Cycle Life of Manganese Dioxide Cathodes for Large-Scale Storage: Effects of Discharge Rate and Depth

The basis materials of batteries for large-scale storage must be low-cost, earth-abundant, and safe at the desired scale. The alkaline MnO₂ cathode, typically found in low-cost consumer primaries, fulfills these requirements. The U.S. Department of Energy ARPA-E goals for large-scale storage are $100 per kWh and a 5000-cycle life. A standard bobbin-style alkaline cell costs roughly $20 per kWh. Thus a shallow-cycled alkaline cell is a sufficiently inexpensive basis for large-scale storage. However, in this case an extended 5000-cycle life becomes the relevant engineering challenge.¹

Discharge of an MnO₂ cathode involves many reactions, both chemical and electrochemical, and only certain of these are fully reversible. For example the proton/electron insertion to MnOOH is usually assumed reversible. However, a subsequent solution-phase reduction to Mn(OH)₂ can then lead to Mn₃O₄, which is considered irreversible.² Confining the cathode cycling window to a well-controlled state-of-charge range involving only MnOOH can in theory result in perfect reversibility. However, current distributions can result in locally high reaction rates in the electrode, and chemical reactions can be challenging to control, as they do not respond directly to the potential.

We report studies of long-term cycling with prismatic zinc-manganese dioxide batteries in which zinc was in excess, and the cycling depth was set as a fraction of the MnO₂ capacity. Cycling results are shown in Figure 1 for a collection of cells in which the cycling depth and discharge rate were varied. Cell energy efficiency was initially between 80-90% for all cells, and was higher when a lower fraction of the MnO₂ capacity was cycled. The discharge end voltage was taken as a general metric of cell health, and in all cases evolved over time, generally decreasing with cell age. Under certain profile conditions, cells cycled uninterrupted for years with stable efficiency. In Figure 1 cell (2) exceeded 3000 cycles, and cell (3) attained 4000 cycles.

In the early stages of cycling, a selection of cells were halted and material changes within the cells were analyzed. This information was combined with that from post mortem analyses of decommissioned cells such as those in Figure 1. The goal was delineation of the processes that most frequently lead to cell failure, and knowledge of the rates at which these processes occur. In most decommissioned cells, anode, cathode, and separator were all found altered. In particular we tracked the formation of ZnMn₂O₄ or hetaerolite in the cell cathodes, to assess the impact of this byproduct on cycling.

References

1. J.W. Gallaway, C.K. Erdonmez, Z. Zhong, M. Croft, L.A. Sviridov, T.Z. Sholklapper, D.E. Turney, S. Banerjee, and D.A. Steingart, J. Mater. Chem. A, 2, 2757 (2014).
2. L. Binder, K. Kordesch, and P. UrdI, J Electrochem Soc, 143(1) 13 (1996).

Acknowledgement: This research was funded by the Advanced Research Projects Agency – Energy (ARPA-E), U.S. Department of Energy, under Award Number DE-AR0000150.

Figure 1. Cycling results for prismatic Zn-MnO₂ batteries. (a) Energy efficiency per cycle. (b) Discharge end voltage. Cycling profile was constant-current/constant-voltage (CCCV) for all cells with a maximum voltage of 1.65 V and a minimum voltage of 1 V, which was in some cases relaxed at the end of cell life. Cycled fraction of MnO₂ capacity and discharge rate were: (1) 10% cycle, 2 h rate, initial build design (2) 10% cycle, 2 h rate (3) 5% cycle, 1 h rate (4) 10% cycle, 1 h rate (5) 10% cycle, 45 min rate.