Direct Emissions Reduction in Plasma Dry Etching Using Alternate Chemistries: Opportunities, Challenges, and Need for Collaboration

Abstract

—Plasma Dry-Etch (DE) is one of the key unit-operations in semiconductor manufacturing that use greenhouse gases (GHG) as feed gas (Donnelly and Kornblit, 2013). The exhaust GHG emission reduction or mitigation is one of the main focuses of scope 1 emission reduction at Micron Technology Inc. The reduction and mitigation approaches have been strategized in focus-tiers in order of proximity to the source of emissions. The focus-tiers upstream of exhaust are avoidance, replacement, reduction and downstream of exhaust are recovery/capture/recycle, abatement. This paper focuses on the replacement focus-tier that pertains to replacing high-emission feed gases (HE gas, feedgas that will produce relatively high kgCO2e through exhaust) with relatively low-emission feed gases (LE gas, feedgas that will produce relatively low kgCO2e through exhaust). The paper presents replacement opportunities and challenges through an evaluation study of Carbonyl Floride (COF2) as a replacement gas for NF3 or CF4 as a DE in-situ plasma chamber cleans gas. In conclusion, direct emissions from DE chamber cleans can be lowered by replacing NF3 and CF4 GHGs with COF2 by 90% or more. However, this replacement would require additional safety measures and abatement in operations due to increased toxicity and reactivity of COF2, along with cost roadmap to make its adoption economically feasible. Similar and possibly additional challenges would arise with other replacement options. To overcome challenges in replacement strategy focus-tier, it will require strong industry level collaboration between chemical suppliers, original equipment manufacturers (OEMs), device manufacturers, semiconductor research and collaboration centers and university research groups.