Research

I am a condensed matter theorist fascinated with the effect of strong electronic correlations on the properties of materials in general, and of nanoscale devices in particular. My motivation to study these complex systems is mainly driven by the curiosity to understand fundamental aspects of quantum mechanics and many-body physics. On the other hand, strong correlations often give rise to intriguing phenomena that could be exploited for new materials and functionality of electronic devices. In the last couple of years my research has focused on strong electronic correlation effects in nanoscale devices, mainly atomic- or molecular junctions comprising transition metal atoms. Such devices are sought for as building blocks for prospective spintronics applications or as ultimately miniaturized magnetic information storage devices. A particularly important example of a phenomenon driven by electronic correlations and frequently observed in nanoscale devices is the Kondo effect. The multi-orbital nature of the Kondo physics in these systems and the interplay between local Kondo physics and the molecular degrees of freedom makes these systems particularly interesting from a fundamental point of view. At the same time this complexity poses a great challenge to the theoretical description of these systems. To this end I have developed a novel ab initio approach, the NanoDMFT method, that allows to calculate the electronic, magnetic and transport properties of nanoscale devices, taking explicitly into account strong electonic correlations effects.

NanoDMFT: Ab initio description of strong electronic correlations at the nanoscale

In the last couple of years I have developed a novel ab initio approach for calculating the electronic structure and transport properties of nanoscale devices explicitly taking into account strong local correlations originating e.g. from the strongly interacting 3d-electrons of transition metal atoms. The approach is basically an adaption of the DFT+DMFT method, originally developed for the realistic description of strongly correlated materials, to the special situation of nanoscale systems. More precisely, Density Functional Theory (DFT) based ab initio quantum transport calculations are combined with Dynamical Mean-Field Theory (DMFT). While the DFT part yields a static mean-field description of the weakly to moderately interacting electrons in the device, DMFT takes care of the strong dynamic correlations originating from strongly interacting localized electrons in the device such as the 3d- and 4f-shells of transition metal atoms. This method forms the basis for the research in the NanoDMFT subgroup that I am directing at the MPI Halle. Current research efforts are the further development of the NanoDMFT method and the application of the method for calculating the electronic and magnetic properties of experimentally relevant systems.

The NanoDMFT method has been implemented within the ANT.G package which interfaces Gaussian in order to implement DFT based quantum transport calculations. The basic NanoDMFT methodology is described in the following papers:

• Towards a full ab initio theory of strong electronic correlations in nanoscale devices, J. Phys.: Condens. Mat. 27, 245606 (2015)
• Dynamical mean-field theory for molecular electronics, Physical Review B 82 (19), 195115 (2010)
• Kondo effect and conductance of nanocontacts with magnetic impurities, Physical Review Letters 103, 016803 (2009)

Kondo effect in atomic and molecular junctions

Nanoscale junctions made from magnetic molecules or atomic constrictions coupled to conducting electrodes are prospective building blocks for ultimately miniaturize spintronics and magnetic information storage applications. But whenever some local magnetic moment is coupled to a sea of conduction electrons the Kondo effect may arise. The Kondo effect is one of the most intriguing phenomena induced by strong electronic correlations, and leads to the screening of the magnetic moment by formation of a total spin-singlet state between local magnetic moment and the surrounding conduction electrons. In view of the prospective applications it is necessary to understand and predict if and how the Kondo effect occurs in magnetic molecular junctions, and how to control it. On the other hand, the multi-orbital nature of the Kondo effect in these systems together with the low symmetry environment and coupling to the organic ligands can give rise to intriguing and novel physical phenomena [1].

Using the NanoDMFT method it is now possible to actually predict the electronic and magnetic properties of molecular junctions made from transition metal complexes under the influence of strong correlation effects as proven by the excellent agreement of our calculations with experimentally measured systems [1,2]. Moreover, these calculations allow to obtain an unprecedented deep understanding of the intricate interplay between multi-orbital Kondo physics and the molecular structure [1]. Often these calculations are also crucial for a proper interpretation of the experimental results in these rather complex systems: In a recent work we could show in collaboration with the experimental group of Richard Berndt at the University of Kiel (Germany) that the actual lineshape of the Kondo feature measured in the dI/dV spectra relates information about the voltage drop within the molecular junction [2].

[1] Kondo effect and spin quenching in high-spin molecules on metal substrates, Phys. Rev. B 88, 134417 (2013)

[2] Shifting the voltage drop in electron transport through a single molecule, Phys. Rev. Lett. 115, 016802 (2015)

Properties of nanoscale quantum magnets