Automated SEM-Guided AFM Scan with Dynamically Varied Scan Speed
Jun Chen,Ji Ge,Brandon K. Chen,Zheng Gong,Chao Zhou,Chaoyang Shi,Changhai Ru,Huayan Pu,Yan Peng,Shaorong Xie,Yu Sun
DOI: https://doi.org/10.1109/marss.2018.8481201
2018-01-01
Abstract:For imaging nano-scaled samples, atomic force microscopy (AFM) and scanning electron microscopy (SEM) represent two complementary imaging techniques. In a hybrid SEM-AFM system, a compact AFM is installed inside the high vacuum chamber of an SEM, where SEM provides largely 2D imaging and material compositions of a sample while AFM is capable of complementarily measuring 3D topography of the sample. Although SEM can achieve real-time imaging (e.g., 20 Hz), AFM scan can take minutes to generate an image, demanding strategies for speeding up AFM measurement. In existing hybrid SEM-AFM systems, SEM and AFM measurements are made independently. This paper presents, for the first time, a technique of using SEM nanoscopic imaging to guide the scan speed of AFM imaging. The dynamic variation of AFM scan speed is based on features identified in SEM imaging. Information/features are extracted from real-time SEM images and quantitated using local entropy and other metrics. The generated feature metric map is used to produce a speed map for varying AFM scan speed at each position on the sample. Experiments were conducted with a new SEM-compatible AFM instrument we recently developed, as the test bed of the SEM-guided AFM scan technique. The results for the samples measured in this work demonstrate that time savings of this technique, compared to traditional AFM scan using a constant speed, were up to 66% with equivalent imaging accuracy obtained with traditional fine scan. With the same time cost of traditional fast scan, the SEM-guided AFM scan technique had an accuracy improvement of 47%.