【作者】 邓斌；

【导师】 顾昌鑫；

【作者基本信息】 复旦大学 ， 物理电子学， 2006， 博士

【摘要】 与计算机技术相结合的计算材料和材料设计是现代材料科学研究的重要方面，本论文应用计算机模拟方法对功能材料的结构和性能等若干问题进行了研究，具体包括以下三个内容：运用第一性原理的方法研究了锂离子电池正极材料的电子结构、导电性能以及它们之间的关系；运用分子动力学方法研究了嵌入轻质小原子或者分子层状石墨体系的热学行为；运用第一性原理的方法研究了单分子科学领域内人工控制化学键的形成及其特性。一．能源材料是材料科学的一个重要分支，也是目前材料科学领域的研究热点之一。本文运用局域密度近似框架内的基于密度泛函理论的第一性原理方法，研究了LiCoO₂及其被非钴金属元素掺杂后LiCo_0.92M_0.08O₂（M=Ni，Zn，Mg，Al，Cr，Mn，Fe，Cu）的电子结构，然后加大了非钴元素掺杂的量，运用相同的方法研究了LiCo_0.67M_0.33O₂（M=Mg，Mn，Ni）的电子结构。计算结果表明，与LiCoO₂的相比，LiCo_0.92M_0.08O₂（M=Ni，Zn，Mg，Cr，Mn，Fe）的态密度和能带结构分布发生了有利于电导率提升的变化；LiCo_0.92M_0.08O₂（M=Al，Cu）的电导率没有得到提升；如果加大非钴元素的掺杂量，LiCo_0.67Mg_0.33O₂相对于LiCoO₂的电导率没有提升，LiCo_0.67Mn_0.33O₂或者LiCo_0.67Ni_0.33O₂相对于LiCoO₂的电导率依旧得到了提升。主要的计算结果与实验事实相符合，因而从理论上证明了掺杂适当数量的非钴原子Ni，Zn，Mg，Cr，Mn或者Fe可以改善LiCoO₂的导电性能；而LiCo_0.92M_0.08O₂（M=Al，Cu）的电导率并没有提高；如果引入非钴元素Mn或者Ni的量至0.33，则LiCoO₂的电导率也可以得到改善。我们把改进的、结合了氧离子的电荷平衡和补偿机制首次用于以上锂离子电池电极材料计算结果的讨论和解释，对探索和开发新的具有优异性能的正极材料具有启发和理论指导意义。二．石墨材料也是一种优质的电极材料，对石墨材料的理论和实验研究一直是电极材料研究领域的热点。本文的第二部分主要内容是运用分子动力学的方法研究了嵌入轻质原子Li和H的层状石墨体系的热学行为，发现：（1）．Li和H在扩散过程中表现出不一样的特性，Li的行为更加复杂；（2）．分别计算了LI-GIC和H-GIC的导带带隙，发现Li-GIC的导带带隙没有发生明显变化，相反，H-GIC的导带带隙变宽了0.1eV，这意味着H-GIC的电导率下降。这些计算结果表明，作为锂离子二次电池电极材料，层状石墨有更大的优势和潜力，是比较好的储锂材料。这些与实验事实也相符合，同时解释了石墨等炭材料储锂有较大的不可逆Li容量的原因。另外，我们还运用分子动力学方法研究了其它轻质小分子如CO₂、H₂O、NH₃等在层状石墨中的热运动，发现这些轻质小分子在层状石墨中的热学行为与H原子在层状石墨中的热学行为非常相似；当层状石墨中的小分子的数目有明显增加时，石墨的晶体结构将会遭到不可逆转的破坏，这一计算结果是与实际情况相符合的。上述计算结果对寻求具有更加优异性能的石墨类电极材料具有指导和启发意义。三．人工控制原子或分子组装成具有复杂功能的材料和器件是人们孜孜以求的目标，其基础科学—单原子分子科学备受关注。受到使用STM人工控制形成化学键的实验事实，也就是借助STM控制CO分子与吸附在Ag（110）表面的过渡金属元素原子Fe和Cu形成化学键的实验事实的启发，在团簇结构模型下，运用第一性原理的方法验证了该实验结果并探求了受控形成化学键的机制和特征，理论计算结果与实验事实完全一致。我们把这一模型推广到其它元素原子，即Sc，V，Cr，Mn，Co，Ni，Zn，Zr，Ag，Au以及一些稀土金属元素原子等。发现所形成的化学键实际上可以被分成下面的三大类。第一类：对于Fe，Co，Ni，Cr等这些d轨道没有全占的原子，形成M（CO）／Ag（110）体系的时候，对应的是S₁结构；形成M（CO）₂／Ag（110）体系的时候，对应的是SS₂结构。对M（CO）／Ag（110）体系而言，没有发现稳定的S₂结构；第二类：对于Cu，Zn，Ag，Au等这些d轨道全被占的原子，吸附在Ag（110）面上的时候，无论是M（CO）／Ag（110）体系还是M（CO）₂／Ag（110）体系，结构都与Cu原子的情形一样，都是S₃类型；第三类：对应于那些具有f轨道的稀土金属原子，它们形成化学键依赖与f轨道上的电子，而由于它们的f轨道在空间的伸展方向极其复杂，所以这类元素原子与CO分子成键后表现出来的情况也就复杂一些，其中的机理还需要进一步的探求。但是通过以上计算，有助于了解和揭示Fe，Cu，Sc，V，Cr，Mn，Co，Ni，Zn，Zr，Ag，Au以及一些稀土金属元素原子在Ag（110）面上与CO分子成键的过程和机理，对人工控制、操纵单原子或者单分子具有理论指导意义。特别需要指出的是，对于在Ag（110）表面稀土金属元素原子和CO分子成键的情况比较复杂，需要在选择适当的方法的基础上对这些体系做更加深入的研究。更多还原

【Abstract】 Computational materials and materials design combined with computer techniques are important contents in materials science. Here three main aspects of the dissertation have been achieved by computer simulations and modeling using first principle theory or molecular dynamics methods: electronic structures, conductive properties and their relationship of the cathode material LiCoO₂ and its doped compounds used in Li-ion rechargeable batteries; thermal behaviors of the systems of layered graphite intercalated with H, Li and other small molecules; the controlled formations of chemical bonds and their characteristics.1. Energy material is an important branch of materials science, and is also a research hotspot. In this dissertation, the electronic structures of LiCoO₂ and its doped compounds LiCo_0.92M_0.08O₂ （M=Ni, Zn, Mg, Al, Cr, Mn, Fe, Cu） have been studied using first principle theory based on density-functional theory （DFT） in local density approximation （LDA）, based on the results of which, the electronic structures of LiCo_0.67M_0.33O₂ （M=Mg, Mn, Ni） have been also studied in the same methods. As the calculated results shown, compared with LiCoO₂, the band structures and the distributions of density of states （DOS） of the doped compounds have been changed for LiCo_0.92M_0.08O₂ （M=Ni, Zn, Mg, Cr, Mn, Fe）, which indicated that the electronic conductivities of these doped compounds have been improved, while the electronic conductivities of LiCo_0.92M_0.08O₂ （M=Al, Cu） have not been improved. The same method is used for LiCoo.67Mo.33O2 （M=Mg, Mn, Ni） with more non-Co atoms doped to LiCoO₂. It is found that the electronic conductivities of the gained LiCo_0.67Mn_0.33O₂ and LiCo_0.67Ni_0.33O₂ have been improved compared with LiCoO₂, while the electronic conductivities of LiCo_0.67Mg_0.33O₂ have not. These facts are in accord with the experimental results. Thus it has been theoretically testified that after doped with Ni, Zn, Mg, Cr, Mn, and Fe in a proper amount, the electronic conductivity of the cathode material LiCoO₂ can be improved, while doped with Al and Cu the electronic conductivity of LiCoO₂ will not be improved. And it is proved that the electronic conductivities of LiCo_0.67Mn_0.33O₂ and LiCo_0.67Ni_0.33O₂ are higher than that of LiCoO₂. The improved mechanism of charge balance and compensation with Oxygen ions taken into account is firstly adopted to explain for these changes of cathode materials used in Li-ion rechargeable batteries and the calculated results may afford enlightenment and guidance for exploring and developing new type cathode materials with high properties and performances.2. Graphite is also a kind of electrode material with high quality and attracts much research attention experimentally and theoretically. The thermal behaviors for the system of graphite intercalated with small atoms or molecules like H and Li have been studied using molecular dynamics （MD） method in this dissertation. And the conductive band gaps for them have been calculated using the extended Huckel method. It is found that the rates of diffusion of both Li and H atoms increase with the increase of the simulation temperature: from 50K to 200K, and the specific diffusive behaviors and rates for Li and H are different according to their trajectories. The thermal behaviors of Li are more complex than that of H. The conductive gap is broadened by about 0.1eV for H-GIC, but it remains unchanged for Li-GIC, which indicates that the addition of Li does not influence the conduction characteristic of graphite while that of H does. These results tell the fact that Li-GIC is proposed to be a favorable material for the electrode, which is consistent with the experimental facts. At the same time, these calculated results have explained why there exists a fairly irreversible capacity of Li for graphite and other carbonaceous materials. Besides, systems of layered graphite intercalated with CO₂, H₂O and NH₃ etc. have also been studied by MD method, the results of which show that the thermal behaviors of these small molecules are similar to those of H and the lattice structures of layered graphite will be destroyed irreversibly if the quantity of these small molecules increases. This is coincident with facts.3. It is a great goal to control atoms or molecules to build up functional structures and devices. The third part of our work is following an experimental fact of the controlled formations of chemical bonds between CO and Fe/Cu adsorbed on Ag （110） using STM. The experimental fact is testified and characterized by the first principle method. And we have extended this method to other series of metal atoms adsorbed on Ag （110）, such as Sc, V, Cr, Mn, Co, Ni, Zn, Zr, Ag, Au and some lanthanons. It is found that according to their electron number in the outermost layer orbital, the calculated results can be classified into three kinds, （i）. With an unfully occupied d orbital, such as Fe, Co, Ni and C, structures for M（CO）/Ag（110） are corresponding to S₁ and structures for M（CO）₂ /Ag（110） are corresponding to SS₂, no stable S₂ structures are found for m（CO）/Ag（110）; （ii）. With a fully occupied d orbital, such as Cu, Zn, Ag and Au, the structures for both m（CO） / Ag （110） and M（CO）₂/Ag（110） are corresponding to S₃; （iii）. With f orbital, such as those lanthanons, the structures for M（CO）/Ag（110） and M（CO）₂/Ag（110） are very complex because of the complicated space extending directions for f and more and deeper investigations into themechanisms for chemical bond formations between lanthanons and CO should be undertaken. The calculated results also show that the chemical bonds between these metal atoms and CO appear different characteristics, for example, different bond angles, different bond energy and different torsions because of their different space extending directions of outmost layer orbital. The frontier orbital theory （FOT） has been adopted to explain these results. All of these results help us to understand the procedures for controlled chemical bond formations and may provide enlightenments in controlling and manipulating single atoms or molecules.更多还原