In the context of global efforts toward carbon neutrality and sustainable energy, the rapid growth of industries such as electric vehicles, energy storage systems, and portable electronics has been largely driven by advancements in li ion battery technology. Li ion batteries are favored for their high energy density, long cycle life, and lightweight design, making them integral to modern energy solutions. However, the operational safety and longevity of li ion batteries are often compromised by parasitic reactions, among which lithium plating is particularly concerning. This phenomenon occurs during charging, especially under harsh conditions like low temperatures or high charging rates, leading to the deposition of metallic lithium on the graphite anode. While some of this plated lithium may re-intercalate during rest or constant-voltage phases, a portion reacts with the electrolyte to form “dead lithium,” resulting in capacity fade and, in severe cases, internal short circuits that can trigger thermal runaway. Thus, developing reliable methods for the online detection of lithium plating is crucial for enhancing the safety and performance of li ion batteries.
Traditional approaches to detecting lithium plating in li ion batteries can be broadly categorized into physical and electrochemical methods. Physical techniques, such as microscopy or spectroscopy, offer direct visualization but are invasive, expensive, and unsuitable for real-time monitoring in field applications. Electrochemical methods, including coulombic efficiency analysis and differential voltage (dV/dQ) curves, provide indirect indicators but often require post-mortem analysis or lengthy procedures, limiting their utility for online detection. In contrast, electrochemical impedance spectroscopy (EIS) has emerged as a promising non-invasive tool due to its sensitivity to interfacial dynamics and reaction kinetics in li ion batteries. By analyzing impedance changes during operation, it may be possible to identify lithium plating onset in real time. This article explores an online detection method for lithium plating in li ion batteries based on electrochemical impedance measurement, leveraging equivalent circuit modeling, custom-designed measurement hardware, and experimental validation under various conditions.
The electrochemical behavior of a li ion battery during charging and discharging can be represented using equivalent circuit models (ECMs), which help interpret impedance spectra. A fractional-order ECM is often preferred for its accuracy in capturing the distributed nature of electrochemical processes. In such a model, key components include: an ohmic resistance $R_0$ representing electrolyte and contact resistances; a constant phase element $Q_{sei}$ in parallel with a resistance $R_{sei}$ to model the solid-electrolyte interphase (SEI) layer; another constant phase element $Q_{ct}$ in parallel with a charge transfer resistance $R_{ct}$ to describe the electrode-electrolyte interface; and a Warburg element $W$ for diffusion effects. The impedance $Z(\omega)$ as a function of angular frequency $\omega$ can be expressed as:
$$Z(\omega) = R_0 + \frac{1}{\frac{1}{R_{sei}} + Q_{sei}(j\omega)^{\alpha_{sei}}} + \frac{1}{\frac{1}{R_{ct}} + Q_{ct}(j\omega)^{\alpha_{ct}}} + W(j\omega)^{-\frac{1}{2}}$$
where $j$ is the imaginary unit, and $\alpha_{sei}$ and $\alpha_{ct}$ are dispersion coefficients. During lithium plating, additional electrochemical reactions occur at the anode. The primary intercalation reaction is:
$$x\text{Li}^+ + x\text{e}^- + \text{Li}_\delta\text{C}_6 \rightarrow \text{Li}_{\delta+x}\text{C}_6$$
Simultaneously, lithium plating involves the reduction of lithium ions to metallic lithium:
$$y\text{Li}^+ + y\text{e}^- \rightarrow y\text{Li}$$
This plated lithium may either re-intercalate later or react with the electrolyte to form dead lithium. From an impedance perspective, the plating reaction introduces an additional parallel charge transfer pathway, effectively reducing the overall charge transfer resistance. This reduction manifests as a decrease in the mid-frequency impedance magnitude, providing a potential signature for lithium plating onset. Thus, monitoring impedance at characteristic frequencies sensitive to charge transfer processes can enable online detection in li ion batteries.
To achieve real-time impedance measurement, a dedicated online measurement device was designed and constructed. The system generates a small-amplitude alternating current (AC) excitation signal—typically 0.05C of the battery’s capacity to ensure linearity—and measures the voltage response. Key components include a direct digital synthesizer (DDS) for signal generation, a bipolar voltage-controlled current source for excitation, a four-wire measurement setup to minimize lead resistance effects, and a digital lock-in amplifier-based processing circuit for precise impedance calculation. The device computes impedance magnitude $|Z|$ and phase angle $\phi$ at selected frequencies by performing cross-correlation between the response and reference signals. Validation tests against a commercial electrochemical workstation showed excellent agreement, with errors below 0.5% for resistive loads and minimal deviations for actual li ion batteries, confirming the device’s reliability for online monitoring.
The experimental platform integrates this impedance measurement device with a battery cycler, a thermal chamber for temperature control, and data acquisition systems. Tests were conducted on 18650-type li ion batteries with NCA cathodes and graphite anodes, having a nominal capacity of 3500 mAh. Prior to experiments, all li ion batteries underwent conditioning cycles to ensure consistency. Static EIS measurements were first performed at various temperatures (-5°C, 5°C, 15°C, 25°C) and states of charge (SOC: 10%, 30%, 50%, 70%, 90%) using a commercial workstation to establish baseline behavior. The results, summarized in Table 1, indicate that impedance decreases with increasing temperature and SOC, primarily due to enhanced charge transfer kinetics. The charge transfer resistance $R_{ct}$, extracted via ECM fitting, shows significant reduction at low SOCs and stabilizes at higher SOCs.
| Temperature (°C) | SOC (%) | $R_0$ (mΩ) | $R_{ct}$ (mΩ) | Notes |
|---|---|---|---|---|
| -5 | 10 | 25.1 | 180.5 | High impedance at low temperature |
| -5 | 50 | 24.8 | 92.3 | Impedance decreases with SOC |
| 5 | 10 | 24.9 | 120.7 | Moderate temperature effect |
| 5 | 50 | 24.7 | 65.4 | Further reduction in $R_{ct}$ |
| 15 | 10 | 24.8 | 85.2 | Kinetic improvement |
| 15 | 50 | 24.6 | 48.9 | Low charge transfer resistance |
| 25 | 10 | 24.7 | 60.1 | Near-optimal conditions |
| 25 | 50 | 24.5 | 32.5 | Minimal impedance |
Dynamic impedance measurements during charging were then conducted using the online device. Under normal conditions (25°C, 0.2C charge), impedance magnitude at various excitation frequencies (5 Hz, 10 Hz, 20 Hz, 30 Hz) decreased gradually with increasing SOC, as shown in Figure 1 (note: figures are not referenced by number in text). The 10 Hz frequency was selected as the characteristic excitation frequency for lithium plating detection, as it lies within the mid-frequency range (1–70 Hz) most sensitive to charge transfer processes in this li ion battery type. The impedance evolution during charging can be modeled empirically as:
$$|Z(t)| = Z_0 \exp(-\beta \cdot \text{SOC}(t)) + \gamma$$
where $Z_0$ is the initial impedance, $\beta$ is a decay constant, and $\gamma$ is an offset. Under normal operation, this decay is smooth, but deviations may indicate side reactions.
To induce lithium plating, tests were performed under low-temperature conditions (-10°C, -5°C, 0°C, 5°C) with varying charge rates (0.1C to 1C). The experimental matrix is outlined in Table 2. For each test, the online impedance measurement device recorded the impedance magnitude at 10 Hz during constant-current charging. The results revealed that in most cases, after an initial gradual decrease, the impedance exhibited an accelerated decay phase—a deviation from the expected trend. This accelerated decay is hypothesized to correspond to the onset of lithium plating, as it aligns with the reduction in charge transfer resistance due to parallel plating reactions. The onset SOC for lithium plating was identified as the point where the impedance decay rate changed abruptly, calculated using the derivative of impedance with respect to capacity:
$$\frac{d|Z|}{dQ} = \lim_{\Delta Q \to 0} \frac{|Z(Q+\Delta Q)| – |Z(Q)|}{\Delta Q}$$
A negative spike in this derivative indicates accelerated decay. For instance, at -10°C and 0.75C charging, the onset occurred at approximately 12% SOC, while at milder conditions (e.g., 5°C and 0.25C), no such feature was observed, suggesting no plating. The extracted onset SOCs are summarized in Table 3, demonstrating that lithium plating initiates earlier at lower temperatures and higher charge rates, consistent with known mechanisms in li ion batteries.
| Temperature (°C) | Charge Rates (C-rate) | Discharge Rate | Purpose |
|---|---|---|---|
| -10 | 0.1C, 0.25C, 0.5C, 0.75C | 0.2C | Study extreme cold effects |
| -5 | 0.1C, 0.25C, 0.5C, 0.75C, 1C | 0.2C | Moderate low temperature |
| 0 | 0.1C, 0.25C, 0.5C, 0.75C, 1C | 0.2C | Near-freezing conditions |
| 5 | 0.25C, 0.5C, 0.75C, 1C | 0.2C | Cool but above freezing |
| Temperature (°C) | Charge Rate (C) | Onset SOC (%) | Impedance Decay Rate Change |
|---|---|---|---|
| -10 | 0.1 | 45 | Yes |
| -10 | 0.25 | 23.5 | Yes |
| -10 | 0.5 | 17 | Yes |
| -10 | 0.75 | 12 | Yes |
| -5 | 0.1 | None | No |
| -5 | 0.25 | 43 | Yes |
| -5 | 0.5 | 32 | Yes |
| -5 | 0.75 | 27 | Yes |
| -5 | 1 | 20 | Yes |
| 0 | 0.1 | None | No |
| 0 | 0.25 | 53 | Yes |
| 0 | 0.5 | 41 | Yes |
| 0 | 0.75 | 31 | Yes |
| 0 | 1 | 28 | Yes |
| 5 | 0.25 | None | No |
| 5 | 0.5 | 46 | Yes |
| 5 | 0.75 | 34 | Yes |
| 5 | 1 | 31 | Yes |
To validate the impedance-based detection method, differential voltage analysis (dV/dQ) was employed as a reference. After each charge-discharge cycle, the discharge voltage curve was analyzed. The presence of a voltage plateau in the early discharge phase, accompanied by a “lithium stripping peak” in the dV/dQ curve, indicates reversible lithium plating. The amount of stripped lithium $Q_{\text{strip}}$ can be estimated from the peak area. Results confirmed that in cases where impedance showed accelerated decay, dV/dQ curves exhibited clear stripping peaks, whereas no peaks appeared when impedance decay was normal. For example, at -10°C and 0.75C charging, the stripped lithium amounted to 480 mAh (14.3% of capacity), while at 5°C and 0.25C, no stripping was detected. This correlation underscores the reliability of the impedance method for online lithium plating detection in li ion batteries.
The underlying mechanism can be further elucidated through the modified equivalent circuit model during plating. When lithium plating occurs, an additional charge transfer resistance $R_{ct,\text{Li}}$ appears in parallel with $R_{ct}$, reducing the overall impedance $Z_{ct}$ of the charge transfer element:
$$\frac{1}{Z_{ct}} = \frac{1}{R_{ct}} + \frac{1}{R_{ct,\text{Li}}} + Q_{ct}(j\omega)^{\alpha_{ct}}$$
Since $R_{ct,\text{Li}}$ is typically smaller than $R_{ct}$ due to the facile plating reaction, $Z_{ct}$ decreases, leading to a drop in total impedance at mid-frequencies. The accelerated decay in impedance magnitude thus serves as a real-time indicator. Moreover, the method is non-invasive and can be integrated into battery management systems (BMS) for continuous monitoring of li ion batteries, enabling proactive safety measures.
In conclusion, this study presents an online detection method for lithium plating in li ion batteries based on electrochemical impedance measurement. By leveraging a custom-designed impedance measurement device and analyzing impedance at a characteristic frequency of 10 Hz, the onset of lithium plating is identified through an accelerated decay signature in the impedance magnitude during charging. Experimental results under various low-temperature and charging-rate conditions demonstrate that this method effectively detects plating, with validation via differential voltage analysis. The findings highlight that lithium plating in li ion batteries is more prone to occur at lower temperatures and higher charge rates, and the impedance-based approach offers a practical, real-time solution for enhancing battery safety. Future work should extend this method to other stress conditions, such as overcharging or high-rate cycling, to develop a comprehensive online monitoring framework for li ion batteries in diverse applications. Ultimately, the integration of such techniques can contribute to the safer and more efficient utilization of li ion batteries across energy storage and transportation sectors.