Carbonate solvents for battery electrolyte: what matters
Technical article · Eapearl Chemical ·
The phrase that circulates about carbonate solvents in cell manufacturing is picturesque but misleading in one respect: no single carbonate is the fluid in question. What goes into a cell is a blend, and almost everything a buyer needs to know follows from why the blend has to exist and what has to be kept out of it.
Identity, and an abbreviation that collides
Dimethyl carbonate is the simplest dialkyl carbonate, molecular formula C3H6O3, molar mass 90.08 g/mol, CAS 616-38-6 and EC 210-478-4. It is a clear, mobile, flammable liquid with a mild odour, and it is sold both as an ordinary industrial solvent and as a purified grade for electrochemical use. Those two things travel under the same chemical name and are not the same product.
One notational trap is worth clearing immediately, because it causes real confusion in specification documents. In cell engineering, EC normally denotes ethylene carbonate. In regulatory documents, EC denotes the European Community inventory number. A line reading EC 210-478-4 and a line reading EC content are talking about entirely unrelated things. Spell the substance out in any document that will be read by someone outside your own team, and keep registry identifiers next to the substance name rather than in a column of their own.
Why an electrolyte is a blend
An electrolyte has to do two jobs that sit in tension. It has to dissolve a lithium salt and hold the ions apart well enough for them to move independently, which calls for a strongly polar medium. It also has to remain thin and mobile so that those ions can travel through a porous separator without excessive resistance, which argues for a small, weakly associated molecule.
The carbonate family happens to contain members at both ends of that trade-off. The cyclic members, ethylene carbonate and its propylene homologue, are highly polar and excellent at salt dissociation, but they are the thicker and less volatile members of the set, and one of them is a solid at ordinary warehouse conditions. The linear members, of which dimethyl carbonate is the smallest, are thin and volatile but weaker solvators.
Blending recovers both properties. It is also generally accepted that the cyclic component participates in forming the passivating film on the negative electrode during the first cycles, though the detail of that chemistry is specific to the electrode materials and the additive package, and it is not something a solvent article can settle. The important commercial point is narrower and safer: the ratio between components is the cell developer’s design variable, and a supplier’s job is to deliver each component to a written specification, not to recommend a formulation.
The linear carbonates are not independent of one another
The three linear carbonates in common use are the dimethyl ester, the mixed ethyl methyl carbonate, CAS 623-53-0, and the symmetrical diethyl carbonate, CAS 105-58-8. They are related by transesterification: the mixed ester can be made from the two symmetrical ones, and the equilibrium runs in both directions.
Two consequences follow for anyone writing a specification. First, a lot of the mixed ester will normally contain some of both symmetrical esters, and a specification that demands their absence is asking for something the chemistry does not readily give. Second, the distribution is not frozen once the drum is filled. Acidic or basic contamination, and a long residence time in a warm store, can move it. If the ratio matters to you, it belongs on a retained-sample test plan rather than on a single certificate read once at receipt.
What a battery grade controls
The gap between a solvent grade and an electrochemical grade is not one figure on an assay line. It is a panel of impurities that are present in trace quantity, each with a defined analytical method, plus controls on how the material is contained and delivered. The panel normally addresses water; protic species such as residual methanol or ethanol carried over from manufacture; acidity, usually expressed as an acid number or as free acid; chloride and other halide; metals, determined by an elemental technique; and non-volatile residue, which is what stays behind when the solvent has gone.
The limits themselves belong in the agreed specification between the two parties, and there is a good reason not to circulate them in general text: they differ between applications, they are tightened as cell designs evolve, and quoting a number out of context invites someone to build a purchase decision on it. What can be said in general is how to read such a panel. Ask which method produced each figure, because different techniques answer subtly different questions. Ask whether the figure is a release limit or a typical value, because a typical value is not a commitment. Ask when the sample was drawn, because for a moisture-sensitive product the interval between filling and analysis is part of the result.
Where contamination actually enters
A purified lot rarely fails because the purification failed. It fails because something was added to it afterwards, and the paths are mundane. Headspace in a part-used container exchanges with humid air every time it is opened. A transfer hose that was cleaned with an aqueous detergent and not fully dried contributes water on first use. Sampling in an open plant undoes in a minute what a drying train achieved. A container previously in another service leaves residue that no visual inspection will show. Carbon steel in a line handling a lot that has begun to develop acidity contributes iron.
The practical answer is to treat the container, the transfer route and the sampling method as part of the specification. Dedicate lines and pumps. Use dry inert gas for blanketing and for displacing headspace. Sample through a closed route into a prepared container rather than by opening a lid. Keep retained samples in the same kind of packaging as the bulk, because a retain stored badly will mislead you later at exactly the moment you need it to be reliable.
Handling, fire and transport
The linear carbonates are flammable liquids and should be handled as such: bonding and earthing on transfer, no ignition sources, ventilation designed for a heavier-than-air vapour, and a fire response plan that assumes an alcohol-resistant foam may be needed. Classification and transport entries differ between the individual substances and again for blends, so the shipping description has to be derived for the material actually being moved rather than copied from a neighbouring product. Where a formal classification is quoted in your own documents, it should be taken from the current safety data sheet for that specific grade and not from a general article.
Qualifying a lot, and the paperwork that goes with it
A workable qualification sequence for an electrochemical-grade solvent looks like this. Agree the specification, including methods, before any sample is sent, so that the sample can be judged rather than admired. Obtain a sample in the intended production packaging, not in a laboratory bottle, since packaging is one of the variables under test. Run your own incoming panel on that sample and compare it against the supplier certificate, method by method. Agree in writing what constitutes a change requiring notification: a different production site, a different purification train, a different container supplier, or a different analytical method all qualify. Then keep retained samples and trend them, because a slow drift is invisible in single certificates and obvious in a series.
Delivery times, container availability and documentation for a specific grade are commercial matters that depend on the lot and the destination, and they are best settled directly rather than assumed. Specifications, method statements and samples for a named application can be requested through our contact page.