Failure analysis of lead‐acid batteries at extreme operating temperatures

The lead‐acid battery system is designed to perform optimally at ambient temperature (25°C) in terms of capacity and cyclability. However, varying climate zones enforce harsher conditions on automotive lead‐acid batteries. Hence, they aged faster and showed lower performance when operated at extremity of the optimum ambient conditions. In this work, a systematic study was conducted to analyze the effect of varying temperatures (−10°C, 0°C, 25°C, and 40°C) on the sealed lead acid. Enersys® Cyclon (2 V, 5 Ah) cells were cycled at C/10 rate using a battery testing system. Environmental aging results in shorter cycle life due to the degradation of electrode and grid materials at higher temperatures (25°C and 40°C), while at lower temperatures (−10°C and 0°C), negligible degradation was observed due to slower kinetics and reduced available capacity. Electrochemical impedance spectroscopy, X‐ray diffraction, and energy‐dispersive X‐ray spectroscopy analysis were used to evaluate the degradation mechanism and chemical and morphological changes.

Temperature plays a key role in battery operation as it affects the cycle life, performance, and available capacity. The PbA battery system is designed to perform optimally at ambient temperature (25°C) for performance, capacity, and cyclability. However, they degraded faster when operated at higher than ambient conditions leading to shorter cycle life due to the degradation of the electrode and grid materials. 3 The oxidation and reduction rates increased significantly at both the Pb anode and PbO 2 cathode, leading to higher discharge capacity at elevated temperatures. 4,9 Besides having a deleterious impact on the cycle life at elevated temperatures, several other impacts include self-discharge reactions, loss of electrolyte, active material shedding, grid corrosion, and loss of mechanical strength of the positive electrode (PbO 2 ). [12][13][14] Shedding or loss of positive active mass particles into the electrolyte could also increase corrosion and macrodefects on the lugs of the negative electrode. 13,14 While operating at a lower temperature, low electrolyte conductivity and active material would result in reduced available capacity. 3 To reduce the corrosion or degradation rate of the PbA battery, limiting the internal temperature to <60°C could minimize electrolyte vaporization. 15 The cell internal pressure should be in an acceptable range for longterm optimum performances. However, it was reported that charging efficiency and cyclability were improved under high internal pressure with the favorable crystal structure of the electrodes. 13 Performance evaluation of the batteries at elevated temperatures and near-freezing temperatures are critical for using these batteries for outdoor energy storage applications in hot and cold conditions. 3 The adverse effects of temperatures also include reduced discharge capacity, increased internal resistance, and self-discharge with increased duration at extreme temperatures. Typical PbA batteries undergo many charge/discharge cycles during their lifetime. 7 Hence it is necessary to understand the complete cycling behavior of PbA batteries in various operating conditions, including incomplete charging, slow discharging, extreme conditions such as temperature fluctuations, and vibrations that cause degradation of internal components leading to failure of the battery.
In this work, a systematic study was conducted to analyze the 2 V/5 Ah Enersys® Cyclon SLA cells cycled at −10°C, 0°C, 25°C, and 40°C, and to minimize the experimentation duration as these conditions are practical for vehicles used or stored in frozen tundra and arid desert climates. To evaluate these conditions, electrochemical impedance spectroscopy (EIS) was carried out to evaluate internal resistance (ohmic and charge transfer) to explain the degradation mechanism of the battery. Further, electrode materials were extracted postcycling analysis for morphology characterization using X-ray diffraction (XRD) and energydispersive X-ray spectroscopy analysis (EDX). SLA batteries were observed to degrade faster at higher temperatures (25°C and 40°C). However, the degradation is minimal at lower temperatures (0 and −10°C) due to less active material and slower kinetics. The impedance value, x axis intercept of the Nyquist plot, was observed to increase with cycling at all temperatures.

| EXPERIMENT SETUP
SLA cells 2 V, 5 Ah (Cyclon Enersys®) were tested in this study at −10°C, 0°C, 25°C, and 40°C ( Figure 1A). Studies on these temperature margins are sufficient to provide F I G U R E 1 (A) Cyclon cell 2 V and 5 Ah, (B) Temperature-controlled environmental chamber. a better understanding for SLA cell performance. The thermal chamber (ESPEC, BTU-433) was used for testing cells in a controlled environment ( Figure 1B). A programmable battery tester (Arbin Instruments, BT-1) was used for charge/discharge cycling studies. The charge and discharge cut-off potential were kept at 2.45 and 1.85 V, with a constant charging and discharging rate of C/10 (0.5 A) and a 5 min rest period between charging and discharging cycles. The discharge rate was taken to study the slow charge-discharge phenomena, which will ensure enough time for the battery materials to equilibrate under the test conditions so that the degradation mechanism can be captured. Discharge capacities were calculated for each cycle from the charge and discharge profiles. EIS spectra were recorded after 20/25 charge-discharge cycles in a fully charged state at room temperature using a potentiostat/galvanostat (PARSTAT 2273; Princeton Applied Research) with a frequency range of 10 Hz to 1 kHz.
Postcycling analysis, electrode materials were extracted in charged and discharged states for chemical and morphological characterization ( Figure 2). The extracted material was thoroughly ground, rinsed with water profusely to remove remnant acid, and dried overnight at 80°C. Phase and crystal structure of pristine and degraded electrodes were examined using XRD (PANalytical X'Pert PRO MRD; Cu Kα radiation; 0.006°/s). Phases of the materials were validated after the rinsing and drying process during the material extraction process. EDX spectral mapping analysis (25 kV) was performed to estimate the chemical composition of electrodes postdegradation.

| RESULT AND DISCUSSION
The charge-discharge profiles at different operating temperatures, electrochemical impedance spectra studies, and chemical and morphological analysis are discussed in the following sections.

| Charge-discharge cycling studies
The discharge profiles of the cycled cells at −10°C, 0°C, 25°C, and 40°C are shown in Figure 3A-D. A sudden drop in cell voltage at the beginning of discharge is due to polarization. All batteries lose charge over time when kept on an open circuit, which is termed as selfdischarge. It was observed that during the initial cycles, the total discharge duration at 40°C and 25°C is about two times as compared to that at 0°C and −10°C, which is due to slower kinetics and lower electrolyte conductivity at subzero temperatures. Moreover, the discharge duration is 2, 2.5, 3.8, and 1.5 h after 120 cycles for 40°C, 25°C, 0°C, and −10°C, respectively. This suggests that F I G U R E 2 Cell disassembling process after cycling studies to extract active electrode materials for X-ray diffraction characterization. the aging of the PbA battery is faster at elevated temperatures (40°C and 25°C) than at subambient temperatures (0°C and −10°C). For cells discharged at 0°C, the total discharging duration remains unchanged throughout the cycling studies. Whereas for cells discharged at −10°C, the discharge duration decreases from 3.5 to 1.5 h, which is relatively faster aging despite low active materials utilization due to reduced initial available capacity. Moreover, during initial cycles, lower capacity is more essentially than the manufactured rated capacity due to capacity fading resulted from longer shelf-life prior to procurement of the cells for this study. The noise observed in discharge profiles is due to temperature fluctuations in the battery during charge-discharge, which are normalized by the thermal chamber keeping the environment around the cells at a predetermined temperature.
Generally, battery life is reduced by 50% for every 10°C/ 18°F increase in temperature. Similarly, at lower temperatures like 0°C and −10°C, the available capacity is reduced due to slower kinetics and lower conductivity leading to slower movement of ions, which is otherwise indicated as increased resistance for charge-transfer reactions. Hence capacity fading at higher temperatures is more significant than that at zero and subzero temperatures. Figure 4A-D illustrates the charging pattern of the cells at −10°C, 0°C, 25°C, and 40°C. Like discharging profiles, during the initial cycles, it was observed that total charging duration is higher for cells discharged at 40°C and 25°C, as compared to that at 0°C and −10°C. The charge durations are 2, 2.5, 4, and 1.5 h after 120 cycles for 40°C, 25°C, 0°C, and −10°C, respectively. Similar charging and discharging patterns corroborate the observed capacity degradation process.
The discharge capacity of the cells calculated from the discharge profiles is illustrated in Figure 5A. It was observed that the discharge capacity is close to the theoretical capacity (5 Ah) for 25°C and 40°C, while the initial discharge capacity at −10°C is higher than the result at 0°C, the discharge capacity drop is much significant in −10°C because lower temperature increases the internal resistance and reduces its capacity. 16 Steep slope confirms that degradation is higher at 40°C due to enhanced aging of the electrodes from faster reaction kinetics at elevated temperatures. The capacity degradation after 130 cycles is 92%, 74%, 69%, and 17% for 40°C, −10°C, 25°C, and 0°C, respectively. Moreover, the cell degradation is insignificant at 0°C even though the total discharge capacity is low. The state of health (SOH) is a measurement that reflects the general condition and ability to deliver the specified performance compared to that of a fresh battery. SOH takes factors  into consideration such as charge acceptance, internal resistance, voltage, and self-discharge. 17,18 It is a measure of the battery's long-term capability and available lifetime energy throughput. The health of the battery tends to deteriorate gradually due to irreversible physical and chemical changes; hence SOH is estimated by multiple methods, including impedance, conductance, and discharge capacity (current/actual). 1,17 SOH was calculated using instant discharge capacity and initial capacity (after first cycle) as illustrated in Figure 5B. SOH was observed to be highest for 0°C ranging from 95% (1st cycle) to 89% (130th cycle). However, the SOH decreased sharply for cells discharged at −10°C (78%-46%), 25°C (67%-28%), and 40°C (73%-10%). Significant SOH loss was observed for 40°C confirming major degradation.

| EIS studies
EIS was measured after every 20/25 charge-discharge cycles at room temperature in the fully charged state. EIS is essential for evaluating the power loss due to cycling and analyzing the failure mechanism. The Nyquist plots of the cells discharged at different temperatures are illustrated in Figure 6A-D. All the cells initially showed similar impedance values (0.05-0.06 Ω) prior to cycling. However, after 130 cycles, the impedance values are significantly higher: 0.41, 0.20, 0.13, and 0.38 Ω at −10°C, 0°C, 25°C, and 40°C, respectively. The increase in impedance is consistent with capacity degradation after 130 cycles, as shown in Figure 5A.
According to the standard equivalent circuit model, the measured impedance data were analyzed using a complex nonlinear least squares (CNLS) fitting method (Figure 7). EIS studies the system response to the application of a periodic small-amplitude AC signal. 2,7 Analysis of the system response contains information about the interface, its structure, and reactions taking place at the interfaces, for example, electrolyte-electrodes. 17 The impedance value, x-intercept from the Nyquist plot ( Figure 6A-D), is observed to increase gradually with cycling for all the temperatures ( Figure 8A). However, the increase is larger at lower temperatures (−10°C and 0°C) than at higher temperatures (25°C and 40°C), even though the total active material availability and utilization are significantly reduced at lower temperatures. The lower real impedance at higher temperatures (25°C and 40°C) essentially means that battery activity is improved despite higher degradation. The ohmic and charge-transfer resistances were calculated by fitting the EIS data to an equivalent circuit model using ZSimpWin software. The ohmic and charge-transfer resistances showed similar behavior as impedance values ( Figure 8B,C). However, it is reported that charge-transfer resistance is critical in representing the degradation mechanism as it reflects the reaction mechanism at electrodes. 7,17,19 The voltage at 50% discharge cycle (V dis,50% ) was analyzed further to evaluate capacity degradation with cycling ( Figure 8D). These values increase with cycles at all temperatures, however, more rapidly at 25°C and 40°C than at 0 and −10°C. Higher battery degradation correlates to higher slope of the V dis,50% versus cycle profiles, which essentially shows that the cell discharge capacity and SOH are decreasing with cycling.

| Chemical and morphological analysis
Two cells were cycled at each temperature for data reliability and XRD analysis. The anode and cathode materials obtained from the fresh and cycled cells in charged and discharged states were analyzed using XRD (Figure 9).
The XRD pattern of the anode after 130 cycles in charged condition contained Pb (cubic) and PbSO 4 (anglesite) peaks, like that for the fresh cells ( Figure 9A). Along with Pb and PbSO 4 , there were PbO and Pb 3 O 4 peaks for the cycled cell at different temperatures ( Figure 9A-D). Pb peak intensity was higher and sharper in the new cell than in the cycled cells, which confirmed that the PbSO 4 was completely converted into Pb through a reduction reaction. However, there was a permanent deposition of PbSO 4 in the cycled cells after 130 cycles due to sulfate hardening. The intensity of PbSO 4 peaks was more intense at 40°C compared to 25°C and the new cell, which could be attributed to poor reversibility with a higher amount of PbSO 4 at the anode in the charge condition ( Figure 9A).
The cathode in charged condition contained PbO 2 and PbSO 4 peaks, like the fresh cells ( Figure 9B). The cell, after 130 cycles, also showed Pb, PbO, and Pb 3 O 4 peaks. The PbO 2 peak intensity was higher and sharper in the fresh cell than in the cycled cell after 130 cycles. PbSO 4 was detected due to permanent deposits (sulfate hardening) through charge/discharge cycling, leading to capacity loss. The PbSO 4 peaks were more intense at 40°C compared to those cycled at 25°C. And the fresh cell due to accelerated kinetics at higher temperatures led to irreversible state. However, there was not much difference observed in the cells at 0°C in the charged state ( Figure 9A) than in the discharged state ( Figure 9C) because at 0°C only half of the cell's capacity was used. The discharge capacity is 4.1 Ah at 25°C and 2.15 Ah at 0°C due to the change in kinetics at 0°C. Similarly, at −10°C, the cathode for the cell after 130 cycles in charged and discharged conditions contained PbO 2 , Pb 3 O 4, and PbSO 4 peaks ( Figure 9B,D). However, at 0°C, the anode, both in charged and discharged conditions, contained Pb (cubic), PbO, and PbSO 4 (anglesite) peaks ( Figure 9A,C and 10).
The anode in discharged condition for the cell after 130 cycles contained the Pb (cubic) and PbSO4 (anglesite) peaks, like that of the fresh cells ( Figure 9C). Along with Pb and PbSO 4 , there were Pb 3 O 2 SO 4 , PbSO 3 , PbO, and Pb 3 O 4 peaks in the cell after 130 cycles. The PbO, PbSO 4 , and Pb 3 O 4 peaks were identified in Figure 9C,D. The PbSO 4 peak intensity was higher and sharper in the fresh cell compared to the cell after 130 cycles, mostly agglomerated and surface-hardened active materials, which confirmed that the Pb was completely converted into PbSO 4 through an oxidation reaction in the fresh cell. 3,14,15,20 Some of the PbSO 4 peaks, such as 26.6°, 29.8°, and 36°are relatively higher in the cycled cell compared to the fresh cells. However, the main peak intensity at 44.4°is more intense in the fresh cell than in cycled cells. This discrepancy observed could be attributed to the compositional difference on the electrode surface and in the bulk. The Pb peaks were more intense at 40°C compared to that at 25°C and fresh cell. This is probably due to the higher deposition of Pb at the anode in the charged condition. 3,20 The cathode in discharged condition contained PbO 2 and PbSO 4 peaks for the cycled cells, like that with the fresh cells ( Figure 9D). Moreover, the cell after 130 cycles also showed Pb 3 O 2 SO 4 , PbSO 3 , Pb, PbO, and Pb 3 O 4 peaks. The PbSO 4 peak intensity was higher and sharper in the new cell compared to the cell after 130 cycles. The PbO 2 detected could be due to the inability of the bulk active material not being able to get reduced during discharge at elevated temperatures, leading to less utilization of active materials and faster capacity fading/degradation in the cycled cell. Obviously, the PbO 2 peaks were more intense at 40 than compared to 25°C and the fresh cell due to a larger deposition of PbO 2 at the cathode in the charged condition. 10,11 There was not much difference observed in the Pb peak intensity in charge condition for the cathode ( Figure 9B) and cell at discharge condition ( Figure 9D) at 0°C. Similar to the anode, this could be because at 0°C, only half of the cell capacity was used due to the slower kinetics at this temperature. In summary, the XRD analysis F I G U R E 10 Energy-dispersive X-ray spectroscopy chemical analysis comparison of the electrodes from cycled cells (130 cycles) at various temperatures and fresh cells (Table 2 data are represented as histograms).
concludes that the major/minor phases in charged and discharged state for electrodes are as illustrated in Table 1.
The energy-dispersive X-ray spectroscopy (EDS) was conducted on anode and cathode active materials to examine sulfur content which is represented as the Pb/S ratio. The Pb/S is the ratio of the composition obtained from the Pb, PbO 2, and PbSO 4 peaks in the EDS spectrum ( Table 2). The error bar is determined by calculating the standard deviation for data collected at different locations on the sample analyzed. Higher values of Pb/S mean lower SO 4 2− content and vice versa.
The Pb/S ratio is indicative of the relative amount of sulfate (SO 4 2− ) getting deposited on the anode and cathode after cycling. The Pb/S ratio of the fresh cell is significantly different due to the complete reversibility of PbSO 4 to Pb as compared to that observed in cycled cells. As indicated from the Pb/S ratio, it can be seen that SO 4 2− deposition and hardening were higher in the anode and cathode in the discharged state, which eventually led to a reduction in the amount of active materials with cycling (130 cycles). 18,19,21 It can be seen that SO 4 2− content increased with decreasing temperature from 40°C to 0°C due to slower kinetics and lower PbSO 4 reversibility to Pb. However, it is the lowest at −10°C compared to cycled and fresh cells, essentially due to reduced active material (~50%) availability, as seen from the charge-discharge characteristics (Figure 3-5). 3,15,22 4 | CONCLUSION The SLA cells were cycled at −10, 0, 25°C, and 40°C to evaluate the performance and degradation mechanism. Discharge profiles demonstrated that aging is faster at elevated temperatures (40°C) than at lower temperatures (−10°C, 0°C, and 25°C). However, capacity degradation is minimal at 0 and −10°C due to reduced active material availability. Moreover, EIS analysis revealed that impedance change was significantly greater at 25°C and 40°C as compared to that at 0°C and −10°C. The charge-transfer resistance is a relatively more governing factor than ohmic resistance for indicating the degradation of the cell. XRD analysis revealed a permanent deposition of SO 4 due to sulfate hardening in the cells after cycling at all temperatures. However, sulfate hardening is significantly higher at 25°C and 40°C as compared to that at 0°C and −10°C, as confirmed by the Pb/S ratio determined from EDS analysis. The study demonstrates that the temperature of operation plays a crucial role in the SOH prediction of SLA batteries.
T A B L E 1 Summary of phases present in the electrode from XRD analysis. Abbreviation: EDS, energy-dispersive X-ray spectroscopy.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.