In response to the difficulties posed by the resolution, line edge roughness, sensitivity (RLS) trade-off to traditional chemically amplified resist (CAR) systems used for extreme ultraviolet lithography, a number of novel resist technologies have been proposed. In this paper, the effect of quencher loading on three resist technologies is analyzed via an error propagation-based resist simulator. In order of increasing novelty as well as complexity, they are: conventional CAR with quencher, CAR with photodecomposable base, and PSCAR 2.0, a CAR system with photodecomposable base as well as an EUV-activated UV-sensitive resist component. Simulation finds the more complicated resist systems trade in an increase in resist stochastics for improved deprotection slopes, yielding a net benefit in terms of line width roughness.

No abstract

Continued lithographic scaling using high-NA EUV scanners requires materials and processes with sufficient resolution and stochastic performance to translate the aerial image into thin film photoresist material. Amongst these key processes is photoresist dissolution that converts latent exposure chemistry in the photoresist into a developed pattern. However, co-optimization of resist materials and the develop process is difficult due to the challenge of directly measuring resist dissolution at the nanometer spatial and sub-second temporal scales on which it occurs. Most metrology of the dissolution process thus ignores either the temporal component by measuring just the final developed structure, or the spatial component as is done in most dissolution rate monitoring experiments. To overcome these challenges, we have developed an in-situ dissolution rate monitoring technique using high-speed atomic force microscopy (AFM). As opposed to pioneering work using AFM to monitor the dissolution process, our technique incorporates the use of a specially-designed flow cell which provides precise control of the time at which developer is introduced to the photoresist material, as well as delivery of nearly full-strength developer in fractions of a second. Our system thus offers the ability to probe the spatially-dependent nature of the dissolution process at conditions close to those in the fab, providing insight into exposure-dependent dissolution rate gradient, material swelling, and other potentially lithographically relevant phenomena such as polymer entanglement. In doing so, we provide another technique to aid in the design and study of photoresist materials for future lithographic nodes.

No abstract