Research Assistant Professor Tillmann Kubis

My research interest includes all topics of equilibrium and non-equilibrium phenomena in nanodevices and molecules. This covers electronic and phonon bandstructures as well as heat, charge and spin transport in nanodevices. I have the honor to lead a great team of talented students. Most of our work enters the multipurpose, nanodevice simulation tool NEMO5.

IWCN 2019 - Workshop on HPC TCAD
invited talk 


ESSCIRC-ESSDERC: Atomistic Green's functions: the beauty of self-energies



The value of nonequilibrium Green's function (NEGF) implementations stands and falls with the quality of its self-energies. This presentation gives an introduction to NEGF. It will be explained how self-energies cause NEGF to fundamentally differ from most other quantum methods. Atomistic examples of phonon and impurity scattering self-energies agree quantitatively with experiments. The latest development in NEGF, the ROBIN method, overcomes notorious shortcomings of periodic boundary conditions in defect and disordered material simulations.

In this talk, an introductory overview of the nonequilibrium Green’s function (NEGF) method will be given. NEGF results in state-of the art semiconductor nanodevices will illustrate the strengths of the method. Unique benefits of the method for molecular chemistry will be highlighted and pathways to extend the NEGF application space to fluids will be sketched.

Since its introduction in the early 1960’s, the NEGF method has been applied on a large variety of many-particle systems, ranging from electron, heat and spin transport in metals, semiconductors and molecular junctions. Typical NEGF applications require a thorough description of coherent quantum mechanical effects such as particle tunneling, confinement and interference as well as incoherent effects such as phase-destructive scattering on thermal vibrations, device imperfections or finite lifetime effects. All of these effects are expected to be equally important in molecular reactions.

The successful isolation of graphene in 2004 opened up the exciting new research field of 2D materials. These materials host a long list of unique mechanical, electrical and chemical features that promise important device applications. Reliable performance predictions of 2D nanodevices must embrace coherent quantum mechanical effects (tunneling, confinement and interferences) atomistic effects (corrugation, subatomic confinement) and incoherent effects (phonon scattering and device imperfections). Subatomic resolution is needed, but techniques must be efficient enough to model real-size devices. Recently, the multipurpose simulation tool NEMO5 was augmented with the maximally localized Wannier function (MLWF) representation. This representation offers a good balance between numerical efficiency and subatomic resolution. MLWF parameterizations are highly transferable and free from ambiguities that have plagued empirical tight binding models.

In this talk, I will briefly discuss the MLWF approach and compare it to DFT and atomistic tight binding. Initial results using the MLWF approach for 2D material based devices will be discussed and compared to experiments. These results unveil systematic band structure changes as functions of the layer thickness and the applied gate potential. The electrostatic response depends on the location of the band edges in the Brillouin zone, their degeneracy and associated wavefunctions. All these properties turn out to be tunable. Scattering rates, mobilities and density of states are tightly bound to such band structure details as well. Even the bandgap is a function of the layer thickness and the applied electric field. Fitting NEMO5’s gate control of bandgaps to experimental data allows us to deduce the layer thickness dependence of the dielectric constant in the 2D materials. The enhancements discussed in this talk provide NEMO5 with the new capabilities needed to play an important role in the exploration of novel 2D devices.

State of the art nanodevices have reached length scales in which a clear distinction of fundamental physics, material science and device engineering is not applicable. Reliable performance predictions of nanodevices have to embrace all these fields: coherent quantum mechanical effects such as tunneling, confinement and interferences, atomistic effects such as strain profiles, alloy compositions and interface relaxations, uncertainty effects such as incoherent scattering on phonons and device imperfections and electrostatic effects have all similarly strong impact on the nanodevice performances. Many of these aspects are described in different physical models and are characterized on different length scales.

In this talk, it will be shown how the concept of self-energies can be used to interface all these fields into the same nanotechnology modeling framework. Self-energies are most commonly used in the quantum transport method of nonequilibrium Green’s functions (NEGF). The NEGF method is widely accepted as the most consistent method for modeling coherent and incoherent effects. Given that this method allows for atomic resolution and the inclusion of strain effects, alloy disorder and electrostatics, it is most often used for charge, heat and spin transport in nanometer scaled systems. Nevertheless, it will be shown how self-energies can get used beyond NEGF. It is also part of this talk how self-energies can set nanotechnology into the context to solve 3 of mankind’s biggest challenges: shortage of energy, shortage of fresh water and the decline of world’s economy.

This website does not reflect the views of Purdue University. 
© 2019 Tillmann Kubis