Sodium-ion batteries (SIBs) are emerging as a more cost-effective and environmentally friendly alternative to the commonly used lithium-ion batteries (LIBs). With energy densities anticipated to reach up to 150 Wh/kg, SIBs have the potential to replace lithium iron phosphate (LFP) batteries in many current applications. While research on SIBs started over 50 years ago, commercial SIB cells have only recently became available. In this study, we conducted a systematic cyclic aging study on two different SIB cell types under varying load rates. Both cell types use hard carbon as anode and layered oxide, namely NaFe¬1/3Ni1/3Mn1/3, as cathode material. Though, we classified these cells as high-power (SIBHP), and high-energy (SIBHE) types based on their maximum charge currents.
Throughout the aging study, some cells showed a sudden increase in temperature accompanied by a voltage drop during charge. This end-of-charge temperature rise (EOCTR) led to a substantial increase in charge capacity, while the discharge capacity remained largely unchanged compared to previous cycles. As a result, the coulombic efficiency decreased significantly, with reductions ranging from few percent points to over 50%. The effect happens at lower state-of-charges when the environment’s temperature decreases or the charge rate increases. Furthermore, we could detect that pausing before the test can increase the severity of the EOCTR.
To evaluate the cause of this effect, we opened the cells under Argon atmosphere. In cells that have shown EOCTR, we observed plated sodium on the anode. This was not the case for cells without such behavior. The combined observation of strong plating, reduced coulombic efficiency and increase in temperature suggest the occurrence of internal short circuits. Nevertheless, providing proof of internal short circuits is challenging.
Here, we applied a physics-based approach to detect the occurrence of internal short circuits. Cells were placed into liquid nitrogen, lowering the temperature below -195°C. At this temperature, the liquid electrolyte freezes, and no electrochemical reactions can occur anymore. This is accompanied by a voltage drop to 0 V. Internal short circuits can then be identified by measuring the resistance between the positive and negative terminals using a standard multimeter. A healthy cell is expected to show high, ideally infinite, resistance, while a cell with an internal short circuit will exhibit a very low, measurable resistance.
We have chosen several cells, which have previously shown the effect of sudden temperature rise. These cells were at rest at room temperature for more than 24h. Then, we put these cells into liquid nitrogen for at least 10 minutes. We monitored their voltage and could see that, according to expectation, it drops to 0V within the first 2 minutes. Then, directly after removing the cells from the liquid nitrogen, we measured their resistance using a multimeter. Results have shown that the cells which have shown the temperature rise, had significant lower cryo-resistances (for instance 10 kOhm versus 1 MOhm in the case of SIBHP). Then, we charged the SIBs with 0.5C at 25°C ambient temperature. Once a temperature of 29°C was reached (which is above the temperature a healthy cell would reach), we stopped the charging process and freezed the cells again as described above. Now, the measured cryo-resistance was strongly decreased down to values below 100 Ohm, clearly indicating strong short circuits within the cell.
After pausing for 1h at 25°C, we freezed the cells again and measured their cryo-resistance. Now, the cells showed cryo-resistances in the range between 1 and 10 kOhm. This procedure is then again repeated in several time steps clearly showing that, without any load, the internal resistance strongly increases again and reaches values close to the initial values after 24 hours. This can be explained such that during charging internal short circuits due to dendrites build. However, when the cell is stored at room temperature, this short circuits degrade over time. As a result, the cryo-resistance increases again, as the internal short circuits reduce.
Concluding, we used an easy, but very reliable method to show that the commercially available sodium-ion cells show internal short circuits. Furthermore, we could directly measure that the internal soft short circuits recede without any external influences at room temperature.