Contract analysis

I. INTRODUCTION Plaintiff Optiwave Semiconductor Corp. ("OWS" or "Plaintiff") respectfully submits this Reply Claim Construction Brief in response to the arguments raised in Defendant Lumos Photonics, Inc.'s ("Lumos" or "Defendant") Responsive Claim Construction Brief, filed December 19, 2022 (Dkt. No. 87) ("Resp. Br."), regarding the construction of eleven disputed claim terms across U.S. Patent Nos. 10,234,567 ("th

II. LEGAL STANDARD FOR CLAIM CONSTRUCTION OWS's Opening Brief set forth the governing legal standards in detail. See Opening Br. at 4–8. OWS briefly reiterates the key principles here, as they are directly implicated by the arguments raised in Lumos's Responsive Brief. Claim construction is a matter of law for the Court to decide. Markman v. Westview Instruments, Inc., 517 U.S. 370, 388–91 (1996); Teva Pharm. USA, In

III. REPLY ARGUMENTS ON DISPUTED TERMS OF THE '567 PATENT A. "Adiabatically Coupled" (Claim 1, and Dependent Claims 4, 7, 12, 15) The parties' competing constructions of the term "adiabatically coupled" are as follows: Lumos's Responsive Brief advances several arguments in support of its unduly narrow construction. Each of these arguments fails for the reasons set forth below.

1. Plain and Ordinary Meaning Supports OWS's Construction. The term "adiabatic" in the context of photonic waveguide coupling is a well-understood and well-established term of art. As OWS demonstrated in its Opening Brief, and as Dr. Whitfield's declaration confirms, a POSITA working in the silicon photonics field would understand "adiabatically coupled" as a functional description of the coupling efficiency — specif

2. The Specification Does Not Require a "Continuous" Taper. A central pillar of Lumos's argument is the assertion that the specification of the '567 Patent exclusively describes a continuous taper, and therefore "adiabatically coupled" must be limited to coupling achieved through such a taper. See Resp. Br. at 12–14. This argument is factually incorrect and legally insufficient. As an initial matter, even if the spec

3. Prosecution History Estoppel Does Not Support Lumos's Narrow Construction. Lumos's estoppel argument constitutes a significant portion of its Responsive Brief's analysis of the "adiabatically coupled" term. See Resp. Br. at 15–20. Lumos contends that during prosecution of the '567 Patent, the claims underwent two office action rejections, and that the "adiabatically coupled" limitation was added to overcome the pr

4. Lumos's Over-Reliance on Dr. Ng's Extrinsic Testimony. Finally, Lumos devotes nearly five pages of its Responsive Brief to the testimony of Dr. Raymond Ng regarding the meaning of "adiabatic" in photonics literature. See Resp. Br. at 20–24. As noted above, Dr. Ng is Of Counsel / Technical Consultant at Brannick Howell & Stein LLP, counsel for Lumos, and a former professor of photonics at the University of Colorado

IV. REPLY ARGUMENTS ON DISPUTED TERMS OF THE '234 PATENT A. "Graded Buffer Layer Having a Continuously Varying Germanium Concentration" (Claim 1, and Dependent Claims 3, 5, 8, 11) The parties' competing constructions of this term are as follows: Lumos's proposed construction suffers from the same fundamental deficiency that afflicts its constructions of other terms: it imports multiple limitations from a preferred em

1. The Claim Language Does Not Require "Monotonic" Increase or Exclude Step Changes. The starting point of claim construction is the claim language itself. Phillips, 415 F.3d at 1312. The relevant claim language is "graded buffer layer having a continuously varying germanium concentration." The claim uses the phrase "continuously varying" — not "monotonically increasing," not "smoothly graded," not "without step chan

2. The Specification Supports a Broader Reading. Lumos argues that the specification of the '234 Patent supports only a smooth, continuous gradient for the "continuously varying" germanium concentration. See Resp. Br. at 28–30. This argument is contradicted by the specification itself. The '234 Patent specification does not limit "continuously varying" to a smooth gradient. To the contrary, the specification describe

3. Dr. Ng's Testimony Improperly Narrows the Claim. As with the '567 Patent term "adiabatically coupled," Lumos relies heavily on Dr. Ng's extrinsic testimony to support its narrow construction of "continuously varying." Dr. Ng's declaration asserts that "continuously varying" in materials science necessarily means a smooth, monotonic gradient without discrete steps, and he cites materials science textbooks in suppor

V. REPLY ARGUMENTS ON DISPUTED TERMS OF THE '678 PATENT A. "Integrated Wavelength-Division Multiplexing Element Monolithically Formed on a Single Silicon Die" (Claim 1, and Dependent Claims 2, 6, 9, 14) The parties' competing constructions are as follows: Lumos's proposed construction is one of the most extreme in this case. It does not merely import a single limitation from the specification; it rewrites the claim b

1. Lumos's Construction Improperly Imports Negative Limitations. The claim recites "integrated wavelength-division multiplexing element monolithically formed on a single silicon die." This language identifies a WDM element that is (a) integrated, (b) monolithically formed, and (c) on a single silicon die. Each of these terms has a well-understood meaning. "Monolithically formed" is a standard term of art in semicondu

2. Prosecution History Estoppel Is Limited in Scope. Lumos's principal argument for its construction of this term rests on prosecution history estoppel. The '678 Patent underwent three office actions during prosecution. During that prosecution, the applicant amended the claims to add the "monolithically formed on a single silicon die" limitation in order to overcome a prior art rejection. The examiner had relied upon

3. Dr. Ng's Testimony on "Monolithically Formed". Dr. Ng opines that "monolithically formed" in the semiconductor industry means fabricated exclusively through front-end-of-line (FEOL) processing within the same silicon substrate. See Ng Resp. Decl. ¶¶ 42–50. This definition is unduly narrow and does not reflect the understanding of a POSITA working in the silicon photonics field. Dr. Whitfield responds comprehensive

VI. REPLY ARGUMENTS ON DISPUTED TERMS OF THE '890 PATENT A. "Pre-Emphasis Equalization Stage Electrically Co-Integrated with the Modulator on a Common Substrate" (Claim 1, and Dependent Claims 5, 8, 13) The parties' competing constructions are as follows: Lumos's proposed construction is the most extreme example of limitation importation in this case. It transforms a straightforward claim requirement — co-integration

1. Lumos's Construction Adds Multiple Unwarranted Process Limitations. The claim recites a "pre-emphasis equalization stage electrically co-integrated with the modulator on a common substrate." The claim language identifies three requirements: (1) a pre-emphasis equalization stage, (2) electrical co-integration with the modulator, and (3) location on a common substrate. OWS's proposed construction — "an equalization

2. Prosecution History Context. The prosecution history of the '890 Patent provides context for the meaning of "co-integrated" but does not support the sweeping limitations Lumos proposes. During prosecution, the "co-integrated" limitation was added to distinguish prior art in which the driver circuit (equalization stage) and the modulator were fabricated on entirely separate substrates and then connected through pac

3. Extrinsic Evidence. Dr. Ng's declaration once again imports rigid academic definitions in support of Lumos's narrow construction. Dr. Ng opines that "co-integrated" in the semiconductor industry necessarily requires fabrication in the same semiconductor process with shared doping profiles and process layers. See Ng Resp. Decl. ¶¶ 55–62. Dr. Whitfield addresses this testimony directly and persuasively. Whitfield De