Flow Injection or Segmented Flow: Which Continuous Flow Analyser Does Your Lab Need?

Anyone specifying an automated wet chemistry analyser runs into the same three terms within about five minutes: continuous flow analysis, segmented flow analysis, and flow injection analysis. They are used interchangeably in brochures and precisely in method standards, which is exactly the wrong way round for a buyer. Getting the distinction straight before you write a specification saves you from an instrument that suits eight of your ten methods and fights you on the other two.

Continuous Flow Analysis Is the Umbrella, Not a Third Option

Continuous flow analysis describes any technique where sample and reagents are pumped through narrow tubing, mixed and reacted on the way, and carried to a flow-through detector. Nothing is decanted into a cuvette. The sample never stops moving. Both segmented flow and flow injection sit inside that description, so in the broad sense a segmented flow analyser and a flow injection analyzer are both continuous flow analysers.

Where it gets slippery is in method standards. ISO documents generally use the abbreviation CFA to mean segmented flow specifically, and name FIA separately alongside it. So ISO 14403 has one part written for flow injection and another for continuous flow, and in that context "continuous flow" means the segmented technique, not the umbrella. A supplier saying "continuous flow analyser" is usually using the broad sense. A method schedule saying "CFA" is usually using the narrow one.

The practical habit worth forming: whenever a specification, a quote or a laboratory scope says continuous flow, confirm in writing which of the two techniques is meant. That one question heads off a surprising amount of trouble at accreditation time.

Segmented Flow Analysis: Air Bubbles and Steady State

Segmented flow is the older of the two, tracing back to Leonard Skeggs and the Technicon AutoAnalyzer in the late 1950s. The defining feature is simple: air bubbles are injected into the flowing stream at regular intervals, chopping it into a train of discrete liquid packets separated by gas.

Those bubbles do two jobs. They scrub the tubing wall as they pass, which limits the film of previous sample left behind and keeps carryover under control. And they act as barriers to axial dispersion, so a sample packet holds its identity no matter how much tubing you send it through. That second property is the important one, because it means the reaction coil can be as long as the chemistry needs, and it can be heated, without the sample smearing into its neighbours.

Give a slow colour-forming reaction enough residence time and it reaches completion, or close to it. The detector then sees a flat-topped plateau rather than a transient spike, and you measure the steady-state signal. That is why segmented flow suits chemistries that take their time: in-line dialysis, heated digestion, distillation steps, and reduction chemistries where the reagent needs to work.

The trade is in operating overhead. Bubbles have to be removed or electronically gated out before the signal is recorded. There is more tubing in the manifold to maintain. Startup takes longer because the whole flow path has to reach a stable baseline before the first result means anything, and the same applies at shutdown. Reagent consumption per determination is higher, because reagent flows continuously whether or not a sample is in front of the detector.

Flow Injection Analysis: Controlled Dispersion Instead of Bubbles

Flow injection, introduced by Ruzicka and Hansen in the mid-1970s, takes the opposite approach. There are no bubbles. A precisely metered plug of sample is injected through a valve into an unsegmented carrier stream that flows continuously past the detector.

The sample plug does disperse as it travels. Flow injection does not try to prevent that. It relies on the dispersion being reproducible, which it is, provided flow rate, tubing geometry and injection volume are held constant. Because dispersion is predictable, the signal shape is predictable, and the measurement can be taken at a fixed point in time rather than waiting for the reaction to finish. The detector sees a transient peak and peak height is the measurement.

Measuring before completion is the source of every advantage flow injection has. Cycle times drop because you are not waiting for equilibrium. Reagent consumption per determination drops with the shorter manifold and shorter run. Sample volume requirements drop, which matters when the sample is a precious extract or a small-volume field collection. The manifold is simpler, the footprint smaller, and the instrument reaches a usable baseline quickly, so stop-start operation is far less painful than it is on a segmented system.

The limitation follows from the same mechanism. If the chemistry is slow, measuring early costs you sensitivity. You can compensate with a heated reactor or a longer coil, but a longer coil in an unsegmented stream increases dispersion, which broadens and flattens the peak. At some point the compensations erode the speed advantage that made flow injection attractive in the first place.

How They Compare: A Side-by-Side Look

Read that table as a set of tendencies, not laws. A well-designed flow injection method on a fast chemistry will out-perform a poorly optimised segmented method on the same analyte, and the reverse is equally true. The tendencies matter most when you are choosing what to build a whole panel around.

Where the Line Blurs: Segmented Flow Injection

The two techniques are not sealed off from each other. Segmented flow injection analysis, usually shortened to SFIA, combines an injected sample plug with air segmentation of the reaction stream. You get the sample economy and valve-controlled precision of flow injection together with the residence time and dispersion control of segmented flow.

This hybrid exists because some chemistries genuinely need both. Total cyanide by in-line ultraviolet digestion is the clearest example. The digestion step needs the sample held in the light path long enough for the UV to break down the metal cyanide complexes, which a plain flow injection manifold does not readily provide. ASTM D7511 is written on segmented flow injection with in-line UV digestion and amperometric detection for exactly that reason.

The consequence for anyone specifying an instrument is worth stating plainly. On a modern platform, the technique is a per-method configuration, not a lifetime commitment made at purchase. The FS3700 supports SFA, FIA, iSFA and SFIA and can run different techniques on different channels simultaneously. A laboratory with a mixed panel therefore does not have to force every method onto one technique to keep the instrument count down.

Method Standards: Which Technique Each One Names

If your laboratory reports against accredited methods, the standard usually settles the technique for you. This table summarises the common ones for water, wastewater and soil work.

Standards are revised and part numbering changes, so check the current edition before you write a technique into a method schedule or a laboratory contract. If your results have to be NATA accredited, verify the specific standard number and part on the laboratory's scope rather than relying on the analyte name alone.

For the cyanide side of that table in more detail, including which method suits an ICMC audit, see our comparison of ASTM D6888 and OIA-1677 and the background on free, WAD and total cyanide.

ISO 14255 and Soil Nitrogen: A Segmented Flow Workhorse

Agricultural and soil laboratories are one of the biggest continuous flow user groups in Australia, and ISO 14255 is the standard most of that work sits on. It covers extraction of air-dry soil with calcium chloride solution and determination of nitrate nitrogen, ammonium nitrogen and total soluble nitrogen in the resulting extract.

Segmented flow suits this work for three reasons that compound. The cadmium reduction chemistry for nitrate and the indophenol chemistry for ammonium both develop over time rather than instantly, so residence time buys sensitivity. The extract carries a substantial dissolved salt load from the calcium chloride, and bubble scrubbing keeps carryover under control across a long run of similar-looking samples. And soil programs are seasonal, arriving as hundreds of extracts in a compressed window, so a long unattended run with a large autosampler is worth more than a fast cycle time on a short batch.

If you are sizing an instrument for that pattern, the two numbers that decide it are throughput per channel and reagent cost per determination across a full season, not the headline samples-per-hour figure. A platform running two channels at 20 to 90 analyses per hour each, method depending, covers most Australian soil laboratories comfortably.

Western Australian Conditions: Salinity, Heat and Sample Loads

Process water in the Eastern Goldfields is hypersaline, often several times seawater strength, and that changes how both techniques behave. The effects differ, which is worth knowing before you commit.

In segmented flow, a heavy dissolved solids load together with the surfactant added to control bubble behaviour can change how the sample wets the tubing wall. Bubble formation becomes less regular, and irregular segmentation shows up as baseline noise and drift over a long run.

In flow injection, the issue is optical rather than mechanical. A dense hypersaline plug injected into a low-ionic-strength carrier creates a refractive index gradient at the boundary between the two, and that gradient bends light in the flow cell. The result, known as the schlieren effect, is a signal artefact that has nothing to do with the analyte concentration. It is well documented and manageable, but only if you know to look for it.

The fix for both is the same, and it is unglamorous. Matrix-match the carrier stream to the samples, prepare calibration standards in a matrix that resembles the site water rather than deionised water, and validate on real samples before the method goes into production. Ambient temperature deserves the same attention. A Goldfields laboratory in February can swing enough to move a reaction coil off its optimum, so a thermostatted reactor earns its cost quickly. Our article on cyanide management in Western Australian gold mining covers the wider set of local challenges.

Decision Framework: Which Technique for Your Situation

The Practical Bottom Line

The choice between flow injection and segmented flow is decided by your chemistry and your sample pattern, in that order. Write out your full method list, mark each one as fast or slow chemistry, note any that need in-line digestion or distillation, and add the expected annual sample count against each. The answer usually declares itself once that list is on paper, because most laboratories find their panel clusters heavily towards one technique with two or three outliers.

Those outliers are what modern platforms are for. Rather than buying a second instrument or forcing an unsuitable technique onto a difficult method, configure the outlier channels differently on the same chassis. That is a materially cheaper answer than it was fifteen years ago, and it is the single biggest change in how these instruments should be specified.

Where the decision genuinely locks you in is accreditation. Method standards name techniques, laboratory scopes name standards, and site licences name laboratory results. Confirm the standard number and part before you commit, verify it against your contracted laboratory's scope, and validate on your own matrix rather than on synthetic standards. That groundwork holds up every time an auditor asks how the number was produced.

This article provides general guidance on technique selection. Confirm the current edition of any standard, and check your site licence conditions and laboratory accreditation scope, before finalising an analytical program.

Frequently Asked Questions

Is a continuous flow analyser the same thing as a segmented flow analyser?
Not quite, and the loose usage causes real confusion. Continuous flow analysis is the umbrella term for any technique that pumps sample and reagents through tubing to a flow-through detector, which covers both segmented flow and flow injection. In ISO method standards, however, the abbreviation CFA is generally used to mean segmented flow specifically, with FIA named separately. When a supplier or a method schedule says continuous flow, confirm which of the two is meant.

Which gives better detection limits, flow injection or segmented flow?
It depends on the reaction kinetics rather than the technique itself. Segmented flow lets the reaction run to or near completion in a long residence coil, so for slow-developing colour chemistries it usually reaches a lower detection limit. For fast chemistries that develop colour within seconds, flow injection is generally comparable because the reaction is close to complete at the point of measurement anyway. Matrix-matched validation on your own samples is the only reliable way to settle it.

Can one instrument run both flow injection and segmented flow methods?
Yes. Modern continuous flow platforms treat the technique as a per-channel configuration rather than a whole-instrument commitment. The FS3700 Chemistry Analyser, for example, supports SFA, FIA, iSFA and SFIA methods, and can run different techniques on different channels at the same time. That means a lab with a mixed panel does not have to compromise every method to suit one technique.

Does ISO 14255 require segmented flow analysis?
ISO 14255 covers the extraction of air-dry soil with calcium chloride solution and the determination of nitrate nitrogen, ammonium nitrogen and total soluble nitrogen in that extract. Continuous flow analysis is the routine determination route for the extract, and segmented flow suits it well because the cadmium reduction and indophenol chemistries both benefit from residence time. Check the current edition of the standard before writing a specific technique into a method schedule.

Is flow injection analysis suitable for cyanide?
Yes, and for available cyanide it is the specified route. ASTM D6888 determines available cyanide by ligand displacement and flow injection analysis. Total cyanide with in-line ultraviolet digestion is handled by the segmented flow injection hybrid under ASTM D7511, because the digestion step needs residence time that a plain flow injection manifold does not provide.

How does high salinity affect flow analysis results?
Both techniques are affected, in different ways. In segmented flow, high dissolved solids and surfactant chemistry can change bubble formation and how the sample wets the tubing wall. In flow injection, a dense hypersaline sample injected into a low-ionic-strength carrier can produce a refractive index artefact, known as the schlieren effect, at the detector. The practical answer for both is to matrix-match the carrier and the calibration standards, and to validate on real site water rather than synthetic standards.

Further Reading

• ISO 14403-1 and ISO 14403-2: Water quality, determination of total cyanide and free cyanide using flow analysis (FIA and CFA)
• ISO 11732: Water quality, determination of ammonium nitrogen using flow analysis (FIA and CFA) and spectrometric detection
• ISO 13395: Water quality, determination of nitrite nitrogen and nitrate nitrogen using flow analysis (FIA and CFA)
• ISO 15681-1 and ISO 15681-2: Water quality, determination of orthophosphate and total phosphorus content by flow analysis
• ISO 14255: Soil quality, determination of nitrate nitrogen, ammonium nitrogen and total soluble nitrogen in air-dry soils using calcium chloride solution as extractant
• ASTM D6888: Standard Test Method for Available Cyanide with Ligand Displacement and Flow Injection Analysis
• ASTM D7511: Standard Test Method for Total Cyanide by Segmented Flow Injection Analysis, In-Line Ultraviolet Digestion and Amperometric Detection
• Segmented Flow Analysis, Walker Scientific glossary
• Flow Injection Analysis, Walker Scientific glossary
• Continuous Flow Analysis, Walker Scientific glossary
• ASTM D6888 or OIA-1677: Which WAD Cyanide Method, Walker Scientific
• How to Choose a Cyanide Analyser for Gold Processing, Walker Scientific