C3.11: Dynamic Response of Molecular Nanostructures

Subproject Leader: Ferdinand Evers

Institut für Nanotechnologie, KIT

 

Contributing Scientists:

Present: Michiel-Jan Van Setten
Past: Alexej Bagrets

 

Figure 1:
Comparison of ionization potentials of different molecules obtained with Hartree-Fock method, DFT (HOMO-energy), GW-approach and the experimental value.
Already to lowest order, G0W0, the improvement above DFT (with standard functionals) is significant.
Figure 2:
(a) Spin transition complex [Fe(bpp)2] with an extra anchor group as used in the experiment.
(b) Sketch of the electro-migrated break junction setup used for contacting the molecule in the Delft group.

Theoretical Spectroscopy of Molecular Nanostructures

We develop, test and apply novel approaches that allow to calculate spectroscopic properties and the dynamical response of molecular systems (including reservoirs) with an accuracy beyond the present standard methods based on ground-state density functional theories (DFT). Such theoretical investigations are of importance for several reasons:

(i) Most experiments probe the response of molecular systems to an external stimulus. The molecule's spectral properties determine these responses and therefore a detailed understanding of the prior helps the interpretation of the latter.

(ii) The excitation spectrum encodes crucial information about the microscopic state of a molecule and its interaction with the environment. To disentangle environmental impact from genuine molecular affairs usually requires a thorough theoretical analysis.

(iii) Current computational approaches (DFT, coupled cluster methods, configuration interaction approaches, etc.) are either not accurate enough or not applicable to intermediate size molecules [1].

Method Development: At the moment our strongest emphasis is on the implementation of the GW-method into the Karlsruhe quantum chemistry package TURBOMOLE. The method has proven to be efficient and precise when calculating elementary excitations (band and spectral gaps) in solids; first applications to molecular systems strongly suggest that it will also be useful for applications in quantum chemistry. Our implementation into the TURBOMOLE package is designed to make use of previous achievements of this code in terms of computational efficiency and therefore holds a promise that we can reach system sizes which could not be treated before using the GW-approach.

Two sidelines of method development are also pursued. First, we develop the GGA+U-method for quantum chemistry applications on magnetic systems. This method can be understood as a strongly simplified variant of GW. It is the analogue of the LDA+U approach, which is well known in solid state applications. Our second sideline in method development relates to ab-intio calculations of the electrical response of molecular metamaterials. The basic idea is to feed (improved) response calculations of individual molecules into computations of optical properties of molecular dielectrics composed of such molecules.

Applications: Our methodological developments are in their prototypical test phase. For more complicated applications to experimentally relevant systems we have been relying on previously existing computational machinery. Areas of active research include:

Hydrogen storage materials: Since very recently the GW-method is available in standard condensed matter codes, e.g., VASP, applications to the formation of hydrides and consecutive comparison with experiments allows to test the method already now. Adopting this strategy, in the GW-approach is employed to calculate optical properties of hydrides that are relevant for hydrogen storage. [2] In this spirit in [3] calculated spectra were used to identify, and exclude, the formation of specific materials

Molecular Instabilities: Systems near instabilities react sensitively to external stimuli. For this reason they are attractive candidates to build molecular materials or devices, like switches, which are supposed to change their properties when tuning an external control parameter. Our focus is on mechanical [4] and magneto-mechanical instabilities which trigger the "spin transition". Specifically, our research has identified a spin transition complex, see Figure 2a, which undergoes a transition from a low spin, S=0, state into a high spin state upon charging. [5] How such transitions would manifest themselves unambiguously in transport experiments in an exciting present days research topic in the field strongly correlated electron systems ("underscreened Kondo effect").

 References

[1] P. Schmitteckert and F. Evers, Exact ground state density functional theory for impurity models coupled to external reservoirs and transport calculations, Phys. Rev. Lett. 100, 086401 (2008); F. Evers and P. Schmitteckert, Broadening of the Derivative Discontinuity in Density Functional Theory, submitted (2011)
[2] M.J. van Setten, R. Gremaud, G. Brocks, B. Dam, G. Kresse, and G.A. de Wijs, Optical properties of sodium alanate, GW0 BSE calculations and thin film measurements, accepted Phys. Rev. B
[3]  M. Gonzalez-Silveira, R. Gremaud, H. Schreuders, M.J. van Setten, E. Batyrev, A. Rougier, L. Dupont, E.G. Bardaji, W. Lohstroh, and B. Dam, The in-situ deposition of alkali (earth) hydride thin films to investigate the formation of Reactive Hydrogen Complexes. Journal of Physical Chemistry C 114, 13895 (2010)
[4] V. Meded, A. Arnold, A. Bagrets and F. Evers, Molecular current switch controlled by pulsed bias voltages, SMALL 5, 2218 (2009)
[5] A. Bernand-Mantel, V. Meded, J. Seldenthuis, A. Beukman, K. Fink, M. Ruben, F. Evers and H. S. J. van der Zant, Spin-coupled double-quantum-dot behavior inside a single molecule transistor, accepted Phys. Rev. B. (2011)

 

List of Publications 2006-2011 as PDF

Subproject Report 2006-2010 as PDF