Henan Normal University

About: Henan Normal University is a education organization based out in Xinxiang, China. It is known for research contribution in the topics: Catalysis & Ionic liquid. The organization has 10863 authors who have published 11077 publications receiving 166773 citations.

Toward Solution Syntheses of the Tetrahedral Au20 Pyramid and Atomically Precise Gold Nanoclusters with Uncoordinated Sites

Abstract: ConspectusA long-standing objective of cluster science is to discover highly stable clusters and to use them as models for catalysts and building blocks for cluster-assembled materials. The discovery of catalytic properties of gold nanoparticles (AuNPs) has stimulated wide interests in gaseous size-selected gold clusters. Ligand-protected AuNPs have also been extensively investigated to probe their size-dependent catalytic and optical properties. However, the need to remove ligands can introduce uncertainties in both the structures and sizes of ligand-protected AuNPs for catalytic applications. Ideal model catalysts should be atomically precise AuNPs with well-defined structures and uncoordinated surface sites as in situ active centers. The tetrahedral (Td) Au20 pyramidal cluster, discovered to be highly stable in the gas phase, provided a unique opportunity for such an ideal model system. The Td-Au20 consists of four Au(111) faces with all its atoms on the surface. Bulk synthesis of Td-Au20 with appropri...

Unveiling lasing mechanism in CsPbBr3 microsphere cavities

Abstract: Recently, the light-matter interaction of perovskite microcavities has been widely explored for its great potential in low-threshold lasing devices. However, the mechanism of perovskite lasing remains unclear to date. In this study, we demonstrated high-quality single-mode excitonic lasing in CsPbBr3 microspheres, providing an ideal platform to study the underlying physics of lasing behavior. We show that the lasing mechanism shifts from the exciton-exciton scattering to the exciton-phonon scattering with the increase in temperature from 77 to 300 K, which was verified by temperature-dependent photoluminescence (PL), time-resolved photoluminescence (TRPL) as well as temperature-dependent Raman spectroscopy. Furthermore, by analyzing PL line width broadening with varied temperatures, we found that two different phonon modes were involved in the exciton-phonon scattering process. The scattering from the low-energy phonon (∼8.6 meV) is the dominant source of exciton-phonon coupling in the intermediate temperature range (77 to 230 K), while the high-energy phonon (∼15.3 meV) dominates from 230 K to room temperature. These results confirm the lasing mechanism in such perovskite-based micro/nano-cavities and significantly influence the development of future low-threshold lasers.

Improve Transformer Models with Better Relative Position Embeddings

Abstract: The transformer model has demonstrated superior results on NLP tasks including machine translation and question answering. In this paper, we argue that the position information is not fully utilized in existing work. For example, the initial proposal of a sinusoid embedding is fixed and not learnable. In this paper, we first review the absolute position embeddings and existing relative position embedding methods. We then propose new methods to encourage increased interaction between query, key and relative position embeddings in the self-attention mechanism. Our most promising approach is a generalization of the absolute position embedding. Our method results in increased accuracy compared to previous approaches in absolute and relative position embeddings on the SQuAD1.1 dataset. In addition, we address the inductive property of whether a position embedding can be robust enough to handle long sequences. We demonstrate empirically that our relative embedding method can be reasonably generalized to and is robust in the inductive perspective. Finally, we show that our proposed method can be effectively and efficiently adopted as a near drop-in replacement for improving the accuracy of large models with little computational overhead.