Coverly (24) provides operational details and performance
characteristics of parallel-plate dialyzers and dilution loops,
both of which are used routinely to achieve 15- to 150-fold
dilution factors in CFA applications. Dilution by continuousflow
dialysis is straightforward. The extent of dilution is
governed by the length and depth of dialyzer channels, donor
and recipient streamflow rates, temperature, and the type
and thickness of the dialysis membrane. The main disadvantage
of dilution by dialysis is temporal variation in analyte
transport rates that can be caused by temperature fluctuations,
ionic strength differences between donor and recipient
streams, and membrane fouling. Dilution loops are auxiliary
analytical cartridges (see the upper section of Figure 1) in
which samples are proportioned into an air-segmented
diluent stream, mixed, and returned to the inlet side of the
pump. A small portion of each diluted sample is aspiratedinto the primary analytical cartridge (lower section of Figure
1), while the remainder flows to waste. Dilution loops are
less affected by temperature fluctuations and sample-tosample
ionic strength variation than dialyzers and were easy
to implement with the CFA hardware used for this work.
Dispersion that otherwise would have occurred when diluted
samples made a second pass through thepumpwas reduced
substantially by maintaining air-segmented flow within the
resample pump tube (24). This result was achieved by
debubbling and immediately rebubbling the dilution loop
stream just before it reached the resample fitting (see fittings
2 and 3 in Figure 1). The volume of these newly injected air
bubbles was about three times greater than normal to ensure
that the resample pump tube would aspirate a portion of
each that was sufficiently large to fully occlude the 1-mm i.d.
conduits of the primary analytical cartridge. The withdrawal
legs of the waste and resample pump tubes connected to
fittings 2 and 3 were shortened to about 3 cm. Slight
adjustment (( 3 mm) of the length of the resample pump
tube might be necessary if resampled air bubbles shear as
they pass under the pump rollers. Results reported in the
remainder of this paper were obtained with the dilution loop
analytical cartridge shown in Figure 1. For some types of
CFA analyzers, substituting a dialyzer for the dilution loop
in the NaR-reduction method analytical cartridge might be
easier to implement and should yield comparable results.