1. Introduction and Research Motivation
The global strategic push toward carbon neutrality has accelerated the electrification of the transportation sector. As a result, the penetration rate of electric vehicles (EVs) has surged dramatically over the past decade. This explosive growth, however, brings forth a critical challenge: the disposal and reuse of retired power batteries. When the capacity of a lithium-ion battery in an EV decays to 70%–80% of its initial value, it is no longer suitable for high-demand automotive applications. Nevertheless, these retired batteries retain substantial residual value and can be repurposed for second-life applications, such as stationary energy storage, low-speed vehicles, and grid peak shaving.
This research focuses on the thermal safety and thermal management of retired ternary lithium-ion batteries with different nickel contents. The high energy density of ternary cathode materials, particularly those with high nickel fractions, comes at the cost of reduced thermal stability. Among the various cathode chemistries, Li(Ni₀.₈Co₀.₁Mn₀.₁)O₂ (NCM811) offers superior specific capacity but exhibits the poorest thermal stability, making it more susceptible to thermal runaway (TR). Therefore, systematic investigation into the thermal-electrochemical behavior of these cells and the development of effective thermal management strategies is of paramount importance for ensuring the safe deployment of second-life EV battery pack systems.
2. Screening and Reusability Assessment of Retired Cells
This study commenced with a comprehensive screening of 150 retired 18650-type cells, comprising three different chemistries: NCM523, NCM622, and NCM811. The screening process involved a multi-stage evaluation protocol, beginning with external visual inspection followed by electrical performance characterization. Each cell’s technical specifications are summarized in Table 1.
Table 1. Technical specifications of cells with different nickel contents
| Items | NCM523 | NCM622 | NCM811 |
|---|---|---|---|
| Nominal voltage/capacity (V/Ah) | 3.6 / 2.6 | 3.6 / 2.75 | 3.6 / 3.05 |
| Charge cut-off voltage (V) | 4.2 | 4.2 | 4.2 |
| Discharge cut-off voltage (V) | 2.75 | 2.5 | 2.5 |
| Maximum discharge rate (C) | 3 | 3 | 2 |
| Mass energy density (Wh/kg) | 180 | 206 | 234 |
| Weight (g) | 46.5 | 49 | 47 |
| Internal resistance (mΩ) | ≤40 | ≤35 | ≤35 |
Through the initial visual screening, cells exhibiting cracks, corrosion, or electrolyte leakage were rejected. The results showed that NCM811 cells had the most severe external damage. Following the full screening cascade, which included voltage and internal resistance verification, 130 cells (86.7%) were deemed qualified. The remaining cells failed due to various reasons: 10 cells (6.7%) displayed physical damage, 7 cells (4.7%) exhibited excessively low or zero voltage, and 3 cells (2.0%) had high internal resistance.
The capacity consistency of the screened cells was evaluated through three consecutive charge-discharge cycles. Figures of merit included the average discharge capacity and the variance within each chemistry group. The NCM523 cells exhibited a mean capacity of 2414.89 mAh, NCM622 reached 2711.16 mAh, while NCM811 delivered 2790.15 mAh. The NCM811 cells showed a capacity advantage of 15.5% over NCM523 and 2.9% over NCM622. The state-of-health (SOH) metrics were equally encouraging, with 90.8% of cells exhibiting a capacity retention rate above 95% and 79.0% achieving a capacity recovery rate exceeding 90%.
Two important parameters for assessing battery health are capacity retention rate and capacity recovery rate, calculated as follows:
$$ \text{Capacity retention rate} = \frac{C_1}{C_0} \times 100\% \quad (2.1) $$
$$ \text{Capacity recovery rate} = \frac{C_2}{C_0} \times 100\% \quad (2.2) $$
where C₀ is the initial capacity before storage, C₁ is the first discharge capacity after storage, and C₂ is the third discharge capacity after storage.
Voltage curve consistency was investigated by monitoring the charge-discharge profiles of the three chemistries. In the charging phase, all cells exhibited a gradual voltage rise. However, NCM811 reached the same voltage plateau faster than NCM622 and NCM523. Conversely, during discharge, the voltage curves remained relatively flat until approximately 2300 seconds, after which the discharge rate notably accelerated, with NCM811 completing the discharge most rapidly. This behavior suggests that higher nickel content correlates with faster charge-discharge kinetics but also accelerated degradation during prolonged cycling.
3. Thermal-Electrochemical Behavior Under Adiabatic Conditions
This section investigates the thermal-electrochemical performances of the EV battery pack cells under adiabatic conditions using an Extended Volume Accelerating Rate Calorimeter (EV-ARC). The experimental protocol involved charging cells to 100% state of charge (SOC) at 1.0C, followed by discharging at 0.5C, 1.0C, and 2.0C rates under two ambient temperatures (25°C and 55°C). The ARC operates by maintaining adiabatic conditions, ensuring the measured temperature rise accurately reflects internal heat generation.
3.1 Temperature Rise at Room Temperature (25°C)
The temperature evolution characteristics at 25 °C under different discharge rates are summarized in Table 2.
Table 2. Temperature rise characteristics at 25°C under different discharge rates
| Discharge Rate | Parameter | NCM523 | NCM622 | NCM811 |
|---|---|---|---|---|
| 0.5C | Max Temp (°C) | 37.0 | 39.0 | 34.9 |
| Average Temp Rise (°C) | 12.0 | 14.0 | 9.9 | |
| Temp Rise Rate (°C/min) | 0.095 | 0.123 | 0.087 | |
| 1.0C | Max Temp (°C) | 42.9 | 47.3 | 45.4 |
| Average Temp Rise (°C) | 17.9 | 22.3 | 20.4 | |
| Temp Rise Rate (°C/min) | 0.285 | 0.388 | 0.355 | |
| 2.0C | Max Temp (°C) | 56.7 | 68.8 | 70.9 |
| Average Temp Rise (°C) | 31.7 | 43.8 | 45.9 | |
| Temp Rise Rate (°C/min) | 1.026 | 1.436 | 1.556 |
At low discharge rates (0.5C and 1.0C), the heat dissipation rate exceeded the heat generation rate due to the extended discharge duration, resulting in minimal temperature differences among the three chemistries. However, at 2.0C discharge, significant divergence emerged. The NCM811 cell exhibited the highest peak temperature of 70.9 °C with a temperature rise rate of 1.5559 °C/min. Remarkably, this represented a 25% increase in maximum temperature and a 51.6% increase in temperature rise rate compared with NCM523. The NCM811 cell also exhibited the shortest discharge duration (29.5 min), indicating that high-nickel cathodes generate more internal heat under high-current conditions.
3.2 Temperature Rise at High Temperature (55°C)
To investigate the elevated-temperature behavior of the EV battery pack cells, similar experiments were performed at 55 °C, and the outcomes are shown in Table 3.
Table 3. Temperature rise characteristics at 55°C under different discharge rates
| Discharge Rate | Parameter | NCM523 | NCM622 | NCM811 |
|---|---|---|---|---|
| 0.5C | Max Temp (°C) | 59.6 | 61.9 | 59.3 |
| Average Temp Rise (°C) | 6.7 | 7.2 | 4.3 | |
| Temp Rise Rate (°C/min) | 0.053 | 0.063 | 0.038 | |
| 1.0C | Max Temp (°C) | 68.4 | 70.6 | 71.1 |
| Average Temp Rise (°C) | 13.4 | 17.2 | 16.1 | |
| Temp Rise Rate (°C/min) | 0.213 | 0.299 | 0.280 | |
| 2.0C | Max Temp (°C) | 78.6 | 91.5 | 96.6 |
| Average Temp Rise (°C) | 23.6 | 36.5 | 41.6 | |
| Temp Rise Rate (°C/min) | 0.761 | 1.217 | 1.460 |
The trends at 55 °C mirrored those observed at 25 °C but were more pronounced. At the 2.0C discharge rate, the NCM811 cell reached a maximum temperature of 96.6 °C, which was 18.0 °C higher than NCM523 and 5.1 °C higher than NCM622. More notably, the temperature rise rate of NCM811 (1.4597 °C/min) was 1.92 times that of NCM523 and 1.2 times that of NCM622. This indicates that the heat generation rate of high-nickel cells escalates disproportionately at elevated operating temperatures, presenting severe challenges for thermal management of high-nickel EV battery pack modules.
The fundamental cause of this behavior can be traced to the charged cathode material. At high SOC, the Ni⁴⁺ ions present in the delithiated NCM lattice are thermodynamically unstable. Upon exposure to elevated temperatures, these highly reactive species readily decompose, releasing oxygen and generating significant amounts of heat. This parasitic side reaction, coupled with accelerated electrolyte decomposition at high temperatures, contributes to the escalating heat generation observed in NCM811 cells.
Furthermore, the average temperature rise under adiabatic conditions directly correlates with the internal resistance and reaction enthalpy of the electrode materials. Ohmic heating, described by Joule’s law:
$$ Q_{\text{ohmic}} = I^2 R_{\text{int}} t \quad (3.1) $$
as well as polarization heat and reaction heat, all scale with internal resistance. As the nickel content increases, the internal resistance typically increases after cycling due to microcrack formation and structural degradation, further amplifying the heat output.
4. Thermal Runaway Characteristics and Mechanism
Thermal runaway (TR) represents the most severe safety failure mode in lithium-ion batteries. To evaluate the thermal abuse tolerance of the different-nickel-content cells, I conducted a series of TR experiments using the ARC in “Heat-Wait-Search” (HWS) mode, heating fully charged cells (SOC = 100%) at a rate of 2 °C/min to trigger TR.
During the TR process, three critical temperatures are defined: T₁ (self-heating onset temperature), T₂ (TR trigger temperature), and T₃ (maximum temperature). The voltage evolution was monitored concurrently with temperature, providing insight into the internal short-circuit behavior preceding TR.
4.1 Thermal Runaway of NCM523 Cell
For the NCM523 cell, the self-heating onset temperature T₁ was detected at 81.3 °C, corresponding to the initiation of solid-electrolyte interphase (SEI) decomposition. In the initial phase, the chamber temperature rose uniformly at 2 °C/min as heat was transferred to the cell by convection. As the temperature reached approximately 119.5 °C, a noticeable perturbation occurred: the voltage abruptly dropped to around 2 V before oscillating. This behavior suggests local separator shrinkage, which transiently closed Li⁺ transport pathways and induced mild internal micro-short circuits. Concurrently, there was a temporary temperature dip, attributed to the safety vent opening. High-temperature gases (including electrolyte vapor and decomposition byproducts) were expelled, carrying away significant sensible heat.
Following venting, the internal active materials were directly exposed to the ambient atmosphere, accelerating parasitic reactions and heat generation. Once the internal temperature exceeded 184.3 °C (T₂), the cell entered an irreversible TR cascade. The peak temperature rise rate reached 4264.71 °C/min, with T₃ momentarily reaching 574.94 °C.
4.2 Thermal Runaway of NCM622 Cell
The NCM622 cell displayed a similar TR evolution pattern. The T₁ was detected at 80.8 °C. The voltage signature during heating showed a similar behavior, dropping to approximately 2 V and then oscillating as the temperature approached the critical threshold. The thermal event at 119.7 °C corresponded to the safety vent opening. The TR trigger temperature T₂ was determined to be 171.7 °C. Once TR was initiated, the temperature escalated at a remarkable peak heating rate of 9069.1 °C/min, reaching a maximum temperature T₃ of 670.63 °C. The temperature increase was 645.63 °C, corresponding to a 25.83-fold elevation above T₂.
4.3 Thermal Runaway of NCM811 Cell
The TR behavior of the NCM811 cell exhibited the most concerning characteristics. The self-heating onset temperature T₁ was detected at 78.4 °C, which was lower than that of both NCM523 (81.3 °C) and NCM622 (80.8 °C). At 120.7 °C, a significant temperature drop was observed due to cap rupture. However, the subsequent TR was catastrophic. The T₂ was reached at only 139.9 °C: notably 47 °C and 31.8 °C lower than for NCM523 and NCM622 respectively. This lower TR trigger temperature signifies reduced thermal stability with increasing nickel content. Following the onset of TR, the temperature rise was explosive, with the peak heating rate soaring to 9420.33 °C/min. The maximum temperature T₃ reached 676.44 °C, an increase of 651.44 °C over T₂, representing a 26.06-fold amplification.
The TR characteristics of all three cells are comprehensively compared in Table 4.
Table 4. Critical TR parameters of NCM523, NCM622, and NCM811 cells
| Parameter | NCM523 | NCM622 | NCM811 |
|---|---|---|---|
| T₁ – Self-heating onset (°C) | 81.3 | 80.8 | 78.4 |
| T₂ – TR trigger temp (°C) | 184.3 | 171.7 | 139.9 |
| T₃ – Peak temperature (°C) | 574.94 | 670.63 | 676.44 |
| Peak heating rate (°C/min) | 4264.71 | 9069.1 | 9420.33 |
| ΔT (T₃ − T₂) (°C) | 390.64 | 498.93 | 536.54 |
The lower T₂ value for the NCM811 cell indicates that high-nickel cathode materials possess a lower activation energy for exothermic decomposition. The mechanism is primarily governed by the release of lattice oxygen from the highly delithiated, nickel-rich cathode at relatively low temperatures. This oxygen then reacts exothermically with the organic electrolyte, accelerating the chain reaction. The overall TR process can be described by the following sequential reactions:
$$ \text{SEI decomposition:} \quad (\text{CH}_2\text{OCO}_2\text{Li})_2 \rightarrow \text{Li}_2\text{CO}_3 + \text{C}_2\text{H}_4 + \text{CO}_2 + \frac{1}{2}\text{O}_2 \quad (4.1) $$
$$ \text{Anode-electrolyte reaction:} \quad 2\text{Li} + \text{C}_3\text{H}_4\text{O}_3 (\text{EC}) \rightarrow \text{Li}_2\text{CO}_3 + \text{C}_2\text{H}_4 \quad (4.2) $$
$$ \text{Cathode decomposition:} \quad \text{Li}_{0.2}\text{Ni}_{0.8}\text{Co}_{0.1}\text{Mn}_{0.1}\text{O}_2 \rightarrow \text{Li}_{0.2}\text{NiO}_2 + \frac{1}{3}\text{Co}_3\text{O}_4 + \frac{1}{5}\text{MnO} + \frac{2}{3}\text{O}_2 \uparrow \quad (4.3) $$
These exothermic reactions are self-sustaining once initiated, releasing substantial heat, generating flammable gases, and potentially leading to fire or explosion.
5. Post-Thermal Runaway Microscopic Analysis
To fundamentally understand the damage mechanisms and correlate the macro-scale TR observations with microstructural degradation, I conducted extensive post-mortem analyses of the cells.
The severity of external damage was found to escalate with nickel content. NCM811 displayed significantly more pronounced damage than NCM523 and NCM622, in complete agreement with the higher T₃ value. The TR process caused severe delamination, electrolyte leakage, and disintegration of the electrode structure. Macroscopic observation of the negative electrodes revealed that NCM811 exhibited the most severe damage, losing its original rolled structure entirely and becoming extremely fragile. In contrast, NCM523 retained much of its structural integrity.
Scanning Electron Microscopy (SEM) analysis was performed on the anode materials before and after TR. Prior to TR, the active materials of NCM523, NCM622 and NCM811 exhibited tightly packed, uniformly distributed granular structures. After the TR event, the surface morphology suffered irreversible damage: active material particles became loose, and significant gaps appeared. The degree of morphological disruption increased with nickel content. For the NCM811 sample, the internal active material structure was completely distorted, and many particles had detached from the copper current collector. The microscopic separation was most severe for NCM811, aligning well with its demonstrated low thermal stability.
The irreversible damage to active materials directly explains the electrochemical performance degradation observed in thermally abused cells. The destruction of the crystal structure and the loss of electrical contact between active particles prevent normal Li⁺ intercalation and de-intercalation, significantly compromising the energy storage capacity. This phenomenon also increases electrode polarization resistance, resulting in higher heat generation during subsequent charge-discharge operations. The detailed microscopic findings reveal that morphological and structural degradation are the root causes of both thermal performance deterioration and capacity fade in post-TR cells.
6. Development of High-Insulation and High-Thermal-Conductivity Composite Phase Change Materials
Managing thermal excursions within a safe temperature range is essential for high-nickel EV battery pack modules. Although phase change material (PCM) cooling is an attractive passive thermal management strategy, the inherent low thermal conductivity (≈0.2 W·m⁻¹·K⁻¹) and poor electrical insulation of commonly used PCMs limit their applicability. To address these bottlenecks, I developed a novel silicon nitride (Si₃N₄)-based composite phase change material (CPCM) using paraffin (PA), expanded graphite (EG), epoxy resin (ER), and Si₃N₄ particles via a melt-gelation method.
The compositions of the six synthesized CPCM groups are detailed in Table 5.
Table 5. Composition and properties of synthesized CPCMs
| Sample | PA Mass (%) | EG Mass (%) | ER Mass (%) | Si₃N₄ Mass (%) | Thermal Conductivity (W·m⁻¹·K⁻¹) | Volume Resistivity (×10¹¹ Ω·cm) | Latent Heat (J·g⁻¹) |
|---|---|---|---|---|---|---|---|
| CPCM-1 | 167 | 3 | 30 | 0 | 1.11 | 3.37 | 174.06 |
| CPCM-2 | 163 | 3 | 30 | 4 | 1.38 | 4.95 | 163.19 |
| CPCM-3 | 159 | 3 | 30 | 8 | 1.56 | 5.95 | 159.37 |
| CPCM-4 | 155 | 3 | 30 | 12 | 1.88 | 6.95 | 154.64 |
| CPCM-5 | 151 | 3 | 30 | 16 | 2.39 | 8.93 | 147.17 |
| CPCM-6 | 147 | 3 | 30 | 20 | 2.53 | 6.44 | 144.58 |
6.1 Thermal Conductivity and Electrical Insulation
Thermal conductivity analysis revealed a monotonic improvement with increasing Si₃N₄ content. The thermal conductivity of CPCM-5 (8wt% Si₃N₄) reached 2.39 W·m⁻¹·K⁻¹, representing a significant improvement over pure paraffin. This finding indicates that the Si₃N₄ particles form an effective thermally conductive percolation network within the PA/EG matrix, substantially reducing interfacial thermal resistance. However, the volume resistivity exhibited an initial enhancement followed by a decrease beyond the 8wt% threshold. The CPCM-5 achieved its maximum volume resistivity of 8.93 × 10¹¹ Ω·cm, which satisfies the stringent criterion for electrical insulation (>10⁹ Ω·cm). This improvement arises because well-dispersed Si₃N₄ particles block the electron transport pathways of the conductive EG network. Beyond the optimum concentration, however, the excessive Si₃N₄ content leads to particle agglomeration and interfacial polarization, which compromise the insulating properties.
6.2 Thermal Stability and Mechanical Performance
Thermogravimetric analysis (TGA) demonstrated enhanced thermal stability with increasing Si₃N₄ content. The residual mass at 800 °C increased linearly with Si₃N₄ addition, confirming that Si₃N₄ acts as a thermally stable barrier. Differential scanning calorimetry (DSC) revealed that the phase change temperature of CPCMs (47 °C–48.06 °C) is well suited for EV battery pack applications. The latent heat values obtained are shown in Table 5.
High-temperature cyclic leakage tests conducted at 70 °C for 12 hours confirmed excellent shape stability, with total leakage rates below 0.2wt% for all samples. The epoxy resin matrix combined with EG’s porous structure effectively encapsulates the paraffin, preventing its flow during the melting process. Mechanical testing showed that with 8wt% Si₃N₄, the tensile and flexural strengths reached their peak values: 79.9% and 43.6% higher than those of CPCM-1, respectively. Further increases in Si₃N₄ content caused particle agglomeration, which weakened the internal structure.
The X-ray diffraction (XRD) patterns of CPCM-1 and CPCM-5 confirmed that all characteristic peaks of PA and EG remained unchanged, with no new peaks emerging. This observation demonstrates that the component materials interact only physically without any chemical reactions. The microstructural analysis via SEM showed that in CPCM-1, the PA and ER filled the pores of the EG network, but many micropores remained visible due to volume contraction during solidification. In CPCM-5, these micropores were effectively filled by the well-dispersed Si₃N₄ particles, creating a more compact and continuous surface structure. However, in CPCM-6, some Si₃N₄ agglomerates were observed, corroborating the mechanical and electrical results.
7. Battery Thermal Management Performance
To validate the practical application of CPCM-5, thermal management experiments were conducted using retired 18650-type NCM811 cells. This chemistry was chosen due to its highest heat generation and thermal risk. I fabricated two types of battery modules with a 6-parallel-1-series (6P1S) configuration: one embedded in CPCM-1 and one in CPCM-5, with a third control module using natural air convection. Each module had a nominal voltage of 3.7 V and a capacity of 18.3 Ah.
The modules were subjected to continuous charge-discharge cycling with a 1.0C charge rate and a 2.0C discharge rate. The peak temperatures were recorded to evaluate cooling efficiency, and the spatial temperature differences were measured to assess temperature uniformity.
7.1 Temperature Rise of Single Cells under Convection and CPCM Cooling
Different discharge rates, from 0.5C to 2.0C, were applied to evaluate the cooling performance under varying heat generation intensities. Experimental results are summarized in Table 6.
Table 6. Cooling module performance at different discharge rates
| Cooling Mode | Discharge Rate | Peak Temperature (°C) | ΔT Max (°C) |
|---|---|---|---|
| Air Cooling | 0.5C | 36.1 | 1.78 |
| 1.0C | 48.9 | 2.95 | |
| 2.0C | 67.7 | 3.6 | |
| PCM Cooling (CPCM-5) | 0.5C | 33.9 | 0.15 |
| 1.0C | 41.6 | 0.23 | |
| 2.0C | 48.2 | 0.37 |
The natural convection module showed a dramatic temperature rise, reaching 67.7 °C at the 2.0C discharge rate, with a ΔTmax of 3.6 °C. In comparison, the CPCM-5 module maintained the peak temperature at just 48.2 °C, which was 40.1% lower than the air cooling module at 2.0C. The latent heat absorption during paraffin melting effectively absorbed the heat generated by the cells, buffering the system temperature increase. The uniformity analysis was equally noteworthy: the CPCM-5 module achieved ΔTmax values of only 0.15 °C, 0.23 °C, and 0.37 °C at the 0.5C, 1.0C, and 2.0C discharge rates respectively, which are five-to-ten-fold better than those for the natural convection module.
The temperature rise curves of CPCM-5 modules were much smoother and more linear during the phase transition region between roughly 35 °C and 47 °C, signifying the onset of paraffin melting and its effective latent heat absorption.
7.2 Module Temperature Uniformity under Cycling Conditions
Continuous charge-discharge cycling at an aggressive 2.0C discharge rate was performed to assess long-term performance. The temperatures measured at three distinct positions on the battery surface were analyzed after each cycle. In the air-cooled module, the maximum surface temperature reached 69.5 °C during the cycles. The CPCM-1 module controlled the temperature to a peak of 56.7 °C, while the CPCM-5 module was the most effective, maintaining the temperature below 48.8 °C.
The maximum temperature remaining stable at approximately 48 °C over multiple cycles demonstrates excellent reusability and cycling stability of CPCM-5. The small temperature difference between consecutive cycles (≤0.7 °C) indicates that the CPCM had fully resolidified between cycles and was ready to absorb heat in subsequent discharge steps, which is crucial for a robust thermal management system.
Analysis of temperature uniformity on the module level revealed the following:
$$ \Delta T_{\text{max, air}} = 3.6\,^\circ\text{C} $$
$$ \Delta T_{\text{max, CPCM-1}} = 2.4\,^\circ\text{C} $$
$$ \Delta T_{\text{max, CPCM-5}} = 1.3\,^\circ\text{C} $$
The significantly reduced spatial temperature gradient achieved by CPCM-5, due to its enhanced thermal conductivity, is crucial for maintaining consistent electrochemical performance across the EV battery pack. Temperature uniformity prevents over-charging or over-discharging of certain cells, mitigates local capacity fade, and curtails the propagation of TR in the event of partial failure.
This substantial mitigation of cell temperature excursions is attributed to the combined effects of the Si₃N₄-modified thermal conduction pathway and the high latent heat of the paraffin. The high thermal conductivity efficiently conducts heat away from hot spots and into the bulk PCM, where it is absorbed by the solid-to-liquid phase transition. The high volume resistivity ensures safe electrical isolation, minimizing the risk of short circuits and enabling direct contact with live components within the EV battery pack.
8. Conclusions
This comprehensive research investigated the thermal-electrochemical characteristics of retired ternary lithium-ion batteries with varying nickel contents and developed a next-generation composite phase change material for the thermal management of EV battery pack modules. The subsequent conclusions can be drawn.
First, the screening of 150 retired cells demonstrated an 86.7% qualification rate and a good state of health. Higher nickel content correlated with higher discharge capacity (NCM811 > NCM622 > NCM523 by 15.5% and 2.9% respectively) but also with lower voltage plateaus and greater self-heating tendencies during discharge.
Second, the adiabatic thermal runaway investigation revealed that nickel content profoundly affects thermal stability. Furthermore, the onset and trigger temperatures for thermal runaway decreased with nickel content, whereas the peak temperature and heat release rate increased. The NCM811 cells displayed the most severe TR behavior, with a trigger temperature of 139.9 °C, a peak temperature of 676.44 °C, and a temperature rise rate of 9420.33 °C/min, far exceeding the performance of NCM523. Electron microscopy studies after thermal abuse confirmed severe structural degradation of the anode and cathode active materials, particularly in NCM811, directly correlating with the reduced thermal stability. These findings underscore the substantial safety challenges of high-nickel chemistries and the imperative for advanced thermal management.
Third, a novel composite phase change material incorporating Si₃N₄ was successfully synthesized and evaluated. With a Si₃N₄ content of 8wt%, the CPCM-5 composite achieved an optimal balance of properties: thermal conductivity of 2.39 W·m⁻¹·K⁻¹, volume resistivity of 8.93 × 10¹¹ Ω·cm, latent heat of 147.17 J·g⁻¹, leakage rate below 0.2wt%, and superior mechanical properties.
Fourth, the experimentally validated CPCM cooling system for a retired 18650-type NCM811 EV battery pack exhibited outstanding performance. Compared to natural convection cooling, the CPCM-5-based module reduced the peak temperature at a 2.0C discharge rate by 40.1% (from 67.7 °C to 48.2 °C) and decreased the maximum temperature difference from 3.6 °C to 0.37 °C. During continuous cycling tests, CPCM-5 maintained the cell temperature below 48.8 °C, demonstrating the viability of this cooling approach for effective EV battery pack thermal management.
Specifically, this work contributes valuable insights into the thermal degradation mechanisms of high-nickel NCM cells and introduces a potential material solution that could facilitate the safe second-life utilization of retired EV battery pack cells by mitigating their propensity toward thermal extremes.