An ongoing challenge for Li-ion automotive batteries is to be able to recharge them quickly, and thus compete with the refueling times of gasoline vehicles. Such fast-charging rates force battery materials to operate at extreme rates. Historically, such high rates were avoided because they led to accelerated degradation and therefore shorter battery life. However, the new targets demand that batteries operate in this regime. Accordingly, there is a renewed scientific interest in examining the fundamental origins of degradation at these faster rates. Since corresponding degradation events are found to be localized, the macroscopic measurements of voltage, current, and temperature do not provide sufficient insights to understand degradation events. In situ or in operando imaging is therefore required to study degradation events while they are happening.

These degradation events are chemical changes such as the formation of a new solid phase due to unwanted side reactions, such as Li plating, and simultaneously result in physical changes. Hence, one can obtain many insights by tracking physical-geometrical evolution during operation. Atomic force microscopy (AFM) is a promising technique to map such geometrical evolution but is constrained by the slower image acquisition times (∼min). An additional challenge is to configure AFM for the unique electrochemical environment of operating cells.

Guangyi Shang, Zhuanfang Bi, and colleagues from Beihang University and Singapore University of Technology and Design have advanced AFM to break these constraints and provide meaningful imaging for fast-charging battery materials. The study was recently published in the Review of Scientific Instruments.

The researchers combined advances in AFM in other scientific disciplines to extend the capabilities of AFM to study electrochemical systems. They borrowed high-speed AFM to improve the image acquisition time from a few minutes to a few seconds. In this setting, a smaller cantilever (i.e., AFM probe) is used since it has a higher resonant frequency, lower spring constant and allows for faster scans. The downside of such a cantilever is its small dimensions. The researchers developed a custom optical detection system to accurately detect the displacements of the smaller cantilever.

A large image window was achieved using a scan stage that was actuated by two piezoelectric stacks to provide accurate, fast, and independently controlled motions in x- and y-coordinates (z-coordinate is the sample height to be mapped using the AFM). A custom electrochemical housing was also built to support the AFM tip holder, electrolyte, reference, and counter electrodes. The working electrode—the sample of interest—was mounted on the sample stage and moved relative to the electrochemical housing. The electrochemical housing had a glass window at the bottom for laser optics to detect the cantilever displacement. Such clever instrumentation allowed faster image acquisition as well as a larger image window compared to the traditional electrochemical AFM. 

Figure a–c shows example scans of LiMn₂O₄—a positive electrode material for Li-ion batteries. LiMn₂O₄ particles are ∼300 nm in size and are spin-coated on a gold foil. During the experiment, these particles are charged and discharged rapidly (in 100 s each) and 10 AFM scans are captured during each operation. The surface topological changes are compared during charging and discharging (see Figure b–c). For clarity, surface evolution along one coordinate axis is shown, while the images capture a 2.5 × 2 μm² window at 0.75 frames per second.

While discussing the relevance of this study, Jonathan Larson from Lawrence Berkeley National Laboratory (not related to the study) says, “Without question, the advancement of basic energy-storage science stands to greatly benefit from the development of further creative approaches that can expand nanoscopic characterization capabilities of energy-storage interfaces in their native environment. This is acutely true for techniques that can simultaneously access combinations of relevant parameter spaces such as time, space, electrochemical condition, chemical content, electrical properties, pressure, etc.” In his opinion, the group has done a nice job of expanding the electrochemical research community's characterization toolbox by integrating a handful of these relevant measurement parameters with their recently developed “electrochemical high-speed AFM.” The utility of the new tool to track the nanoscopic structure of electrochemically active material as a function of both time and electrochemical potential is well conveyed in their article, as well as the video provided in the supplemental material. Larson is curious to see the future work that follows.

The researchers are excited about various future investigations for this technique. Other positive electrode materials such as lithium iron phosphate and lithium nickel manganese cobalt oxide will be studied, particularly over larger voltage windows where they have been known to undergo morphological changes. Another interesting possibility is to study negative electrode materials like graphite and silicon, which are known to show side reactions and volume expansion. Such in situ imaging studies will finally allow researchers to verify degradation predictions of physics-based battery models, especially at faster rates relevant to next-generation batteries.

Originally published in the December 2020 issue of MRS Bulletin.