【作者】 汪丽莉；

【导师】 石兢； 刘惠军；

【作者基本信息】 武汉大学 ， 凝聚态物理， 2009， 博士

【摘要】 本文采用基于密度泛函理论的第一性原理计算方法,研究了几种块体和低维热电材料的电子结构以及输运特性。从理论上预测了对这几种材料如何掺杂可以最大限度地优化其热电性能,以及体系低维化后对热电性能的影响,对相关实验研究有一定的指导意义。我们详细研究了Half-Heusler体系MNiSn和MCoSb系列化合物(M=Ti、Zr、Hf)的晶体结构和输运特性,讨论了这两种体系的电子结构和输运性质随p型及n型掺杂的变化关系,从理论上指出掺杂什么元素和对应的掺杂浓度可能获得较大的功率因子,可望大大减少实验研究所需的各种资源。我们还利用密度泛函微扰理论研究了TiNiSn和TiCoSb两种化合物的声子谱并讨论了其晶格热导率随晶体尺寸的变化关系。我们还计算了Bi2Te3化合物的电子结构和输运性质,结果表明对该体系n型和p型掺杂的最佳浓度大致在化学势不超过0.6 eV的位置,并且高温情况下p型掺杂将得到高于n型掺杂的功率因子。另外,我们还设计了一系列Bi和Sb纳米管并计算了它们的电子结构和输运性质。结果表明,Bi纳米管的能量较为稳定,基本不随其螺旋角和直径的大小而发生变化；直径相同但螺旋角明显不同的Bi(3,3)、Bi(4,2)、Bi(5,0)纳米管都是半导体；螺旋角固定时Bi纳米管的带隙值均随管径的增大而减小。Sb纳米管的总能会随着螺旋角和管径的大小而发生变化；螺旋角为0°的Sb纳米管表现出半导体特征,但螺旋角为30°时变成金属；半导体特性的Sb(n,0)纳米管直径增大后会转变成为金属,而所有Sb (n, n)纳米管都是金属性的。我们还讨论了Bi(10,0)和Sb(6,0)纳米管的功率因子随化学势和温度的变化关系。计算表明Bi(10,0)纳米管在p型掺杂的情况下表现出较好的电输运性能,而Sb(6,0)纳米管则在n型掺杂的情况下表现出较好的输运性能。在此基础上,如果可以通过调节两个纳米管之间的功率因子和热导率分别得到p型和n型低维纳米热电材料,将有望实现纳米级的热电器件。对碳纳米管掺杂V族(As、Sb、Bi)原子的计算表明,无论其螺旋角是0°,19°还是30°,直径为4 A的三种典型碳纳米管掺杂后均变成金属。对于一系列(n,0)碳纳米管的掺杂在管径逐步增大、掺杂浓度逐步降低的过程中,反应所需要的能量是逐步减小的。(8,0)和(10,0)碳纳米管在As原子掺杂时仍然表现出半导体特性,并且带隙随着掺杂原子的引入变窄。对其电输运性能的研究结果表明,它们在200K-400K范围内具有很高的功率因子。如果可以有效地降低其热导率,碳纳米管有望成为新一代高性能的常温热电材料。更多还原

【Abstract】 The aim of this thesis is to investigate the electronic, transport and other physical properties of some bulk and low-dimensional thermoelectric materials from first-principles calculations. In combination with the Boltzmann theory, we are able to predict a large number of potential candidates which may exhibit higher thermoelectric performance. Our theoretical work serves as a guide to the experimental study searching for high-performance thermoelectric materials.We first focus on the half-Heusler compounds TiNiSn and TiCoSb. At room and elevated temperature, it is found that the p-type doping may be more favorable than the n-type doping for both compounds. The optimized doping concentration corresponds to a chemical potential of -0.34 eV and -0.61 eV for the TiNiSn and TiCoSb, respectively. By using the density functional perturbation theory (DFPT), the phonon dispersion relations of TiNiSn and TiCoSb are obtained and the corresponding thermal conductivity are examined with respect to the size of these two compounds.The electronic structure and transport properties of bulk Bi2Te3 are then investigated. Within the rigid-band picture, the optimal doping atom and the corresponding doping concentration are obtained. We find that the power factor of p-type doped Bi2Te3 is higher than that of n-type at high temperature. On the other hand, we construct a series of Bi and Sb nanotube structures. It is found that most of the Bi-tubes are semiconducting and the energy gap decreases when the diameters of these tubes are increased. The power factor of the Bi(10,0) and Sb(6,0) tubes will be enhanced for the p-and n-type doping, respectively. If the power factor and thermal conductivity between them can be adjusted experimentally, one can realize thermoelectric devices with nanosize.We also discuss the possibility to using carbon nanotubes as thermoelectric materials. When doped with As/Sb/Bi elements, there are obvious changes of the band structures of three kinds of 4 A carbon nanotubes. Moreover, the band gap of (8,0) and (10,0) carbon nanotubes decrease when doped with As atom. The relatively higher power factor of these two nanotubes suggests that carbon nanotubes may be a new kind of high performance thermoelectric materials if one can effectively reduce their thermal conductivity.更多还原