Results and discussion
Since its introduction in 1935 the Karl Fischer
reaction has been a subject matter of many researchers
and (despite still existing open questions and contro-versial discussions about the true reaction mechanism
[16]) the method is widely and very successfully em-ployed for the determination of water in a variety of materials. Convenient to use commercially available
one-component reagent solutions have become the
most preferred form of the KF-reagent today [3] and
have also been used in the present work. In conven-tional volumetric batch titration, the active iodine
(determining the titre of the KF solution) is adjusted
according to the desired working range and added
dropwise to the sample solution until the endpoint
is reached. In the adaptation of the KF method to
common FIA the procedure is significantly different,
in that a small amount of sample is either injected
directly into the KF-reagent (single-line manifolds
[7–9,11]) or injected into a carrier and merged with
the reagent solution (dual-line manifold [12,14]). In
both cases, the reaction takes place during travelling
to the detector where the decrease of the iodine con-tent or the alteration of the iodine to iodide ratio is
measured by different detection techniques. Hence,
optimisation of a FIA-method desires consideration
of the chemical conditions as well as the operational
parameters of the flow system.
In contrast to most of the other authors having dealt
with the application of the KF method to FIA, we se-lected a dual-line manifold configuration (see Fig. 1).
This has the advantage that the entire sample plug (ir-respective of the injection volume) is mixed with the
same amount of KF-reagent. Additionally, the actual
concentration of the KF-reagent coming in contact
with the sample can be simply altered by changing
the flow rate ratio of carrier to reagent stream. Both
spectrophotometric and zero-current potentiometry
have been applied which will be discussed separately
in subsequent sections.