PTAB

IPR2021-00752

Microchip Technology Inc v. HD Silicon Solutions LLC

Log In

1. Case Identification

2. Patent Overview

  • Title: Metal Stack for Local Interconnect Layer
  • Brief Description: The ’033 patent discloses a method of forming a local interconnect layer for integrated circuits. The method involves depositing a first film comprising titanium nitride (TiN) over an oxide layer, followed by depositing a second film comprising tungsten (W) over the TiN film to form a conductive metal stack.

3. Grounds for Unpatentability

Ground 1: Obviousness over Trivedi and Joo - Claims 2-6, 15, and 16 are obvious over Trivedi in view of Joo.

  • Prior Art Relied Upon: Trivedi (Patent 5,847,463) and Joo (Patent 6,534,401).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Trivedi taught the foundational method of forming a local interconnect with a TiN barrier layer and an overlying tungsten-silicide layer. Joo was argued to supply the missing elements required by the dependent claims and independent claim 15. Specifically, Petitioner asserted that Joo taught that both TiN and tungsten layers can be deposited by physical vapor deposition (PVD), a known sputtering process, which would render their in-situ deposition in the same tool (claims 2 and 15) obvious. Joo also allegedly disclosed specific thickness ranges for TiN and tungsten layers (e.g., TiN <100Å, W 200-700Å) that overlap with or otherwise render obvious the claimed thicknesses in claims 4-6 and 16 (e.g., total stack ≤600 Angstroms, TiN film ≤300 Angstroms).
    • Motivation to Combine: A person of ordinary skill in the art (POSITA) would combine Trivedi and Joo because both address fabricating similar semiconductor interconnect structures. A POSITA would be motivated to modify Trivedi’s process by applying Joo’s teachings on using the same deposition method (PVD) for both layers and adopting Joo’s layer thicknesses. The motivation would be to improve manufacturing efficiency, reduce complexity, and obtain a predictable, functional interconnect with desirable properties like low resistance.
    • Expectation of Success: A POSITA would have a reasonable expectation of success, as the combination involved applying known deposition techniques and established thickness parameters from Joo to the similar base process of Trivedi to achieve predictable improvements in a well-understood field.

Ground 2: Obviousness over Trivedi and Yang - Claim 8 is obvious over Trivedi in view of Yang.

  • Prior Art Relied Upon: Trivedi (Patent 5,847,463) and Yang (a 1997 IEDM Technical Digest publication).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner asserted that Trivedi disclosed all limitations of independent claim 1. Yang was introduced to teach the specific limitation of dependent claim 8: a metal stack having a sheet resistance equal to or less than about 10 Ohm/sq. Petitioner argued that while Trivedi’s stack inherently has some sheet resistance, Yang explicitly disclosed TiN/W stacks with specific layer thicknesses and processing that result in a sheet resistance of 2.06 Ω/sq, well below the claimed threshold.
    • Motivation to Combine: A POSITA would combine Trivedi and Yang as both describe highly similar interconnect structures featuring a tungsten-based layer over a TiN barrier. A POSITA would be motivated to modify Trivedi's tungsten-silicide layer to the pure tungsten layer taught by Yang and apply Yang's thickness parameters. The goal would be to achieve the known and desirable property of low sheet resistance, a critical performance metric for interconnects that Yang explicitly studied and quantified.
    • Expectation of Success: Success would be expected because sheet resistance is a predictable physical property dependent on material choice and layer thickness. Yang provided explicit, successful data for a nearly identical material stack, making the outcome of the combination predictable.

Ground 3: Obviousness over Trivedi and Lien - Claims 2, 11, and 12 are obvious over Trivedi in view of Lien.

  • Prior Art Relied Upon: Trivedi (Patent 5,847,463) and Lien (Patent 6,103,623).

  • Core Argument for this Ground:

    • Prior Art Mapping: Petitioner argued Trivedi taught the base method of claim 1, while Lien provided teachings for claims 2, 11, and 12. For claim 2, Lien was asserted to teach in-situ deposition of a TiN layer and a tungsten layer. For claim 11 (etching the second film with a fluorine-based etchant), Lien was argued to teach using SF6 (a fluorine-based etchant) to etch a tungsten layer. For claim 12 (etching the first film with a chlorine-based etchant), Lien allegedly taught using a chlorine-based etchant in a reactive ion etching (RIE) process to remove a TiN layer.
    • Motivation to Combine: A POSITA would combine Trivedi and Lien to improve Trivedi's fabrication process with known, alternative techniques. Lien provided methods for in-situ deposition and specific etchants (fluorine- and chlorine-based) that a POSITA would find advantageous for improving process efficiency or using less corrosive etching techniques (Lien's RIE vs. Trivedi's wet etch).
    • Expectation of Success: The combination was argued to be predictable, as it involved substituting known, discrete process steps from Lien into Trivedi's similar framework to achieve the well-understood benefits associated with those steps.

  • Additional Grounds: Petitioner asserted additional obviousness challenges based on Trivedi alone; Trivedi in view of Joo and Yang; Trivedi in view of Johnson; and Trivedi in view of Joo and Johnson, relying on similar arguments of simple substitution and optimization of a known process.

4. Arguments Regarding Discretionary Denial

  • Petitioner argued that discretionary denial under §314(a) based on the Fintiv factors would be inappropriate. The core arguments were that no trial date had been set in the parallel district court litigation, and a final written decision in the IPR would likely issue before any potential trial. Petitioner also asserted that the petition was filed early in the litigation timeline, before significant investment by the parties or the court. Furthermore, the petition challenges all 17 claims of the ’033 patent, making it likely to address issues and claims not presented in the district court. Finally, Petitioner contended that the strength of its seven invalidity grounds weighs strongly in favor of institution.

5. Relief Requested

  • Petitioner requests institution of inter partes review and cancellation of claims 1-17 of Patent 6,774,033 as unpatentable.