Study warns electron microscopes may be distorting key images of next-generation battery materials

Inside a glovebox at the University of Chicago, a researcher lifts a wafer-thin sliver of lithium metal onto a tiny grid and seals it into a microscope holder flooded with argon gas. Outside that controlled bubble, the metal would start reacting with air in seconds. Inside a powerful electron microscope, a strong beam could quietly rearrange its chemistry.

For years, images captured under those imperfect conditions have been treated as hard evidence of what happens inside cutting-edge lithium and sodium batteries. A new study argues many of those pictures are at best incomplete—and sometimes misleading.

A microscope that can change what it measures

A team led by the University of Chicago’s Pritzker School of Molecular Engineering reports that common transmission electron microscopy (TEM) workflows can chemically and structurally damage some of the most important materials in next-generation batteries. The peer-reviewed paper, published March 2 in Joule, finds that standard preparation and imaging techniques frequently alter lithium and sodium metals and their compounds before or during imaging—and that researchers often have no way to tell.

The group also demonstrated a revised workflow that allows high-resolution images of lithium metal at room temperature without significantly changing it, and proposed reporting standards they hope will become a new baseline.

“There is no standard framework,” said co-first author Zhao Liu of Thermo Fisher Scientific. “So even with exactly the same materials, people may get a very different result.”

Why lithium and sodium are so hard to image

Lithium and sodium are alkali metals that react readily with air and moisture, forming new compounds on their surfaces in moments. That reactivity helps enable high energy storage, but it also makes these materials notoriously difficult to study.

Over the past decade, TEM—including cryogenic TEM—has become central to battery research. The technique can resolve structures down to the atomic scale and map chemical elements across extremely thin layers in a cell. Scientists have used it to visualize lithium dendrites and to probe the solid electrolyte interphase (SEI), a fragile nanometer-thick layer that influences battery life and safety.

The new work suggests, however, that the microscope and the handling steps leading to it may be reshaping samples in ways that can be mistaken for real battery behavior.

Three common workflows, three pathways to artifacts

To test the scope of the problem, the team prepared multiple sets of lithium- and sodium-based materials under controlled conditions, then varied only how samples were moved from an inert glovebox into the TEM and how they were exposed to the electron beam.

They evaluated three widely used workflows:

1. Cryogenic transfer using liquid nitrogen. While rapid cooling is intended to protect delicate structures, the team found that extreme cold can cause water from the surrounding environment to condense and freeze on the sample surface during transfer. For reactive lithium and sodium, that ice can react with the metal and leave residues that may be imaged as if they were native features.

2. Inert-gas mounting with brief air exposure during insertion. A common practice is to mount a sample under inert gas but still accept a short exposure to air while inserting the holder into the microscope. The researchers found that even this short interval can substantially change lithium metal.

“Lithium metal is so reactive that after about 15 seconds in air, the structure already looks very different,” said co-first author Shuang Bai, a postdoctoral researcher at Argonne National Laboratory and research associate at UChicago, in a university release.

1. A sealed, room-temperature holder enabling inert-gas transfer end-to-end. Using a sealed holder designed to maintain an inert environment from glovebox to microscope vacuum, the team reported they could move lithium metal into the TEM with negligible exposure to oxygen or water.

Under those sealed conditions, the researchers found that pure lithium metal can tolerate higher electron-beam doses at room temperature than many researchers assume, challenging the view that lithium must be imaged cryogenically to survive in a TEM.

“The big surprise is that lithium metal itself is relatively robust,” Bai said. “It’s the compounds on top of it—the SEI—that are extremely beam-sensitive.”

Beam damage that can masquerade as battery chemistry

The SEI forms spontaneously when lithium or sodium reacts with an electrolyte and often contains compounds such as lithium carbonate (Li₂CO₃) and lithium fluoride (LiF). The team found these compounds to be far more beam-sensitive than the underlying metal.

By varying electron dose and dose rate, they showed that LiF can decompose under intense beam exposure, producing metallic lithium that can then react with residual gas inside the microscope to form lithium oxide. Such beam-driven reaction chains can mimic processes that might occur during battery cycling, potentially confusing interpretation.

“What you see in the microscope can be something you just created,” Liu said. “If you don’t know the dose, you don’t know whether you changed it.”

A push for standards—and basic reporting

The team reviewed published TEM studies of lithium and sodium battery materials and found that most did not report key imaging parameters such as electron dose or dose rate. Without those details, they argue, it is hard to determine whether images reflect native structures or artifacts.

Their proposed guidelines include recommendations for inert-gas transfer for pure lithium and sodium metals, strict dose limits for SEI components, and more comprehensive documentation of experimental conditions. They call on researchers to report sample storage, glovebox conditions, focused ion beam preparation, transfer workflows, beam energies, imaging temperatures, and dose metrics.

“Right now, you use your method, I use my method,” Liu said. “It’s really hard to compare.”

Why it matters beyond academia

The work is part of the Energy Storage Research Alliance (ESRA), a U.S. Department of Energy “Energy Innovation Hub” launched in 2024 and led by Argonne National Laboratory. The hub is funded at $125 million to coordinate research on safer, higher-energy, longer-duration batteries using abundant materials.

Y. Shirley Meng, the Liew Family Professor in Molecular Engineering at UChicago and ESRA’s director, is the corresponding author of the Joule paper and called the findings a “milestone.”

“If we are going to rely on batteries to decarbonize our energy system, we must first have confidence in the science at the nanoscale,” Meng said in a university statement. “That starts with understanding when our tools are telling us the truth.”

Battery companies use microscopic images to guide choices about electrolytes, protective coatings, and operating conditions for electric vehicles and grid storage. If some images have been shaped by air exposure or beam damage, the study suggests, design decisions based on them may need re-examination.

The proposed standards could also influence journal policies and funding expectations, similar to the detailed metadata requirements common in structural biology. Still, implementing the full workflow may be difficult for smaller laboratories, given the cost of inert-gas transfer holders, advanced TEM systems with dose control, and cryogenic capabilities.

The authors argue that even if some labs cannot adopt every component immediately, consistent reporting of actual conditions would still be a major step toward reproducibility.

For researchers working at the intersection of electron microscopy and reactive battery materials, the study offers a checklist: How long did the sample see air? Was the dose measured? Could an apparent new phase be the result of beam-induced reactions?

Until those questions are answered, the authors contend, pictures on a microscope screen may not be a reliable map of the materials expected to power future cars—and power grids.