The Analytical Column

Water gets into everything. Into fuel tanks, into lubricating oils, into medicines and into the solvents labs use every day. A little is often harmless; too much can corrode an engine, spoil a drug, or ruin a chemical reaction. So labs need to know exactly how much is there, sometimes down to around ten parts per million. The method most of them use is called Karl Fischer titration, after the chemist who invented it in 1935.

The counting rule

Karl Fischer titration works because of one useful reaction: in the right mix of chemicals, iodine gets used up when water is there to react with it (other substances can use some up too, which is why some samples need special care). In the usual reagents, one molecule of iodine is used up for every molecule of water. So if you can count the iodine, you have counted the water.

Two ways to deliver the iodine

There are two versions. In the first, the iodine is added from a burette, like any other titration. It suits samples with a lot of water, from about a tenth of a percent up to pure water, and the liquid's strength has to be checked regularly because it changes as it picks up moisture.

In the second, the iodine is made inside the cell by an electric current, and the instrument simply counts the electricity it took. Because the charge needed for each molecule is fixed by nature, there is no strength to check in the liquid itself, though labs still test the whole cell with a known water standard. This version is built for traces of water, down to around ten parts per million.

Water you did not put in

Here is the catch. A Karl Fischer cell is never perfectly dry. Water creeps in through seals and from the air, and the instrument sees it as a slow, steady use of iodine even with no sample in it. That background is called drift. A good analyst waits for it to settle low and steady before measuring, and the instrument subtracts it afterwards.

That subtraction rests on one assumption: that the drift during the measurement is the same as it was before. When the drift is high and the sample holds very little water, the correction can be bigger than the water itself, and the result comes out negative. A negative result is not a measurement. It is a sign to check the cell, the drift and the sample weight before trying again.

When the sample fights back

Some samples react with the reagents on their own. Ketones and aldehydes, common in solvents and fragrances, react with the methanol in ordinary reagents and make extra water as they do, so the endpoint keeps fading and the result comes out high. Special reagents made without methanol solve that. For samples that will not dissolve or that misbehave, there is another trick: heat the sample in a sealed vial and sweep its water into the cell with dry gas, so the sample itself never touches the reagent.

Why this matters to you

The next time a fuel, an oil or a medicine passes a "water content" test, this is very likely how it was measured. It is a method built on a simple count, made trustworthy by the analyst who checks the background, watches for samples that cheat, and treats a strange result as a question rather than an answer.

References

1. "Karl Fischer titration", Wikipedia (the reaction and its history).
2. Metrohm, "Karl Fischer titration: When to use volumetry or coulometry", 2023; and "Frequently asked questions in Karl Fischer titration - Part 2", 2020.
3. "Water determination in aldehydes and ketones", Honeywell application note, The Analytical Scientist, 2017.