Multiscale Theoretical Study On Graphene Growth And Application

Since the last century,with the establishment of quantum mechanics,people’s understanding of the microscopic world has become deeper and more intuitive.The Schrodinger equation provides us with the basic method for calculating the behavior of microscopic particles.The Dirac equation further considers the possible relativistic effect of the microscopic world.These two equations can in principle explain all phenomena in the electronic world.However,the complex multi-electron coupling makes the analytical solution of the equation difficult to obtain.In order to solve the problem of material calculations,people introduced adiabatic approximation and single-electron approximation above the basic equations of quantum mechanics,making the first-principles calculation a powerful tool for exploring the nature of matter.In recent decades,tremendous progress has been made in computational chemistry,computational materials science,and computer science.In theoretical chemistry,density functional theory(DFT),Kohn-Sham equations,and exchange-correlation energy approximating methods have made it possible for researchers to accurately calculate the properties of materials such as mechanical property,heat,light emitting and adsorption,electronic structure,and acoustics.In the field of computer science,the hardware and software of computers have constantly been updated and replaced,greatly increasing the computing power and making simulation of the nature of complex systems possible.The development of these fields enables us not only to explore existing materials or to predict the physicochemical properties of unsynthesized materials,but also allow us to conduct a first-principle theoretical exploration of the microscopic kinetic mechanism of the material synthesis process.On the other hand,since the beginning of this century,with the successful stripping of monolayer graphene,two-dimensional materials have become a research hotspot in scientific research.Due to their unique structural characteristics,two-dimensional materials have relatively smaller charge shielding and a larger specific surface area than the bulk,making them often have unique electronic and catalytic properties,opening up another path for material design and application.The star material graphene has a very strong mechanical stability,high carrier mobility,and high degree of transparency.These unique properties make it widely used in electronic devices and chemical catalysis applications.The early graphene was obtained by mechanical stripping method,its crystal structure is complete,and its electronic properties keep almost perfect.However,this method cannot be used to produce a large area of graphene,and the stripping efficiency is also much low,which has prompted people to find other large-scale production methods that can be applied to industrial production.The chemical vapor phase synthesis of graphene has been developed.It can grow large scale graphene on a metal substrate under high temperature.The synthesis process of graphene is generally performed in a quartz transistor,with a furnace outside the tube,controlling the growth temperature,and a metal film as growth substrate in the tube.Methane and hydrogen are introduced into the tube.Methane molecules adsorb and dissociate on the metal surface in a high temperature environment,and carbon aggregates to form graphene.The commonly used metal substrate material is a thin foil of copper or nickel,and the growth temperature is generally 1300K.After the graphene growth is completed,the graphene can be transferred onto a non-metal substrate such as silicon dioxide by etching the metal substrate.This method can grow large graphene with grain boundaries,but it is difficult to control large single crystal formation.In addition,the grown graphene contains many defects,and the electron mobility such as carrier mobility is not as good as that of the graphene obtained by the mechanical lift-off method.In this work,the chemical vapor deposition(CVD)mechanism of graphene on copper surface,the fitting of copper-carbon interaction potential and the application of graphene as anode material are studied.The main content is divided into the following sections:The first chapter briefly introduces the main theoretical calculation methods of this paper,including density functional theory(DFT),transition state calculation method,Monte Carlo method and neural network method.In chapter two,the concentrations of adsorbed species in the growth of graphene on the copper surface were studied.The high temperature environment during the growth process and the difficulty in detecting hydrogen atoms make it impossible to experimentally obtain the concentration of hydrocarbon species on the copper surface.To understand the growth mechanism of graphene,it is necessary to answer questions like:what hydrocarbon species is mainly on the copper surface,how high the total carbon concentration is,which species has higher concentrations and higher contributions,what species have lower contribution and lower hydrogen content,how high the concentration of H adatom is and what effect the concentration has on the edge configuration of the graphene.In this chapter,we used first-principles calculations conbined with Monte Carlo simulations to answer these questions.In the third chapter,the attachment and detachment of hydrocarbon species on the edge of graphene were studied.Based on the second chapter,we specifically discussed the process of attachment and detachment of hydrocarbon species on graphene edges.We found that it is not enough to consider only the attachment process when considering the growth of graphene.The concentration of hydrocarbon species on copper surface is very low.After one hydrocarbon species attached onto graphene edge,averagely other hydrocarbon species will not attach at the same position for a long time.Therefore,the stability after attachment will greatly affect the graphene growth process.We further discovered the role of hydrogen in stablizing graphene edges.In Chapter 4,we try to fit the first-principles potential surface of a copper-carbon system with neural networks.We find that the neural network potential obtained by fitting can describe structures similar to the training set well,but the energy description is often poor for structures with large deviations from the training set.The neural network potential may have some non-physical valleys on potential energy surface,making molecular dynamics simulation using this neural network potential easy to fall into a strange configuration.In order to solve this problem,we introduced the grand canonical Monte Carlo method(GCMC)and used the neural network potential to run GCMC simulation for generating structures in the valleys.Then we performed first-principles calculations and rejoin the training set to optimize the neural network potential.This cycling process can continuously optimize the neural network potential to obtain a robust force field.In Chapter 5,we studied the application of graphene in ion battery electrodes.The generally used negative electrode material for commercial lithium ion batteries is graphite,and the two-dimensional counterpart of graphite is graphene.People discussed the possibility of two-dimensional materials as anode materials for ion batteries and found that graphene has weaker interaction with lithium than graphite.The binding energy of lithium atoms and graphene is less than the binding energy of lithium metal,which means that graphene is not a suitable Li ion anode material.We studied the influencing factors of the interaction between graphene and lithium atoms,and found that the interaction between extended graphene and lithium atoms is greater,and discussed the possibility of B2S materials,which is similar to the extended graphene,as anode materials for ion batteries.