公开(公告)号：US20250003866A1

公开(公告)日：2025-01-02

申请号：US18461487

申请日：2023-09-05

Applicant: HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY

Inventor： Xiuguo Chen ,  Shuo Liu ,  Shiyuan Liu

IPC: G01N21/21 ,  G01B11/24 ,  G06N3/0464 ,  G06N3/084

Abstract: Disclosed is a method for measuring optical properties and geometric properties of a thin film material, which belongs to the field of ellipsometry. The method realizes the automation of the ellipsometry data analysis with accurate results and includes using a spline model and a forward optical property model to generate a training set; training a neural network model; sequentially inputting preliminary results of geometric parameters and spline parameters of a material obtained by inputting the measured optical characterization quantity into the neural network into the spline model and the forward optical property model to obtain a theoretical optical characterization quantity. After optimizing the neural network model by using the deviation between the theory and the measured optical characterization quantity, the predicted value of the output is the final geometric and optical property parameters of the material.