Configuration-Specific Insight into Single-Molecule Conductance and Noise Data Revealed by the Principal Component Projection Method

Abstract

We explore the merits of neural network boosted, principal-component-projection-based, unsupervised data classification in single-molecule break junction measurements, demonstrating that this method identifies highly relevant trace classes according to the well-defined and well-visualized internal correlations of the data set. To this end, we investigate single-molecule structures exhibiting double molecular configurations, exploring the role of the leading principal components in the identification of alternative junction evolution trajectories. We show how the proper principal component projections can be applied to separately analyze the high- or low-conductance molecular configurations, which we exploit in 1/f-type noise measurements on bipyridine molecules. This approach untangles the unclear noise evolution of the entire data set, identifying the coupling of the aromatic ring to the electrodes through the π orbitals in two distinct conductance regions, and its subsequent uncoupling as these configurations are stretched.

Total variation denoising-based method of identifying the states of single molecules in break junction data

“…Estimating the plateau lengths is critical in inferring the stabilities of individual single-molecule junction states and molecular junction structures. Furthermore, identifying regions of identical conduction states via plateau analysis facilitates the evaluation of the noise magnitude within each conduction state [ 26 , 35 ]. In a single-molecule junction, the conductance and its variation are crucial in representing the junction-specific state.…”

Making the Most of Nothing: One-Class Classification for Single-Molecule Transport Studies

“…In Figure , we present the RT-bpy data set. The experimental setup for gathering the data is explained in the Supporting Information and in previous publications. , After preprocessing, it consists of 9787 blank traces and 20,129 traces from an experiment with a molecule. In Figure A,B, we show 10 traces from a blank experiment and an experiment with an added molecule, respectively.…”

“…To minimize the impact of background signal on the accuracy of results, a number of different methods to filter blank traces have been developed, including clustering techniques, − principal component analysis method, − and custom filtering algorithms. − The widespread use of filtering blank by several different group traces underlines the need for a structured and robust approach to this problem.…”