Commentary: Flow-injection analysis—an idea incomplete?

The ultimate aim of any instrumentation must be 'sample in-result out in 1-2 min.'. This specification has been the reason for the phenomenal success of flame analysis, it is the reason for the now remarkable success of the near infra-red analysis for protein, fat etc. and for the continuing research after more than 60 years into all methods of electrode measurements. Yet Flow-Injection Analysis (FIA) which has often been publicized in the same terms, has still to make a significant impact in the analytical laboratory if by 'impact' we mean systems sold in comparison with the number of Technicon AutoAnalyzers or similar continuous flow systems. A conservative estimate suggests that more than 50 000 continuous flow units are now in the field after about 25 years. A fully automated FIA-type system was, after all, commercially available in 1959 and described in great detail by Jonnard [1] for the analysis of chromate, protein, urinary glucose, red blood cells and haemoglobin. The fundamental theory of FIA, that is the dispersion of a sample slug in a flowing stream, was first developed by Taylor [2] in 1953 and subsequently by Aris [3] in 1956. The concept was reintroduced into electro-chemistry by Pungor's school in 1970 and, since 1976, widely expanded by Ruzicka and his colleagues in the academic press. Today at least four firms market FIA systems actively in Europe and the USA. Hence the very limited acceptance of FIA by practising analysts cannot be blamed on a lack of history, academic support or commercial exploiters. The goal ofFIA is certainly achievable-theoretically. Given completely turbulent flow (Reynold's number greater than 1000), a sample injected into such a stream would be recorded as a complete square wave on any detector. Samples can be injected at hundreds/h, each peak could be monitored for quality of the assay and so duplicates would be unnecessary. Micro samples could be used with high precision and if the reactions were relatively simple, increased reaction rates would often be possible using high temperatures with the high pressures necessary. Separations using packed columns would open new possibilities without significant loss of the square wave. The price? A high/very-high pressure system of narrow bore tubes (0.2 ram?) with velocities in the order of 500cm/s and possible (but not certainly) high reagent consumption. To my knowledge, the obvious technological problems have never been solved on a commercial instrument. In contrast, .FIA uses conventional …