1.5. Optimization of carrier flow rate and sample volume

Good peak reproducibility was achieved when samples were injected into the air flow as carrier. When samples were injected into 0.14 mol L-1 HNO3, used as the carrier, there was a rise in the baseline (as expected, due to increase of the blank), followed by a fall due to cool­ing of the metal or ceramic tubes. This cooling was significant, since no transient signals were obtained following injection of standards, indicating that the temperature within the tubes was insufficient to atomize the analyte, which remained dispersed in the carrier solu­tion. This confirmed the findings of earlier work that the use of air (or other gas) as the carri­er avoids dilution and dispersion of the sample. Here, all analyses were performed using air as the carrier, not only because it was less expensive than use of a solution, and minimized waste generation, but also because it enabled the TS-FF-AAS system to be used to determine copper, which would not have been possible using a solution as the carrier.

Figure 2 shows the influence of the carrier (air) flow rate, in the range 9.0-18.0 mL min-1, on the absorbance values obtained using 50 |jL of a standard of 200 |jg Cu L-1in 0.14 mol L-1 HNO3, using both tubes. In the case of the metal tube, lower absorbance values were ob­tained at low flow rates, because the sample arrived slowly at the atomizer, increasing the measurement duration and resulting in an unpredictable and erratic vaporization. Hence, as the flow rate was increased, the absorbance also increased due to a more homogeneous va­porization of the sample [23,27,58].

This increase proceeded up to a carrier flow rate of 12.0 mL min-1, above which there was no significant variation in absorbance. The highest absorbance value was obtained at a flow rate of 18.0 mL min-1, which was therefore selected as the best flow rate to use with the metal tube.

When the ceramic tube was used, maximum absorbance was achieved at a carrier flow rate of 9.0 mL min-1. At higher flow rates, the residence time of the liquid in the heated section of the ceramic capillary was considerably diminished, reducing the time available for evapora­tion of the liquid, so that the sample was not delivered in the form of vapor/aerosol, but rather as a flow of liquid. The temperature within the tube decreased, and the color of the tube changed from ruby-red to opaque grey. It was also possible to see droplets emerging from the atomizer tube. Hence, the absorbance values did not increase, while greater varia­bility in the signal resulted in elevated standard deviation values. A flow rate of 3.0 mL min-1was selected, at which the absorbance signal was maximized, and the standard devia­tion was minimized.

The sample volume was varied between 50 and 200 |jL, using carrier flow rates of 18.0 and 9.0 mL min-1 for the metal and ceramic tubes, respectively. The results (Figure 3) re­vealed that for both tubes a sample volume of 50 |jL generated the highest absorbance value, with a low standard deviation, reflecting good repeatability in the experimental measurements. When 100 |jL of sample was used, there was a slight cooling of the ceram­ic capillary, and consequently of the atomization tubes, while there was no increase in the absorbance values. At a sample volume of 200 |jL, the ceramic capillary and the tube were substantially cooled, and there was no homogeneous thermospray formation, with erratic generation of droplets that acted to disperse the light radiation (probably to a large degree, since the deuterium lamp was unable to fully correct the resulting background signal). The unpredictable atomization resulted in very high standard deviation values. Using air as the carrier, a sample volume of 50 |jL was selected for the subsequent meas­urements, due to greater atomization homogeneity, satisfactory absorbance for a 30 mg Cu L-1 standard, and a low SD value.