关联材料结构优化新算法—LDA + eDMFT方法的基本理论探讨

Basic Theory Discussion of LDA + eDMFT: A New Algorithm for Structural Optimization of Correlated Materials

摘要: 理论上对晶体结构进行预测是凝聚态物理和材料科学中最基本的挑战之一。目前对弱关联材料的基于密度泛函理论(DFT)的第一性原理晶体结构预测已经变得足够准确，可以和实验结果进行比较，并且结合动力学平均场理论(DMFT)的DFT + DMFT方法使得对给定晶体结构的关联材料电子性质的预测成为可能。然而目前的LDA + DMFT方法，由于它的交换关联函数的半局部近似，使得它无法对关联电子材料的结构进行优化，如对莫特绝缘材料和关联金属的结构预测往往失败。因此，对这些材料的晶体结构预测必须发展新的有效算法。本文主要介绍目前国际上在统一的Luttinger-Ward泛函框架下可以对关联电子材料进行晶体结构优化的新算法—DFT-eDFMT方法的主要原理及基本应用。

Abstract: Theoretical prediction of crystal structure is one of the most fundamental challenges in condensed matter physics and materials science. At present, the first-principles crystal structure prediction based on density functional theory (DFT) for weakly correlated materials has become accurate enough to be compared with experimental results, and DFT + DMFT method combined with kinetic mean field theory (DMFT) makes it possible to predict the electronic properties of related materials for a given crystal structure. However, the current LDA + DMFT method is unable to optimize the structure of the associated electronic materials due to its semi-local approximation of the exchange correlation function. It has been demonstrated that the prediction of the structure of Mott insulation materials and associated metals often fails. Therefore, it is necessary to develop new efficient algorithms for crystal structure prediction of these materials. This paper mainly introduces the principle and application of DFT-eDMFT, a new algorithm for crystal structure optimization of related electronic materials under the unified Luttinger-Ward functional framework.