Abstract An automatic detection method of crystal defect positions including interface dislocations and a strain imaging method for an arbitrarily oriented atomic array were developed based on the sampling moire technique. An automatic defect location detection algorithm was proposed by defining a defect detection coefficient. As an application, a transmission electron microscope (TEM) image of a Ge/Si crystal structure was investigated, and the atomic arrays including defects were in-situ amplified in a large field of view. The locations of defects and interface dislocations were automatically detected. The strain distribution of this structure was presented and the complex interface curve was plotted. All the positions and atom distribution trends of common dislocations in Ge and Si arrays, and the Ge/Si interface dislocations were verified successfully from the enlarged TEM images. These dislocations were caused by the mismatch between n-column and (n−1) or (n+1)-column homogeneous or heterogeneous atoms, where n is an integer greater than two. This study provides a useful way for characterizing crystal defects and strain distribution in semiconductors and metals.

中文翻译：

---

通过 TEM 图像的莫尔相处理自动检测 Ge/Si 异质结构中的缺陷位置，包括界面位错和应变测量

摘要 基于采样云纹技术，提出了包括界面位错在内的晶体缺陷位置自动检测方法和任意取向原子阵列的应变成像方法。通过定义缺陷检测系数，提出了一种自动缺陷定位检测算法。作为应用，研究了 Ge/Si 晶体结构的透射电子显微镜 (TEM) 图像，并在大视野中原位放大了包括缺陷的原子阵列。自动检测缺陷和界面错位的位置。给出了该结构的应变分布并绘制了复杂的界面曲线。Ge和Si阵列中常见位错的所有位置和原子分布趋势，从放大的 TEM 图像中成功验证了 Ge/Si 界面位错。这些位错是由 n 列与 (n-1) 或 (n+1) 列同质或异质原子之间的不匹配引起的，其中 n 是大于 2 的整数。这项研究为表征半导体和金属中的晶体缺陷和应变分布提供了一种有用的方法。