Total Reflexion X-Ray Fluorescence Analysis for Ca, Cu, Fe and K in Wines. Analytical Testing Respect to Traditional Techniques as Flame Atomic Absorption and Emission Spectrometry

The analysis of macro and trace elements is an important information for assuring the quality of red and white wines. Several metals as As, Br, Cd and Zn (1) are harmful to human health, as well as, other metals such as Ca, Cu, Fe, and K, are of crucial importance for quality in wine production. Among the most common analytical techniques that are applied in wine analysis are the flame atomic absorption and emission spectrometry (FAAS and FAES, respectively), and also, the multi-elemental analytical techniques as ICP-MS and ICP-OES (2). Normally for this techniques, a sample pretreatment by classical wet digestion procedures with concentrate acids or ultraviolet irradiation are required. In the last years, the techniques so-called total reflection X-ray fluorescence (TXRF) has been investigated and applied for multi-elemental analysis in wines (3), and in general, has been well received by the analytical community because offers lower operating costs and almost no use of sample digestion is required. However, also is well know the overlap of the fluorescence emission lines, which severely limits its application, for example, Cd and As in wines could not be analyzed because its Lα-line at 3.133 keV has an overlap with strong K Kα-line. This work presents the results of accuracy for K, Ca, Cu, Fe and K in wines by TXRF. An analytical testing of the accuracy respect the traditional techniques applied in wines as flame atomic absorption and emission spectrometry, were realized. A benchtop total reflection X-ray fluorescence instrument, model S2 PICOFOX and a red wine certified reference for enology laboratory (MATERIAL, TITRIVIN BTN, Batch T03121217B), were used. For the metal analysis in wine both FAAS and FAES a conventional pretreatment was used. For TXRF 10 µL of an internal standard solution (Ga, 100 mg/L) were added to a sample aliquot of 1 mL and homogenized, and finally 5 µL this solution were then applied on to quartz glass carrier and carefully dried on heating plate. The samples were analyzed in triplicated for 1000 seconds each. The results show that for a Ca, Cu, Fe and K by TXRF determination a good accuracy is found (no statistically significant difference at 95% confidence between the reference value and that found by TXRF). On the other hand, when performing the analyzes of Ca and K, both by FAAS and by FAES, statistically significant differences were evidenced. Major antecedents of these results will be announced in the presentation of this work.