A Transformer-Based Sequence Prediction Architecture for Etch Profile Evolution in Semiconductor Manufacturing
- 3 March 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Semiconductor Manufacturing
- Vol. 39 (2) , 248-255
Abstract
The precise control of three-dimensional feature profiles during plasma etching is a fundamental challenge in nanoscale device fabrication, directly impacting performance and yield. While conventional physics-based simulations can capture the underlying plasma-surface interactions, their prohibitive computational cost limits their use for large-area analysis or rapid process development. To bridge this gap, we introduce EtchFormer, a Transformer-based deep learning framework engineered as a high-speed virtual metrology engine for 3D etch profile prediction. Crucially, the model is trained on a comprehensive dataset generated by the physically-calibrated, multi-scale TCAD simulations described in our methodology. This dataset, comprising over 50,000 time-resolved frames from thousands of etching sequences, embeds the fundamental physics of ion-enhanced etching, surface passivation, and byproduct transport across a wide spectrum of industrial process conditions and complex nanoscale patterns. The EtchFormer framework demonstrates high-fidelity reconstruction of entire etch sequences (Structural Similarity Index > 0.945) and, critically, exhibits robust long-range forecasting capabilities by accurately extrapolating profile evolution beyond the temporal range of the training data. Validated on complex layouts, EtchFormer reduces the prediction time for a complete sequence from hours to subseconds (~0.3 s). This capability provides a foundation for the virtual metrology, enabling the exploration of complex process-pattern dependencies, the prediction of structural anomalies, and the accelerated design and optimization of etching processes.Keywords
Funding Information
- National Key Research and Development Program of China (2022YFF0605800)
- Zhejiang Key Research and Development Project (2024SJCZX0029)
- Zhejiang Key Research and Development Project (2024SJCZX0030)
- Zhejiang Key Research and Development Project (2024SJCZX0031)
- Scientific Research Fund of Liaoning Provincial Education Department
This publication has 20 references indexed in Scilit:
- Optimizing Plasma Etching: Integrating Precise Three-Dimensional Etching Simulation and Machine Learning for Multi-Objective OptimizationIEEE Access, 2024
- MR-Transformer: Multiresolution Transformer for Multivariate Time Series PredictionIEEE Transactions on Neural Networks and Learning Systems, 2023
- Optical Proximity Correction Using Machine Learning Assisted Human DecisionIEEE Photonics Journal, 2022
- Video prediction by efficient transformersImage and Vision Computing, 2022
- VPTR: Efficient Transformers for Video Prediction2022
- Machine Learning for CAD/EDA: The Road AheadIEEE Design & Test, 2022
- TCAD augmented generative adversarial network for hot-spot detection and mask-layout optimization in a large area HARC etching processPhysics of Plasmas, 2022
- Hotspot Detection via Multi-task Learning and Transformer Encoder2021
- From IC Layout to Die Photograph: A CNN-Based Data-Driven ApproachIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2020
- Machine Learning-Guided Etch Proximity CorrectionIEEE Transactions on Semiconductor Manufacturing, 2016