Remote and infield analysis is developing into a crucial aspect of modern analytical chemistry. Chemical sensors, which are small measuring devices consisting of a recognition layer, a transducer, and the measurement electronics, have proven to be highly suitable tools for this purpose. Although usually they are constructed for the detection of very limited amounts of chemical compounds and often do not reach the detection limits of modern analytical instruments, they show very desirable advantages: the systems are easy to operate, mostly rugged, small, of good value, and can be manufactured by established technological methods. Owing to their size and ruggedness, they can even be operated in harsh environments and often can be constructed to be suitable for remote measurements, where several sensors are interrogated by a central analytical site.
Devices being used include optical fibers, field‐effect transistors (FETs), electrochemical cells (be they potentiometric or amperometric), and mass‐sensitive components of all kinds. The chemically sensitive coatings range from bare transducer surfaces to dyes reacting with analyte or sterically specific binding sites (e.g. host–guest complexes). A seeming drawback is the fact that only few sensors react specifically toward one defined component but normally show a sensitivity pattern toward chemically related substances. By using an array of several components followed by modern methods of data evaluation (such as neural networks), a group of analytes can be detected and characterized simultaneously. This article introduces device and recognition layer principles as well as covering field sensors published during the last one and a half decades based on all the different kinds of transducers and sensor layers. Some of the devices (such as the lambda probe) are already marketed, others are developed to a stage where field measurements have already been carried out, but which are not market‐ready.