Low-voltage SEM of air-sensitive powders: from sample preparation to micro/nano analysis with secondary electron hyperspectral imaging

Powder materials are used in all corners of materials science, from additive manufacturing to energy storage. Scanning electron microscopy (SEM) has developed to meet morphological, microstructural and bulk chemical powder characterization requirements. These include nanoscale elemental analysis and high-throughput morphological assays. However, spatially localized powder surface chemical information with similar resolution to secondary electron (SE) imaging is not currently available in the SEM. Recently, energy filtered (EF-) SEM has been used for surface chemical characterization by secondary electron hyperspectral imaging (SEHI). This review provides a background to existing powder characterization capabilities in the low voltage SEM provided by SE imaging, EDX analysis and BSE imaging and sets out how these capabilities could be extended for surface chemical analysis by applying SEHI to powders, with particular emphasis on air and beam sensitive powder surfaces. Information accessible by SEHI, its advantages and limitations, is set into the context of other chemical characterization methods that are commonly used for assessing powder surface chemistry such as by Auger electron spectroscopy (AES) and X-ray photoelectron spectroscopy (XPS). The applicability of existing powder preparation methods for SEM to SEHI is also reviewed. An alternative preparation method is presented alongside first examples of SEHI characterization of powder surfaces. The commercial powder materials used as examples were carbon-fiber/polyamide composite powder feedstock (CarbonMide®) used in additive manufacturing and powders consisting of lithium nickel cobalt oxide (NMC). SEHI is shown to differentiate bonding present at carbonaceous material surfaces and extract information about the work function of metal oxide surfaces. The surface sensitivity of SEHI is indicated by comparison of pristine powders to those with surface material added in preparation. A minimum spatial localization of chemical information of 55 nm was achieved in differentiating regions of NMC surface chemistry by distinct SE spectra.