Electrochemical Stabilization of Self-Extinguishing Electrolyte Solutions with Trimethyl Phosphate By Adding Potassium Salts

Introduction

   Safety, specifically flammability, is still an important concern in lithium-ion batteries, especially as they are now widely used in higher-power and energy applications, such as electric-load leveling and electric vehicle systems. Organophosphorus compounds, such as phosphates, phosphonates, and phosphazenes, are widely known as excellent flame retardants. In our previous research, an ethylene carbonate (EC)-based solution including 50vol% trimethyl phosphate (TMP) self-extinguished after ignition in air. However, TMP has a higher donor number than EC and tends to co-intercalate with lithium ions into the graphite anode, resulting in continuous decomposition and large irreversible capacity loss during the initial cycle [1]. Takeuchi et al. have improved the compatibility of TMP with graphite by introducing a salt composed of a stronger Lewis acid than Li⁺, such as Ca²⁺. This suppression was derived from the solvation structure of Li⁺, which was altered by the stronger Lewis acid, Ca²⁺[2].

   In this study, however, we found that an electrolytic solution including 50vol% TMP could also improve coulombic efficiency in the initial cycle by adding a weaker Lewis acid than Li⁺, such as Na⁺ or K⁺, to the solution. In particular, when potassium bis(trifluoromethanesulfonyl)amide (KTFSA) as a potassium salt was added to the electrolyte, the coulombic efficiency was significantly improved.

Results and Discussion

   Figure 1 shows the first charge-discharge curves in 1.0 mol dm^-3 LiPF₆ EC+EMC+TMP (1:2:3 v/v/v%) with no additives, 0.5 mol dm^-3 sodium bis(trifluoromethanesulfonyl)amide (NaTFSA), and KTFSA. The graphite electrodes were charged and discharged at a constant current of 0.1 mA cm^-2 to 0.01 V and 1.5 V, respectively. The charging capacity was more than 600 mAh g^-1 in the electrolyte solution without additives. However, the discharge capacity was only 109 mAh g^-1. Therefore, the charging capacity exceeded the theoretical capacity of graphite of 372 mAh g^-1, which was derived from reductive decomposition in the graphite electrode. This decomposition can be attributed to co-intercalation of Li⁺with TMP into the graphite interlayer [1].

   Although Na⁺ is a weaker Lewis acid than Li⁺, the electrolyte solutions added NaTFSA could suppress the co-intercalation in the charging process. As a result, the discharge capacity also increased more than that of electrolyte solutions without additives. However, the charging capacity exceeded the theoretical capacity of graphite because it was considered to be difficult to completely suppress co-intercalation of TMP with Li⁺.

   On the basis of the above results, the electrolyte solution added KTFSA consisting of K⁺ (which is a weaker Lewis acid than Na⁺) was not expected to be suppressed the co-intercalation in the charging process, but the charge-discharge curve was more stabilized than that of the electrolyte solution added NaTFSA. Moreover, three plateaus were observed under 0.2 V in the charging and discharging process resulting from the structural transition of the graphite intercalation compound. The result indicates that Li⁺ can intercalate-deintercalate into the graphite without decomposition. It was considered that this improvement was not derived from the solvation of Li⁺, which was altered by the Lewis acid K⁺. Therefore, the factor derived from this improvement was elucidated by analysis of the electrode after charging or discharging in an electrolyte solution.

References

[1] H. Nakagawa, M. Ochida, Y. Domi, T. Doi., S. Tsubouchi, T. Yamanaka., T. Abe, and Z. Ogumi, J. Power Sources., 212,148 (2012).

[2] S. Takeuchi, S. Yano, T. Fukutsuka, K. Miyazaki, and T. Abe, J. Electrochem. Soc., 159, A2089 (2012).