Molecular electronics is an intriguing area of research and it has received a considerable amount of attention in the recent years. This attention is partially fuelled by possible application of single molecules as microelectronic devices and by fundamental research about charge transport processes. In many areas of molecular electronics, large data sets are recorded to account for the large fluctuations on the nanoscale.
Over the course of this research project, a new, machine-learning-based approach to data analysis was developed, namely MPVC. It is demonstrated herein, how this fundamentally new approach to data analysis can provide greater insight into complex single molecule conductance data. At the same time, this approach does not make assumptions about the data.
The new analysis approach was then applied to single molecule conductance data of organometallic molecules. These types of molecules have received considerable attention lately because of their potential applications as molecular wires or single molecular switches. It is demonstrated how MPVC analysis can provide additional insights into these data.
Finally, results on single molecular conductance data for four ring-shaped, organometallic compounds will be presented. This is the first time that single molecule conductance data has been reported on for ring-shaped compounds. Ring-shaped molecules are of particular interest for the molecular electronics community because of their potential application as single molecular switches. Their spatially separated current pathways could provide a testbed for charge transport studies with a special focus on quantum interference.
Overall, the work presented within this thesis demonstrates the use of machine learning techniques in molecular electronics and provides novel insights into the charge transport on the nanoscale with a special focus on organometallics and is thus a significant contribution to the field.