`
`
`
`
`
`
`
`
` Paper No. __
`
`
`
`
`
`
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`____________
`
`TAIWAN SEMICONDUCTOR MANUFACTURING CO., LTD,
`Petitioner,
`
`v.
`
`GODO KAISHA IP BRIDGE 1,
`Patent Owner.
`____________
`
`Case IPR2017-018431
`Patent 7,893,501
`____________
`
`PATENT OWNER’S SUR-REPLY
`PURSUANT TO PAPER NO. 40
`
`
`
`1 IPR2017-01844 has been consolidated with this proceeding. See Paper 10 at 3.
`
`
`
`TABLE OF CONTENTS
`
`PETITIONER’S IMPROPER NEW REPLY ARGUMENT FAILS .............. 1
`A.
`Silicon Oxide Spacers Cannot Be Thermally Grown “On Top of
`The Silicide Layer” ............................................................................... 1
`B. Misra Teaches Only Silicon Nitride Spacers 23 ................................... 2
`PATENT OWNER MAINTAINS ITS IMPROPER NEW
`ARGUMENT OBJECTION ............................................................................ 4
`
`
`
`I.
`
`II.
`
`
`
`
`i
`
`
`
`TABLE OF AUTHORITIES
`
`
`CASES
`Ariosa Diagnostics v. Verinata Health, Inc. (“Ariosa”),
`805 F.3d 1359 (Fed. Cir. 2015) .............................................................................. 4
`In re: NuVasive, Inc. (“NuVasive”),
`841 F.3d 966 (Fed. Cir. 2016) ................................................................................ 4
`
`
`
`
`
`ii
`
`
`
`I.
`
`PETITIONER’S IMPROPER NEW REPLY ARGUMENT FAILS
`The Reply cited Misra 6:54-58 (not cited in the Petition) to support the new
`
`argument that Misra teaches that spacers 23 can be “made of a thermally grown
`
`silicon dioxide2 rather than silicon nitride.” Reply at 4 (citing Ex. 1232, ¶10 where
`
`Shanfield quotes Misra 6:54-58 in full and opines “[t]his disclosure [Misra 6:54-
`
`58] refers to a silicon oxide formed on top of the silicide layer” from “thermally-
`
`driven diffusion of excess silicon in the silicide.”). Petitioner and Shanfield are
`
`wrong about what this newly cited section of Misra discloses.
`
`A.
`
`Silicon Oxide Spacers Cannot Be Thermally Grown “On Top of
`The Silicide Layer”
`First, spacers 23 cannot be formed “on top of silicide layer” 18 because the
`
`top of silicide layer 18 is not exposed when spacers 23 are formed. Misra, Figs. 5-
`
`7. As shown in Figs. 2-3, “silicide layer 18 within the opening 24 is etched away”
`
`and “remove[d]” so sacrificial oxide 25 can thermally grow on the silicon
`
`substrate. Id. at 5:43-67, 6:34-36. Spacers 23 are formed “on top of the sacrificial
`
`oxide 25” in locations from which silicide layer 18 was removed—not “on top of
`
`the silicide layer” as Shanfield erroneously alleged. Id. at 6:37-42.
`
`
`2 The Reply and Ex. 1232 refer to silicon oxide and silicon dioxide. Both would be
`
`understood by a POSA to refer to SiO2 in this field. Regardless, arguments in this
`
`brief apply to whichever oxide of silicon allegedly replaces silicon nitride (“SiN”).
`
`1
`
`
`
`Second, silicon oxide spacers cannot grow by “thermally-driven diffusion”
`
`on oxide 25. See Sec. I.B. Thus, “thermally-driven diffusion” cannot form silicon
`
`oxide spacers 23 in Misra’s spacer embodiment (Fig. 7). Compare Ex. 1232, ¶10.
`
`Third, SiN layer 20 “is deposited overlying … silicide [] 18.” Misra 5:20-22,
`
`Figs. 2-7. Because SiN is a diffusion barrier, preventing oxidation of the silicide,
`
`silicon oxide cannot thermally grow on top of silicide 18 which is covered by SiN.
`
`Fourth, Shanfield mischaracterizes Misra as disclosing an inoperable device.
`
`If spacers 23 were somehow thermally grown silicon oxide on top of silicide layer
`
`18 as Shanfield suggests (i.e., if in Fig. 7 spacers 23 were silicon oxide, and
`
`elements 25 were silicide), the silicide under the spacers would electrically connect
`
`(short) gate 28b to source 26 and drain 28, rendering the device inoperable. Id.
`
`6:49-54 (spacers required to isolate the gate from silicide).
`
`B. Misra Teaches Only Silicon Nitride Spacers 23
`“After formation of the spacers,” ion implantation is used to dope region 31.
`
`Misra 6:42-58. A POSA would have understood this also “dopes” spacers 23 and
`
`decreases their ability to isolate the gate. Misra teaches the damaged SiN spacers
`
`may be replaced with (1) new SiN spacers or (2) new SiN spacers “composited
`
`with a sidewall thermal growth” (i.e., prior to re-depositing SiN spacers, silicon
`
`oxide is thermally grown on oxide 25 and exposed portions of the lateral surfaces
`
`of silicide 18). In both cases, silicon nitride covers the sides of Misra’s gate 28b.
`
`2
`
`
`
`For the first case, Misra 6:54-58 removes and replaces SiN spacers with SiN
`
`because doping (id. at 6:42-45) damages the spacers—decreasing their isolating
`
`properties. For that same reason, Misra removes doping-damaged silicon oxide 25
`
`not covered by spacers 23 and replaces it with silicon oxide 27. Id. at 6:42-49, 59-
`
`61. Thus, silicon nitride covers the sides of Misra’s gate 28b.
`
`For the second case, a POSA would have rejected Shanfield’s interpretation,
`
`because it is not physically possible to thermally grow silicon oxide spacers 23 and
`
`have an operable device. The surfaces exposed by removing SiN spacers 23 after
`
`the doping step (i.e., the step shown in Fig. 5, but with spacers 23 removed) cannot
`
`thermally grow silicon oxide spacers 23: (a) the silicon oxide surfaces 22 (green
`
`below) cannot because there is insubstantial free silicon; (b) SiN surfaces 20
`
`(purple below) cannot because SiN is a diffusion barrier; (c) increasing the
`
`thickness of oxide 25 (orange) to form spacers 23 is not possible as the oxide
`
`consumes silicon further into the doped regions 26, 28, and 31 (orange arrows)
`
`which would render the device inoperable; and (d) thermally growing spacers 23
`
`from exposed lateral surfaces of silicide 18 would also require consuming silicon
`
`from doped regions 26, 28, and 31 which would render the device inoperable.
`
`Thus, a POSA understood any thermally grown silicon oxide would need to be
`
`composited with deposited silicon nitride to form spacers 23. Id. 6:49-54. Silicon
`
`nitride would thus cover the sides of Misra’s gate 28b. For the same reasons, the
`
`3
`
`
`
`lateral surface of SiN layer 20 will be exposed while forming the composite spacer
`
`such that SiN layer 20 and SiN spacers 23 will be contiguous layers of a SiN film.
`
`
`
`II.
`
`PATENT OWNER MAINTAINS ITS IMPROPER NEW ARGUMENT
`OBJECTION
`Petitioner changed its theory of unpatentability to argue Misra 6:54-58
`
`discloses embodiments where spacers 23 are not formed of SiN. Reply at 4. This
`
`new theory relies on a part of Misra not cited in the Petition. Shanfield admitted
`
`his initial declaration did not cite Misra 6:54-58, and did not argue Misra discloses
`
`spacers 23 of a material other than SiN. Ex. 2232 at 36:18-37:7, 38:3-8; Ex. 2210
`
`at 287:21-288:4. This argument cannot be considered because (1) it is
`
`meaningfully distinct and (2) the Board denied Patent Owner expert testimony to
`
`rebut it. Papers 40, 42; Ariosa at 1367 (improper to cite “previously unidentified
`
`portions of a prior-art reference to make a meaningfully distinct contention”);
`
`NuVasive at 970-73 (APA requires PO have “arguments and evidence”).
`
`Dated: October 5, 2018
`
`
`
`
`Respectfully submitted,
`Godo Kaisha IP Bridge 1
`
`By /Gerald B. Hrycyszyn/
`Gerald B. Hrycyszyn, Reg. No. 50,474
`
`4
`
`
`
`CERTIFICATE OF SERVICE UNDER 37 C.F.R. §42.6 (e)(4)
`
`I certify that on October 5, 2018 I will cause a copy of the foregoing
`
`document, including any exhibits referred to therein, to be served via electronic
`
`mail, as previously consented to by Petitioner, upon the following:
`
`David L. Cavanaugh
`
`David.Cavanaugh@wilmerhale.com
`
`Dominic E. Massa
`
`Dominic.Massa@wilmerhale.com
`
`Michael H. Smith
`
`MichaelH.Smith@wilmerhale.com
`
`Date: October 5, 2018
`
`/Faye Miller/
`Faye Miller
`Paralegal
`WOLF GREENFIELD & SACKS, P.C.
`
`