Drying Solvents

Electrochemistry in non-aqueous media (acetonitrile, THF, DME, propylene carbonate) depends on solvent dryness. This page covers lab techniques for removing water. It frames these methods for electrolyte and solvent prep instead of general synthesis.

Why residual water matters here

• **Potential window.** Aprotic solvents have a wide electrochemical window. Water does not. Water electrolyzes at a modest overpotential. Dissolved water narrows the aprotic window. You get parasitic O2 and H2 evolution. It eats into the window you need.
• **Side reactions.** Water reacts with electrogenerated intermediates. These include radical anions, radical cations, and strong reductants or oxidants at the electrode. Water can hydrolyze moisture-sensitive supporting-electrolyte salts. LiPF6 salts are the classic case. Hydrolysis produces HF. HF then attacks everything else in the cell.
• **Reference electrode stability.** Aqueous reference electrodes (SCE, Ag/AgCl) contain water. They leak water into a non-aqueous cell. Leakage contaminates the bulk electrolyte. It drifts the junction potential. See Reference Electrodes for workarounds (double junctions, non-aqueous pseudo-references).

Drying methods and agents

Use a combination of methods from quick to thorough:

• **Anhydrous salt desiccants.** Add anhydrous MgSO4, CaCl2, Na2SO4, or CaSO4 to the solvent. Let it hydrate. Filter or decant. These are cheap and fast. They are fine for a first pass. They are not sufficient for electrochemistry-grade solvent.
• **Sodium/benzophenone ketyl still.** Use this for ethers (THF, 1,4-dioxane, DME) and hydrocarbons. Distill solvent over sodium dispersion with a trace of benzophenone. Sodium reduces benzophenone to the ketyl radical. The ketyl radical is deep blue/purple. The color persists only when the system is dry and oxygen-free. The still acts as its own indicator. Safety note: only feed the still solvent that is already reasonably dry. Sodium metal in wet solvent is a fire/pressure hazard, not just an inefficiency.
• **Molecular sieves.** Sieves adsorb water into pores. The pores admit H2O but exclude most solvent molecules. This method is effective but slow. It takes 24+ hours to equilibrate. It is not compatible with ketones (acetone, MEK, etc.). Ketones undergo base-catalyzed aldol condensation on the sieve surface. Use sieves as a final polishing step. They also make good storage desiccants to keep dried solvent dry between uses.

Molecular sieves: 3A vs 4A, and activating them

Pore size determines selectivity. 3A sieves have ~3 Ã… pores. 4A sieves have ~4 Ã… pores. Water has kinetic diameter ~2.8 Ã…. It fits through either sieve. The question is what else enters with it.

• **3A** excludes most small polar solvent molecules (methanol, ethanol) along with water. It is the safer default when you do not want the sieve to adsorb your solvent.
• **4A** has more room. It also takes up ethanol and similar small molecules. Use 4A for solvents where this does not matter. Avoid it for solvents you do not want the sieve to absorb.

Sieves ship labeled "dry" but rarely are. They hold a large amount of adsorbed water out of the box. Activating them requires heating under vacuum (a few mbar) to high temperature (up to ~320 °C). This drives water off. Do not exceed this temperature. The alumina/silica lattice can crack. Heating in dry hot gas works too. It is less complete.

Quantifying dryness: Karl Fischer targets

Visual inspection cannot confirm dryness for clean voltammograms. Karl Fischer (KF) titration quantifies water content. It is an iodometric titration. Iodine oxidizes SO2 stoichiometrically with water. The titer is proportional to water. Run KF in dried methanol as the working medium. Pre-titrate to remove residual water from the medium. Otherwise you titrate solvent along with sample.

Targets depend on application:

• **Battery-grade carbonate electrolytes** (LiPF6 in EC/DMC mixes) require < 20 ppm water. Above this limit, HF from salt hydrolysis causes degradation. This is an industry-standard target (see Metrohm's application notes).
• **General non-aqueous electrochemistry** (CV, EIS on aprotic electrolytes) does not always need single-digit ppm. The tens-of-ppm range is a reasonable target before trusting potential window measurements near solvent limits.
• A well-run sodium/benzophenone still followed by sieves reaches single-digit ppm for ethers. This is as dry as bench chemistry gets.

Attribution and sourcing note

This page paraphrases the Sciencemadness wiki's Drying solvents page. It covers drying agents, the sodium/benzophenone ketyl still, molecular sieve behavior, regeneration, and KF titration. The content is for an electrochemistry-prep context. No content is quoted verbatim. The live wiki was down for maintenance when this page was written. Content was cross-checked via cache excerpts. The 20 ppm battery-electrolyte figure is not from Sciencemadness. See Metrohm's water-in-battery-materials application note.

See Standard Operating Procedure (cell assembly, inert-gas sparging) for keeping dried solvent dry during experiments. See Conical Joints (ground-glass hardware). See Fumehood (volatile solvent handling).