Karl Fischer titration (revision 22)

Karl Fischer titration is a titrimetric method for determining the water content of a sample, based on the stoichiometric oxidation of sulfur dioxide by iodine in the presence of water. It is the reference technique for moisture determination in lyophilised pharmaceutical solids and is specified in pharmacopoeial moisture chapters worldwide — USP <921> and Ph. Eur. 2.5.12.^[1]

The method matters to readers of a certificate of analysis for a specific and frequently misunderstood reason: a chromatographic purity figure is a statement about the relative composition of the material that dissolved, not about how much peptide the vial contains. Water, counterions and residual salts occupy mass. A vial of material reported at 99% purity by HPLC may nevertheless be 6% water and 8% trifluoroacetate by mass, in which case its peptide content is closer to 85%.^[2]

Two variants are in routine use. Volumetric titration delivers iodine from a burette and suits samples containing milligram quantities of water; coulometric titration generates iodine electrochemically at the anode and suits the microgram-to-low-milligram range typical of a 10 mg peptide sample.^[3]

The underlying reaction is the oxidation of sulfur dioxide by iodine, which consumes water stoichiometrically:

H₂O + I₂ + SO₂ + CH₃OH + 2\,RN → 2\,RNH⁺I^- + RNH⁺CH₃SO₄^-

Methanol participates directly, forming a methyl sulfite intermediate that is the species actually oxidised; the base (historically pyridine, now typically imidazole) buffers the liberated acid and keeps the reaction in the pH window where the stoichiometry holds at 1 mol iodine per mol water.^[3]

Two features follow from the stoichiometry and are the practical basis of the method:

• the endpoint is detectable amperometrically, because excess iodine depolarises a double-platinum indicator electrode;
• one faraday of charge liberates one mole of iodine, so in the coulometric variant the charge passed is a direct measure of water — no titrant standardisation is required.

Away from the buffered pH window the stoichiometry drifts, which is the reason strongly acidic or basic samples require a modified reagent rather than a correction factor.

For a research peptide vial nominally containing 5–10 mg of solid, only the coulometric variant is realistic: the total water present may be as little as 300 µg, which is below the practical resolution of a volumetric titration on that sample mass. Certificates reporting water content on 10 mg vials by volumetric titration should be read with that in mind.^[3]