As societal needs move towards more sustainable and efficient energy landscapes, electrical energy storage is taking a center stage. The need for devices with greater safety, higher energy and power densities, and longer lifetimes is ever increasing, with applications ranging from stationary storage for intermittent sources (e.g. wind and solar) and grid load-leveling, to electrification of transportation and consumer electronics.

The development of longer cycle life Li-ion batteries is an important goal for the large-scale deployment of electrified transportation.  A key requirement for these devices is an improved solvent-electrolyte system, capable of withstanding the extreme potentials and temperature ranges expected for modern Li-ion cells.

We have undertaken fundamental experimental and computational studies to determine the significance of decompositions of linear and cyclic carbonates to commonly observed nucleophilic species in cycled Li-ion battery electrolytes. The formation of these nucleophiles has been widely reported, as have the straightforward follow-up reactions.^1,2,3, although, the mechanism(s) by which these nucleophiles are formed has not been settled.

Moreover, the relevant kinetic aspects of these nucleophilic reactions have not been widely explored, nor have possible reactions that proceed at higher temperature or longer timescales. Since these reactions are generally in competition with one another, and Li-ion electrolyte formulations contain multiple solvents in varying concentrations, a view of the relative rates of these reactions is an important piece of formulating a long cycle life electrolyte.

Rather than allow these species to react in multiple experimental studies, across temperatures and concentration ranges, computational methods are well suited to determine the relevant reaction kinetics and thermodynamics with a high degree of confidence. For example, products can be oxidized at relatively low cathode potentials, or they can attack cyclic and linear solvents to generate a host of oligomers and other small molecules.^4,5 Many of the products of these reactions can also be soluble and electrochemically unstable.⁶

In this report, we will present headspace GC-MS data for a cycled EC-EMC + LiPF₆ electrolyte, with consideration toward species found at the anode and cathode. The observed products will then be compared with data from computational studies of electrochemical and chemical reactions of Li-ion battery solvents.

From the computational data, we may conclude the most likely pathways for nucleophile generation and follow up solvent decompositions under typical operating conditions of a Li-ion cell. The focus will be on reasonable routes to observed products, many of which have been proposed as the product of solvent oxidation. Solvent transesterfiications, and formation of lithium ethylene dicarbonate (LEDC), poly(ethylene oxide) oligomers and poly(ethylene carbonate-co-ethylene oxide) oligomers will be discussed.

Figure 2. Cycle of chemical / electrochemical solvent decompositions.

References:

1. S. E. Sloop, J. B. Kerr, and K. Kinoshita, J. Power Sources, 119-121, 330–337 (2003).

2. K. Xu et al., J. Phys. Chem. B, 110, 7708–7719 (2006).

3. D. M. Seo et al., ECS Electrochem. Lett., 3, A91–A93 (2014).

4. G. Gachot et al., J. Power Sources, 178, 409–421 (2008).

5. J.-C. Lee and M. H. Litt, Macromolecules, 33, 1618–1627 (2000).

6. J. Xu, R. D. Deshpande, J. Pan, Y.-T. Cheng, and V. S. Battaglia, J. Electrochem. Soc., 162, A2026–A2035 (2015).