一种用于周期性体系的高效缩放反向自旋MP2方法
An Efficient scaled opposite-spin MP2 method for periodic systems
AI总结:
本文提出SOS-RILT-MP2算法,结合RI与拉普拉斯变换,实现O(N4)缩放的高效周期体系SOS-MP2计算,在苯分子晶体中显著提升效率,适用于大晶胞与缺陷材料研究。
摘要:
我们开发了SOS-RILT-MP2,一种高效的基于高斯基的周期性缩放反向自旋二阶Møller-Plesset微扰理论(SOS-MP2)算法,该算法利用分辨率恒等近似(RI)结合拉普拉斯变换技术(LT)。在我们之前的工作[J. Chem. Phys. 157, 174112 (2022)]中,我们表明,与常规MP2和一些领先的密度泛函相比,SOS-MP2对12种简单半导体和绝缘体的晶格常数、体模量和内聚能给出了更好的预测。在这项工作中,我们提出了一种高效的SOS-MP2算法,其缩放为O(N4),其中N是晶胞中的原子数,并且随着布里渊区中k点数量的增加,缩放性降低。我们在分子和固态体系上实现并测试了我们的算法,通过系统地增加原子数、基组大小和k点采样的密度,确认了预测的缩放行为。以苯分子晶体为案例研究,我们证明了SOS-RILT-MP2与常规MP2相比实现了显著提高的效率。这种高效的算法未来可用于研究具有大晶胞的复杂材料以及缺陷结构。
英文摘要:
We develop SOS-RILT-MP2, an efficient Gaussian-based periodic scaled opposite-spin second-order Møller-Plesset perturbation theory (SOS-MP2) algorithm that utilizes the resolution-of-the-identity approximation (RI) combined with the Laplace transform technique (LT). In our previous work [J. Chem. Phys. 157, 174112 (2022)], we showed that SOS-MP2 yields better predictions of the lattice constant, bulk modulus, and cohesive energy of 12 simple semiconductors and insulators compared to conventional MP2 and some of the leading density functionals. In this work, we present an efficient SOS-MP2 algorithm that has a scaling of O(N4) with the number of atoms N in the unit cell and a reduced scaling with the number of k-points in the Brillouin zone. We implemented and tested our algorithm on both molecular and solid-state systems, confirming the predicted scaling behavior by systematically increasing the number of atoms, the size of the basis set, and the density of k-point sampling. Using the benzene molecular crystal as a case study, we demonstrated that SOS-RILT-MP2 achieves significantly improved efficiency compared to conventional MP2. This efficient algorithm can be used in the future to study complex materials with large unit cells as well as defect structures.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。