基于多项式求根的双厚度透射法确定液体光学常数

光谱反演法可实现液体光学常数(消光系数和折射率)的确定,其中双厚度透射法是有代表性的光谱反演法之一。由于液体自身无法形成确定的形状,需要盛放在透明容器(液体池)中,因而通过实验测量得到的光谱透射率包含了液体池光学常数的影响,这使得基于双厚度透射法所建立的光谱透射率方程极为复杂,难以求得解析解,通常采用反演法来计算液体的光学常数。现有的反演法存在如下问题:一是反演迭代耗费时间,二是反演迭代会引入误差,三是反演法得到的液体折射率存在二值问题。为解决上述问题,基于三层介质结构(液体池),考虑光在两种介质分界面上的多次反射,建立液体厚度满足整数比的光谱透射率方程组。通过代数运算获得与消光系数有关的多项式方程,求解并选择大于0小于1的实数根来计算消光系数;另求解关于液体池光学窗口反射率的一元二次方程,选择大于0小于1的根来计算液体和容器分界面的反射率,进而计算得到液体折射率的两个值。再用另一材料制作的液体池测量液体的光谱透射率,然后结合已经得到的消光系数作相关计算,会另外得到液体折射率的两个值,从四个值中选择相同者即为液体的折射率。作为应用示例,选用文献中水在0.5~1.0μm光谱范围的光学常数作为“理论值”、光学常数已知的石英玻璃和有机玻璃作为液体池材料。在不考虑仪器测量误差的情况下,将上述文献数据代入光谱透射率方程,计算得到的透射率作为“实验数据”。然后用多项式求根的方法确定水的光学常数,所得结果与“理论值”完全符合。模拟过程和计算结果表明:新方法是可用的,该方法解决了反演法的计算耗时、迭代误差以及折射率的二值问题,为液体光学常数的确定提供了一个新选择。

基于多项式求根的双厚度透射率模型确定光学常数

为解决光谱反演法确定物质光学常数的一些问题,基于传统的双厚度透射率模型,建立厚度分别为L和2L的光谱透射率方程,通过代数运算获得与衰减系数有关的八次多项式方程,求解并选择其大于0小于1的实数根来计算衰减系数和消光系数;再求解关于界面反射率的一元二次方程,选择其大于0小于1的根来计算折射率。在确定光学常数的过程中,新方法没有反演误差和迭代计算耗时问题。利用已知文献中庚烷的光学常数验证新方法的可靠性,并分析了双厚度不满足2倍关系时对计算结果的影响,结论是第二厚度2L的相对误差不超过1%时,消光系数的计算误差不超过2.03%,不考虑3个强吸收点时,折射率的计算误差不超过1%。

基于多项式求根的双厚度透射率模型确定透明固体光学常数

为解决光谱反演法确定透明固体光学常数的一些问题,如存在反演误差、计算耗时等。本文基于传统的双厚度透射率模型,建立了厚度满足整数比的两个光谱透射率方程。通过代数运算获得了与消光系数有关的多项式方程,求解并选择大于0小于1的实数根来计算消光系数;然后求解关于界面反射率的一元二次方程,选择大于0小于1的根来计算折射率。在确定光学常数的过程中,新方法没有反演误差、迭代计算耗时及多值问题。作为应用示例,利用已知文献中的双厚度透射率实验数据计算了CaF_(2)和Si的光学常数,并和文献的结果进行了比较。结果表明,新方法优于传统的光谱反演法,新方法为透明固体光学常数的高精度确定提供了新选择。

基于虚拟光源的实时半透明材质渲染

半透明材质渲染是实时渲染领域的重要研究部分。针对透射渲染依赖于准确的透射厚度计算,往往受限于场景模型和光照的复杂度问题,提出一种基于虚拟光源来计算半透明材质渲染中的透射厚度的方法。即在场景中增加一个虚拟光源,并在虚拟光源处使用排序算法来计算场景的深度信息。在计算从真实光源到着色点处的透射厚度时,提出在两者世界空间连线的直线段上进行采样,统计在物体内部的采样点占总采样点数目的比例,乘以直线段的长度得到估计结果。并且,当场景中存在多个真实光源时,基于采样的方法能够复用虚拟光源中存储的场景深度信息。该算法能够有效地提升透射厚度计算的准确率,也能减轻场景中光源数量增多带来的显存开销问题。实验证明该方法能在效率、效果和显存开销之间取得较好的平衡。

Fracture toughness and fatigue crack growth analysis of 7050-T7451 alloy thick plate with different thicknesses

The strength, fracture toughness and fatigue crack growth resistance of 7050-T7451 aluminum alloy plate with different thicknesses （35 mm and 160 ram） were investigated by means of optical microscope, scanning electron microscope and transmission electron microscope. The results show that thicker plate has lower strength and fracture toughness but higher fatigue crack growth resistance, by comparison to the thinner plate. The drop of strength is mainly attributed to grain coarsening in the thicker plate, and the increased degree of recrystallization results in the loss of Kio However, the coarsened grains in the thicker plate make cracks deflected and closure effect enhanced due to surface roughness increased. For both of plates, in the fracture surface subjected plain strain, a transition from transgranular dimpled fracture to intergranular dimpled fracture is observed during the fracture process.

Uniform AlF_3 thin layer to improve rate capability of LiNi_(1/3)Co_(1/3) Mn_(1/3)O_2 material for Li-ion batteries

LiNi1/3Co1/3Mn1/3O2 was coated with uniform nano-sized AlF3 layer by chemical precipitation method to improve its rate capability.The samples were characterized by X-ray diffractometry (XRD),transmission electron microscopy (TEM),energy dispersive spectroscopy (EDS),charge-discharge cycling,cyclic voltammetry (CV),and electrochemical impedance spectroscopy (EIS).Uniform coated layer with a thickness of about 3 nm was observed on the surface of LiNi1/3Co1/3Mn1/3O2 particle by TEM.At 0.5C and 2C rates,1.5% (mass fraction) AlF3-coated LiNi1/3Co1/3Mn1/3O2/Li in 2.8-4.3 V versus Li/Li+ after 80 cycles showed less than 3% of capacity fading,while those of the bare one were 16.5% and 45.9%,respectively.At 5C rate,the capacity retention of the coated sample after 50 cycles maintained 91.4% of the initial discharge capacity,while that of the bare one decreased to 52.6%.EIS result showed that a little change of charge transfer resistance of the coated sample resulting from uniform thin AlF3 layer was proposed as the main reason why its rate capability was improved obviously.CV result further indicated a greater reversibility for the electrode processes and better electrochemical performance of AlF3-coated layer.

Effect of Er ion implantation on the physical and electrical properties of TiN/HfO_2 gate stacks on Si substrate

In this paper, we report the fabrication, electrical and physical characteristics of TiN/HfO2/Si MOS capacitors with erbium （Er） ion implantation. It is demonstrated that the fiat band voltage can be reduced by 0.4 V due to the formation of Er oxide. Moreover, it is observed that the equivalent oxide thickness is thinned down by 0.5 nm because the thickness of interfacial layer is significantly reduced, which is thought to be attributed to the strong binding capability of the implanted Er atoms with oxygen atoms. In addition, cross-sectional transmission electron microscopy experiment shows that the HfO2 layer with Er ion implantation is still amorphous after annealing at a high temperature. This Er ion implantation technique has the potential to be implemented as a band edge metal gate solution for NMOS without a capping layer, and may also satisfy the demand of the EOT reduction in 32 nm technology node.

Convenient and inexpensive determination of optical constants and film thickness of blended organic thin film

This work presents the study of optical constants and film thickness of blended organic thin films, emphasizing on the modeling procedure with modified genetic algorithm aided by absorption or transmittance spectra of both pure materials and the blends. Taking the blending of copper phthalocyanine(Cu Pc) and fullerene(C60) as an example, a simple, convenient and low-cost method for the determination of the optical constants and film thickness of blended organic thin films was demonstrated. New scheme for optical modeling of blended organic thin film was proposed by introducing peak energies of Cody-Lorentz oscillators of the pure materials, which were determined by fitting the film absorption of pure materials. These oscillators of pure materials could be recognized in the transmittance spectrum of their blends, and were further used as the initial searching ranges in the simulation of blended films. As a result, the constraint bounds of the unknown parameters were significantly reduced and modeling efficiency as well as fitting accuracy was improved. For instance, the fitting of the transmittance curves of blended films with different blending ratios reached reliable results in comparison with extinction coefficients obtained from experiment.