Li-ion batteries(LIBs)have demonstrated great promise in electric vehicles and hybrid electric vehicles.However,commercial graphite materials,the current predominant anodes in LIBs,have a low theoretical capacity of o...Li-ion batteries(LIBs)have demonstrated great promise in electric vehicles and hybrid electric vehicles.However,commercial graphite materials,the current predominant anodes in LIBs,have a low theoretical capacity of only 372 mAh·g?1,which cannot meet the everincreasing demand of LIBs for high energy density.Nanoscale Si is considered an ideal form of Si for the fabrication of LIB anodes as Si–C composites.Synthesis of nanoscale Si in a facile,cost-effective way,however,still poses a great challenge.In this work,nanoscale Si was prepared by a controlled magnesiothermic reaction using diatomite as the Si source.It was found that the nanoscale Si prepared under optimized conditions(800°C,10 h)can deliver a high initial specific capacity(3053 mAh·g?1 on discharge,2519 mAh·g?1 on charge)with a high first coulombic efficiency(82.5%).When using sand-milled diatomite as a precursor,the obtained nanoscale Si exhibited a well-dispersed morphology and had a higher first coulombic efficiency(85.6%).The Si–C(Si:graphite=1:7 in weight)composite using Si from the sand-milled diatomite demonstrated a high specific capacity(over 700 mAh·g?1 at 100 mA·g?1),good rate capability(587 mAh·g?1 at 500 mA·g?1),and a long cycle life(480 mAh·g?1 after 200 cycles at 500 mA·g?1).This work gives a facile method to synthesize nanoscale Si with both high capacity and high first coulombic efficiency.展开更多
随着互联网技术的迅速发展和普及,越来越多的用户开始通过社会网络进行各种信息的分享与交流。网络中同一用户可能申请多个不同账号进行信息发布,这些账号构成了网络中的关联用户。准确、有效地挖掘社会网络中的关联用户能够抑制网络中...随着互联网技术的迅速发展和普及,越来越多的用户开始通过社会网络进行各种信息的分享与交流。网络中同一用户可能申请多个不同账号进行信息发布,这些账号构成了网络中的关联用户。准确、有效地挖掘社会网络中的关联用户能够抑制网络中的虚假信息和不法行为,从而保证网络环境的安全性和公平性。现有的关联用户挖掘方法仅考虑了用户属性或用户关系信息,未对网络中含有的多类信息进行有效融合以及综合考虑。此外,大多数方法借鉴其他领域的方法进行研究,如去匿名化问题,这些方法不能准确解决关联用户挖掘问题。为此,文中针对网络关联用户挖掘问题,提出了基于多信息融合表示学习的关联用户挖掘算法(Associated Users Mining Algorithm based on Multi-information fusion Representation Learning,AUMA-MRL)。该算法使用网络表示学习的思想对网络中多种不同维度的信息(如用户属性、网络拓扑结构等)进行学习,并将学习得到的表示进行有效融合,从而得到多信息融合的节点嵌入。这些嵌入可以准确表征网络中的多类信息,基于习得的节点嵌入构造相似性向量,从而对网络中的关联用户进行挖掘。文中基于3个真实网络数据对所提算法进行验证,实验网络数据包括蛋白质网络PPI以及社交网络Flickr和Facebook,使用关联用户挖掘结果的精度和召回率作为性能评价指标对所提算法进行有效性验证。结果表明,与现有经典算法相比,所提算法的召回率平均提高了17.5%,能够对网络中的关联用户进行有效挖掘。展开更多
基金the National Natural Science Foundation of China(No.51572238)Zhejiang Provincial Natural Science Foundation(No.LY19E020013)the Joint Research Project of Zhejiang University with Zotye Automobile Corporation Limited on Si-Based Anode Materials(No.P-ZH-2018-003).
文摘Li-ion batteries(LIBs)have demonstrated great promise in electric vehicles and hybrid electric vehicles.However,commercial graphite materials,the current predominant anodes in LIBs,have a low theoretical capacity of only 372 mAh·g?1,which cannot meet the everincreasing demand of LIBs for high energy density.Nanoscale Si is considered an ideal form of Si for the fabrication of LIB anodes as Si–C composites.Synthesis of nanoscale Si in a facile,cost-effective way,however,still poses a great challenge.In this work,nanoscale Si was prepared by a controlled magnesiothermic reaction using diatomite as the Si source.It was found that the nanoscale Si prepared under optimized conditions(800°C,10 h)can deliver a high initial specific capacity(3053 mAh·g?1 on discharge,2519 mAh·g?1 on charge)with a high first coulombic efficiency(82.5%).When using sand-milled diatomite as a precursor,the obtained nanoscale Si exhibited a well-dispersed morphology and had a higher first coulombic efficiency(85.6%).The Si–C(Si:graphite=1:7 in weight)composite using Si from the sand-milled diatomite demonstrated a high specific capacity(over 700 mAh·g?1 at 100 mA·g?1),good rate capability(587 mAh·g?1 at 500 mA·g?1),and a long cycle life(480 mAh·g?1 after 200 cycles at 500 mA·g?1).This work gives a facile method to synthesize nanoscale Si with both high capacity and high first coulombic efficiency.
文摘随着互联网技术的迅速发展和普及,越来越多的用户开始通过社会网络进行各种信息的分享与交流。网络中同一用户可能申请多个不同账号进行信息发布,这些账号构成了网络中的关联用户。准确、有效地挖掘社会网络中的关联用户能够抑制网络中的虚假信息和不法行为,从而保证网络环境的安全性和公平性。现有的关联用户挖掘方法仅考虑了用户属性或用户关系信息,未对网络中含有的多类信息进行有效融合以及综合考虑。此外,大多数方法借鉴其他领域的方法进行研究,如去匿名化问题,这些方法不能准确解决关联用户挖掘问题。为此,文中针对网络关联用户挖掘问题,提出了基于多信息融合表示学习的关联用户挖掘算法(Associated Users Mining Algorithm based on Multi-information fusion Representation Learning,AUMA-MRL)。该算法使用网络表示学习的思想对网络中多种不同维度的信息(如用户属性、网络拓扑结构等)进行学习,并将学习得到的表示进行有效融合,从而得到多信息融合的节点嵌入。这些嵌入可以准确表征网络中的多类信息,基于习得的节点嵌入构造相似性向量,从而对网络中的关联用户进行挖掘。文中基于3个真实网络数据对所提算法进行验证,实验网络数据包括蛋白质网络PPI以及社交网络Flickr和Facebook,使用关联用户挖掘结果的精度和召回率作为性能评价指标对所提算法进行有效性验证。结果表明,与现有经典算法相比,所提算法的召回率平均提高了17.5%,能够对网络中的关联用户进行有效挖掘。
