Investigation of Impedance-Based Parameters in Metal-O2 Batteries for Next Generation of Battery Management Systems

The need for increasing driving range in electric vehicles (EV) has led to research in high capacity batteries beyond Li-ion. Metal-O2 batteries, like Li-O2 and Na-O2, have high theoretical energy densities and thus a possibility to increase the capacity of automotive batteries significantly compared to current Li-ion batteries1,2.

A battery management system (BMS) typically uses a combination of coulomb counting and calibration based on open circuit voltage (OCV) measurements that depend on the state of charge (SOC). Calibration is needed due to the accumulation of errors in the coulomb counting3. Common for most metal-O2 batteries is that the chemistry is unchanged during discharge and charge (assuming no degradation).

This means that the OCV does not change as function of SOC. Furthermore, constant current measurements show a flat discharge plateau in large parts of the discharge period. This has been shown for Li-O2 and Na-O2 batteries2,4 and in some cases also for Mg-O2 and Al-O2 5,6. Taking the well studied Li-O2 battery as an example, the dominating process during discharge is reduction of oxygen to produce Li2O2 on top of an existing Li2O2 layer.

As this process continues during the entire discharge, both OCV and discharge potential is constant until the end of discharge, where other processes become limiting, as shown in figure 1. New methods have to be developed to overcome the constant OCV and flat discharge plateau that otherwise would complicate both battery management and accurate online prediction of available capacity in metal-O2 batteries.

Furthermore, the increased capacity predicted for the metal-O2 batteries will result in longer discharge periods without charging, thus increasing the need for accurate calibration of the capacity in the range between ~10% and ~90% SOC. In the following we propose a method to accurately predict the capacity of metal-O2 batteries using impedance spectroscopy for calibration of the SOC tracking algorithm as well as gauging the degradation of the battery materials.

This method can easily be implemented in an automotive BMS with only a few extra components and preliminary experiments have shown that the impedance measurements can be performed both during rest periods and under load conditions. This makes the method applicable not only for EVs but for batteries in a large range of electrical devices as the measurements can be performed when needed, thus maintaining a high level of accuracy for the SOC estimation and state of degradation. [1] Christensen, J., Albertus, P., Sánchez-Carrera, R.

S., Lohmann, T., Kozinsky, B., Liedtke, R., et al. (2012). A Critical Review of Li/Air Batteries. Journal of the Electrochemical Society, 159(2), R1. doi:10.1149/2.086202jes [2] Hartmann, P. (2012). A rechargeable room-temperature sodium superoxide (NaO2) battery. Nature Materials, 12(3), 228–232. doi:10.1038/nmat3486 [3] Ng, K.

S., Moo, C.-S., Chen, Y.-P., & Hsieh, Y.-C. (2009). Enhanced coulomb counting method for estimating state-of-charge and state-of-health of lithium-ion batteries. Applied Energy, 86(9), 1506–1511. doi:10.1016/j.apenergy.2008.11.021 [4] McCloskey, B. D., Garcia, J. M., & Luntz, A. C. (2014). Chemical and Electrochemical Differences in Nonaqueous Li–O 2and Na–O2 Batteries.

The Journal of Physical Chemistry Letters, 5(7), 1230–1235. doi:10.1021/jz500494s [5] Shiga, T., Hase, Y., Kato, Y., Inoue, M., & Takechi, K. (2013). A rechargeable non-aqueous Mg–O2 battery. Chemical Communications, 49(80), 9152–9154. doi:10.1039/c3cc43477j [6] Revel, R., Audichon, T., & Gonzalez, S. (2014).

Non-aqueous aluminium-air battery based on ionic liquid electrolyte. Journal of Power Sources, 272(c), 415–421. doi:10.1016/j.jpowsour.2014.08.056 [Figure]