Tim Zuehlsdorff

Research

My main research interest lies in developing computational techniques capable of predicting the optical properties of large systems, ranging from semiconductor nanoparticles to pigment-protein complexes. These types of systems generally present a challenge for computational modelling: While the optical absorption of semiconductor nanoparticles is strongly influenced by quantum confinement and surface effects, the coupling of a pigment to light is dependent on its protein environment. In each case, accurate predictions of absorption spectra can only be achieved by a quantum mechanical treatment of systems containing thousands of atoms.

For the purpose of modelling optical properties of large systems, I am making use of the time-dependent extension of density-functional theory (TDDFT). TDDFT has the advantage of providing a good balance between computational cost and achievable accuracy, making it well-suited to large scale applications that cannot be treated by more advanced quantum chemistry methods. However, in its standard formulation, TDDFT exhibits a cubic scaling of computational effort with the system size studied, making the treatment of systems containing thousands of atoms prohibitively expensive computationally. For this reason, my research is focused on applying and extending techniques that have been developed to allow for a linear scaling of computational effort with system size in standard ground-state density functional theory to the TDDFT. The methodological developments that originate from my research are part of the linear-scaling DFT code ONETEP.