Ionic liquids consisting of bis(fluorosulfonyl)imide (FSI − ) anion show promise as electrolytes for Li-ion-based electric storage devices, as they exhibit relatively low viscosity, high chemical stability, and form robust solid−electrolyte interphase (SEI) protecting liquid electrolyte from further breakdown on the electrode. These ionic liquids have been reported to inhibit dendrite formation on lithium metal and lithiated graphite electrodes, which also relates to the unusual SEI properties. In this study, we examine the chemistry aspects that may account for this behavior. Radiolysis was used to induce redox reactions of FSI − anions in model systems, and matrix isolation electron paramagnetic resonance was used to identify radical (ion) intermediates generated in these reactions. Our results suggest that qualitative differences between such ionic liquid electrolytes versus common carbonate electrolytes reflect ease of mineralization of the reduced anion without the concurrent generation of organic radicals and/or elimination of gaseous products in side reactions of the corresponding radical intermediates.

Room temperature ionic liquids (IL) find increasing use for the replacement of organic solvents in practical applications, including their use in solar cells and electrolytes for metal deposition, and as extraction solvents for the reprocessing of spent nuclear fuel. The radiation stability of ILs is an important concern for some of these applications, as previous studies suggested extensive fragmentation of the constituent ions upon irradiation. In the present study, electron paramagnetic resonance (EPR) spectroscopy has been used to identify fragmentation pathways for constituent anions in ammonium, phosphonium, and imidazolium ILs. Many of these detrimental reactions are initiated by radiation-induced redox processes involving these anions. Scission of the oxidized anions is the main fragmentation pathway for the majority of the practically important anions; (internal) proton transfer involving the aliphatic arms of these anions is a competing reaction. For perfluorinated anions, fluoride loss following dissociative electron attachment to the anion can be even more prominent than this oxidative fragmentation. Bond scission in the anion was also observed for NO(3)(-) and B(CN)(4)(-) anions and indirectly implicated for BF(4)(-) and PF(6)(-) anions. Among small anions, CF(3)SO(3)(-) and N(CN)(2)(-) are the most stable. Among larger anions, the derivatives of benzoate and imide anions were found to be relatively stable. This stability is due to suppression of the oxidative fragmentation. For benzoates, this is a consequence of the extensive sharing of unpaired electron density by the π-system in the corresponding neutral radical; for the imides, this stability could be the consequence of N-N σ(2)σ(*1) bond formation involving the parent anion. While fragmentation does not occur for these "exceptional" anions, H atom addition and electron attachment are prominent. Among the typically used constituent anions, aliphatic carboxylates were found to be the least resistant to oxidative fragmentation, followed by (di)alkyl phosphates and alkanesulfonates. The discussion of the radiation stability of ILs is continued in the second part of this study, which examines the fate of organic cations in such liquids.

Whereas there are numerous experimental and computational studies of electrochemical reduction leading to the formation of solid-electrolyte interface (SEI) in lithium-ion batteries, so far there have been no direct spectroscopic observations of radical intermediates involved in the SEI formation. In Part 1 of this series, radiolysis and laser photoionization of carbonate electrolytes are used to observe and identify these reaction intermediates using electron paramagnetic resonance spectroscopy. Our study indicates that the suggested scenarios for electrolyte reduction require elaboration. In particular, we establish the occurrence of efficient H abstraction and 1,2-migration involving radicals generated through the reductive ring-opening. Instead of the primary radicals postulated in the current models, secondary and tertiary radicals are generated, biasing the subsequent chemistry to radical disproportionation. The consequences of this bias for radical and anionic polymerization are examined, and it is suggested that branching and the formation of a polymer network is favored. We argue that this chemistry accounts for some of the heretofore unexplained properties of SEI, including the dramatic difference in solvent permeability for SEIs derived from ethylene carbonate and propylene carbonate.