圓二色性光譜(Circular Dichroism Spectroscopy)，簡稱CD光譜，是一種光譜技術，用於研究螺旋狀結構的生物大分子（如蛋白質和核酸）和其他螺旋結構的化學物質，在CD光譜實驗中，會將圓偏振光照射到樣品上，測量左旋圓偏振光和右旋圓偏振光在不同波長下吸收與散射效應整體的差異，而通過分析圓二色性光譜，可以獲得有關螺旋結構物質的結構資訊，例如:蛋白質的二級結構、核酸的構型等。
一般來說，將左右圓偏振光照射到手性金屬電漿奈米結構上，會產生強烈的CD光譜反應，因此基於這個機制，我們設計一個金屬的阿基米德螺旋(Archimedes spiral)光柵結構，並以左右圓偏振波照射去獲得在阿基米德螺旋間產生的CD光譜，論文中將根據改變阿基米德螺旋結構的幾何參數，以及對應幾何參數不同所量測到的CD特徵頻譜，來建立以CD特徵頻譜資料預測幾何結構參數的逆向神經網路，而有了逆向神經網路後，我們也能建立一個以幾何結構參數來預測CD光譜資料的前向神經網路，並且結合這兩種網路的模型，我們能夠用來取代傳統的光學量測與模擬工具(FDTD)，來對CD光譜進行結構的最佳化預測，此外，本篇論文的深度神經網路架構主要採用卷積層（Convolution Layer, CL）與全連接層（Full Connected Layer, FCL）為核心。其中，卷積層主要負責特徵提取，即直接對輸入的特徵頻譜進行處理和學習，而全連接層則為特徵被卷積層提取和學習後，將這些特徵作為輸入，進行分類或回歸。最後，將評估逆向神經網路與前向神經網路的預測效果。

Circular Dichroism Spectroscopy (CD spectroscopy) is a powerful tool in studying helical structures in macromolecules and other chemical structures. In this study, we focus on the CD spectral response of chiral metal plasmonic nanostructures. By designing a metallic Archimedean spiral grating structure, we observe its interaction with circularly polarized waves. An inverse and a forward neural network are established for bi-directional prediction - estimating geometric structure parameters from CD spectral data, and vice versa. The deep neural network architecture is composed of a convolution layer for feature extraction and a fully connected layer for performing classification or regression tasks.
The predictive performance of these neural networks is thoroughly evaluated.

[1] Stern B, Ji XC, Okawachi Y, Gaeta AL, Lipson M. Battery-operated integrated frequency comb generator. Nature 562, 401–405 (2018).
[2] Cheng QX, Bahadori M, Glick M, Rumley S, Bergman K. Recent advances in optical technologies for data centers: a review. Optica 5, 1354–1370 (2018).
[3] Walmsley I. Photonic quantum technologies. Proceedings of SPIE 11844, 11844OF (2021).
[4] Tittl A, Leitis A, Liu MK, Yesilkoy F, Choi DY et al. Imagingbased molecular barcoding with pixelated dielectric metasurfaces. Science 360, 1105–1109 (2018).
[5] RUI Guanghao, ZHAN Qiwen. Nanophotonic Methods for Chiral Sensing and Characterization（Invited）［J］,Acta Photonica Sinica,2022,51（5）：0551301.
[6] Yifei Mao, Yun Zheng, Can Li, Lin Guo, Yini Pan, Rui Zhu, Jun Xu, Weihua Zhang, and Wengang Wu, Programmable Bidirectional Folding of Metallic Thin Films for 3D Chiral Optical Antennas,Adv.Mater.2017,29,1606482.
[7] Shuang Liang, Wei Zhang, Jing Yuan, Jianli Jiang, Jisong Qian, Jing Shu, and Liyong Jiang, “In-plane gap plasmon induced strong circular dichroism in double-layer Archimedean planar metamaterials”, Optical Materials Express 2870, Vol. 8, No. 9,1 Sep 2018.
[8] Shuo Liu, Xiuguo Chen, Tianjuan Yang, Chunfu Guo, Jiahao Zhang, Jianyuan Ma, Chao Chen, Cai Wang, Chuanwei Zhang, and Shiyuan Liu, “Machine learning aided solution to the inverse problem in optical scatterometry: Supplementary Material”, Measurement,Volume 191, 15 March 2022, 110811.
[9] Yang Zhen-zhen, Kung Nan, Fan Lu, Kang Bin, “Review of Image Classification Algorithms Based on Convolutionl Neural Networks ”,Journal of Signal
Processing ,Vol.34,No.12,Dec.2018.
[10] Feng Yu , Yi Benshun, Wu Chenyue, Zhang Yungang,“Pulmonary Nodule Recognition
Based on Three-Dimensional Convolution Neural Network”, Acta Optica Sinica,
Vol.39,No.6, June,2019.
[11] 吳民耀、劉威志,表面電漿子理論和模擬,物理雙月刊28卷2期,pp486-496 (2006).
[12] C. Bohren and D. Huffman, “Absorption and Scattering of Light by Small Particles”
Wiley,New York (1983).
[13] 邱國斌、蔡定平,金屬表面電漿子簡介,物理雙月刊28卷2期,pp473-485 (2006).
[14] Stefan A. Maier, Surface Plasmon Polaritons at Metal / Insulator Interfaces, Plasmonics: Fundamentals and Applications,pp 21-37 (2007).
[15] Henrique T. M. C. M. Baltar, Krystyna Drozdowicz-Tomsia and Ewa M. Goldys,
Propagating Surface Plasmons and Dispersion Relations for Nanoscale Multilayer
Metallic-Dielectric Films, Plasmonics – Principles and Applications,CH6,pp135-155
(2012).
[16] “Chiral” 檢自 https://en.wikipedia.org/wiki/Chirality.
[17] Zhao, Yang, et al. "Chirality detection of enantiomers using twisted optical metamaterials." Nature communications 8 (2017): 14180.
[18] Hengjie Zhou, Shaojian Su, Huanxi Ma, Zeyang Zhao, Zhili Lin, Weibin Qiu, Pingping Qiu, Beiju Huang, and Qiang Kan, “Chiral graphene plasmonic Archimedes spiral nanostructure with tunable circular dichroism and enhanced sensing performance”, Optics Express 31954, Vol. 28, No. 21,12 October 2020.
[19] "Activation Functions." ML Cheatsheet, Read the Docs, 2023, ml-cheatsheet.readthedocs.io/en/latest/activation_functions.html.
[20] 葛德彪、閻玉波,電磁波時域有限差分方法(第二版),pp1-154 (2005).
[21] Y. Lecun, L. Bottou, Y. Bengio and P. Haffner, "Gradient-based learning applied to document recognition," in Proceedings of the IEEE, vol. 86, no. 11, pp. 2278-2324, Nov. 1998.
[22] S. Liu , X. Chen, T. Yang , C. Guo , J. Zhang , J. Ma, C. Chen, C. Wang, C. Zhang and S. Liu, “Machine learning aided solution to the inverse problem in optical scatterometry,” Measurement 191, 110811 (2022).
[23] J. Zhuang, T. Tang, Y. Ding et al., “Adabelief Optimizer: Adapting Stepsizes by the
Belief in Observed Gradients,” Advances in Neural Information Processing Systems, vol. 33, pp. 18795-18806, 2020.