{"title":"Deep learning with slow feature analysis for silicon single crystal growth state identification in Czochralski process","authors":"Jun-Chao Ren, Ding Liu, Zi-Xing Huang, Yin Wan","doi":"10.1016/j.jcrysgro.2025.128346","DOIUrl":null,"url":null,"abstract":"<div><div>The preparation of semiconductor silicon single crystal for advanced chip process requirements is the top priority of the current integrated circuit industry development. In this process, real-time and accurate identification of the growth state of Czochralski silicon single crystals (Cz-SSC) is the key to ensuring stable and improved crystal quality. Due to the switching of operating conditions, environmental disturbances and high dependence on operator experience, the process dynamics are highly complex, and minor anomalies are often masked by normal operating conditions, posing a serious challenge to real-time crystal growth state identification. To solve this problem, this study proposes a Cz-SSC growth state identification method that integrates slow feature analysis (SFA) and deep learning. Firstly, SFA extracts slow features reflecting the nature of process evolution at the time scale to reduce the masking effect of anomalies on minor anomalies from the source; subsequently, a multi-scale one-dimensional convolutional neural network (MS-1DCNN) is designed and a cross-attention mechanism is introduced for features extracted from convolutional kernels of various scales to achieve the weighted fusion of cross-scalar information, thus comprehensively capturing the discriminative patterns at different time scales. Finally, the experimental results demonstrate that the proposed method achieves superior performance in crystal growth state recognition, outperforming other approaches in terms of overall accuracy, recall, and F1-score. This Cz-SSC growth state identification method provides an effective solution for fine control of the semiconductor SSC manufacturing process.</div></div>","PeriodicalId":353,"journal":{"name":"Journal of Crystal Growth","volume":"670 ","pages":"Article 128346"},"PeriodicalIF":2.0000,"publicationDate":"2025-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Crystal Growth","FirstCategoryId":"88","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0022024825003008","RegionNum":4,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"CRYSTALLOGRAPHY","Score":null,"Total":0}
引用次数: 0
Abstract
The preparation of semiconductor silicon single crystal for advanced chip process requirements is the top priority of the current integrated circuit industry development. In this process, real-time and accurate identification of the growth state of Czochralski silicon single crystals (Cz-SSC) is the key to ensuring stable and improved crystal quality. Due to the switching of operating conditions, environmental disturbances and high dependence on operator experience, the process dynamics are highly complex, and minor anomalies are often masked by normal operating conditions, posing a serious challenge to real-time crystal growth state identification. To solve this problem, this study proposes a Cz-SSC growth state identification method that integrates slow feature analysis (SFA) and deep learning. Firstly, SFA extracts slow features reflecting the nature of process evolution at the time scale to reduce the masking effect of anomalies on minor anomalies from the source; subsequently, a multi-scale one-dimensional convolutional neural network (MS-1DCNN) is designed and a cross-attention mechanism is introduced for features extracted from convolutional kernels of various scales to achieve the weighted fusion of cross-scalar information, thus comprehensively capturing the discriminative patterns at different time scales. Finally, the experimental results demonstrate that the proposed method achieves superior performance in crystal growth state recognition, outperforming other approaches in terms of overall accuracy, recall, and F1-score. This Cz-SSC growth state identification method provides an effective solution for fine control of the semiconductor SSC manufacturing process.
期刊介绍:
The journal offers a common reference and publication source for workers engaged in research on the experimental and theoretical aspects of crystal growth and its applications, e.g. in devices. Experimental and theoretical contributions are published in the following fields: theory of nucleation and growth, molecular kinetics and transport phenomena, crystallization in viscous media such as polymers and glasses; crystal growth of metals, minerals, semiconductors, superconductors, magnetics, inorganic, organic and biological substances in bulk or as thin films; molecular beam epitaxy, chemical vapor deposition, growth of III-V and II-VI and other semiconductors; characterization of single crystals by physical and chemical methods; apparatus, instrumentation and techniques for crystal growth, and purification methods; multilayer heterostructures and their characterisation with an emphasis on crystal growth and epitaxial aspects of electronic materials. A special feature of the journal is the periodic inclusion of proceedings of symposia and conferences on relevant aspects of crystal growth.