`
`———————
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`———————
`
`General Electric Company,
`Petitioner,
`
`v.
`
`United Technologies Corporation,
`Patent Owner
`
`———————
`
`Case IPR2016-01289
`Patent 7,060,360 B2
`
`———————
`
`
`PATENT OWNER’S SURREPLY BRIEF ON REMAND
`
`
`
`GE’s remand arguments confirm that its proposed combination cannot satisfy
`
`the Federal Circuit’s construction of “bond layer.” Many of GE’s arguments conflict
`
`with the Federal Circuit’s express construction and its holding that adherence is an
`
`“additional” requirement beyond the bond layer’s location and composition. United
`
`Techs. Corp. v. Gen. Elec. Co., 757 F. App’x 971, 973-74 (Fed. Cir. 2019) (original
`
`emphasis). For example, GE relies on the Examiner and Eaton to imply that
`
`Terentieva’s protective coating is a “bond layer” either in the proposed combination
`
`or in Terentieva itself merely because it is an “intermediate layer.” Paper 32 at 1-2.
`
`Similarly, GE argues that UTC “cannot now argue that certain compositions for the
`
`bond layer are within the scope of the claims and others are not, based on an adhesion
`
`quality lacking any quantitative scope.” Id. at 4; see also id. at 5 (arguing that “bond
`
`layer” does not require “‘improved’ . . . adhesion” because “Terentieva’s coating
`
`layer . . . falls within the composition of the claimed ‘bond layer.’”).
`
`These arguments directly conflict with the Federal Circuit’s holding. If every
`
`intermediate layer were a “bond layer,” as GE implies, or if every composition
`
`within the scope of the claimed genus would work as a “bond layer,” then there
`
`would be no additional requirement of adherence. But the Federal Circuit directly
`
`rejected GE’s position, holding that “those additional requirements do not mean that
`
`the ‘bond layer’ does not also have to bond.” United Techs., 757 F. App’x at 974
`
`(original emphasis). GE’s continuing to advance these rejected arguments confirms
`
`1
`
`
`
`that GE cannot carry its burden under the Federal Circuit’s construction.
`
`GE also tries to rewrite the Federal Circuit’s construction to require only
`
`“adherent properties.” Paper 32 at 1-2. Yet, the only “adherent properties” to which
`
`GE points relate to the protective coating’s ability “‘to adhere to the surface of the
`
`material to be protected,’” id. (citing GE-1005, 4:30-33), not whether it is “designed
`
`to adhere another layer to a substrate,” as the Federal Circuit’s construction requires.
`
`GE tries to elide this deficiency by invoking Terentieva’s optional outer layer.
`
`Paper 32 at 1-2. Yet, GE cites no evidence that the protective coating was “designed
`
`to adhere” this optional layer—which is used in none of Terentieva’s embodiments,
`
`GE-1005.004-006—to a substrate. Terentieva itself is silent about whether the
`
`protective coating plays a role in any adhesion between the optional layer and a
`
`substrate, and Dr. Clarke’s testimony that “Terentieva does not describe or suggest
`
`that the protective coating was designed to adhere another layer (e.g., the refractory
`
`layer) to the substrate” is unrebutted because GE opted to put forth no contrary
`
`expert testimony. UTC-2001, ¶¶ 92-93. That is dispositive because expert testimony
`
`is required to establish obviousness where the technology is beyond the grasp of a
`
`layperson, as it is here. Alexsam, Inc. v. IDT Corp., 715 F.3d 1336, 1347-48 (Fed.
`
`Cir. 2013). GE relies entirely on the Examiner’s assumption that an intermediate
`
`layer is “considered to effectively function as a bond layer,” Paper 32 at 1 (citing
`
`Ex. 1002.032), but that reasoning is foreclosed by the Federal Circuit’s holding.
`
`2
`
`
`
`Further, even if Terentieva’s protective coating could work as a bond layer in
`
`some systems, GE cannot show that it would work in the relevant context—namely,
`
`in the combination asserted in GE’s petition (Terentieva’s protective coating directly
`
`between Eaton’s BSAS layer and a substrate). Pet., 23. Obviousness is evaluated
`
`“with respect to the subject matter as a whole, not separate pieces of the claim.”
`
`Sanofi-Synthelabo v. Apotex, Inc., 550 F.3d 1075, 1086 (Fed. Cir. 2008). Whether
`
`the protective coating could work as a bond layer in another context is irrelevant.
`
`GE’s only substantive attempt to show that Terentieva’s protective coating
`
`could work as a bond layer with Eaton’s BSAS is its argument—made for the first
`
`time in its reply remand brief—that “a POSITA would expect the CTEs of
`
`Terentieva’s coating layer and Eaton’s BSAS to be compatible.” Paper 32 at 3-4.
`
`The experts agree that compatible CTEs are necessary for a bond layer to work.
`
`UTC-2001, ¶ 46; UTC-2014, 48:5-49:4. Consistent with Eaton and the POSITA’s
`
`knowledge, the ’360 Patent explicitly identifies mismatched CTE as contributing to
`
`the loss of mechanical properties. GE-1001, 1:35-39 [cited in Paper 12 at 3].
`
`However, GE’s evidence purporting to show compatible CTEs between
`
`Terentieva’s protective coating and Eaton’s BSAS is speculative and unsupported.
`
`GE first points to Dr. Glaeser’s testimony about “combinations of Si and MoSi2,”
`
`see GE-1003, ¶ 57, but Dr. Glaeser never addressed the effects of Titanium—
`
`particularly TiSi2, a significant component of the protective coating. Id.; GE-
`
`3
`
`
`
`1005.002 (2:57-67). GE’s attorney argument cannot fill this gap not only because
`
`expert testimony is necessary to explain the impact of TiSi2, see e.g. Alexsam, 715
`
`F.3d at 1347-48, but also because the evidence undercuts GE’s attorneys’ assertions.
`
`GE argues that any combination of Mo, Ti and/or Si would have a similar
`
`CTE merely because Ti and MoSi2 have similar CTEs in isolation. Paper 32 at 4-5
`
`(citing GE-1016.007). That argument is specious. The evidence GE cites itself
`
`disproves this inference, showing that compositions with Mo, Ti, and/or Si can have
`
`vastly different CTEs (e.g., Ti5Si3 has a CTE of 3.05, which is far less than the 8.25
`
`of MoSi2). GE-1016.007. Indeed, GE’s table shows no known value for the TiSi2
`
`contained in Terentieva because “?” is the entry next to TiSi2. Id. Furthermore, the
`
`undisputed evidence establishes that TiSi2’s CTE is “significantly higher” than Si or
`
`MoSi2. UTC-2013, ¶ 16; UTC-2023 at 1204 & Fig. 10. Thus, GE’s assertion that
`
`combining Ti with Si or MoSi2 would yield a similar CTE is unsupported.
`
`GE’s other attempts to address Terentieva’s Titanium are also flawed. First,
`
`that some Ti-containing compositions work as bond layers for BSAS does not mean
`
`that Terentieva’s specific composition would. UTC-2013, ¶¶ 15-19. Second, that a
`
`composition’s CTE can be “tuned . . . to some extent,” GE-1031, 44:20-46:6
`
`(emphasis added), does not suggest that CTE mismatch can always (or even usually)
`
`be overcome through modifications. Further, GE has not proposed how or why a
`
`POSITA would have tried to tune the CTE of Terentieva’s protective coating. GE
`
`4
`
`
`
`relies entirely on the Board’s previous obviousness findings, Paper 32 at 4, but those
`
`findings were based on GE’s now-rejected claim construction, and GE never
`
`addresses a POSITA’s expectations regarding adherence between Terentieva’s
`
`protective coating and BSAS, and fails to provide any evidence that tuning Si content
`
`would solve any CTE mismatch. Id. GE’s lack of evidence is especially problematic
`
`given the experts’ agreement that CTE is just one of many factors that influence
`
`compatibility between layers. UTC-2001, ¶¶ 44-49; UTC-2013, ¶¶ 7-12; UTC-
`
`2014, 47:14-49:19; GE-1003, ¶¶ 45-46 [all cited in Paper 12 at 33].1
`
`Finally, GE does not argue that it can show that Terentieva’s protective
`
`coating provides enhanced adherence. Instead, GE asserts that the Federal Circuit’s
`
`construction imposes no such requirement. See Paper 32 at 5. The plain language of
`
`the court’s construction requires a layer “designed to adhere another layer to a
`
`substrate.” The extrinsic evidence relied upon by the Federal Circuit and common
`
`sense both show that a layer is not “designed to adhere another layer to a substrate”
`
`if it provides no better adherence than would exist without it. See Paper 29 at 6-7.
`
`Moreover, GE provides no alternative test for whether a layer is designed to adhere
`
`another layer to a substrate. Paper 32 at 5. Accordingly, GE cannot satisfy the
`
`“designed to adhere” requirement of Federal Circuit’s claim construction.
`
`
`1 The Federal Circuit did not reach UTC’s challenge to the Board’s obviousness
`
`analysis on appeal, and UTC may renew those arguments in any further appeal.
`
`5
`
`
`
`Respectfully submitted,
`
`
`
`/W. Karl Renner/
`
`W. Karl Renner, Reg. No. 41,265
`Lauren A. Degnan, Reg. No. 40,584
`Timothy W. Riffe, Reg. No. 43,881
`David L. Holt, Reg. No. 65,161
`Attorneys for Patent Owner
`
`
`
`
`
`
`
`
`
`
`May 22, 2019
`
`
`
`
`Date:
`
`
`
`
`
`
`Customer Number 26171
`Fish & Richardson P.C.
`Telephone: (202) 783-5070
`Facsimile: (877) 769-7945
`
`6
`
`
`
`CERTIFICATE OF SERVICE
`
`Pursuant to 37 CFR § 42.6(e)(4), the undersigned certifies that on May 22,
`
`2019, a complete and entire copy of this Patent Owner’s Surreply Brief on Remand
`
`was provided via email, to the Petitioner by serving the email correspondence
`
`address of record as follows:
`
`Anish Desai
`Weil Gotshal & Manges LLP
`767 Fifth Avenue
`New York, NY 10153-0119
`
`Brian E. Ferguson
`Megan H. Wantland
`Weil Gotshal & Manges LLP
`2001 M Street, NW, Suite 600
`Washington, DC 20006
`
`Email: GE.WGM.Service@weil.com
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`/Edward G. Faeth/
`Edward G. Faeth
`Fish & Richardson P.C.
`3200 RBC Plaza
`60 South Sixth Street
`Minneapolis, MN 55402
`(202) 626-6420
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`7
`
`