Experimental evaluation of interfaces using atomic-resolution high angle annular dark field (HAADF) imaging

详细信息	查看全文 | 推荐本文 |

• 作者：Paul D. Robb^a ; ^b ; ^{p.robb@physics.gla.ac.uk} ; Michael Finnie^a ; Paolo Longo^a ; Alan J. Craven^a ; ^b
• 关键词：HAADF imaging ; Interfaces ; Image processing
• 刊名：Ultramicroscopy
• 出版年：2012
• 期刊代码：79_03043991
• 类别：ph
• 出版时间：March, 2012
• 卷：114
• 期：Complete
• 页码：11-19
• 文件大小：1151 K

摘要

Aberration-corrected high angle annular dark field (HAADF) imaging in scanning transmission electron microscopy (STEM) can now be performed at atomic-resolution. This is an important tool for the characterisation of the latest semiconductor devices that require individual layers to be grown to an accuracy of a few atomic layers. However, the actual quantification of interfacial sharpness at the atomic-scale can be a complicated matter. For instance, it is not clear how the use of the total, atomic column or background HAADF signals can affect the measured sharpness or individual layer widths. Moreover, a reliable and consistent method of measurement is necessary. To highlight these issues, two types of AlAs/GaAs interfaces were studied in-depth by atomic-resolution HAADF imaging. A method of analysis was developed in order to map the various HAADF signals across an image and to reliably determine interfacial sharpness. The results demonstrated that the level of perceived interfacial sharpness can vary significantly with specimen thickness and the choice of HAADF signal. Individual layer widths were also shown to have some dependence on the choice of HAADF signal. Hence, it is crucial to have an awareness of which part of the HAADF signal is chosen for analysis along with possible specimen thickness effects for future HAADF studies performed at the scale of a few atomic layers.