Sequential injection analysis and flow injection analysis

TUTORIAL
LESSON 4
Dispersion

1. Lesson 1: Introduction
2. Lesson 2: Fundamentals of Flow Injection Analysis
3. Lesson 3: Membrane Sampling Devices
4. Lesson 4: Dispersion
5. Lesson 5: Enrichment
6. Lesson 6: Chemistry
7. Lesson 7: Sequential Injection Analysis
8. Lesson 8: Zone Fluidics
9. Bibliography

The flowing stream in a FIA/SIA system has two main purposes. One is to deliver the sample zone to the detector. The second is to intermingle the sample zone with carrier enroute to the detector. If the carrier contains reagent, this process brings the analyte and reagent together to promote chemistry that generates a detectable product; it also causes dilution of the sample.

This process is commonly referred to as dispersion in FIA, or controlled dispersion. The theory of dispersion is well treated in a number of texts and other references on FIA (see references at end of lesson). Dispersion is a very complex process, and not easily defined, but we will take our shot at it. We will define dispersion as the dynamic but reproducible intermingling of sample zone with a reagent zone and/or carrier caused by flow patterns created by the dynamics of fluid flow through narrow bore tubing. The following Figure attempts to depict dispersion in a sample zone shortly after injection into a FI stream.

A key word in the definition of dispersion is reproducible. While the dispersion is dynamic and never reaches equilibrium or steady state before the sample zone reaches the detector, it is reproducible at any given instant in time if the factors that affect dispersion are held constant.

These factors include flow rate, tubing ID, type of reactor (e.g. coil, knotted, static mixer, straight, serpentine), length of tubing, and internal architecture of components such as valves, detectors, and connectors, all of which are readily controlled. Thus, the degree of dilution and chemistry caused by the dispersion process during transport of the sample zone from injector to detector can be controlled so that it is reproducibly the same for calibrants and samples. This allows calibration of the system and use of the calibration to quantify the unknown concentration in the samples.

Dispersion is a generally a double-edged sword because, on the one hand it promotes chemistry between analyte and reagent which enhances sensitivity and, on the other hand, it causes dilution, lowering sensitivity, increasing peak broadening, and reducing sample throughput (there are, of course, some cases, e.g. with concentrated samples, where dilution is desirable). In general, initially (following injection of the sample) the chemistry effect predominates, leading to a net increase in sensitivity with increased dispersion, but at some point dilution has the predominant effect and sensitivity drops. Therefore, in developing a new methodology, the analyst must find a set of conditions that gives the best balance between enhancement of chemistry and dilution for the application of interest. This is part of the concept of controlling dispersion to achieve the desired results in FIA.

There are actually two types of dispersion in FIA/SIA, axial and radial. Axial dispersion occurs in the direction of stream flow and causes greater dilution and peak broadening than radial dispersion. Its effect can be dramatic; thus, in a typical FIA manifold, an injected sample zone 1 cm long can be stretched to over a meter in length by dispersion by the time it passes through the detector. Axial dispersion predominates in a straight tube. Radial dispersion is caused by flow patterns in the stream that circulate normal to the direction of flow, and thus cause mixing with minimum dilution and peak broadening. Turns in the flow path, e.g. coiling the tubing, and in particular where frequent and sharp changes in the direction of the turns occurs, e.g. with serpentine ("figures of 8") and knotted tubing configurations, promote radial dispersion. For this reason, serpentine reactors and knotted reactors have gained in use over coiled reactors coiled reactors since they lead to greater sensitivity and narrower peaks.

The radial flow patterns created by turns in the flow path are illustrated in the following Figure. As the Figure shows, when the direction of the flow reverses, the radial flow pattern also reverses in direction.

The reversal in radial flow with each reversal in the flow direction of the carrier stream is very efficient in mixing sample with the carrier/reagent with minimal dilution. The effect on peak shape is dramatic, as illustrated in the following Figure, which compares peaks obtained with the same length of tubing reactor for a normal coiled reactor, a serpentine reactor with 42 turns and a serpentine reactor with 92 turns.

Developing a new FIA/SIA methodology, or modifying an existing one, involves manipulating the factors that control dispersion until a combination is found which generates the desired results. The mathematical relationships between most of these factors and dispersion have been worked out and can be found in references, but in practice, it is rare for analysts to use these to select parameters for a methodology. More commonly, they depend on experience, intuition, and trial-and-error. A tubing ID of 0.8 mm appears to have been adopted by most for FIA/SIA, although 0.5 mm ID tubing is also frequently used. For the other parameters, e.g. flow rate, type of reactor, tubing length, a "best guess" of the right combination is generally made, some experiments performed, and then adjustments made until a combination is found which provides acceptable results.

REFERENCES
1.) J. Ruzicka and E. H. Hansen, " Flow Injection Analysis - 2nd edition", J. Wiley and Sons (1988).
2.) B. Karlberg and G. E. Pacey, " Flow Injection Analysis. A Practical Guide", Elsevier (1989).

This completes this session of our Web Tutorial.