Lithium metal is an ideal electrode material for future rechargeable lithium metal batteries. However, the widespread deployment of metallic lithium anode is significantly hindered by its dendritic growth and low Coulombic efficiency, especially in ester solvents. Herein, by rationally manipulating the electrolyte solvation structure with a high donor number solvent, enhancement of the solubility of lithium nitrate in an ester‐based electrolyte is successfully demonstrated, which enables high‐voltage lithium metal batteries. Remarkably, the electrolyte with a high concentration of LiNO3 additive presents an excellent Coulombic efficiency up to 98.8 % during stable galvanostatic lithium plating/stripping cycles. A full‐cell lithium metal battery with a lithium nickel manganese cobalt oxide cathode exhibits a stable cycling performance showing limited capacity decay. This approach provides an effective electrolyte manipulation strategy to develop high‐voltage lithium metal batteries.

The insertion/deinsertion mechanism enables plenty of charge-storage sites in the bulk phase to be accessible to intercalated ions, giving rise to at least one more order of magnitude higher energy density than the adsorption/desorption mechanism. However, the sluggish ion diffusion in the bulk phase leads to several orders of magnitude slower charge-transport kinetics. An ideal energy-storage device should possess high power density and large energy density simultaneously. Herein, surface-modified Fe O quantum dots anchored on graphene nanosheets are developed and exhibit greatly enhanced pseudocapacitance via fast dual-ion-involved redox reactions with both large specific capacity and fast charge/discharge capability. By using an aqueous Na SO electrolyte, the oxygen-vacancy-tuned Fe O surface greatly enhances the absorption of SO anions that majorly increase the surface pseudocapacitance. Significantly, the Fe O -based electrode delivers a high specific capacity of 749 C g at 5 mV s and retains 290 C g at an ultrahigh scan rate of 3.2 V s . With a novel dual-electrolyte design, a 2 V Fe O /Na SO //MnO /Na SO asymmetric supercapacitor is constructed, delivering a high energy density of 75 W h kg at a power density of 3125 W kg .

Visually rich documents (VRDs) are ubiquitous in daily business and life. Examples are purchase receipts, insurance policy documents, custom declaration forms and so on. In VRDs, visual and layout information is critical for document understanding, and texts in such documents cannot be serialized into the one-dimensional sequence without losing information. Classic information extraction models such as BiLSTM-CRF typically operate on text sequences and do not incorporate visual features. In this paper, we introduce a graph convolution based model to combine textual and visual information presented in VRDs. Graph embeddings are trained to summarize the context of a text segment in the document, and further combined with text embeddings for entity extraction. Extensive experiments have been conducted to show that our method outperforms BiLSTM-CRF baselines by significant margins, on two real-world datasets. Additionally, ablation studies are also performed to evaluate the effectiveness of each component of our model.