基于热-流-力耦合计算的多芯片组件布局构形优化

基于构形理论和多物理场耦合数值计算方法,建立了自然对流条件下均匀产热的多芯片组件模型,给定印刷电路板面积和芯片总占地面积为约束条件,分别以最高温度、最大应力和最大形变为优化目标,以芯片个数及芯片长宽比为设计变量,研究了芯片布局演化对系统性能的影响.结果 表明:不同优化目标下,最优构形均为芯片长宽比为2.1的8芯片布局方式,多芯片组件的最高温度、最大应力和最大形变分别最多可降低16.5％,28.3％和26.9％.对芯片个数和芯片长宽比双自由度的优化效果要明显优于仅对芯片长宽比的单自由度优化.     

Improving Ideality of P-Type Organic Field-Effect Transistors via Preventing Undesired Minority Carrier Injection

Electron injection plays a crucial role in arousing the double-slope characteristics for p-type organic field-effect transistors (OFETs) with narrow-bandgap organic semiconductors (OSCs). This issue will not only result in the misrepresentation of OFET performance but also may cause device instability, hence impeding their further development in real-world applications. A facile and highly efficient approach is developed to circumvent this issue by implementing modification on the electrode/organic semiconductor interface. An ultrathin layer of wide-bandgap OSC with suitable energy levels is introduced to block the undesirable electron injection. By this means, typical double-slope behaviors and bias stress stability in the p-type OFETs can be significantly improved. Using 2,8-difluoro-5,11-bis(triethylsilylethynyl) anthradithiophene-based OFETs the double-slope behaviors of as-fabricated devices are effectively converted to near-ideal behaviors after modification, leading to a dramatic improvement of average reliability from 65.11% to 91.76%. Furthermore, the positive drift of transfer curves under prolonged bias stress is also successfully suppressed. This strategy demonstrates good universality and can provide a new guideline for the fabrication of OFETs with ideal behaviors.     

水热法制备K0.5Bi0.5TiO3陶瓷粉体及其介电性能研究

采用水热法合成了K0.5Bi0.5TiO3(KBT)无铅压电陶瓷粉体.运用X线衍射(XRD)分析了KOH浓度及反应时间对产物物相的影响;利用扫描电镜(SEM)和透射电镜(TEM)观察研究了粉体的形貌.结果表明,增大碱浓度(4～12 mol/L)和延长反应时间都有利于KBT纳米粉体的合成.在220℃,KOH的起始浓度12 mol/L水热合成48 h的纯KBT相呈立方形结构,晶型规整,晶粒尺寸约为100nm.由水热法制备的粉体在1060℃烧结,获得介电损耗为0.05的致密压电陶瓷材料.     

厚膜混合集成电路可靠性技术

总结了混合集成电路的主要失效模式,在此基础上,主要从混合集成电路的设计和工艺两方面分析了其产生的原因,通过设计、工艺、原材料和元器件等方面采取对策和措施,达到提高混合集成电路可靠性的目的.     

Thermal Characterization of 1.3 μm InAsP/InGaAsP Ridge Waveguide MQW Lasers Based on Spectroscopy Method

An experimental way to analyze the thermal characterization of semiconductor lasers based on spectroscopy method under pulse driving conditions has been developed. By using this way the thermal characteristics of strain compensated 1.3 μm InAsP/InGaAsP ridge waveguide MQW laser diodes have been investigated. Results show that by measuring and analyzing the lasing spectra under appropriate driving parameters and temperature ranges, the thermal resistance of the laser diodes could be deduced easily. A higher thermal resistance of 640 K/W has been measured on a narrow ridge laser chip without soldering. Other thermal and spectral properties of the lasers have also been measured and discussed.     

基于分段凿孔的极化码级联方案

极化码拥有出色的纠错性能,但编码方式决定了其码长不够灵活,需要通过凿孔构造码长可变的极化码。该文引入矩阵极化率来衡量凿孔对极化码性能的影响,选择矩阵极化率最大的码字作为最佳凿孔模式。对极化码的码字进行分段,有效减小了最佳凿孔模式的搜索运算量。由于各分段的第1个码字都会被凿除,且串行抵消译码过程中主要发生1位错,因此在各段段首级联奇偶校验码作为译码提前终止标志,检测前段码字的译码错误并进行重新译码。对所提方法在串行抵消译码下的性能进行仿真分析,结果表明,相比传统凿孔方法,所提方法在10^–3误码率时能获得约0.7 dB的编码增益,有效提升了凿孔极化码的译码性能。     

Models of Restoration Probability in WDM Networks Employing Active Restoration

The ever-increasing demand for network bandwidth makes network survivability an issue of great concern. Lightpath restoration is a valuable approach to guaranteeing an acceptable level of survivability in WDM optical networks with better resource utilization than that of its protection counterpart. Active restoration (AR) is a newly proposed lightpath restoration scheme [M. Mostafa et al. OSA Journal of Optical Networking, vol. 3, no. 4, pp. 247–260] that combines the best of protection and reactive restoration while avoiding their shortcomings. In this paper, we conduct detailed performance analysis on the restoration probability of AR-based WDM networks. In particular, analytical models of restoration probability are developed respectively for networks with full-wavelength conversion capability and for networks without wavelength conversion capability under different backup path searching schemes. Based on the new models, we investigate the effects of wavelength availability, wavelength conversion capability, path length as well as backup path seeking methods on the restoration probability.     

基于联合区间偏最小二乘判别分析的猪肉近红外光谱定性建模分析

为解决偏最小二乘判别分析(PLSDA)建模时光谱区域中的噪声及冗余信息干扰问题,提出一种基于联合区间偏最小二乘判别分析(Si PLSDA)算法,并将该算法应用于猪肉近红外光谱的定性建模分析。Si PLSDA利用联合区间偏最小二乘回归(Si PLS)进行光谱特征区域筛选,在筛选出来的光谱区域内建立数据的定性预测模型。采用Antaris II快速傅里叶变换近红外光谱分析仪获取波数范围为10000~4000 cm-1的猪肉样本近红外光谱,采用标准正态变量变换(SNV)进行近红外光谱的预处理,用Si PLSDA建立猪肉近红外光谱的定性模型。实验结果表明,Si PLSDA建立的预测模型对猪肉储藏时间的识别率达到93.94%,高于基于全光谱区域建立的PLSDA预测模型的识别率。     

基于SOI材料的刻蚀光栅分波器的制作工艺

研究了基于绝缘材料上的硅（SOI）材料的平面波导刻蚀光栅分波器的主要制作工艺．利用电感耦合等离子体刻蚀（ICP）技术，在SOI材料上制作了垂直度大于89°的光滑的光栅槽面．氧化抛光后刻蚀侧壁的表面均方根粗糙度（RMS）有3nm的改善，达到7.27nm（采样面积6.2μm×26μm）．通过采用集成波导拐弯微镜代替弯曲波导使1×4分波器的器件尺寸仅为20mm×2.5mm．测试结果表明器件实现了分波功能．     

1064nm RCE探测器光电响应特性分析

对1064nm谐振腔增强型（RCE）光电探测器(PD)的光电响应特性进行了分析研究．利用MBE生长技术得到有源区分别为量子阱和量子点的1064nm RCE探测器的外延片,并对制作的探测器进行了各种光电特性测试．结果表明量子阱结构的RCE探测器量子效率峰值达到57%,谱线半宽6～7nm,峰值波长1059nm;而量子点结构的RCE探测器量子效率峰值达到30%,谱线半宽5nm,峰值波长1056nm．通过分析量子效率和吸收系数之间的关系,对两种结构器件的吸收进行了比较,发现虽然量子点探测器的吸收小,但通过合理设计共振腔等方法也可以达到较高的量子效率．两种结构的器件都有很好的Ｉ-Ｖ特性．