To understand materials and multi-material systems (e.g., liquid/metal batteries) towards advanced applications, it is necessary to characterize both the individual material components and their interfacial interactions. For example, because battery stacks are comprised of different material types (i.e., metals, polymers, liquid electrolytes), which may evolve as the battery cycles (e.g., generating solid-electrolyte interphase) Zhang et al., 2020Kim et al., 2020, high-quality, targeted characterization of the interfaces and morphology are crucial to understanding and improving battery performance Ding et al., 2020. Many methods have been used to characterize multiphase materials, including cryogenic scanning electron microscopy (cryo-SEM) and focused ion beam (cryo-FIB), cryogenic transmission electron microscopy (cryo-TEM), microcomputed tomography (MicroCT), neutron scattering, and X-ray microanalysis Xiang et al., 2020. However, there are challenges as well as benefits to each individual approach: while cryo-SEM, FIB, TEM workflows can achieve high (sub-nm) resolution Zachman et al., 2018, they require the batteries to be disassembled and can only observe small (several µm) regions; in contrast, while MicroCT is non-destructive, its resolution is relatively poor (~ 700 nm) Frisco et al., 2017. Thus, combining the strengths of these methods has the potential to radically advance comprehensive battery characterization.

Because batteries are a high-consequence exemplar of a multiphase system, in 2019 we were specifically concerned with obtaining large-area cross-sections of lithium-metal (Li-metal) coin cells in half cell format; however, no single technique available at the time was able to deliver nondestructive characterization with nanoscale resolution. Cryogenic cross-sectioning was a feasible approach to freezing the liquid electrolyte and preserving interfaces (effectively converting a multiphase system into a solid system). However, to cross-section a frozen, intact coin cell, it is necessary to section through the 250 µm-thick stainless-steel casing and then obtain a smooth cross-sectioned surface of the fragile battery stack. Traditional Ga⁺ FIB technology has been enhanced by the introduction of plasma FIB (PFIB), which uses higher beam currents and sputter yields to enable larger-volume applications. However, a PFIB-only workflow cannot feasibly be used to access the buried regions of interest (ROI) in an intact coin cell for three reasons: (1) the workflow would be low throughput; (2) the high beam currents are likely to induce thermal damage; and (3) each individual component of the stack would have differential sputter yields, so any single sputter yield setting will necessarily introduce undesirable surface artifacts (e.g., curtaining, damage) in other stack layers. Alternatively, ultrashort pulsed lasers (UPL) could be used for rapid and athermal material removal and had been demonstrated for material processing, micromachining, tissue modification Vogel & Venugopalan, 2003Shirk & Molian, 1998Förster et al., 2021Fazio et al., 2020, and spectroscopy applications; however, UPL-prepared cross-sections tend to have observable aperiodic laser texturing, better known as “laser-induced periodic surface structures” LIPSS ([ Bonse et al., 2012, ]), which can limit the quality of characterization and may obscure features in the ROI.

To exploit the combined strengths of UPL and PFIB, we developed techniques using coincident UPL, PFIB, and SEM (collectively, “Laser PFIB”), to enable rapid analysis of large 2D areas and 3D volumes under cryogenic conditions while preserving multiphase material interfaces for nanoscale characterization Echlin et al., 2015Randolph et al., 2018. The Laser PFIB workflow occurs in two steps: sample sectioning and sample polishing. Sample sectioning: once the sample reaches cryogenic temperatures, the UPL is used to rapidly remove material from the sample and obtain a relatively flat cross-section. Specifically, the UPL’s high average power at femtosecond (fs) pulse durations can enable athermal ablation that carves through the stainless steel without thermally damaging the more beam-sensitive ROI. However, this rapid, coarse material removal tends to leave some observable LIPSS, which can be mitigated (but not eliminated) by adjusting the laser parameters. Sample polishing: PFIB is used to remove the LIPSS in a targeted ROI, polishing the sample surface to leave a very smooth cross-section surface and reveal smaller scale features. Because the bulk removal has already been done, the PFIB milling can be done without the use of high-dose beam conditions, mitigating beam-induced damage. The combined cryo-UPL/PFIB technique improves sample quality compared to either UPL or PFIB alone. While the results of our previous efforts have already been published, here we offer greater technical detail related to the holders, sample preparation, and parameter optimization. Our goal is to highlight both the capabilities available at the time and the potential advancements enabled by more recent technology gains, enhancing the generalizable utility of this technique for future 2D and 3D multiphase system characterization.