Defect-rich MoS2 ultrathin nanosheets are synthesized on a gram scale for electrocatalytic hydrogen evolution. The novel defect-rich structure introduces additional active edge sites into the MoS2 ultrathin nanosheets, which significantly improves their electrocatalytic performance. Low onset overpotential and small Tafel slope, along with large cathodic current density and excellent durability, are all achieved for the novel hydrogen-evolution-reaction electrocatalyst.

Molybdenum disulfide (MoS2) has emerged as a promising electrocatalyst for catalyzing protons to hydrogen via the so-called hydrogen evolution reaction (HER). In order to enhance the HER activity, tremendous effort has been made to engineer MoS2 catalysts with either more active sites or higher conductivity. However, at present, synergistically structural and electronic modulations for HER still remain challenging. In this work, we demonstrate the successfully synergistic regulations of both structural and electronic benefits by controllable disorder engineering and simultaneous oxygen incorporation in MoS2 catalysts, leading to the dramatically enhanced HER activity. The disordered structure can offer abundant unsaturated sulfur atoms as active sites for HER, while the oxygen incorporation can effectively regulate the electronic structure and further improve the intrinsic conductivity. By means of controllable disorder engineering and oxygen incorporation, an optimized catalyst with a moderate degree of disorder was developed, exhibiting superior activity for electrocatalytic hydrogen evolution. In general, the optimized catalyst exhibits onset overpotential as low as 120 mV, accompanied by extremely large cathodic current density and excellent stability. This work will pave a new pathway for improving the electrocatalytic activity by synergistically structural and electronic modulations.

Existing person re-identification benchmarks and methods mainly focus on matching cropped pedestrian images between queries and candidates. However, it is different from real-world scenarios where the annotations of pedestrian bounding boxes are unavailable and the target person needs to be searched from a gallery of whole scene images. To close the gap, we propose a new deep learning framework for person search. Instead of breaking it down into two separate tasks-pedestrian detection and person re-identification, we jointly handle both aspects in a single convolutional neural network. An Online Instance Matching (OIM) loss function is proposed to train the network effectively, which is scalable to datasets with numerous identities. To validate our approach, we collect and annotate a large-scale benchmark dataset for person search. It contains 18, 184 images, 8, 432 identities, and 96, 143 pedestrian bounding boxes. Experiments show that our framework outperforms other separate approaches, and the proposed OIM loss function converges much faster and better than the conventional Softmax loss.