Integration of III-Vs on silicon: from the micron to the nanometre regime

几十年来，硅CMOS技术的复杂性一直在不断增加，这是由于对更高速度和存储容量的需求不断增长。 在大多数情况下，这些需求可以通过微型化的惊人进步来满足，这种进步在一代人之前是不可想象的。然而，最近，鉴于对功耗、数据传输和迄今为止无法进入的领域中的新应用的关注，越来越迫切地需要附加功能。新的功能必然意味着硅技术将扩展到其他半导体材料，这些材料的光学和电学性能超出了单独的Si。鉴于其优异的光学和电学性能，GaAs是最重要的III-V族半导体之一。本计画的目的是探索以金属有机化学气相沈积法（MOCVD）在图案化矽及锗/矽基板上磊晶成长砷化镓半导体堆叠。这将是在硅CMOS平台上实现基本光学器件子组件（如波导、光栅耦合器和激光器）单片集成的第一个必要步骤。这里的挑战是要克服的晶格参数和热性能的大的失配所产生的晶体缺陷的问题，迄今为止，阻碍了砷化镓硅技术的应用。 在我们的方法中，外延将在专门设计的衬底图案上进行，具有大的纵横比和从几微米到几十纳米的横向尺寸。这将允许探索作为图案尺寸的函数的位错结构反相域的演变，其定义了从基本上刚性到顺应性衬底制度的过渡。衬底图案化将包括蚀刻的Si特征的侧壁钝化，允许使用选择性外延技术来精确定位化合物半导体堆叠，所述化合物半导体堆叠直接在Si上或选择性生长的Ge缓冲结构上。图案几何形状和缺陷结构之间的关系将通过结构、分析和光学技术来研究，例如透射电子显微镜（TEM）、原子探针测量、高分辨率X射线衍射（HR-XRD），包括同步加速器纳米衍射、显微拉曼和光致发光光谱。我们打算设计特殊的测试结构，以证明不匹配的硅衬底上的外延III-V族材料的器件质量。这些将允许评估激光作用的光学泵浦作为一个初步的步骤，对制造单片集成激光器CMOS处理的硅晶片。

During decades the complexity of silicon CMOS technology has continuously been increasing, driven by ever rising demands for higher speed and storage capacity. For the most part, these demands could be met by breath taking advances in miniaturization to a degree unthinkable a generation ago. Lately, however, in view of concerns about power consumption, data transmission, and new applications in hitherto inaccessible fields, additional functionalities are becoming more and more urgently needed. New functionalities necessarily imply silicon technology to be extended to other semiconducting materials with optical and electrical properties beyond those of Si alone.In view of its excellent optical and electrical properties GaAs is one of the most important III-V semi-conductors. It is the goal of this project to explore the epitaxial growth by metal organic chemical va-pour deposition (MOCVD) of semiconductor stacks based on GaAs on patterned Si and Ge/Si sub-strates. This will be the first necessary step towards the monolithic integration of elementary optical-device subcomponents, such as waveguides, grating couplers and lasers on a Si CMOS platform. The challenge here is to overcome the problem of crystal defects arising from the large mismatch of lattice parameters and thermal properties which has so far hindered GaAs-on-Si technology to be applied. In our approach the epitaxy will be carried out on specially designed substrate patterns with large aspect ratios and lateral dimensions ranging from several microns to tens of nanometers. This will permit to explore the evolution of dislocation structures anti-phase domains as a function of the pattern size, which defines the transition from an essentially rigid to a compliant substrate regime. The substrate patterning will include sidewall passivation of the etched Si features, permitting the use of selective epitaxy techniques for precise positioning of the compound semiconductor stacks either directly on Si or on selectively grown Ge buffer structures.The relation between pattern geometry and defect structure will be studied by structural, analytical and optical techniques, such as transmission electron microscopy (TEM), atom probe measurements, high-resolution X-ray diffraction (HR-XRD) including synchrotron nanodiffraction, micro-Raman and photoluminescence spectroscopy. We intend to design special test structures in order to prove the device quality of epitaxial III-V material on mismatched Si substrates. These will permit to assess lasing action by optical pumping as a preliminary step towards the fabrication of monolithically integrated lasers on CMOS-processed Si-wafers.