Dependence of Discharge Product Morphology on Discharge Rate in Li-O­2 Batteries

Li-O₂ batteries have the potential to achieve a much higher (3-5x)¹ specific energy than current state of the art Li-ion batteries, which could allow battery-powered electric vehicles to compete with traditional gas-powered cars.² The greatest barriers to Li-O₂ technology are limited cycle life and low battery efficiency. A better understanding of the mechanism of discharge product formation and dissolution would help to overcome these barriers. Controlled observation of discharge-product (Li₂O₂) morphology may provide insight into reaction mechanisms.

Several reports of the morphology of electrochemically formed Li₂O₂ show the size and shape of Li₂O₂ particles and relate these characteristics to battery performance^3,4,5. Typically, the discharge products are discussed as either small particles⁴ or nanosheets⁵ on the order of 10 nm, or larger toroidal shapes on the order of 100 nm. The small particles have a lower overpotential during recharge than the larger particles. If Li₂O₂ is formed by an elementary electron exchange, then the formation of large particles suggests a discharge mechanism where electrons travel over hundreds of nanometers, which suggests a complex charge transfer mechanism since bulk Li₂O₂is an electrical insulator.

In this talk we will report a study of the rate dependence of Li-O₂ discharge-product morphology during first discharge. At each of four rates ten independent Li-O₂ cells were discharged to provide a statistically significant assessment of discharge capacity. Figure 1 shows the variation of discharge capacity with discharge rate with the inset showing the discharge curves at 0.2 mA/cm².

Cathodes with near-average capacity were selected for further characterization. X-ray diffraction showed that Li₂O₂ was the primary crystalline discharge product at all rates. Scanning electron microscopy (SEM) showed that the Li₂O₂ deposits comprised roughly cylindrical particles at lower rates, and needle-like particles at the highest rate.  The cylinders had constant radii at different rates, which were comparable to the lengths of the needles. The heights of cylindrical products appeared to rise with decreasing rate. Figure 2 shows the volume of the Li₂O₂ particles as a function of discharge rate. The inset SEM images show the cylinders at end-of-discharge within 1 µm x 1 µm areas. The characteristic shapes (and aspect ratios) of particles yielded by image processing of > 25 particles from each electrode are also shown schematically. A nucleation and growth mechanism for Li₂O₂will be proposed and the impacts of this mechanism on battery performance will be discussed.

1. Lu, Y. C., B. M. Gallant, et al. (2013). Energy & Environmental Science 6(3): 750-768.

2. Christensen, J., P. Albertus, et al. (2012). J. Electrochem. Soc. 159(2): R1-R30.

3. Adams, B. D., C. Radtke, et al. (2013). Energy & Environmental Science 6(6): 1772-1778.

4. Gallant, B. M., D. G. Kwabi, et al. (2013). Energy & Environmental Science 6(8): 2518-2528.

5. Xu, J. J., Z. L. Wang, et al. (2013). Nature Communications 4.