Water determination in samples with high sugar and protein content

The aim of water determination should be detection of water and nothing but the water. Large number of methods use heating where the result shows loss of all volatile compounds and not only water. The result of these techniques is not water content, but the mass loss. One of the best techniques for determination of water content is Karl Fischer titration, based on a chemical reaction selective for water. Determination of water content by heating, in samples that are rich in proteins and sugars is especially hard, because of the Maillard reaction. During the Maillard reaction, water is produced, and it is hard to determine water which is originally from sample and water that is produced by Maillard reaction. In this investigation we used samples of different types of condensed milk - rich in proteins and sugars. Samples were measured in ten probes, and by four methods: Karl Fischer titration with different solvents, Classical Oven, IR drying and Oven Sample Processor. Classical Karl Fischer titration was used as a reference method. The best method was Classical Karl Fischer titration, because of precision, trueness and duration of measurement. Usage of boiling methanol for extraction is not recommended. Due to a small amount of sample, contribution in the amount of water from the Maillard reaction is not significant. The best method for measurement is KF titration