Working standard solutions were prepared from a stock 1000 mg L-1 copper standard solu­tion (spectroscopic grade), by dilution in 0.14 mol L-1 HNO3 (Synth).

The HEAF samples were prepared by mixing the fuel with an equal volume of 0.14 mol L-1 HNO3,with final volumes of 50 mL [3—5]. Subsequent quantification employed the standard additions procedure.

1.2. Assembly of the TS-FF-AAS system

A schematic diagram of the TS-FF-AAS system is shown in Figure 1.

It is recommended that the Inconel tube should only be positioned above the burner head after lighting the flame, to avoid the possibility of an explosion within the tube due to gas accumulation. The TS-FF-AAS system was therefore first assembled, after which the spec­trometer flame was ignited immediately after opening the gas valves to avoid any explosion risk. This procedure facilitated the positioning of the tube above the burner head, which was performed while the flame was extinguished. All analyses employed a fixed volume of sam­ple, injected into the flow of air as the carrier, since previous work has shown that injection using carrier solutions results in greater sample dilution and dispersion [40,45,54,58].

The sample was introduced into the system using a manual Rheodyne valve (Figure 1), after which it was transported to the ceramic capillary in the flow of air. Since the capillary was heated simultaneously with the metal or ceramic reactor tube, the liquid was partially va­porized, forming a thermospray, and atomization occurred on arrival in the tube, generating a transient signal that was captured and stored by the software. The determination em­ployed the height of the transient signal peak.

1.3. Optimization of carrier flow rate and sample volume

The influences of the carrier flow rate (in the range 9.0-18.0 mL min-1) and the sample vol­ume (50, 100, and 200|jL) were evaluated using a standard of 200|jg Cu L-1.

1.4. Construction of analytical curve

After optimization of the system, analytical curves were constructed in the concentration range 0.1-0.4 mg Cu L-1 in 0.14mol L-1 HNO3. Additions of analyte were made to the sample mixed with an equal volume of 0.14 mol L-1 HNO3. The detection limit (DL) was calculated from 12 blank readings for each type of tube (metal or ceramic).