Figure 4 illustrates the results obtained for the analytical curve in the concentration range 0.1-0.4 |jg Cu L-1 in 0.14 mol L-1HNO3, using the optimized conditions of the TS-FF-AAS sys­tem. The transient signals were repeatable, and (for both tubes) the curve was linear in the concentration range studied. A two-fold greater sensitivity was achieved using the ceramic tube.

Figure 5 illustrates the results obtained for the analytical curves constructed using concen­trations of Cu in the range 100-400 |jg L-1, with additions of analytein 0.14 mol L-1 HNO3 to equal volumes of sample, under the optimized TS-FF-AAS system conditions. The presence of 75.8 |jg Cu L-1 in the sample was calculated from curve (a), obtained using the ceramic tube. This value was slightly above the detection limit (Table 1), although below the concen­tration of the first point of the analytical curve. In the case of the metal tube (curve (b)), a Cu concentration of 80.0 |jg L-1 was below the detection limit for this tube, but was nevertheless in agreement with the result obtained for the ceramic tube.

A

0.30-| 0.25 0.20 0.15

0.10 0.05 0.00

0 100 200 300 400 500

[Cu] / mg L1

Figure 5. Regression lines fitted to the analytical curves of Cu in 1:1 mixtures of fuel samples and standards prepared in 0.14 mol L-1 HNO3, obtained using the ceramic tube (a) and the metal tube (b) Equations of the lines: A = 1.00×10-2 + 1.32×10-4 (Cu) (ceramic tube); A = 1.16×10-2 + 1.45×10-4 (Cu) (metal tube).

The analytical parameters obtained for the determination of Cu under the optimized condi­tions of the TS-FF-AAS system are provided in Table 2. The analytical curves were linear for a concentration range of 100-400 jag Cu L-1 in 0.14 mol L-1 HNO3. The system could be con­sidered to be sensitive, with characteristic concentrations of 8 and 15 jag Cu L-1 for the ce­ramic and metal tubes, respectively, and analysis frequencies (using HNO3 medium) of 26
and 100 determinations per hour, respectively. Better analytical performance of the system was achieved using the ceramic tube, compared to the metal tube. The data showed that the TS-FF-AAS technique was more sensitive than FAAS, with nine-fold (ceramic tube) and five-fold (metal tube) increases in sensitivity, relative to FAAS with pneumatic nebulization, for which the characteristic concentration was 77 |jg L-1. The increase in power of detectio- nobtained using the ceramic tube was around twice that for the metal tube. The sensitivity for determination of copper using the ceramic tube was therefore two-fold that obtained us­ing the metal tube.

2. Conclusions

The TS-FF-AAS system can be used to determine copper at low concentrations, using either metal (Inconel) or ceramic (Al2O3) tubes as atomizers. Following optimization considering the most important experimental variables affecting atomization, these systems provided significantly improved detection limits for Cu determination, with nine-fold (ceramic tube) and five-fold (metal tube) increases in sensitivity, compared to traditional FAAS with pneu­matic nebulization. The TS-FF-AAS technique is simple, fast, effective, and inexpensive. It requires low volumes of sample (as little as 50 |jL) and reagents, and reduces waste genera­tion. The method offers a useful new alternative for the determination of copper in alcohol.

Acknowledgments

The authors thank UFOP and CNPq for financial assistance.

Author details

Fabiana Aparecida Lobo1, Fernanda Pollo2, Ana Cristina Villafranca2 and Mercedes de Moraes2

1 UFOP — Universidade Federal de Ouro Preto, Brazil

2 UNESP — Universidade Estadual Paulista, Brazil