Investigation of a chemically treated InP(100) surface during hydrophilic wafer bonding process

Surface micro-roughness, surface chemical properties, and surface wettability are three important aspects of wafer surfaces during a wafer cleaning process, which determine the bonding quality of ordinary direct wafer bonding. In this study, InP wafers are divided into four groups and treated by different chemical processes. Subsequently, the characteristics of the treated InP surfaces are carefully studied by X-ray photoelectron spectroscopy (XPS), atomic force microscopy (AFM), and contact angle measurements. The optimal wafer treatment method for wafer bonding is determined by comparing the results of the processes as a whole. This optimization is later evaluated by a scanning electronic microscope (SEM), and the ridge waveguide 1.55 mu m Si-based InP/InGaAsP multi-quantum-well laser chips are also fabricated. (c) 2005 Elsevier B.V. All rights reserved.

Dual-wavelength distributed feedback laser for CWDM based on non-identical quantum well

Using non-identical quantum wells as the active material, a new distributed-feed back laser is fabricated with period varied Bragg grating. The full width at half maximum of 115 nm is observed in the amplified spontaneous emission spectrum of this material, which is flatter and wider than that of the identical quantum wells. Two wavelengths of 1.51 mu m and 1.53 mu m are realized under different work conditions. The side-mode suppression ratios of both wavelengths reach 40 dB. This device can be used as the light source of coarse wavelength division multiplexer communication systems.

Growth and properties of magnetron cosputtering grown Mn(x)Ge(1-x)on Si(001)

We have grown MnxGe1-x films (x=0, 0.06, 0.1) on Si (001) substrates by magnetron cosputtering, and have explored the resulting structural, morphological, electrical and magnetic properties. X-ray diffraction results show there is no secondary phase except Ge in the Mn0.06Ge0.94 film while new phase appears in the Mn0.1Ge0.9 film. Nanocrystals are formed in the Mn0.06Ge0.94 film, determined by field-emission scanning electron microscopy. Hall measurement indicates that the Mn0.06Ge0.94 film is p-type semiconductor and hole carrier concentration is 6.07 X 10(19) cm(-3) while the MnxGe1-x films with x=0 has n-type carriers. The field dependence of magnetization was measured using alternating gradient magnetometer, and it has been indicated that the Mn0.06Ge0.94 film is ferromagnetic at room temperature. (c) 2005 Elsevier Ltd. All rights reserved.

Quantum waveguide theory for hole transport in mesoscopic structures

A quantum waveguide theory is proposed for hole transport in the mesoscopic structures, including the band mixing effect. We found that due to the interference between the 'light' hole and 'heavy' wave, the transmission and reflection coefficients oscillate more irregularly as a function of incident wave vector geometry parameters. Furthermore conversion between the heavy hole and light hole states occurs at the intersection. (C) 2003 Elsevier Ltd. All rights reserved.

Structural characteristic of cubic GaN nucleation layers on GaAs(001) substrates by MOCVD

The structural characteristic of cubic GaN (C-GaN) nucleation layers on GaAs(0 0 1) substrates by metalorganic chemical vapor deposition was in detail investigated first by X-ray diffraction (XRD) measurements, using a Huber five-circle diffractometer and an intense synchrotron X-ray source. The XRD results indicate that the C-GaN nucleation layers are highly crystallized. Phi scans and pole figures of the (1 1 1) reflections give a convincing proof that the GaN nucleation layers show exactly cubic symmetrical structure. The GaN(1 1 1) reflections at 54.74degrees in chi are a measurable component, however (002) components parallel to the substrate surface are not detected. Possible explanations are suggested. The pole figures of {1 0 (1) over bar 0} reflections from H-GaN inclusions show that the parasitic H-GaN originates from the C-GaN nucleation layers. The coherence lengths along the close-packed [1 1 1] directions estimated from the (1 1 1) peaks are nanometer order of magnitude. (C) 2002 Elsevier Science B.V. All rights reserved.

Correlated structural and optical investigation of terbium-doped zinc oxide nanocrystals

Tb3+-doped zinc oxide nanocrystals with a hexagonal wurzite structure were successfully prepared by reaction between Zn-O-Tb precursors and LiOH in ethanol. Good incorporation of Tb3+ in ZnO nanocrystals is proved by XRD, FTIR, PL and PLE measurements. The presence of acetate complexes to zinc atoms on particle surfaces is disclosed by FTIR results. Emission from both Tb3+ ions and surface states in ZnO matrix, as well as their correlation were observed. The luminescence mechanism is discussed. (C) 2000 Elsevier Science B.V. All rights reserved.

Optical and electrical investigation of low dimensional self-assembled InAs quantum dot field effect transistors

Self-assembled InAs QD dot-in-a-well (DWELL) structures were grown on GaAs substrate by MBE system, and heterojunction modulation-doped field effect transistor (MODFET) was fabricated. The optical properties of the samples show that the photoluminescence of InAs/GaAs self-assembled quantum dot (SAQD) is at 1.265 mu m at 300 K. The temperature-dependence of the abnormal redshift of InAs SAQD wavelength with the increasing temperature was observed, which is closely related with the inhomogeneous size distribution of the InAs quantum dot. According to the electrical measurement, high electric field current-voltage characteristic of the MODFET device were obtained. The embedded InAs QD of the samples can be regard as scattering centers to the vicinity of the channel electrons. The transport property of the electrons in GaAs channel will be modulated by the QD due to the Coulomb interaction. It has been proposed that a MODFET embedded with InAs QDs presents a novel type of field effect photon detector.

Lateral intersubband photocurrent study on InAs/InAlas/InP self-assembled nanostructures

We present lateral intersubband photocurrent (PC) study on self-assembled InAs/InAIAs/InP(001) nanostructures in normal incidence. With the help of interband excitation, a broad PC signal has been observed in the photon energy range of 150-630 meV arising from the bound-to-continuum intersubband absorption in the InAs nanostructures. The large linewidth of the intersubband PC signal is due to the size inhomogeneity of the nanostructures. With the increase of the interband excitation the intersubband PC signal firstly increases with a redshift of PC peak and reaches its maximum, then decreases with no peak shift. The increase and redshift of the PC signal at low excitation level can be explained by the state filling effect. However, the decrease of PC signal at high excitation level may be due to the change of the mobility and lifetime of the electrons. The intersubband PC signal decreases when the temperature is increased, which can be explained by the decrease of the mobility and lifetime of the electrons and the thermal escape of electrons.

Corrugated surfaces formed on GaAs (331)a substrates: The template for laterally ordered InGaAs nanowires

Morphology evolution of high-index (331)A surfaces during molecular beam epitaxy (MBE) growth have been investigated in order to uncover their unique physic properties and fabricate spatially ordered low dimensional nanostructures. Atomic Force Microscope (AFM) measurements have shown that the step height and terrace width of GaAs layers increase monotonically with increasing substrate temperature in conventional MBE. However, this situation is reversed in atomic hydrogen-assisted MBE, indicating that step bunching is partly suppressed. We attribute this to the reduced surface migration length of Ga adatoms with atomic hydrogen. By using the step arrays formed on GaAs (331)A surfaces as the templates, we fabricated laterally ordered InGaAs self-aligned nanowires.

Tuning of emission wavelength of InAs/GaAs quantum dots sandwiched by combination layers

We investigate about controlling of photoluminescence (PL) wavelengths of InAs/GaAs self-assembled quantum dots (QDs) sandwiched with combination strained-buffer layer (CSBL) and combination strained-reducing layer (CSRL). The emission peak position of QDs is red-shifted to 1.37 mu m. The density of the QDs is increased to 1.17x10(10) cm(-2). It is indicated that optical properties of QDs could be improved by optimizing of the buffer and covering layers for the QDs. These results may provide a new way to further developing GaAs-based 1.3 mu m light sources.

Surface morphology evolution of strained InAs/GaAs(331)a films

Surface morphology evolution of strained InAs/GaAs(331)A films was systematically investigated in this paper. Under As-rich conditions, InAs elongated islands aligned along [1 (1) over bar0] are formed at a substrate temperature of 510 degrees C. We explained it as a result of the anisotropic diffusion of adatoms. Under In-rich conditions, striking change has occurred with respect to the surface morphology of the InAs layers. Instead of anisotropic InAs elongated islands, unique island-pit pairs randomly distributed on the whole surface were observed. Using cooperative nucleation mechanisms proposed by Jesson et al. [Phys. Rev. Lett. 77, 1330 (1996)], we interpret the resulting surface morphology evolution.

Continuous-wave electrically injected InP/GaInAsP equilateral-triangle-resonator lasers

Equilateral-triangle-resonator (ETR) microlasers with an output waveguide connected to one of the vertices of the ETR are fabricated using standard photolithography and inductively-coupled-plasma etching techniques. Continuous-wave electrically injected 1550 nm ETR laser with side length ranged from 15 to 30 tm are realized at room temperature.

曙光5000A天体大规模数值模拟软件性能测试

在国产百万亿次超级计算机曙光5000A上进行了天体大规模数值模拟软件的性能和可扩展测试实验,详细介绍了软件中的测试程序以及测试环境和过程,并对测试结果进行了分析.对于80×80×50的网格规模,采用每节点4进程测试了16～128个处理器核,每节点8进程、16进程分别测试了16～512个处理器核,相对加速比最终分别达到5.33、10.48和12.57,并行效率分别达到66.66%、32.58%和32.29%.对于160×160×100的网格规模,测试了每节点16进程的64～8192个核的性能,最大相对加速比为12.46,并行效率为9.73%.测试结果表明,曙光5000A具有良好的性能,测试结果对软件下一步的优化研究具有重要的指导意义.

密码Hash函数相关问题研究

密码Hash函数是信息安全密码学的一个重要研究内容，是一类广泛应用的密码算法，用于把任意长度的字符串压缩成特定长度的字符串，同时需要在各种应用环境下满足一定的安全要求如抗碰撞，抗原象等。Hash函数广泛应用于数字签名、可证明安全、密码算法的构造以及重要的安全协议中。对Hash函数进行研究、分析Hash函数的安全性、构造安全高效的Hash算法有着重要意义。 本文研究了Hash函数的安全性质、设计结构以及常用分析方法，研究了Hash函数扩散层部件的设计，并且对MAME压缩函数算法进行了分析，取得了如下研究结果： (1) 研究了密码Hash函数的安全性质、设计结构、设计原理和常用分析方法，归纳总结了51个SHA-3候选算法的设计特点、设计原理和实现效率，研究了最新的分析进展，总结了新的攻击方法如REBOUND攻击等。NIST仿照AES的征集过程的SHA-3竞赛，目标是选出新的Hash函数标准SHA-3。进入第一轮的候选算法有51个，经过筛选选出其中的14个作为当前第二轮的候选算法。这些新Hash算法是由世界各国密码学家精心设计，是Hash函数领域最新设计思想的集体展示，当中涌现出很多新的设计结构和设计方法，同时激励密码学家发展新的分析方法。 (2) 设计并实现了了有限域上的扩散层构造算法以及扩散层分支数测试的算法，并针对多元域上的扩散层矩阵，本文使用编码理论，利用GRS码和柯西矩阵等设计了多元域扩散层矩阵的构造算法；使用有限域上的高斯消元法和线性码的性质设计了多元域扩散层矩阵的分支数的检测；设计了高效的二元域扩散层矩阵分支数测试算法。 (3) 针对MAME压缩函数算法进行差分分析，MAME算法是SHA-3候选算法Lesamnta的前身，于CHES 2007上提出的面向硬件有效实现的Hash算法。本文利用差分攻击对MAME算法进行分析，首先针对MAME的结构性质利用对通用Feistel结构的攻击方法构造了22轮差分攻击，碰撞攻击的复杂度为2^97，（第二）原象攻击的复杂度为2^197；对23轮的差分攻击需要的预计算是2^64张表，每张表的大小为2^64；对24轮的差分攻击需要的预计算是2^128张表，每张表的大小为2^64。针对24轮差分攻击很大的内存复杂度，我们利用了算法的细节特性，改进了差分攻击，新的差分不需要预计算的辅助内存，（第二）原象的复杂度为2^224。

面向Feistel密码的基于故障传播模式的差分故障分析

密码算法是信息安全研究的核心内容之一，其实际安全性不仅依赖于密码自身的数学特性，也依赖于具体的实现特性。基于实现的密码分析是一种有别于传统密码分析的新型密码分析方法，它利用算法实现时的信息泄露来恢复秘密信息。差分故障分析(Differential Fault Analysis，简称DFA)就是这样一类重要的密码分析方法。 现代密码学中，密码设计通常基于混淆和扩散这两大基本原则。对于一个分组密码而言，选择一个合适的轮函数并进行若干次迭代可以提供必要的混淆和扩散。因此，目前流行的分组密码均为迭代型密码，所采用的典型结构包括Feistel结构、SPN结构和广义Feistel结构等。这些密码结构及其所采用的基础密码组件(例如，S盒和P置换等)的性质，完全决定了故障在传播过程中所呈现的一些模式。直观上，这种内在特征可以用于挖掘DFA攻击和密码结构之间的关系。因此，完全可能利用这种特征来建立一种面向密码结构的系统化DFA攻击方法。 本文主要研究面向Feistel密码的差分故障分析方法，并探讨这类分析方法与已有可证明安全性理论分析结论之间的关系。为此，引入了故障传播路径(Fault Propagation Path，简称FPPath)和故障传播模式(Fault Propagation Pattern，简称FPPattern)的概念，给出了适用于Feistel结构的 FPPath 和 FPPattern 计算方法，建立了与已有可证明安全性理论结果之间的关系。在此基础上，提出了一种面向Feistel密码的基于故障传播模式的 系统化差分故障分析方法。使用该方法，可编程实现FPPath和FPPattern的自动计算，这将有助于针对Feistel密码的自动化DFA攻击的实施。这种情形下，可将FPPath的长度视作评估DFA攻击有效性的一种度量指标。此外，该系统化方法的必然结果是攻击性能的显著提高：不但攻击轮数有所减少，而且故障植入点数量也会减少，这将迅速降低实施一次成功攻击所需的故障密文数。最后，为验证该方法的正确性和有效性，以Camellia密码算法为具体实例，进行了相关模拟攻击实验研究，并给出了相应的数据复杂度分析和时间复杂度分析。通过充分利用Camellia算法中P置换的性质，在不需要穷举搜索的情况下，新攻击方法仅需要6个故障密文即可完全恢复出128位密钥，而成功恢复出192位或256位密钥所需要的故障密文数则为22个。结果表明，基于FPPattern的DFA方法要优于所有已有同类方法。