EUV Mask Manufacturing Readiness in the Merchant Mask Industry

摘要

As nodes progress into the 7nm and below regime, extreme ultraviolet lithography (EUVL) becomes critical for all industry participants interested in remaining at the leading edge. One key cost driver for EUV in the supply chain is the reflective EUV mask. As of today, the relatively few end users of EUV consist primarily of integrated device manufactures (IDMs) and foundries that have internal (captive) mask manufacturing capability. At the same time, strong and early participation in EUV by the merchant mask industry should bring value to these chip makers, aiding the wide-scale adoption of EUV in the future. For this, merchants need access to high quality, representative test vehicles to develop and validate their own processes. This business circumstance provides the motivation for merchants to form Joint Development Partnerships (JDPs) with IDMs, foundries, Original Equipment Manufacturers (OEMs) and other members of the EUV supplier ecosystem that leverage complementary strengths. In this paper, we will show how, through a collaborative supplier JDP model between a merchant and OEM, a novel, test chip driven strategy is applied to guide and validate mask level process development. We demonstrate how an EUV test vehicle (TV) is generated for mask process characterization in advance of receiving chip maker-specific designs. We utilize the TV to carry out mask process "stress testing" to define process boundary conditions which can be used to create Mask Rule Check (MRC) rules as well as serve as baseline conditions for future process improvement. We utilize Advanced Mask Characterization (AMC) techniques to understand process capability on designs of varying complexity that include EUV OPC models with and without sub-resolution assist features (SRAFs). Through these collaborations, we demonstrate ways to develop EUV processes and reduce implementation risks for eventual mass production. By reducing these risks, we hope to expand access to EUV mask capability for the broadest community possible as the technology is implemented first within and then beyond the initial early adopters.