基于云模型量子进化算法的薄膜微观结构表征  

Characterization of Microstructure of Thin Film Based on Cloud Model Quantum Evolution Algorithm

作　　者：王一名[1] 张超[1] 孙秀平[1] 匡尚奇[1] 杨海贵 WANG Yiming;ZHANG Chao;SUN Xiuping;KUANG Shangqi;YANG Haigui(School of Science,Changchun University of Science and Technology,Changchun 130022;Advanced manufacturing technology for optical systemslaboratory,Changchun Institute of Optics,Fine Mechanics and Physics,Chinese Academy of Sciences,Changchun 130033)

机构地区：[1]长春理工大学理学院,长春130022 [2]中国科学院长春光学精密机械与物理研究所光学系统先进制造技术重点实验室,长春130033

出　　处：《长春理工大学学报（自然科学版）》2018年第4期20-24,30,共6页Journal of Changchun University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金青年基金项目(61405189);吉林省科技发展计划项目(20150101019JC;20170312024ZG)

摘　　要：掠入射X射线反射(GIXR)是光学薄膜表征的优选方法,但它需要建立相应的物理模型并采用一定的算法进行拟合求解。针对目前普遍使用的进化算法存在着种群规模大、收敛速度慢和拟合精度差的问题,本文将云模型与量子进化算法相结合,提出了一种基于云模型量子进化算法(CQEA)的薄膜微观结构表征方法,并将该算法应用于Si单层膜和周期极紫外Mo/Si多层膜的表征之中,并将其拟合结果与基于进化算法的拟合结果进行对比。分析表明,CQEA具有种群规模小、收敛速度快和拟合精度高的优势,同时给出的薄膜结构参数值精度更高。相关研究工作体现了CQEA应用于薄膜微观结构表征中的可行性与优越性,为基于GIXR的薄膜微观结构表征提供一种更优选的拟合求解方法。The grazing incidence X-ray reflection（GIXR） is regard as the optimal characterization method of optical thin film. However,the fitting of GIXR data depends on proper structural model of thin film and a powerful search algorithm. So far the evolutionary algorithm（EA） is widely used for the fitting of GIXR data of thin film,but it has several shortcomings such as mass computation,slow convergence and poor accuracy. Therefore,a cloud model quantum evolutionary algorithm（CQEA） has been developed,and this algorithm is used to fit GIXR data of optical thin film. In order to test the feasibility of CQEA in the fitting of GIXR data,this algorithm has been applied to fit the GIXR curves of Si single layers and periodic Mo/Si multilayers;and the microstructures of these optical coatings have been obtained. Furthermore,a comparison between the fittings of GIXR data based on EA and CQEA has been made;and it is found that CQEA has the advantages of small population,higher efficiency and higher accuracy compared with EA;and then this mathematical procedure based on CQEA can supply the film structure parameters in a higher precision. These researches demonstrate a great potential of CQEA in the curve fitting of GIXR data of optical coatings,which can supply a precise characterization of microstructure of optical thin film.

关 键 词：薄膜光学 薄膜表征 掠入射X射线反射 云模型量子进化算法 极紫外多层膜 单层膜 

分 类 号：O434.1[机械工程—光学工程]