The origin of cycle life degradation of a lead-acid battery under constant voltage charging

Received Sep 23, 2020 Revised Feb 19, 2021 Accepted Mar 6, 2021 Due to its low cost and recycle-ability, the lead-acid battery is widely used in mobile and stationary applications. Despite much research on lead-acid batteries, the effect of charging voltage on the degradation mechanism requires further investigation. In particular, the origin of cycle life degradation remains unclear. In the present work, by using electrochemical tests and materials characterization, we studied the effect of charging voltage at voltages slightly higher than the open-circuit potential (OCP) i.e., 103107% OCP, on the battery life cycle. The highest degradation was observed at 105% OCP charging voltage. Based on the materials characterization results, we found that the degradation of a lead-acid battery is influenced by the amount of hard sulfate and the sulfate particles' size.


INTRODUCTION
The development of renewable energy is one of the factors in the growing global demand for energy storage. The share of renewable energy is expected to rise from 25% in 2015 to 85% in 2050, with the highest predicted proportion generated by wind power and photovoltaic (PV) sources [1]. Wind and solar power sources are intermittent energy sources. This intermittent energy production has to be stabilized using either energy from the grid (for an interconnected power plant) or energy storage (for a stand-alone power system) [2], [3]. In combination with energy storage, e.g., a battery, a controller balances the power generation intermittency and the load dynamics [4]- [7].
A battery is an energy storage device that stores electrical energy in the form of chemical energy. The lead-acid battery is one of the most widely used rechargeable/secondary batteries. This battery type is well established for industrial applications (e.g., stationary units and start-stop in automobiles) and has been successfully applied to utility energy storage [8], [9]. Lead-acid batteries are widely used because of their low cost, reliability, simple manufacturing, and an established method for lead recycling [10]- [12].
The mode of charging and discharging depends on the application. For example, renewable energy storage requires the battery to be at a partial state of charge. In other instances, in automotive startinglighting-igniting, the battery primarily operates at a float charging. It is well-known that charge-discharge modes will cause different deterioration types, resulting in a variation in the battery's lifetime [13], [14]. The variety of lead-acid battery failure modes can be divided into three categories, i) positive active mass degradation and loss of adherence, ii) sulfation of the active material, and iii) other factors, such as the  [21]. Despite a large amount of research on the various modes of charging, the effect of charging voltage on the battery failure mode (particularly the relationship between the charging voltage and their impact on battery degradation) requires further investigation. The present work involves analyzing the effect of the charging voltage at slightly above open-circuit potential (OCP) on the cycle life and the material degradation. We mostly focus on the materials characterization aspect to better understand the degradation process.

RESEARCH METHOD
In this study, a modified battery taken from a commercial VRLA (valve regulated lead acid) battery consisting of six cells with a capacity of 12V-3.5Ah was used. The experimental setup preparation was performed by removing the third and fourth cells. Thus, each battery was converted into two batteries containing two cells each (4V-1.17Ah capacity) to match the potentiostat's capability. A metrohm autolab PGSTAT302N having ±10V voltage range was used to perform charging and discharging (electrochemical test). Figure 1 shows the electrochemical test scheme and setup using the experiment. The test was carried out with the chrono charge-discharge potentiostatic method.   Table 1 shows the list of tested samples and the condition for experiments. The charging voltage of the lead-acid battery depends on the application and can be higher than the open-circuit voltage. The OCP or open-circuit voltage of a measured lead-acid battery cell was used as a charging-discharging reference [22]- [24]. A potential of 4.28V was achieved during the OCP measurements. All samples were discharged at a set voltage of 97.3% OCP. The first, second, and third samples were charged using different voltages of 103.3% OCP, 105.3% OCP, and 107.3% OCP. The charge-discharge process was carried out for 40 cycles, and each charging/discharging time was set for 3600 seconds. An example of the charge and discharge process is shown in Figure 2. The discharge capacity (C d ) of the samples was obtained by integrating the discharge current (I) with respect to time (t), as shown in (1) [25]. (1) X-ray diffraction (XRD) analysis was performed to investigate the new phase formed on the negative and positive electrodes of samples examined during the electrochemical tests. The XRD data were collected using an X-ray diffractometer PANalytical X'pert PRO with a Cu-Kα radiation source. The morphological changes due to the new phase formed on the negative and positive electrodes were characterized using a CARL ZEISS EVO MA10 scanning electron microscope (SEM). The SEM was used to evaluate the surface characteristics, including the particle size distribution, of both the positive and negative electrodes [20], [26]- [31].  Figure 3 shows the degradation of the discharge capacity of a battery throughout the cycle of the electrochemical test. The percentage of capacity degradation is calculated from the difference between the maximum capacity at the 1 st cycle and the latest cycle. The degradation of capacity from the highest to the lowest is 33.84%, 32.98%, and 24.96% is observed from the sample at charging voltage of 105.3% OCP, 103.3% OCP, 107.3% OCP, respectively.

Phase structural characterization
To gain further insight into the degradation mechanism, XRD patterns of the samples before and after the electrochemical test are shown in Figure 4 for the positive electrode and in Figure 5 for the negative electrode. The XRD spectrum in Figure 4 (a) reveals that the original positive active materials were welldefined PbO 2 crystals. Upon cycling, PbSO 4 formation occurred; the PbSO 4 was stable upon cycling. According to the XRD spectra b to c shown in Figure 4, the fraction of PbSO 4 increased. The sulfation in this positive electrode decreased because the charging voltage is high. According to Figure 5 (a), the active material is comprised of well-crystallized Pb for the negative electrode before the charging cycle. The sulfation also occurs upon cycling. Thus, sulfation was observed in both negative and positive active materials; this observation was also made in previous work [32].

Microstructural characterization
SEM analysis on the positive and negative active materials was conducted on the sample after cycling using a similar sample to that studied in the XRD analysis. The results of SEM image analysis on the discharged active materials are shown in Figure 6. It can be seen that, before the cycle test, the positive active materials consist of PbO 2 crystal. After the cycling at different voltages, the sulfate particles start to appear, with the size being charging voltage dependence (Figure 6 (b) to Figure 6 (d). The SEM image for the negative active materials is shown in Figure 7. A similar trend was found; the higher the charge voltage, the smaller the lead sulfate formed. It can be seen that the lead sulfate formed on the negative electrode of the 103.3% OCP test tends to be small, and that of the 105.3% OCP test tends to be large. In contrast, at higher charging voltage, the sulfate particle becomes smaller. The lead sulfate, which does react back to pure Pb, slowly converts into a stable crystalline form of sulfate that no longer dissolves on recharging. More crystalline lead sulfate in the active material reduces the reversible active phase, resulting in battery degradation. 20   The particle size can be measured statistically analyzed from an SEM image using linear intercept methods. The resulting plot of the particle size is shown in the Appendix. A direct comparison of the average particle size at each electrode is presented in Table 2. The table shows the lead sulfate average particle size before and after 40 times the charge and discharges cycling. The particle size at the positive and negative electrode was initially 0.353 and 0.655m, respectively. After the cycle test, the particle size increases to approximately 1 m, with the size being dependent on the charging voltage. The smallest particle size for the cycled samples was observed for the battery that was charged at 107.3% OCP. The largest particle size was

CONCLUSION
Due to its scalable and mature manufacturing and recycling process, the lead-acid battery is widely applied in mobile and stationary applications. The charging voltage and scheme determine the cycle of the battery. It is commonly believed that the amount of stable sulfate particles causes lead-acid battery degradation. Based on the results of an electrochemical and materials characterization study, we showed direct evidence of sulfate particle size's influence on the degradation mechanism. The capacity degradation depends on the charging voltages, with the degradation after 40 cycles of 33.84%, 32.98%, and 24.96% at the charging voltage of 105.3% OCP, 103.3% OCP, and 107.3% OCP, respectively. This degradation was found to originate from the size and amount of lead sulfate formed upon cycling. The phase fraction measurement confirmed this cause of degradation via XRD measurements and the particle size via SEM measurements. From those results, it can be concluded that degradation depends not only on the number of sulfate particles but also on the particles' size. The present work explains the sulfation process and should apply to many leadacid systems. Further research investigating various charging schemes related to the sulfate particle dissolution mechanism is required.

ACKNOWLEDGMENTS
Authors thank Departemen Teknik Mesin ITS for funding this present work through Skema Penelitian Mandiri Departemen year 2017.