Gauge-origin dependence in electronic g-tensor calculations.

J Chem Phys

Chair of Theoretical Chemistry and Center for Integrated Protein Science Munich (CIPSM), Department of Chemistry, University of Munich (LMU), Butenandtstr. 7, 81377 Munich, Germany.

Published: June 2018

We present a benchmark study on the gauge-origin dependence of the electronic g-tensor using data from unrestricted density functional theory calculations with the spin-orbit mean field ansatz. Our data suggest in accordance with previous studies that g-tensor calculations employing a common gauge-origin are sufficiently accurate for small molecules; however, for extended molecules, the introduced errors can become relevant and significantly exceed the basis set error. Using calculations with the spin-orbit mean field ansatz and gauge-including atomic orbitals as a reference, we furthermore show that the accuracy and reliability of common gauge-origin approaches in larger molecules depends strongly on the locality of the spin density distribution. We propose a new pragmatic ansatz for choosing the gauge-origin which takes the spin density distribution into account and gives reasonably accurate values for molecules with a single localized spin center. For more general cases like molecules with several spatially distant spin centers, common gauge-origin approaches are shown to be insufficient for consistently achieving high accuracy. Therefore the computation of g-tensors using distributed gauge-origin methods like gauge-including atomic orbitals is considered as the ideal approach and is recommended for larger molecular systems.

Download full-text PDF

Source
http://dx.doi.org/10.1063/1.5028454DOI Listing

Publication Analysis

Top Keywords

common gauge-origin
12
gauge-origin dependence
8
dependence electronic
8
electronic g-tensor
8
g-tensor calculations
8
calculations spin-orbit
8
spin-orbit field
8
field ansatz
8
gauge-including atomic
8
atomic orbitals
8

Similar Publications

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!