Comment: Comments on ‘Flow-injection analysis— an idea incomplete?’

it has required the perception of workers such as Rfi.i6ka and Hansen [6] and Stewart et al. I-7] for the realization that this apparently simple operation was capable ofa remarkable degree of exploitation. It was, in fact, the prolific work of Rfl2i6ka and Hansen [1 and 2] that led us to investigate the potential of FIA in clinical chemistry. Using simple systems constructed from items ofequipment commonly used in clinical analyses, we have been able to carry out analyses with economy, speed and accuracy and with negligible carry-over. For those familiar with continuous-flow analysis, the speed of FIA is startling: with fast reactions the analysis may be complete 10 after introduction of the sample. Holy suggests that there is a need in FIA for turbulent flow in which the detector may recognize the sample as a square wave. Apart from the difficulty of achieving turbulent flow, it is problematical whether or not flow injection would work effectively under such conditions. Mixing of the sample slug with the carrier stream of reagent depends on the fact that it moves under conditions of laminar flow. The slug adopts a hollow bullet shape and travels along a boundary layer which forms a constantly replenished source of reagent. With the narrow-bore tubing in use, radial diffusion of reagent into slug is extremely rapid. These conditions would not be obtained with turbulent flow. Holy's commentary [5] contains a number of other misleading statements regarding the practice of FIA. Contrary to his assertions, we find from our own experience and by recourse to the literature that FIA calibration curves are usually linear, that triplicate (or duplicate) measurements are unnecessary and that separation steps (when necessary) can be carried out at least as fast as when using segmented-flow analysis (SFA). If a single-channel FIA system is compared with a comparable SFA system for carrying out an assay in which the sample reacts rapidly with reagent, it will be evident that: () (b) (c) The FIA system will be ready for use almost immediately , whilst the SFA system requires several minutes of operation before its base-line is established. The FIA peaks will be available first-usually within 30 s of sample injection, as against some minutes for SFA. The FIA system will consume less reagent and normally less sample (even after allowing for the small volume which is needed to wash the valve through). The technique known as …