Critical Issues of Fluorinated Alkoxyborate-Based Electrolytes in Magnesium Battery Applications

Abstract

Development of non-corrosive but highly efficient electrolytes has been long-standing challenges in magnesium rechargeable battery (MRB) research fields. The surface passivation film formed upon undesired reactions of magnesium metal with conventional electrolytes indeed inhibits efficient interfacial electrochemical reactions. In recent years, the new class of electrolyte materials has been reported, and these electrolytes especially incorporating the weakly coordinated anions (WCAs) have overcome some serious problems associated with conventional electrolytes.1 –3 Among the reported WCA-based electrolytes, certain fluorinated alkoxyborate-based electrolytes, e.g. Mg[B(HFIP)4]2/glyme (HFIP = hexafluoroisopropoxy-group), are regarded as promising candidates for practical MRB materialization owing to the excellent electrochemical activity and synthetic accessibility. We herein report several critical issues associated with the use of Mg[B(HFIP)4]2/glyme electrolytes in MRBs. The electrolyte containing Mg[B(HFIP)4]2 prepared from Mg(BH4)2 requires several ten-cycles of pre-activation process. The reductive decomposition of glymes (not [B(HFIP)4]−) in the electrolyte solutions takes place upon just contacting the electrolytes with magnesium metal. Non dendritic short-circuit should also be solved for practical application. The Mg[B(HFIP)4]2 salt was synthesized according to the procedure reported elsewhere,3 and the electrolyte solutions were prepared by dissolving pre-weighted Mg[B(HFIP)4]2 in a series of glymes; monoglyme (G1), diglyme (G2), and triglyme (G3). The coordination and dissolution states of the electrolytes were evaluated by the combination of crystallography, vibrational spectroscopy, and first-principle calculations. The soak test was performed by just soaking mechanically polished Mg ribbon into the prepared electrolytes for three days. Electrochemical magnesium deposition/dissolution cycles were evaluated using a separable cell, in which the AZ31 alloy and copper foils were served as the counter and working electrodes, respectively, and the glass fiber (GA200) was served as a separator. A constant cathodic/anodic current of 0.5 mA cm-2 was applied for 30 minutes in each cycle at 30 °C. The morphological and chemical analysis of the soaked Mg ribbon, cycled AZ31, and cycled separators were conducted by the scanning electron microscope (SEM), the energy dispersive X-ray spectroscopy (EDX), and X-ray photoelectron spectroscopy (XPS). The SEM observation on the soaked magnesium metal clearly indicated that the Mg[B(HFIP)4]2-based electrolytes can be reacted with magnesium metal even though no current applied. The soaked metal surface was unevenly reacted with the electrolytes – partly corroded and decomposition products found while other part kept pristine. The EDX on the soaked metal revealed that the decomposition products were consisting of Mg, C, and O. No F and B signal was detected, indicating decomposition of not [B(HFIP)4]− but glymes taking place upon just contacting electrolyte solutions with magnesium. Glymes are known to be stable toward reduction. The reason why rather glymes were reductively decomposed by magnesium would be the dissociation state of the salt and the strong electric field produced by Mg2+. As evidenced by the Raman spectra, Mg[B(HFIP)4]2 is highly dissociated in glymes, and Mg2+ was totally surrounded by glymes. The electric field of Mg2+ strongly polarizes the coordinated glymes, thereby resulting in lowering their LUMO energy level and consequently making glymes reductively fragile. This hypothesis was supported by the first-principle calculations. On the other hand, non-dendritic short-circuit of the cells upon repeated deposition/dissolution cycling is also problematic. The SEM images of the cycled separator obtained by the focused-ion-beam (FIB) processing clearly indicates the presence of the granular magnesium deposits from top to bottom of the separator matrix. The uneven distribution of the reactive site of magnesium metal probably causes such observations. The synthetic issues of the Mg[B(HFIP)4]2 salt will also be discussed.Acknowledgement: This study was financially supported by JST ALCA-SPRING program (Grant Number JPMJAL1301). The authors also thank to NIMS Battery Research Platform for the Raman, SEM, EDX, FIB, and XPS measurements.References Tutusaus et al., Angew. Chem. Int. Ed., 2015, 54, 7900. B. Arnold et al., ACS Energy Lett., 2016, 1, 1227. Z.-Karger et al., J. Mater. Chem. A, 2017, 5, 10815.