First-Principles Studies Of Environmental Effects In Electron Transport And Surface Reaction

With the development and improvement of density functional theory and related algorithms,atomic-scale computational simulation has become one of the important means to study the physical and chemical properties of materials.The environment around an electron or atom in a material often directly determines its final properties.For environmental effects,a simple harmonic approximation is usually used.For example,the motion of the lattice around the electron is often treated by simple harmonic motion,or directly simulates the real motion of the atom,such as molecular dynamics.This paper mainly applies the first-principles to study the effects of environmental on electron transport and high temperature surface reactions.Using a simple harmonic approximation,the vibration of the lattice can be understood as a phonon.The transport properties of the material at room temperature mainly come from the interaction of electrons and phonons.A comprehensive consideration of the scattering of all phonon modes is a prerequisite for its calculation.The correlation properties of phonon determined can be calculated by means of density functional perturbation theory combined with Wannier-based interpolation technique.In addition to considering the simple harmonic approximation,we also studied the influence of the environment on the high temperature surface reaction by using the first-principles molecular dynamics technique.By taking the high temperature dissociation of methane as an example,we discussed the relationship between the change of local environmental and the free energy of the reaction.This article mainly includes the following contents:The first and second chapters mainly describe some calculation methods used in this paper.Firstly,we introduced the basic concepts of density functional theory and some problems that should be paid attention to.Then the density functional perturbation theory(DFPT)based on density functional theory is further introduced.The phonon-related properties can be obtained by DFPT.With the Wannier-based interpolation technique,the Brillouin region can be finely and quickly sampled.It is possible to calculate the convergent transport properties.After the potential energy surface related information is obtained by the first-principle method,some thermodynamic properties such as free energy of the system can be quickly obtained by using some enhanced sampling algorithms such as MetaDynamics.In the third chapter,we study the effects of environment on electron transport under simple harmonic approximation.We calculate the phonon-related properties of the single layer Ca2N,an electene material obtained after the theoretical prediction.We found the electron-phonon coupling matrix elements of single layer Ca2N is very small compared to other two-dimensional materials due to its unique two-dimensional electron gas in free space(2DEG-FS)characteristics,and its carrier mobility(189 cm2V-1s-1)is also higher than the other common metals.By comparison with graphene with the same electron concentration,it is found that the low group velocity is the main reason for the difference of mobility between the two materials.And we predict that the mobility of single-layer Ca2N can be increased about one order of magnitude by the hole doping.The single layer of Ca2N may become a BCS superconductor under 4.7 K.In the fourth chapter,we discuss the effects of the environment on high temperature surface reactions in addition to the harmonic approximation.We discuss the feasibility of the minimum energy path(MEP)in predicting the reaction at high temperature by using the first-principles molecular dynamics to study the dissociation of methane at high temperature.By comparing the first-principles molecular dynamics with the minimum energy path method,we found that for the species with strong steric hindrance effects such as CH2 and CH3,the results of the free energy obtained by the addition of harmonic vibration correction and the results of direct molecular dynamics are very close,and even at high temperatures the MEP model can still provide similar results.However,for species with weak steric hindrance,such as CH not only at 1300K,even at 300K MEP can not give reasonable results and give a large deviation.By analyzing the changes in the coordination number around the species,we conclude a linear correlation between the change in the local environment and the energy barrier.Combined with the results of molecular dynamics,we found that CH concentration decreased during graphene growth and C2 became the main feeding species.In the fifth chapter,we try to develop a new enhanced sampling algorithm based on neural network and simulated annealing algorithm.The neural network has a strong numerical fitting ability,and does not need many adjusted parameters.When the number of reaction coordinates increases,the amount of calculation does not increase exponentially.At the same time,the simulated annealing method can quickly evolve to the state of the system we need by warming up,and finally slowly cool down to the simulated temperature.By combining the above two methods,our test results show that this new method can achieve fast and efficient sampling of a simple system,and at the same time,after the end of the simulation,the free energy surface can be obtained without excessive subsequent processing.