Search

Developing advanced bifunctional water splitting electrodes that can perform both hydrogen evolution reaction (HER) and oxygen evolution reaction (OER) is critically important for sustainable hydrogen production because of the simplicity in implementing the system where only one type of electrode material is required. However, the cell voltage of currently available electrolyzers using bifunctional electrodes is still high because maintaining high intrinsic OER and HER activity simultaneously while keeping good charge and mass transport is a challenging task. Herein, the dense ternary NiMoFe alloy nanoparticles are in situ precipitated on the highly conductive hierarchical MoO2 nano-pillar arrays (NiMoFe NPs@MoO2 NPAs) to form an ultra-hydrophilic and super-aerophobic structure by the reductive annealing process-induced phase separation method. Benefiting from the intrinsically high OER and HER activity, abundant active sites and excellent charge and mass transport, the as-obtained NiMoFe NPs@MoO2 NPAs exhibit ultra-low overpotentials of 79 mV for hydrogen evolution and 246 mV for oxygen evolution at 100 mA cm−2. The assembled two-electrode cell delivers the industrially relevant current densities of 100 and 1000 mA cm−2 at record low cell voltages of 1.52 and 1.66 V while maintaining an almost unchanged potential over 1000 h water splitting. The present method of in situ constructing multielement alloy nanoparticles on hierarchical structure also shows great simplicity and broad applicability potential for preparing high-performance electrocatalysts beyond water splitting. [ABSTRACT FROM AUTHOR]

Liquid-phase shear exfoliation (LPSE) is a potential method of large-scale production of good-quality graphene. Obtaining few-layer (<10) graphene with large lateral sizes and high yields remains a challenging issue. Here, we report an improved LPSE processing method involving the addition of mica plates, which serve as bifunctional agents. With the addition of 100 g L−1 of mica, the LPSE (9000 rpm/0.24 mm/1 h) of expanded graphite offers ∼100% few-layer graphene with average lateral sizes (ALSs) of ∼2.8 μm and yields of up to 27% (the highest ever reported). The graphene produced by LPSE without any additives, however, contains only ∼77% few-layer nanosheets, and has ALSs of ∼1.3 μm and yields of 8%. Significantly, the addition of SiO2 microspheres as a comparison has almost no influence on the LPSE. The mica plates control the vector of fluids in LPSE and guide the oriented arrangement of the layered materials (e.g., expanded graphite), leading to significantly improved LPSE efficiency. 2D transition metal dichalcogenides (TMDs) with improved yields were also successfully prepared using LPSE with the addition of mica. [ABSTRACT FROM AUTHOR]

Deep learning has greatly promoted intelligent fault diagnosis for bearings and has led to the development of many intelligent fault recognition methods based on deep neural networks. However, bearing fault signals in industrial applications are susceptible to noise interference, and intelligent recognition methods for bearing faults under strong noise interference are lacking. An intelligent fault pattern recognition model termed IFPR-MDL, which is based on a Transformer assisted by depthwise separable convolutional (DSC) and bidirectional long short-term memory (BiLSTM) networks, is designed to identify bearing faults under strong noise conditions. In this model, multiscale features are first extracted by a multiscale feature extraction module designed on the basis of DSC; then, contextual information is further extracted by a BiLSTM, and information from different scales is integrated by Transformer encoders. The DSC network facilitates model parameter reduction, the BiLSTM network has expertise in mining temporal features, and the Transformer network excels at capturing long-range feature correlations. This model fully utilises the advantages of these models and complements them to improve the anti-interference ability of bearing fault pattern recognition. Tests on several public datasets show that the proposed model has higher recognition accuracy, stronger generalisation ability, and noise resistance in strong noise situations. [ABSTRACT FROM AUTHOR]