Poor result repeatability achieved in water determination with volumetric Karl Fischer titration.

The result repeatability achieved in water determination according to volumetric Karl Fischer titration is poor. What affects the result repeatability and what measures can be recommended?

Possible reasons:

Air bubbles capture in burette and tubes they may lead to inaccurate dosing thus affecting accuracy. Due to inhomogeneous water distribution in the sample specimen, the sample amount used is either too small or not representative and this leads to inconsistent results. The Platinum pins of the indicator sensor may be dirty which influences the endpoint potential, therefore the termination criteria are inconsistent from one sample to the other. The titration stand is not properly protected against moisture from the ambient. The solvent in the Karl Fischer titration cell has been used too long or for too many samples, i.e. the capacity to titrate water or completely dissolve the sample is inadequate.

Recommendations:

Rinse the burette and tubes until all air bubbles are removed. Verify this frequently during the course of the day especially if you are using the one-component KF reagent. This reagent solution contains dissolved sulfur dioxide that may evaporate caused by temperature variations. This leads to gas bubble formation in the tubes. Increase the sample amount in order to ensure that approximately 10 mg of water is present per sample. Samples with low water content are highly hygroscopic, therefore the samples must be taken quickly with a syringe or spatula that is completely dried. In order to achieve a representative sample the sample must be homogenized prior to titration Clean the platinum pins of the sensor with paper tissue or if necessary generate the sensor by cleaning the sensor by soaking it in 50% HNO₃ for 10 minutes, rinse thoroughly with deionized water to remove all traces of acid and then rinse with vessel solution or dry methanol. Check whether the titration stand is completely tight; replace the molecular sieves and silica gel in drying tube on the solvent and titrant bottle. Change the solvent after 4-5 samples have been tested, because if the dissolving capacity of the solvent is exhausted, the water will no longer be completely released from the sample and that would lead to incorrect results.