Mit decodes acoustic signatures of battery formation and degradation to aid development of new monitoring devices

Before batteries run out of power, suddenly fail, or catch fire, they often emit subtle sounds over time, which are characteristic markers of the deterioration process within the battery's internal structure. However, so far, no one has been able to accurately interpret the meaning of these sounds or distinguish between ordinary background noise and important signs of potential faults.

Image Source: Massachusetts Institute of Technology

According to foreign media reports, a research team from the Department of Chemical Engineering at the Massachusetts Institute of Technology (MIT) conducted a detailed analysis of the sounds emitted by lithium-ion batteries. They successfully correlated specific acoustic patterns with particular degradation processes occurring within the battery. The research findings were published in the journal Joule. This new discovery is expected to lay the foundation for the development of relatively simple, completely passive, and non-destructive monitoring devices. Such devices could continuously monitor the health of battery systems in electric vehicles or grid-scale energy storage facilities, thereby predicting effective service life and providing early warnings of failures.

Bazant stated, "In this study, through rigorous scientific work, our team successfully deciphered the acoustic emission signals. We were able to classify these signals as being caused by bubbles generated by side reactions, or by cracks formed due to the expansion and contraction of active materials, and even identify the characteristics of these signals amidst noisy data."

Samantaray explained, "The core of this work lies in exploring a non-destructive method to study the internal mechanisms of batteries during charging and discharging. There are currently a few research methods available, but most are costly and difficult to apply to conventional battery forms."

To achieve the analysis objectives, the research team conducted electrochemical testing and acoustic emission recording simultaneously under real charging and discharging conditions. Through refined signal processing, they established a correlation between the electrical and acoustic data. Samantaray pointed out, "As a result, we have developed a cost-effective and efficient method to truly understand the gas generation and fracture processes of materials."

Gas generation and material fracture are the two core mechanisms of battery degradation failure. Monitoring the sounds emitted by the battery can detect and distinguish these processes, which will be an important tool for battery system managers.

Previous methods only monitored sound and recorded the time points when the overall sound level exceeded a threshold. However, Bazant pointed out that this study simultaneously monitored voltage, current, and acoustic characteristics. "We found that acoustic emissions occur at specific potentials (voltages), which helps identify the specific processes that trigger acoustic emissions.”

After the testing is completed, the research team will disassemble the battery and observe under an electron microscope whether the materials have fractured.

In addition, the research team also employed wavelet transform technology. This method can encode the frequency and duration of each captured signal, forming unique feature markers, making it easier to extract signals from background noise. "No one has tried this method before," Bazant emphasized, "This is undoubtedly another major breakthrough."

Bazant pointed out that acoustic emission technology is widely used in engineering fields, such as monitoring early signs of damage in structures like bridges. "It's an excellent way to monitor the state of systems," Bazant stated, "because these acoustic emissions continue to exist whether you listen to them or not." By listening, people are able to gain insight into internal mechanisms that are invisible to the naked eye.

When discussing batteries, Bazant stated, "We often find it difficult to accurately interpret voltage and current data to precisely understand the internal condition of the battery. Acoustic emission technology provides us with another window of observation, enabling us to assess the battery's health status and remaining service life, as well as ensuring operational safety."

In a related paper published in the Journal of Energy Storage, in collaboration with researchers from Oak Ridge National Laboratory, the team demonstrated that acoustic emissions can provide early warning of thermal runaway, which, if not detected in time, could lead to fires. The new research shows that these sound waves can also be used to detect gas generation before combustion. Bazant stated, "It's like observing the initial tiny bubbles long before the kettle boils."

The next step will be to develop practical and affordable monitoring systems based on new insights into the correlation between specific sounds and particular conditions. For example, the team has received funding from Tata Motors to develop a battery monitoring system for their electric vehicles. "We now understand which indicators to focus on and how to relate them to lifespan, health status, and safety," said Bazant.

Samantaray pointed out that one potential application of this new discovery is "as a laboratory tool to help teams developing new materials or testing new environments detect gas generation or active substance fracture without disassembling the battery."

Bazant believes that the system is also valuable for quality control in battery manufacturing, and explains: "The most expensive and speed-limiting part of battery production is often the formation cycling process." This process involves repeatedly charging and discharging the battery to condition it, during which some chemical reactions release gases. Bazant states that the new system can detect the formation characteristics of these gases, "and by sensing these characteristics, it might be possible to distinguish well-formed batteries from poorly-formed ones earlier in the manufacturing stage, before the actual battery lifespan begins."