Multireference Equation of Motion Coupled Cluster Study of Atomic Excitation Apectra

Variants of the family of the recently developed Multireference Equation of Motion Coupled Cluster (MR-EOM-CC) approaches are applied to the atomic excitation spectra of the first-row transition metals, for which experimental data are readily available. The first part of my research is to apply MR-E...

Full description

Bibliographic Details
Main Author: Liu, Zhebing
Format: Master Thesis
Language:English
Published: University of Waterloo 2015
Subjects:
Online Access:http://hdl.handle.net/10012/9381
Description
Summary:Variants of the family of the recently developed Multireference Equation of Motion Coupled Cluster (MR-EOM-CC) approaches are applied to the atomic excitation spectra of the first-row transition metals, for which experimental data are readily available. The first part of my research is to apply MR-EOM-CC approaches to the atomic excitation spectra of the neutral and the +1 and +2 Cr, Mn, Fe and Co atoms using ACESII program. Scalar relativistic effects are considered but spin-orbit coupling is not included. The computational results are compared to J-averaged experimental values and the errors are typically below 0.1 eV. The second part of this project is to combine MR-EOM approach with a simple treatment of spin-orbit coupling and the resulting MR-EOM-CC-SOC approach can be used to the atomic excitation spectra and the splitings in each L-S multiplet of the Cr, Mn, Fe and Co neutral atoms using the ORCA program. The errors of the spin-orbit calculations in the ORCA program are somewhat larger compared to the J-averaged calculations run in ACESII program and this is still under investigation. From both parts of this research, one can easily find that upon introducing additional cluster operators T, S, X, D and U in different variations of MR-EOM, the dimension of the final diagonalization space is greatly reduced, the overall cost of the calculation is significantly reduced, and finally, the accuracy of the calculations is, perhaps surprisingly, improved.