In order to design molecular electronic devices with high performance and stability, it is crucial to understand their structure-to-property relationships. Single-molecule break junction measurements yield a large number of conductancedistance traces, which are inherently highly stochastic. Here we propose a weakly supervised deep learning algorithm to classify and segment these conductance traces, a method that is mainly based on transfer learning with the pretrain-finetune technique. By exploiting the powerful feature extraction capabilities of the pretrained VGG-16 network, our convolutional neural network model not only achieves high accuracy in the classification of the conductance traces, but also segments precisely the conductance plateau from an entire trace with very few manually labeled traces. Thus, we can produce a more reliable estimation of the junction conductance and quantify the junction stability. These findings show that our model has achieved a better accuracy-to-manpower efficiency balance, opening up the possibility of using weakly supervised deep learning approaches in the studies of single-molecule junctions.

In molecular electronics, electrode-molecule anchoring strategies play a crucial role in the design of stable and high-performance functional single-molecule devices. Herein, we employ aromatic pyrazine as anchors to connect a...

Electron propagation through a molecular device is determined by its quantum electronic structure. We employ molecular conductance orbitals (MCOs) to predict and interpret quantum interference (QI), which contain more information about the electrodes compared with molecular orbitals (MOs) of an isolated molecule. The phases, amplitudes, and alignment of MCOs determine whether they interfere constructively or destructively, which can be seen directly from projection transmissions and QI maps. We apply this intuitive method to butadiene, benzene, and cyclopentadienyl (Cp) anion so that we can elucidate the mechanism of QI among the whole energy range beyond the Fermi level and demonstrate the unique characteristics of MCOs.

One of the key issues for the use of organic radicals in high-performance molecular electronic, spintronic and thermoelectric devices is the retention, at room temperature, of their open-shell character when they are in contact with the metal electrodes of a junction. By means of first-principles quantum transport calculations, we have investigated the stability and the electronic transport properties of single-molecule junctions incorporating a Blatter radical (BR) with thiomethyl linker groups. Our calculations suggest that the BR can unequivocally retain its open-shell nature even when bound to gold electrodes through undercoordinated gold adatoms. Such a state cannot be destroyed by the surrounding solvent molecules and the applied bias. The calculated low-bias conductance values of junctions with BR, an oxidized BR derivative, and a closed-shell stilbene molecule are in quantitative agreement with the available experimental measurements. The small conductance of the BR can be associated with the localization of its π-type singly occupied and singly unoccupied molecular orbitals (SOMO and SUMO), leading to an asymmetrically weak electronic coupling to the two gold electrodes, even though these two orbitals lie very close to the Fermi energy. Furthermore, the nonmonotonic modulation of the junction conductance controlled by a charge gate can be used in experiments to further verify the presence of the unpaired electron. This proof may be realized in practice with an electrostatic solidstate or an electrochemical gate. Our findings deepen the current understanding of the radical−metal interfacial properties and facilitate the design of future radical-based multifunctional molecular devices.

No abstract