Precise Extraction of Charge Carrier Mobility for Organic Transistors

Unreliable mobility values, and particularly greatly overestimated values and severely distorted temperature dependences, have recently hampered the development of the organic transistor field. Given that organic field‐effect transistors (OFETs) have been routinely used to evaluate mobility, precise parameter extraction using the electrical properties of OFETs is thus of primary importance. This review examines the origins of the various mobilities that must be determined for OFET applications, the relevant extraction methods, and the data selection limitations, which help in avoiding conceptual errors during mobility extraction. For increased precision, the review also discusses device fabrication considerations, calibration of both the specific gate‐dielectric capacitance and the threshold voltage, the contact effects, and the bias and temperature dependences, which must actually be handled with great care but have mostly been overlooked to date. This review serves as a systematic overview of the OFET mobility extraction process to ensure high precision and will also aid in improving future research.     

Two-step pyrolytic engineering to form porous nitrogen-rich carbons with a 3D network structure for Zn-air battery oxygen reduction electrocatalysis

It is of great urgency to design inexpensive and high-performance oxygen reduction reaction (ORR) electrocatalysts derived from biowastes as substitutes for Pt-based materials in electrochemical energy-conversion devices. Here we propose a strategy to synthesize three-dimensional (3D) porous nitrogen-doped network carbons to catalyze the ORR from two-step pyrolysis engineering of biowaste scale combined with the use of a ZnCl₂ activator and a FeCl₂ promotor. Electrochemical tests show that the synthesized network carbons have exhibited comparable ORR catalytic activity with a half-wave potential (~0.85 V vs. RHE) and outstanding cyclical stability in comparison to the Pt/C catalyst. Beyond that, a high electron transfer number (~3.8) and a low peroxide yield (<7.6%) can be obtained, indicating a four-electron reaction pathway. The maximum power density is ~68 mW cm^?2, but continuous discharge curves (at a constant potential of ~1.30 V) for 12 h are not obviously declined in Zn-air battery tests using synthesized network carbons as the cathodic catalyst. The formation of 3D porous structures with high BET surface area can effectively expose the surface catalytic sites and promote mass transportation to boost the ORR activity. This work may open a new idea to prepare porous carbon-based catalysts for some important reactions in new energy devices.     

Disease Ionomics: Understanding the Role of Ions in Complex Disease

Ionomics is a novel multidisciplinary field that uses advanced techniques to investigate the composition and distribution of all minerals and trace elements in a living organism and their variations under diverse physiological and pathological conditions. It involves both high-throughput elemental profiling technologies and bioinformatic methods, providing opportunities to study the molecular mechanism underlying the metabolism, homeostasis, and cross-talk of these elements. While much effort has been made in exploring the ionomic traits relating to plant physiology and nutrition, the use of ionomics in the research of serious diseases is still in progress. In recent years, a number of ionomic studies have been carried out for a variety of complex diseases, which offer theoretical and practical insights into the etiology, early diagnosis, prognosis, and therapy of them. This review aims to give an overview of recent applications of ionomics in the study of complex diseases and discuss the latest advances and future trends in this area. Overall, disease ionomics may provide substantial information for systematic understanding of the properties of the elements and the dynamic network of elements involved in the onset and development of diseases.     

改进HSR-FCN的服装图像识别分类算法研究

目前网络上的服装图像数量增长迅猛，对于大量服装图像实现智能分类的需求日益增加。将基于区域的全卷积网络（Region-Based Fully Convolutional Networks，R-FCN）引入到服装图像识别中，针对服装图像分类中网络训练时间长、形变服装图像识别率低的问题，提出一种新颖的改进框架HSR-FCN。新框架将R-FCN中的区域建议网络和HyperNet网络相融合，改变图片特征学习方式，使得HSR-FCN可以在更短的训练时间内达到更高的准确率。在模型中引入了空间转换网络，对输入服装图像和特征图进行了空间变换及对齐，加强了对多角度服装和形变服装的特征学习。实验结果表明，改进后的HSR-FCN模型有效地加强了对形变服装图像的学习，且在训练时间更短的情况下，比原来的网络模型R-FCN平均准确率提高了大约3个百分点，达到96.69%。