Tubes, connecting short, dispersion

Figure 9.4 Column for two-phase dispersion measurements. To emphasize the tracer spread outside the column, most of the long column has been omitted in this graph. (A) The 10 pi injection loop (B) a thin needle is located inside the T-connector and reducing coimector, (C) the needle is inserted into the bed of particles (D) the gas-liquid separatory (E) detail of the separator, including the relevant liquid volume (F) a short tube connects the separator to the flow cell. (Source Marquez et al. [20]. Reproduced with permission of John Wiley Sons.)...

The curve in Figure IB is probably more useful from a practical point of view. Although the standard deviations of any dispersion process are not additive, they do give a better impression of the actual dispersion that a connecting tube alone can cause. It is clear that a tube 10 cm long and 0.012 cm I.D. can cause dispersion resulting in a peak with a standard deviation of 4 pi. Now, a peak with a standard deviation of 4 pi would have a base width of 16 pi and, in practice, many short... 

However, when assessing the length of tube that can be tolerated, it must be remembered that the 10% increase in variance that can be tolerated before resolution is seriously denigrated involves all sources of extra-column dispersion, not just for a connecting tube. In practice, the connecting tube should be made as short as possible and the radius as small as... 

When the residence time becomes shorter, this approach becomes questionable for several reasons. For example, the asymptotic state may not have been reached yet, or the peaks may be unsymmetrical. These "short-time" situations may be encountered when trying to apply chromatographic concepts to the study of dispersion in connecting tubes, or in some apphcations, such as hollow-fiber liquid chromatography. Shankar and Lenhoff [77] have derived a solution in the time domain, using series expansion. This solution can be implemented by numerical computation for the determination of concentration profiles inside a tube coated with a retentive layer, when the fluid flow is laminar. This solution is valid for systems that are either short or long after the Taylor-Aris definition. 

Although these are the most important considerations in designing a parallel reactor module, another important point is the analysis. While in principle each analytical instrument is suitable and can be connected to the exit of such a multiple-pass reactor, one must ensure that the instrument works with relatively small amounts of sample, as well as low flow rates. For the model reaction under investigation in our reactor, the CO-oxidation, non dispersive IR is used. As C02 concentrations are relatively high under our conditions, the analysis chamber can be kept short, and purge times are therefore also short. Analysis times are around 4 minutes, this being determined mainly by the purge times of the tubes and the sample chamber. However, any analytical techniques, such as mass spectrometry, GC, etc., can in principle be used in connection with this set-up. 

Fig.l. (a) FI A manifold for simultaneous determinations of pH and pCa in serum samples as operated under conditions of limited dispersion. S, point of injection (30 fil) pH, flow through capillary glass electrode FC, flow through cell containing a PVC-membrane-based calcium selective electrode and the common reference electrode, details of which are shown below (b). The carrier solution (A and B) is THIS buffer of pH 7.4, the connecting tubes (a and b) being made as short as possible. The carrier solution supplied via line B to the tip of the reference electrode is included in order to stabilize the reference electrode junction potential for sera measurements. In (c) is shown the potentio-metric determination of the ionized calcium content in 6 serum samples, bracketed by two calibration runs of aqueous calcium standards (0.5 to 5 mM), all assays made in triplicate. ... 

The curves in Figure 8 clearly demonstrate the relative performance of tubes of different geometry. Although the coiled tube has a high variance, the peak is almost symmetrical, but obviously, the curve from the serpentine form is both better in shape and contributes less variance. Nevertheless, the dispersion is still very large when compared with the dispersion from the small bore column previously discussed. Ipso facto connecting tubes should be made as short as possible whatever their geometric form. 

---