Researchers at the US Department of Energy’s (DOE) Argonne National Laboratory have developed and demonstrated an innovative set of methods to evaluate long-term aging in real-world battery cells. The methods are based on a phenomenon called nuclear magnetic resonance (NMR), commonly used in medical imaging. This is the first NMR spectroscopy capability that can track in fine detail how the chemistry of commercial pouch battery cells evolves over years of operation.

A paper on the subject was published in the Journal of Power Sources.

NMR spectroscopy is a nondestructive, noninvasive technique that relies on magnetic properties of atomic nuclei to study the chemical environments in a sample. A radio-frequency field is applied to a sample immersed in a strong magnetic field, causing the sample to absorb energy. Then, the radio-frequency field is removed, and a probe measures the energy released when the nuclei return to their lower energy state. The measurements provide insights about the atomic and molecular structures and reactions, including those in battery materials.

Wang et al.

Argonne’s new NMR capability is available for use by battery researchers and manufacturers.

Today’s lithium-ion batteries work by electrolytes transporting lithium ions back and forth between two electrodes, converting stored energy into electricity. Most lithium-ion batteries in electric vehicles have anodes (negative electrodes) made of graphite. However, new electrode materials with higher energy densities, such as silicon, are needed for longer driving ranges.

The application of NMR to batteries has been limited to date. But with our powerful new capability, I hope that it will become ‘bread and butter’ for researchers and manufacturers who want to probe the long-term evolution of their batteries without opening them up. We can study technologies that are already or nearly commercialized.

—Baris Key, an Argonne chemist and one of the study’s authors

Before silicon can be fully utilized in the anode, there are several technical challenges to solve. When a silicon-anode battery cell is charging, lithium ions bond with silicon to form compounds known as lithium silicides. This causes the anode to expand in volume by as much as 400%. When the cell discharges, lithium exits the anode, causing it to contract.

The expansion and contraction can cause the silicon anode to crack. Additionally, the lithium silicides are highly reactive, resulting in a much less stable interface with the cell’s electrolyte.

In the Argonne study, researchers developed and applied the NMR spectroscopy technique to observe the fate of lithium atoms in silicon-anode cells as they were charged and discharged, then allowed to rest over seven months. The technique is similar to magnetic resonance imaging, or MRIs, used in medicine to create detailed images of the body.

What we did in our study was like taking MRIs of operating battery cells, except that we didn’t produce images of the cells. Instead, the output was information on how the lithium chemical environment in the cells changed due to charging, discharging, resting and aging.

This information allowed us to determine where the lithium atoms go, how they interact with other atoms, how many lithium atoms are involved in those interactions and whether there is any associated degradation. Our goal was to understand why the silicon anodes degrade over time.

—Evelyna Wang, an Argonne postdoctoral appointee and the study’s main author

To understand better how the cells age under real-world conditions, the team applied the NMR technique while the cells were operating. This operando approach enables real-time observation of structural and electronic changes within the cell. In contrast, typical battery aging experiments evaluate chemical dynamics after operation and cell disassembly. The operando NMR method can provide an accurate picture of aging in electric vehicle batteries and other real-world devices.

Another important aspect of simulating real-world conditions was the cells themselves. Argonne’s Cell Analysis, Modeling and Prototyping facility fabricated the cells using a process comparable to commercial battery manufacturing. As a result, the cells were more standardized and had much better sealing and contacts than typical laboratory-made cells.

The team made an important discovery: After the cells charged, many lithium atoms were getting trapped in the anode. During discharge, lithium atoms remained in the anode in the form of lithium silicides rather than being removed and transported to the cathode (positive electrode).

The trapped lithium silicides accumulated in the anode, depleting the total amount of lithium available for cycling the cells. They also reacted with the electrolyte. The trapped molecules and reactions contributed to reductions in the cell’s energy-storage capacity.

The Argonne team also found that adding a magnesium salt to the electrolyte decreased the amount of trapped lithium silicides. These findings are likely to inspire new lines of research to identify different chemical additives, electrolyte formulations and silicon materials that can limit the formation of trapped lithium silicides.

A key advantage of NMR spectroscopy is that it is highly sensitive to the behavior of light elements like lithium, silicon, carbon and hydrogen that other characterization methods cannot easily probe.

The new NMR methods are thus not limited to silicon-anode batteries. They can easily be applied to other emerging battery technologies like sodium-ion and solid-state. They can also probe aging in other battery components like cathodes and electrolytes.

Resources

• Evelyna Wang, Marco-Tulio F. Rodrigues, Sohyun Park, Fulya Dogan, Baris Key, Operando NMR characterization of cycled and calendar aged nanoparticulate silicon anodes for Li-ion batteries, Journal of Power Sources, Volume 604, 2024, 234477, ISSN 0378-7753, doi: 10.1016/j.jpowsour.2024.234477