PTAB

IPR2020-00786

SMIC Americas v. Innovative Foundry Technologies LLC

Key Events
Petition
petition Intelligence

1. Case Identification

2. Patent Overview

  • Title: In-Situ Nitride/Oxynitride Processing With Reduced Deposition Surface Pattern Sensitivity
  • Brief Description: The ’226 patent describes semiconductor devices with enhanced channel carrier mobility. The technology involves using a stressed silicon nitride liner over transistor structures to induce strain in the channel and, optionally, adding a transitional silicon oxynitride layer to reduce surface pattern sensitivity during the deposition of an overlying dielectric layer.

3. Grounds for Unpatentability

Ground 1: Claims 3-5 and 7 are obvious over Ito and Ko

  • Prior Art Relied Upon: Ito (Application # 2002/0081794) and Ko (Patent 7,101,742).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Ito disclosed the foundational semiconductor device of claim 1, including a plurality of transistors on a substrate with a conformal stressed nitride liner and an overlying dielectric layer filling the gaps. Petitioner asserted that Ko supplied the remaining limitations for the dependent claims, specifically teaching the use of a strained silicon (Si) layer on a silicon-germanium (SiGe) layer (claim 3) to improve carrier mobility for both P-channel (claim 4) and N-channel (claim 5) transistors. Ko also taught using composite sidewall spacers comprising an oxide liner and a nitride layer (claim 7) to improve device isolation.
    • Motivation to Combine: A Person of Ordinary Skill in the Art (POSITA) would combine Ko’s well-known strained Si/SiGe substrate with Ito’s device to achieve Ito's goal of improving carrier mobility, as this was a conventional technique for performance enhancement. Additionally, as both Ito and Ko recognized the trend of shrinking transistor sizes, a POSITA would have been motivated to incorporate Ko’s composite spacers into Ito's device to address the known problems of leakage and element overlap that arise from miniaturization.
    • Expectation of Success: Petitioner contended success was expected because both references operate in the same field of semiconductor fabrication, address improving transistor performance, and use similar structures and manufacturing processes. Ko’s semiconductor device includes many of the same elements in a similar configuration as Ito’s, ensuring predictable integration.

Ground 2: Claims 8 and 9 are obvious over Brigham, Smith, and Chou

  • Prior Art Relied Upon: Brigham (Patent 6,046,494), Smith (Patent 6,255,233), and Chou (Patent 6,348,389).

  • Core Argument for this Ground:

    • Prior Art Mapping: Petitioner asserted that Brigham, like Ito, taught a base semiconductor device consistent with claim 1, including transistors with a conformal stressed silicon nitride layer and an overlying interlevel dielectric. The combination with Smith was alleged to render claim 2 obvious by teaching the insertion of a silicon oxynitride layer between the nitride and dielectric layers to solve adhesion problems. Finally, Petitioner argued that Chou and Smith taught the specific thickness ranges recited in claims 8 and 9. Chou disclosed an oxynitride layer thickness of 20 to 1,000 Å, overlapping the claimed 20 to 40 Å range (claim 8), while Smith disclosed a dielectric layer thickness of 5,000 to 30,000 Å, encompassing the claimed 6,500 to 7,500 Å range (claim 9).
    • Motivation to Combine: A POSITA would combine Smith with Brigham to solve the well-known problem of poor adhesion between nitride and oxide layers, which adversely affects transistor performance. The addition of a transitional oxynitride layer as taught by Smith was a known solution to reduce interfacial stress. A POSITA would then look to a reference like Chou to optimize the thickness of this oxynitride "buffer" layer, making its selection a routine design choice. Similarly, selecting a dielectric thickness from the range taught by Smith was argued to be an obvious optimization for providing sufficient device isolation.
    • Expectation of Success: Success was expected as all three references are in the same technical field and address common fabrication challenges. Petitioner argued that combining the layers as taught by the references was predictable, especially since Smith and Chou describe using the same chemical vapor deposition techniques as Brigham and note the benefits of forming the layers in a single processing chamber to avoid contamination.
  • Additional Grounds: Petitioner asserted additional obviousness challenges, including claims 1 and 6 over Ito alone; claim 2 over Ito and Smith; claims 1 and 6 over Brigham alone; claim 2 over Brigham and Smith; and claims 3-5 and 7 over Brigham and Ko. These grounds relied on similar arguments about combining known elements for their predictable functions.

4. Key Claim Construction Positions

  • Petitioner argued for a construction of the terms “a conformal stressed nitride liner” and “a conformal silicon oxynitride layer.”
    • Petitioner proposed that a "conformal" layer was understood by a POSITA to mean a layer of uniform thickness that follows the contours of the underlying topography. This construction was asserted to be critical, as Petitioner argued the prior art liners and layers (e.g., in Ito and Brigham) were deposited using processes that result in such uniform, contour-following coverage, thereby meeting the claim limitation.

5. Arguments Regarding Discretionary Denial

  • Petitioner argued that institution should not be denied based on parallel district court proceedings.
    • Petitioner asserted that although parallel litigation involving the ’226 patent existed, the courts had not yet invested significant time or resources in analyzing the prior art invalidity issues presented in the petition. Therefore, discretionary denial under 35 U.S.C. §314(a), considering the Fintiv factors, would be inappropriate and would not serve the interests of judicial economy.

6. Relief Requested

  • Petitioner requests institution of an inter partes review (IPR) and cancellation of claims 1-9 of the ’226 patent as unpatentable.