Green Synthesis Strategy of Template-Free S-Doped g-C3N4 with Cystine for Efficient Photocatalytic Degradation of RhB.

Constructing a nanostructure with a high surface area and regulating the band gap by nonmetallic doping are two effective methods for improving the photocatalytic activity of catalysts. A green template-free synthesis strategy of S-doped g-C3N4 nanosheets is proposed via doping cystine as both the structural additive and S source. The features of S-doped samples (GCN-x%) were systematically studied, including morphology and textural and photoelectric properties, which demonstrated that the introduction of cystine and simple manipulation of the preparation process could realize self-exfoliation of g-C3N4 into nanosheets. The GCN-3% sample showed a surface area (131.88 m2·g-1) 10.7 times enlarged compared with bulk g-C3N4 (bulk-phase carbon nitride). Obvious redshift on the absorption edge induced by S doping can be observed, revealing a narrowed band gap and enhanced efficiency of photogenerated charge carrier separation. The DFT calculation results also verified that the introduced C-S site could lead to polarization of the local electric field and thus decrease the bandgap of g-C3N4 nanosheets. GCN-3% showed a 99.3% photocatalytic degradation ratio of rhodamine B in 60 min at a rate of 0.17 min-1. By scavengers experiment revealed that superoxide anion (·O2-) radicals and holes (h+) were vital active components during the photocatalytic degradation.

Incorporating multi-level CNN and attention mechanism for Chinese clinical named entity recognition.

Named entity recognition (NER) is a fundamental task in Chinese natural language processing (NLP) tasks. Recently, Chinese clinical NER has also attracted continuous research attention because it is an essential preparation for clinical data mining. The prevailing deep learning method for Chinese clinical NER is based on long short-term memory (LSTM) network. However, the recurrent structure of LSTM makes it difficult to utilize GPU parallelism which to some extent lowers the efficiency of models. Besides, when the sentence is long, LSTM can hardly capture global context information. To address these issues, we propose a novel and efficient model completely based on convolutional neural network (CNN) which can fully utilize GPU parallelism to improve model efficiency. Moreover, we construct multi-level CNN to capture short-term and long-term context information. We also design a simple attention mechanism to obtain global context information which is conductive to improving model performance in sequence labeling tasks. Besides, a data augmentation method is proposed to expand the data volume and try to explore more semantic information. Extensive experiments show that our model achieves competitive performance with higher efficiency compared with other remarkable clinical NER models.