净分析信号算法用于近红外模型优化的研究

将净分析信号概念引入多元复杂的银杏叶提取物分析中,利用净分析信号算法对近红外光谱图进行处理,并用净分析算法结合二维相关光谱算法对模型进行了优化,最终建立槲皮素、山柰素和异鼠李素三种组分同时测定的近红外定量模型。利用两种净分析信号算法(Lorber的NAS算法和Goicoechea 与 Olivieri的HLA/GO算法)作为光谱预处理方法,用一系列主因子重构后的光谱图建立校正模型,并通过净分析信号-二维相关光谱算法对图谱中主对角线处的波段进行选择,优化模型建立所采用的波段。所采用的两种基于净分析信号的预处理方法成功的应用于银杏叶中三种黄酮：槲皮素,山奈素和异鼠李素含量的同时测定。净分析信号预处理方法显著降低了建模所需主因子数,提高模型的稳健性。净分析信号-二维相关光谱算法能够优化建模波段,提高模型的预测能力。结论：净分析信号算法能够应用于近红外光谱模型预处理和优化过程,在复杂多组分且干扰未知植物药含量分析的研究中具有实际应用价值。     

Vibrational properties of graphene and graphene layers

The phonon dispersions of graphene and graphene layers are theoretically investigated within fifth‐nearest‐neighbor force‐constant approach. Based on their symmetry groups, the number of Raman‐ and infrared‐active modes at the Γ point is given. Interatomic force constants are recalculated by fitting them to experimental phonon energy dispersion curves. Wavenumbers of optically active modes are presented as a function of number of layers (n). Our calculated results reproduce well the experimental data of G peak for graphene (1587 cm⁻¹) and graphite (1581.6 cm⁻¹) and clearly give the relation that ω_G = 1581.6 + 11/(1 + n^1.6). Copyright © 2009 John Wiley & Sons, Ltd.     

Facile method for CLSM imaging unfunctionalized Au nanoparticles through fluorescent channels

The microscopic visualization of metal nanoparticles has become a useful tool for the investigation of their applications in cell labeling and the study of their bio-effects. In the current study, we have developed a facile method with confocal laser scanning microscope (CLSM) to observe unfunctionalized Au nanoparticles through fluorescent channels. The sharp reflected signal and photostable property of the metal nanoparticles makes the present method very ideal for fluorescent co-localization, real-time imaging, and further quantitative analysis. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users. Lan Yuan and Wei Wei contributed equally to this study.     

X-ray and DFT Study of Glaucocalyxin A Compound with Cytotoxic Activity

采用单晶X射线衍射技术对化合物的晶体结构进行了研究. 研究结果表明,Glaucocalyxin A分子含3个六元环和1个五元环,其中六元环均为椅式构象,五元环为扭曲信封式构象. Glaucocalyxin A晶体结构属正交晶系. 采用密度函数理论和Hartree-Fock方法对该化合物晶体的键长、键角和两面角进行计算,并与X衍射测定值进行比较,结果表明理论值与实验值符合相当好. 应用规范不变原子轨道,分别在不同水平上,计算了Glaucocalyxin A的1H和13C NMR化学位移,并对理论计算值的误差     

Extended Power Law and Hall Anomaly of High-Temperature Superconductors

Starting from the free energy of a moving vortex, we obtain the extended power law form of longitudinal resistivity pxx and the analytical form of Hall resistivity pxy.Based on them, we obtain a scaling relation which agrees well with the experimental data of different kinds of high-temperature superconductors. Furthermore, the theoretical results well fit the scaling results.     

Characterization of rapidly solidified powder of high-speed steel

Ionization and excitation cross sections as well as electron-energy spectra and stopping powers of the alkali metal atoms Li, Na, K, and Rb colliding with antiprotons were calculated using a time-dependent channel-coupling approach. An impact-energy range from 0.25 to 4000 keV was considered. The target atoms are treated as effective one-electron systems using a model potential. The results are compared with calculated cross sections for antiproton-hydrogen atom collisions.     

Sintering characteristics of LiZn ferrites fabricated by a sol-gel process

Ultra-fine powders of LiZn ferrites with composition of Li_0.4Zn_0.2Fe_2.4O₄ were synthesized by a sol-gel process. X-ray diffraction pattern reflected that the synthesized powders were single spinel structure, with crystallite size of about 35 nm. Then the powders were granulated, pressed and sintered at different temperatures. The sintered samples were investigated by means of characterizing microstructures and magnetic properties by scanning electron microscope and B-H analyzer, respectively. The results indicate that when compared with a traditional ceramic process, the sol-gel process could slightly bring down the sintering temperature of LiZn ferrite whereas the microstructures are not homogeneous in the sintered samples. The sintering mechanisms of LiZn ferrites sintered at 1360 °C were studied, which reveal that during sintering, solid mass transfer is dominant in the LiZn ferrites fabricated by a traditional ceramic process while in the gel-derived ferrites, gas mass transfer is dominant.     

多孔硅拉曼光谱随激发功率变化的研究

用阳极氧化法新制备了多孔硅样品,以457．5nm固体激光器为激发光源,在不同激发功率下,获得了拉曼谱图和一些谱峰参量随激光功率的变化关系。解释了520cm^-1和300cm^-1附近拉曼峰随功率变化的一系列可逆的实验现象：随激光功率升高出现的红移和非对称性展宽,主要是由于样品局域平均粒径变小而受量子限域效应的影响导致的;样品局域平均粒径在表面上的二维减小与随激光功率升高而导致的局域温升并不违背基本的热力学定律;高功率时520cm^-1附近双峰的出现是由于多孔硅样品局域平均粒径达到一定阈值而导致的纵模和横模双声子模的分裂。     

Coarse-Grained Pruning of Neural Network Models Based on Blocky Sparse Structure

Deep neural networks may achieve excellent performance in many research fields. However, many deep neural network models are over-parameterized. The computation of weight matrices often consumes a lot of time, which requires plenty of computing resources. In order to solve these problems, a novel block-based division method and a special coarse-grained block pruning strategy are proposed in this paper to simplify and compress the fully connected structure, and the pruned weight matrices with a blocky structure are then stored in the format of Block Sparse Row (BSR) to accelerate the calculation of the weight matrices. First, the weight matrices are divided into square sub-blocks based on spatial aggregation. Second, a coarse-grained block pruning procedure is utilized to scale down the model parameters. Finally, the BSR storage format, which is much more friendly to block sparse matrix storage and computation, is employed to store these pruned dense weight blocks to speed up the calculation. In the following experiments on MNIST and Fashion-MNIST datasets, the trend of accuracies with different pruning granularities and different sparsity is explored in order to analyze our method. The experimental results show that our coarse-grained block pruning method can compress the network and can reduce the computational cost without greatly degrading the classification accuracy. The experiment on the CIFAR-10 dataset shows that our block pruning strategy can combine well with the convolutional networks.     

基于Willgerodt-Kindler反应的两种二茂铁羧酸的合成及机理研究

以Na2S•9H2O为催化剂, 通过Willgerodt-Kindler反应分别合成了二茂铁乙酸和1,1’-二茂铁二乙酸, 用IR, 1H NMR, MS和元素分析对产物分别进行了结构表征, 并利用GC-MS研究了该反应机理, 推测反应是由多硫阴离子进攻羰基α碳所引发, 同时讨论了影响反应的各种因素.