flow injection analysis
/ floh in-JEK-shun uh-NAL-uh-sis /
Picture a stream of water flowing steadily through a narrow tube. Every so often you squirt a small slug of sample into the moving stream; as it travels, it mixes with reagents, develops a colour, and sweeps past a detector that takes its reading — then the next sample follows behind. This conveyor-belt of liquid is flow injection analysis.
Technically, flow injection analysis injects a precise, small volume of sample into a continuously flowing, unsegmented carrier stream. On the way to the detector the sample plug merges with reagents and reacts in a reproducible, only partly mixed way; because the timing and dispersion are identical for every injection, the method does not need the reaction to reach completion — it just needs to be consistent.
Its great virtue is throughput and economy: many samples per hour, tiny reagent use, and easy automation, all from simple tubing and pumps. The trade-off is that it is built for repetitive, well-defined measurements of one or a few species; it is not a general-purpose separation tool, and matrix differences between samples must be controlled because the reaction is deliberately stopped short of equilibrium.
A water utility uses flow injection analysis to test hundreds of samples a day for nitrate: each sample plug merges with a colour-forming reagent in the tubing and reaches a spectrophotometer just as the colour peaks, giving one nitrate result every few seconds.
Reproducible timing, not complete reaction, is what makes FIA work.
Unlike older segmented-flow systems, flow injection deliberately uses no air bubbles to divide samples; it relies on precise timing alone, which is why every standard and sample must be handled identically.