Application
Concentration, as it actually moves
Scheduled sampling gives you the concentration at the moment of sampling and an assumption for every moment in between. A 0.1 mL flow cell on a recirculation loop gives you the curve.
The measurement problem
Most streams are characterised by a titration or a chromatogram taken on a schedule. Between readings, composition is inferred from process settings — feed rate, temperature, elapsed time — which is a model of the process rather than a measurement of it.
The gap matters most where the stream is being replenished, consumed or diluted continuously. A bath that is used and topped up, a blend that is being trimmed to target, a reactor being fed: in each case the variable you are controlling is the one you are not watching.
What a live reading changes
- A continuous trace of dissolved concentration, updated every second.
- Drift detected as it starts rather than at the next scheduled check.
- A control input over OPC UA, so the reading can close a loop rather than fill a log.
- Calibration against your own reference method — usually HPLC — so the number means the same thing as the one you already trust.
- Demonstrated detection down to 0.05 m%, with the limit depending on the analyte and the path length chosen.
The instrument reads through a flow cell fed by a recirculation loop or a representative sampling point — particles <15 µm, viscosity under 10,000 cP, pH 2–10.
What has actually been measured
- Glucose in aqueous solution from 0.1 g/L to 100 g/L, quantified with a PLS model — the same model has since been used to predict glucose in customers' cell-culture media.
- Weak but detectable glucose signal still present at 0.1 g/L, at the low end of that range.
- Acetone in a toluene matrix down to 0.05 mass%, read in the 1140–1300 cm⁻¹ window, with reproducible separation between neighbouring sub-percent concentrations.
Both are published application notes. The detection limit that applies to your stream depends on the analyte and the path length chosen, which is what the feasibility check settles.
Is this your measurement?
Send the stream, the concentration range and the decision it blocks. Where it makes sense we test on your own process samples first.