Strain Analysis by Nano-Beam Electron Diffraction using convergent electron nanoprobes

半导体晶体中应变场的测量是理解新型GeSi/Si或（Al）GaN/蓝宝石异质结构物理过程的先决条件。本文首先研究了GeSi/Si孤岛上的应变特征，以便对锗在Si上的生长有一个基本的了解。其次，作为高性能计算机处理器的基本模块，gesi - mosfet中的应变场和局部Ge组成将被测量，因为电子特性极大地依赖于应变分布。第三，将研究GaN/蓝宝石假衬底的应变，因为该层的应力减小效率在很大程度上决定了后续InGaN/GaN LED有源层的效率和生长。为了可靠地获取这些应变场，需要开发一种利用衍射光束位置进行应变测量的扫描透射电子显微镜（STEM）方法。特别是，在每个扫描点获得会聚束电子衍射（CBED）模式时，在样品上扫描STEM探针。通过专用算法，可以准确地检测到圆盘状CBED反射，精度达到7x10-4。为了大幅减少实验数据量，这些算法将在控制所使用的FEI Titan显微镜的软件中实现，以便在STEM探针扫描时进行原位CBED模式处理。这不仅可以实现二维应变映射，而且可以在数据采集后立即提供应变结果。利用该方法在0.5nm空间分辨率下进行2D映射的全部能力的一个主要先决条件是CBED模式采集的急剧加速，这是传统相机无法实现的。因此，延迟线探测器的测试，修改和最终安装计划在项目的开始。为了研究微米范围内大扩展的应变场，将在实验和模拟中发展stem -进动方法。在进动模式下，STEM探针的入射在一个圆锥上旋转，而由此产生的衍射图样的位移得到补偿。通过这种方式，可以消除CBED强度的强烈变化。此外，高阶反射的高应变灵敏度可以被利用，因为它们看起来更亮。此外，由于后放大镜造成的衍射图样的畸变要在应变测量中加以测量和考虑。这对于随后确定包括剪切在内的全应变张量是很重要的。最后，该技术将与所谓的z对比信号相结合，对化学成分和试样的应变敏感。因此，同时测量应变和z -对比度将提高成分测量的精度，因为应变和成分引起的z -对比度变化可以分开。

The measurement of strain fields in semiconductor crystals is a prerequisite for understanding physical processes in novel GeSi/Si or (Al)GaN/sapphire based heterostructures. This proposal firstly deals with the strain characterization in GeSi/Si islands for a basic understanding of Ge growth on Si. Secondly, strain fields and local Ge composition in GeSi-MOSFETs, being the basic module of high-performance computer processors, are to be measured as electronic properties drastically depend on the strain distribution. Thirdly, GaN/sapphire pseudosubstrates will be investigated with respect to strain since the efficiency of stress reduction in this layer highly determines efficiency and growth of subsequent InGaN/GaN LED active layers.In order to access these strain fields reliably, a scanning transmission electron microscopy (STEM) method for strain measurement is to be developed which exploits the positions of diffracted beams. In particular, a STEM probe is scanned over the specimen while a convergent beam electron diffraction (CBED) pattern is acquired at each scan point. With dedicated algorithms the disc-shaped CBED reflections can be detected accurately, leading to precisions of 7x10-4. To reduce the amount of experimental data drastically, these algorithms are to be implemented in a software for controlling the FEI Titan microscope used, so as to perform in-situ CBED pattern processing while the STEM probe is scanning. This not only enables 2D strain mapping but also provides the strain result immediately after data acquisition. One main prerequisite to exploit the full capability of the method for 2D mapping as to the spatial resolution of 0.5nm already achieved before is a drastic acceleration of CBED pattern acquisition which is not possible with conventional cameras. Therefore, test, modification and final installation of a delay-line detector are planned at the beginning of the project.To study strain fields of large extensions in the micron range, a STEM-precession method will be developed experimentally and in simulations. In precession mode, the incidence of the STEM probe is rotating on a cone while the resulting shift of the diffraction pattern is compensated for. In this way, strong variations of CBED intensities can be eliminated. In addition, the high strain sensitivity of high order reflections can be exploited as they appear brighter.Furthermore, distortions of the diffraction pattern due to post-magnifying lenses are to be measured and accounted for in strain measurements. This is important for the subsequent determination of the full strain tensor including shear.Finally, this technique is to be combined with the so-called Z-contrast signal, being sensitive to the chemical composition and to strain of the specimen. Consequently, the simultaneous measurement of both strain and Z-contrast will improve the precision of composition measurements since strain- and composition-induced Z-contrast variations can be separated.