Computational E‑Beam Lithography enables reliable nanofabrication by simulating the underlying physics of electron scattering and resist development. This modeling allows for accurate control over critical dimensions, further refined by optimizing beam shot placement to minimize line‑edge roughness and maximizing exposure throughput to mitigate tool drift. However, the cleanroom is an inherently noisy and stochastic environment where parameter uncertainties, such as beam blur, and unmodeled fluctuations in tool stability and chemistry limit absolute predictive accuracy. Recognizing the necessary balance between model complexity and practical usability, we resort to calibration methods to move predictions even closer to experimental results. This approach enables the transition from proof‑of‑concept hero devices to mature processes, significantly reducing the experimental iterations required for fully functional devices.
Speaker Bio
Marvin Zai holds a Ph.D. in EE from Imperial College London and is an applications engineer at GenISys.
· SEM image analysis, computational e‑beam lithography, numerical analysis and algorithms (DSA, C++)
· Software development for laboratory management systems (Cogiten Corp)
· Tech lead at a MIT startup (USA) and Omron Corp (Japan) for the Corning 2D Optical Cross‑connect, a MEMS actuator for a tunable VCSEL, the US Genomics rapid DNA sequencing chip (fab only), and a microfluidic mixer using flexural plate waves
· Owner of Inductive‑Coupled Plasma (ICP) tool for Bosch process, developed processes for deep silicon etching and wet etching of functional oxides and platinum films
· Materials science and tech transfer: created a new chemical route using metal organics for synthesizing functional oxide films for MEMS (Imperial College, London). Microstructural (XRD, SEM, TEM, XTEM) and electrical characterization of thin ceramic films. Technology transfer of PhD work to class 100 cleanrooms at Omron Corp. (Tsukuba, Japan).
Livestream the event on zoom (Yale login required)