Theoretical Investigation of Spin-Dependent Transport from Adsorbates to Magnetic Substrates
Final Report Abstract
The present project focused on the study of self-assembled monolayers (SAMs) of organic molecules specifically for use in electronic or spintronic applications. Thereby, we followed a two lines of research. First, in a collaboration with colleagues from experimental physics, we elucidated the formation, structure and stability of such monolayers. The first surprise there was that the commonly used thiolate connectors are not particularly well suited for the formation of SAMs on at least Ni substrates. The second line of research which was aimed at establishing a comprehensive modeling strategy for charge transfer from such SAMs to the substrate, was delayed, though, through the unforeseen departure of this part’s main PI. Before his departure, a first benchmark study of the static background time-dependent density functional theory (SB-TD-DFT) revealed some shortcomings of this approach, specifically with respect to absolute charge transfer rates. In a later study, conducted under the supervision of Dr. Harald Oberhofer, who had stepped in after the departure of the PI, the source of this error was finally shown to be an insufficient sampling of the substrates electronic structure. Employing a much improved method for the calculation of electronic couplings, developed in the course of this project, we could trace the error to a lack of sampling of the metal’s Brillouin zone, which explains why it is present in both cluster models and Γ-point descriptions of extended surfaces. This was also somewhat surprising, given that much of the available literature described studies on exactly such model systems. The new family of electronic coupling calculation methods, termed the projection-operator diabatization version 2 (POD2), was specifically formulated to correct the main shortcoming of its predecessor, which due to a necessary orthogonalization step of the involved orbitals was unable to fully spatially separate initial and final states. This, in general, leads to an underestimation of the calculated couplings and, even more importantly, to a complete lack of convergence with basis set size. This was a well known problem, with common wisdom being to just conduct the calculations at a much too small basis set size. This option, though is not available for density functional theory codes that employ naturally more disperse basis sets such as numeric atomic orbitals. The new POD2 methods completely remove this shortcoming and open up the approach to a much wider user base. The calculated couplings then need be evaluated further in order to yield e.g. charge transfer rates comparable with experiment. In that respect, we could show that direct time propagation of the charge or the spin in the SB-TD-DFT method does not yield any improved results over a perturbative treatment with a generalization of Fermi’s golden rule. In summary, our project yielded both insight into the formation and stability of self assembled monolayers and necessary methodological advances towards a comprehensive set of theoretical tools to study charge transfer from adsorbates to extended metallic surfaces.
Publications
- “Charge transfer dynamics from adsorbates to surfaces with single active electron and configuration interaction based approaches”, Chem. Phys. 446, 24–29 (2015)
R. Ramakrishnan and M. Nest
(See online at https://doi.org/10.1016/j.chemphys.2014.11.005) - “Thiolatebonded self-assembled monolayers on ni (111): bonding strength, structure, and stability”, J. Phys. Chem. C 119, 15455–15468 (2015)
F. Blobner, P. N. Abufager, R. Han, J. Bauer, D. Duncan, R. Maurer, K. Reuter, P. Feulner, and F. Allegretti
(See online at https://doi.org/10.1021/acs.jpcc.5b04351) - “Improved projection-operator diabatization schemes for the calculation of electronic coupling values”, J. Chem. Theory Comput. 16, 7431–7443 (2020)
S. Ghan, C. Kunkel, K. Reuter, and H. Oberhofer
(See online at https://doi.org/10.1021/acs.jctc.0c00887)
