Large volume tomography using plasma FIB-SEM: A comprehensive case study on black silicon

Yu Zhang; Charlie Kong; Giuseppe Scardera; Malcolm Abbott; David N R Payne; Bram Hoex

doi:10.1016/j.ultramic.2021.113458

Large volume tomography using plasma FIB-SEM: A comprehensive case study on black silicon

Ultramicroscopy. 2022 Mar:233:113458. doi: 10.1016/j.ultramic.2021.113458. Epub 2021 Dec 14.

Authors

Yu Zhang¹, Charlie Kong², Giuseppe Scardera¹, Malcolm Abbott¹, David N R Payne³, Bram Hoex¹

Affiliations

¹ School of Photovoltaic and Renewable Energy Engineering, University of New South Wales, Sydney, NSW, 2052, Australia.
² Electron Microscope Unit, University of New South Wales, Sydney, NSW, 2052, Australia.
³ School of Photovoltaic and Renewable Energy Engineering, University of New South Wales, Sydney, NSW, 2052, Australia; School of Engineering, Macquarie University, Sydney, NSW, 2109, Australia. Electronic address: yu.zhang@unsw.edu.au.

PMID: 34929560
DOI: 10.1016/j.ultramic.2021.113458

Abstract

The xenon plasma focused ion beam and scanning electron microscopy (PFIB-SEM) system is a promising tool for 3D tomography of nano-scale materials, including nanotextured black silicon (BSi), whose topography is difficult to measure with conventional microscopy techniques. Advantages of PFIB-SEM include high material removal rates, precise control of milling parameters and automated slice-and-view procedures. However, there is no universal sample preparation procedure nor is there an established ideal workflow for the PFIB-SEM slice-and-view process. This work demonstrates that specimen preparation, including the orientation of the volume of interest, is critical for the quality of the final reconstructed 3D model. It thoroughly explores three unique configurations incrementally optimized for higher total throughput. All three sampling configurations are applied to a resin-embedded BSi sample to determine the most favourable workflow and highlight each approach's advantages and disadvantages. The reconstructed 3D models of the BSi surface obtained are shown to be qualitatively closer to the topography measured directly by SEM. The height distribution data extracted from the rendered 3D models reveal a higher structure depth compared to that obtained from an atomic force microscopy measurement. Furthermore, the work demonstrates how samples with different rigidity react to long-term ion-beam interaction, as both amorphous (resin) and crystalline (Si) material is present in the tested specimen. This study improves the understanding of sample-beam interaction and broadens the utility of the 3D PFIB-SEM for more complicated sample structures.