Lithium-ion batteries power everything from electric vessels to smartphones and home energy systems. Yet, like all technology, they age.
In my ten years of experience in power testing across industries from automotive and marine to aerospace and grid storage I’ve seen the same question come up time and again: “How long will this battery actually last?” Companies invest millions in battery systems, only to face unexpected degradation, premature failures, and costly warranty disputes. The core issue is visibility. Traditional capacity testing shows when performance is declining, but not why. Without insight into the underlying mechanisms, it’s impossible to prevent or mitigate the damage. The result? Higher costs, longer development cycles, and increased risk in mission-critical applications.
Electrochemical Impedance Spectroscopy (EIS) has emerged as a vital tool to monitor, diagnose, and even predict battery health. By measuring how a battery responds to electrical signals across a range of frequencies, EIS provides insight into the inner workings of the cell, revealing which degradation mechanisms are active and how they will evolve over time.
This article explores the EIS, its applications in understanding battery aging, and concrete examples of how researchers link impedance signatures to specific physical processes.
Battery degradation is multi-faceted. Electrodes accumulate structural damage, solid-electrolyte interphase (SEI) layers grow, lithium may plate under unfavourable conditions, and ion transport can slow due to blocked pores or particle cracking. These mechanisms manifest as capacity fade, reduced power delivery, and sometimes hazardous conditions.
For manufacturers, fleet operators, and energy system designers, understanding which degradation mechanisms are active and when can inform material selection, thermal management, and charging strategies. Predictive insights also allow proactive maintenance, extending battery life and improving safety.
EIS provides this insight by offering a non-destructive, frequency-resolved window into a battery’s true quality and state. Unlike conventional voltage or current measurements, EIS can separate fast, medium, and slow processes and track how each evolves during the life of the cell.
EIS applies a small alternating current or voltage across a battery and measures the resulting response. By sweeping across frequencies from high to low, it produces an impedance spectrum that encodes information about multiple processes occurring inside the battery.
This frequency decomposition allows engineers to assign changes in the spectrum to specific physical mechanisms. For example, a growing mid-frequency semicircle may indicate SEI thickening, while increased low-frequency impedance could reflect particle cracking or diffusion limitations.
Beyond visual inspection, EIS can be coupled with modeling to produce quantitative insights. Equivalent circuit models (ECMs) and physics-based impedance models map spectral features to resistances, capacitances, and diffusion coefficients. More advanced approaches, like the distribution of relaxation times (DRT), can separate overlapping processes that traditional Nyquist plots might obscure.
Dynamic techniques, such as dynamic EIS (DEIS), allow measurement while the battery is operating, providing time-resolved insight into fast or transient degradation events. Machine-learning models can also ingest entire spectra to predict capacity fade or remaining useful life, even when the precise physical attribution of features is complex.
Despite its power, EIS has limitations. The technique is sensitive to measurement conditions like state-of-charge, temperature, prior cycling, and perturbation amplitude all affect the spectrum. Model fitting can be ambiguous; multiple equivalent circuits can reproduce similar spectra, making physical interpretation challenging. Commercial batteries with proprietary formulations and complex geometries need further complicates analysis, as a single spectrum represents a cell-wide average. Finally, low-frequency measurements are time-consuming, which can limit field or real-time applications.
Nevertheless, methodological advances like DRT, physics-informed modeling, DEIS, and hybrid machine-learning approaches, mitigate many of these challenges, making EIS an increasingly reliable and practical tool for battery diagnostics.
Understanding degradation in controlled laboratory conditions is one thing. Knowing what happens under actual driving is another. Researchers used EIS modeling to simulate how different mechanisms manifest under realistic drive cycles, for example the mix of acceleration, cruising, regenerative braking, and fast charging that batteries experience daily (3).
They discovered that SEI growth, lithium plating, and particle cracking each produce distinct signatures in specific frequency ranges. By measuring impedance on cells cycled under standard drive profiles, they validated these signatures and created diagnostic frameworks for identifying which mechanisms dominate under different usage patterns (3).
Batteries don't fail from a single cause like protective films grow, charge-transfer gets harder, and lithium diffusion slows down, all happening at once. Traditional analysis couldn't effectively separate these overlapping effects, but the Distribution of Relaxation Times method changed that.
By decomposing complex spectra into individual peaks with characteristic timescales, researchers now quantify exactly how much each mechanism contributes to decline (1). When an EV battery loses capacity, engineers can determine whether 40% comes from protective film growth, 35% from reduced charge transfer, and 25% from diffusion limitations, and then prioritize solutions accordingly.
Researchers developed pattern-recognition systems that identify characteristic impedance signatures associated with specific failure trajectories. By correlating these patterns with post-failure analysis of thousands of cells, they created a diagnostic database of what failure looks like in impedance terms (2).
Another study analyzed thousands of EIS spectra from commercial Li-ion cells, feeding them into a machine-learning pipeline. The model learned which spectral patterns predicted capacity loss and remaining life, without manual circuit fitting. This work demonstrated that EIS contains rich predictive information, enabling SoH forecasting even when exact physical interpretation is challenging
Battery degradation isn't just chemistry. Mechanical damage plays a huge role. Research on graphite anodes revealed that physical structure breaks down over repeated lithium insertion and extraction cycles. Bonds break, surfaces become disordered, and this structural damage creates new reactive sites that accelerate chemical degradation (4).
A mathematical approach analyzed Nyquist plots to extract chord lengths, semicircle intercepts, and slopes as proxies for ohmic, interfacial, and diffusion resistances. Trends in these geometric features reliably reflected SEI growth and transport limitations. This method offers a practical, fast way to monitor battery health when full circuit modeling is impractical, such as in high-throughput testing or field applications
Many dangerous battery conditions develop during normal operation, not while resting. But conventional EIS requires equilibrium, making it blind to these critical events. Dynamic EIS solves this by performing impedance analysis fast enough to capture what's happening during actual charge and discharge cycles.
This real-time capability detects impedance anomalies associated with internal short circuits, lithium plating during aggressive fast charging, and early stages of thermal runaway—catching problems while they're still developing.
Electrochemical impedance spectroscopy is transitioning from telling us what went wrong after batteries fail to predicting failures before they happen and guiding interventions that prevent them.
For battery manufacturers, EIS is becoming a quality control tool that catches problematic cells before they leave the factory. Testing every cell could flag the 2% with impedance signatures indicating premature failure before they reach customers.
For automotive companies, EIS-based diagnostics promise smarter battery management systems that optimize charging and operation based on each battery's actual internal condition. Your EV could adjust its fast-charging strategy based on real-time impedance measurements showing early signs of lithium plating.
For marine companies, EIS-based diagnostics promise intelligent battery management systems that optimize charging and operation based on each battery's actual internal condition in the demanding maritime environment. Vessels can adjust their fast-charging strategies at port based on real-time impedance measurements showing early signs of lithium plating, while onboard systems continuously monitor battery health during long voyages where saltwater exposure, vibration, and temperature fluctuations create unique degradation challenges. For electric ferries, hybrid propulsion systems, and all-electric vessels, EIS diagnostics enable predictive maintenance scheduling that prevents at-sea failures and maximizes operational availability, that are critical factors where downtime means missed sailings and lost revenue.
For grid storage operators, where a single failed system can cost millions, EIS provides predictive maintenance capabilities that schedule service when needed, preventing costly failures and fire hazards.
For researchers and engineers, EIS provides the mechanistic understanding needed to design fundamentally better batteries. When you can see which degradation processes limit lifespan under real conditions, you can engineer solutions that address root causes.
The technology continues evolving rapidly. Machine learning algorithms are getting better at automatically identifying degradation patterns. Dynamic EIS methods are becoming fast enough for real-time onboard diagnostics. Physics-based models are becoming sophisticated enough to extract quantitative materials properties from impedance measurements. Integration with battery management systems is deepening, bringing laboratory diagnostic capabilities into production vehicles and stationary storage.
As batteries continue to power the clean energy transition, advanced testing tools Terma Battery Cell Tester will remain a cornerstone of innovation by helping us build devices that last longer, perform better, and operate safely under a wide range of conditions. Its modular architecture scales from single-channel benchtop configurations for research labs to multi-channel systems for high-throughput manufacturing lines.
By making sophisticated EIS analysis accessible and practical for industrial applications, Terma enables the battery industry to transform scientific insights into real-world improvements in quality, reliability, and sustainability.