Fabrication and Characterization of Ni Thin Films Using Direct-Current Magnetron Sputtering

Ni films are deposited by using ultra high vacuum dc magnetron sputtering onto silicon substrates at room temperature, and the high-quality and high-density films are prepared. The parameters, such as thickness, density and surface roughness, are obtained by using small-angle x-ray diffraction (XRD) analyses with the Marquardt gradient-expansion algorithm. The deposition rate is calculated and the Ni single layer can be fabricated precisely. Based on the fitting results, we can find that the surface roughness of the Ni films is about 0.7nm, the densities of Ni films are around 97% and the deposition rate is 0.26nm/s. The roughness of the surface is also characterized by using an atomic force microscope (AFM). The changing trend of the surface roughness in the simulation of XRD is in good agreement with the AFM measurement.     

软X射线多层膜透射式起偏器和检偏器设计

叙述了软X射线波段多层膜透射式起偏器和检偏器的设计原理，提出了设计准则，通过该准则设计的多层膜偏振元件能够同时满足具有较大的光通量和偏振度要求．同时还优化了不同膜厚比情况下软X射线多层膜偏振元件的性能，讨论了在多层膜制作过程中，由于表界面粗糙度和扩散的不同对元件性能的影响，这些工作为制作软X射线多层膜透射式起偏器和检偏器提供了设计基础。     

多时变时滞细胞神经网络的全局指数稳定性

针对一类多时变时滞细胞神经网络,利用Young不等式和Halanay不等式技术,给出了保证平衡点惟一性和全局指数稳定性的几个充分判据.所得到的全局指数稳定判据完全独立于时滞,不要求时变时滞的可微性和神经元激励函数的严格单调性,且通过几个注释说明本文的结果改进和扩展了现有一些文献中的结果.仿真例子证明了本文结果的有效性.     

一类时滞混沌神经网络的全局渐近同步

应用驱动—响应同步方法,研究了一类时滞混沌神经网络的全局渐近同步问题.基于分散控制策略,通过构造适当的Lyapunov-Krasovskii泛函,给出了保证两个具有相同结构但初始条件不相同的时滞混沌神经网络全局渐近同步的控制律设计方法.所得到的控制律不仅易于实现,而且具有高可靠性等特点,克服了传统的集中控制的不足.仿真示例验证了该方法的有效性.     

北京同步辐射装置3W1B软X射线光束线偏振特性测量

应用自行研制的Mo/Si周期多层膜作为起偏器和检偏器的光学元件, 测量了北京同步辐射装置3W1B束线的偏振状态. 通过数据分析, 得到了3W1B软X射线的有关偏振参数, 在86—89eV能区经过起偏器后的偏振度超过98%, 圆偏振分量介于1%—3%之间.     

Multilayers with high-reflectivity at 19.5 nm and low-reflectivity at 30.4 nm

The multilayer （ML） mirror with high-reflectivity （HR） at a specific emission line of 19.5 nm （Fe line） and low-reflectivity （LR） at 30.4 nm （He line） is needed to be designed and fabricated for observing the image of sun. Based on a variety of optimizations utilized different structures, the design is performed and the final results demonstrate that the reflectivity at 30.4 nm does not achieve minimum value when the reflectivity at 19.5 nm reaches the maximum value. The tradeoff should be done between the HR at 19.5 nm and LR at 30.4 nm. One optimized mirror is fabricated by direct current magnetron sputtering and characterized by grazing-incident X-ray diffraction （XRD） and synchrotron radiation （SR）. The experimental results demonstrate that the ML achieves the reflectivity of 33.3% at 19.5 nm and of 9.6× 10-4 at 30.4 nm at the incident angle of 13°.     

B_4C/Mo/Si high reflectivity multilayer mirror at 30.4 nm

The B4C/Mo/Si high reflectivity multilayer mirror was designed for He-Ⅱradiation (30.4 nm) using the layer-by-layer method. The theoretical peak reflectivity was up to 38.2% at the incident angle of 5℃. The B4C/Mo/Si multilayer was fabricated by direct current magnetron sputtering and measured at the National Synchrotron Radiation Laboratory (NSRL) of China. The experimental reflectivity of the B4C/Mo/Si multilayer at 30.4 nm was about 32.5%. The promising performances of the B4C/Mo/Si multilayer mirror could be used for the construction of solar physics instrumentation.     

激光驻波场中钠原子沉积图样的理论研究

原子在近共振激光驻波场中受光的作用力汇聚沉积，形成以激光半波长为周期的光栅结构. 利用半经典模型对钠原子束在不饱和驻波场中的运动行为进行了理论研究，用数值方法模拟出在考虑横向温度、纵向速度和自由飞程后钠原子沉积的光栅结构，通过半高宽、对比度对最终沉积图样进行表征，分析了这些因素对沉积的影响，为正在开展的原子光刻实验提供了必要的理论依据. 关键词： 原子沉积 半经典模型 激光驻波场     

一类混沌神经网络的全局同步

研究了一类时滞混沌神经网络的全局同步问题.应用驱动-响应同步方法和线性矩阵不等式技术，给出了时滞混沌神经网络全局同步的充分条件和同步控制器设计方法，而且所得到的控制器易于实现.仿真示例验证了本文方法的有效性. 关键词： 混沌神经网络 同步 驱动-响应法 线性矩阵不等式     

Detection of subsurface defects of fused silica optics by confocal scattering microscopy

<正>A non-destructive technique for subsurface measurements is proposed by combining light scattering method with laser confocal scanning tomography.The depth and distribution of subsurface defect layers are represented in term of scattered light intensity pattern,and three types of fused silica specimens are fabricated by different grinding and polishing processes to verify the validity and effectiveness.By using the direct measurement method with such technique,micron-scale cracks and scratches can be easily distinguished,and the instructional subsurface defect depths of 55,15,and 4μm are given in real time allowing for an in-process observation and detection.