geometry of a waveguide may determine the appropriate PCM thickness to be applied in order to obtain a near-zero effective TOC in a PCM amorphous state and / or a PCM crystalline state. In some examples, a 290 nm core width and 100 nm thick SbS layer deposited on the waveguide provides a temperature-dependent wavelength shift of the ring resonator with 0 pm / K for a-SbS. In some examples, a 290 nm core width and 100 nm thick SbS layer deposited on the waveguide provides a temperaturedependent wavelength shift of the ring resonator with 6 pm / K for c-SbS.

[0147] In some embodiments, an additional layer of oxide cladding may be added. In some examples, an additional layer of VO2 may be added to a waveguide geometry. In some examples, a 600 nm core width and 1550 nm thick VO2 layer deposited on the waveguide provides a temperature-dependent wavelength shift of the ring resonator with 0 pm / K for a-SbS. In some examples, a 600 nm core width and 1550 nm thick VO2 layer deposited on theDocket No. 0077145-08201waveguide provides a temperature-dependent wavelength shift of the ring resonator with 3 pm / K for c-SbS.

[0148] In this way, applying a layer of PCM with an oxide cladding may provide a lower temperature-dependent wavelength shift for PCM in a crystalline state while ensuring a reliable switching behavior of the PCM layer.

[0149] FIGs. 9A-9B are schematic illustrations of waveguide devices arranged in accordance with examples described herein. Waveguides and / or systems described herein may be used to implement and / or may be implemented by the devices shown and described with reference to FIGS. 9A-9B. For example, all or a portion of system 100 of FIG. 1 may be used to implement portions of the waveguide structures in FIGS. 9A-9B. In some examples, the systems of FIGS. 6A-6B may be used to implement the waveguide structures in FIGS.9A-9B.

[0150] FIGS. 9A-9B are exemplary. Additional, fewer, and / or different components may be present in other examples. The number of laser wavelengths, inputs, outputs, and waveguide structures shown may vary in other examples - any number may be used.

[0151] In FIG. 9A, a MRR may include a core 902, PCM 904, and negative TOC oxide 906.

[0152] In some embodiments, tuning may narrow linewidth for resonant devices. In some examples, PCM 904 and negative TOC oxide 906 may coat an entire ring of a MRR. In some examples, the PCM 904 may trim the resonances of the MRR.

[0153] In some embodiments, a negative TOC oxide 906 may have multiple layers. In some examples, the negative TOC oxide 906 may have a passivation layer and an oxide cladding layer with a negative TOC. In some examples, the oxide cladding layer may be thicker than the passivation layer. In some examples, the oxide cladding layer may be thinner than the passivation layer. In some embodiments, the passivation layer may be AI2O3, HfO2, SiN, or SiO2. In some embodiments, the negative TOC oxide 906 may have one or more passivation layers, oxide cladding layers with a negative TOC, or a combination thereof.

[0154] In some embodiments, a composite waveguide structure including the core, a PCM layer, and an engineered cladding (e.g., negative TOC oxide cladding) may enable programming of a non-zero thermal drift. A thermal drift may be a change in performance parameters (e.g., voltage, current, resistance) of a semiconductor device due to temperature fluctuations. A non-zero thermal drift may indicate that a controlled dependence with temperature change.

[0155] In some embodiments, a high-density architectures such as CPO with integrated lasers (e.g., Indium Phosphide lasers) may exhibit a characteristic thermal wavelength shiftDocket No. 0077145-08201that differs from the intrinsic shift of silicon. In some examples, configuring the material volume ratios and the state of the PCM layer may allow for the effective TOC to be tuned to synchronize its resonance shift with the emission drift of a coupled laser source.

[0156] In some embodiments, a high-TOC cladding (e.g., optical polymers) may be employed to bridge the drift gap between disparate material systems that may have an amplified sensitivity. In some examples, thermal tracking may allow for spectral alignment across a wide temperature range without power-intensive active feedback loops.

[0157] FIG. 9B illustrates a MZI with portions of an arm coated with PCM and negative TOC oxide. In some examples, a negative TOC oxide 914 may coat an entire MZI. In some examples, a negative TOC oxide cladding may coat a portion of the MZI.

[0158] In some embodiments, MZI with an upper arm 908 and lower arm 910. The lower arm 910 is coated with PCM 912 and negative TOC oxide 914. In some embodiments, the PCM 912 and negative TOC oxide 914 may be coated on the upper arm 908. In some embodiments the PCM 912 and negative TOC oxide 914 may be coated on both the upper arm 908 and lower arm 910. In some embodiments, the PCM 912 and negative TOC oxide 914 may be coated on different and / or alternative arms.

[0159] In some embodiments, a negative TOC oxide 914 may have multiple layers. In some examples, the negative TOC oxide 914 may have a passivation layer and an oxide cladding layer with a negative TOC. In some examples, the oxide cladding layer may be thicker than the passivation layer. In some examples, the oxide cladding layer may be thinner than the passivation layer. In some embodiments, the passivation layer may be AI2O3, HfO2. SiN, or SiO2. In some embodiments, the negative TOC oxide 414 may have one or more passivation layers, oxide cladding layers with a negative TOC, or a combination thereof.

[0160] The techniques described herein to control phase of propagating light while maintaining independence from temperature variation can be utilized for athermalization of other active and / or passive devices.

[0161] In some examples, devices may be Arrayed Waveguide Gratings (AWGs), optical delay lines, beam-forming networks, and / or polarization controllers. In some examples, devices may be athermalize interferometric devices such as power-splitters / combiners, beamformers, optical switches, and / or programmable photonic processors.

[0162] In some embodiments, MZI-based multiplexing or demultiplexing devices may operate without tuning in Coarse-WDM (CWDM) wavelength grids. In some embodiments, active tuning may be utilized to reduce wavelength spacing. In some embodiments, active tuning may be utilized to switch to a DWDM system.Docket No. 0077145-08201

[0163] PCM 912 may be used to control phase differences between the arms. In some examples, a negative TOC oxide 914 may obtain a phase shift between MZI upper arm 908 and lower arm 910. In some examples, a combination of waveguide length difference and negative TOC oxides 914 may provide a phase shift between MZI upper arm 908 and lower arm 910.

[0164] In some embodiments, MZI-based devices with a balanced MZI may omit utilizing a negative TOC material. In some embodiments, adjusting widths of arms in unbalanced MZI may obtain an athermal response.

[0165] FIG. 10 is a schematic illustration of a closed-loop tuning system arranged in accordance with examples described herein.

[0166] The system 1002 of FIG. 10 includes waveguide system 1006. The waveguide system 1006 includes PCM material 1008 and diode 1010. The waveguide system 1006 may be used to modulate laser light from an input port, depicted as a multi-wavelength input light source in FIG. 10 to an output port, depicted as modulated light output in FIG. 10. The system 1002 further includes data 1014 coupled to driver 1012. The driver 1012 is coupled to drive the waveguide system 1006. The system 1002 further includes a heater driver 1016, including a port 1018 coupled to circuitry 1022 coupled to controller 1020 which is coupled to control diode 1010 of the waveguide system 1006.

[0167] The tuning system of FIG. 10 may be utilized to tune systems described herein, including waveguide and PCM-including systems. For example, the tuning system of FIG. 10 may be used to tune system 100 of FIG. 1 and / or the devices shown and described with reference to FIGS. 6A-6B.

[0168] FIG. 10 is exemplar^'. Additional, fewer, and / or different components may be used in other examples.

[0169] During operation, data 1014 may be provided to driver 1012. The data 1014 may be digital data (e.g., binary data). The driver 1012 may be a high-speed driver. The driver 1012 may be implemented using circuitry, such as one or more digital-to-analog converters and / or amplifiers. The driver 1012 may be coupled to the waveguide system 1006. For example, an output of the driver 1012 may be coupled to one side of the diode 1010. The driver 1012 may be coupled to a port of the waveguide system 1006 used to control modulation.

[0170] During a tuning phase of operation, the driver 1012 may be grounded. During tuning, data 1014 may be decoupled from the waveguide system 1006.

[0171] The heater driver 1016 may be used during tuning to deliver a voltage across the diode 1010. In some examples, the heater driver 1016 may deliver 10 V across the diode 1010Docket No. 0077145-08201which may have a forward-bias resistance of 100 Q. Other parameter values may be used in other examples. During tuning, a phase of the PCM material 1008 may be selected and / or laser wavelengths tuned or other parameters of the waveguide system 1006 adjusted to achieve a resonance of the waveguide system 1006 at a predetermined frequency or frequencies. The heater driver 1016 may be implemented using circuitry’. For example, the heater driver 1016 may include circuitry 1022, which may be analog circuitry and / or controller 1020, which may be a digital controller. The circuitry71022 may include, for example, one or more amplifiers, transistors, and / or analog-to-digital converters. In some examples, a series of transistors may be used, such as six thick-oxide transistors. In some embodiments, the transistors may have a voltage that maintains device stress based on device specification. In some examples, the transistors may be rated for 1.8 V.

[0172] One or more signals provided at port 1018 may control the heater driver 1016 and may set an optical power provided to and / or by the waveguide system 1006.

[0173] In some examples, the diode 1010 may operate to heat (e.g., tune) PCM material 1008, provide modulation, and / or a combination thereof. In some examples, the PCM material 1008 when providing modulation may be driven with up to 3V in reverse bias (e.g., depletion mode) for high baud rate operation based on electro-optical plasma-dispersion effects in silicon.

[0174] FIG. 11 is a schematic illustration of circuitry to perform programming and modulation for a shared PN junction device in accordance with examples herein.

[0175] FIG. 11 includes a series arrangement of three transistors - transistor 1102. transistor 1104, and transistor 1106 - coupled between two reference voltages (e.g., a modulator bias voltage and ground). The diode 1112 is coupled between transistor 1104 and transistor 1102. The heater driver 1108 is coupled to one side of the diode 1112 and may provide a programming signal during operation. The modulation driver 1110 is coupled to another side of the diode 1112 and may provide a modulation pulse during operation.

[0176] The circuitry of FIG. 11 may be utilized to program and / or perform modulation using systems and / or devices described herein, such as the optical devices and / or waveguide systems shown and described with reference to FIG. 1. FIGS. 6A-6B, and / or FIGS. 9A-9B. The circuitry of FIG. 11 may be used to implement and / or may be implemented by the circuitry’ shown and described with reference to FIG. 10 In some examples. For example, the heater driver 1108 may be used to implement and / or may be implemented using heater driver 1016. The modulation driver 1110 may be used to implement and / or may be implemented using driver 1012 of FIG. 10.Docket No. 0077145-08201

[0177] FIG. 11 is exemplary. Additional, fewer, and / or different components may be used in other examples.

[0178] The stack of transistors including transistor 1102, transistor 1104, and transistor 1106 may be provided to bias the diode 1112. The transistor 1102 may receive a programming enable signal at a control input of the transistor 1102 (e.g., the gate of transistor 1102). The programming enable signal may be used to place systems described herein into a programming mode. The programming enable signal may be provided by one or more controllers. The programming enable signal may be used to turn on the transistor 1102. Turning on the transistor 1102 may allow a programming signal to be applied to the diode 1112 from the heater driver 1108. The programming signal may be, for example, one or more programming pulses. In programming mode, the diode 1112 may be in a forward bias mode. Responsive to the programming signal provided by the heater driver 1108, a state of PCM material may be selected and / or a resonant frequency of a waveguide system including the diode 1112 may be selected.

[0179] The transistor 1104 and / or transistor 1106 may receive a modulation enable signal at a control input of transistor 1104 and / or transistor 1106. The modulation enable signal may be used to place systems described herein into modulation mode. The modulation enable signal may be provided by one or more controllers. The modulation enable signal may be used to turn on transistor 1104 and / or transistor 1106. Turning on transistor 1104 and / or transistor 1106 may allow a modulation signal to be applied to the modulation input 112. The modulation signal may be a modulation pulse. The modulation signal may be provided by modulation driver 1110. The modulation signal may be applied to the diode 1112 in a reverse bias mode. The modulation signal may be used to modulate light output from optical device(s) or waveguide systems including the diode 1112.

[0180] Accordingly, during operation one or more enable signals may be used to place a diode in a forward bias mode. While in forward bias mode, a driver may provide programming signals to select a state of PCM material for a waveguide system and / or to select a resonant frequency of the waveguide system.

[0181] Following programming, one or more enable signals may be used to place the diode in a reverse bias mode. While in reverse bias mode, a driver may provide modulation signals to modulate light output from the waveguide system. The modulation may occur in accordance with data to be communicated from the waveguide system.

[0182] Examples of controllers described herein, such as controller 1020 of FIG. 10 may utilize a slower clock rate compared to a digital controller clock (e.g., 100MHz). In someDocket No. 0077145-08201examples, the controller 1020 may be a low power controller. In some examples, the controller may consume 10 fj / b. In some examples, the controller may be time-multiplexed among multiple channels to amortize power and area. The actual number of channels a controller may depend on controller clock frequency, tuning-loop latency, number of MRMs, wavelength plan (e.g., DWDM grid), system architecture, and / or combinations thereof. In some examples, there may be 16 channels. In some examples, there may be 4, 8, 16, 32, 64, 128, or 256 channels.

[0183] In some embodiments, tuning may occur to align laser and devices. In some examples, a tuning loop may occur when a new laser module is added. In some examples, a tuning loop may occur when a new laser module is replaced.

[0184] Examples described herein accordingly may include waveguide systems which may be controlled. The waveguide systems (e.g., MRM systems described herein) may be used to form optical interconnects, such as DWDM optical interconnects. The interconnects may be used to transmit data, using optical energy, from one location to another (e.g., from one chip to another and / or from one package to another).

[0185] From the foregoing it will be appreciated that, although specific embodiments have been described herein for purposes of illustration, various modifications may be made w hile remaining with the scope of the claimed technology.

[0186] Various operations may be described as multiple discrete operations in turn, in a manner that may be helpful in understanding embodiments; however, the order of description should not be construed to imply that these operations are order dependent.

[0187] The description may use the terms “embodiment” or “embodiments,” which may each refer to one or more of the same or different embodiments. Furthermore, the terms “comprising.” “including,” “having,” and the like, as used with respect to embodiments, are synonymous.

[0188] As used herein, the term “substantially” in reference to a given parameter, property , or condition means and includes to a degree that one of ordinary skill in the art would understand that the given parameter, property, or condition is met with a small degree of variance, such as, for example, within acceptable manufacturing tolerances. By way of example, depending on the particular parameter, property, or condition that is substantially met, the parameter, property, or condition may be at least 90% met, at least 95% met. or even at least 99% met.

[0189] Examples described herein may refer to various components as “coupled” or signals as being “provided to” or “received from” certain components. It is to be understood that inDocket No. 0077145-08201some examples, the components are directly coupled one to another, while in other examples the components are coupled with intervening components disposed between them. Similarly, signal may be provided directly to and / or received directly from the recited components without intervening components, but also may be provided to and / or received from the certain components through intervening components.

Claims

Docket No. 0077145-08201CLAIMSWhat is claimed is:

1. An apparatus comprising:an input port configured to receive optical input energy;a modulation input configured to receive modulation data;an optical device coupled to the input port and the modulation input, the optical device comprising:one or more waveguides;a phase-change material (PCM) coating at least a portion of the waveguide; and an oxide cladding;wherein either the PCM or the cladding, or both have a negative thermo-optic coefficient, andwherein the optical device is configured to utilize the waveguide to modulate the input energy in accordance with the modulation data to generate modulated output energy; and an output port coupled to the optical device, the output port configured to provide the modulated output energy.

2. The apparatus of claim 1, wherein the optical device comprises a micro-ring modulator.

3. The apparatus of claim 1, wherein control of the PCM material is used to configure the resonant wavelength of the optical device.

4. The apparatus of claim 3, wherein the control is applied through the modulation input.

5. The apparatus of claim 3, wherein the control is applied through separate input from the modulation input.

6. The apparatus of claim 1 wherein the waveguide comprises silicon.

7. The apparatus of claim 1 wherein an effective thermo-optic coefficient of an optical mode in the waveguide is reduced due to the phase-change material, the cladding, or both.

8. The apparatus of claim 1, wherein the phase-change material has a thickness, based in part on a width of the waveguide, to cause the effective thermo-optic coefficient of an optical mode in the waveguide to be zero.

9. The apparatus of claim 1, wherein the oxide cladding comprises TiCh.Docket No. 0077145-0820110. The apparatus of claim 1. wherein the oxide cladding comprises one or more layers.

11. The apparatus of claim 10, wherein the oxide cladding comprises a first material and a second material, the first material is an AI2O3, HfCh, SiN, SiCh, or combinations thereof.

12. The apparatus of claim 2, wherein the micro-ring resonator further comprises:a diode, the diode configured to program the phase-change material in a forward bias mode and to perform the modulation in a reverse bias mode.

13. The apparatus of claim 2, wherein the micro-ring resonator further comprises:a first diode configured to program the phase-change material; anda second diode configured to perform the modulation.

14. The apparatus of claim 13, wherein the diode comprises a P-N junction or a P-i-N junction.

15. The apparatus of claim 1, wherein the waveguide, the PCM layer, and oxide cladding are configured such that the TOC of the waveguide is substantially zero over a target temperature range.

16. The apparatus of claim 1, wherein the thicknesses and TOCs of the core material, the PCM layer, and the cladding layer are configured such that the effective thermo-optic coefficient of the optical waveguide structure is substantially zero over a target temperature range.

17. The apparatus of claim 1, further comprising:a control interface configured to transition the PCM coating between at least two states to tune the phase or resonant wavelength of the optical waveguide structure.

18. The apparatus of claim 17, wherein the optical device comprises a micro-ring resonator, and wherein the control interface is configured to shift the resonant peak of the micro-ring resonator by altering the state of the PCM layer without increasing the thermal sensitivity of the resonator19. The apparatus of claim 2, wherein the micro-ring modulator comprises a Mach-Zehnder Interferometer (MZI) having at least two arms, wherein at least one arm includes the PCM layer and a cladding with a negative TOC, enabling athermal phase-shifting for switching or modulation.Docket No. 0077145-0820120. The apparatus of claim 1, wherein the one or more waveguides are arranged in a series configuration to form a Wavelength Division Multiplexing (WDM) multiplexer, wherein each structure is independently tunable via its respective PCM layer.

21. The apparatus of claim 1. wherein the one or more waveguides are arranged in a parallel configuration to form a Wavelength Division Multiplexing (WDM) demultiplexer, wherein each structure is independently tunable via its respective PCM layer.

22. A method comprising:applying a control signal to a PCM material included in an optical device to place the PCM material into a particular state, wherein either the PCM or a cladding of the optical device, or both have a negative thermo-optic coefficient; andutilizing the optical device to modulate input optical energy in accordance with data while the PCM material remains persistently in the particular state.

23. The method of claim 22, wherein the particular state remains persistently defined by the state of the PCM without continuous application of the control signal.

24. The method of claim 22, wherein the optical device comprises a micro-ring modulator.

Images