530
Multi-Cell Stack of Nonaqueous Lithium-Air Battery
We have recently developed a multi-cell stack of nonaqueous lithium-air battery and have demonstrated the discharge/charge cycle. Cells were stacked in parallel, similar to the conventional lithium-ion cell stack, but with ~0.5 mm porous current collector between adjacent cathode layers. In this case, air is passively taken up from the periphery of the stack through pores of the current collector. Note that air (oxygen) is naturally breathed in by discharging and breathed out by charging without any additional mechanics like pumps. We believe that such an air-breathing passive system is indispensable to realize both high energy density and low cost. In addition, mass production process for such a cell stack will be quite compatible with the conventional lithium-ion battery. Each cell of the stack is 20 × 20 to 20 × 42.5 mm2 in area and ~0.5 mm thick, and consists of layers of Li metal anode, separator and porous carbon cathode. The porous carbon cathode was prepared by painting Ketjen black paste on a carbon paper. Ether-based electrolytes such as 1 M LiCF3SO3/TEGDME were used. The cell stack assembly and measurements were done in a dry room with the supply-air dew point of −90°C.
Figure 1 shows discharge and charge profiles for a two-cell stack with cathode layers inside the stack. Stable discharge at a constant current of 1 mA (current density of ~60 μA/cm2) was observed at ~2.7 V, and lasted for 58 h until the voltage dropped to 2.5 V. The energy density for this stack was roughly estimated to be ~150 Wh/kg although the cell stack design had not been optimized yet. By charging at the same current density, the voltage was gradually increased from ~3 V to over 4 V. These discharge/charge profiles are very similar to those for a single cell, which suggests that air is well breathed in and breathed out from the periphery through the narrow paths between cathode layers. We confirmed that the discharge current density can be increased to 0.5 mA/cm2although the voltage is lowered (2.45 V). Preliminary results for ten-cell stack (~6 mm thick in total) also showed similar cycle profiles.
XRD, XANES and FIB-SEM observations showed that Li2O2 was predominantly grown on the carbon cathode during discharge, and decomposed by charging. The observed Li2O2 layer as thick as 200 nm (Fig. 2) strongly suggests that the grown Li2O2 is rather conductive. These results will also be discussed in the presentation.