Theoretical Developments for the Real-Time Description and Control of Nanoscale Systems

Sammanfattning: In this thesis we focus on improvements of the description of the electron-electron correlation effects in nonequilibrium nanosystems. We mainly focus on developments of two nonequilibrium methods, namely the formalism of Nonequilibrium Green’s Function and Time Dependent Density Functional Theory and we explore the possibility to improve existing approximations in these theories. A smaller part of the thesis is devoted to the Exact Diagonalization method which provides a numerically exact description of small systems.Paper I: We review the current methods for description of correlated materials in nonequilibrium and their connection to pump-probe spectroscopy.Paper II: We propose a hybrid method for the real time dynamics of strongly correlated materials which includes memory effects beyond the adiabatic local density approximation.Paper III: We study the dynamics of desorption of a molecule from a surface with different levels of approximation for both the nuclear and the electronic part. We compare a full quantum mechanical treatment to the Ehrenfestapproximation for the molecule and perturbative approximations for the electrons.Paper IV: We develop a theory of current-induced forces within Adiabatic Ehrenfest Dynamics which includes effects of electron-electron interactions. We study a dependence of the electronic friction on interaction strength.We also benchmark it against nonadiabatic Ehrenfest dynamics.Paper V: We study the competition of interaction and disorder in systems with steady state currents - in transport and ring geometries. We exactly define the exchange-correlation screening of the disorder by the interactioneffects via Kohn–Sham construction of DFT.Paper VI: We study a competition between Kondo and RKKY interaction in small clusters of Periodic Anderson Model (ring geometries), we construct a nonequilibrium Doniach-phase like diagram. We then determine anoptimal pulse to induce transitions with the highest fidelity.