Improved Simulation Method Enhances Plasma Modeling for Chip Manufacturing

Researchers from Princeton University, in collaboration with the University of Alberta, Los Alamos National Laboratory, and Applied Materials Inc., have developed a significantly improved method for simulating inductively coupled plasmas used in microchip manufacturing. These plasmas play a crucial role in etching patterns onto silicon, a key process in producing faster and more efficient chips. Traditional simulations using kinetic models have struggled with performance and reliability due to the complexity of tracking individual particles and the demand for real-time computation. The new method provides greater stability and accuracy by modifying how electric fields—specifically the solenoidal electric field—are calculated. Implemented as a particle-in-cell code, this approach ensures better energy conservation, avoiding the build-up of numerical errors that can distort physical simulations. The code now runs in two spatial dimensions, making it a practical and scalable tool for industrial use. The breakthrough, supported under a U.S. Department of Energy agreement, promises advances not only in semiconductor technology but also in fusion research, where understanding plasma behavior is equally essential.