First-Principles Study on Non-Collinear Magnetic States and Magneto-Strain Effects

非共线磁态和磁应变效应的第一性原理研究

基本信息

批准号：
12640349
负责人：
SUZUKI Naoshi
金额：
$ 1.54万
依托单位：
Osaka University
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (C)
财政年份：
2000
资助国家：
日本
起止时间：
2000 至 2002
项目状态：
已结题

项目摘要

(1) First-principles study on noncollinear magentismWe have carried out an ab initio molecular dynamics of liquid oxygen, a molecular fluid in which the individual O_2 units carry a molecular magnetic moment. In addition to the atomic and electronic structures, our simulation describes the evolution of the noncollinear magnetic structure. The atomic structure shows a strong preference for parallel alignment of first-neighbor molecules. The magnetic structure shows strong short range antiferromagnetic correlations, in agreement with spin-polarized neutron diffraction data. The short range correlations, observed in both structural and magnetic properties, primarily result from appropriate trajectories of colliding O_2 molecules. Our simulation also reveals the occurrence of several long-living O_4 units which survive for time periods longer than four times the average residence time observed during collisions.(2) Development and implementation of density functional perturbation theory : Lattice dynamics from first principlesWe have implemented the variational formulation of density functional perturbation theory (DFPT) to investigate the harmonic lattice dynamics. Response of the charge density and wave functions to atomic displacements within linear response theory, is computed by minimizing the second-order derivative of total energy. The implementation is based on a plane-wave basis set and pseudopotentails. For the optimization of the first-order wave functions, the damped molecular dynamics (MD) scheme and the Lagrange multiplier method is adopted for the first time. We find that this method is very robust and gives a stable convergence of the wave functions. The computational efforts to calculate the first-order wave functions are comparable to the ground state calculations, similar to other iterative methods. Our method is simple and suitable for the implementation of DFPT to the MD code which is based on the original Car-Parrinello MD scheme.

(1)非共线磁力的第一性原理研究我们对液氧进行了从头算分子动力学研究。液氧是一种分子流体，其中单个O_2单元带有分子磁矩。除了原子和电子结构之外，我们的模拟还描述了非共线磁结构的演变。原子结构显示出对第一相邻分子平行排列的强烈偏好。磁结构显示出强烈的短程反铁磁相关性，与自旋极化中子衍射数据一致。在结构和磁性方面观察到的短程相关性主要是由碰撞 O_2 分子的适当轨迹引起的。我们的模拟还揭示了几个长寿 O_4 单元的出现，它们的存活时间超过碰撞期间观察到的平均停留时间的四倍。(2) 密度泛函微扰理论的发展和实现：来自第一原理的晶格动力学我们已经实现了密度泛函微扰理论 (DFPT) 的变分公式用于研究调和晶格动力学。线性响应理论中电荷密度和波函数对原子位移的响应是通过最小化总能量的二阶导数来计算的。该实现基于平面波基组和赝势。对于一阶波函数的优化，首次采用阻尼分子动力学（MD）格式和拉格朗日乘子法。我们发现这种方法非常鲁棒，并且可以稳定地收敛波函数。计算一阶波函数的计算量与基态计算相当，与其他迭代方法类似。我们的方法简单，适合对基于原始Car-Parrinello MD方案的MD代码实施DFPT。