Probing the Lithium-Sulfur Redox Reactions By Rotating-Ring Disk Electrode Measurements

Abstract

Introduction Rechargeable lithium-sulfur batteries promise to provide a 2 to 3 times higher specific energy than conventional lithium rechargeable batteries. Detailed mechanistic understanding of sulfur redox reactions is critical for developing efficient and stable lithium-sulfur batteries. It is proposed that the solvent´s ability to solvate the primary polysulfide reduction products will influence the reduction/oxidation process.1 However, in lots of studies, the maximum reduction charge obtained is ~4 e-/S8 for potentials as low as +1 to +1.5 V vs. Li/Li+, i.e., significantly lower than the 12 to 16 e-/S8 which are obtained for long-term discharge curves (fraction of hours to hours) in battery cells using electrodes based on high-surface area carbons.2-5 Here, we examine the differences in sulfur charge/discharge behavior in low (DOL:DME) vs. high-dielectric solvents (DMSO) and over short and long time scales using rotating ring disk voltammetry (RRDE) on flat model electrode6 and galvanostatic characterization in a specially designed Li-sulfur catholyte battery cell, respectively. Results and Discussion Figure 1 shows the ring (Fig. 1a), disk (Fig. 1b) current and kinetic current (Fig. 1c) of sulfur reduction in DMSO and DOL:DME(1:1). By using Levich-Koutecky slopes, the number of electrons exchanged during the reaction can be obtained. For DMSO (blue lines/symbols in Fig. 1), the first reduction wave (~ -0.9 to ~ -1.5 VFc corresponds to a 2-electron reduction of S8 to S8 2- and has a Tafel slope of ~90 mV/decade. Considering the similar Tafel slope observed for the 1st and 2nd reduction waves, we hypothesize that the 2nd reduction wave is likely the series of two 2-electron processes with the second step being rate limiting (i.e., S8 + 2 e- --> S8 2- followed by S8 2- + 2 e- --> S8 4- --> 2 S4 2-). For DOL:DME (red lines/symbols in Fig. 1) only one reduction step corresponding to ~5 e-/S8 is observed. This could be related to the solvation ability of the solvent. The two well-separated reduction waves observed in DMSO can be rationalized by strongly solvated and thus energetically stabilized S8 2- anions so that their further electrochemical reduction requires a significantly more cathodic potential. On the contrary, if S8 2- is weakly solvated and its formation is thus energetically unfavorable, its formation might be delayed until more cathodic potentials are reached at which point multiple electron transfer reactions might occur. The influence of the solvent’s solvation power on the sulfur reduction processes and the implications for Li-S batteries will be discussed Acknowledgments The authors are grateful to Dr. Juan Herranz-Salaner for RRDE cell design and scientific discussion, Dr. Himendra Jha for experimental assistance and scientific discussion, and Johannes Hattendorff for the measurement of solvent viscosity. Financial support of this research by the Bavarian Ministry of Economic Affairs and Media, Energy and Technology performed under the auspices of the EEBatt project. Figure 1. Capacitively, non-Ohmically corrected ring (a) and disk (b) currents recorded at 50 mV/s in Ar-saturated 4 mM S8 – 0.2 M LiClO4 DMSO (Blue) and 4 mM S8 – 1.0 M LiTFSI DOL:DME(1:1) (Red) at rotation rates between 100 and 1600 rpm and continuously holding the Au ring electrode at 0.494 VFc for DMSO and at 0.557 VFc for DOL:DME. (c) Kinetic currents measured in DMSO and DOL:DME. 60 mV/dec and 120 mV/dec lines are provided as a guide-to-the-eye for comparison. Note that 0 VFc = +3.756 VLi in DMSO and 0 VFc = +3.283 VLi in DOL:DME.