SEM Observation of Lithium Electrodeposition Behavior in Molten Lithium Fluorosulfonyl(trifluorosulfonyl)Amide

Abstract

Dendritic growth of lithium during charging is one of the largest obstacles against the widely spreading commercialization of the battery with lithium metal anode.1 Generally, one of reasons why the dendritic lithium is deposited is insufficient mass transfer under diffusion-limited condition, as is often the case for silver or zinc electrodeposition. Some researchers have reported lithium electrodeposition from such a viewpoint.2,3 Sano et al. has also examined the lithium electrodeposition in some binary ionic liquids, and reported that diffusion-limitied condition enhances the dendritic growth.4 On the other hand, Aurbach's group proposed that the non-uniformity of surface film formed by the reaction between the electrolyte and lithium metal also can cause the dendritic growth.1 The mechanism of lithium dendritic growth is now under discussion, and the behavior of dendrite growth remains to be clarified, for example, whether the lithium grows from the bottom or on the top.1-5 In this study our group focused on single lithium molten salt, because in single lithium molten salts, unlike the conventional lithium salt solution or binary ionic liquid containing lithium salt, the lithium ion concentration is thought to be constant at any time and location even during electrodeposition, so that mass transfer of lithium ion is usually sufficient at the deposit site during the electrodeposition. Recently, Kubota et al firstly reported the synthesis of an intermediate temperature lithium molten salt, lithium (fluorosulfonyl)(trifluoromethylsulfonyl)amide (Li[FTA] (Li[fTfN])), with a melting point of 100 ºC,5,6 which can be used with metallic lithium unlike most single lithium molten salts with a melting point exceeding that of metallic lithium (180 ºC). In the present work, lithium was electrodeposited in Li[FTA], without gradient of lithium ion concentration, in order to understand further the dendrite growth behavior and its mechanism. Li[FTA] was used as electrolyte without any solvent. Two electrode cells with a 1-mm-gap spacer were constructed, according to the previous report,4 in an Ar-filled glove box using nickel foil as a working electrode and lithium foil as a counter electrode. A current density from 50 to 1000 μA cm-2 was applied for charge density of 1 A s cm-2 at a temperature of 150 ºC. Lithium electrodeposited on working electrode was observed by SEM after washing with dimethyl carbonate or tetrahydrofuran. Figure 1 shows the SEM images of the working electrode after electrodeposition at a current density of (a) 200 and (b) 1000 μA cm-2 for 1 A s cm-2 at 150 ºC, respectively. The current was continually flowed through the cell even at the current density of 1000 μA cm-2, although the diffusion co-efficient of lithium ion at 150 ºC is very small. Note that the current cannot be continuously flowed with the same cell for the binary ionic liquid (PP13[TFSA]+Li[TFSA] (100:10 wt) (PP13; N-methyl-N-propylpiperidinium, TFSA; bis(trifluoromethanesulfonyl)amide)), lithium ion diffusion coefficient (room temperature) of which is similar, since the lithium ion concentration at the working electrode surface gradually decreases to almost zero with such a high current density. The shapes of the deposits are dendritic in the two conditions. The sizes of the deposits are smaller and the number of the deposits is larger when the current density was high. The same tendency was also confirmed with the binary ionic liquid.4 The number of the deposits in a unit area was rather small compared with literature,4 which is thought to arise from small charge-transfer resistance leading the small deposition overpotential. Dendritic lithium was observed in this experiment, where there is no change in the lithium-ion concentration in the electrolyte on the working electrode surface during deposition. A more detailed discussion on the mechanism for the dendrite formation is currently underway. 1. D. Aurbach et al., Solid State Ionics, 148, 405 (2002). 2. K. Nishikawa et al., J. Electroanal. Chem., 661, 84 (2011). 3. J. Steiger et al., Electrochem. Commun., 50, 11 (2015). 4. H. Sano et al., J. Electrochem. Soc., 161, A1236 (2014). 5. J. Yamaki et al., J. Power Sources, 74, 219 (1998). 5. K. Kubota et al., Chem. Lett., 39, 1303 (2010). 6. K. Kubota and H. Matsumoto, J. Phys. Chem. C, 117, 18829 (2013). Figure 1