基于机器学习力场的固态电解质界面相的热传导研究.docxVIP

PAGE

PAGE2

摘要

固态电解质界面相是指金属负极与电解液在第一次充放电时形成的一层同时具备一定导电能力与防止电池内部反应致使电池分解的膜状物质，其性质特点是科研工作者研究的热点，而在其中固态电解质界面相热导率是提升锂离子电池综合性能的重要一环，也因此成了在传统锂离子电池能量密度受限的情况下电池性能提高的重要突破口。在这项工作中，我们将采用机器学习力场使用MTP势能从分子层面对固态电解质界面相的热导率变化进行模拟分析，通过对于氟化锂、氧化锂和硫化锂这三种固态电解质界面相常见组分的细致分析，发现声子的变化对热导率的大小有着重要影响，其具体表现为随着温度上升其热导率在不断降低。提高声子寿命、降低声子散射率、增加低频去声子态密度、降低高频区声子态密度，适当增加固态电解质界面相厚度均有利于提高热导率。这为日后提高固态电解质界面相热导率研究提供理论参考。

关键词：固态电解质界面相；热导率；声子；机器学习

PAGE

PAGE3

Abstract

Solidelectrolyteinterfacephasereferstoafilmlikesubstanceformedbythemetalnegativeelectrodeandelectrolyteduringthefirstchargeanddischarge,whichsimultaneouslyhascertainconductivityandpreventsinternalreactionsinthebatteryfromcausingbatterydecomposition.Itspropertiesandcharacteristicsareahotresearchtopicforresearchers,andthethermalconductivityofthesolidelectrolyteinterfacephaseisanimportantlinkinimprovingthecomprehensiveperformanceoflithium-ionbatteries,Therefore,ithasbecomeanimportantbreakthroughforimprovingbatteryperformanceunderthelimitedenergydensityoftraditionallithium-ionbatteries.Inthiswork,wewillusemachinelearningforcefieldsandMTPpotentialenergytosimulateandanalyzethethermalconductivitychangesofthesolidelectrolyteinterfacephasefromthemolecularlevel.Throughdetailedanalysisofthecommoncomponentsoflithiumfluoride,lithiumoxide,andlithiumsulfideinthethreesolidelectrolyteinterfacephases,itisfoundthatthechangesinphononshaveasignificantimpactonthemagnitudeofthermalconductivity,whichismanifestedbythecontinuousdecreaseofthermalconductivityastemperatureincreases.Improvingphononlifetime,reducingphononscatteringrate,increasinglow-frequencyphonondensityofstates,reducinghigh-frequencyphonondensityofstates,andappropriatelyincreasingthethicknessofthesolidelectrolyteinterfacephaseareallbeneficialforimprovingthermalconductivity.Thispr