First Principles Theory of Complex Compounds

9802076皮克特这个奖项是为了支持对当前感兴趣的各种材料的电子结构和性质的理论研究。最近发现的产生巨大活性的材料的例子包括：化合物Sr14-xCaxCu24O41，它有Cu-O链和梯子，但没有层，但在特定的压力范围内x的范围很窄；自旋1/2层状钒酸盐CaVnO2n+1，它表现出自旋-Peierls行为(n=3)或自旋-带隙行为(n=2和n=4)；Fe-Si系统FenSim，其中n=1，m=2和n=2，m=1成员是相当标准的金属，而n=m=1成员是非常不寻常的‘Kondo绝缘体’，能隙为60meV；以及填充的Skutterudite系统，如CE1-xLaxCo4Sb12，在热管理应用中显示出出色的热电行为。当然，高温超导铜酸盐由于表观d波序参数，复杂的磁通晶格行为，部分是由于它们的短相干长度，以及在相当大程度上由于微观配对机制的持续难以捉摸，继续引起人们的兴趣。所谓的“巨磁电阻”锰氧化物材料引起了广泛的兴趣，这种材料表现出大量的电荷、自旋和结构有序，以及不寻常的场驱动重入行为。在一些情况下，新材料的发展是理论和实验之间密切相互作用的结果，不仅对性质的解释是理论的作用，而且还直接表明了什么有望成为有趣的东西，在哪里寻找它，等等。这些新材料的共同特征是：(1)它们是多组分的，性质强烈地依赖于化学计量；(Ii)它们的晶体结构具有低对称性，表明不寻常的化学键模式；(Iii)它们的性质是非常规的，往往随着材料参数的微小变化(掺杂水平、温度、压力，甚至离子大小有时是至关重要的)而变化很大；以及(Iv)它们涉及巡回载流子和状态的相互作用，在某种意义上是局域化的。正是这最后一个特征，以及其他复杂的情况，提供了这种特殊的行为。对于试图描述系统的理论家来说，单胞的复杂性通常不会造成根本的困难。复杂的金属间化合物、四元半导体和半导体超晶格通常可以用唯象方法和密度泛函理论中的局域密度近似相结合的方法来处理。同样，简单晶格中的局域原子态(即，盐岩结构中的NiO)可能会导致非常有趣的关联电子行为，但它们根本不具备更复杂的晶格中存在的自由度。在许多这样的系统模型中，将继续需要哈密顿量来研究它们的热力学和光谱性质。然而，从第一原理出发，迫切需要解决伴随着在组合巡回/局域载流子系统中出现的相关电子行为的单位电池的复杂性。这里提出的工作将把各种方法--参数化紧束缚哈密顿量、精确的局部自旋密度计算、“超越LDA”的创新和精确的对角化方法--应用于这样的复杂化合物，期望同时增加对观察到的行为的理解，并扩展第一原理方法的能力。这项理论研究将研究当前感兴趣的复杂材料，这些材料表现出定域的电子之间的竞争，即固定在原子或分子上的电子和非定域的电子之间的竞争。这种竞争，连同其他材料性质，如温度、压力和浓度，可以产生非常罕见的L效应，这是根本感兴趣的，可能导致技术应用。***

9802076 Pickett This award is in support of theoretical research on the electronic structure and properties of various materials of current interest. Examples of recently discovered materials that have generated great activity include: the compund Sr14- xCaxCu24O41, which has Cu-O chains and ladders but no layers, yet superconducts for a narrow range of x in a specific range of pressure; the class of spin-1/2 layered vanadates CaVnO2n+1, which show spin-Peierls behavior (n=3) or spin-gap behavior with a wide range of gaps (n=2 and n=4); the Fe-Si system FenSim, for which the n=1, m=2 and n=2, m=1 members are rather standard metals whereas the n=m=1 member is a very unusual `Kondo insulator' with an energy gap of 60meV; and the filled skutterudite systems such as Ce1-xLaxCo4Sb12 which show superb thermoelectric behavior for heat management applications. The high temperature superconducting cuprates, of course, continue to cause fascination due to apparent d-wave order parameter, the complex flux lattice behavior arising in part from their short coherence length, and in no small part due to the continuing elusiveness of the microscopic pairing mechanism. A widespread fascination has arisen from the so-called `colossal magnetoresistance' manganite materials which show numerous charge, spin, and structural orderings, and unusual field-driven reentrant behavior. In several cases the development of novel materials has resulted from close interplay between theory and experiment, with not only interpretation of properties being the role of theory but also the direct indication of what promises to be interesting, where to look for it, and so on. Features that are common to many of these novel materials are: (1) they are multicomponent, with properties being strongly dependent on stoichiometry; (ii) their crystal structures have low symmetry, indicative of unusual chemical bonding patterns; (iii) their properties are unconventional, and often vary considerably wit h small changes in material parameters (doping level, temperature, pressure, even ionic size is sometimes crucial); and (iv) they involve the interplay of itinerant carriers and states which are in some sense localized. It is this last feature, in concert with the other complications, that provides the peculiar behavior. Complexity of the unit cell alone usually does not cause fundamental difficulty for theorists trying to descibe the systems. Complex intermetallics, quaternary semiconductors, and semiconductor superlattices can usually be handled well by a combination of phenomenological methods and the local density approximation within density functional theory. Likewise, localized atomic states in simple cell lattices (viz. NiO in the rocksalt structure) may lead to very interesting correlated electron behavior, but they simply do not possess the degrees of freedom that are present in more complex lattices. In many of these systems model Hamiltonians will continue to be required to address their thermodynamic and spectral properties. There is however serious need to address the complexities of the unit cell concomitantly with the correlated electron behavior that arises in combined itinerant/loccalized carrier systems from first principles. The work proposed here will apply a combination of methods - parametrized tight-binding Hamiltonians, precise local spin density calculations, `beyond LDA' innovations, and exact diagonalization methods - to such complex compounds with the expectation of simultaneously increasing understanding of the observed behavior and extending the capabilities of first principles methods. %%% This theoretical research will investigate complex materials of current interest which display a competition between electrons which are localized, i.e., fixed to atoms or molecules, and ones which are not. This competition, along with other materials properties such as temperature, pressure and concentration, can produce highly unusua l effects which are of fundamental interest and which may lead to technological applications. ***