强关联f电子材料磁激发谱的第一原理计算方法研究

Strongly correlated f-electron materials usually exhibit complex ground state electronic structures and exotic low-energy collective excitations. Since the last decades, they have attracted a lot of attentions. It is well known that the inelastic neutron scattering could be used to detect the magnetic excitation spectra of strongly correlated f-electron materials, and some remarkable results have been obtained. However, theoretical descriptions for the magnetic excitation spectra of strongly correlated f-electron materials are not well established so far. This proposal will concentrate on the theoretical methods for magnetic excitations and spin dynamics. The major tasks are as follows: (1) We would like to implement a new first-principles approach for calculating the magnetic excitation spectra of strongly correlated f-electron materials. This approach will rely on the combination of density functional theory and dynamical mean-field theory. The long-standing problems, such as the evaluation of high-frequency asymptotic behavior of two-particle Green’s function, analytical continuation of dynamical magnetic susceptibility, and numerical stability of Bethe-Salpeter equation iteration solver, will be carefully tackled. (2) The magnetic excitation spectra of some typical strongly correlated f-electron materials, including Mobius topological insulator CeNiSn, Weyl-Kondo semimetal Ce3Bi4Pd3, heavy fermion compound URu2Si2, Mott insulator CmO2, and mixed-valence element Cf will be calculated by the new approach. The relationship between the f electronic configurations and spin dynamics will be revealed. This project will enrich our understandings about the unusual characters of 4f and 5f electrons. The ability of first-principles many-body approach for studying the intriguing properties of strongly correlated f-electron materials will be enhanced as well.

强关联f电子材料的基态电子结构异常复杂，其低能集体激发行为更是异彩纷呈，长期以来备受关注。非弹性中子散射可以用于测量磁激发谱，目前进展十分显著，但是相应的理论计算工作却踯躅不前。本项目拟聚焦于磁激发谱与自旋动力学的第一原理计算方法，研究以下内容: (1)以密度泛函理论结合动力学平均场理论为基础，重点突破双粒子格林函数的高频行为模拟、磁化系数的解析延拓、Bethe-Salpeter方程的迭代求解等关键技术，发展一套适用于强关联f电子材料的磁激发谱的第一原理计算方法。(2)计算典型强关联f电子材料(包括拓扑绝缘体CeNiSn、Weyl-Kondo半金属Ce3Bi4Pd3、重费米子化合物URu2Si2、Mott绝缘体CmO2、变价元素Cf)的磁激发谱，探索f电子结构与自旋动力学的联系。本项目的研究成果将促进对f电子反常行为的了解，提升第一原理计算方法对于强关联材料电子结构的描述能力。

强关联f电子材料展现出丰富多姿的量子物理特性，但是由于实验技术以及理论计算方法的限制，人们对其电子结构，能带结构仍然知之甚少。在本项目的支撑下，我们完成了以下研究工作：(1)结合了密度泛函理论与动力学平均场方法，并以此为基础，发展了一套计算磁激发谱的后处理程序。它以格林函数为输入，可以输出布里渊区中任意高对称点的磁激发谱。(2)为了解决磁激发谱以及格林函数的解析延拓问题，发展了一套名为ACFLOW的解析延拓软件包，它支持最大熵方法、随机解析延拓方法、随机优化方法等主流的解析延拓方法。(3)使用密度泛函理论并结合动力学平均场方法研究了一系列Ce基重费米子材料的电子结构以及奇异的量子物理特性。所涉及的Ce基材料包括：Ce、Ce3Al、CeX(X = N、P、As、Sb)等等。(4)使用密度泛函理论并结合动力学平均场方法研究了一系列锕系材料的电子结构以及奇异的量子物理特性。所涉及的锕系材料包括：Pu、Cm、Cf、CmO2、PuBx、PuIn3、PuSn3、Pu3Ga等等。(5)预测FeGe是一个Kagome-Hund金属，研究了典型Kagome金属CoSn的电子结构随温度的演化趋势。(6)使用线性响应方法标定了主要锕系元素的相互作用参数，为后续计算奠定基础。(7)使用光电子能谱以及第一原理计算方法研究了La薄膜的电子结构。