( 您好!臺灣時間:2021/05/07 23:27
字體大小: 字級放大   字級縮小   預設字形  


論文名稱(外文):Electrolyte molecule design by using machine learning
指導教授(外文):Wen-Dung Hsu
外文關鍵詞:lithium-ion batteryelectrolytemachine learning
  • 被引用被引用:0
  • 點閱點閱:44
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
近年來,由於電動汽車的蓬勃發展,高電壓鋰離子電池近期獲得眾多的關注。但 是傳統的碳酸酯溶劑的電解質對高壓電極的穩定性較差,因此設計適合的電解液 分子成為主要的研究目標。本研究討論運用機器學習為高電壓鋰離子電池設計新 型電解液的方法。首先為電解質分子的選擇設置了不同的篩選標準。在這些標準 中,電解質分子的兩個重要性質是還原電位(RP)以及氧化電位(OP)。此兩性 質決定電解質的工作電壓。電解質在此兩個電壓區間中不會發生氧化還原反應, 我們建立機器學習模型使其能夠根據分子的結構預測 OP / RP。以此我們便能快 速地篩選大量的分子。此外,我們開發一套程式,可以通過添加官能團自動修飾 給定的分子。此程式至目前為止已經產生了超過一億六千萬個不同的分子。另外, 我們利用生成模型 — 變分自動編碼器來進行逆向設計以產生理想的分子。機 器學習技術帶來了一種新的方式來創建全新的電解質分子,無需進行複雜的計算 和昂貴的實驗。該方法可以為設計分子的第一步。
Recently, high-voltage Lithium-ion batteries (LIBs) has gained a lot of attraction because of the potential application in electric vehicles. However, conventional carbonate solvent-based electrolytes are less stable against high-voltage cathodes. In this thesis, machine learning to design new electrolytes for high-voltage Lithium- ion battery have been discussed. Different filters are set up for electrolyte molecular selection. Among these filters, the two important parameters for characterizing electrolytes are reduction potential (RP) and oxidation potential (OP). Machine learning models are trained to predict OP/RP from molecular structures. Additionally, we develop an algorithm which automatically modify given molecules by adding functional groups. So far about 160 million distinct molecules are generated. The machine learning model allows us to screen these generated molecules at a large scale and select ideal molecules. Furthermore, the generative model, variational autoencoder, was applied to carry out inverse design to generate ideal molecules. The machine learning techniques lead to a new way to create brand new functional molecules for electrolyte without doing complicated calculations and costly experiments. The methods could serve as a first step design towards further investigation into other necessary properties of electrolytes for high- performance LIBs.
中文摘要 I
Abstract II
Table of contents III
List of tables V
List of figures VI
Chapter 1 Introduction 1
Chapter 2 Literature Review 3
2.1 Lithium-ion battery electrolyte 3
2.2 Experiments, first-principles calculation, and machine learning 5
2.3 Molecular design and machine learning 6
Chapter 3 Calculation methods and Machine learning 11
3.1 Introduction to first-principles calculation 11
3.1.1 Introduction to Density functional theory (DFT) 11
3.1.2 The variational principle for ground state 12
3.1.3 Hartree fock theory 13
3.1.4 The Hohenberg-Kohn (HK) theorem 13
3.1.5 The Kohn-Shan equations 14
3.1.6 Basis set 15
3.1.7 Solvent Effect 18
3.2 The conception of machine learning 18
3.2.1 The introduction of Artificial Neural network 20
3.2.2 Convolution neural network 24
3.2.3 Recurrent neural network (RNN) 25
3.2.4 Probability theory for machine learning 28 Bayes Theorem 28 Entropy 28 Cross entropy 29 Kullback-Leibler Divergence 29
3.2.5 Autoencoder 29
3.2.6 Variational autoencoder 30
3.2.7 Data splitting 32
3.2.8 MAE and RMSE 33
3.3 Molecular representation 34
3.3.1 SYBYL 34
3.3.2 Extended Connectivity Fingerprint (ECFP) 35
3.3.3 Simplified molecular-input line-entry system (SMILES) 37
3.3.4 Coulombic matrix (CM) 40
3.4 The calculation of oxidative potential and reductive potential 40
3.5 Molecular structures and OP and RP 43
3.6 Molecular modification algorithm development 45
Chapter 4 Results and Discussions 47
4.1 Property predictors building and selection 47
4.2 The different between Gaussian03 and Q-chem 58
4.3 Molecular generation from functional group modification 62
4.4 Molecular generation from Variational autoencoder 69
Chapter 5 Conclusions 74
Reference 75
Appendix 78
1.Sun, Y., N. Liu, and Y. Cui, Promises and challenges of nanomaterials for lithium-based rechargeable batteries. Nature Energy, 2016. 1(7): p. 16071.
2.Qu, X., et al., The Electrolyte Genome project: A big data approach in battery materials discovery. Computational Materials Science, 2015. 103: p. 56-67.
3.An, S.J., et al., The state of understanding of the lithium-ion-battery graphite solid electrolyte interphase (SEI) and its relationship to formation cycling. Carbon, 2016. 105: p. 52-76.
4.Aurbach, D., et al., Design of electrolyte solutions for Li and Li-ion batteries: a review. Electrochimica Acta, 2004. 50(2-3): p. 247-254.
5.Shao, N., et al., Electrochemical windows of sulfone-based electrolytes for high-voltage Li-ion batteries. The Journal of Physical Chemistry B, 2011. 115(42): p. 12120-12125.
6.Zhang, Z., et al., Fluorinated electrolytes for 5 V lithium-ion battery chemistry. Energy & Environmental Science, 2013. 6(6).
7.Hu, L., Z. Zhang, and K. Amine, Fluorinated electrolytes for Li-ion battery: An FEC-based electrolyte for high voltage LiNi0. 5Mn1. 5O4/graphite couple. Electrochemistry communications, 2013. 35: p. 76-79.
8.Dalavi, S., et al., Effect of added LiBOB on high voltage (LiNi0. 5Mn1. 5O4) spinel cathodes. Electrochemical and Solid State Letters, 2011. 15(2): p. A28.
9.Xu, M., et al., Tris (pentafluorophenyl) phosphine: An electrolyte additive for high voltage Li-ion batteries. Electrochemistry communications, 2012. 18: p. 123-126.
10.Wang, R., C. Buhrmester, and J. Dahn, Calculations of oxidation potentials of redox shuttle additives for Li-ion cells. Journal of the Electrochemical Society, 2006. 153(2): p. A445-A449.
11.Borodin, O., W. Behl, and T.R. Jow, Oxidative stability and initial decomposition reactions of carbonate, sulfone, and alkyl phosphate-based electrolytes. The Journal of Physical Chemistry C, 2013. 117(17): p. 8661-8682.
12.Zhang, X., J.K. Pugh, and P.N. Ross, Computation of thermodynamic oxidation potentials of organic solvents using density functional theory. Journal of The Electrochemical Society, 2001. 148(5): p. E183-E188.
13.Maine, E. and E. Garnsey, Commercializing generic technology: The case of advanced materials ventures. Res. Policy, 2006. 35(3): p. 375-393.
14.Virshup, A.M., et al., Stochastic voyages into uncharted chemical space produce a representative library of all possible drug-like compounds. J Am Chem Soc, 2013. 135(19): p. 7296-303.
15.Samuel, A.L., Some studies in machine learning using the game of checkers. IBM Journal of research and development, 1959. 3(3): p. 210-229.
16.Rupp, M., et al., Fast and accurate modeling of molecular atomization energies with machine learning. Phys Rev Lett, 2012. 108(5): p. 058301.
17.Hansen, K., et al., Machine Learning Predictions of Molecular Properties: Accurate Many-Body Potentials and Nonlocality in Chemical Space. J Phys Chem Lett, 2015. 6(12): p. 2326-31.
18.Zunger, A., Inverse design in search of materials with target functionalities. Nature Reviews Chemistry, 2018. 2(4): p. 1-16.
19.Kuhn, C. and D.N. Beratan, Inverse Strategies for Molecular Design. 1996.
20.Broach, J. and J. Thorner, High-throughput screening for drug discovery. Nature, 1996. 384(6604): p. 14-16.
21.Hoelder, S., P.A. Clarke, and P. Workman, Discovery of small molecule cancer drugs: successes, challenges and opportunities. Molecular oncology, 2012. 6(2): p. 155-176.
22.Xiao, D., et al., Inverse Design and Synthesis of acac-Coumarin Anchors for Robust TiO2 Sensitization. J. Am. Chem. Soc., 2011. 133(23): p. 9014-9022.
23.S. A. Lopez, B.S.-L., J. de Goes Soares, A. Aspuru-Guzik, Design principles and top non-fullerene acceptor candidates for organic photovoltaics. Joule, 2017. 1(4): p. 857–870.
24.Gómez-Bombarelli, R., et al., Design of efficient molecular organic light-emitting diodes by a high-throughput virtual screening and experimental approach. Nature Materials, 2016. 15(10): p. 1120-1127.
25.Gómez-Bombarelli, R., et al., Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules. ACS Central Science, 2018. 4(2): p. 268-276.
26.Kang, S. and K. Cho, Conditional Molecular Design with Deep Generative Models. Journal of Chemical Information and Modeling, 2018. 59(1): p. 43-52.
27.Goodfellow, I.J., et al., Generative Adversarial Networks. 2014.
28.Kadurin, A., et al., druGAN: An Advanced Generative Adversarial Autoencoder Model for de Novo Generation of New Molecules with Desired Molecular Properties in Silico. Mol Pharm, 2017. 14(9): p. 3098-3104.
29.Olivecrona, M., et al., Molecular de-novo design through deep reinforcement learning. Journal of Cheminformatics, 2017. 9(1): p. 1-14.
30.Popova, M., O. Isayev, and A. Tropsha, Deep reinforcement learning for de novo drug design. Science Advances, 2018. 4(7): p. 7885.
31.Kingma, D.P. and M. Welling, Auto-Encoding Variational Bayes. 2013.
32.Sanchez-Lengeling, B. and A. Aspuru-Guzik, Inverse molecular design using machine learning: Generative models for matter engineering. Science, 2018. 361(6400): p. 360-365.
33.Rogers, D. and M. Hahn, Extended-Connectivity Fingerprints. 2010.
34.Weininger, D., SMILES, a chemical language and information system. 1. Introduction to methodology and encoding rules. J. Chem. Inf. Comput. Sci., 2002. 28(1): p. 31-36.
35.Weininger, D., SMILES, a chemical language and information system. 1. Introduction to methodology and encoding rules. Journal of chemical information and computer sciences, 1988. 28(1): p. 31-36.
36.Landrum, G., RDKit: Open-source cheminformatics.
37.Bjerrum, E.J. and B. Sattarov, Improving chemical autoencoder latent space and molecular de novo generation diversity with heteroencoders. Biomolecules, 2018. 8(4): p. 131.
38.Blum, L.C. and J.-L. Reymond, 970 Million Druglike Small Molecules for Virtual Screening in the Chemical Universe Database GDB-13. J. Am. Chem. Soc., 2009. 131(25): p. 8732-8733.
39.Gómez-Bombarelli, R., et al., Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules. 2018.
第一頁 上一頁 下一頁 最後一頁 top
系統版面圖檔 系統版面圖檔