Electrolyte Decomposition - Battery Design

When thinking about electrolyte decomposition we should consider this in stages, and the electrolyte type and additives in use shift how badly each stage bites:

decomposition of electrolyte in bulk storage
decomposition of electrolyte prior to cell formation
in use decomposition / aging
decomposition of the electrolyte during thermal runaway

Bulk Storage of Electrolyte

The optimum solution is to use electrolyte immediately and not store it long term. However, this is often not possible due to a number of reasons. An investigation of ‘shelf-life’ was conducted by Gorman et al [1] and is summarised as:

non-aqueous electrolytes are prone to moisture absorbtion
water as contaminant impacts performance and lifetime of the resultant cell
water at concentrations as low as approximately 10 ppm can promote undesirable side reactions
manufacturers typically state 6 month shelf life for electrolytes
cell manufacturers typically use electrolytes within a 6–10 weeks
electrolytes observed to be sensitive to temperature and moisture exposure
HF production and the associated acidification seen even in dry room storage of electrolytes

The HF seen even in dry-room storage comes from salt hydrolysis: LiPF6, the salt in almost every conventional carbonate electrolyte, reacts with trace water even at parts-per-million levels. The reaction produces HF together with fluorophosphate species such as HPO2F2, and the exact product mix shifts depending on which solvent the salt is dissolved in [2]. HF is corrosive on both electrodes, so this is not a cosmetic side reaction, it is the same chemistry that later attacks the cathode surface and etches the SEI once the cell is built.

Prior to Cell Formation

Once the cell is filled but before the first charge has built a protective SEI, freshly exposed electrode surfaces are in direct contact with electrolyte with no barrier layer between them. Any elevated temperature at this point, whether from the filling process, formation ovens or transport, also brings the salt itself into play: LiPF6 on its own is only stable up to around 107°C in a dry, inert atmosphere, where it dissociates into solid LiF and gaseous PF5 [3]. With even a few hundred ppm of water present, as is typical before formation, that onset temperature drops sharply because the water reacts directly with the salt to form POF3 and HF instead.

Aging of Electrolyte during Cell Use

Electrolyte decomposition refers to the breakdown of electrolyte components (solvents and salts) in a battery, leading to the formation of solid electrolyte interphase (SEI) layers and irreversible capacity loss. Stockhausen et al [4] traced which of these mechanisms actually dominate capacity loss at different stages of electrochemical aging, and found that the balance shifts as a cell gets older rather than staying fixed. Two mechanisms run throughout normal use, one at each electrode.

At the cathode, the carbonate solvents oxidise at high potential. Using two-compartment cells that let the decomposition products from each electrode be identified separately, operando pressure and NMR measurements have traced this oxidative decomposition, showing that it generates CO2 and other gaseous species, with the rate climbing sharply as voltage increases [5]. This is one of the reasons high-voltage and high-nickel cathode chemistries push formulators toward more oxidation-resistant solvent blends: transition metal dissolved from the cathode can also catalyse further electrolyte breakdown, compounding the problem.

At the anode, the same family of solvents decomposes in the opposite direction. On first charge, electrons transferred from the polarised graphite (or silicon) surface reduce solvent molecules such as ethylene carbonate, and the resulting products build the SEI [6]. This first pass is necessary rather than a fault, but the SEI is not a one-time, static layer: it cracks and re-forms as the electrode expands and contracts on cycling, and the gas released each time it does, ethylene and CO among them, is a significant contributor to cell swelling over life.

Thermal Runaway

Under thermal runaway, every pathway above accelerates at once rather than staying an independent, slow background process. The salt decomposition threshold discussed under Prior to Cell Formation is passed within seconds, releasing PF5, which is a strong Lewis acid and goes on to open the ring of cyclic carbonate solvents such as EC [3]. That reaction, on top of the accelerating oxidative and reductive decomposition already running at the electrodes, is what generates the bulk of the flammable, corrosive gas vented during a runaway event.

Electrolyte Type and Additives

Because these decomposition modes trace back to specific chemistry, electrolyte type and additive choice can suppress some of them almost entirely while leaving others untouched:

SEI-forming additives such as FEC and VC (see Electrolyte Additives) push the anode-side SEI toward a thinner, more flexible, LiF-rich film that resists cracking on cycling, which slows ongoing reductive decomposition and the gas generation that comes with it.
Thermally stabilising additives, such as LiDFOB-type additives, improve the thermal stability of the salt system, pushing back the temperature at which salt decomposition accelerates.
Moving away from LiPF6 altogether removes hydrolysis and LiPF6 thermal decomposition as failure modes specifically, since both are properties of that anion. Ionic liquid electrolytes and other high-concentration electrolyte designs take this route, though each introduces its own trade-offs in cost, conductivity or wettability rather than eliminating decomposition outright.