We have discussed several approaches to determining water content including gravimetrÂc methods, electrical impedance techniques, and spectroscopic methods. In addition, for specificity and sensitivity, a chemical method, the Karl Fischer titration (KF) is the method of choice for many applications.
KF is a chemical reaction of water with iodine in the presence of base (e.g., pyridine), solvent (typically methanol), sulfur dioxide, and buffering. KF has the appeal of a molecularly quantitative chemical reaction; if the number of molecules of iodine used in the reaction is known, the number of molecules of water can be determined.
KF is used to determine water content in an immense variety of materials such as solvents, oils, gasses, natural products, and lyophilized material. Typical uses in critical areas of contamination control include lyophilized materials and assorted solid reagents for pharmaceutical applications and organic solvents used in aerospace and wafer fabrication. KF is appealing in that it is specific for water and is a direct (extractive) measurement. Gravimetric methods, on the other hand, while direct, are non-specific. That is, weight loss after drying the sample could be due to outgassing of other solvents or of plasticizers, not only to water content.
KF determinations may be coulombic or volumetric. Coulombic measurements are typically used for microgram to milligram levels; volumetric measurements, for higher levels of water content, up to 100%. Coulombs are the product of current (amps) and titration time in seconds; these can be quantitatively related to the number of molecules of iodine. Volumetric determinations measure, as one might expect, the volume of iodine added. In typical analytical systems, the volumetric endpoint (i.e., the absence of unreacted water and presence of free iodine) is determined not by an operator peering intently at a glass burette but rather by a decrease in voltage needed to maintain a pre-specified current.
Automatic KF titrators, for which there are several commercial sources, improve convenience and consistency. KF determinations, however, are by no means trivial. Samples must be collected and extracted appropriately; analysis requires a skilled, thoughtful, and often innovative chemist. To avoid interferences, the analyst must be aware of the sample composition. The minutiae of ramifications of collection and extraction techniques, standardization, titration methods, temperature, agitation, and reagent provides for ongoing spirited discussion and a host of technical papers; all are of great fascination to analytical chemists.
For the rest of us, a few illustrative highlights will suffice. Water itself, when inadvertently introduced, interferes with accuracy. Humidity must be controlled during testing, and spurious, atypical humidity must be controlled during sample collection and handling. Complete extraction of water can be a challenge. If the water is bound (as in some biological samples), it may not be available to react with the KF reagent. One approach, sometimes used with lyophilized samples, is to subject the sample to vigorous agitation or to ultrasonic action prior to analysis.
Side reactions impact accuracy. For example, ketones (such as acetone) and aldehydes can react with methanol to form acetals or ketals, with the release of water. Falsely elevated water levels can also result from the presence of a host of readily oxidized chemicals; copper and tin salts, for example, can react with iodine. In addition, the KF reaction must be held within a limited pH range, to avoid factors which overpower the buffering system and may skew the results.
Because KF requires an understanding of sample composition and potential interferences, partnering with your analytical chemist is imperative. It is crucial to communicate pertinent factors such as the type of sample (solid, liquid, gas), approximate level of water expected, and materials of construction or chemical components of the mixture. If there are unknowns, initial testing can sometimes reveal biased determinations (high or low); steps can then be taken to correct the problem. For example, if there are excess oxidizing agents, water can be evaporated from the solvent, then collected for analysis in an inert stream of gas. For ongoing analysis protocols, it is important to explain changes in the product or process variables, because an apparent change in water could actually signal some other product or process modification.
The authors appreciate the comments of Eric Andersen, Manager, Materials and Processes Department of Northrop-Grumman, Navigation Systems Division.In the next column, we will discuss details of the primary Karl-Fisher technique and two additional secondary on-line techniques, near infrared reflectance and millimeter wave.
In the next column, we will discuss more of the common secondary methods for moisture measurement.
Next month: A discussion of the most common techniques for moisture measurement.