The long-range periodically ordered atomic structures in intermetallic nanoparticles(INPs)can significantly enhance both the electrocatalytic activity and electrochemical stability toward the oxygen reduction reaction...The long-range periodically ordered atomic structures in intermetallic nanoparticles(INPs)can significantly enhance both the electrocatalytic activity and electrochemical stability toward the oxygen reduction reaction(ORR)compared to the disordered atomic structures in ordinary solid-solution alloy NPs.Accordingly,through a facile and scalable synthetic method,a series of carbon-supported ultrafine Pt_3Co_(x)Mn_(1-x)ternary INPs are prepared in this work,which possess the"skin-like"ultrathin Pt shells,the ordered L1_(2) atomic structure,and the high-even dispersion on supports(L1_(2)-Pt_3Co_(x)Mn_(1-x)/~SPt INPs/C).Electrochemical results present that the composition-optimized L1_(2)-Pt_3Co_(0.7)Mn_(0.3)/~SPt INPs/C exhibits the highest electrocata lytic activity among the series,which are also much better than those of the pristine ultrafine Pt/C.Besides,it also has a greatly enhanced electrochemical stability.In addition,the effects of annealing temperature and time are further investigated.More importantly,such superior ORR electrocatalytic performance of L1_(2)-Pt_3Co_(0.7)Mn_(0.3)/~SPt INPs/C are also well demonstrated in practical fuel cells.Physicochemical characterization analyses further reveal the major origins of the greatly enhanced ORR electrocata lytic performance:the Pt-Co-Mn alloy-induced geometric and ligand effects as well as the extremely high L1_(2) atomic-ordering degree.This work not only successfully develops a highly active and stable ordered ternary intermetallic ORR electrocatalyst,but also elucidates the corresponding"structure-function"relationship,which can be further applied in designing other intermetallic(electro)catalysts.展开更多
加性分位数回归为非线性关系的建模提供一种灵活、鲁棒的方法.拟合加性分位数模型的方法通常使用样条函数逼近分量,但需要先验的选择节点,计算速度较慢,并不适合大规模数据问题.因此文中提出基于融合Lasso的非参数加性分位数回归模型(No...加性分位数回归为非线性关系的建模提供一种灵活、鲁棒的方法.拟合加性分位数模型的方法通常使用样条函数逼近分量,但需要先验的选择节点,计算速度较慢,并不适合大规模数据问题.因此文中提出基于融合Lasso的非参数加性分位数回归模型(Nonparametric Additive Quantile Regression Model Based on Fused Lasso,AQFL),是在融合Lasso罚和l_(2)罚之间折衷的可对加性分位数回归模型进行估计和变量选择的模型.融合Lasso罚使模型能快速计算,并在局部进行自适应,从而实现对所需分位数甚至极端分位数的预测.同时结合l_(2)罚,在高维数据中将对响应影响较小的协变量函数值压缩为零,实现变量的选择.此外,文中给出保证收敛到全局最优的块坐标ADMM算法(Block Coordinate Alternating Direction Method of Multipliers,BC-ADMM),证明AQFL的预测一致性.在合成数据和碎猪肉数据上的实验表明AQFL在预测准确性和鲁棒性等方面较优.展开更多
基金supported by the National Key Research and Development Program of China(2021YFB4001301)the Science and Technology Commission of Shanghai Municipality(21DZ1208600)the Oceanic Interdisciplinary Program of Shanghai Jiao Tong University(SL2021ZD105)。
文摘The long-range periodically ordered atomic structures in intermetallic nanoparticles(INPs)can significantly enhance both the electrocatalytic activity and electrochemical stability toward the oxygen reduction reaction(ORR)compared to the disordered atomic structures in ordinary solid-solution alloy NPs.Accordingly,through a facile and scalable synthetic method,a series of carbon-supported ultrafine Pt_3Co_(x)Mn_(1-x)ternary INPs are prepared in this work,which possess the"skin-like"ultrathin Pt shells,the ordered L1_(2) atomic structure,and the high-even dispersion on supports(L1_(2)-Pt_3Co_(x)Mn_(1-x)/~SPt INPs/C).Electrochemical results present that the composition-optimized L1_(2)-Pt_3Co_(0.7)Mn_(0.3)/~SPt INPs/C exhibits the highest electrocata lytic activity among the series,which are also much better than those of the pristine ultrafine Pt/C.Besides,it also has a greatly enhanced electrochemical stability.In addition,the effects of annealing temperature and time are further investigated.More importantly,such superior ORR electrocatalytic performance of L1_(2)-Pt_3Co_(0.7)Mn_(0.3)/~SPt INPs/C are also well demonstrated in practical fuel cells.Physicochemical characterization analyses further reveal the major origins of the greatly enhanced ORR electrocata lytic performance:the Pt-Co-Mn alloy-induced geometric and ligand effects as well as the extremely high L1_(2) atomic-ordering degree.This work not only successfully develops a highly active and stable ordered ternary intermetallic ORR electrocatalyst,but also elucidates the corresponding"structure-function"relationship,which can be further applied in designing other intermetallic(electro)catalysts.
文摘列表L(2,1)-标号是一个重要的可以应用到信道分配问题中的优化问题,k-L(2,1)-标号是指对于一个平面图G满足映射ϕ :V (G)→{0,1,…,k},使得若d(u,v)=1,则|ϕ(u)−ϕ(v)|≥2;若d(u,v)=2,则|ϕ(u)−ϕ(v)|≥1,其中d(u,v)是图中点u和点v之间的距离。记λ(2,1)l(G)=min{k|G有一列k-L(2,1)-标号}是列表L(2,1)-标号数。在2018年,Zhu and Bu等人在全局最优化杂志中得出这样一个结论:对于不含4-圈和6-圈的平面图G有λ(2,1)l(G)≤max{Δ+15,29}。本文改进了这个结论的上界λ(2,1)l(G)≤max{Δ+12,24}。
文摘加性分位数回归为非线性关系的建模提供一种灵活、鲁棒的方法.拟合加性分位数模型的方法通常使用样条函数逼近分量,但需要先验的选择节点,计算速度较慢,并不适合大规模数据问题.因此文中提出基于融合Lasso的非参数加性分位数回归模型(Nonparametric Additive Quantile Regression Model Based on Fused Lasso,AQFL),是在融合Lasso罚和l_(2)罚之间折衷的可对加性分位数回归模型进行估计和变量选择的模型.融合Lasso罚使模型能快速计算,并在局部进行自适应,从而实现对所需分位数甚至极端分位数的预测.同时结合l_(2)罚,在高维数据中将对响应影响较小的协变量函数值压缩为零,实现变量的选择.此外,文中给出保证收敛到全局最优的块坐标ADMM算法(Block Coordinate Alternating Direction Method of Multipliers,BC-ADMM),证明AQFL的预测一致性.在合成数据和碎猪肉数据上的实验表明AQFL在预测准确性和鲁棒性等方面较优.
