請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/84798
標題: | 探討鋰離子在雙(三氟甲基磺醯)氨基/聚乙二醇/三氧化鉬複合材料的傳輸機制: 機器學習分子勢能參數化與分子動力學研究 Parameterization of Machine Learning Hybrid Potential for Investigating Lithium Ions Transportation in LiTFSI/PEO/MoO3 Composites |
作者: | Chia-Pin Mao 毛家斌 |
指導教授: | 張建成(Chien-Cheng Chang) 張建成(Chien-Cheng Chang | mechang@iam.ntu.edu.tw | ), |
關鍵字: | 第一原理計算,分子動力學,鋰離子電池,固態電解質,聚合物電解質, ab initio calculation,molecular dynamics,lithium ion battery,solid-state electrolyte,polymer electrolyte, |
出版年 : | 2022 |
學位: | 碩士 |
摘要: | 鋰離子電池 (Lithium ion battery, LIB) 因為高能量密度及市場使用普及等優點,是現今最有前景的儲能設備之一,以固態電解質取代目前使用液態電解質能減少安全性問題,且有機會提高電池之能量密度。固態電解質可分為無機電解質與聚合物電解質,兩者各有優缺點。而為了改善兩者的缺點,科學家投入了大量的研究。其中,一些研究提出將帶有氧空缺的無機材料加入聚合物電解質中可以提升聚合物電解質的離子導電性,並提出氧空缺影響鋰離子傳輸的機制。 中研院的朱治偉博士團隊在實驗中將帶有氧空缺的無機材料MoO3加入聚合物電解質PEO/LiTFSI中,發現確實可以提升PEO/LiTFSI的鋰離子傳導性,因此我們想以模擬的方式探討PEO/LiTFSI/MoO3複合固態電解質中鋰離子的傳輸能力及傳輸機制,驗證其傳輸機制是否與先前研究提出的論述一致。 由於本研究的尺度較大,較適合利用分子動力學模擬。然而對於本研究所探討的PEO/LiTFSI/MoO3複合固態電解質,目前並沒有可以直接使用的勢能函數。因此吾人設計一套結合多種函數型態的混合勢能,來描述固態電解質中三種主成分之間的作用力。其中,描述PEO/LiTFSI與MoO3之間的作用勢能函數無法由先前的研究取得,因此吾人將以Lennard-Jones勢能作為勢能函數模型,並利用機器學習訓練將此勢能函數參數化。待得到完整的混合勢能之後,對PEO/LiTFSI/MoO3系統進行分子動力學模擬,求得系統在不同MoO3氧空缺濃度的情況下鋰離子的自擴散係數,分析MoO3之氧空缺影響鋰離子傳輸的機制。由於加入無機物後,聚合物電解質之機械性質可能提升,因此吾人也對複合固態電解質進行拉伸試驗模擬,並求得其楊氏模數。 經過模擬分析後,我們發現加入PEO/LiTFSI的MoO3奈米帶表面氧空缺濃度為10 %時,其鋰離子擴散係數為最高的7.73×10-8 cm2/s,提升約10%。而氧空缺對鋰離子傳輸的加速機制,我們認為是因為奈米帶表面氧空缺吸引TFSI-陰離子,驅使離奈米帶較遠處之陰離子往奈米帶方向移動分布,使得該處的電解質中鋰離子與TFSI-離子質心間的距離能變得更遠,連帶增加鋰離子的擴散速率。我們也發現加入MoO3奈米帶能提升複合固態電解質的機械性質,其楊氏模數由0.76 GPa變為1.30 GPa,增加約69.35%。鋰離子擴散係數與機械性質的提升能讓固態電解質更有機會取代液態電解質,作為鋰離子電池中傳輸離子的材料。 Lithium ion batterys (LIBs) due to its high energy density and markpopularity, are one of the best energy storage devices. By using solid electrolytes, we can avoid the hazzard of liquid electrolytes and have the opportunity to improve the energy density of the battery. Solid electrolytes can be divided into inorganic electrolytes and polymer electrolytes, each of which has advantages and disadvantages. To improve the shortcomings of both, scientists have invested a lot of research. Among them, some studies have proposed that adding inorganic materials with oxygen vacancies into polymer electrolytes can improve the ionic conductivity of polymer electrolytes, and proposed the mechanism by which oxygen vacancies affect lithium ion transport. The research team of Dr. Chu from from the Academia Sinica introduced the inorganic material MoO3 with oxygen vacancies into the polymer electrolyte PEO/LiTFSI in the experiment, and found that the lithium ion conductivity of PEO/LiTFSI can indeed be improved. Therefore, we wanted to investigate the transport ability and transport mechanism of lithium ions in PEO/LiTFSI/MoO3 composite solid electrolytes by means of simulation, and verify whether the transport mechanism is consistent with the discussion proposed in previous studies. The scale of the model in this study is large, which is more suitable for the use of molecular dynamics simulation. However, for the PEO/LiTFSI/MoO3 composite solid electrolyte discussed in this study, there is currently no potential energy that can be directly used. Therefore, we design a mixed potential energies combining various potential functions to describe the forces between the three components in the solid electrolyte. Among them, the potential energy function describing the interaction between PEO/LiTFSI and MoO3 cannot be obtained from previous studies, so we will use the Lennard-Jones potential energy as the potential function model and use machine learning to train this potential function parameterization. After obtaining the complete potential energy, molecular dynamics simulation of the PEO/LiTFSI/MoO3 system was performed to obtain the self-diffusion coefficient of lithium ions under different MoO3 oxygen vacancy concentrations, and to analyze the influence of MoO3 oxygen vacancies on lithium ion transport mechanism. Since the mechanical properties of the polymer electrolyte may be improved after adding inorganic substances, we also simulated the tensile test of the composite solid electrolyte and obtained its Young’s modulus. After simulation and analysis, we found that when the oxygen vacancy concentration on the surface of MoO3 nanobelts introduced into PEO/LiTFSI was 10 %, the lithium ion diffusion coefficient was the highest at 7.73×10-8 cm2/s, which was increased by about 10%. As for the acceleration of lithium ion transport by oxygen vacancies, we believe that the oxygen vacancies on the nanobelt surface attract TFSI- anions, driving the anions farther from the nanobelt to move toward the nanobelt, so that the lithium ions and TFSI in the electrolyte there are moved towards the nanobelt. The distance between ions becomes farther, which increases the diffusion rate of lithium ions. We also found that introducing MoO3 nanobelt into PEO/LiTFSI can improve its mechanical properties (its Young's modulus changes from 0.76 GPa to 1.30 GPa, an increase of about 69.35%). The improvement of lithium ion diffusion coefficient and mechanical properties can give solid electrolytes more opportunities to replace liquid electrolytes as ion transport materials in lithium ion batteries. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/84798 |
DOI: | 10.6342/NTU202203012 |
全文授權: | 同意授權(限校園內公開) |
電子全文公開日期: | 2022-09-02 |
顯示於系所單位: | 應用力學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
U0001-3108202212422000.pdf 授權僅限NTU校內IP使用(校園外請利用VPN校外連線服務) | 2.99 MB | Adobe PDF | 檢視/開啟 |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。