Karl Fischer: volumetric or coulometric?
A practical rule: coulometric below about 1% water, volumetric above it. Here is why, and what goes wrong at the boundary.
Karl Fischer titration measures water specifically, which is why it outlives loss-on-drying for anything that contains a volatile other than water. Choosing between the two formats is mostly a question of how much water you expect.
The rule of thumb
- Coulometric - samples from about 10 ppm to 1% water. The iodine is generated electrochemically, so there is no titre to standardise and no burette to drift.
- Volumetric - samples from about 0.1% water upwards. You dose a reagent of known titre from a burette.
At the boundary, choose by sample size. If you can weigh in enough sample to deliver 100-1000 µg of water into a coulometric cell without overwhelming it, stay coulometric.
What actually goes wrong
Most bad Karl Fischer data comes from three places, none of them the reagent:
- Atmospheric ingress. The titration cell is a humidity trap. Check your drift before every run - a stable cell should sit below 10 µg/min.
- Side reactions. Ketones and aldehydes form acetals with methanol and release water, giving a slow creeping endpoint that never settles. Use a ketone-specific working medium.
- Incomplete extraction. A solid that does not dissolve only gives up its surface water. Consider an oven module and carrier gas rather than fighting the solvent.
Reagent handling
Karl Fischer reagents are hygroscopic by design. Keep the bottle closed, use a proper drying tube, and do not decant into an intermediate container. Titre drops measurably within days of opening if the closure is poor.
Our HydraTrace range covers both formats - Compozit 1 and Compozit 5 for volumetric work, Coulomet AG as a fritless coulometric anolyte. All three ship with a lot-specific water-equivalent value.