3.1. Temperature and voltage variations in TR experiments
Figure 3 presents photographs of the batteries after thermal runaway (TR), revealing viscous internal substances at the safety valve outlets of batteries with 50% state of charge (SOC) and 25% SOC. Unlike the findings from other literature [22] that report evidence of melted aluminum in fresh batteries after TR experiments, no traces of melted aluminum were observed in Fig. 3. The surface temperature of the batteries subjected to cycling aging did not reach the melting point of aluminum at 660°C, indicating that the scale of thermal runaway events in cycled-aged batteries was smaller than that in fresh batteries.
Figure 4 illustrates the temperature and voltage profiles of batteries during the thermal runaway (TR) process under different state of charge (SOC) conditions. Analyzing the curve for 100% SOC in Fig. 4a, the TR process can be divided into four stages, characterized by three critical temperatures [23].
In the first stage, characterized by a self-heating rate of less than 0.02℃/min, the battery exhibits a gradual decline in capacity and a slight decrease in voltage [23].
In the second stage, denoted by T1 as the initiation temperature of self-heat generation in the battery [23], the solid electrolyte interface (SEI) decomposes as the temperature rises. Additionally, the anode, no longer protected by the SEI membrane, begins to react with the electrolyte [24,25], resulting in a gradual increase in battery temperature due to the heat released by these reactions.
In the third stage, denoted by T2 as the triggering temperature where the self-heat generation rate of the battery exceeds 1℃/s, various chemical reactions inside the battery generate a significant amount of gas at elevated temperatures, resulting in an increase in internal pressure [26]. When the internal pressure reaches a critical threshold, safety valves open to release the pressure. As the gas is expelled, the battery temperature slightly decreases. Subsequently, the chemical reactions of the internal materials, such as SEI decomposition, cathode decomposition, anode-electrolyte reactions, etc., release heat [26], causing the battery temperature to slowly rise again. Moreover, with the temperature further increasing, the chemical reaction rate accelerates, leading to a faster temperature rise in the battery [27].
In the fourth stage, T3, which is defined as the highest temperature point reached during the thermal runaway process [23], represents the peak temperature of battery TR. During the T2 to T3 stage, i.e., the fourth stage, the internal chemical reactions within the battery intensify, including decomposition reactions at the cathode, reactions between the anode and electrolyte, and electrolyte decomposition reactions, among others [26]. These reactions release a significant amount of heat, resulting in a rapid increase in battery temperature。
As evident from Fig. 4, during the process of heating until the end of thermal runaway, the temperature data collected from different thermocouples of batteries at the same time point show that TS > TP > TN > TE, where TS represents the surface temperature, TP represents the positive electrode temperature, TN represents the negative electrode temperature, and TE represents the electrolyte temperature. The fluctuations in the temperature curves of TP, TN, and TE are more pronounced with decreasing SOC. The video and temperature curve of the safety valve nozzle indicate that as the SOC decreases, more materials are ejected from the battery before thermal runaway, Video in Supplementary file.
Based on the findings from the literature [28], it has been demonstrated that the cathode materials undergo various chemical reactions within the temperature range of 100–300℃. The graphite crystal structure experiences collapse and deformation, while LiC6 and LixSi react with the electrolyte, leading to the release of heat [29]. Moreover, both crystalline and amorphous silicon undergo transformation into amorphous phases during the lithium intercalation process, followed by subsequent conversion into crystalline compounds [30]. Furthermore, within the temperature range of 110–300℃, the cathode undergoes a series of phase transitions from layered structure to M3O4 spinel, and then to rock-salt structure [29,31], These two processes are associated with the release of a significant amount of oxygen, which reacts with the electrolyte and releases a substantial amount of heat [20].
Figure 5 presents the variations of TS,max, TP,max, TN,max, and TE,max with respect to State of Charge (SOC). It is evident from the graphical representation that the values of TS,max, TP,max, TN,max, and TE,max exhibit an increasing trend with an increase in SOC. The general trend of the maximum temperature among the thermocouples in the battery at different SOC levels is observed as follows: TS,max > TP,max > TN,max > TE,max.
3.2. Strategies for Addressing Pressure Rise Rate in Experimental Contexts
Influence of Battery Rupture on the Thermodynamic Environment of Sealed Compartment: Time Rate of Change of Pressure as Depicted in Fig. 6.
The pressure evolution inside the sealed chamber, resulting from the battery rupture and the consequent significant changes in the thermodynamic environment, is presented in Fig. 6. As shown in the figure, the pressure initially increases slowly due to the temperature rise in the chamber. Subsequently, a rapid pressure rise is observed, followed by a quick decrease until it reaches a nearly constant value. The primary reason for the internal pressure increase in the battery is gas generation, including H2, CO2, CO, CH4, C2H4, and C2H6. Furthermore, the gas mixture also contains electrolyte vapor, hydrogen fluoride (HF), and other gases, which are detailed in reference [32]. In addition to electrolyte vaporization induced by physical changes (90–248°C), new gases are generated through chemical reactions, which can be explained by thermal decomposition and reactions between the electrolyte, binder, and electrode materials, as summarized in public literature [33].
To obtain a standardized pressure rise rate curve for lithium-ion batteries (LIBs) during thermal runaway, this study employs the concept of LIB eruption index (\({\text{K}}_{\text{LIB}}\)) proposed in reference [28]. Formula (1) is utilized to calculate \({\text{K}}_{\text{LIB}}\), which yields a time-varying curve as shown in Fig. 6. The maximum value of \({\text{K}}_{\text{LIB}}\) is obtained using formula (2) during the calculation process.
$${\text{K}}_{\text{LIB}}=\left(\text{dP}/\text{dt}\right){\text{V}}^{1/3} \left(1\right)$$
In this context, the rate of change of pressure with respect to time, dP/dt represents the pressure rise rate during lithium-ion battery (LIB) thermal runaway, while \(V\) denotes the volume of the sealed chamber.
$${\text{K}}_{\text{LIB},\text{max}}={\left(\text{dP}/\text{dt}\right)}_{\text{max}}{\text{V}}^{1/3} \left(2\right)$$
In this study, \({\text{K}}_{\text{LIB},\text{max}}\) refers to the maximum value of KLIB, while \({\left(\text{dP}/\text{dt}\right)}_{\text{max}}\) represents the maximum rate of pressure rise within the chamber. To ensure a valid comparison of \({\text{K}}_{\text{LIB},\text{max}}\) values for batteries with different initial SOC levels, measurements should be conducted under the same conditions using containers with the same shape and volume. The \({\text{K}}_{\text{LIB},\text{max}}\) value for the 100% SOC battery was determined to be 0.2886 kPa\(\cdot\)s−1\(\cdot\)m− 1, 0.1345 kPa\(\cdot\)s−1\(\cdot\)m− 1 for the 75% SOC battery, 0.0632 kPa\(\cdot\)s−1\(\cdot\)m− 1 for the 50% SOC battery, and 0.1266 kPa\(\cdot\)s−1\(\cdot\)m− 1 for the 25% SOC battery. As observed, \({\text{K}}_{\text{LIB},\text{max}}\) values decrease with decreasing SOC levels. Figure 6 illustrates that \({\text{K}}_{\text{LIB}}\) curves of the battery safety valve spray are more sensitive to SOC. Additionally, the high-speed camera video shows that lower SOC levels result in higher amounts of materials released prior to thermal runaway. These findings can provide a basis for the design of BTMS.
3.3. Influence of SOC on Mass Loss Rate and Gas Generation
An essential finding of this study is the relationship between the magnitude of thermal runaway events and the quantity of material ejected. A comparison of the mass characteristics and gas generation rates before and after battery experiments indicates that more severe thermal runaway events are often associated with larger mass ejections. As shown in Fig. 7, it can be observed that with an increase in State of Charge (SOC), both the mass loss rate and gas generation rate increase, but the normalized gas generation rate decreases. However, compared to the fresh battery (TR), both the mass loss rate and gas generation rate are reduced in the battery with increased SOC.
The experimental findings reveal that TR (thermal runaway) generates a significant amount of gas. Furthermore, the electrolyte in the battery can vaporize at high temperatures, resulting in the formation of large molecular organic compounds[27]. Despite the four batteries being identical, there are significant differences in the gas generation rates. The gas generation equation for the battery is formulated as follows[27].
$$PV=nRT \left(3\right)$$
$$n=\frac{{P}_{2}{V}_{2}}{R{T}_{2}}-{n}_{0} \left(4\right)$$
In the equation, the variable n represents the gas generation rate, \({P}_{2}\) denotes the real-time pressure inside the experimental chamber after thermal runaway (TR), \({V}_{2}\) refers to the volume of the experimental chamber, R represents the ideal gas constant, \({T}_{2}\) represents the ambient temperature inside the experimental chamber after stabilization, and \({n}_{0}\) represents the initial volume of gas inside the chamber.
Due to the rapid increase in temperature and pressure in the battery after thermal runaway (TR), accurate gas measurements can only be obtained when the temperature and pressure inside the experimental chamber are stabilized. Table 2 presents the experimental data obtained after achieving stable temperature and pressure conditions inside the experimental chamber.
Table 2
Experimental Findings at Varying State of Charge (SOC)
SOC | Battery residual weight (g) | Gas production (mol) |
100% | 623.3 | 2.45 |
75% | 648.5 | 2.06 |
50% | 663.6 | 1.35 |
25% | 675.4 | 1.52 |
In order to analyze the gas generation rates of the four batteries, the gas generation rates were normalized, and the mass loss and normalized gas generation rates were plotted as shown in Fig. 7, following the methodology described in reference[27].
Equation (5) for calculating the mass loss rate is shown below:
$$K=\frac{{m}_{e}}{{m}_{r}}\times 100\% \left(5\right)$$
In the equation, K represents the mass loss rate, \({m}_{e}\) denotes the initial mass of the battery, and \({m}_{r}\) refers to the residual mass of the battery after thermal runaway (TR).