臺灣博碩士論文加值系統

隨著二維（2D）材料其優越的性能與其具有晶圓級合成方法而引起了越來越多的關注，引起了人們的極大興趣並引發了相應器件應用的革命。然而目前對於二維材料實現奈米結構的大面積表徵、精準度、智能自動化和高效率的檢測還未達工業級水準的成功應用，因此在這我們透過了大數據分析與深度學習成功開發一套可見光超頻譜影像技術自動識別二維材料的光學層數特徵分析。
對於分類演算法的部分，我們嘗試提出決策樹演算法(Decision tree, DT)、深度神經網絡(Deep Neural Networks, DNN)、一維卷積神經網絡(1D Convolutional Neural Network,1D-CNN)與三維卷積神經網絡(3D Convolutional Neural Network,3D-CNN)模型以探討模型識別的準確性與二維材料光學特徵之間的關聯性，實驗結果表明三維卷積神經網絡在泛化能力優於其它分類模型，該模型適用空間域和頻譜域的特徵輸入，因此我們深入了解二維材料的生長與形貌演變機制，以方便我們對雲端資料庫內的數據進行資料探勘(Data Mining)，本研究優勢是能將提取算法之數據(薄膜邊緣區域的光學形態特徵值和頻譜特性)，應用於光學影像並結合超頻譜影像技術。
這樣的方法與先前研究的差異在於本研究無須特定基板，並且可經由自動光圈快門所給予在樣本上不同動態範圍(Dynamic Range, DR)區間之增益影像，因此無須調整到相同色對比條件下的成像品質，也不用傳統影像處裡過程。本實驗以二硫化鉬作為測試目標，該檢測系統具有以下優勢:
1.準確定量分析所呈現的覆蓋率(例如:薄膜厚度的識別、殘留物/污染物的存在)
2.可應用於無須特定或厚度的基板
3.最大視野(Field of View, FOV)識別範圍可達到1.6 mm x 1.2 mm
4.最佳識別解析度可達~100 nm
5.最佳檢測時間 30 sec/image
6.人性化視覺的非破壞性光學檢測

Two-dimensional (2D) materials with wafer-scale synthesis methods and fascinating properties, have attracted significant interest and triggered revolutions in corresponding device applications. However, the large-area characterization, precision, intelligent automation and high-efficiency detection of nanostructures for two-dimensional materials have not yet reached the industrial level. Therefore, we have successfully developed a set of visible hyperspectral imaging technology to analyze the optical layer characteristics of two-dimensional materials through big data analysis and deep learning.
For the part of the classification algorithm, we try to propose Decision tree(DT), DNN, 1D-CNN and 3D-CNN models to explore the correlation between the accuracy of model recognition and the optical characteristics of 2D materials. The experimental results show that the generalization ability of the 3D-CNN is better than other classification models, and the model is applicable to the feature input of the spectral–spatial information. Therefore, we have a deep understanding of the growth and morphology evolution mechanism of 2D materials, so that we can carry out Data Mining on the data in the cloud database. The research advantage is that the extraction algorithm can be applied to optical images combined with hyperspectral imaging techniques to include optical morphological and spectral characteristics from the edge regions of the film.
The difference between this method and the previous research is that this study does not need to the specific substrate and the image can be given to different DR intervals on the sample via the auto iris shutter, so there is no need to adjust the different color contrast of imaging quality, and does not use the traditional image processing. This experiment uses MoS_2 as the test target. The system has the following advantages:
1.Accurate quantitative analysis of coverage (e.g., identification of thickness, presence of residues/ impurity)
2.Applied to substrates without specific or thickness.
3.Maximum FOV recognition range up to 1.6 mm x 1.2 mm
4.Best recognition resolution up to ~100 nm
5.Best detection time 30 sec/image
6.Non-destructive optical inspection of humanized vision

目錄 - 4 -
圖目錄 - 7 -
表目錄 - 13 -
第一章 緒論 - 14 -
1-1 二硫化鉬薄膜之特性與應用 - 14 -
1-2 研究之重要性及先前技術 - 15 -
1-2-1 現有儀器分析MoS2薄膜厚度 - 15 -
1-2-2 目前光學顯微鏡針對MoS2進行層數之研究 - 17 -
1-2-3 透過機器學習識別MoS2層數之方法 - 18 -
1-3 超頻譜研究目的與方法 - 20 -
1-4研究動機 - 20 -
1-5 章節的架構規劃 - 21 -
第二章 理論與文獻探討 - 31 -
2-1 CVD成長二硫化鉬之探討 - 31 -
2-1-1 二硫化鉬(MoS2)成長機制 - 31 -
2-1-2 二硫化鉬(MoS2)的生長形貌演變 - 33 -
2-2 超頻譜影像 - 33 -
2-2-1 超頻譜影像之數據結構 - 34 -
2-2-2 分析二硫化鉬薄膜層數之光譜研究 - 34 -
2-3 機器學習與深度學習之相異點 - 36 -
2-4 機器學習-決策樹（DT）模型理論與應用 - 38 -
2-5 深度學習之模型基礎理論 - 39 -
2-5-1 深度神經網絡(Deep Neural Networks, DNN) - 39 -
2-5-2 一維卷積神經網絡(1D - CNN) - 40 -
2-5-3 三維卷積神經網絡(3D - CNN) - 41 -
第三章 實驗架構與方法 - 59 -
3-1 實驗樣品的製備 - 59 -
3-2 系統裝置和程序 - 59 -
3-3 可見光超頻譜影像技術(VIS-HSI) - 60 -
3-4 特徵數據處理過程 - 65 -
3-4-1 程序處理之架構 - 65 -
3-4-2 數據集劃分定義 - 66 -
3-4-3 資料預處理 - 67 -
3-5 模型建置與超參數 - 69 -
3-5-1 DT+PCA - 69 -
3-5-2 DNN - 69 -
3-5-3 1D-CNN - 70 -
3-5-4 3D-CNN - 70 -
第四章 結果與討論 - 83 -
4-1 顯微鏡影像光強度探討 - 83 -
4-2 RGB與頻譜特徵含量差異 - 84 -
4-3 ML與DL模型之泛化能力比較 - 85 -
4-4 三種DL模型在不同倍率下的結果 - 86 -
4-4-1 X100倍 - 86 -
4-4-2 X40倍 - 87 -
4-4-3 X10倍 - 88 -
4-4-4 大面積覆蓋率分析 - 88 -
4-4-5 三種倍率及模型之差異性 - 89 -
4-5 結果分析與儀器驗證 - 89 -
第5章 結論與未來展望 - 118 -
參考文獻 - 119 -
附錄一 - 125 -

[1]B.Radisavljevic, A.Radenovic, J.Brivio, V.Giacometti, and A.Kis, “Single-layer MoS2 transistors,” Nat. Nanotechnol., vol. 6, no. 3, pp. 147–150, 2011.
[2]K. F.Mak, C.Lee, J.Hone, J.Shan, and T. F.Heinz, “Atomically thin MoS2: A new direct-gap semiconductor,” Phys. Rev. Lett., vol. 105, no. 13, pp. 2–5, 2010.
[3]C.Ataca, H.Şahin, and S.Ciraci., “Stable, Single-Layer MX2 Transition-Metal Oxides and Dichalcogenides in a Honeycomb-Like Structure.pdf,” J. Phys. Chem. C, vol. 116, p. 8983−8999, 2012.
[4]Wu, W. et al., “High mobility and high on/off ratio field-effect transistors based on chemical vapor deposited single-crystal MoS2 grains,” Appl. Phys. Lett. 102, 142106, 2013.
[5]M. Sup Choi et al., “Controlled Charge Trapping by Molybdenum Disulphide and Graphene in Ultrathin Heterostructured Memory Devices, ” Nature Communications, vol. 4,p. 1624, Mar.2013.
[6]C.-H. Lee et al., “Atomically Thin p-n Junctions with Van Der Waals Heterointerfaces,” Nature Nanotechnology, vol. 9, no. 9, pp.676-681, Aug. 2014.
[7]44 Zhao, J.; Li, N.; Yu, H.; Wei, Z.; Liao, M.; Chen, P.; Wang, S.; Shi, D.; Sun, Q.; Zhang, G., “Highly Sensitive MoS2 Humidity Sensors Array for Noncontact Sensation,” Adv. Mater.2017, 29, 1702076, 10.1002/adma.201702076
[8]Sarkar, D. et al., “MoS2 field-effect transistor for next-generation label-free biosensors,” ACS Nano 8, 3992–4003, 2014.
[9]Y. J. Park et al., “All MoS2 based large area, skin-attachable active-matrix tactile sensor,” ACS Nano, vol. 13, no. 3, pp. 3023–3030, 2019.
[10]Huang B, Clark G, Navarro-Moratalla E, Klein DR, Cheng R, Seyler KL, et al., “ Layer-dependent ferromagnetism in a van der Waals crystal down to the monolayer limit,” Nature, 546:270-273, 2017.
[11]Geim AK, Novoselov KS., “ The rise of graphene,” Nat. Mater. 2007;6:183-191.
[12]Son, Y.; Wang, Q. H.; Paulson, J. A.; Shih, C. J.; Rajan, A. G.; Tvrdy, K.; Kim, S.; Alfeeli, B.; Braatz, R. D.; Strano, M. S., “ Layer number dependence of MoS2 photoconductivity using photocurrent spectral atomic force microscopic imaging,” ACS Nano 2015, 9, 2843–2855.
[13]C.Lee, H.Yan, L. E.Brus, T. F.Heinz, J.Hone, andS.Ryu, “Anomalous lattice vibrations of single- and few-layer MoS2,” ACS Nano, vol. 4, no. 5, pp. 2695–2700, 2010.
[14]H.Li et al., “From bulk to monolayer MoS2: Evolution of Raman scattering,” Adv. Funct. Mater., vol. 22, no. 7, pp. 1385–1390, 2012.
[15]Lei S, Ge L, Najmaei S, George A, Kappera R, Lou J, et al. “Evolution of the electronic band structure and efficient photo-detection in atomic layers of InSe, ” ACS Nano 2014;8:1263-1272.
[16]G.Ye et al., “Defects Engineered Monolayer MoS2 for Improved Hydrogen Evolution Reaction,” Nano Lett., vol. 16, no. 2, pp. 1097–1103, 2016.
[17]A. Zobel, A. Boson, P.M. Wilson, et al., “Chemical vapour deposition and characterization of uniform bilayer and trilayer MoS2 crystals,” J. Mater. Chem. C, 4 (47) (2016)
[18]Serna, M. I.; Yoo, S. H.; Moreno, S.; Xi, Y.; Oviedo, J. P.; Choi, H.; Alshareef, H. N.; Kim, M. J.; Minary-Jolandan, M.; Quevedo-Lopez, M. A., “Large-Area Deposition of MoS2 by Pulsed Laser Deposition with in Situ Thickness Control ” ACS Nano 2016, 10, 6054– 6061.
[19]Kuang-I Lin, Yen-Hung Ho, Shu-Bai Liu, Jian-Jhih Ciou, Bo-Ting Huang, Christopher Chen, Han-Ching Chang, Chien-Liang Tu, and Chang-Hsiao Chen, “Atom-Dependent Edge-Enhanced Second-Harmonic Generation on MoS2 Monolayers,” Nano Letters, 10.1021/acs.nanolett.7b04006 (2018).
[20]H.Li et al., “Optical identification of single- and few-layer MoS2 sheets,” Small, vol. 8, no. 5, pp. 682–686, 2012.
[21]H.Li et al., “Rapid and Reliable Thickness Identi fi cation of Two-Dimensional Nanosheets Using Optical Microscopy,” ACS Nano, no. Xx, pp. 10344–10353, 2013.
[22]Y. Y.Wang et al., “Thickness identification of two-dimensional materials by optical imaging,” Nanotechnology, vol. 23, no. 49, p. 495713, 2012.
[23]Y. Li, N. Dong, S. Zhang, K. Wang, L. Zhang, and J. Wang, “Optical identification of layered MoS2 via the characteristic matrix method,” Nanoscale 8(2), 1210–1215 (2016).
[24]Lin, X. et al, “Intelligent identification of two-dimensional nanostructures by machine-learning optical microscopy,” Nano Research 11, 6316–6324 (2018).
[25]G. A. Wilkin, X. Huang, “K-means clustering algorithms: Implementation and comparison,” Proc. Int. Multi-Symp. Comput. Comput. Sci., pp. 133-136, 2007-Aug.
[26]Y. Li, Y. Kong et al., “Rapid identification of two-dimensional materials via machine learning assisted optic microscopy,” J. Materiomics (2019).
[27]Simone Arrigoni, Giovanni Turra and Alberto Signoroni, “ Hyperspectral image analysis for rapid and accurate discrimination of bacterial infections: A benchmark study,” Computers in Biology and Medicine, 88, (60), (2017).
[28]G. Hughes, “On the mean accuracy of statistical pattern recognizers,” IEEE Trans. Inf. Theory, vol. IT-14, no. 1, pp. 55-63, Jan. 1968.
[29]B. Krishnapuram, L. Carin, M. A. T. Figueiredo, A. J. Hartemink, “Sparse multinomial logistic regression: Fast algorithms and generalization bounds,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 27, no. 6, pp. 957-968, Jun. 2005.
[30]Q. Wang, Z. Meng, X. Li, “Locality adaptive discriminant analysis for spectral–spatial classification of hyperspectral images,” IEEE Geosci. Remote Sens. Lett., vol. 14, no. 11, pp. 2077-2081, Nov. 2017.
[31]Q. Wang, J. Lin, Y. Yuan, “Salient band selection for hyperspectral image classification via manifold ranking,” IEEE Trans. Neural Netw. Learn. Syst., vol. 27, no. 6, pp. 1279-1289, Jun. 2016.
[32]Y. Yuan, J. Lin, Q. Wang, “Dual-clustering-based hyperspectral band selection by contextual analysis,” IEEE Trans. Geosci. Remote Sens., vol. 54, no. 3, pp. 1431-1445, Mar. 2016.
[33]H.Li et al., “Optical identification of single- and few-layer MoS2 sheets,” Small, vol. 8, no. 5, pp. 682–686, 2012.
[34]H.Li, J.Wu, Z.Yin, and H.Zhang, “Preparation and Applications of Mechanically Exfoliated Single-Layer and Multilayer MoS2 and WSe2 Nanosheets,” ACC. Chem. Res., vol. 47, no. 4, pp. 1067–1075, 2014.
[35]Y.Zhao et al., “Interlayer Breathing and Shear Modes in Few-Trilayer MoS2 and WSe2,” Nano Lett., vol. 13, no. 3, pp. 1007–1015, 2013.
[36]J.Jeon et al., “Layer-controlled CVD growth of large-area two-dimensional MoS2 films,” Nanoscale, vol. 7, no. 5, pp. 1688–1695, 2015.
[37]D.Dumcenco et al., “Large-Area Epitaxial Monolayer MoS2,” ACS Nano, vol. 9, no. 4, pp. 4611–4620, 2015.
[38]Lee, Y-H. et al., “Synthesis of large-area MoS2 atomic layers with chemical vapor deposition,” Adv. Mater. 24, 2320–2325 (2012).
[39]Najmaei, S. et al., “Vapour phase growth and grain boundary structure of molybdenum disulphide atomic layers,” Nature Mater. 12, 754–759 (2013).
[40]A. M.van derZande et al., “Grains and grain boundaries in highly crystalline monolayer molybdenum disulphide,” Nat. Mater., vol. 12, no. 6, pp. 554–561, 2013.
[41]G.Eda, H.Yamaguchi, D.Voiry, T.Fujita, M.Chen, and M.Chhowalla, “Photoluminescence from chemically exfoliated MoS2,” Nano Lett., vol. 11, no. 12, pp. 5111–5116, 2011.
[42]P. J.Mulhern, “Lithium intercalation in crystalline LixMoS2,” Can. J. Phys., vol. 67, p. 1049, 1989.
[43]A. N.Enyashin and G.Seifert, “Density-functional study of LixMoS2 intercalates (0⩽x⩽1),” Comput. Theor. Chem., vol. 999, no. 343, pp. 13–20, 2012.
[44]Kang, J.; Liu, W.; Banerjee, K., “High-Performance MoS2 Transistors with Low-Resistance Molybdenum Contacts,” Appl. Phys.Lett. 2014, 104, 093106.
[45]Radisavljevic, B., Radenovic, A., Brivio, J., Giacometti, V. & Kis, A., “ Single-layer MoS2 transistors,” Nature Nanotechnol. 6, 147–150 (2011)
[46]Kufer, D.; Konstantatos, G. Highly Sensitive, “Encapsulated MoS2 Photodetector with Gate Controllable Gain and Speed,” Nano Lett. 2015, 15, 7307−7313.
[47]Cheng, R.; Li, D.; Zhou, H.; Wang, C.; Yin, A.; Jiang, S.; Liu, Y.; Chen, Y.; Huang, Y.; Duan, X., “Electroluminescence and Photocurrent Generation from Atomically Sharp WSe2/MoS2 Heterojunction p-n Diodes,” Nano Lett. 2014, 14, 5590−5597.
[48]Wang, T.; Zhu, R.; Zhuo, J.; Zhu, Z.; Shao, Y.; Li, M., “ Direct Detection of DNA Below ppb Level Based on Thionin-Functionalized Layered MoS2 Electrochemical Sensors,” Anal. Chem. 2014, 86, 12064−12069
[49]Mak, K. F., McGill, K. L., Park, J. & McEuen, P. L., “ The valley Hall effect in MoS2 transistors,” Science 344, 1489–1492 (2014).
[50]Ji, Q.; Zhang, Y.; Zhang, Y.; Liu, Z. , “Chemical Vapour Deposition of Group-VIB Metal Dichalcogenide Monolayers: Engineered Substrates from Amorphous to Single Crystalline Chem,” Soc. Rev. 2015, 44, 2587– 2602
[51]Zhou, D.; Shu, H.; Hu, C.; Jiang, L.; Liang, P.; Chen, X., “Unveiling the Growth Mechanism of MoS2 with Chemical Vapor Deposition: From Two-Dimensional Planar Nucleation to Self-Seeding Nucleation,” Cryst. Growth Des. 2018, 18, 1012– 1019.
[52]Zhu D, Shu H, Jiang F, Lv D, Asokan V, Omar O, Yuan J, Zhang Z and Jin C., “Capture the growth kinetics of CVD growth of two-dimensional MoS2,” NPG 2D Mater. Appl. 18(2017).
[53]Wang, S. et al, “Shape Evolution of Monolayer MoS2Crystals Grown by Chemical Vapor Deposition,” Chem. Mater. 26, 6371–6379 (2014).
[54]Wang, L.; Chen, F.; Ji, X., “Shape Consistency of MoS2 Flakes Grown Using Chemical Vapor Deposition,” Appl. Phys. Express 2017, 10, 065201
[55]Z. Wei, Q. Yunfeng, F. Wei, C. Jingxin, Y. Huihui, W. Shiwei, J. Dechang, Z. Yu, H. PingAn., “Controlled growth of six-point stars MoS2 by chemical vapor deposition and its shape evolution mechanism,” Nanotechnology, 28 (2017), Article 395601
[56]Geng,D. et al., “Controlled growth of single-crystal twelve-pointed graphene grains on a liquid Cu surface,” Adv Mater. 26 6423–9(2014)
[57]H. E. Lin, P. A. N. Quan, D. I. Wei, and L. I. Yuan, “Research Advance on Target Detection for Hyperspectral Imagery,” Chinese J. Electron., vol. 37, no. 9, pp. 2016–2024, 2009.
[58] P. Shippert, “Introduction to hyperspectral image analysis,” Online J. Sp. Commun., p. 3, 2003.
[59] T. E. Bell, “Remote sensing,” IEEE Spectr., vol. 32, no. 3, pp. 24–31, 1995.
[60] Zhang, X., Chen, S., Xu, F. Combining Raman Imaging and Multivariate Analysis to Visualize Lignin, Cellulose, and Hemicellulose in the Plant Cell Wall. J. Vis. Exp. (124), e55910, doi:10.3791/55910 (2017).
[61]P. Shippert, “Introduction to hyperspectral image analysis,” Online J. Sp. Commun., 2003.
[62]H.Zhang et al., “Measuring the Refractive Index of Highly Crystalline Monolayer MoS2 with High Confidence,” Sci. Rep., vol. 5, no. 1, p. 8440, 2015.
[63] Zhang, H.; Ma, Y. G.; Wan, Y.; Rong, X.; Xie, Z. A.; Wang, W.; Dai, L., “Measuring the refractive index of highly crystalline monolayer MoS2 with high confidence,” Sci. Rep. 2015, 5, 8440.
[64] Palik, E. D., “Handbook of Optical Constants of Solids,” Elsevier: Amsterdam, 1997.
[65] Benameur MM, Radisavljevic B, Heron JS, Sahoo S, Berger H, Kis A., “Visibility of dichalcogenide nanolayers,” Nanotechnology 2011;22:125706.
[66] Blake P, Hill EW, Castro Neto AH, Novoselov KS, Jiang D, Yang R, et al., “ Making graphene visible,” Appl. Phys. Lett, pp. 91, 2007.
[67] A. Castellanos-Gomez, J. Quereda, H. P. van der Meulen, N. Agraït and G. Rubio-Bollinger., “Spatially resolved optical absorption spectroscopy of single-and few-layer MoS2 by hyperspectral imaging,” Nanotechnology, 2016, 27, 115705
[68] P. Louridas, C. Ebert, "Machine learning", IEEE Softw., vol. 33, no. 5, pp. 110-115, Sep./Oct. 2016.
[69] Y. LeCun, Y. Bengio, G. Hinton, "Deep learning", Nature, vol. 521, pp. 436-444, May 2015.
[70] T. Poggio, S. Smale., "The mathematics of learning: dealing with data" Not Am Math Soc, 50 (5) (2003), pp. 537-544
[71] Y. Xin, L. Kong, Z. Liu, Y. Chen, Y. Li, H. Zhu, C., "Wang Machine learning and deep learning methods for cybersecurity", IEEE Access, pp. 35365-35381, 6 (2018)
[72] Kotsiantis SB., "Decision trees: a recent overview", Artif Intell Rev. 2013;39(June 2011):261–83.
[73] W. Li, J. Du, and B. Yi, “Study on classification for vegetation spectral feature extraction method based on decision tree algorithm,” inInternational Conference on Image Analysis and Signal Processing (IEEE, 2011), pp. 665–669.
[74] Ghose, M. K.; Pradha R.; Ghose, S. S., “Decision Tree Classification Of Remotely Sensed Satellite Data Using Spectral Separability Matrix,” International Journal of Advanced Computer Science and Applications, Vol. 1, No.5, pp. 93-101, 2010.
[75] Z. Lin, Y. Chen, X. Zhao, G. Wang, "Spectral-spatial classification of hyperspectral image using autoencoders", Proc. 9th Int. Conf. Inf. Commun. Signal Process. (ICICS), pp. 1-5, Dec. 2013.
[76] Y. Chen, H. Jiang, C. Li, X. Jia, P. Ghamisi., “Deep feature extraction and classification of hyperspectral images based on convolutional neural networks,” IEEE Trans. Geosci. Remote Sens., vol. 54, no. 10, pp. 6232-6251, Oct. 2016.
[77] Y. Li, H. Zhang, Q. Shen., “Spectral-spatial classification of hyperspectral imagery with 3D convolutional neural network,” Remote Sens., vol. 18, no. 7, pp. 1527-1554, 2006.
[78] M. Anderson, R. Motta, and S. C. M. Stokes, “A Standard Default Color Space for the Internet - sRGB,” Color Imaging Conf., vol. 8, pp. 238–245, 1996.
[79] D. Đorđević, A. Hladnik, and A. Javoršek, “Performance Of Five Chromatic Adaptation Transforms Using Large Number Of Color Patches,” Acta Graph. Znan. časopis …, vol. 20, pp. 9–19, 2010.
[80] T. Smith and J. Guild, “The C.I.E. colorimetric standards and their use,” Trans. Opt. Soc., vol. 33, no. 3, pp. 73–134, 1931.
[81] Hotelling, Harold, “The Generalization of Student's Ratio,” Ann. Math. Statist. no. 3, 360-378, Feb. 1931.
[82] P. Evennett, “Kohler illumination: a simple interpretation,” Proc. R. Microsc. Soc. 28, 10–13, 1994.
[83] R. K.Attota, “Step beyond Kohler illumination analysis for far-field quantitative imaging: angular illumination asymmetry (ANILAS) maps,” Opt. Express24(20), 22616–22627.OPEXFF1094-4087, 2016.
[84] Hsiang-Chen Wang, Shih-Wei Huang, Jhe-Ming Yang, Guan-Huang Wu, Ya-Ping Hsieh, Shih-Wei Feng, Min Kai Lee and Chie-Tong Kuo, “Large-area few-layered graphene film determination by multispectral imaging microscopy,” Nanoscale, 7, 9033 - 9039, 2015.