PRIVILEGED AND CONFIDENTIAL — ATTORNEY WORK PRODUCT
CALDERWOOD & HARKNESS LLP One Embarcadero Center, 34th Floor San Francisco, California 94111
Re: Deposition Preparation Outline — Dr. Kevin Fong, CTO & Co-Founder, Optiwave Semiconductor Corp.; Named Inventor on U.S. Patent Nos. 10,234,567 ("'567 Patent"), 10,891,234 ("'234 Patent"), and 11,312,890 ("'890 Patent")
Deposition Context: Claim construction discovery deposition, to be taken before close of claim-construction-related discovery on October 31, 2022. Opposing counsel: Victoria "Tori" Ashworth and Daniel Kowalski of Brannick Howell & Stein LLP ("BHS").
Right-click to update Table of Contents
I. PURPOSE AND OBJECTIVES
A. Overview
This outline has been prepared to assist in the substantive preparation of Dr. Kevin Fong for his deposition during the claim construction phase of Optiwave Semiconductor Corp. v. Lumos Photonics, Inc., Case No. 5:22-cv-03417-EJD (N.D. Cal.). The deposition is being taken by counsel for Lumos Photonics, Inc. ("Lumos") as part of claim-construction-related discovery permitted under the Patent Local Rules for the Northern District of California. Because this deposition is governed by the Patent Local Rules and the Court's scheduling order, the scope of permissible examination is limited to claim-construction-related topics: the meaning of disputed claim terms as they would be understood by a person of ordinary skill in the art ("POSITA"), the state of the art at the time of each claimed invention, and the relevant prosecution histories of the patents-in-suit. Counsel should object to, and if necessary seek a protective order regarding, any questioning that strays into infringement analysis, damages, willfulness, or other topics reserved for fact and expert discovery.
B. Primary Objectives
The primary objectives for Dr. Fong's deposition are as follows:
1. Elicit testimony consistent with OWS's proposed claim constructions. Dr. Fong, as a named inventor on three of the four asserted patents and a recognized expert in silicon photonics, is uniquely positioned to provide testimony about how a POSITA would understand the disputed claim terms. His testimony should affirmatively support the claim constructions set forth in OWS's Preliminary Claim Construction Contentions (served September 12, 2022) and OWS's Responsive Claim Construction Contentions (served October 3, 2022). In particular, his testimony on the terms "adiabatically coupled," "graded buffer layer having a continuously varying germanium concentration," and "pre-emphasis equalization stage electrically co-integrated with the modulator on a common substrate" should be carefully developed to align with OWS's positions.
2. Prepare Dr. Fong for anticipated cross-examination on prosecution history. Brannick Howell & Stein will attempt to use prosecution history — particularly claim amendments and applicant remarks made during prosecution of the '567 and '890 Patents — to narrow the scope of the disputed terms. Dr. Fong must be prepared to address prosecution history questions without conceding that amendments were made to impose structural limitations beyond those reflected in OWS's proposed constructions. He should understand the contours of the prosecution history estoppel doctrine at a high level so that he can frame his answers appropriately.
3. Ensure consistency with OWS's claim construction positions. Dr. Fong's testimony will in all likelihood be quoted extensively in the parties' opening and responsive claim construction briefs and may be cited at the Markman hearing. It is therefore essential that his testimony be precise, consistent with OWS's legal positions, and not susceptible to being taken out of context to support Lumos's narrower proposed constructions.
C. Markman Hearing Context
The Markman hearing is currently scheduled for January 30, 2023, before the Honorable Edward J. Torvelis. Opening claim construction briefs are due December 12, 2022; responsive briefs are due January 9, 2023; and reply briefs are due January 23, 2023. Dr. Fong's deposition testimony will be available for citation in all briefing rounds. Any damaging admissions made at deposition could be leveraged by Lumos in its opening brief or, more critically, in its responsive brief where OWS would have limited opportunity to rehabilitate the testimony. This underscores the importance of thorough preparation.
D. Dr. Fong's Background — Summary
Dr. Kevin Fong is a co-founder of Optiwave Semiconductor Corp. ("OWS"), which he established in 2014 together with Dr. Priya Ramanathan (CEO). He currently serves as Chief Technology Officer. Dr. Fong holds a Ph.D. in Electrical Engineering with a specialization in photonics from Stanford University and was formerly a principal engineer at a major semiconductor foundry prior to co-founding OWS. His credentials as a person of ordinary skill in the art are exceptionally strong — indeed, they significantly exceed the typical POSITA profile. This is both an advantage and a potential risk: his testimony carries considerable weight, but BHS may attempt to characterize his understanding as that of a leading expert rather than a POSITA, or may try to use his deep technical knowledge to draw him into concessions on technically nuanced points. Dr. Fong should be instructed to frame his testimony consistently in terms of what a person of ordinary skill in the art would understand, not merely what he personally believes.
II. WITNESS BACKGROUND AND CREDENTIALS
A. Education and Professional History
Dr. Kevin Fong received his Ph.D. in Electrical Engineering from Stanford University, with a dissertation focused on integrated photonic devices and waveguide design for silicon-on-insulator platforms. His doctoral work, conducted under the supervision of Professor David A.B. Miller in the Ginzton Laboratory, included pioneering research on mode coupling in heterogeneous waveguide structures, which laid the groundwork for much of the technology later developed at OWS.
Prior to co-founding OWS, Dr. Fong served as a principal engineer at a major semiconductor foundry, where he was responsible for developing photonic process design kits and silicon photonics integration schemes for volume manufacturing. His experience at the foundry provided him with deep familiarity with the practical constraints of semiconductor fabrication, including process uniformity, doping profiles, lithographic resolution, and wafer-scale integration — knowledge that would prove directly relevant to the inventions claimed in the patents-in-suit.
Dr. Fong's publication record includes over forty peer-reviewed journal articles and conference papers in the fields of integrated photonics, silicon photonics device design, and optoelectronic integration. He has been an invited speaker at major industry conferences including the Optical Fiber Communication Conference (OFC), the IEEE International Electron Devices Meeting (IEDM), and the European Conference on Integrated Optics (ECIO). He is a Senior Member of the IEEE.
B. Role at OWS
Dr. Fong co-founded Optiwave Semiconductor Corp. in 2014, together with Dr. Priya Ramanathan, who serves as Chief Executive Officer. OWS is headquartered at 2775 Augustine Drive, Suite 400, San Jose, California 95054. Dr. Fong serves as Chief Technology Officer and in that capacity oversees all technology development for OWS's silicon photonics integrated circuits, including device design, process integration, testing, and validation.
OWS is a fabless semiconductor design company, meaning it designs its chips in-house but contracts fabrication to external foundries. OWS's primary manufacturing partner is Pinnacle Semiconductor Foundry Co., Ltd., which fabricates OWS's designs on a 45nm silicon-on-insulator ("SOI") process at its fabrication facility in Hsinchu, Taiwan. The choice of the 45nm SOI process node is significant to the inventions claimed in the patents-in-suit, as it enables the monolithic co-integration of photonic devices (waveguides, modulators, photodetectors) with electronic circuits (driver amplifiers, transimpedance amplifiers) on a common substrate.
Under Dr. Fong's technical leadership, OWS has developed and commercialized a number of key products, including:
• The OWS-400ZR coherent optical transceiver chipset, targeting 400 Gb/s data-center interconnect applications compliant with the OIF 400ZR Implementation Agreement;
• The OWS-800LR coherent optical transceiver chipset, targeting 800 Gb/s long-reach intra-data-center and campus networking applications; and
• Integrated silicon photonics engines for co-packaged optics ("CPO"), designed for integration with next-generation network switch ASICs.
OWS reported fiscal year 2022 revenue of approximately $187.4 million and currently employs approximately 620 people, including a research and development team of approximately 180 engineers and scientists.
C. OWS Patent Portfolio
OWS's patent portfolio comprises 87 issued U.S. utility patents and 14 pending U.S. utility patent applications. The portfolio is concentrated in three core technology areas: waveguide coupling structures, germanium-on-silicon photodetector architectures, and modulator driver circuit architectures. The four patents asserted in this litigation — the '567, '234, '678, and '890 Patents — are core assets within this portfolio and cover fundamental aspects of OWS's commercial products.
OWS's ownership is structured as follows: approximately 62% of the company's equity is held by Vantage Point Capital Partners Fund IV, LP, which made an initial equity investment of $74.5 million across OWS's Series C and Series D financing rounds. The remaining equity is held by the founders (including Dr. Fong and Dr. Ramanathan), members of the management team, and early angel investors.
Caution: If asked at deposition about the Clearpath IP Valuation Services, Inc. patent portfolio valuation (portfolio-level valuation of $215 million; four asserted patents valued collectively at $42.8 million), this information is privileged and protected work product. Dr. Fong should be instructed not to disclose this information, and defending counsel should assert privilege.
D. Inventorship
Dr. Fong is a named inventor on three of the four patents asserted in this litigation:
• U.S. Patent No. 10,234,567 ("'567 Patent") — "Integrated Waveguide Coupler with Tapered Mode Converter for Silicon Photonics Platforms." Co-inventors: Dr. Priya Ramanathan and Hideo Tanaka.
• U.S. Patent No. 10,891,234 ("'234 Patent") — "Germanium-on-Silicon Photodetector with Graded Buffer Layer for Enhanced Responsivity." Co-inventor: Dr. Amara Osei.
• U.S. Patent No. 11,312,890 ("'890 Patent") — "Driver Circuit Architecture for High-Speed Silicon Photonic Mach-Zehnder Modulators." Co-inventor: Rajesh Iyer.
Dr. Fong is not a named inventor on U.S. Patent No. 11,045,678 ("'678 Patent"), which is titled "Multi-Channel Wavelength Division Multiplexing Transceiver with Integrated Laser Array." The inventors of the '678 Patent are Dr. Priya Ramanathan, Dr. Amara Osei, and Liwei Zhang. Dr. Fong should not be asked to provide inventor testimony regarding the '678 Patent, and any such questions should be objected to on foundation grounds.
E. Preparation Points for Background Testimony
Prior to the deposition, the following preparation points should be reviewed with Dr. Fong:
1. Document Review: Confirm that Dr. Fong has reviewed the complete patent specifications, all asserted claims, and the full prosecution file histories (including all office actions, responses, amendments, applicant remarks, examiner interviews, and notices of allowance) for each of the three patents on which he is a named inventor.
2. POSITA vs. Personal Understanding: Confirm that Dr. Fong understands the distinction between his personal understanding of the claim terms and the legally relevant standard — i.e., the understanding of a person of ordinary skill in the art. His testimony should be framed in terms of POSITA understanding whenever possible, while acknowledging that his personal understanding is informed by his extensive expertise.
3. General Deposition Conduct: Remind Dr. Fong of the following standard deposition guidelines:
• Answer only the question asked. Do not volunteer additional information.
• Listen carefully to the full question before answering. If a question is unclear or compound, ask for clarification.
• Do not guess or speculate. If he does not know the answer to a question, he should say so.
• Do not discuss litigation strategy, attorney communications, settlement discussions, or privileged information.
• If defending counsel objects to a question, wait for the objection to be stated and for instructions before answering.
• If he needs a break, he may request one, but he should not request a break in the middle of a pending question.
4. Protective Order Compliance: Remind Dr. Fong of the Stipulated Protective Order entered by the Court in September 2022, which governs the treatment of confidential information produced by both parties. Dr. Fong should not disclose or discuss any materials designated as "CONFIDENTIAL" or "HIGHLY CONFIDENTIAL — ATTORNEYS' EYES ONLY" by Lumos during the deposition, except in accordance with the terms of the Protective Order.
III. U.S. PATENT NO. 10,234,567 ('567 PATENT) — INTEGRATED WAVEGUIDE COUPLER
A. Background of the Invention
The '567 Patent is titled "Integrated Waveguide Coupler with Tapered Mode Converter for Silicon Photonics Platforms." The application was filed on March 14, 2017, and the patent issued on March 19, 2019. The named inventors are Dr. Kevin Fong, Dr. Priya Ramanathan, and Hideo Tanaka. Hideo Tanaka was employed at OWS as an engineer from approximately 2016 to 2019. Mr. Tanaka's prior employment and prior patent work is discussed in detail in Section III.E below.
The asserted claims of the '567 Patent are Claims 1, 4, 7, 12, and 15. Claim 1 is the only independent claim among the asserted claims; Claims 4, 7, 12, and 15 are dependent claims that further narrow the scope of Claim 1 with additional structural and functional limitations.
The '567 Patent is directed generally to an integrated optical coupler that achieves efficient, low-loss optical coupling between a silicon waveguide and a silicon nitride waveguide on a common photonic chip. This waveguide coupling function is critical in modern silicon photonics platforms because silicon and silicon nitride waveguides have different optical mode profiles, and transitioning light between the two materials without excessive loss is a fundamental engineering challenge.
B. Genesis of the Invention
Dr. Fong should be prepared to testify about the genesis of the invention claimed in the '567 Patent. The technical problem that motivated the invention was the challenge of achieving low-loss optical coupling between a silicon (Si) waveguide and a silicon nitride (SiN) waveguide within an integrated photonics platform. Silicon waveguides offer high index contrast and compact bending radii, making them ideal for dense photonic integration. Silicon nitride waveguides, by contrast, offer lower propagation losses and broader transparency windows, making them preferable for certain functions such as wavelength filtering and delay lines. Many advanced silicon photonics designs require both types of waveguides on the same chip, and efficient coupling between them is essential.
At the time of the invention, existing approaches to Si-to-SiN waveguide coupling suffered from significant insertion losses, particularly at the mode transition interface. The prior art included both direct butt-coupling approaches (which were highly lossy) and evanescent coupling approaches (which required precise alignment tolerances that were difficult to achieve in volume manufacturing). The '567 Patent addresses this problem through the use of a tapered mode converter that achieves an adiabatic optical coupling — that is, a coupling in which the optical mode transitions from the silicon waveguide to the silicon nitride waveguide without significant power loss.
Dr. Fong should be prepared to describe the timeline of conception and reduction to practice of the invention. In particular, he should be able to describe:
• When the technical problem was first identified within OWS's development program;
• When the concept of using an adiabatic taper for Si-to-SiN coupling was first conceived;
• When prototype devices were first fabricated (through Pinnacle Semiconductor Foundry) and tested;
• The results of initial testing and any subsequent design iterations.
Dr. Fong should also be prepared to describe the respective contributions of each co-inventor:
• Dr. Kevin Fong — his contributions to the conception of the adiabatic coupling approach and the overall device architecture;
• Dr. Priya Ramanathan — her contributions (which may include system-level architecture and integration strategy);
• Hideo Tanaka — his specific contributions to the taper design and simulation work.
The role of Hideo Tanaka is an area of particular sensitivity and is discussed further in Sections III.D and III.E below. Tanaka is the inventor of U.S. Patent No. 9,456,012, which issued on September 6, 2016 — before Tanaka joined OWS and before the '567 Patent application was filed. Lumos has cited Tanaka's U.S. Patent No. 9,456,012 as prior art against the '567 Patent under 35 U.S.C. § 102(a)(1). Dr. Fong must be prepared to clearly articulate the distinction between the invention claimed in the '567 Patent and Tanaka's earlier work.
The key differentiator between the '567 Patent and Tanaka's prior U.S. Patent No. 9,456,012 is the adiabatic coupling approach used in the '567 Patent as contrasted with the non-adiabatic coupling approaches described in Tanaka's earlier patent. This distinction is at the heart of the claim construction dispute over the term "adiabatically coupled."
C. Key Disputed Claim Term: "adiabatically coupled"
1. Claim Language
Claim 1 of the '567 Patent (independent) recites, in relevant part:
"An integrated optical coupler comprising a tapered silicon waveguide having a tip width of less than 200 nanometers adiabatically coupled to a silicon nitride waveguide layer…"
The term "adiabatically coupled" is the principal disputed claim term in the '567 Patent. The parties' proposed constructions are as follows:
2. Competing Constructions
OWS's Proposed Construction:
"Coupled such that optical mode transitions occur without significant power loss."
Lumos's Proposed Construction:
"Coupled through a gradual, continuous taper that prevents excitation of higher-order modes throughout the entire transition region."
The difference between these two constructions is significant. OWS's construction defines "adiabatically coupled" in functional terms — the coupling must achieve mode transitions without significant power loss. This functional construction encompasses multiple structural implementations, including both continuous tapers and segmented (discrete-step) tapers, so long as the result is a low-loss mode transition. Lumos's construction, by contrast, imposes a specific structural requirement — that the taper be "gradual" and "continuous" — and adds a specific functional criterion — that higher-order modes must be suppressed "throughout the entire transition region." Lumos's narrower construction is designed to exclude segmented taper designs from the scope of the claims.
3. Preparation of Dr. Fong's Testimony
Dr. Fong should be prepared to testify as follows:
(a) His understanding of "adiabatically coupled" as used in the specification:
Dr. Fong should testify that the '567 Patent specification describes the adiabatic coupling concept in functional terms — i.e., as a coupling mechanism that achieves mode transitions without significant power loss. While the specification includes embodiments that feature a continuous taper, the adiabatic coupling concept is described in the specification as a performance characteristic of the coupler, not as a structural mandate for a particular taper geometry. Dr. Fong should be able to point to specific passages in the specification that describe the adiabatic principle in these functional terms.
(b) POSITA understanding of "adiabatic" in the silicon photonics field:
Dr. Fong should testify that, in the silicon photonics field as understood by a person of ordinary skill in the art at the time of the invention (circa 2016–2017), the term "adiabatic" when applied to waveguide coupling referred to a mode transition that avoids significant power loss by ensuring that the optical mode evolves smoothly from one waveguide to another. This understanding is rooted in the physics of guided-wave optics, where "adiabatic" transitions are those in which the system remains in its lowest-order eigenstate as the waveguide geometry changes — the hallmark of which is low loss and suppression of higher-order mode excitation. Critically, a POSITA would understand that "adiabatic" describes the performance outcome of the coupling, not the specific structural mechanism by which that outcome is achieved. Both continuous tapers and properly designed segmented tapers can achieve adiabatic coupling.
(c) State of the art regarding waveguide coupling in 2016–2017:
Dr. Fong should be prepared to describe the state of the art in waveguide coupling as of the relevant time period. Both continuous tapers and segmented tapers were known approaches for achieving waveguide coupling. The term "adiabatic" was used in the technical literature to describe the performance characteristic of the coupling (i.e., low-loss mode transition), and was not limited in the literature to a single structural geometry. Dr. Fong should be familiar with representative publications from the period that use "adiabatic" in this functional sense, including publications describing segmented or stepped taper designs that achieve adiabatic coupling.
(d) The specification describes adiabatic coupling in functional terms:
The specification contains language describing the adiabatic coupling principle as follows: it discusses the gradual evolution of the optical mode field from the silicon waveguide to the silicon nitride waveguide, and characterizes the coupling as "adiabatic" based on the outcome of that evolution — namely, that the mode transition occurs without significant power loss to higher-order modes or radiation modes. The specification does not contain language limiting the adiabatic concept to continuous tapers only.
4. Anticipated BHS Attack Lines
Counsel should prepare Dr. Fong for the following anticipated lines of attack from BHS:
(a) Continuous taper argument: BHS will attempt to pin Dr. Fong to a definition of "adiabatic" that requires a "continuous" taper geometry. They may ask questions designed to elicit testimony such as: "Isn't it true that an adiabatic transition, by definition, requires that the waveguide geometry change gradually and continuously?" or "Would a segmented or stepped taper that introduces abrupt changes in waveguide width be considered 'adiabatic'?" Dr. Fong should be prepared to respond that while a continuous taper is one way to achieve adiabatic coupling, the term "adiabatic" in the silicon photonics field is a description of performance, not structure. A segmented taper with sufficiently small steps can also achieve adiabatic coupling — the question is whether the mode evolves smoothly and without significant loss, not whether the physical geometry is mathematically continuous.
(b) Prosecution history arguments: BHS will likely point to the prosecution history of the '567 Patent, in which the claims were narrowed to include the "adiabatically coupled" limitation specifically to overcome the Tanaka (U.S. 9,456,012) prior art. BHS will argue that this amendment constitutes a prosecution history estoppel that limits the scope of "adiabatically coupled" to exclude the type of non-adiabatic coupling described in Tanaka's prior patent. They may further argue that the amendment should be read to limit "adiabatic" to the specific structural implementation described in the specification (i.e., a continuous taper), because that is the embodiment that most clearly distinguishes the claimed invention from the prior art.
Dr. Fong should be prepared to testify that the addition of the "adiabatically coupled" limitation during prosecution was intended to distinguish the functional performance of the claimed coupler from the prior art — i.e., the claimed coupler achieves low-loss mode transitions, whereas the Tanaka prior art coupler does not. The amendment was not intended to limit the structural geometry of the taper to a continuous taper only. The prosecution record, including the applicant's remarks accompanying the amendment, characterizes the distinction in functional terms.
(c) Scope of surrender: BHS may argue that under the doctrine of prosecution history estoppel, particularly the Festo presumption, the scope of surrender extends to all subject matter between the original claim scope and the amended claim scope, unless one of the Festo exceptions applies. BHS may contend that by adding the "adiabatically coupled" limitation, OWS surrendered coverage of non-continuous (segmented) taper designs. Dr. Fong should be prepared to explain that the amendment was directed to distinguishing the functional behavior of the coupler (adiabatic vs. non-adiabatic), not to limiting the structural geometry.
5. Redirect Topics
If Dr. Fong's cross-examination testimony on this point requires rehabilitation on redirect, the following topics should be addressed:
• The specification's language describing the adiabatic principle in functional terms, with specific citations to column and line numbers;
• Representative technical literature from the 2016–2017 period demonstrating that "adiabatic" was used in the field to describe the performance outcome of a coupling regardless of whether the taper geometry was continuous or segmented;
• The applicant's remarks during prosecution that characterize the distinction between the claimed invention and the Tanaka prior art in functional (not structural) terms.
D. Prosecution History of the '567 Patent
The '567 Patent application underwent two office action rejections during prosecution before being allowed. The prosecution history is directly relevant to the claim construction dispute because the "adiabatically coupled" limitation was added by amendment during prosecution to overcome prior art.
1. Overview of Prosecution
The original claims as filed on March 14, 2017, did not include the term "adiabatically coupled." The claims recited an integrated optical coupler with a tapered silicon waveguide coupled to a silicon nitride waveguide layer, but without the "adiabatically" qualifier.
The first office action rejected the claims over Tanaka, U.S. Patent No. 9,456,012, under 35 U.S.C. § 102(a)(1). The applicant responded with an amendment adding the "adiabatically coupled" limitation to Claim 1 and corresponding dependent claims, accompanied by remarks distinguishing the claimed invention from Tanaka's disclosure.
The second office action maintained certain rejections, and the applicant responded with further argument (but no additional substantive amendments to the claim language at issue). The claims were subsequently allowed.
2. Dr. Fong's Involvement in Prosecution Decisions
Dr. Fong should be prepared to discuss his involvement in the prosecution of the '567 Patent, to the extent such discussion does not implicate privileged attorney-client communications. Specifically, BHS may ask whether Dr. Fong consulted with Ted Marchetti (VP of IP, Reg. Pat. Atty. No. 58,304) regarding the claim amendments made during prosecution. Dr. Fong may testify that he worked with Marchetti on technical aspects of the patent application, including reviewing claims and providing input on the technical distinctions between the claimed invention and the cited prior art. However, Dr. Fong should not disclose the substance of any privileged communications with Marchetti regarding prosecution strategy or legal advice.
3. Key Points for Testimony
Dr. Fong should be prepared to testify that:
• The distinction made during prosecution between the claimed invention and Tanaka's prior art was focused on the functional adiabatic coupling characteristic of the claimed coupler — i.e., the ability to achieve low-loss mode transitions — as distinguished from the prior art coupler, which did not achieve adiabatic performance;
• The amendment was not intended to impose a specific structural limitation (such as requiring a continuous taper) but rather to recite the functional characteristic that distinguishes the inventive coupler from the prior art;
• He was involved in reviewing the technical accuracy of the applicant's remarks, and those remarks are consistent with this understanding.
4. Coordination with Internal Analysis
The testimony to be elicited from Dr. Fong on prosecution history should be coordinated with the internal memorandum on rebuttal of prosecution history estoppel arguments currently being prepared by Christy Bauer for the December 2022 briefing deadline. Any positions taken by Dr. Fong in deposition testimony should be consistent with the legal arguments that will be advanced in briefing.
5. Review of Applicant's Remarks
Prior to the deposition, the team should carefully review the applicant's remarks filed during prosecution for any statements that could be used against OWS under the doctrine of prosecution history estoppel. In particular, the team should identify any language in which the applicant characterized the "adiabatic" limitation in structural terms (e.g., by reference to a "continuous" or "gradual" taper) rather than functional terms. If such language exists, Dr. Fong should be prepared to address it and, if necessary, to explain that the structural description was provided as an example of one embodiment, not as a definition of the claim term.
E. Relationship with Hideo Tanaka's Prior Art
1. Background. U.S. Patent No. 9,456,012 (hereinafter "Tanaka") issued on September 6, 2016. Hideo Tanaka is the sole inventor of this patent, which he conceived and developed before joining OWS. Tanaka subsequently joined OWS as an engineer in approximately 2016 and became a co-inventor on the '567 Patent, which was filed on March 14, 2017.
The chronology here is important and will be of significant interest to BHS:
2. Preparation of Dr. Fong's Testimony
Dr. Fong should be prepared to testify about the following:
(a) When OWS first became aware of Tanaka's prior patent: Dr. Fong should provide a clear answer about when he and OWS became aware of U.S. Patent No. 9,456,012. This may have occurred during the hiring process, as part of OWS's standard prior art review, or during the course of developing the '567 Patent invention.
(b) How the '567 Patent invention builds upon or differs from Tanaka's earlier work: The '567 Patent invention differs from Tanaka's U.S. 9,456,012 in that the '567 Patent employs an adiabatic coupling approach to achieve low-loss mode transitions between Si and SiN waveguides, whereas Tanaka's earlier patent describes a non-adiabatic coupling structure that does not achieve the same level of mode transition efficiency. Dr. Fong should be prepared to articulate this distinction in clear, technically accurate terms.
(c) Tanaka's specific contributions to the '567 Patent versus his earlier independent work: Dr. Fong should be prepared to explain what Tanaka contributed to the '567 Patent as a co-inventor, and how those contributions are distinct from the subject matter of Tanaka's earlier patent. The '567 Patent reflects collaborative work among Fong, Ramanathan, and Tanaka at OWS, building on the collective expertise of the team. Tanaka's earlier patent reflects his independent work before joining OWS and does not disclose the adiabatic coupling approach claimed in the '567 Patent.
3. Anticipated BHS Arguments on Obviousness
Although the deposition is limited to claim construction topics, BHS may attempt to probe the relationship between the '567 Patent and Tanaka's prior art in a manner designed to lay the groundwork for later obviousness arguments. In particular, BHS may ask whether the '567 Patent is an obvious modification of Tanaka's U.S. 9,456,012, especially when combined with Chen & Liu, "Segmented Taper Coupler Design for Low-Loss Fiber-to-Chip Coupling," Journal of Photonic Integration, Vol. 14, No. 3, pp. 217–229 (2016). This article has been cited by Lumos under 35 U.S.C. § 103 in combination with Tanaka.
If BHS pursues this line of questioning, defending counsel should consider objecting as beyond the scope of claim-construction-related discovery. However, to the extent questions about the state of the art and the relationship between the claimed invention and the prior art are permissible in the claim construction context, Dr. Fong should be prepared to:
• Distinguish the inventive contribution of the '567 Patent from both the Tanaka patent and the Chen & Liu reference;
• Explain that the Chen & Liu reference addresses fiber-to-chip coupling (a different application than the Si-to-SiN on-chip coupling addressed by the '567 Patent);
• Explain that the combination of Tanaka and Chen & Liu would not have been obvious to a POSITA at the time of the invention, because the two references address different coupling challenges and a POSITA would not have been motivated to combine them in the manner suggested.
F. Other Claim Terms from the '567 Patent
1. "Tapered silicon waveguide having a tip width of less than 200 nanometers"
This claim limitation specifies that the silicon waveguide taper terminates in a tip with a width of less than 200 nanometers. This term is expected to be less heavily contested than "adiabatically coupled," but Dr. Fong should nonetheless be prepared for questions about the significance of the 200nm threshold. He should be able to explain that:
• The 200nm tip width is significant because it ensures that the optical mode is substantially expelled from the silicon waveguide at the tip, facilitating efficient coupling to the underlying silicon nitride waveguide;
• The choice of 200nm as the threshold was based on simulations and experimental data showing that effective mode transfer occurs when the silicon waveguide width is reduced below this value on the 45nm SOI platform;
• A POSITA would understand that the 200nm value is an upper bound on tip width, and that narrower tips (e.g., 100–150nm) would also fall within the claim scope.
2. Additional Disputed Terms
Dr. Fong should be familiar with any other disputed terms from the '567 Patent identified in the Joint Claim Construction and Prehearing Statement filed on October 14, 2022. As of the preparation of this outline, the primary disputed term from the '567 Patent is "adiabatically coupled." However, the team should review the Joint Statement for any additional terms and prepare Dr. Fong accordingly.
IV. U.S. PATENT NO. 10,891,234 ('234 PATENT) — GERMANIUM-ON-SILICON PHOTODETECTOR
A. Background of the Invention
The '234 Patent is titled "Germanium-on-Silicon Photodetector with Graded Buffer Layer for Enhanced Responsivity." The application was filed on September 8, 2017, and the patent issued on January 14, 2021. The named inventors are Dr. Kevin Fong and Dr. Amara Osei, who is a principal engineer at OWS.
The asserted claims of the '234 Patent are Claims 1, 3, 5, 8, and 11. Claim 1 is the primary independent claim; Claims 3, 5, 8, and 11 depend therefrom.
The '234 Patent is directed to a germanium-on-silicon (Ge-on-Si) photodetector structure that employs a graded silicon-germanium (SiGe) buffer layer between the silicon substrate and the germanium absorber layer. The purpose of the graded buffer layer is to manage the lattice mismatch between silicon (lattice constant ~5.431 Å) and germanium (lattice constant ~5.658 Å). Without a buffer layer, direct epitaxial growth of germanium on silicon results in high threading dislocation densities that degrade photodetector performance — particularly responsivity (the ratio of photocurrent to incident optical power) and dark current (leakage current in the absence of light). The graded buffer layer gradually transitions the lattice constant from that of silicon to that of germanium, reducing dislocation density and thereby enhancing device performance.
B. Genesis of the Invention
Dr. Fong should be prepared to describe the genesis of the invention, including the technical problem it addressed and the inventive contributions of each co-inventor.
The core technical problem was achieving enhanced responsivity in germanium-on-silicon photodetectors while maintaining compatibility with standard CMOS fabrication processes — specifically, the 45nm SOI process used at Pinnacle Semiconductor Foundry. Prior approaches to Ge-on-Si photodetectors suffered from high threading dislocation densities, which degraded both responsivity and dark current performance. The '234 Patent addresses this problem through a graded SiGe buffer layer that reduces dislocation density by providing a controlled transition in lattice constant.
Dr. Fong should be able to describe:
• His specific contributions to the invention: conceptualization of the graded buffer approach, design of the buffer layer composition profile, and analysis of device performance;
• Dr. Amara Osei's contributions: Dr. Osei's expertise in epitaxial growth and materials science, her contributions to the design and optimization of the deposition process for the graded buffer layer;
• The timeline of conception and reduction to practice, including when the graded buffer concept was first conceived, when prototype devices were first fabricated, and when successful performance results were obtained;
• The fabrication of the graded buffer structures through Pinnacle Semiconductor Foundry on the 45nm SOI process, and the practical significance of how the "continuously varying" germanium concentration was achieved in actual fabrication — this point is directly relevant to the claim construction dispute discussed below.
C. Key Disputed Claim Term: "graded buffer layer having a continuously varying germanium concentration"
1. Claim Language
Claim 1 of the '234 Patent recites, in relevant part, a photodetector structure including a "graded buffer layer having a continuously varying germanium concentration" disposed between the silicon substrate and the germanium absorber layer.
2. Competing Constructions
OWS's Proposed Construction:
"A buffer layer in which the germanium concentration changes across the layer thickness."
Lumos's Proposed Construction:
"A buffer layer in which the germanium atomic fraction increases monotonically and without step changes from the silicon substrate interface to the absorber layer."
3. Significance of the Construction Dispute
The difference between these constructions has direct implications for infringement. Lumos's LumosEdge and LumosBridge photodetector products use a step-graded SiGe buffer layer with four discrete composition steps, at germanium atomic fractions of 0.25, 0.50, 0.75, and 1.0 (i.e., pure germanium). Under OWS's broader construction, which encompasses buffer layers where the germanium concentration "changes across the layer thickness" — including both continuously graded and step-graded implementations — Lumos's step-graded buffer would fall within the claim scope. Under Lumos's narrow construction, which requires that the germanium atomic fraction increase "monotonically and without step changes," Lumos's four-step buffer would be excluded because it employs discrete steps rather than a smooth, continuous grading profile.
4. Preparation of Dr. Fong's Testimony
Dr. Fong should be prepared to testify on the following points:
(a) Understanding of "continuously varying germanium concentration" as used in the specification:
Dr. Fong should testify that the term "continuously varying germanium concentration" as used in the '234 Patent specification describes a buffer layer in which the germanium concentration changes as a function of position across the thickness of the layer. The specification contemplates that the concentration profile may take various forms, and it is not limited to a single mathematical function (e.g., a linear ramp).
(b) Specification describes both continuously graded and step-graded embodiments:
This is a critically important point. The '234 Patent specification contains embodiments describing both "continuously graded" and "step-graded" buffer layers. Both approaches are discussed as means for managing the lattice mismatch between silicon and germanium. Dr. Fong should be prepared to identify the specific passages in the specification where step-graded embodiments are described, and to explain that the inventors understood and intended the patent to cover both approaches.
However, Dr. Fong should also acknowledge — because BHS will certainly press this point — that the claim language itself uses the phrase "continuously varying," rather than a broader phrase such as "graded" or "varying." BHS will exploit the distinction between the broad specification disclosure and the more specific claim language. Dr. Fong should be prepared to address this by explaining that the term "continuously varying," as understood by a POSITA, does not require a mathematically continuous function but rather encompasses gradual changes in concentration across the buffer layer thickness — including changes achieved through a series of fine incremental steps.
(c) POSITA understanding of "continuously varying":
A person of ordinary skill in the art would understand "continuously varying germanium concentration" to encompass buffer layers in which the germanium concentration changes across the layer thickness in a manner that provides a graded transition from silicon to germanium. A POSITA would not interpret "continuously varying" as requiring a strict mathematical continuity — i.e., a function with no discontinuities at any scale — because such mathematical continuity is impossible to achieve in practice. In actual semiconductor fabrication, even nominally "continuous" grading is achieved through molecular beam epitaxy (MBE) or chemical vapor deposition (CVD) processes that deposit material in discrete atomic layers. The distinction between a "continuously graded" buffer and a "step-graded" buffer is therefore one of degree, not kind. A POSITA would understand that a step-graded buffer with sufficiently small steps (including, for example, the four discrete steps used in Lumos's products) may constitute a "continuously varying" concentration profile, particularly when the steps are small enough that the buffer layer provides the same functional benefit — reduced threading dislocation density — as a nominally continuous grading.
(d) State of the art circa September 2017:
Both continuously graded and step-graded SiGe buffer layers were well-known techniques for managing lattice mismatch in germanium-on-silicon systems. The two approaches were widely recognized in the field as alternative implementations of the same graded buffer concept. The term "continuous grading" was used in the literature somewhat loosely, often to describe buffer layers that were in fact fabricated using many small discrete steps.
5. Anticipated BHS Attack Lines
(a) Specification vs. claim language distinction: BHS will press on the distinction between the specification, which describes both continuously graded and step-graded embodiments, and the claim language, which uses only the term "continuously varying." BHS will argue that the applicants deliberately chose "continuously varying" over a broader term (such as "graded") to distinguish the specific approach claimed from the step-graded approach described in the specification. They will argue that the claim should be construed according to its plain meaning and that "continuously varying" excludes discrete step changes.
(b) Plain meaning argument: BHS may argue that the ordinary and customary meaning of "continuously varying" — both in everyday English and in technical usage — requires a smooth, uninterrupted change, and that a step-graded profile with four discrete steps is not "continuously varying" by any reasonable interpretation.
(c) Extrinsic evidence: BHS may introduce extrinsic evidence, including technical dictionaries, textbooks, and expert testimony, defining "continuously varying" as requiring a smooth change without discontinuities. For example, they may cite semiconductor physics textbooks that distinguish between "continuously graded" and "step-graded" buffer layers as different approaches.
(d) Prosecution history (limited): The '234 Patent was allowed on the first office action after an examiner interview, and no narrowing amendments were made during prosecution. This means there is no prosecution history estoppel argument available to Lumos. However, BHS may argue that this cuts against OWS as well: because the claims were not narrowed, the applicants chose the specific language "continuously varying" at the outset, which should be given its plain meaning.
6. Redirect Topics
If Dr. Fong's cross-examination testimony requires rehabilitation:
• Point to the specification's description of both continuously graded and step-graded embodiments, arguing that the specification is the best evidence of the meaning of the claim terms;
• Explain that the inventors' intent, as reflected in the specification, was to cover the functional benefit of graded buffers generally;
• Explain that in practical semiconductor fabrication, even "continuous" grading is achieved through many small incremental steps at the molecular deposition level — the distinction between "continuous" and "step-graded" is inherently one of degree;
• If Pinnacle Semiconductor Foundry's process for fabricating the "continuously varying" buffer in OWS's own products involves discrete deposition steps, this fact supports the interpretation that "continuously varying" encompasses step-wise grading.
D. Prosecution History of the '234 Patent
1. Overview. The '234 Patent was allowed on the first office action after an examiner interview. No narrowing amendments were made during prosecution. This prosecution history is notable for what it does not contain: there is no prosecution history estoppel argument available to Lumos with respect to the '234 Patent claims.
2. Examiner Interview
The examiner interview is potentially significant. Dr. Fong should be prepared to discuss what was communicated to the examiner during the interview, to the extent he was involved. In particular:
• Who participated in the interview from OWS's side (e.g., Ted Marchetti, Dr. Fong, Dr. Osei)?
• What technical arguments were presented to the examiner?
• Were any representations made about the meaning of "continuously varying germanium concentration"?
The examiner interview summary, which should be part of the prosecution file history, should be reviewed carefully for any statements that could be relevant to the meaning of the disputed claim term. If the examiner interview summary contains statements suggesting that "continuously varying" was understood to mean something specific (e.g., excluding step-graded buffers), this could be used against OWS at the Markman hearing.
3. Absence of Narrowing Amendments
The absence of narrowing amendments means that BHS cannot rely on prosecution history estoppel to narrow the claims. However, as noted above, this also means that BHS may argue that the claim language should be given its plain meaning, without any basis for broadening the interpretation beyond what the words "continuously varying" would naturally convey to a POSITA.
E. Prior Art Cited Against the '234 Patent
1. Nordlicht EP Application
European Patent Application EP 3,234,567 A1, published August 2, 2017, by applicant Nordlicht Photonische GmbH, has been cited by Lumos under 35 U.S.C. § 102(a)(2) as prior art against the '234 Patent. Dr. Fong should be prepared to: