Ultra-thin plan-view lamella made by focused ion beam

Abstract

Understanding and controlling microstructure is critical for a wide range of functional materials, where properties such as ferroelectricity, conductivity, and catalytic activity are tightly coupled to nanoscale features like grain size, phase distribution, and interfaces. However, rigorous microstructural characterization is often hindered by the ultrathin geometries, nanocrystalline domains, and structural polymorphism inherent to these materials. Hafnium zirconium oxide (Hf₀.₅Zr₀.₅O₂, HZO) is an exemplar case: a leading ferroelectric oxide for next-generation embedded memory, storage and DRAM like applications, whose performance is strongly microstructure-dependent. Among available techniques, transmission electron microscopy (TEM) provides the most direct access to HZO’s nanometer-scale structure—TEM imaging reveals grain morphology, while nanobeam electron diffraction (NBED) distinguishes among its subtle polymorphs. However, conventional cross-sectional focused-ion-beam (FIB) lamellae often obscure such details due to limited field of view and through-thickness grain overlap. In contrast, plan-view lamellae isolate the target ultrathin film, enabling single-grain NBED and large-area imaging. In this study, we detail a refined focused-ion-beam (FIB) workflow for producing electron-transparent, plan-view lamellae from ∼10 nm-thick HZO films grown on Si substrates. The pivotal step is a low-kV fine-thinning sequence designed to maximize the preserved HZO area while minimizing ion-induced damage. This progress is tracked in situ by monitoring contrast changes in the scanning electron microscope (SEM) and verified ex situ with energy-dispersive X-ray spectroscopy (EDS). Referenced to an adjacent unthinned region, quantitative scanning transmission electron microscopy (STEM) imaging combined with EDS confirms a reproducible taper from the native 10 nm thickness to 3–4 nm at the lamella edge, all while maintaining atomic-column integrity. This workflow enables unambiguous polymorph identification and provides a robust platform for correlating microstructure with ferroelectric functionality in HZO and related ultrathin oxides—an essential step toward evaluating the performance of fluorite-based ferroelectric devices.