UNITED STATES DISTRICT COURT
EASTERN DISTRICT OF TEXAS
MARSHALL DIVISION
APPENDIX OF EXTRINSIC EVIDENCE IN SUPPORT OF PLAINTIFF OPTIWAVE SEMICONDUCTOR CORP.'S OPENING CLAIM CONSTRUCTION BRIEF
Filed: November 15, 2024
Alan Ngo (Cal. Bar No. 247813) ango@calderwoodharkness.com Marcus Delgado (Cal. Bar No. 261504) mdelgado@calderwoodharkness.com Wendy Schultz (Cal. Bar No. 305119) wschultz@calderwoodharkness.com Omar Siddiqui (D.C. Bar No. 1089322) osiddiqui@calderwoodharkness.com CALDERWOOD & HARKNESS LLP 350 South Grand Avenue, 48th Floor Los Angeles, California 90071 Telephone: (213) 555-4200 Facsimile: (213) 555-4201
1201 F Street NW, Suite 700 Washington, D.C. 20004 Telephone: (202) 555-8100 Facsimile: (202) 555-8101
James "Jim" Callahan (Tex. Bar No. 03871542) jcallahan@whitmorecallahan.com WHITMORE & CALLAHAN LLP 110 North College Avenue, Suite 800 Tyler, Texas 75702 Telephone: (903) 555-6300 Facsimile: (903) 555-6301
Attorneys for Plaintiff Optiwave Semiconductor Corp.
LIST OF EXTRINSIC EVIDENCE EXHIBITS
CROSS-REFERENCE TABLE: EXHIBITS TO DISPUTED CLAIM TERMS
INTRODUCTION
Plaintiff Optiwave Semiconductor Corp. ("OWS" or "Plaintiff") respectfully submits this Appendix of Extrinsic Evidence in support of its Opening Claim Construction Brief, filed concurrently herewith on November 15, 2024. This Appendix compiles, organizes, and presents for the Court's convenience the extrinsic evidence referenced in OWS's Opening Brief in connection with the construction of the twelve remaining disputed claim terms at issue in this action.
As set forth in the Joint Claim Construction Statement filed by the parties on October 28, 2024, the parties initially identified fourteen claim terms as potentially disputed. During the course of the meet-and-confer process, the parties reached agreement on the construction of two of those terms—"optical signal path" (as used in claims 1 and 9 of U.S. Patent No. 10,234,847) and "transceiver module" (as used in claims 1, 3, and 7 of U.S. Patent No. 10,891,193)—leaving twelve terms for the Court's resolution at the Markman hearing currently scheduled for February 7, 2025. This Appendix addresses extrinsic evidence pertinent to each of those twelve disputed terms.
The extrinsic evidence compiled herein consists of four categories of materials: (1) technical dictionary definitions from leading reference works in the fields of photonics engineering, semiconductor technology, optoelectronics, and thermal management; (2) excerpts from relevant IEEE standards and glossary documents reflecting the consensus understanding of the professional engineering community; (3) textbook excerpts from authoritative treatises on silicon photonics, waveguide design, thermal management, and grating coupler architectures; and (4) peer-reviewed journal articles from leading scientific and engineering publications, including the Journal of Lightwave Technology, IEEE Photonics Technology Letters, Nature Photonics, Optics Express, and other recognized journals in the field.
The extrinsic evidence assembled in this Appendix is submitted consistent with the framework established by the Federal Circuit in Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005) (en banc). As the Phillips court explained, while intrinsic evidence—the claims, specification, and prosecution history—is the primary tool for claim construction, extrinsic evidence such as "dictionaries, learned treatises, and expert testimony" may be used to assist the Court "in understanding the technology and the meaning of terms to one of skill in the art." Id. at 1318–19. The Court may consider extrinsic evidence to "shed useful light on the relevant art," provided that such evidence is not used to contradict the meaning of claim terms as informed by the intrinsic record. Id. at 1319; see also Ridgeway & Co. v. Conceptronic, Inc., 90 F.3d 1576, 1584 (Fed. Cir. 1996). This Appendix is offered in that spirit—to illuminate the objective understanding of a person of ordinary skill in the art ("POSITA") regarding each of the disputed claim terms.
As defined in the Declaration of Dr. Caroline Ashford, Ph.D., filed concurrently herewith, a POSITA for purposes of the patents-in-suit is a person with a Master of Science or Doctor of Philosophy degree in electrical engineering, photonics, or a closely related discipline, and at least three years of professional or research experience in the design of silicon-photonic integrated circuits. Dr. Ashford's declaration explains the basis for this definition and applies it in analyzing the twelve disputed claim terms. The Declaration of Dr. Jian-Wei Huang, Ph.D., also filed concurrently, provides additional technical analysis and opinions on the remaining disputed terms from the perspective of a POSITA. Both declarations reference materials compiled in this Appendix.
This Appendix of Extrinsic Evidence should be read in conjunction with OWS's Opening Claim Construction Brief, the Appendix of Intrinsic Evidence (filed concurrently), and the Declarations of Dr. Ashford and Dr. Huang. Together, these filings present a comprehensive evidentiary record for the Court's consideration in construing the disputed claim terms.
SECTION I
INDEX OF DISPUTED CLAIM TERMS AND CORRESPONDING EXTRINSIC EVIDENCE
The following master cross-reference table identifies each of the twelve disputed claim terms, the patent or patents in which each term appears, the specific claims at issue, and the exhibit numbers of the extrinsic evidence cited in OWS's Opening Claim Construction Brief in support of its proposed construction of each term. This table is intended to provide the Court with a single reference point for identifying the extrinsic evidence relevant to each disputed term.
Table of Disputed Terms and Supporting Extrinsic Evidence Exhibits
The Court should note that certain patent-family relationships are relevant to the interpretation of related terminology across the patents-in-suit. Specifically, the '847 Patent and the '562 Patent are in the same patent family, with the '847 Patent being a continuation of Application No. 14/927,351, from which the '562 Patent issued. The '193 Patent and the '015 Patent share a common priority document, Provisional Application No. 62/499,117, filed April 7, 2017. These familial relationships inform how a POSITA would understand related terminology across the patents, as terms used consistently across related patents should be construed consistently. See NTP, Inc. v. Research in Motion, Ltd., 418 F.3d 1282, 1293 (Fed. Cir. 2005) (noting that related patents may inform claim construction).
SECTION II
TECHNICAL DICTIONARY DEFINITIONS
This Section presents the technical dictionary definitions referenced in OWS's Opening Claim Construction Brief. For each disputed term, OWS has identified definitions from recognized reference works in the relevant technical fields that confirm the ordinary meaning of the term as understood by a POSITA. These definitions are organized by disputed claim term. The full excerpted pages from each reference work are reproduced behind the corresponding exhibit tab.
Subsection II.A: Definitions Relevant to "Waveguide Multiplexer"
('847 Patent, Claims 1, 3, 5, 9, 14)
The '847 Patent, titled "Multi-Wavelength Silicon Photonic Transceiver with Waveguide-Based Multiplexing," was filed on September 14, 2016 and issued on March 19, 2019. The patent claims a silicon photonic transceiver architecture employing a "waveguide multiplexer" to combine optical signals of different wavelengths onto a single output waveguide. The parties dispute the scope of the term "waveguide multiplexer" as it appears in independent claims 1, 9, and 14, as well as dependent claims 3 and 5.
OWS proposes that "waveguide multiplexer" be construed to mean "an optical device that combines multiple optical signals of different wavelengths onto a single waveguide." Defendant Lumicore Technologies, Inc. ("Lumicore") proposes the narrower construction "an arrayed waveguide grating that combines optical signals using a slab coupler and phase-shifted waveguide array." As set forth in OWS's Opening Brief, Lumicore's proposed construction improperly imports a specific embodiment—the arrayed waveguide grating ("AWG")—into a claim term that the intrinsic record uses as a genus term encompassing multiple species of multiplexer architectures. The extrinsic evidence presented below confirms that a POSITA at the time of the '847 Patent filing would have understood "waveguide multiplexer" to be a broad term encompassing multiple architectures, not limited to AWGs.
Exhibit EE-1: *The Photonics Engineering Dictionary* (3rd ed. 2015). Full Bibliographic Citation:
Reginald T. Hartwell & Susan M. Okafor, eds., The Photonics Engineering Dictionary (3rd ed., Prentice-Hall Technical Reference, 2015), p. 847.
Relevant Excerpt:
"Waveguide multiplexer. An optical device that combines multiple optical signals of different wavelengths onto a single waveguide. Waveguide multiplexers are commonly employed in wavelength-division multiplexing (WDM) systems to aggregate multiple data channels for transmission over a shared optical path. Common implementations include arrayed waveguide gratings (AWGs), echelle diffraction gratings, cascaded Mach-Zehnder interferometers (MZIs), and ring-resonator-based multiplexers. The term is generic and encompasses any waveguide-based device performing the multiplexing function, regardless of the specific optical mechanism employed."
Relevance to Disputed Term: This definition directly supports OWS's proposed construction of "waveguide multiplexer" as "an optical device that combines multiple optical signals of different wavelengths onto a single waveguide." The definition is drawn from a leading technical dictionary published one year prior to the '847 Patent filing date and thus reflects the understanding of a POSITA at the relevant time. Critically, the dictionary expressly states that the term is "generic" and "encompasses any waveguide-based device performing the multiplexing function, regardless of the specific optical mechanism employed." This language directly contradicts Lumicore's proposed construction, which would limit the term to a single implementation—the AWG. The dictionary's enumeration of AWGs, echelle gratings, cascaded MZIs, and ring-resonator-based multiplexers as "common implementations" confirms that the term serves as a genus encompassing multiple species.
Exhibit EE-2: *IEEE Standard Dictionary of Electrical and Electronics Terms* (7th ed. 2014). Full Bibliographic Citation:
Institute of Electrical and Electronics Engineers, IEEE Standard Dictionary of Electrical and Electronics Terms (7th ed., IEEE Press, 2014), p. 712.
Relevant Excerpt:
"Multiplexer (optical). A device that combines two or more optical signals, typically at different wavelengths, onto a common output port or waveguide. In wavelength-division multiplexing (WDM) systems, the multiplexer performs the function of wavelength combining, directing each input wavelength channel to a shared output. The device may be implemented in free-space optics, fiber optics, or integrated waveguide platforms. When implemented in a waveguide platform, the device is commonly referred to as a 'waveguide multiplexer.' See also: arrayed waveguide grating; echelle grating; wavelength-division multiplexing; optical add-drop multiplexer."
Relevance to Disputed Term: The IEEE dictionary definition confirms that the term "multiplexer" in the optical context refers broadly to any device that combines optical signals at different wavelengths onto a common output, and that when such a device is implemented on a waveguide platform, it is referred to as a "waveguide multiplexer." The definition does not limit the term to any single implementation architecture. Instead, it cross-references multiple specific implementations—including AWGs and echelle gratings—as distinct entries, reinforcing the understanding that "waveguide multiplexer" is a higher-level genus term. This is consistent with OWS's proposed construction and inconsistent with Lumicore's attempt to equate "waveguide multiplexer" with "arrayed waveguide grating." The IEEE dictionary is a universally recognized reference in the electrical engineering community and would be a standard resource consulted by a POSITA.
Exhibit EE-3: *Handbook of Optoelectronics and Fiber Optics* (2nd ed. 2016). Full Bibliographic Citation:
Lawrence K. Chen & Maria A. Roldán, eds., Handbook of Optoelectronics and Fiber Optics (2nd ed., Cobalt Ridge & Co.-Interscience, 2016), pp. 423–425.
Relevant Excerpt:
"Waveguide-based multiplexing approaches represent a critical building block in integrated photonic systems for telecommunications and data communications. The term 'waveguide multiplexer' is used in the art as an umbrella term to describe any integrated optical device that combines multiple wavelength channels into a single output waveguide. The principal architectures employed for waveguide multiplexing include:
(1) Arrayed waveguide gratings (AWGs): These devices utilize a phased array of waveguides with carefully designed path-length differences to achieve constructive interference at the output for each wavelength channel. AWGs are the most widely deployed waveguide multiplexer architecture in commercial WDM systems.
(2) Echelle diffraction gratings: Planar grating structures etched into the waveguide layer that use diffraction to separate or combine wavelength channels. Echelle gratings offer advantages in compactness relative to AWGs for certain channel counts.
(3) Cascaded Mach-Zehnder interferometers (MZIs): A series of interferometric stages, each tuned to separate or combine a specific wavelength pair, cascaded to handle multiple channels. MZI-based multiplexers offer reconfigurability advantages.
(4) Ring-resonator-based multiplexers: Arrays of microring resonators, each resonant at a specific wavelength, that selectively couple individual wavelength channels from separate input waveguides onto a common bus waveguide.
Each of these architectures performs the same fundamental function—wavelength multiplexing in the waveguide domain—and all are properly classified under the generic term 'waveguide multiplexer.' "
Relevance to Disputed Term: This excerpt from a leading optoelectronics handbook, published in the same year as the '847 Patent filing, expressly describes "waveguide multiplexer" as an "umbrella term" encompassing AWGs, echelle gratings, cascaded MZIs, and ring-resonator multiplexers. The passage enumerates four distinct architectures, all classified under the same generic term. This directly supports OWS's proposed construction and undermines Lumicore's effort to narrow the term to AWGs alone. As Dr. Ashford explains in her declaration, a POSITA reading the '847 Patent at the time of its filing would have understood "waveguide multiplexer" in precisely this broad sense—as a functional term describing the wavelength-combining function, not a structural term dictating a specific implementation.
Subsection II.B: Definitions Relevant to "Integrated on a Silicon Substrate"
('193 Patent, Claims 1, 3, 7, 10, 15, 18)
The '193 Patent, titled "Coherent Optical Transceiver with Silicon Photonic Integration," was filed on April 7, 2017 (claiming priority to Provisional Application No. 62/499,117, filed April 7, 2017) and issued on January 12, 2021. The patent claims a coherent optical transceiver in which various optical and electronic components are "integrated on a silicon substrate." The parties dispute whether the term "integrated on a silicon substrate" encompasses only monolithic fabrication or also includes heterogeneous and hybrid integration approaches.
OWS proposes that "integrated on a silicon substrate" be construed to mean "formed on or within a silicon-based wafer using semiconductor fabrication processes." Lumicore proposes the narrower construction "monolithically fabricated on a single silicon die without hybrid bonding or heterogeneous integration." As demonstrated by the extrinsic evidence below, Lumicore's proposed construction is inconsistent with the ordinary meaning of the term as understood by a POSITA.
Exhibit EE-7: *Silicon Photonics: An Introduction* (2013 ed.). Full Bibliographic Citation:
Graham T. Predding & Andrew P. Knights, Silicon Photonics: An Introduction (Crowne Albany Industries, 2013), pp. 12–14.
Relevant Excerpt:
"Silicon photonics encompasses the design and fabrication of photonic devices and systems using silicon as the primary platform material. The phrase 'integrated on a silicon substrate' refers broadly to devices that are formed on or within a silicon-based wafer using semiconductor fabrication processes. This definition encompasses several distinct integration paradigms:
Monolithic integration: All photonic and electronic components are fabricated directly on a single silicon wafer using a unified CMOS-compatible process flow. This approach offers the highest potential for integration density and manufacturing scalability, but it is subject to material constraints—silicon's indirect bandgap limits its utility as a light source.
Heterogeneous integration: Non-silicon materials (such as III-V compound semiconductors, germanium, or lithium niobate) are bonded to a silicon substrate at the wafer or die level, and the bonded materials are subsequently processed into functional devices. The resulting structure is considered 'integrated on a silicon substrate' because the silicon wafer serves as the mechanical and optical platform.
Hybrid integration: Separately fabricated components—such as laser dies, photodetector chips, or electronic ICs—are assembled onto a silicon photonic substrate using flip-chip bonding, micro-transfer printing, or other packaging techniques. Although the components are not fabricated monolithically, the final assembly is 'integrated on a silicon substrate' in the sense that the silicon substrate serves as the common platform for all functional elements.
In the literature and in industry practice, all three approaches are recognized as valid forms of 'integration on a silicon substrate.' "
Relevance to Disputed Term: This foundational textbook defines "integrated on a silicon substrate" as encompassing monolithic, heterogeneous, and hybrid integration approaches. The textbook was published in 2013, four years before the '193 Patent filing, and represents the state of knowledge that a POSITA would have brought to the patent's claims. The definition directly supports OWS's proposed construction and refutes Lumicore's attempt to limit the term to monolithic fabrication. As the textbook explains, the key criterion is that devices are "formed on or within a silicon-based wafer using semiconductor fabrication processes"—precisely the construction OWS proposes. Lumicore's exclusion of hybrid bonding and heterogeneous integration from the scope of "integrated on a silicon substrate" contradicts the established usage in the art.
Exhibit EE-8: *The Semiconductor Technology Handbook* (4th ed. 2016). Full Bibliographic Citation:
Richard P. Vandermeir, The Semiconductor Technology Handbook (4th ed., Sterling Reach Industries Professional, 2016), pp. 568–571.
Relevant Excerpt:
"Integration of optical and electronic devices on silicon substrates has been pursued through multiple approaches, each representing a valid path toward realizing high-performance photonic integrated circuits (PICs). The term 'integrated on a silicon substrate' is used throughout the semiconductor industry to describe any configuration in which functional devices are formed on, bonded to, or assembled upon a silicon wafer that serves as the primary structural and electrical platform.
Monolithic integration represents the ideal of a single, unified fabrication process, but practical considerations—including the difficulty of achieving high-quality light emission from silicon—have driven the development of heterogeneous and hybrid integration techniques. In heterogeneous integration, epitaxial layers of III-V materials are bonded to processed or unprocessed silicon wafers, followed by lithographic patterning and etching to define the III-V devices in situ. In hybrid integration, separately fabricated optical and electronic dies are assembled onto a silicon interposer or photonic substrate using solder bumps, copper pillars, or adhesive bonding.
All of these paradigms produce devices that are properly described as 'integrated on a silicon substrate.' The distinguishing feature is the use of a silicon wafer as the integration platform, not the specific technique by which individual components are formed or attached."
Relevance to Disputed Term: This excerpt from a leading semiconductor technology reference work confirms that the term "integrated on a silicon substrate" is not limited to monolithic fabrication. The handbook explicitly identifies heterogeneous integration and hybrid integration as valid approaches that fall within the scope of the term. Published one year before the '193 Patent filing date, this reference reflects the contemporaneous understanding of a POSITA. Lumicore's proposed construction—which would exclude hybrid bonding and heterogeneous integration—contradicts the plain meaning of the term as defined in this authoritative source.
Exhibit EE-9: IEEE Standards Glossary for Photonic Integrated Circuits (2017). Full Bibliographic Citation:
Institute of Electrical and Electronics Engineers, IEEE Standards Glossary for Photonic Integrated Circuits (IEEE Photonics Society, 2017), p. 34.
Relevant Excerpt:
"Integration on a silicon substrate. The process of forming, bonding, or assembling photonic and/or electronic devices on a silicon wafer that serves as the primary platform for the integrated circuit. Integration on a silicon substrate may be achieved through monolithic fabrication, heterogeneous integration (e.g., wafer bonding of III-V materials to silicon), or hybrid integration (e.g., flip-chip assembly of discrete components onto a silicon photonic interposer). The term is not limited to any single integration methodology."
Relevance to Disputed Term: The IEEE glossary definition is directly on point and explicitly states that "integration on a silicon substrate" is "not limited to any single integration methodology." This glossary was published by the IEEE Photonics Society in the same year as the '193 Patent's earliest effective filing date and reflects the consensus understanding of the engineering standards community. It confirms that a POSITA would have understood the term to encompass monolithic, heterogeneous, and hybrid integration—consistent with OWS's proposed construction.
Subsection II.C: Definitions Relevant to "Micro-Channel Heat Sink"
('562 Patent, Claims 1, 4, 6, 11)
The '562 Patent, titled "Thermally Managed Photonic Integrated Circuit with Embedded Micro-Channel Cooling," was filed on November 22, 2014 and issued on June 13, 2017. The patent was originally assigned to Solara Optics, Inc. and was acquired by OWS pursuant to an Asset Purchase Agreement executed on June 15, 2018 for $34.5 million (covering a portfolio of twelve patents including the '562 Patent). The patent claims a photonic integrated circuit with an embedded "micro-channel heat sink" for thermal management. The parties dispute the scope of the term "micro-channel heat sink."
OWS proposes that "micro-channel heat sink" be construed to mean "a heat-dissipating structure containing channels with a hydraulic diameter of less than 1 mm." Lumicore proposes the narrower construction "a closed-loop liquid-cooling system with channels having a hydraulic diameter between 10 µm and 200 µm." As set forth below, the extrinsic evidence demonstrates that Lumicore's proposed construction improperly imports two limitations not found in the claim language or the intrinsic record: (1) a requirement that the heat sink operate as a "closed-loop liquid-cooling system," and (2) a narrow hydraulic diameter range (10 µm to 200 µm) that is far more restrictive than the conventional definition of "micro-channel."
Exhibit EE-13: *Thermal Management in Microelectronics: A Reference Guide* (2nd ed. 2015). Full Bibliographic Citation:
David B. Tuckerman & Michael J. Ellsworth, Thermal Management in Microelectronics: A Reference Guide (2nd ed., Aldermere Partners LLC, 2015), pp. 211–213.
Relevant Excerpt:
"Micro-channel heat sink. A heat-dissipating structure containing channels with a hydraulic diameter of less than 1 mm (1000 µm). Micro-channel heat sinks exploit the high surface-area-to-volume ratio of small-diameter channels to achieve efficient heat transfer from a solid substrate to a coolant fluid. The coolant may be a gas (e.g., air in forced-convection micro-channel heat sinks) or a liquid (e.g., water, fluorocarbon, or dielectric fluid). Micro-channel heat sinks are classified as:
• Passive: Heat dissipation occurs via natural convection or conduction through the channel structure without active fluid circulation.
• Active, open-loop: A coolant fluid is pumped through the channels in a single pass and is not recirculated.
• Active, closed-loop: A coolant fluid is continuously recirculated through the channels by a pump, with heat rejected through a remote heat exchanger.
All three configurations are recognized as 'micro-channel heat sinks' in the art, provided the channel hydraulic diameter is less than 1 mm."
Relevance to Disputed Term: This definition from a leading thermal management reference directly supports OWS's proposed construction. The definition sets the upper bound for micro-channel hydraulic diameter at 1 mm (1000 µm), not at the 200 µm ceiling proposed by Lumicore. Moreover, the definition expressly encompasses passive, open-loop, and closed-loop configurations, thereby refuting Lumicore's proposed requirement of a "closed-loop liquid-cooling system." The reference was published one year after the '562 Patent filing date, reflecting the understanding of a POSITA at the relevant time.
Exhibit EE-14: *Fundamentals of Heat Dissipation in Integrated Circuits* (2014). Full Bibliographic Citation:
Priya R. Narayan & Thomas L. Bergman, Fundamentals of Heat Dissipation in Integrated Circuits (Springer, 2014), pp. 187–189.
Relevant Excerpt:
"Micro-channel heat sinks represent a versatile class of thermal management structures suitable for high-heat-flux applications in semiconductor and photonic devices. The term 'micro-channel heat sink' describes a heat sink in which the coolant channels have a hydraulic diameter in the micro-scale range—conventionally defined as less than 1 mm. The coolant flowing through these channels may be air, a single-phase liquid (such as deionized water or a fluorinated fluid), or a two-phase fluid undergoing phase change within the channels. The choice of coolant and flow configuration (open-loop, closed-loop, or passive) depends on the thermal requirements and system-level constraints, but all such configurations are properly classified as micro-channel heat sinks so long as the channel dimensions fall within the micro-scale regime."
Relevance to Disputed Term: This excerpt confirms that micro-channel heat sinks can operate with air, single-phase liquid, or two-phase fluid, and are not limited to closed-loop liquid systems as Lumicore contends. The textbook was published in the same year as the '562 Patent filing and is contemporaneous with the POSITA's understanding at the relevant time. The definition is consistent with OWS's proposed construction and contradicts Lumicore's narrower proposal.
Exhibit EE-15: ASME Standard Terminology for microscale Heat Transfer (2013). Full Bibliographic Citation:
American Society of Mechanical Engineers, ASME Standard Terminology for microscale Heat Transfer (ASME, 2013), p. 18.
Relevant Excerpt:
"Micro-channel. A channel with a hydraulic diameter less than 1 mm. See also: mini-channel (hydraulic diameter between 1 mm and 3 mm); macro-channel (hydraulic diameter greater than 3 mm). The term 'micro-channel heat sink' refers to a heat sink incorporating one or more micro-channels through which a coolant may flow for purposes of thermal management."
Relevance to Disputed Term: The ASME standard establishes the conventional engineering definition of "micro-channel" as a channel with a hydraulic diameter less than 1 mm. This definition is consistent with OWS's proposed construction and directly contradicts Lumicore's proposed upper limit of 200 µm. The ASME standard is a recognized authority in the mechanical and thermal engineering community and would be among the first resources consulted by a POSITA seeking the ordinary meaning of the term "micro-channel."
Subsection II.D: Definitions Relevant to "Polarization-Diverse Grating Coupler"
('015 Patent, Claims 1, 5, 8, 12, 16)
The '015 Patent, titled "Polarization-Diverse Grating Coupler for High-Bandwidth Silicon Photonic Transceivers," was filed on August 30, 2019 and issued on August 2, 2022. The patent claims a silicon photonic transceiver incorporating a "polarization-diverse grating coupler" for coupling light from an optical fiber to on-chip waveguides while handling both orthogonal polarization states. The parties dispute the scope of the term.
OWS proposes that "polarization-diverse grating coupler" be construed to mean "a grating coupler capable of coupling light of two orthogonal polarization states." Lumicore proposes the narrower construction "a two-dimensional grating coupler with a polarization-splitting element that simultaneously couples TE and TM modes." Lumicore's proposed construction imposes two unjustified limitations: (1) it requires a "two-dimensional" grating structure, and (2) it requires "simultaneous" coupling of TE and TM modes. As the extrinsic evidence demonstrates, neither limitation is inherent in the term as understood by a POSITA.
Exhibit EE-19: *Photonic Integrated Circuits: Design Principles* (2018). Full Bibliographic Citation:
Henrik Lund Rasmussen & Yuki Tanaka, Photonic Integrated Circuits: Design Principles (Cambridge University Press, 2018), pp. 302–304.
Relevant Excerpt:
"Polarization-diverse grating coupler. A grating coupler capable of coupling light of two orthogonal polarization states (typically the TE and TM modes, or equivalently the x- and y-polarized modes of an optical fiber). In silicon photonic circuits, single-polarization grating couplers efficiently couple only one polarization state (usually TE), and any power in the orthogonal polarization is lost. A polarization-diverse grating coupler overcomes this limitation by providing coupling paths for both polarization states.
Several architectures exist for achieving polarization diversity in grating couplers:
(a) Two-dimensional (2D) grating couplers: A grating with periodicity in two orthogonal directions, which inherently splits the incoming light into two polarization components directed into two separate on-chip waveguides.
(b) One-dimensional (1D) grating couplers with external polarization rotators: A pair of 1D grating couplers, each optimized for a single polarization, combined with a polarization beam splitter and a polarization rotator to convert the TM-coupled component to TE for on-chip processing.
(c) Hybrid architectures: Combinations of grating elements and polarization management components that achieve polarization diversity through sequential or parallel processing of the two polarization states.
All of these architectures are classified as 'polarization-diverse grating couplers' because they achieve the common function of coupling both orthogonal polarization states from a fiber to on-chip waveguides."
Relevance to Disputed Term: This definition from a leading photonic circuit design textbook directly supports OWS's proposed construction. The definition identifies three distinct architectures—2D grating couplers, 1D grating couplers with polarization rotators, and hybrid architectures—all classified under the same term "polarization-diverse grating coupler." Notably, the definition does not require simultaneous coupling; in architecture (b), the two polarization states are coupled sequentially through separate grating elements. This contradicts Lumicore's proposed construction, which requires "simultaneous" coupling. The textbook was published one year before the '015 Patent filing date and reflects the understanding of a POSITA at the relevant time.
Exhibit EE-20: *Optical Waveguide Theory and Technology* (2017 ed.). Full Bibliographic Citation:
Kazuo Inoue, Optical Waveguide Theory and Technology (rev. ed., Oxford University Press, 2017), pp. 456–458.
Relevant Excerpt:
"Polarization-diverse coupling is a critical requirement for coherent optical receivers and for systems in which the polarization state of the input signal is unknown or variable. In integrated photonics, polarization diversity is commonly achieved at the fiber-to-chip coupling stage using grating couplers designed to handle both polarization states. Two principal approaches are employed:
Approach 1 — Two-dimensional gratings: A 2D grating coupler with a square or near-Beaumont Corporation lattice diffracts incoming light into two orthogonal directions based on polarization. Each diffracted beam enters a separate on-chip waveguide, achieving polarization splitting and coupling in a single element. This approach offers compactness but may suffer from higher insertion loss compared to optimized 1D designs.
Approach 2 — Polarization-diversity circuits using 1D gratings: A polarization beam splitter at the fiber-to-chip interface separates the incoming light into two orthogonal polarization components, each of which is directed to a dedicated 1D grating coupler optimized for that polarization. A polarization rotator converts one component's polarization to match the preferred on-chip mode (typically TE). This approach allows independent optimization of each grating for lower insertion loss, at the cost of increased chip area.
Both approaches constitute 'polarization-diverse grating couplers,' and the choice between them depends on system requirements and design trade-offs."
Relevance to Disputed Term: This textbook excerpt reinforces the understanding that "polarization-diverse grating coupler" encompasses both 2D and 1D grating-based architectures. The passage makes clear that the defining feature is the capability to handle both polarization states, not the specific structural implementation. This is consistent with OWS's proposed construction and contradicts Lumicore's requirement that the term be limited to 2D gratings.
Subsection II.E: Definitions Relevant to Remaining Disputed Terms (Terms 5–12)
The following subsections present technical dictionary definitions for each of the eight remaining disputed claim terms. For each term, OWS identifies the relevant dictionary source, provides a full bibliographic citation and the pertinent excerpt, and explains how the definition supports OWS's proposed construction.
Exhibit EE-25: Definition of "Evanescent Mode Coupler" (Term 5). Full Bibliographic Citation:
Reginald T. Hartwell & Susan M. Okafor, eds., Dictionary of Optoelectronic Engineering (5th ed., Prentice-Hall Technical Reference, 2015), p. 312.
Relevant Excerpt:
"Evanescent mode coupler (also: evanescent field coupler, directional coupler). An optical coupler that transfers optical energy between two adjacent waveguides via overlap of their evanescent fields. The evanescent field of a guided mode extends beyond the physical boundary of the waveguide core into the surrounding cladding. When two waveguides are placed in close proximity, the evanescent fields overlap, enabling power transfer between the waveguides. The coupling efficiency depends on the separation distance, interaction length, and refractive index contrast. Evanescent mode couplers are fundamental building blocks in photonic integrated circuits and are used for power splitting, wavelength filtering, and mode coupling applications."
Relevance to Disputed Term: This definition supports OWS's proposed construction of "evanescent mode coupler" as "an optical coupler that transfers optical energy between adjacent waveguides via evanescent field overlap." The definition does not impose limitations on the specific waveguide geometry, material system, or application context—it is a functional definition based on the physical mechanism of evanescent field coupling. Lumicore's proposed construction, which would require "a directional coupler with a coupling length equal to one-half the beat length for complete power transfer," improperly imports a specific design parameter (coupling length equal to half the beat length) that characterizes only one operating point of the device and is not part of the ordinary meaning of the term.
Exhibit EE-26: Definition of "Optical Phase Array" (Term 6). Full Bibliographic Citation:
James W. Connolly & Arjun Mehta, eds., Photonics Terminology and Standards Handbook (Optical Society of America, 2016), p. 578.
Relevant Excerpt:
"Optical phase array (also: optical phased array, OPA). An array of optical emitting or receiving elements whose relative phase can be controlled—typically by electro-optic, thermo-optic, or mechanical means—to steer, shape, or focus an optical beam. Optical phase arrays are the photonic analog of radio-frequency phased-array antennas and are used in applications including free-space optical communications, LiDAR, optical interconnects, and reconfigurable beam forming. The array elements may be waveguide apertures, grating emitters, or discrete emitting/receiving elements arranged in a one-dimensional or two-dimensional pattern."
Relevance to Disputed Term: This definition supports OWS's proposed construction of "optical phase array" as "an array of optical emitting or receiving elements whose relative phase can be controlled to steer or shape an optical beam." The definition encompasses range element types and phase-control mechanisms, consistent with the broad claim language of the '193 Patent. Lumicore's proposed construction, which would limit the term to "a one-dimensional array of silicon waveguide emitters controlled by thermo-optic phase shifters," improperly imports limitations drawn from a single preferred embodiment described in the '193 Patent specification.
Exhibit EE-27: Definition of "Resonant Cavity Photodetector" (Term 7). Full Bibliographic Citation:
Henrik Lund Rasmussen & Yuki Tanaka, Advanced Nanophotonic Devices (Cambridge University Press, 2017), p. 215.
Relevant Excerpt:
"Resonant cavity photodetector (also: resonant-cavity-enhanced photodetector, RCE-PD). A photodetector incorporating a resonant optical cavity to enhance light absorption at one or more target wavelengths. The resonant cavity—typically a Fabry-Pérot cavity formed by two reflective surfaces bracketing the absorbing layer—causes the optical field to make multiple passes through the absorbing region, thereby increasing the effective absorption length and improving the quantum efficiency at the resonant wavelength. Resonant cavity photodetectors offer the advantages of wavelength selectivity, enhanced responsivity, and potentially higher bandwidth compared to conventional bulk-absorber photodetectors."
Relevance to Disputed Term: This definition supports OWS's proposed construction of "resonant cavity photodetector" as "a photodetector incorporating a resonant optical cavity to enhance light absorption at one or more target wavelengths." The definition does not require a specific type of reflector (e.g., distributed Bragg reflectors vs. metal mirrors) or a specific absorbing material (e.g., germanium vs. InGaAs), consistent with the broad claim language of the '015 Patent. Lumicore's proposed construction would limit the term to "a germanium photodetector with distributed Bragg reflectors forming a Fabry-Pérot cavity," improperly importing material and structural limitations from the preferred embodiment.
Exhibit EE-28: Definition of "Ring Resonator Filter" (Term 8). Full Bibliographic Citation:
Institute of Electrical and Electronics Engineers, IEEE Std 2884-2017 — Standard Terminology for Nanophotonic Circuits (IEEE, 2017), p. 42.
Relevant Excerpt:
"Ring resonator filter. A wavelength-selective filter comprising a ring-shaped optical waveguide evanescently coupled to one or more bus waveguides. The ring resonator selectively transfers optical power at wavelengths that are resonant with the ring (i.e., wavelengths for which the round-trip phase accumulation in the ring is an integer multiple of 2π) from the input bus waveguide to the output (drop) bus waveguide. Non-resonant wavelengths pass through the input bus waveguide unaffected. Ring resonator filters are used in wavelength-division multiplexing systems for channel add/drop, wavelength routing, and spectral filtering."
"Resonant cavity photodetector. See resonant-cavity-enhanced photodetector. A photodetector in which an optical resonant cavity structure is used to enhance the interaction between incident light and the absorbing medium, resulting in improved quantum efficiency at the cavity's resonant wavelength(s)."
Relevance to Disputed Term: The IEEE standard definition of "ring resonator filter" is consistent with OWS's proposed construction: "a wavelength-selective filter comprising a ring-shaped optical waveguide evanescently coupled to one or more bus waveguides." Lumicore's proposed construction would add a requirement that the ring resonator filter include "a tuning element for adjusting the resonant wavelength." The IEEE definition does not include any such requirement—a ring resonator filter may be fixed-wavelength or tunable, and the addition of a tuning element is an optional feature, not a definitional requirement. The definition of "resonant cavity photodetector" in the same standard is also consistent with OWS's proposed construction of Term 7, as noted above.
Exhibit EE-29: Definition of "Electro-Absorption Modulator" (Term 9). Full Bibliographic Citation:
Priya R. Narayan, Principles of Optical Modulation and Detection (Springer, 2016), pp. 134–136.
Relevant Excerpt:
"Electro-absorption modulator (EAM). An optical modulator that modulates light intensity by electrically controlling the optical absorption of a semiconductor material. The modulation mechanism exploits the Franz-Keldysh effect (in bulk semiconductors) or the quantum-confined Stark effect (QCSE, in quantum-well structures), both of which cause the absorption edge of the semiconductor to shift under an applied electric field. When the applied voltage shifts the absorption edge to overlap with the signal wavelength, the material becomes absorbing and the transmitted light intensity decreases. EAMs offer advantages of high modulation bandwidth, low drive voltage, and compact size, and are widely used in high-speed optical communication systems."
Relevance to Disputed Term: This definition supports OWS's proposed construction of "electro-absorption modulator" as "an optical modulator that modulates light intensity by electrically controlling the optical absorption of a semiconductor material." The definition encompasses both Franz-Keldysh and QCSE mechanisms and does not limit the term to any specific material system or device geometry. Lumicore's proposed construction, which would require "a quantum-well structure operating via the quantum-confined Stark effect," improperly excludes bulk-semiconductor EAMs that operate via the Franz-Keldysh effect—devices that are well within the scope of the term as understood by a POSITA.
Exhibit EE-30: Definition of "Mode-Selective Coupler" (Term 10). Full Bibliographic Citation:
Lawrence K. Chen & Maria A. Roldán, eds., Fiber Optic Reference Handbook (8th ed., Cobalt Ridge & Co., 2015), p. 389.
Relevant Excerpt:
"Mode-selective coupler. An optical coupler designed to selectively transfer a particular spatial mode of light between waveguides while suppressing or rejecting other modes. Mode-selective couplers exploit the difference in propagation constants between spatial modes to achieve selective coupling. Applications include mode-division multiplexing (MDM), mode filtering, and mode conversion in multimode waveguide systems."
Relevance to Disputed Term: This definition is consistent with OWS's proposed construction of "mode-selective coupler" as "an optical coupler designed to selectively transfer a particular spatial mode of light between waveguides." Lumicore's proposed construction would require "an asymmetric directional coupler with waveguides of different widths designed to phase-match only the fundamental mode of one waveguide to a higher-order mode of the other." While an asymmetric directional coupler is one specific implementation of a mode-selective coupler, the term itself is broader and encompasses other architectures such as adiabatic couplers, Y-junction mode sorters, and tapered couplers, all of which can achieve mode-selective coupling.
Exhibit EE-31: Definition of "Buried Oxide Cladding Layer" (Term 11). Full Bibliographic Citation:
Andrei V. Petrov, Integrated Photonic Circuit Fabrication (Ashford Industries, 2014), pp. 78–79.
Relevant Excerpt:
"Buried oxide (Box) cladding layer. In silicon-on-insulator (SOI) photonic circuits, the buried oxide layer is the layer of silicon dioxide (SiO₂) formed beneath the silicon waveguide core and above the silicon handle wafer. This layer serves as an optical cladding, confining light within the silicon waveguide core by providing a low-refractive-index boundary (n ≈ 1.45 for SiO₂ vs. n ≈ 3.48 for silicon at 1550 nm). The box layer also provides electrical isolation between the active device layer and the silicon handle wafer. Typical box layer thicknesses for photonic applications range from 1 µm to 3 µm, with thicker layers providing better optical isolation from the high-index silicon handle."
Relevance to Disputed Term: This definition supports OWS's proposed construction of "buried oxide cladding layer" as "a layer of silicon dioxide formed beneath the waveguide core and above the silicon handle wafer, serving as an optical cladding." Lumicore's proposed construction, which would require "a thermally grown silicon dioxide layer with a thickness of exactly 2 µm," improperly imports a specific thickness value drawn from a single embodiment in the '562 Patent specification and improperly limits the formation method to thermal growth when other methods (e.g., SIMOX, wafer bonding) are equally valid.
Exhibit EE-32: Definition of "Germanium-on-Insulator Photodiode" (Term 12). Full Bibliographic Citation:
Hideki Yamamoto & Lisa M. Firth, Nanoscale Silicon Device Engineering (Springer, 2016), pp. 412–414.
Relevant Excerpt:
"Germanium-on-insulator (GOI) photodiode. A photodiode formed from a germanium (Ge) active region disposed on an insulating layer, typically the buried oxide (Box) layer of a silicon-on-insulator (SOI) substrate. Germanium-on-insulator photodiodes exploit the strong optical absorption of germanium in the near-infrared wavelength range (approximately 1.1 µm to 1.6 µm), making them suitable for telecommunications-wavelength detection in silicon photonic circuits. The insulating layer beneath the germanium serves both as an optical cladding and as an electrical isolation layer. GOI photodiodes may be fabricated by selective epitaxial growth of germanium on a patterned SOI wafer or by wafer bonding of a germanium layer onto an oxidized silicon substrate."
Relevance to Disputed Term: This definition supports OWS's proposed construction of "germanium-on-insulator photodiode" as "a photodiode formed from a germanium active region disposed on an insulating layer." The definition encompasses multiple fabrication methods (selective epitaxial growth and wafer bonding) and does not require any specific device geometry (e.g., lateral PIN vs. vertical PIN). Lumicore's proposed construction, which would require "a vertical PIN photodiode with an epitaxially grown germanium absorber on a silicon dioxide buried oxide layer," improperly imports structural and fabrication limitations from a preferred embodiment.
SECTION III
IEEE STANDARDS EXCERPTS
This Section presents excerpts from IEEE standards documents referenced in OWS's Opening Claim Construction Brief. IEEE standards are developed through consensus-based processes involving industry experts and represent the collective understanding of practicing engineers and scientists. As such, they are highly probative of the meaning that a POSITA would ascribe to technical terms. See Tex. Digital Sys., Inc. v. Telegenix, Inc., 308 F.3d 1193, 1203 (Fed. Cir. 2002) (recognizing industry standards as relevant extrinsic evidence for claim construction).
Subsection III.A: IEEE Standards Relevant to Waveguide Multiplexing and Optical Transceiver Technology
Exhibit EE-4: IEEE 802.3-2015 Standard — Ethernet Physical Layer Specifications. Full Bibliographic Citation:
Institute of Electrical and Electronics Engineers, IEEE Std 802.3-2015 — IEEE Standard for Ethernet, Section 5, Clause 95 (400 Gb/s Ethernet) and Clause 120 (Physical Medium Dependent Sublayer for 400 Gb/s over WDM), pp. 642–678 (IEEE, 2015).
Relevant Excerpts:
Clause 95.3.2 — Optical Transceiver Architecture:
"The 400GBASE-FR8 and 400GBASE-LR8 physical layer specifications employ wavelength-division multiplexing (WDM) to combine eight optical lanes, each operating at approximately 50 Gb/s, onto a single fiber. The transmitter module includes a waveguide multiplexer that combines the eight wavelength channels (centered on the wavelengths specified in Table 95-8) onto a common output waveguide connected to the fiber interface. The waveguide multiplexer may be implemented using any architecture capable of combining the specified wavelength channels with the insertion loss, channel isolation, and passband characteristics defined in this standard."
Clause 95.3.2.1 — Multiplexer Implementation Notes:
"This standard does not mandate a specific multiplexer architecture. Compliant implementations may employ arrayed waveguide gratings (AWGs), thin-film filter cascades, echelle gratings, or other waveguide-based multiplexing structures, provided they meet the performance specifications set forth in Table 95-9. The term 'waveguide multiplexer' as used in this standard refers generically to any waveguide-integrated device performing the wavelength-combining function."
Relevance to Disputed Terms: The IEEE 802.3 standard's use of "waveguide multiplexer" is directly relevant to the construction of that term in the '847 Patent. The standard explicitly states that it "does not mandate a specific multiplexer architecture" and that compliant implementations may use AWGs, echelle gratings, or "other waveguide-based multiplexing structures." This confirms that the term "waveguide multiplexer" is understood in the industry—including in the formal standards context—as a generic, architecture-neutral term. Lumicore's proposed construction, which would limit the term to AWGs, is inconsistent with the IEEE standard's express language. The 2015 edition of the standard predates the '847 Patent filing by approximately one year and reflects the contemporaneous industry understanding.
Exhibit EE-5: IEEE Std 2883-2016 — Standard for Silicon Photonics Terminology and Definitions. Full Bibliographic Citation:
Institute of Electrical and Electronics Engineers, IEEE Std 2883-2016 — Standard for Silicon Photonics Terminology and Definitions (IEEE Photonics Society, 2016), pp. 14, 28, 31, 45.
Relevant Excerpts:
Section 3.47 — Waveguide multiplexer:
"A photonic device implemented in a waveguide platform that combines multiple optical signals of different wavelengths onto a single output waveguide. The term encompasses all waveguide-based multiplexer architectures, including but not limited to arrayed waveguide gratings, echelle gratings, ring resonator cascades, and Mach-Zehnder interferometer cascades."
Section 3.22 — Integration on a silicon substrate:
"The fabrication, bonding, or assembly of photonic and/or electronic components on a silicon wafer serving as the primary platform. Integration on a silicon substrate may be monolithic (all components fabricated in a single process flow), heterogeneous (non-silicon materials bonded to silicon at the wafer or die level), or hybrid (separately fabricated components assembled onto a silicon platform)."
Section 3.33 — Polarization-diverse grating coupler:
"A grating coupler architecture designed to couple light of both orthogonal polarization states (TE and TM) from an optical fiber to on-chip waveguides. Implementations include two-dimensional grating structures and combinations of one-dimensional gratings with polarization management elements."
Relevance to Disputed Terms: This IEEE standard, specifically developed for silicon photonics terminology, provides definitions for three of the four principal disputed terms at issue. Each definition is consistent with OWS's proposed constructions and contradicts Lumicore's narrower proposals. The standard was published in 2016, contemporaneous with the '847 Patent filing and one year before the '193 Patent filing, and reflects the consensus understanding of the silicon photonics community at the relevant time.
Exhibit EE-10: IEEE Photonics Standards — Definitions of Integration Approaches. Full Bibliographic Citation:
Institute of Electrical and Electronics Engineers, IEEE Photonics Society Technical Report on Photonic Integration Terminology (IEEE Photonics Society, 2016), pp. 7–12.
Relevant Excerpts:
"The photonic integration community employs three principal approaches for realizing photonic integrated circuits on silicon substrates:
Monolithic integration: All photonic components (waveguides, modulators, photodetectors) and, optionally, electronic components are fabricated on a single silicon wafer using a unified, typically CMOS-compatible, fabrication process. Monolithic integration offers the highest potential for integration density, yield, and cost reduction at scale.
Heterogeneous integration: Materials not native to the silicon platform—most commonly III-V compound semiconductors such as InP or GaAs—are bonded to a silicon substrate at the wafer or die level. After bonding, the non-silicon materials are processed lithographically to define functional devices (lasers, amplifiers, etc.) that are thereby 'integrated on' the silicon substrate.
Hybrid integration: Separately fabricated and tested components (e.g., laser chips, driver ICs) are assembled onto a silicon photonic interposer or substrate using advanced packaging techniques such as flip-chip bonding, micro-bump interconnects, or photonic wire bonding. The resulting multi-chip module is considered 'integrated on a silicon substrate' because the silicon platform provides the optical routing, passive components, and interconnection infrastructure.
All three approaches are recognized as valid forms of 'integration on a silicon substrate' in the photonics community. The choice among them is driven by performance, cost, and manufacturing maturity considerations."
Relevance to Disputed Term: This IEEE technical report provides a comprehensive and authoritative overview of integration approaches on silicon substrates, confirming that the term "integrated on a silicon substrate" encompasses monolithic, heterogeneous, and hybrid integration. The report was published in 2016, one year before the '193 Patent filing, and represents the consensus understanding of the IEEE Photonics Society membership. Lumicore's proposed construction, which would exclude heterogeneous and hybrid integration, is directly contradicted by this authoritative source.
Exhibit EE-21: IEEE Standards Relating to Polarization Handling in Photonic Circuits. Full Bibliographic Citation:
Institute of Electrical and Electronics Engineers, IEEE Photonics Society Standards on Polarization Management in Integrated Photonic Circuits (IEEE, 2017), pp. 22–25.
Relevant Excerpts:
"Section 4.3 — Polarization Diversity at the Fiber-Chip Interface:
Polarization diversity is achieved in integrated photonic circuits by employing grating couplers or edge couplers capable of handling both orthogonal polarization states of light entering from a single-mode optical fiber. A 'polarization-diverse grating coupler' is defined as any grating-based coupling structure that provides coupling paths for both TE- and TM-polarized light.
Section 4.3.1 — Two-Dimensional Grating Couplers: