In this work,silicon@reduced graphene oxide/pyrolytic carbon nanofibers(Si@RGO/C NFs) composite with double modified layer is prepared through electrospinning,stabilization and carbonization.In this composite,polyethylene oxide-polypropylene oxide-polyethylene oxide(P123,a non-ionic surfactant) is introduced as the dispersant,which can make silicon nanoparticles evenly dispersed in electrospinning solution to prevent it from agglomeration.Graphene modified layer can buffer the volumetric expansion of silicon nanoparticles,prevent direct contact between silicon and electrolyte as well as enhance the electrical conductivity.Moreover,carbon fibers synthesized by electrospinning can encapsulate silicon@graphene composite internally to form a double modified layer.This composite with double modified layer can further alleviate the volume change of silicon nanoparticles and avoid direct contact between silicon and electrolyte to form a stable interface.Owing to the above-mentioned merits,the Si@RGO/C NFs composite exhibits excellent cyclic stability and superior rate performance.Particularly,it maintains a specific capacity of 929 mA h g~(-1) with the retention ratio of 83.1% after 100 cycles at 0.5 A g~(-1) and delivers an outstanding rate capability of 1003 mA h g~(-1) at 2 A g~(-1). 相似文献
非平稳多变量时间序列(non-stationary multivariate time series, NSMTS)预测目前仍是一个具有挑战性的任务.基于循环神经网络的深度学习模型,尤其是基于长短期记忆(long short-term memory, LSTM)和门循环单元(gated recurrent unit, GRU)的神经网络已获得了令人印象深刻的预测性能.尽管LSTM结构上较为复杂,却并不总是在性能上占优.最近提出的最小门单元(minimal gated unit, MGU)神经网络具有更简单的结构,并在图像处理和一些序列处理问题中能够提升训练效率.更为关键的是,实验中我们发现该门单元可以高效运用于NSMTS的预测,并达到了与基于LSTM和GRU的神经网络相当的预测性能.然而,基于这3类门单元的神经网络中,没有任何一类总能保证性能上的优势.为此提出了一种线性混合门单元(MIX gated unit, MIXGU),试图利用该单元动态调整GRU和MGU的混合权重,以便在训练期间为网络中的每个MIXGU获得更优的混合结构.实验结果表明,与基于单一门单元的神经网络相比,混合2类门单元的MIXGU神经网络具有更优的预测性能. 相似文献
In order to improve the rate capability and mitigate the voltage decay of lithium rich layered oxides, chromium doped cathode materials Li1.2[Mn0.54Ni0.13Co0.13]1?xCrxO2 (x = 0,0.003,0.005,0.007) is synthesized via co-precipitation method followed by calcination. As a result, Cr-doped samples present shortened voltage platform at 4.5 V, indicating that the escape of lattice oxygen is suppressed by Cr doping. Among all samples, Li1.2[Mn0.54Ni0.13Co0.13]1?xCrxO2 (x = 0.005) sample shows the greatest rate capability (119.3 mAh g?1 at 10 C) and the minimum voltage drop (0.6167 V) after 200 cycles at 1.0 C. The reasons for the improved electrochemical performance are the enhanced structural stability, the stronger CrO bond than MnO band and the decreased metal–oxygen covalency induced by Cr doping. 相似文献
Photodynamic therapy (PDT) is a clinical treatment for cancer or non-neoplastic diseases, and the photosensitizers (PSs) are crucial for PDT efficiency. The commonly used chemical PSs, generally produce ROS through the type II reaction that highly relies on the local oxygen concentration. However, the hypoxic tumor microenvironment and unavoidable dark toxicity of PSs greatly restrain the wide application of PDT. The genetically encoded PSs, unlike chemical PSs, can be modified using genetic engineering techniques and targeted to unique cellular compartments, even within a single cell. KillerRed, as a dimeric red fluorescent protein, can be activated by visible light or upconversion luminescence to execute the Type I reaction of PDT, which does not need too much oxygen and surely attract the researchers’ focus. In particular, nanotechnology provides new opportunities for various modifications of KillerRed and versatile delivery strategies. This review more comprehensively outlines the applications of KillerRed, highlighting the fascinating features of KillerRed genes and proteins in the photodynamic systems. Furthermore, the advantages and defects of KillerRed are also discussed, either alone or in combination with other therapies. These overviews may facilitate understanding KillerRed progress in PDT and suggest some emerging potentials to circumvent challenges to improve the efficiency and accuracy of PDT. 相似文献
Adjustable fluidic structures play an important role in microfluidic systems. Fracture of multilayered materials under applied tension has been previously demonstrated as a convenient, simple, and inexpensive approach to fabricate nanoscale adjustable structures; here, it is demonstrated how to extend this concept to the microscale. This is achieved by a novel pairing of materials that leverages fracture mechanics to limit crack formation to a specified region, allowing to create size‐controllable and adjustable microfluidic structures. This technique can be used to fabricate “normally closed” microfluidic channels that are completely reversible, a feature that is challenging to achieve in conventional systems without careful engineering controls. The adjustable microfluidic channels are then applied to mechanically lyse single cells, and subsequently manipulate the released nuclear chromatin, creating new possibilities for epigenetic analysis of single cells. This simple, versatile, and robust technology provides an easily accessible pathway to construct adjustable microfluidic structures, which will be useful in developing complex assays and experiments even in resource‐limited settings. 相似文献