[8p-B11-1~16] 8.2 Plasma deposition of thin film, plasma etching and surface treatment

6:15 PM - 6:30 PM JST(9:15 AM - 9:30 AM UTC)

[8p-B11-16] Surrogate Model Development and Evaluation for Initial-Value Prediction in Plasma Simulations

〇Koki Hayashi¹, Daiki Kawahito¹, Yukiya Saito¹, Hironori Moki¹ (1.Tokyo Electron Ltd.)

Keywords:

machine learning,low pressure plasma

In predicting spatial plasma distributions inside semiconductor manufacturing equipment, physics-based simulations offer high accuracy but require substantial computational cost, whereas surrogate models are fast but limited in extrapolation capability and output physical quantities. We propose a method that uses surrogate predictions as the initial distribution for plasma simulations. Using the CCP GEC reference cell, we evaluate the relationship between surrogate accuracy and computational acceleration, and demonstrate its effectiveness for continuous condition exploration and its applicability to design and process optimization.