Review of Thermal Runaway Monitoring, Warning and Protection Technologies for Lithium-Ion Batteries

Abstract

Due to their high energy density, long calendar life, and environmental protection, lithium-ion batteries have found widespread use in a variety of areas of human life, including portable electronic devices, electric vehicles, and electric ships, among others. However, there are safety issues with lithium-ion batteries themselves that must be emphasized. The safety of lithium-ion batteries is receiving increasing amounts of attention as incidents such as fires and explosions caused by thermal runaway have caused significant property damage and fatalities. Thermal runaway can easily occur when lithium-ion batteries experience issues such as electrical abuse and thermal abuse. This study compares various monitoring, warning, and protection techniques, summarizes the current safety warning techniques for thermal runaway of lithium-ion batteries, and combines the knowledge related to thermal runaway. It also analyzes and forecasts the future trends of battery thermal runaway monitoring, warning, and protection.

1. Introduction

Since the 20th century, there has been a growing trend toward using clean, non-polluting, affordable, and easily accessible electric energy instead of highly polluting and expensive traditional energy sources, as the global community has become increasingly concerned with environmental and energy issues. Due to their advantages over other high-energy secondary batteries, such as high energy density, low self-discharge, good cycling performance, and environmental friendliness, lithium-ion batteries are widely used in the power supply of portable electronic communication devices such as cell phones and laptops, as well as in the energy supply of new electric vehicles. These advancements have significantly altered most industries, including land and marine transportation, and will soon become the norm. This will play a significant role in the renewable technology field and in the development of the next-generation power system. The suppression of thermal runaway events has emerged as a key issue in improving the safety of lithium-ion batteries; however, due to the nature of the safety-critical applications of the batteries themselves, the development of methods to detect and evaluate the safety of lithium-ion batteries has recently become a major challenge for industry and academia [1].

A lithium-ion battery may experience mechanical abuse if the external stress causes the battery casing to deform or become punctured by some sharp items (such as nails). Additionally, due to overcharge, overdischarge, and external short circuit, lithium-ion batteries may not adhere to their fundamental electrical properties. This could result in electrical abuse of lithium-ion batteries. Additionally, lithium-ion batteries may overheat as a result of internal electrochemical side reactions or external heat sources heating up, leading to thermal abuse of lithium-ion batteries. All of these elements have the potential to cause a succession of exothermic reactions in lithium-ion batteries within minutes, which would cause a sharp rise in the internal temperature of battery. This would then likely result in a thermal runaway event, which would typically result in smoke, fire, or possibly an explosion [2,3,4].

Thermal runaway events represent a serious hazard to human life and property in the form of smoke, fires, and explosions [5,6]. However, large-scale distributed energy management systems around the world are currently working toward achieving good safety and reliability. Therefore, a number of international organizations and committees, including the International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC), have created and published some authoritative test specifications to gauge the safety of lithium-ion batteries in an effort to enhance their performance and safety as well as allay concerns about thermal runaway [7,8]. To ensure the quality and safety of lithium-ion batteries used, these specifications mandate that lithium-ion batteries must pass a number of safety verification tests, such as overcharge, overdischarge, overheat, mechanical shock, and other tests. They also stipulate that the test cannot expose the battery to any secondary hazards (leakage, fire, explosion, etc.).

Lithium-ion batteries could still experience thermal runaway issues even if they entirely complied with all specifications and testing criteria since actual operating circumstances and surroundings can be more challenging and harsher than many test situations. Abuse from a variety of sources is frequently inevitable. The lithium-ion battery pack of a Tesla Model X, for instance, deformed and ruptured after colliding with a concrete bulkhead on U.S. Highway 101 in the San Francisco Bay Area in early 2018. A new electric vehicle in Fujian, China, unexpectedly caught fire in August 2020 while charging at a charging station. A 13-ton Megapack huge lithium-ion battery that was being tested for the Australian battery project caught fire in July 2021.

The development and acceptance of electric vehicles, electric ships, and lithium-ion batteries has all been significantly hampered by the growing serious safety issue with lithium-ion batteries. Consequently, research on lithium-ion battery thermal runaway characterization, particularly for equipment using lithium-ion batteries as a power source, can prevent casualties and property damage caused by a lack of timely warning of thermal runaway. This is essential for addressing lithium-ion battery safety issues and advancing the sustainable development of industry.

There are two main areas that can be improved upon to increase the safety performance of lithium-ion batteries and decrease the likelihood of dangerous accidents. In order to increase the safety performance of the battery from the battery’s own perspective, one method is to improve the production process within the lithium-ion battery itself by adding electrolyte additives, improving electrode materials, and improving the separator preparation process; an alternative method is to take changes in voltage, internal resistance, temperature, and other parameters accompanying the process of thermal runaway of lithium-ion batteries into consideration from the perspective of monitoring and warning to achieve the purpose of reducing casualties. As a result, this article provides a thorough assessment of the literature on current thermal runaway monitoring, warning, and protection techniques, as well as an analysis and forecast of the likely direction of future development for battery monitoring, warning, and protection technologies.

2. Lithium-Ion Battery Thermal Runaway Monitoring Technology

The following elements are primarily monitored for determining the lithium-ion battery’s operational parameters: battery voltage, operating current, internal resistance, and ambient temperature. In order to determine whether thermal runaway occurs during the charging and discharging cycle of the battery, the monitoring of voltage, temperature, and internal resistance can be used.

2.1. Monitoring Techniques Based on External Battery Parameters

The most popular technique for keeping track of lithium-ion battery health is the battery management system (BMS). It uses temperature and voltage sensors that are already incorporated into the device as its primary measurement instruments. The system can measure the surface temperature and terminal voltage of each lithium-ion battery in real time if enough sensors are installed [9]. Once an abnormal signal is detected, the BMS can immediately trigger an alarm.

2.1.1. Voltage

A multi-voltage sensor interleaved voltage topology measurement system was presented by Xia et al. [10] to precisely measure the terminal voltage of each lithium-ion battery, as is shown in Figure 1. In order to monitor the voltage anomalies of each lithium-ion battery in a series battery pack, the method first constructs a voltage sensor topology for redundant lithium-ion batteries and then uses clever algorithms, control circuits, and exact voltage thresholds. The voltage sensor value will vary if a lithium-ion battery experiences thermal runaway as a result of overcharging or overdischarging. If the voltage sensor value is below the minimum threshold, the lithium-ion battery is regarded as overdischarged. If not, it is deemed to be overcharged.

Locating a defective cell within a battery pack is a key advantage of multi-voltage sensor-based approaches over temperature- or gas-based monitoring systems, and numerous sensors can prevent false alarms of battery failure caused by individual sensor failures. However, the cost is typically significant because so many voltage sensors are employed. Moreover, a significant amount of sensing data place a heavy computational demand on the digital signal processor of BMS. Large lithium-ion battery systems, such as those used in electric vehicle battery packs, for instance, typically contain thousands of cells. When each battery cell is fitted with a voltage sensor, the BMS receives a significant amount of sensor data. The enormous volume of real-time data may consume a significant amount of BMS processing resources and storage space, jeopardizing the ability and robustness of the BMS to successfully manage charging and discharging behavior.

As a result, installing a single voltage sensor that is shared by a number of battery cells is standard practice in the industry. For instance, the Deluxe Tesla Model S battery pack contains 7104 cells total among 14 series modules, each of which contains 6 series cells and 74 parallel cells. However, only 84 voltage sensors are needed instead of 7104 because all parallel cells share a voltage sensor. A battery monitoring board (BMB) installed on one end of each module measures the temperature of the module’s positive and negative terminals in addition to the battery voltage. A daisy-chain cable assembly with a 10-pin Molex connector on top of the BMB connects all of the modules to a central BMS board at the end of the main battery assembly, which then communicates with other vehicle components via the controller area network (CAN) bus.

2.1.2. Temperature

The idea behind temperature sensors is to use voltage signals to monitor temperature signals. Thermistors, thermocouples, analog temperature sensors, and digital temperature sensors are examples of common temperature sensors. However, the same drawbacks apply to all of these temperature sensor types: poor detection accuracy and sensitivity to environmental changes.

Nascimento et al. [11] used fiber Bragg grating sensors and K-type thermocouples to monitor the surface temperatures of the top, middle, and bottom of a lithium-ion battery under normal and abusive operating conditions at various discharge multiplicities (0.53 C, 2.67 C, and 8.25 C), respectively, in order to increase the detection accuracy of surface temperature and to improve the reliability of monitoring with surface temperature data, as depicted in Figure 2. The findings demonstrate that while both sensors are capable of detecting the surface temperature of lithium-ion battery in real time, the fiber Bragg grating sensor has a higher temperature sensitivity and better resolution than that of the K-type thermocouple. This enhances the accuracy and dependability of the BMS system to monitor the temperature of lithium-ion batteries.

Since then, a network of fiber sensors has been created based on the fiber Bragg grating sensor enabling real-time, in situ, and multipoint monitoring of the surface temperature distribution on a smartphone lithium-ion battery, as shown in Figure 3. In order to simulate the lithium-ion battery response in the dry, moderate, and cold regions, various temperature and relative humidity conditions are taken into account. It was discovered that the surface temperature of the smartphone’s lithium-ion battery was nearly twice as high as it would be under normal circumstances due to the greater discharge rate and dry climate. The temperature clearly decreased from the top to the bottom of the battery cell, which was another notable tendency. The top of the cell, which is close to the electrodes, had the highest temperature, which was followed by the middle top, with the temperature at the bottom of the middle being somewhat higher than the middle. The lowest temperature was found near the bottom of the lithium-ion battery, compared to the other regions [12].

K-type thermocouples were used by Feng et al. [13] to measure the temperature of the battery during thermal runaway. These thermocouples were positioned in the bottom, side, and center of the battery’s two pockets as well as close to the safety valve, as is shown in Figure 4. Ultimately, the findings revealed that the battery’s internal temperature was approximately 870 °C, significantly higher than the ambient temperature. Based on the calculated temperature difference and the recorded data, it was discovered that 97% of the time during the test period, the temperature difference inside the battery stayed below 1 °C, while when thermal runaway occurred, the temperature difference reached its highest level, approximately 520 °C.

At the moment, the most common method of monitoring battery temperature is to install thermocouples on the surface of the battery, with three or more thermocouples installed at the top, middle, and bottom of the battery, respectively. According to research results thus far, three thermocouples are sufficient to reflect the real surface temperature of the battery. However, the internal temperature of the battery is more relevant to the real condition of the battery. It is generally believed that the battery temperature is fairly uniform when it is operating normally; nevertheless, in the event of a battery failure or a nearby heat source that causes the battery to heat up quickly, the difference between the surface and internal battery temperatures can be as high as 40–50 °C [14,15,16,17,18]. Therefore, in addition to the surface temperature measurements in the BMS, additional detection methods are required to monitor the internal battery temperature.

2.2. Monitoring Technology Based on Internal Battery Parameters

It is more accurate to utilize the interior temperature as a parameter to determine the state of the lithium-ion battery in a fully enclosed state rather than simply monitoring the voltage and surface temperature for the lithium-ion battery. Nowadays, electrochemical impedance spectroscopy (EIS) and integrated fiber optic sensors are the two main techniques used to monitor the internal condition of lithium-ion batteries.

2.2.1. Embedded Fiber Optic Sensor

According to a method Du et al. [19] proposed for estimating the internal core temperature of lithium-ion batteries based on fluorescence lifetime measurement, an apparatus with a nickel-coated fluorescent fiber was made in order to reliably monitor the internal core temperature of lithium-ion batteries. A source driving circuit, an optical coupling system, a fluorescence signal detecting and processing system, a display system, and a fluorescence excitation source with a wavelength of 470 nm make up the apparatus. When the rare-earth material on the fluorescent probe is exposed to the UV source, it excites fluorescence and emits afterglow, the decaying life of which is a single-value function of temperature, i.e., the higher the temperature, the shorter the decaying life. The fluorescent probe is nickel-coated and buried inside the cell core.

Scientists have discovered that the fiber Bragg grating sensor can accurately track and identify internal strain and temperature variations in lithium-ion batteries. However, Nascimento et al. [20] found that the monitoring method based on a single Bragg fiber could be flawed because the two signals could interfere with one another, making it impossible to quickly and precisely identify the internal temperature of a lithium-ion battery during a thermal runaway. This is precisely because they were able to monitor both signals. Therefore, a method for accurately measuring the internal temperature of lithium-ion batteries was proposed by coupling a fiber Bragg grating to the Fabry–Perot cavities and then subtracting the internal strain signal monitored by the Fabry–Perot cavities from the signal detected by the fiber Bragg grating. This was carried out based on the fact that Fabry–Perot cavities are extremely sensitive to strain but not to temperature, as is shown in Figure 5.

This method requires certain damaging alterations to the lithium-ion battery, which can weaken its structural integrity when subjected to harsh circumstances, making it mainly unsuitable for commercial and military applications. As a result, non-invasive monitoring methods for lithium-ion batteries are a prominent topic in today’s study.

2.2.2. Electrochemical Impedance Spectroscopy (EIS)

Researchers have developed an EIS-based technique for monitoring internal battery temperature because numerous studies have demonstrated a correlation between electrochemical impedance spectra and internal battery temperature. EIS is a popular monitoring method that is used to describe the electrochemical activity of lithium-ion batteries without harming the battery itself. The basic idea behind this technology is that a lithium-ion battery’s impedance reduces as its interior temperature rises. The ohmic impedance $R_b$, the impedance of lithium ions through the solid electrolyte interface(SEI) $R_{sei}$, the charge transfer impedance $R_{ct}$, and the lithium ion diffusion impedance W are the four main components of the electrochemical impedance spectrum of a lithium-ion battery, which is schematically represented in Figure 6 [21].

The internal resistance and capacitance of lithium-ion batteries, as well as their thermophysical characteristics and thermal behavior models, have traditionally been determined using EIS-based analysis [17,22,23]. The relationship between EIS, the state of health (SOH) and internal temperature has recently been thoroughly studied by researchers [24,25,26,27,28,29,30]. The benefits of EIS over embedded sensors include not requiring physical insertion of thermocouples or other instruments, which could compromise the structural integrity of the battery; being extremely sensitive and quickly reflecting the current condition of battery; and finally, being appropriate for batteries of any size and shape.

Lithium-ion battery misuse and capacity deterioration can be tracked using the SOH measurement provided by EIS [31,32,33,34,35]. In order to compare the SOH of individual lithium-ion batteries and battery modules under normal charge, discharge, and abuse from overcharging, Love et al. [31,32] employed EIS. The findings of this study demonstrated that the method can track the SOH of both individual batteries and battery modules, as well as overcharged cells, suggesting that the method has promising future development.

Batteries made of various materials and forms have also benefited from the application of internal temperature monitoring technology of EIS. Spinner et al. [36] used single-point electrical impedance measurements to monitor the transient internal temperature of commercial 18650-type lithium-ion batteries. The results of the study showed a correlation between the internal temperature of the battery and thermal runaway, and it was observed that the impedance response is related to the SEI, which will subsequently lead to battery deactivation once the SEI decomposes. A method to determine the internal temperature of a lithium-ion battery based on the electrochemical impedance intercept frequency was proposed by Raijmakers et al. [29] after analyzing the variation of the EIS of the battery with temperature and state of charge (SOC), defined as the frequency when voltage and current are in phase. The mechanism is that the SOC of lithium-ion battery is unrelated to this impedance intercept frequency, which only relates to its internal temperature. Therefore, an accurate estimate of the internal temperature can be derived by EIS probing as long as a stable SEI exists at the anode of the active lithium-ion battery. Using an electrochemical impedance spectrum that was produced from a 1 kHz sine wave stimulation signal, Schwarz et al. [37] obtained the estimation of the interior core temperature of lithium-ion battery. This method was used to construct and incorporate a device into the BMS. Lastly, the results demonstrate that the interior core temperature of a lithium-ion battery may be accurately measured in real time.

Monitoring a number of impedance spectrum characteristics enables the EIS-based approach for monitoring and detecting the internal temperature of lithium-ion batteries. The choice of the excitation source, including the mode and frequency of the excitation signal, is crucial for correctly forecasting the internal temperature of lithium-ion batteries and the ensuing thermal runaway [38]. Around 60–135 °C is the temperature at which the SEI layer begins to break down, at which point the lithium-ion battery becomes vulnerable to thermal runaway.

2.3. Summary

In conclusion,

Table 1. Advantages and disadvantages of different thermal runaway monitoring methods.

Method		Advantages	Disadvantages
External parameters	Terminal voltage	Monitor the voltage in real- time Capable of locating faulty battery Easy to operate, low cost	Complex topology of voltage sensors, high cost
	Surface temperature	Monitor the surface temperature in real time Easy to operate, low cost	Significant temperature difference from the internal temperature, Inaccurate for determining the state of the battery
Internal parameters	Embedded optical fiber sensors	Monitor the internal core temperature of the battery directly	High cost Higher requirements for battery packaging
	Electrochemical impedance spectroscopy analysis	Predict the internal core temperature of the battery without complex hardware Online monitoring of battery status	Fail to monitor large-scale batteries quickly and effectively

lists the benefits and drawbacks of two methods for detecting external conditions, such as the monitoring of surface temperature and terminal voltage, as well as two methods for detecting internal conditions, such as the use of an embedded fiber optic sensor for direct temperature measurement and the use of an EIS analysis method for indirect temperature estimation.

3. Lithium-Ion Battery Thermal Runaway Warning Technology

3.1. Thermal Runaway Warning Technology Based on Lithium-Ion Battery Voltage

Voltage is a crucial signal for BMS monitoring, and the key to early warning is the analysis of voltage change during thermal runaway. The time delay between voltage drop and temperature rise, which is about 15 s, as shown in Figure 7, was demonstrated by Feng et al. [13] using a large accelerated calorimeter on a large-capacity lithium-ion battery. Because this time period is favorable for thermal runaway early warning, it can be used as a signal for a thermal runaway warning system. Meanwhile, using the tiny current pulse discharge approach, it has been discovered that the resistance of the battery gradually increases as the battery temperature rises. By using the modest current pulse discharge approach, it was discovered that the battery resistance rose as the battery temperature rose. In-depth research on this phenomenon by Ren et al. [39] established a connection between electrical signal changes brought on by internal short circuits and temperature increases brought on by thermal runaway. When an irregular signal is discovered, the voltage sensors incorporated into the BMS may promptly raise an alarm and do a good job of monitoring the terminal voltage of battery [40].

Mao et al. [41] conducted an extensive and methodical study on the electrochemical and thermal behavior of lithium-ion batteries from overcharge to thermal runaway, and divided the overcharge behavior into four stages by examining the voltage and temperature changes of lithium-ion batteries during overcharge and thermal runaway, as shown in Figure 8. This study was carried out in order to use battery voltage as a warning parameter. The voltage increases to 4.74 V, the temperature gradually climbs to 31 °C, and the battery charge condition increases to 136% in the first stage. In the second stage, the battery keeps growing as the voltage increases to $V_{max}$ (5.07 V) and the temperature quickly increases to 39 °C. The battery temperature rises to 89.5 °C and the voltage drops to 4.47 V in the third stage due to severe electrode damage; in the fourth stage, the voltage rises to 4.93 V and the temperature reaches 97.9 °C; and finally, the battery ruptures at roughly 174% SOC and subsequently catches fire at 174.33% SOC. They came to the conclusion that the highest value of $V_{max}$ may be employed as a crucial point for warning in order to prevent thermal runaway.

In conclusion, voltage can be used as a warning parameter for thermal runaway, but because it is related to the nature of the battery body and changes there are caused by other conditions as well as thermal runaway, it is less frequently used as a separate warning parameter. Voltage will, however, become more crucial in the assessment of battery warning as battery packing, grouping, and other technologies advance continuously.

3.2. Thermal Runaway Warning Technology Based on Lithium-Ion Battery Temperature

Lithium-ion batteries can experience thermal runaway, which is characterized directly by a significant rise in internal temperature and indirectly by a rise in surface temperature. Thermal runaway is produced by the degradation of internal microstructure and a number of side reactions.

A multi-stage warning system for 18650 lithium-ion batteries and battery packs was created by Yang et al. [42]. The law of lithium-ion battery heat production was investigated using charge/discharge cycle testing on 18650 lithium-ion batteries at various multipliers and real-time thermocouple monitoring of the battery surface temperature. Battery capacity started to degrade when battery temperature reached 55 °C, and the trend of increase in temperature was slower from 55 to 80 °C, particularly in the stage of 70–80 °C, according to the analysis of the test findings. Finally, a three-stage warning system was decided on with a first-level warning temperature of 55 °C, a second-level warning temperature of 70 °C, and a third-level warning temperature of 80 °C, as is shown in Figure 9. The internal temperature of the battery is actually close to 100 °C when the surface temperature is 80 °C, at which point the SEI coating begins to decompose. When more heat is continuously accumulated, the battery’s risk of thermal runaway increases significantly. The gadget demonstrates outstanding early warning performance for the unusually hot lithium-ion battery and has the advantages of high efficiency, convenience, and quick response.

However, the biggest issue with using temperature as an early warning parameter is that the thermocouple or temperature sensor’s measurement of the battery’s internal and external temperature has a certain error, which will cause the battery to experience thermal runaway before the set warning temperature and ultimately result in early warning failure.

In order to monitor the battery temperature, Li et al. [43] introduced a resistance temperature detector (RTD) that was mounted behind the electrode current collector of CR2032 coin cells, as is shown in Figure 10. The results revealed that the temperature measured in this way was, on average, 5.8 °C higher than the external RTD, with a detection speed that was almost ten times faster, preventing thermal runaway events without interfering with the operation of the LIBs. Future temperature monitoring techniques must be more accurate to increase the success rate of thermal runaway warnings.

When using temperature as a parameter for early warning, the biggest issue is that the thermocouple or temperature sensor’s ability to accurately measure both the battery’s interior and exterior temperatures suffers from a certain error. This causes the battery to experience thermal runaway before the temperature is set to warn, which ultimately results in early warning failure.

In their research of battery early warning, Zhang et al. [44] discovered that a ternary lithium battery’s surface temperature is only 56.3 °C when it deforms and catches fire, indicating that temperature is not an appropriate basis for detecting fires in lithium-ion batteries and that a more effective method to track the battery’s actual temperature is required.

Grandjean et al. [45] performed a simulation study on the thermal characteristics of LiFePO₄ lithium-ion batteries with a capacity of 20 Ah and discovered that the temperature difference between the internal temperature and the battery’s surface could reach 20 °C in the large multiplier discharge state. They came to the conclusion that it was challenging to accurately reflect the lithium-ion battery’s true state by measuring the battery surface temperature. Additionally, the conventional technique of employing thermocouples to monitor the surface temperature of the battery in order to predict thermal runaway has a certain time delay due to the presence of heat conduction. Researchers have suggested that in order to provide early warning of thermal runaway in lithium-ion batteries, it is necessary to detect the battery’s interior temperature more precisely.

Parhizi et al. [46] developed a temperature tracking model of the internal battery based on heat conduction analysis and experimentally validated it using lithium-ion batteries with two different cathode materials based on the thermal characteristics of lithium-ion batteries and the kinetic characteristics of chemical reactions during thermal runaway. It is not logical to merely use the surface temperature measurements to monitor the thermal runaway of the lithium-ion battery because simulations and experiments have shown that the maximum internal core temperature of the lithium-ion battery during thermal runaway is nearly 500 °C higher than the surface temperature. An internal battery failure symptom that is frequently observed is an excessively high internal temperature of the battery. When the battery’s internal temperature exceeds 90 °C and 130 °C, there is a risk of explosion and the possibility of thermal runaway. As a result, it is important to keep an eye on the battery’s interior temperature.

A lithium-ion battery internal state monitoring scheme based on an embedded collapsible Bragg fiber sensor was proposed by Raghavan et al. [47] to address the challenge of monitoring the internal core temperature of lithium-ion batteries as is shown in Figure 11. When the internal stress or temperature of the battery changes, the Bragg fiber refractive index and refracted light wavelength will change. By measuring the change of refracted light wavelength, the internal stress and temperature of the battery are then determined.

3.3. Thermal Runaway Warning Technology Based on EIS

The EIS-based thermal runaway warning approach is also a frequently explored area of current research since it has been demonstrated by numerous studies that there is a relationship between electrochemical impedance spectra and internal battery temperature.

The AC impedance behavior of a commercial sealed lithium-ion battery at 10–40 °C and different SOC was examined by Suresh et al. [25]. The data were composed of two capacitive parts in the frequency range of 100 Hz to 10 mHz and an inductive component in the range of 100 kHz to 100 Hz. The impedance parameters were assessed after the data were processed using an equivalent circuit and a nonlinear least-squares fitting method. The findings demonstrate that the SEI layer impedance in lithium-ion batteries is temperature-sensitive rather than cell-charge- or charge-transfer-resistance-sensitive. According to impedance characteristics, a method developed by Schmidt et al. [16] can be used to estimate the average internal temperature of a battery even when the temperature inside the battery is not constant or changes abruptly. The results of the discussion on the impact of SOC on internal temperature estimation at various frequencies revealed that SOC considerably influences impedance temperature estimation results at low frequencies, while having less of an impact at high frequencies. Based on the monotonic link between the impedance phase shift and the internal cell temperature suggested by the prior study [27], Zhu et al. [48] further developed an impedance-based temperature estimation method taking into account the electrochemical imbalance generated by current excitation.

Srinivasan et al. [24,49] proposed an early warning method for the thermal runaway of lithium-ion batteries based on rapid monitoring of cell impedance. The Solartron analytical electrochemical interface and frequency response analyzer were used to measure the cell voltage, the phase-shift, and the amplitude |Z| at 5 Hz. The internal temperature of the battery is monitored and the likelihood of thermal runaway of the battery is predicted using the phase shift, which is marginally connected with the battery capacity and substantially correlated with the internal temperature T. Figure 12 illustrates how the temperature varies gradually before a lithium-ion battery thermal runaway, but the impedance phase shift will appear abnormal, leading experts to assume that monitoring the internal impedance can successfully produce a thermal runaway warning. It is not appropriate to use the internal resistance of the battery as the determining factor of the thermal runaway of the battery and should be used in conjunction with other parameters to determine whether the thermal runaway of the battery is or is not occurring. This is because the sudden change in the internal resistance of the battery is not always caused by thermal runaway of the battery, but can also be caused by an external disturbance of the battery, such as poor contact. For transient internal temperature monitoring of commercial 18650-type lithium-ion batteries, Spinner et al. [36] used a single-point electrical impedance measurement technique based on electrochemical impedance spectroscopy. This allowed for the attainment of impedance spectra in the temperature range of −10 °C to 95 °C and 0% to 100% SOC charge, and it demonstrated the correlation between the internal temperature of the battery and the imaginary part of the impedance. This investigation expands on findings that Srinivasan [24,49], Schmidt [16], Love [31,32], and others previously published. It expands the capabilities of the approach as a potential crucial tool for next-generation battery diagnostics by becoming the first single-point impedance test for interior temperature monitoring up to 95 °C.

Carkhuff et al. [50] constructed a tiny, low-power, multi-frequency (1–1000 Hz) impedance battery management system for numerous batteries of various capacity using a multi-frequency impedance meter. Phase shift and amplitude monitoring capabilities of the sensor enable simultaneous monitoring of the interior temperature of each cell. By monitoring and correcting mismatches and other electrical and thermal irregularities that occur in individual cells, the BMS maintains battery safety and efficiency without costing more money, taking up more space, or using more power than a typical BMS. The interior temperature of lithium-ion batteries may now be measured using a novel technique, according to Raijmakers et al. [29]. Due to the close relationship between zero-intercept frequency (ZIF) and the internal battery temperature and the fact that ZIF is unaffected by SOC, the internal battery temperature is calculated using the ZIF method. ZIF is the frequency at which the imaginary part of the impedance in the electrochemical impedance spectrum is zero. Then, a non-ZIF technique based on the internal battery temperature estimating method is provided [51], which can successfully prevent thermal runaway, in order to avoid the interference of the internal battery current and increase the prediction accuracy of the model. It is easy and practical to apply this impedance-based sensorless temperature measurement in a variety of stationary, portable, and high-power equipment, such as electric vehicles, electric ships, etc. Richardson et al. [30,52] looked into a novel technique for calculating the interior temperature distribution of a cylindrical battery by combining measurements of EIS and surface temperature. The model is effective enough to be used in the BMS of an electric car and does not require knowledge of the thermal characteristics of the cell, heat production rate, or thermal boundary conditions. Internal thermocouple measurements, which have not previously been shown for impedance-based temperature estimate, are used to validate the model. Then, using a thermal model and EIS measurements, a technique to estimate the interior and surface temperatures of the battery is suggested.

Based on earlier research, Lyu et al. [53] extended their proposal and confirmed the characteristics of the dynamic impedance slope in the 30–90 Hz region that gradually changed from negative to positive during overcharging. Additionally, using 70 Hz as an example, the thermal runaway mishap can be avoided when the battery does not release gas, bulge, or burn by monitoring the 70 Hz dynamic impedance and stopping the charging when the impedance slope turns positive since there is enough time to take the necessary precautions to stop the thermal runaway because of the warning period of 580 s prior to the thermal runaway. This property, which enables the overcharge warning and thermal runaway prediction of lithium-ion batteries, is simple to recognize and does not necessitate the use of sophisticated mathematical models and parameters. Additionally, a large-scale use of the early warning technique based on this slope shift in conjunction with an online dynamic impedance monitoring equipment is possible to prevent thermal runaway incidents, as is shown in Figure 13.

In conclusion, there is currently no standard way for EIS-based internal battery temperature prediction, and each method will rely on intricate mathematical models. The method described by Lyu that does not depend on mathematical models and parameters has a good chance of development since researchers in the field of EIS-based thermal runaway prediction think that the reliability of the warning is more significant than the accuracy.

3.4. Summary

The advantages and disadvantages of the early warning approaches based on battery voltage, battery temperature, and EIS are summed up in

Table 2. Advantages and disadvantages of different thermal runaway warning methods.

Method	Advantages	Disadvantages
Based on battery voltage	Voltage can be monitored in real time Faulty batteries can be located Predict the state of charge of the battery in real time	Voltage variations are complex and poorly regulated Thermal runaway warning is lagging behind Influenced by other external factors
Based on battery temperature	Surface temperature can be monitored in real time Predict the state of health of the battery in real time Directly measure internal battery temperature	Large temperature difference from internal temperature Low accuracy of thermal runaway prediction Thermal runaway warning is lagging behind Used in combination with EIS technology or embedded fiber optic sensor technology
Based on EIS	Sensitive to temperature changes Independent of SOC The original structure of the battery will not be destroyed	Influenced by other external factors Complex calibration process as different lithium-ion battery systems have different parameters of impedance Relies on more complex mathematical models

.

4. Lithium-Ion Battery Thermal Runaway Protection Technology

Lithium-ion battery thermal runaway events can have serious consequences; therefore, in order to avoid them, not only do we need to monitor their status in real time to detect thermal runaway, but we also need to take certain precautions to stop it from spreading once it does. Enhancing the safety of lithium-ion batteries at the cell level (internal protection) and using cooling or barrier technologies throughout the battery (external protection) are the two most typical ways to slow down the thermal runaway propagation process.

4.1. Lithium-Ion Battery Thermal Runaway Internal Protection Technology

In order to increase the safety of lithium-ion batteries themselves, a more rigorous preparation procedure and better battery materials must be used in battery manufacturing. This is due to the operating principle and thermal runaway mechanism of lithium-ion batteries.

The primary components of the cathode, anode, separator, and electrolyte make up the internal structure of lithium-ion battery. The separator, one of the crucial four core materials, serves as a migration channel for lithium ions, allowing the lithium ions in the electrolyte to freely pass through the micro-pores during charging and discharging to ensure the normal operation of the battery. It also serves to isolate the cathode and anode to prevent a short circuit by contacting the two electrodes. The essential inner layer components, particularly puncture resistance, self-shutdown, and high-temperature resistance of the separator, play a significant role in safety performance of the battery and have an impact on the resistance, capacity, and life of the battery [54,55,56].

4.1.1. Electrolyte

In lithium-ion batteries, the electrolyte serves as the medium for the transportation of lithium ions, and it is often made up of high-purity organic solvents, lithium salts of electrolytes, and associated additives. Lithium hexafluorophosphate (${\mathrm{L}\mathrm{i}\mathrm{P}\mathrm{F}}_6$) is commonly used as a solute in the current electrolyte for lithium-ion batteries, while carbonate is typically used as a solvent because both substances are flammable and combustible. In order to increase the thermal robustness of lithium-ion batteries and lower the risk of thermal runaway, electrolytes are typically adjusted in the ways listed below to address the thermal runaway issue.

• Improve the thermal stability of lithium salts

${\mathrm{L}\mathrm{i}\mathrm{P}\mathrm{F}}_6$, lithium hexafluoroarsenate (${\mathrm{L}\mathrm{i}\mathrm{A}\mathrm{s}\mathrm{F}}_6$), lithium perchlorate (${\mathrm{L}\mathrm{i}\mathrm{C}\mathrm{I}\mathrm{O}}_4$), lithium perchlorate (${\mathrm{L}\mathrm{i}\mathrm{B}\mathrm{F}}_4$), and others are the most frequently utilized lithium salts in lithium-ion batteries. The fact that these lithium salts are all very unstable to heat is one of the factors contributing to the low safety of electrolytes. The severe toxicity of ${\mathrm{L}\mathrm{i}\mathrm{A}\mathrm{s}\mathrm{F}}_6$ restricts its use even though the mixture of ${\mathrm{L}\mathrm{i}\mathrm{A}\mathrm{s}\mathrm{F}}_6$ and tetrahydrofuran (THF) has good electrochemical characteristics and thermal stability. The thermal stability of some novel lithium salts, such as lithium bis(fluorosulfonyl)imide (LiFSI) and lithium bis(trifluoromethanesulphonyl)imide (LiTFSI), is hampered by their poor aluminum collector passivation [57]. Therefore, it is essential to modify the lithium salts in order to enhance the performance of electrolyte.

First, it is possible to think about substituting the unstable lithium salts with others, such as lithium bis(oxalate)borate (LiBOB) and lithium difluoro(oxalate) borate (LiDFOB), which have been shown to form passivation films on collector surfaces and may be connected to the breakdown of B-O compounds [58,59]. Figure 14 illustrates the ultra-concentrated lithium bis(fluorosul-fonyl)amide (LiFSA)electrolyte that Wang et al. [60] utilized for batteries. They discovered that it has a stable cycling performance, good multiplicative performance, and low flammability.

The concentration of the lithium salts can be changed to improve performance, and it has been discovered that high concentration electrolyte (HCE) lessens the flammability of the electrolyte. This is a result of the high concentration of lithium salts, where the majority of solvent molecules combine with cations to form a solventized structure in the HCE system. This decreases the amount of free solvent molecules and creates a special solventized structure, which significantly inhibits the interfacial reactions between the solvent and the electrode and lowers the flammability of the cell [61]. As an illustration, by raising the concentration of ${\mathrm{L}\mathrm{i}\mathrm{P}\mathrm{F}}_6$ in the electrolyte to 2.5 M, the electrolyte demonstrated a noticeably longer ignition time, shorter self-extinguishing time (SET), and better cycling performance in the cell as a result of the improved shuttling of abundant lithium ions between the cathode and anode [62]. Aluminum collectors can develop a protective coating of lithium fluoride and be protected from corrosion by using a high concentration of LiTFSI electrolyte. According to Liang et al. [63], a high concentration (2.3 ${\mathrm{m}\mathrm{o}\mathrm{l} \mathrm{k}\mathrm{g}}^{-1}$) of LiTFSI can be employed as an electrolyte that is non-flammable and has good thermal stability. The dissolution numbers of ethylene carbonate (EC) and ethylene glycol dimethyl ether (DME) dropped and increased, respectively, with increasing salt content, according to the Raman spectroscopy results. It follows that the lithium ions attach to fewer EC but more DME molecules in the high concentration electrolyte, enhancing the thermal stability and nonflammability. As seen in Figure 15, this electrolyte not only exhibits outstanding thermal stability but also electrochemical properties that are equivalent to those of traditional carbonate-based electrolytes. High viscosity, poor wettability, and high cost still prevent their use in commercial lithium-ion batteries, despite the fact that highly concentrated electrolytes can lower flammability and increase electrochemical performance.

2.

An additive for flame retardancy

The most flammable component of the electrolyte is the organic carbonate solvent, and the basic idea behind this technique is to incorporate flame-retardant chemicals into organic electrolytes to increase their capacity to neutralize combustion radicals and thus reduce their flammability [64,65]. Commonly used flame retardants on the market include phosphorus-based additives such as tributyl phosphate (TBP), triphenyl phosphate (TPP), triethyl phosphate (TEP), etc. Liu et al. [64] created a shell out of a poly(vinylidene fluoride-hexafluoropropylene) (PVDF-HFP) and employed TPP as the inner core. The PVDF-HFP shell melts and the TPP within leaks into the electrolyte when the battery temperature is too high, which can prevent the monomer from burning. Traditional carbonic acid electrolyte (propylene carbonate (PC) as solvent and ${\mathrm{L}\mathrm{i}\mathrm{P}\mathrm{F}}_6$ as lithium salt) was given a fluorinated diethyl carbonate co-solvent by Pham et al. [65], which could reduce the flammability of electrolyte by mixing fluorine and hydrogen radicals during combustion. Pires et al. [66] used triphosphite as a supplement to the electrolyte. The cycling performance of the battery was found to be improved by partial fluorination of alkyl phosphate, which also raised the nonflammability of the cathode and electrolyte and increased the safety of lithium-ion batteries. Jin et al. [67] used a synthetic bifunctional additive as a flame-retardant film-former in ${\mathrm{L}\mathrm{i}\mathrm{P}\mathrm{F}}_6$ electrolyte to measure the flame retardancy of the additive. They measured the electrolyte’s self-extinguishing time to achieve this. The results showed that dimethyl allyl phosphonate enhanced the electrolyte’s thermal stability and had a favorable electrochemistry with graphite anodes. Additionally, as flame-retardant additives to the electrolyte, Zeng et al. [68] used large quantities of LiFSI salts and phosphate organic solvents. During the combustion of the electrolyte, the solvent molecules and lithium ions form a solventized shell, retaining the electrolyte’s nonflammability and enhancing the coulombic efficiency and cycling stability of the lithium-ion battery.

Additionally, it has been demonstrated that novel electrolyte additives are effective flame-retardants, even enhancing the electrochemical performance of batteries under high pressures. Due to their lower saturated vapor pressure, which effectively prevents the evaporation of solvents from the electrolyte and reduces the risk of combustion in combustible solvents, it is thought that their excellent flame retardancy is firstly caused by the prevention of combustion propagation through the generation of reactive radicals by P radical capture reactions [69]. Dagger et al. [70] investigated the flame-retardant effect of five flame-retardant additives (tris(2,2,2-trifluoroethyl)phosphate(TFP), tris(2,2,2-trifluoroethyl)phosphite (TTFPi), bis(2,2,2 trifluoroethyl)methylphospho-nate (TFMP), (ethoxy)pentafluorocyclotriphosphazene(PFPN) and (phenoxy)pentafluoro-cyclotriphosphazene(FPPN)) in electrolytes, and found that fluorinated cyclic phosphoramidites (PFPN and FPPN) outperformed the others in terms of electrolyte safety and electrochemical performance, as shown in Figure 16, and concluded that phosphorus-halogen synergistic. Furthermore, nitrogen can produce a protective carbon layer during the combustion process by creating ammonia, thus limiting the oxygen supply.

3.

Non-flammable electrolyte

The flame-retardant ingredient just reduces the flammability of the electrolyte. Thus, the essential fix for lithium-ion battery thermal runaway is the use of non-flammable or even non-combustible electrolytes. Utilizing flame-retardant elements or completely swapping out volatile solvents is a successful method for creating electrolytes that are non-flammable. When heated, typical carbonate-based solvents release hydrogen radicals. These radicals then interact with oxygen to produce oxygen radicals, which can spark the production of further free radicals and eventually cause fires. Introducing hydrogen or oxygen radical scavengers is one efficient technique to stop this free radical chain reaction. Most scientists concur that compounds containing fluorine or phosphorus can act as powerful radical scavengers when electrolytes break down. The electrolyte decomposition products include fluorinated and phosphorus radicals, which can interact with hydrogen radicals to stop a previous chain reaction of radicals and prevent the burning of the electrolyte solvent [71,72].

The most prevalent type of flame-retardants, renowned for a wide range of uses, are organic compounds that include phosphorus. They have low toxicity, low volatility, and good thermal stability [73]. It has been demonstrated that a number of nonflammable electrolytes based on organic solvents, phosphonitrile, and fluorinated phosphoric acid work well in lithium-ion batteries as well as sodium-ion and potassium-ion batteries [74,75]. As flame-retardant electrolyte components, trimethyl phosphate (TMP), TEP, and tripropyl phosphate (TPrP) have drawn a lot of interest. TEP and TMP decreased the SET time at low concentrations, but Xu et al. [75] discovered that 40 vol% was still enough to achieve nonflammability. Alkyl phosphates, while possessing strong oxidative stability, are incompatible with graphite, which lowers cycle rates. For lithium-ion batteries with high flash points and non-flammable hydrofluoric ether as the electrolyte, Fang et al. [76] used a mixture of diethylene glycol diethyl ether and non-flammable methyl-nonafluorobutyl ether, which has a better electrochemical performance and is more safe compared to conventional electrolytes. Chung et al. [77] created a novel electrolyte by replacing one hydrogen atom in the electrolyte with a fluorinated methyl group. This electrolyte not only reduced flammability but also significantly improved battery cycling performance. As shown in Figure 17, the complete cell performance improved with the addition of vinyl carbonate (VC) additive and remained nonflammable.

Due to their non-flammability, fluorinated and phosphate-based electrolytes are a possible route to safer battery electrolytes. However, in the creation and research of these electrolytes, the environmental sustainability of fluorinated compounds and the long-term stability of phosphate-based electrolytes must be taken into account.

4.

Solid electrolyte

Solid electrolytes have the benefits of non-flammability, minimal leakage, and long life over liquid electrolytes, which significantly raises the safety of lithium-ion batteries [78]. Due to their great electrical conductivity, sulfide electrolytes are currently a frequently discussed issue. ${\mathrm{L}\mathrm{i}}_{10}{\mathrm{G}\mathrm{e}\mathrm{P}}_2{\mathrm{S}}_{12}$, a new three-dimensional framework structure published by Kamaya et al. [79], exhibits an incredible high lithium ionic conductivity of 12 ${\mathrm{m}\mathrm{S} \mathrm{c}\mathrm{m}}^{-1}$ at ambient temperature. Its conductivity surpasses that of liquid organic electrolytes and is the greatest ever recorded for a solid electrolyte. This new solid-state battery electrolyte is stable, non-volatile, non-explosive, and has excellent electrochemical features such as high conductivity and wide potential window. It is also simple to mold, shape, and integrate during manufacture.

The main disadvantage of an all-solid polymer electrolyte with no liquid component is its low ionic conductivity at ambient temperature. The simplest and most practical technique to increase the conductivity of polymer electrolytes is to use non-aqueous phase organic solvents as plasticizers. Gel polymer electrolytes are the end result of this shift from traditional liquid electrolytes to all-solid electrolyte intermediates [80]. Gel polymer electrolytes typically contain plasticizers, lithium salts, and polymers with a special microporous structure that allows ions to move between the liquid electrolyte and the microporous structure. Stephan et al. [81] reviewed five organic polymers, including poly(ethylene oxide) (PEO), poly(vinylidene fluoride) (PVDF), poly(acrylonitrile) (PAN), poly(methyl methacrylate) (PMMA), and PVDF-HFP, with PEO being one of the most widely studied polymer electrolytes.

However, the high crystallinity of PEO results in its low ionic conductivity at room temperature, which impairs the stability of the electrochemical performance and reduces the lifetime of the cell. Gel polymer electrolytes, which are the current transition to all-solid electrolytes, maintain the dual benefits of liquid electrolytes and all solid electrolytes, such as high ionic conductivity and good thermal stability. However, there are still many factors to take into account, including electrolyte compatibility, mechanical strength of the polymer matrix, and cycle life.

4.1.2. Separator

In addition to keeping the cathode and anode of the battery apart to prevent short circuits by making contact with the two poles, the separator serves as a conduit for lithium ions from the electrolyte, one of the essential elements of the lithium-ion battery. The integrity of lithium-ion batteries during operation can be ensured by its good chemical characteristics and mechanical strength, which also helps to prevent spontaneous combustion, which has a significant impact on the safety of lithium-ion batteries [82].

Currently, polyolefin separator, which can be classified into polyethylene (PE) and polypropylene (PP) microporous films based on the base material, is the industry standard for lithium-ion battery separator. It is popular due to its high-cost performance, good mechanical properties, and electrochemical stability. Higher standards for comprehensive power lithium batteries are being put forth with the development of new electric vehicles, including high-temperature resistance, high pore size uniformity, and high discharge rate. Due to their high-temperature resistance line, electrolyte wetness, puncture resistance, and subpar oxidation resistance, the present PP/PE polyolefin separators pose serious safety risks and cannot be employed in power batteries [83].

There are now numerous techniques to increase separator safety, which are as follows.

• High performance of polyolefin separator

Currently, dry and wet techniques are primarily used to create lithium-ion battery separators. The production of separators using the dry process is solvent-free, pollution-free, and beneficial to the environment. The separator experiences nearly minimal thermal contraction during the transverse heating process, homogeneous microporous size distribution, high microporous conductivity, and other hot spots since only longitudinal stretching takes place. The separator, on the other hand, is not stretched transversely during the manufacturing process and is prone to cracking transversely when in use, which results in a relatively high likelihood of internal micro-short circuits in mass-produced batteries and low battery safety and reliability. Wet process separators have strong puncture resistance and biaxial tensile strength, and the finished product can be quite thin. However, the wet process uses a lot of solvents, which can lead to environmental pollution; in comparison to the dry process, the equipment is complex, expensive, takes a long time to complete a cycle, has high costs, and uses a significant amount of energy. In addition, the wet process can only produce thin single-layer PE material films, with a melting point of only 130 °C and poor thermal stability [84,85,86].

The following characteristics are primarily expressed in polyolefin separator high performance technology.

(1) Control of thermal shrinkage.

PE and PP, the basic materials for polyolefin separators, have melting values of 130–140 °C and 150–160 °C, respectively. These crystalline polymers are molded under the combined action of a temperature field and tensile force field. The molecular chain segments in the aggregated state structure of the polyolefin become unoriented when the temperature to which the separator is exposed increases to the softening temperature, which presents itself macroscopically as a drop in size known as thermal shrinkage. When heated at 90 °C for two hours, the separator of lithium-ion batteries shrinks transversely by 2.5% and longitudinally by 4%. Heat treatment can lessen thermal shrinkage by accelerating the secondary crystallization of polymer, changing the molecular chain orientation into a crystalline orientation, removing the internal tension of the separator, and improving crystallinity. Li et al. [87] demonstrated that heat treatment at 90 °C for 30 min could enhance the lithium-ion battery’s overcharge safety performance, lengthen the internal short-circuit failure time and increase the heat dissipation efficiency.

(2) High mechanical strength.

The mechanical strength of the separator includes tensile strength and puncture strength, which, on the one hand, can increase the deformation ability of the battery when subjected to external forces, and, on the other hand, can prevent the separator from being pierced by cell burrs or dendrites [88]. According to research by Xu et al. [89] on the impact of separators with various physical characteristics on the safety of lithium-ion batteries, the higher the mechanical strength, the safer the battery is under the same thickness conditions. A wet separator is highly favored in the field of power batteries due to its superior tensile strength and puncture strength when compared to a dry separator for various preparation methods and raw materials. This is because the molecular weight of the raw material is higher than that of the latter.

(3) Self-shutdown function.

The lithium-ion battery separator has a very low hole closure temperature and high film-breaking characteristics, which can cause the separator holes to close before the battery temperature increases too much, ensuring that the ion channels in the battery are closed. This property is known as the self-shutdown function. The melting point of raw materials and the breaking temperature of separators are closely related, and the PP separator has a greater breaking temperature than the PE separator. Because the existing dry separator has inadequate transverse tensile strength, the wet separator has poor thermal stability and cannot fulfill the needs of high-energy, high-power lithium-ion batteries. Due to this, Celgard [56] proposed integrating the attributes of dry and wet microporous membrane technology, using the qualities of good flexibility and low closed pore temperature and polypropylene (PP) with high mechanical properties and high fusing temperature into a lithium battery separator, successfully producing an “ABA” sandwich structure multi-layer separator. This lithium-ion battery separator has a low closed cell temperature of 130 °C and a high fusing temperature of 160 °C, which significantly improves the safety performance of batteries. Hu et al. [90] used high-density polyethylene(HDPE) and PP as raw materials using triple co-extrusion and bi-directional stretching technology successfully studied and prepared a PP/PE/PP three-layer composite separator; the closed pore temperature of the PE separator is lower than that of the PP separator. When the temperature reaches 130 °C, the PE layer melts and closes the pores of the separator, blocking the migration of lithium ions in the electrolyte, while the PP layer remains intact to avoid internal short circuit.

2.

Surface modification of polyolefin separator

The internal resistance of lithium-ion batteries can easily increase due to the high crystallinity of the polyolefin separator, low surface energy, small polarity, poor affinity with electrolytes, lack of wettability and liquid retention, and poor surface contact with positive and negative electrode sheets. At the same time, severe shrinkage will occur at temperatures above 120 °C, which will easily lead to the two poles contacting and forming a short circuit, which may cause safety accidents. In order to fully improve the overall performance of polyolefin separators and guarantee the integrity of the separator in the event of thermal runaway of lithium-ion batteries, polyolefin separator modification technology combines the benefits of flexible organic materials and inorganic materials containing multiple hydrophilic groups.

(1) Inorganic material coating.

Nano alumina (${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$) is commonly used commercially to create ceramic coatings for polyolefin separators because of its excellent thermal and chemical stability, low cost, and ease of accessibility. In their study of ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ single-sided coated PE separators for lithium-ion batteries, Lei et al. [91] found that the ceramic-coated separators had a lower internal resistance of the cell and a discharge temperature rise of about 3 °C. ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ was coated on both sides of a PE separator by Yao et al. [92], and the results revealed that after baking at 150 °C for two hours, the thermal shrinkage of the ceramic-coated separator was less than 5%, significantly enhancing the thermal safety performance of lithium-ion batteries. Due to its low specific gravity and low hardness, boehmite (${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$-${\mathrm{H}}_2\mathrm{O}$), a material comparable to alumina, has also steadily made its way onto the market. By altering the boehmite shape, An et al. [93] successfully created ceramic separators with moisture contents below 600 ppm.

(2) Organic material coating.

The interfacial characteristics and heat resistance of polyolefin separators can be enhanced via organic coating, and polymer systems such as PVDF and its copolymers, polyimide (PI), aramid, polyethylene terephthalate (PET), PEO, and cellulose have emerged as new hot areas for the development of lithium-ion battery separators [94]. PVDF coating is one of the commercial coating technologies that has matured and been accepted by many lithium-ion battery companies. An et al. [95] demonstrated that a PVDF-coated separator can increase electrolyte absorption/retention ability, improve interfacial bonding between the electrode and separator, increase the electrode’s hardness, make the battery thinner and stronger, and simultaneously increase battery safety. By effectively using the impregnation approach, Xiong et al. [96] created an ethyl cellulose-coated polyolefin separator, and both its closed pore temperature and thermal shrinkage rate dramatically decreased. In order to address the safety issues brought on by separator shrinkage in lithium-ion batteries at high temperatures, Song et al. [97] successfully prepared a polyimide-coated polyolefin separator by impregnation. This separator has a significantly improved mechanical strength and reduced high-temperature shrinkage to within 10% at 140 °C. Aramid can endure temperatures of up to 400 °C thanks to its heat resistance and fire-retardant qualities. A high-performance composite separator with aramid-coated coating has closed-hole qualities, heat resistance, wettability, and the ability to absorb and hold liquid. The basic technology is currently mastered by Teijin, Toray, and Sumitomo Chemical of Japan, and Shanghai Research Institute of Chemical Industry Co. of China.

(3) Multi-layer coating.

Multi-layer coating entails coating the first layer (for instance, the ${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ layer) first, the second layer (for instance, the PVDF layer), on top of that, and then preparing the multi-layer composite separator. By baking at 180 °C and baking composite separators with a PVDF/${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$/PE separator/${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$/PVDF multilayer structure, An et al. [98] created composite separators that had a thermal shrinkage of only 7.8% and no melt collapse at 200 °C.

(4) Mixed coating.

In order to create a composite separator, organic and inorganic materials are mixed to create a homogeneous slurry, which is then coated on top of the substrate for the separator. When a PVDF/${\mathrm{A}\mathrm{l}}_2{\mathrm{O}}_3$ hybrid slurry was applied to a PE separator and coated on both sides, Jeong et al. [99] found that the electrical conductivity increased to 0.719 × 10³ ${\mathrm{S} \mathrm{c}\mathrm{m}}^{-1}$ and the composite separator’s shrinkage was reduced from 94% to 74%.

The separator is the fundamental component of a lithium-ion battery with the highest technical barrier and the final localization, and it plays a significant role in the safety and cost performance of lithium-ion batteries. From an economic standpoint, polyolefin separator still holds a dominant position in the field of lithium-ion battery separator research at this time, and the primary research focus is on enhancing the safety and high performance of polyolefin separator. The development of the polymeric separator from a simple structure to a high-order complex structure, to its high mechanical strength, high heat resistance angle, and to protect the safety of the separator for lithium-ion batteries, is influenced by the screening and compounding of separator raw material, post-treatment process improvement, and surface modification. At this point, new substrate materials for separators are also a hot topic in laboratory research, with some studies concentrating on inorganic materials such as hydroxyapatite and sodium titanate to enhance the heat resistance and flame retardancy of separators under ultra-high temperature conditions. Traditional separator materials have also changed from polyolefins to organic materials such as PVDF, PET, PI, and fibrin.

4.1.3. Electrodes

If appropriate changes are made to the electrode, a crucial component of lithium-ion batteries, the risk of thermal runaway can be reduced. One of the main causes of the thermal runaway process in lithium-ion batteries is the extra oxygen produced during the cathodic reaction [100]. One of the main reasons for separator deformation is lithium dendrites, which are created when lithium particles accumulate on the anode surface [101].

In order to raise the pyrolysis temperature, it is therefore effective to dope the cathode material at the atomic level to increase the cathode’s thermal stability. Applying a protective coating to the cathode surface is a popular technique that can successfully prevent the cathode material from directly contacting the electrolyte, thus minimizing side reactions that produce heat [102,103,104]. Additionally, cathode safety can be significantly increased by altering the electrode with PTC materials [105,106].

The unequal anode current distribution caused by the instability and inhomogeneity of SEI is the most fundamental reason of lithium dendrite formation on the anode surface in terms of material attributes. Lithium dendrite formation can be reduced by enhancing the cathode materials [107,108], such as silicon-based materials and graphite-carbon materials, or by applying a protective coating to the anode surface. The unequal distribution of anode current close to the anode causes lithium ions to be converted to lithium particles and deposited on the anode surface to create lithium dendrites when lithium-ion batteries are misused, such as through overcharging or overdischarging. The most typical remedy is to add electrolyte additives to stabilize the contact between the anode and the electrolyte in order to prevent the growth of lithium dendrites caused by uneven SEI production.

4.2. Lithium-Ion Battery Thermal Runaway External Protection Technology

A better cooling design to efficiently disperse heat and insulate all surrounding cells from thermal propagation is only one example of a system-level protection solution. By accelerating the cooling of the lithium-ion battery with air, liquids, and phase change materials, efficient heat dissipation can decrease the severity of thermal runaway. By obstructing the heat transfer path to surrounding healthy cells, isolation of thermal propagation can stop the thermal runaway process and avert a domino effect or chain reaction.

4.2.1. Battery Management Technology

The monitoring technology for lithium-ion batteries is currently developing and is gaining popularity. Setting up a monitoring and alarm system is one of the more popular ways to obtain thermal runaway warning. A thermal runaway monitoring and warning system based on PLC was developed by Shao et al. [109] and is mostly used in electric vehicles. Figure 18 shows the specific control flow chart, which can calculate the time when the battery is about to undergo thermal runaway reaction based on the collected temperature value and the rate of change. It can track the internal temperature change of a single cell and set 80 °C as the alarm signal for thermal runaway. The system can assess the situation and sound an early alarm If it notices that the battery pack’s interior temperature has reached 80 °C. Additionally, the technology can forecast when a thermal runaway would occur based on how quickly the battery pack is heating up, protecting both drivers and passengers.

Attention is also being paid to a different approach to battery pack equalization. The goal of battery pack equalization technology is to transfer and distribute electrical energy evenly throughout each and every cell in the battery pack. Different battery health states result from variations in each individual cell’s voltage, internal resistance, capacity, etc., or from varied depths of charge and discharge. There are two types of equalization technology: energy dissipation type and energy non-dissipation type. The energy non-dissipation kind, which transfers power using capacitance or inductance to balance the power of each battery pack, is more energy-efficient than the other two. Niu [110] employs an energy non-dissipative equalization circuit based on the SOC equalization technique. The experiment demonstrates that the equalization effect is effective, and the voltage is monitored by an LTC6802 chip. Furthermore, the circuit operates steadily while taking real time and safety into consideration, successfully preventing the growth of battery pack inconsistency and extending the battery pack’s useful life. Speltino et al. [111] used the extended Kalman filter SOC algorithm and validated its simulation results for constant-current charging and periodic step-current discharging; the system is able to equalize the single cell charge state during bi-directional operation and continuously track the individual cell charge state. The results show that the algorithm boosts the equalization effect and also fulfills the objective of lowering the discrepancies between individual cells.

4.2.2. Cooling Technology

Battery cooling technologies include air cooling technology, liquid cooling technology and phase change material cooling technology.

Research on thermal management using gas media is currently mainly focused on cell organization, gas flow channel, gas flow direction, gas rate, and various optimization techniques. The Z-type battery thermal management system was developed by Sun et al. [112], utilizing a tapered inlet and outlet duct design and an analytical design of experiments technique. This led to the development of Z-type flow battery modules with improved thermal performance. A J-type battery thermal management system, which has one more outlet than the typical U-type and Z-type battery thermal management systems, was proposed by Liu et al. [113]. The adjustable flexibility of battery thermal management is substantially increased by installing a control valve at the outlet, which can be utilized to modify the opening degree of the two outlets under various operating conditions. The benefits and drawbacks of U-, Z-, and J-type battery thermal management methods are listed in

Table 3. Battery thermal management system for different models.

Models	Illustrations	Advantages	Disadvantages
U-type		Small pressure drop, low energy consumption	Bad temperature field consistency, higher maximum temperature
Z-type		Good temperature field consistency	High pressure drop, high energy consumption
J-type		Small pressure drop, low energy consumption Good temperature field consistency	Complex structure, high sealing requirements

. The ideal cooling performance for the cubic arrangement of the battery module was found by Wang et al. [114] after studying the cooling performance of the battery module in various configurations and taking cost and cooling effect into account. The locations of the cooling system’s inlets and outlets as well as the number of inlets and outlets were all varied in the designs that Peng et al. [115] for a variety of cell configurations. The findings indicated that an arrangement with a smaller aspect ratio was more advantageous to enhance the cooling system’s performance. The cooling capabilities of aligned, staggered, and crossed battery modules under various airflow rates were studied by Fan et al. [116]. The aligned design, followed by the staggered and then the cross arrangement, is determined to have the optimum cooling performance and temperature uniformity. By varying the diameter of the input channel at a fixed flow rate, Yang et al. [117] were able to rationally design the layout of the cell module and ultimately produce a greater cooling effect.

In liquid cooling technology, according to Hirano et al. [118], a module with ten lithium-ion batteries was given a direct contact liquid cooling system. Since the liquid will directly contact the battery, it must have high electric resistance, be non-flammable, and be ecologically acceptable. In this study, hydrofluoroether and perfluoroketone were tested. The masses then flowed back to the module for cyclic operation after being vaporized and cooled in a heat exchanger outside the module. The liquid-cooled battery thermal management system was thoroughly examined by Smith et al. [119] from the perspective of the entire vehicle and practicality. For a module with eight square cells, a computational fluid dynamics model was created. A cold plate with a liquid channel was then inserted underneath the battery module, and to close the air gap between the cells and the cold plate, the soft material with strong thermal conductivity was linked. Mondal et al. [120] created cold plate modules with various structural forms and work fluids in the form of nanofluids, such as air, water and a combination of ethylene glycol and water. Water and other liquid work fluids can freeze in low-temperature situations and cease to function; however, an ethylene glycol-water solution has antifreeze qualities and is employed in the Tesla Model S’s battery thermal management system. The ratio of the two liquids can be changed to achieve the desired freezing temperature, and when the glycol ratio reaches 60%, the combined solution will not freeze at −45 °C. Tang et al. [121] created a mini-channel and a water cooling method for the heat dissipation of lithium-ion battery pack. Three distinct water-cooling techniques are created, and it is ultimately discovered that a bigger contact area and more evenly distributed water-cooling plates may successfully stop the battery’s temperature rise.

For phase change materials (PCMs) cooling technology, Ling et al. [122] used a cylindrical heater as the heat source while analyzing the performances of PCMs changed depending on the type of paraffin used, the paraffin mass fraction in composites, and the packing density of the composites. The PCMs compounded with 75% mass fraction graphite and paraffin with a melting point of 44 °C were discovered to have the best cooling effect. Furthermore, liquid cooling and phase change cooling can be combined to investigate the impacts of various liquid flow rates, expanded graphite mass fractions, and density of composite phase change material on the cooling effect. In order to further improve the efficiency of heat transfer and increase the uniformity of the internal temperature, Wu et al. [123] examined the features of battery pack temperature variation under high temperature environments (25–40 °C). The findings demonstrate that compared to the single cooling technique, the composite thermal management system can still ensure that the battery pack temperature is kept within the safe temperature range of 37–43 °C while the ambient temperature increases from 25 °C to 40 °C.

4.2.3. Blocking Technology

The “Lithium-ion Batteries for Power Storage” GB/T 36276 [124], published in 2018, explicitly states the requirements and test methodologies for thermal runaway propagation of lithium-ion batteries in response to the risk of this occurrence. Since blocking technology can prevent the diffusion of heat and high-temperature materials during thermal runaway, it is vital to create thermal runaway propagation suppression technology for lithium-ion batteries, which will reduce the development of thermal runaway events. Thermal runaway propagation in the battery module is mostly caused by the heat transfer process in the battery pack. There is a dearth of research on the use of barriers to prevent the spread of thermal runaway, and there is no reliable screening technique for choosing barriers.

In their study of the effect of epoxy resin plates of various thicknesses on series and parallel modules, Chen et al. [125] discovered that epoxy resin plates could successfully reduce the intensity of thermal runaway in batteries, lower the maximum temperature of thermal runaway cells, and lengthen the time it took for thermal runaway to propagate between adjacent cells. In order to stop the spread of thermal runaway by lowering the thermal conduction and thermal radiation capacity between cells, Berdichevsky et al. [126] inserted insulating and thermal insulation plates between various layers of cells. Larsson et al. [127] added liquid-cooled and insulated panels between the cells to use numerical simulations to confine thermal runaway to several cells. Mehta et al. [128,129,130] reduced the thermal impact of the thermally runaway cell on neighboring cells by using an expansion material placed between the inner surface of the cell shell and the matching electrode assembly to expand and absorb heat when the cell is heated. Zhang et al. [131] compared the behavioral traits of soft pack type lithium-ion batteries during thermal runaway propagation at various gaps and discovered that, as the distance increases, the time needed for thermal runaway propagation adjacent to the battery becomes longer, while the intensity of thermal runaway occurring in the battery is relatively weaker, which is related to the deformation state during thermal runaway of lithium-ion battery.

In their study of a 10s4p structured 18650 lithium-ion battery pack, Wilke et al. [132,133] induced thermal runaway in one of the cells after filling the pack’s gap with phase change material. They discovered that the presence of the phase change material effectively reduced the thermal impact of the thermally runaway cell on the neighboring cells. Even though adding insulation materials with lower thermal conductivity between cells helps prevent heat transmission, it considerably improves the circumstances for the cells to dissipate heat and has a tendency to lead to heat accumulation. In order to prevent the heat accumulation phenomenon, the phase change material can absorb a specific amount of heat under the heat situation. Unfortunately, the latent heat of phase shift has a limit, making it challenging to completely absorb the significant quantity of heat emitted by the thermally runaway battery. Phase change materials are mostly utilized to lower the cyclic heat output of battery because they also somewhat reduce the ability of the battery and environment to exchange heat. To increase the safety of lithium-ion batteries in air travel, Yi et al. [134] conducted a study on the usefulness of aerogel mats of various thicknesses in preventing the spread of thermal runaway of lithium-ion batteries with various SOCs. The experiment’s results demonstrate that lithium-ion battery thermal runaway could be effectively stopped using aerogel mats of various thicknesses. Only the battery that was directly in contact with the heating rod experienced thermal runaway over the course of the experiment, and the temperature of the batteries next to it remained within a safe range. Although the thinner aerogel felt can also stop the spread of thermal runaway, the temperature of the battery next to it has already risen above the safe level, and an irreversible reaction has taken place inside the battery.

The lithium-ion battery must be put out using fire extinguishing technology if the thermal runaway of the battery is not promptly prevented and controlled. In a review of the effectiveness of the most widely used fire extinguishing agents, Yuan et al. [135] introduced a number of typical extinguishing agents and their fire extinguishing mechanisms, summarized their fire extinguishing effects, and discovered that water-based extinguishing agents performed the best overall. Cui et al. [136] provided a description of the fine water mist fire extinguishing mechanism, followed by a discussion of the impact of internal and external factors on the effectiveness of the fine water mist fire extinguishing, including fine water mist characteristics, additives, obstacles, ventilation conditions, fuel type, and flame scale. They also reviewed the research on the use of fine water mist technology in battery fires. The future development direction and research ideas for fine water mist fire extinguishing technology are then presented based on the current research trends, and the development prospect of its use in the field of battery fires is anticipated. The readers are referred to several review papers that were previously cited for a more thorough evaluation and comparison of the available fire suppression techniques.

However, with lithium-ion batteries, not only may there be a fire, but there could also be an explosion, and the fine water mist fire extinguishing system is significantly longer than other gas fire extinguishing systems. A fire extinguishing time that is too long will also influence the fire suppression effect, as would the combustion of lithium-ion batteries and the explosion of the flame created by nearby quickly ignited flammable objects. Therefore, the primary inhibitory mechanism for asphyxiation and cooling inert gas extinguishing agent, such as carbon dioxide, heptafluoropropane, can better prevent the combustion and explosion of the battery. When the carbon dioxide stored in liquid form in the gas tank is sprayed out, it will quickly vaporize, producing a large amount of gas carbon dioxide, reducing the concentration of oxygen in the combustion area, and reducing the amount of oxygen in the combustion area. Since the vaporization process will absorb a lot of heat, it will have a better cooling effect, reducing the risk of a lithium battery explosion, and it will not pollute the environment. Additionally, heptafluoropropane, a better-performing halon substitute, has the physical effects of reduced temperature, reduced oxygen concentration, chemical inhibition, and superior environmental friendliness. Heptafluoropropane extinguishing chemical excels at putting out liquid, electrical, solid surface, or fusible solid fires by using total flooding due to its superior gas-phase electrical insulation, it can take into account the security of crucial locations including computer rooms, dangerous chemical storage rooms, communication equipment, and generator rooms. In comparison to typical high-pressure fire extinguishing agents, heptafluoropropane has the benefit of high insulation, and the process of burning lithium-ion battery fire will not cause harm to other components. Inert gas extinguishing agents such as carbon dioxide and heptafluoropropane are more suitable as extinguishing agents to prevent the combustion and explosion of lithium-ion batteries when compared to fine water mist, dry powder, and aerosol extinguishing agents. However, if a thermally runaway battery still has enough energy to manufacture combustible material, it may still offer a risk of delayed explosion; hence, more research on battery explosion suppression is required.

5. Conclusions and Outlook

The safety concerns brought on by lithium-ion battery thermal runaway occurrences are receiving increasing amounts of attention as lithium-ion batteries grow in popularity, having emerged as a frequently discussed topic in current research.

Lithium-ion battery thermal runaway monitoring and warning systems currently in use rely on keeping an eye on specific characteristic defect signals, such as terminal voltage, temperature, internal resistance, etc. Enhancing the precision of voltage sensors and temperature sensors in the BMS can enhance the accuracy of thermal runaway detection for external monitoring techniques such as terminal voltage and surface temperature monitoring. Additionally, the sensor array architecture can be improved from both a hardware and software standpoint to obtain the best intelligent monitoring with the fewest possible sensors, which will also save expenses. With regard to the internal detection method, the internal state warning method and BMS can be combined to create a more precise thermal runaway warning model for lithium-ion batteries, and the detection resolution and high-temperature resistance of the embedded sensors can be improved at the same time.

In order to prevent casualties and property damage, lithium-ion batteries must be monitored for thermal runaway and given an early warning. To stop or slow the spread of thermal runaway after it has occurred, some actions are still required. The study of the evolution and mechanism of thermal runaway has recently gone more in-depth, but there are still numerous issues with the monitoring and warning systems for lithium-ion batteries. To effectively avoid the development of fire, the three methods of external protection technology can be combined and used when thermal runaway occurs. First, by using battery management technology to detect the early warning signal of the system, the precise location where the abnormality occurs can be determined. Secondly, utilizing blocking technology to keep the number of thermal runaway modules within a certain range could be a useful approach. These modules are then cooled down by using air cooling, liquid cooling, or phase change material cooling technology, which can successfully prevent fire accidents and achieve safe thermal runaway response. Finally, it is possible to put out the fire and minimize damage by combining various fire suppression techniques with various application scenarios if all other last-resort prevention and control measures are unsuccessful and the battery still experiences thermal runaway.

The most important issue is the enhancement of the safety performance of the battery itself, in addition to the aforementioned methods of monitoring, warning, and protection. High ionic conductivity, suitability for the majority of anode and cathode materials, high lithium salt solubility, consistent electrochemical performance, low toxicity, and environmental protection are all characteristics of a high-quality lithium-ion battery electrolyte. The present electrolyte modification techniques still have a number of flaws, and the majority of them cannot simultaneously satisfy the requirements listed above. Although there are still some issues, the gel polymer electrolyte enhanced by solid electrolytes retains the benefits of both liquid and solid electrolytes and has a promising future. Existing commercial separators are not strong enough from a mechanical perspective to withstand rigid collisions; therefore, new high-temperature-resistant, high-strength separators should be actively developed with the goal of meeting the battery’s fundamental performance requirements. The current excellent performance of the separator, as well as modification technologies for the desired high mechanical qualities and thermal stability, provides the opportunity. The optimum separator is crucial to the energy density, power density, cycle life, and safety of the battery. A thorough understanding of the separator’s fundamental properties, failure mechanisms, and other factors is necessary to ensure battery safety. At the same time, new intelligent separator alternative materials, failure detection methods, and other cutting-edge technologies are being actively developed to address the current issue of mechanical and thermal failure of the separator and enhance battery safety. Concerning the electrode, a key future research path is the hunt for a cathode battery that is more heat-resistant and better able to prevent the growth of lithium dendrites.