摘要
为了计算纳米粒子大尺寸聚集体的表面局域电磁场分布并快速对其增强效果进行评价,利用软件中脚本语言编写局部亚网格程序来实现对纳米粒子大尺寸聚集体模型的非均匀网格离散,并结合时域有限差分(FDTD)方法实现三种尺寸聚集体模型的电磁场仿真;使用K均值聚类算法对计算出的电场数据进行聚类分析,最终得到能够反映金纳米粒子大尺寸聚集体所有“热点”位置处电磁增强效果的平均增强因子。结果表明,使用亚网格离散的金纳米球二聚体仿真模型后的内存占用减少了81%且仿真速度提高1倍,有效提升FDTD的仿真效率;另外,通过K均值聚类算法并根据三种尺寸的金纳米粒子聚集体电磁数据,可以得到与传统积分法计算的平均增强因子(AEF 1)增减规律相同的增强因子AEF 2。
In order to calculate the surface of the nanoparticles large size aggregate local electromagnetic fields distribution and to evaluate its effect,the scripting language is used to write the local and the grid in the software program to implement the nanoparticles large size aggregate model of discrete non-uniform grid,combined with the finite difference time domain(FDTD)method to implement three sizes aggregate model of electromagnetic field simulation.The K-means clustering algorithm is used to cluster the calculated electric field data,and finally the average enhancement factor is obtained which could reflect the electromagnetic enhancement effect at all“hot spots”locations of the large-size aggregation of gold nanoparticles.The results show that the memory usage of the sub-grid discrete gold nanosphere dimer simulation model is reduced by 81%,and the simulation speed is doubled,which effectively improves the simulation efficiency of FDTD.In addition,the enhancement factor AEF 2,which is the same as the average enhancement factor(AEF 1)calculated by traditional integration method,can be obtained by K-means clustering algorithm and electromagnetic data of the aggregate of gold nanoparticles of three sizes.
作者
武忠义
史晓凤
马丽珍
马君
Wu Zhongyi;Shi Xiaofeng;Ma Lizhen;Ma Jun(Optics and Optoelectronics Laboratory of Qingdao,Ocean University of China,Qingdao,Shandong 266100,China)
出处
《激光与光电子学进展》
CSCD
北大核心
2021年第21期285-291,共7页
Laser & Optoelectronics Progress
基金
国家自然科学基金(40906051,41476081)
山东省重点研发计划(2016GSF115020,2019GHY112027)
山东省自然科学基金面上项目(ZR2020MF121)。
关键词
表面光学
表面增强拉曼基底
纳米粒子大尺寸聚集体
局部亚网格
K均值聚类
平均增强因子
optics at surfaces
surface-enhanced Raman substrate
large size aggregates of nanoparticles
local sub-grid
K-means clustering
average enhancement factor
