Fast and direct screening of copper in micro-volumes of distilled alcoholic beverages by high-resolution continuum source graphite furnace atomic absorption spectrometry

Accepted Manuscript Fast and direct screening of copper in micro-volumes of distilled alcoholic beverages by high-resolution continuum source graphite furnace atomic absorption spectrometry Zsolt Ajtony, Nikoletta Laczai, Gabriella Dravecz, Norbert Szoboszlai, Áron Marosi, Bence Marlok, Christina Streli, László Bencs PII: DOI: Reference:

S0308-8146(16)30992-X http://dx.doi.org/10.1016/j.foodchem.2016.06.090 FOCH 19431

To appear in:

Food Chemistry

Received Date: Revised Date: Accepted Date:

10 December 2015 15 June 2016 26 June 2016

Please cite this article as: Ajtony, Z., Laczai, N., Dravecz, G., Szoboszlai, N., Marosi, A., Marlok, B., Streli, C., Bencs, L., Fast and direct screening of copper in micro-volumes of distilled alcoholic beverages by high-resolution continuum source graphite furnace atomic absorption spectrometry, Food Chemistry (2016), doi: http://dx.doi.org/ 10.1016/j.foodchem.2016.06.090

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Fast and direct screening of copper in micro-volumes of distilled alcoholic beverages by high-resolution continuum source graphite furnace atomic absorption spectrometry

Zsolt Ajtony1, Nikoletta Laczai2, Gabriella Dravecz2, Norbert Szoboszlai3, Áron Marosi1, Bence Marlok2, Christina Streli4, László Bencs2,‡

1

Institute of Food Science, University of West Hungary, H-9200 Mosonmagyaróvár, Lucsony utca 15-17, Hungary 2

Institute for Solid State Physics and Optics, Wigner Research Centre for Physics, Hungarian Academy of Sciences, P.O. Box 49, H-1525 Budapest, Hungary

3

Laboratory of Environmental Chemistry and Bioanalytics, Department of Analytical

Chemistry, Institute of Chemistry, Eötvös Loránd University, P.O. Box 32, H-1518 Budapest, Hungary 4

‡

Atominstitut, Technical University of Wien, A-1020 Vienna, Stadionallee 2, Austria

Corresponding author. E-mail: [email protected]; Phone: +36-1-392-2222/1684; Fax: +36-1-392-2223.

1

Abstract – HR-CS-GFAAS methods were developed for the fast determination of Cu in domestic and commercially available Hungarian distilled alcoholic beverages (called pálinka), in order to decide if their Cu content exceeds the permissible limit, as legislated by the WHO. Some microliters of samples were directly dispensed into the atomizer. Graphite furnace heating programs, effects/amounts of the Pd modifier, alternative wavelengths (e.g., Cu I 249.2146 nm), external calibration and internal standardization methods were studied. Applying a fast graphite furnace heating program without any chemical modifier, the Cu content of a sample could be quantitated within 1.5 min. The detection limit of the method is 0.03 mg/L. Calibration curves are linear up to 10-15 mg/L Cu. Spike-recoveries ranged from 89% to 119% with an average of 100.9±8.5%. Internal calibration could be applied with the assistance of Cr, Fe, and/or Rh standards. The accuracy of the GFAAS results was verified by TXRF analyses.

Keywords: distilled alcoholic beverages, heavy metal content, rapid furnace program, calibration methods, electrothermal atomic absorption spectrometry, graphite furnace electrothermal vaporization.

2

1. Introduction Double-distilled alcoholic beverages, made from various fruits and grains after fermentation, are popular drinks worldwide. The distillation process, however, can yield high copper content in the final product, due to the contact of the distillate with the copperware distillation equipment. According to present-day legislations, the Cu content of any drink used for human consumption should not exceed the safety limit of 2 mg/L (WHO, 2004). Sadhra et al. (2007) surveyed the dietary exposure to Cu in the European Union, and assessed its risk to the population. Besides the intake of Cu from regular daily meals, they have found significant intake from alcoholic beverages as well. As a consequence, there is a certain demand for the fast and direct determination of Cu in distilled drinks, in order to check, whether they exceed the permissible limit. The Cu content of various dietary matrices, particularly drinks, has been reported worldwide, determined by means of flame atomic absorption spectrometry (FAAS) (Onianwa et al., 2001; Karadjova et al., 2002; Huguet, 2004; Soufleros et al., 2005; Navarro-Alarcón et al., 2007; Kostić et al., 2010; Raposo et al., 2012; Bonić et al., 2013; Rodríguez-Solana et al., 2014), graphite furnace atomic absorption spectrometry (GFAAS) (Bermejo et al., 2001; Karadjova et al., 2002; Adam et al., 2002; Galani-Nikolakaki et al., 2002; Lara et al., 2005; Catarino et al., 2005; Jurado et al., 2007; Miranda et al., 2010), inductively coupled plasma optical emission spectrometry (ICP-OES) (Murányi and Papp, 1998; Kokkinofta et al., 2003; Lara et al., 2005; Reche et al., 2007; Moreno et al., 2008), inductively coupled plasma mass spectrometry (ICP-MS) (Lachenmeier et al., 2009; Flores et al., 2009; Rodríguez-Solana et al., 2014), anodic stripping voltammetry (Agra-Gutiérrez et al., 1999), kinetic complexformation spectrophotometry (Mitić et al., 2009), size-exclusion chromatography high performance liquid chromatography coupled to GFAAS (Bermejo et al., 2001), total reflection X-ray fluorescence spectrometry (TXRF) (Galani-Nikolakaki et al., 2002), and high

3

temperature liquid chromatography coupled to ICP-OES (Terol et al., 2011). The methods published in the field have also been reviewed in the literature (Pyrzyńska, 2004; Stafilov and Karadjova, 2009; Szymczycha-Madeja et al., 2015). For the fast screening of Cu in alcoholic beverages, implying low-volume sample consumption (for economic reasons), it is necessary to use a selective and highly sensitive technique.

High-resolution

continuum

source

graphite

furnace

atomic

absorption

spectrometry (HR-CS-GFAAS) is an emerging technique of instrumental analysis (Welz et al., 2005; Welz, 2005), which meets the latter requirements. It has high absolute sensitivity (detection at pg-ng levels), and utilizes micro-amounts of samples, e.g., 10-100 µL for solution and 0.01-10 mg for solid (powder) introduction. So far, HR-CS-AAS methods have not often been reported for the analysis of food products, but for distilled alcoholic beverages applying FAAS and internal standardization (Raposo et al., 2012). However, this and the above listed methods need relatively high sample amounts, i.e., some milliliters, and/or the stepping of the monochromator optics over the spectral lines of analytes and internal standards, which prolongs the analysis time. In another study, a HR-CS-GFAAS method was elaborated for the quantitation of Cd, Cr, Cu, Pb and Ni in acid-digested tea samples using external calibration (Zhong et al., 2016). Caldas et al. (2007, 2009) utilized external calibration and internal standards for the simultaneous determination of As, Cu and Pb in sugarcane spirits using a four-beam line-source GFAAS spectrometer. They also studied the effects of W permanent modifier for the same set of analytes with the co-injection of PdMg(NO3)2 (Caldas et al., 2007). Chen et al. (2014) elaborated a sequential method to quantitate Cu, Cd, Cr, Pb, Mn and Al in cooking wine by HR-CS-GFAAS after acidification and dilution of the samples. These methods, however, could not be adapted to the present analytical task, since they utilize too sensitive analytical lines, thus they require sample dilutions to a high extent, especially, those with enhanced Cu content, such as the set in this

4

study. Moreover, they utilize chemical modifiers and/or conventional graphite furnace heating programs, which involve a long total analysis time and increase the cost of the analysis. The aim of the present work was to develop and apply a rapid and low-cost HR-CSGFAAS method with direct sample injection and low sample consumption (e.g., some microliters) for the screening of the Cu content in Hungarian distilled alcoholic beverages (called pálinka), in order to check, if they meet the WHO standard. For this purpose, various experimental conditions/methods were studied, such as alternative analytical lines, regular and fast graphite furnace heating programs, utilization of chemical modifiers at various amounts, external calibration and internal standardization. Moreover, a TXRF method has been applied to verify the accuracy of the GFAAS results.

2. Experimental 2.1. Instrumentation All the GFAAS experiments were performed on an Analytik Jena Model ContrAA700 tandem spectrometer (Analytik Jena AG, Jena, Germany) equipped with a transversally heated graphite tube atomizer (THGA) and an MPE-60 (Analytik Jena) autosampler. The spectrometer consists of a xenon short-arc lamp working in hot-spot mode as a primary radiation source, a high-resolution Echelle double monochromator, and a charge coupled device (CCD) array detector with 588 pixels. A part (200) of these pixels is applied for detection purposes, while the rest is utilized for internal, spectral correction. The monochromator consists of a pre-dispersing prism and a high-resolution echelle grating, both arranged in Littrow mounting. The xenon lamp was used with a lamp current of 13 A, but without hot-spot tracking. A less intense, alternative spectral line of Cu (249.2146 nm, resolution: 1.53 pm/pixel) was selected for the analyses.

5

Peak volume selected absorbance (PVSA) (Welz et al., 2005), i.e., the integrated absorbance (Aint) of the central pixel (CP) plus adjacent ones (CP±1), corresponding to a spectral range of ~4.6 pm has been utilized for signal evaluation. The method of iterative background correction (IBC) was selected for the assessment of the non-specific absorption. The assignment of analytical lines and molecular bands were mostly based on the application of ASpect CS 2.1.1.0 Software® (Analytik Jena AG), and also using various spectral tables (Zaidel et al., 1962; NIST, 2015). Pyrolytic graphite coated graphite tubes with a dosing hole and PIN-graphite platforms (Analytik Jena Part No.: 407-A81.026) were applied throughout the study. The graphite tubes were conditioned each day before the first analytical run using the formatting furnace heating program offered by the ContrAA software. This step is also important for true-temperature calibration of the THGA. High-purity argon (5N) from Messer-Hungary was applied as a graphite furnace purge gas. The lifetime of the graphite tubes was in the range of 800-850 firings. The TXRF analyses were performed by using an Atomika Model 8030C spectrometer (ATOMIKA Instruments GmbH, Oberschleissheim, Germany) equipped with a 2.5 kW X-ray tube made of a Mo-W-alloy anode and a double-W/C multilayer monochromator, adjusted to reach an excitation energy of 17.4 keV (Mo-Kα). The characteristic emission from elements present in the samples was detected by a Si(Li) detector with an active area of 80 mm2 and a resolution of 150 eV at 5.9 keV. The measurements were performed operating the X-ray tube at 50 kV accelerating voltage. The tube-current was adjusted automatically as a trade-off between the detector dead time and total analysis time. The Kα line of Cu at 8.047 keV was used for the determinations. The acquisition time was set to 500 s.

2.2. Materials and methods

6

All reagents were of analytical grade or better (supplier: Reanal, Budapest, Hungary, if not stated otherwise). Ion-exchanged, deionized water from an ELGA Purelab Option-R7 unit (ELGA LabWater/VWS(UK) Ltd., High Wycombe, UK), or from a MilliQ unit (Millipore, Bedford MA, USA), each with a resistivity of 18 MΩ cm, was applied for sample/standard preparation and dilution using calibrated volumetric flasks and micropipettes. Before use, each flask, micropipette tip and autosampler cup was cleaned by means of soaking first in 1:1 diluted solution of concentrated (cc.) HCl and then in 1 mol/L HNO3. After then they were rinsed with high-purity water several times and dried in an electric drier oven at 50 oC. Aliquots of 20 µL samples/standards and 5 µL of the Pd chemical modifier (as nitrate, take from a 1000 mg/L Pd stock, Alpha Ventron) were dispensed by the MPE autosampler. Spike experiments were conducted by means of dosing Cu-standards, along with the chemical modifier and the selected sample, mixed on the PIN-platform of the graphite tube. For common analyses, the samples were injected into the PIN-platform without any sample preparation step. The standard solutions of Cu were prepared via dilutions from a 1000 mg/L Cu stock solution (Alpha Ventron), and adding 1 mL cc. HNO3 for acidification and 50 mL absolute ethanol, and then it was made up to 100 mL final volume. Multi-element internal standards of Au, Cr, Fe, Na, Os, and Rh were prepared at concentrations of 0.4, 7.8, 0.25, 6.9, 3.0, and 8.0 mg/L, respectively, from 1000 mg/L single element stocks (supplier: Alpha Ventron/Merck). For the TXRF analysis, a 100 µL aliquot of each sample along with 2 µL of 10 mg/L Ga, applied as internal standard, was dispensed onto a quartz sample holder (reflector plate) of the TXRF spectrometer. The loaded plates were put under a laminar flow exhaust hood and dried at 80 °C until complete evaporation of the solvent, then subjected to TXRF analyses.

3. Results and discussion 7

3.1. Optimization of the graphite furnace heating program The drying stages of the graphite furnace heating programs were optimized with the utilization of the built-in video camera of the THGA, to assure smooth and complete removal of the solvent and the organic matrix. As the next optimization step, the Welz-functions of pyrolysis and atomization curves were recorded at Cu I 249.2149 nm with the use of standards of 1 mg/L Cu with 5 µg Pd and without modifier (Fig. 1), using fast and regular graphite furnace heating programs, with an atomization temperature (Tat) of 2200 oC and a pyrolysis temperature (Tpyr) of 800 oC. In the presence of the Pd modifier, the pyrolysis curve of Cu shows an increase from 400 oC to 700 oC, then declines slightly from 700 oC to 1100 o

C, develops a plateau between 1100 oC and 1400 oC, and steeply drops above 1400 oC. The

atomization curve of Cu sharply increases at Tat higher than 1500 oC. A detectable Cu signal (just distinguishable from the background) can already be observed for Tat between 1300 oC and 1400 oC, corresponding to the appearance temperature (Tapp) of Cu atoms. The atomization curve has two plateau section, one is between 2000-2200 oC, but after then, it slightly drops, showing likely evaporation losses of Cu, e.g., via diffusion and/or convection at the ends and/or the dosing hole of the atomizer. When applying fast and regular graphite furnace heating programs without any chemical modifier, the optimal Tpyr and Tat is much lower, i.e., 900/1100 oC and 1800/2000 oC, respectively. Finally, 900 oC and 2200 oC as optimal Tpyr and Tat, respectively, were selected for the analysis. The pyrolysis and atomization curves were recorded for candidate internal standard elements (i.e., Au, Cr, Fe, Os, Na, and Rh) too (Fig. 2), in order to find compromise (optimal) Tpyr and Tat values between any of them and Cu. The pyrolysis curve of Au declines at Tpyr higher than 1500 oC, while those of Cr, Fe and Rh start to decline at around 1600 oC, 1700 oC and 1800 oC, respectively. These maximum allowable Tpyr values are higher than those observed for the analyte. Consequently, the optimal Tpyr of Cu has to be applied for analyses 8

with internal standardization, in order to avoid analyte losses. The atomization curves for the candidate internal standard elements showed a different pattern than those of Cu (Figs. 1-2). Several rather refractory elements have not developed any plateau on the atomization curve, even at as high temperature as 2600 oC. For Au, Cr, Fe, Os, Na, and Rh, the optimal Tat values (corresponding to Aint signal maximum) were found to be 2100, 2400, 2100, 2600, 2200, and 2600 oC, respectively. On the other hand, for Os and Rh, instead of the optimal, the maximum allowable temperature for THGA furnaces should be applied as a compromise Tat, i.e., 2400 o

C, which condition is, however, causes a severe loss of sensitivity.

3.2. Transient spectra Typical 3D, time- and wavelength-resolved spectra of standards and samples at and around the analytical line of Cu are depicted on Fig. S1. The spectra observed in the presence of the Pd modifier contain two additional lines, i.e., Fe I 249.0645 nm and Fe I 249.1156 nm (Fig. S1). These lines are due to traces of Fe contamination present in the sample and the modifier solutions. Interestingly, double peaks of the atomization transients could also be observed. The applied solution sample introduction results in a near uniformly distributed analyte/modifier concentration on the PIN-platform. Thus, the double-peak effect is likely due to the action of the chemical modifier, which exhibits vaporization of Cu in two different chemical forms, e.g., reduced Cu, and/or CuI/CuII oxides. In graphite atomizers, Cu has been reported to vaporize mainly as its oxide, although it can be reduced on the surface of the THGA tubes too, as is observed for end-heated atomizers (Frech et al., 1985). This would eventually result in atomization transients consisting of overlapping, or even separated Cu peaks (Fig. S1). Typical 3D transient spectra for the multi-element standard with and without the analyte are depicted on Fig. 3. The medium intense atomization transient of Cu can be seen at

9

249.2146 nm (Fig. 3a). Clearly distinguishable and fairly intense absorption transients of Fe appear at 248.975 nm, 249.0645 nm, and 249.1156 nm. The absorption transients of Au, Cr, Rh and Na are much weaker, thus it is rather difficult to distinguish them from the background, though they are present at fairly high concentrations. The Au line at 249.120 nm is partly overlapped by one of the Fe lines as above (Fig. 3b). Moreover, the Os peak appears at the side of the Rh transient (at 249.230 nm), apparently with low intensity, which is due to the low atomization efficacy of this element at the applied compromise Tat (2400 oC) (Fig. 3a) and even at 2600 oC (Fig. 3b).

3.3. Search for apt internal standard In HR-CS-GFAAS, the utilization of any element as an internal standard for calibration is possible, if it has at least one spectral line situated in the selected spectral range of the CCD around the spectral line of the analyte (Resano et al., 2013). For example, for the current analytical task and the involved spectrometer, it should be within λCu±150 pm. Besides, the vaporization and atomization behavior of the internal standard in the graphite furnace should be similar to that of the analyte. In the case of the Cu I 249.2146 nm analytical line, several spectral lines of candidate internal standard elements are present in the selected spectral range of the CCD, i.e., from 249.0613 nm to 249.3664 nm (Table 2). On the base of the fairly high relative intensities of these lines, observed in arcs and sparks (Zaidel et al., 1962), the candidates for internal standardization are Au, Cr, Fe, Os, Na and Rh. The utilization of some other elements as internal standards, having also a few spectral lines in the desired wavelength range, like As, Cd, Ni, Sn, W and Zn were ruled out, due either to the low relative intensities of their nearby spectral lines (Table 2), and/or their different vaporization behavior in the graphite furnace compared to that of the analyte. Additionally, a few of these

10

elements can form molecular species (e.g., SnO) with considerable band absorption over the concerned spectral range. As a first step, using multi-element standards, the Aint signal intensity was determined on the spectral lines of Au, Cr, Fe, Os, Na and Rh. The Aint signal was observed to be fairly low for Na and Os, even applying high concentrations (e.g., some mg/L levels) in the standards, as compared to those observed for Cr, Fe and Rh. These and the above results indicate that a part of the spectral lines of candidate elements for internal standards are with fairly low intensities, even when present at high concentrations in the sample/standard solutions, while other lines are being (partly) overlapped. Therefore, as an accurate analytical approach, the application of Cr, Fe and Rh as single or even concurrent internal standards can be recommended.

3.4. Calibration and analytical performance The calibration curves, recorded with the use of regular graphite furnace heating program and Pd modifier, are linear up to 20 mg/L Cu (Fig. S2a). The use of fast furnace heating programs without any modifier brings more uncertainty to the calibration, in terms of a widened confidence interval of the calibration curve (Figs. S2b and S2c). Additionally, the upper limit of the dynamic range became lower (i.e., 15 mg/L). On the other hand, by means of using second order fitting to the calibration points, one can extend the upper limit of the dynamic range up to 20-25 mg/L Cu. This concentration range is enough wide to cover not only the Cu content in the present set of alcoholic beverages, but for instance, those, which are made with the assistance of similar, double-distillation technologies (e.g., GalaniNikolakaki et al., 2002; Soufleros et al., 2005). The correlation coefficients (R values) of the linear fittings to the calibration points are not worse than 0.9996. The limit of detection (LOD) of the method, (expressed as three time the standard deviation (SD) of the blank

11

determined in multiple runs (n=11) and divided with the slope of the linear calibration), is 0.03 mg/L. It is important to note here that by the currently applied HR-CS-AAS spectrometer, one can further extend the linear range of the calibration, for example, by means of using ±2, or even higher number of pixels around the CP of the CCD for signal evaluation. In order to check the accuracy of the determinations, spike experiments were conducted with the addition of a 5 mg/L Cu standard to each sample. The calculated recoveries for Cu range from 89% to 119% with an average of 100.9±8.5% (Table 3), which are acceptable values for the present, fast screening analysis based on direct dispension of low sample volumes. The utilization of internal standards was performed in a couple of samples by adding the multi-element standard solution as specified above (Section 2.2). Amongst the candidate elements, only Cr, Fe and Rh were found to be as usable internal standards, providing fairly accurate results for the set of the study samples. It should be noted, however, that the use of Fe as an internal standard could lead to inaccurate analytical results, which arise from its enhanced concentration in a few samples, as visualized in Fig. S1. On the other hand, for Rh standard, an increased Tat of 2600 oC should be used as optimal to attain enough sensitivity as as mentioned above. Consequently, for the purpose of internal calibration, Cr and Fe as concurrent standards could be recommended. These elements were also applied for the present set of samples as follows below.

3.5. Analytical results The Cu content of a large set of alcoholic beverages (39 samples), originating from domestic and industrial distilleries in different regions of Hungary have been quantitated by means of various GFAAS and TXRF methods (Table 3). In general, satisfactory agreement was achieved between the results of these methods. The Cu content of three of the studied 12

samples was only below the LOD of the current GFAAS methods, while in the rest samples, it ranged from 0.05 mg/L to 22 mg/L (average: 4.46 mg/L, median: 2.56 mg/L). It is worth noticing that twenty-one out of the 39 study samples had Cu content above the threshold value of the WHO. According to a recent review (Szymczycha-Madeja et al., 2015), the Cu content of various alcoholic beverages fluctuates over a wide range, i.e., from some µg/L to some tens mg/L. Consequently, the present samples fall rather in the category of drinks with outstandingly high Cu content.

4. Conclusions The rapid HR-CS-GFAAS determination of Cu in distilled alcoholic beverages with high Cu content (i.e., in the range of 0.03-25 mg/L) can be performed on a less intense spectral line of Cu (249.2146 nm) by means of directly dispensing small sample aliquots of 15-20 µL, applying fast graphite furnace heating programs without or with Pd modifier and matrix-matched external calibration. This method avoids any sample pre-treatment and dilution. It also allows performing analyses in the concentration range around the WHO’s safety limit for Cu in drinks. Accurate calibration via addition of Cr, Fe and/or Rh as internal standards could also be possible. In a couple of cases, the use of the chemical modifier is problematic, due to its enhanced Fe blank values. The relatively high Fe content of a few samples makes the usability of Fe for internal standardization rather difficult. More than 50% of the presently studied samples had Cu content above the permissible concentration of Cu in drinks. These results raise a concern on the negative health effects of Cu for a considerable part of the Hungarian population, who are regular consumers of these types of popular alcoholic drinks. The HR-CS-GFAAS methods developed in this study have been found to be simpler, faster and more cost-efficient, compared to former methods reported in the literature. In

13

general, various drinks for human consumption often have highly enhanced Cu content, for instance, as reported for wines (Galani-Nikolakaki et al., 2002), marc distillates (Soufleros et al., 2005), aniseed sprits (Jurado et al., 2007), plum brandies (Bonić et al., 2013), and for teas, cocoa and coffee (Onianwa et al., 2001). The currently developed method has of certain relevance in food quality control, since it can be easily adapted to various dietary/drink samples with the important aim of the fast screening of the Cu content, in order to determine, if it is below the threshold value of the WHO.

Acknowledgments – This work was partly supported by the Research Infrastructural Developments of the Hungarian Academy of Sciences (project No.: IF-037/2013).

14

References

Adam, T., Duthie, E., & Feldmann, J., (2002). Investigations into the use of copper and other metals as indicators for the authenticity of Scotch whiskies. Journal of the Institute of Brewing, 108, 459-464.

Agra-Gutiérrez, C., Hardcastle, J.L., Ball, J. C., & Compton, R. G., (1999). Anodic stripping voltammetry of copper at insonated glassy carbon-based electrodes: Application to the determination of copper in beer. Analyst, 124, 1053-1059.

Bermejo, P., Peña, E., Fompedriña, D., Domínguez, R., Bermejo, A., Fraga, J. M., Cocho, J. A., (2001). Copper fractionation by SEC–HPLC and ETAAS: Study of breast milk and infant formulae when used in lactation of full-term newborn infants. Analyst, 126, 571-576.

Bonić, M., Tešević, V., Nikićević, N., Cvejić, J., Milosavljević, S., Vajs, V., Mandić, B., Urošević, I., Veličković, M., & Jovanić, S., (2013). The contents of heavy metals in Serbian old plum brandies. Journal of the Serbian Chemical Society, 78, 933-945.

Caldas, N. M., Filho, V. R. A., Neto, J. A. G., (2007). Direct and simultaneous determination of As, Cu and Pb in Brazilian cane spirits by graphite furnace AAS using tungsten permanent modifier with co-injection of Pd/Mg(NO3)2. Atomic Spectroscopy, 28, 189-194.

Caldas, N. M., Oliveira, S. R., Neto, J. A. G., (2009). Feasibility of internal standardization in the direct and simultaneous determination of As, Cu and Pb in sugar-cane spirits by graphite furnace atomic absorption spectrometry. Analytica Chimica Acta, 636, 1-5.

15

Catarino, S., Pimentel, I., Curvelo-Garcia, A. S., (2005). Determination of copper in wine by ETAAS using conventional and fast thermal programs: Validation of analytical method. Atomic Spectroscopy, 26, 73-78.

Chen, A.-H., Chen, S.-L., Wu, Y.-H., Zhang, C.-L., (2014). Direct and sequential determination of six metal elements in cooking wine by HR-CS-GFAAS. Advanced Materials Research, 1033-1034, 658-662.

Flores, C. R., Figueroa, J. A. L., Wrobel, K., & Wrobel, K., (2009). ICP-MS multi-element profiles and HPLC determination of furanic compounds in commercial tequila. European Food Research and Technology, 228, 951-958.

Frech, W., Lundberg, E., & Cedergren, A., (1985). Investigation of some methods used to reduce interference effects in graphite furnace atomic absorption spectrometry. Progress in Analytical Atomic Spectroscopy, 8, 257-370.

Galani-Nikolakaki, S., Kallithrakas-Kontos, N., & Katsanos, A. A., (2002). Trace element analysis of Cretan wines and wine products. Science of the Total Environment, 285, 155-163.

Huguet, M. E. R., (2004). Monitoring of Cd, Cr, Cu, Fe, Mn, Pb and Zn in fine Uruguayan wines by atomic absorption spectroscopy. Atomic Spectroscopy, 25, 177-184.

16

Jurado, J. M., Martín, M. J., Pablos, F., Moreda-Piñeiro, A., & Bermejo-Barrera, P. (2007). Direct determination of copper, lead and cadmium in aniseed spirits by electrothermal atomic absorption spectrometry. Food Chemistry, 101, 1296-1304.

Karadjova, I., Izgi, B., & Gücer, S., (2002). Fractionation and speciation of Cu, Zn and Fe in wine samples by atomic absorption spectrometry. Spectrochimica Acta Part B 57 581-590.

Kokkinofta, R. Petrakis, P.V., Mavromoustakos, T., & Theocharis, C. R., (2003). Authenticity of the traditional Cypriot spirit "zivania" on the basis of metal content using a combination of coupled plasma spectroscopy and statistical analysis. Journal of the Agricultural and Food Chemistry, 51, 6233-6239.

Kostić, D., Mitić, S., Miletić, G., Despotović. S., & Zarubica, A., (2010). The concentrations of Fe, Cu and Zn in selected wines from South-East Serbia. Journal of the Serbian Chemical Society, 75, 1701-1709.

Lachenmeier, D. W., Sarsh, B., & Rehm, J., (2009). The composition of alcohol products from markets of Lithuania and Hungary, and potential health consequences: Pilot Study. Alcohol and Alcoholism, 44, 93-102.

Lara, R., Cerutti, S., Salonia, J. A., Olsina, R. A., & Martinez, L. D., (2005). Trace element determination of Argentine wines using ETAAS and USN-ICP-OES, Food and Chemical Toxicology, 43, 293-297.

17

Miranda, K., Dionísio, A. G. G., Pereira-Filho, E. R., (2010). Copper determination in sugarcane spirits by fast sequential atomic absorption spectrometry using internal standardization. Microchemical Journal, 96, 88-101.

Mitić, S. S., Micić, R. J., & Simonović, R. M., (2009). Analytical application of food dye Sunset Yellow for the rapid kinetic determination of traces of copper(II) by spectrophotometry. Food Chemistry, 117, 461-465.

Moreno, I. M., González-Weller, D., Gutierrez, V., Marino, M., Cameán, A. M., González, A. G., & Hardisson, A., (2008). Determination of Al, Ba, Ca, Cu, Fe, K, Mg, Mn, Na, Sr and Zn in red wine samples by inductively coupled plasma optical emission spectroscopy: Evaluation of preliminary sample treatments. Microchemical Journal, 88, 56-61.

Murányi, Z., & Papp, L., (1998). Enological metal speciation analysis. Microchemical Journal, 60, 134-142.

Navarro-Alarcón, M., Velasco, C., Jodral, A., Terrés, C., Olalla, M., Lopez, H., & Lopez, M. C., (2007). Copper, zinc, calcium and magnesium content of alcoholic beverages and byproducts from Spain: nutritional supply. Food Additives and Contaminants, 24, 685-694.

NIST Spectral Database: http://www.nist.gov/pml/data/asd.cfm (accessed: June 14th, 2016).

Onianwa, P.C., Adeyemo, A. O., Idowu, O. E., & Ogabiela, E. E., (2001). Copper and zinc contents of Nigerian foods and estimates of the adult dietary intakes. Food Chemistry, 72, 8995.

18

Pyrzyńska, K., (2004). Analytical methods for the determination of trace metals in wine. Critical Reviews in Analytical Chemistry, 34, 69-83.

Raposo Jr., J. L., de Oliveira, A. P., Jones, B. T., & Neto, J. A. G., (2012). Internal standardization combined with dilute-and-shoot preparation of distilled alcoholic beverages for Cu determination by high-resolution continuum source flame atomic absorption spectrometry. Talanta, 92, 53-57.

Reche, R. V., Leite Neto, A. F., Da Silva, A. A., Galinaro, C. A., De Osti, R. Z., & Franco, D. W., (2007). Influence of type of distillation apparatus on chemical profiles of Brazilian cachaças. Journal of the Agricultural and Food Chemistry, 55, 6603-6608.

Resano, M., Flórez, M. R., & García-Ruiz, E., (2013). High-resolution continuum source atomic absorption spectrometry for the simultaneous or sequential monitoring of multiple lines. A critical review of current possibilities. Spectrochimica Acta Part B, 88, 85-97.

Rodríguez-Solana, R., Salgado, J. M., Domínguez, J. M., & Cortés, S., (2014). Assessment of minerals in aged grape marc distillates by FAAS/FAES and ICP-MS. Characterization and safety evaluation. Food Control, 35, 49-55.

Sadhra, S. S., Wheatley, A. D., & Cross, H. J., (2007). Dietary exposure to copper in the European Union and its assessment for EU regulatory risk assessment. Science of the Total Environment, 374, 223-234.

19

Soufleros, E. H., Natskoulis, P., & Mygdalia, A. S., (2005). Discrimination and risk assessment due to the volatile compounds and the inorganic elements present in the Greek marc distillates Tsipouro and Tsikoudia. Journal International des Sciences de la Vigne et du Vin, 39, 31-45.

Stafilov, T. & Karadjova, I., (2009). Atomic absorption spectrometry in wine analysis – A review. Macedonian Journal of Chemistry and Chemical Engineering, 28, 17-31.

Szymczycha-Madeja, A., Welna, M., Jamroz, P., Lesniewicz, A., & Pohl, P., (2015). Advances in assessing the elemental composition of distilled spirits using atomic spectrometry. TrAC –Trends in Analytical Chemistry, 64, 128-136.

Terol, A., Paredes, E., Maestre, S. E., Prats, S., & Todolí, J. L., (2011). Alcohol and metal determination in alcoholic beverages through high-temperature liquid-chromatography coupled to an inductively coupled plasma atomic emission spectrometer. Journal of Chromatography A, 1218, 3439-3446.

Welz, B., (2005). High-resolution continuum source AAS: the better way to perform atomic absorption spectrometry. Analytical and Bioanalytical Chemistry, 381, 69-71.

Welz, B., Becker-Ross, H., Florek, S., Heitmann, U., (2005). High-Resolution Continuum Source AAS: The Better Way to Do Atomic Absorption Spectrometry. Weinheim, Germany: Wiley-VCH Verlag GmbH & Co. KGaA.

20

World Health Organization (WHO), (2005). Copper in drinking water, Background document for development of WHO Guidelines for Drinking-water Quality, WHO/SDE/WSH/03.04/88.

Zaidel, A. N., Prokofjev, V. K., Raiskii, S. M., & Shreider, E. Ya., (1962). Tables of Spectral Lines. (in Russian). (1st ed.). Moscow: Gos. Izd. Fiz. Mat. Lit.

Zhong, W.-S., Ren, T., Zhao, L.-J., (2016). Determination of Pb (lead), Cd (cadmium), Cr (chromium), Cu (copper), and Ni (nickel) in Chinese tea with high-resolution continuum source graphite furnace atomic absorption spectrometry. Journal of Food and Drug Analysis, 24, 46-55.

21

Figure caption

Fig. 1. Pyrolysis and atomization curves recorded for 1 mg/L Cu in a mixture of 1:1 etanolwater, preserved with 0.144 mol/L HNO3 (a) with 5 µg Pd (as nitrate) modifier, (b) without modifier using normal furnace program, and (c) using fast furnace program without modifier, recorded at Tat=2200 oC and Tpyr=800 oC, respectively

Fig. 2. Pyrolysis and atomization curves recorded for a multi-element standard (0.4 mg/L Au, 1 mg/L Cu, 0.7 mg/L Cr, 0.25 mg/L Fe, 6.9 mg/L Na, 3 mg/L Os, 8 mg/L Rh) diluted with a 1:1 etanol-water mixture (acidity: 0.144 mol/L HNO3) without modifier using the fast furnace program

Fig. 3. Typical 3D transient spectra obtained for the vaporization and atomization of Au, Cr, Fe, Na, Os and Rh; (a) with Cu (Tpir=400 oC, Tat=2400 oC) and without Cu (Tpir=400 oC, Tat=2600 oC).

22

Table 1 Conventional and rapid (modified parameters in parentheses) graphite furnace heating program used without and with the chemical modifier

Step

Temperature

Ramp

Hold

Internal

(oC)

Time

Time

Furnace Gas

(oC/s)

(s)

Flow Rate (dm3/min)

Drying 1

80

6

50 (10)

2

Drying 2

90

3

20 (5)

2

Drying 3

110

5

10 (5)

2

Pyrolysis 1

350

50

20 (5)

2

Pyrolysis 2

800

300

10 (5)

2

Gas adaption

800

0

5

0

Atomization

2200a/2400 b

2500

5

0

2450

500

4

2

Cleaning a b

– single element determinations; – for multi-element determinations with the internal standard.

23

Table 2 Spectral lines of various elements covered by the optics of the ContrAA-700 spectrometer in the vicinity of the Cu I 249.2146 nm analytical line (width of the spectral window of the CCD: ±150 pm) Spectral line

Literature data on relative

(nm)

line strength Aa

S/GDa

b

Zn I 249.332

25

-

-

As I 249.291

25

5

40

Fe I 249.2644

3

-

-

Cr I 249.257

50

-

36

W I 249.2367

12

8

-

Os I 249.2367

50R

8R

-

Rh I 249.2299

100

10

30

Cu I 249.2146

200R

50

2000R

Fe I 249.1984

10

-

-

Sn I 249.178

12

5

-

Os I 249.169

-

-

290

Zn I 249.148

100

50

-

Cr I 249.135

30

1

19

Au I 249.120

10

5

-

Ni I 249.1187

20

-

-

Cd I 249.116

3

(2)

-

Fe I 249.1156

150R

10

13200

W I 249.1843

9

-

-

Os I 249.102

-

-

290

Na I 249.0733

3R

-

-

Ni I 249.0700

40

-

-

Fe I 249.0645

200R

10

18200

Note: values determined in arc (A), spark (S), or in Geissler discharge tube (GD); spectral lines chosen for the study of internal standardization are underlined. a

– Zaidel et al., (1962); b – NIST, (2015). 24

Table 3 The Cu content of distilled alcoholic beverages from Hungarian regions determined by means of various methods, and GFAAS recoveries for 5 mg/L Cu spikes

Sample No., Name/Type, City of Origin (alcohol degree in %)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Sárkányölő cuvee*, Szalafő (54) Peach, Komarno (45) Pear, Győrszentiván (47) C. Savignon*, Gencsapáti (47.5) Blackcurrant, Pannonhalma (40) Pear, Tényő (45) Strawberry - 2013, Gencsapáti (47.5) Mixed fruit, Kunsági (37) Strawberry – 2012, Gencsapáti (45) Mixed fruit, Győrújbarát (50) Plum, Almásfüzitő (50) Mulberry, Pázmándfalu (43) Merlot*, Zalaegerszeg (47) Mulberry, Naszály (44) Williams, Panyola (52) Peach, Győr (52) Peach, Győrszemere (45) Apricot, Nyergesújfalu (50) Peach, Gyirmót (49) Domestic*, Vasi Hegyhát (50.5)

Average Cu concentration ±SD (mg/L) _______________________________________________________ GFAAS TXRF _____________________________________________ Regul. GF prog. Fast GF prog. Fast GF prog. Fast GF prog. with int. std. without Pd + 5 µg Pd + 10 µg Pd. 5.57±0.09 n.d. 10.9±0.50 0.42±0.08 0.14±0.01 3.62±0.03 0.43±0.05 8.78±0.12 2.37±0.01 0.28±0.06 19.9±0.88 10.7±0.74 2.24±0.13 1.48±0.11 0.48±0.09 n.d. 4.74±0.05 n.d. 1.84±0.03 0.80±0.08

5.09±0.14 n.d. 7.87±0.19 0.45±0.10 0.10±0.02 3.04±0.06 0.33±0.04 7.27±0.04 2.22±0.06 0.25±0.04 16.7±0.07 8.93±0.18 2.13±0.06 1.20±0.08 0.45±0.06 n.d. 4.12±0.03 n.d. 1.71±0.12 0.78±0.13

25

5.62±0.05 n.d. 8.58±0.23 0.41±0.11 0.14±0.02 3.75±0.02 0.47±0.04 7.89±0.04 2.30±0.02 0.27±0.02 16.7±0.17 9.5±0.12 2.16±0.01 1.19±0.01 0.71±0.03 n.d. 4.01±0.04 n.d. 1.95±0.03 0.70±0.04

n.a. n.d. n.a. n.a. n.a. n.a. n.a. 10.1±2.8 n.a. n.a. n.a. 9.30±1.35 n.a. 1.41±0.47 n.a. n.a. 4.30±0.20 0.04±0.01 1.62±0.12 1.46±0.32

n.a. 0.06±0.01 12.4±0.65 0.79±0.06 0.08±0.02 n.a. 0.43±0.04 n.a. n.a. n.a. 23.7±1.2 n.a. 2.59±0.19 1.19±0.14 0.42±0.05 0.04±0.01 4.22±0.33 0.03±0.01 1.51±0.12 0.58±0.04

Recovery (%) GFAAS

110 100 94 94 103 83 95 88 110 106 119 97 95 91 108 99 107 100 112 96

21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39

Cherry – 2012, Sopron, (47) Pear - 2009, Hosszúpereszteg (48) Peach, Kőszeg (49) Cherry, Nyúl (47) Currant, Pápa (44) Plum, Székesfehérvár (50) Plum, Karcag (51) Marc, Öreghegy (40) Peach-honey, Szombathely (47) Apple, Bezenye (50) Mixed fruit, Újszilvás (50) Pear, Győrszentiván (52) Marc, Szőce (51) Marc, Kiskunfélegyháza (52) Apple, Budapest (43) Marc, Esztergom (52) Peach-honey, Koroncó (49) Plum, Békéscsaba (47) Apricot, Nyék (50)

4.51±0.15 21.6±0.85 10.1±0.40 8.04±0.31 0.11±0.01 3.49±0.10 3.49±0.02 0.46±0.10 0.31±0.04 1.31±0.03 0.38±0.01 1.16±0.06 6.37±0.45 4.15±0.04 0.19±0.01 10.8±0.10 2.07±0.39 2.75±0.07 3.57±0.14 * Note: – wine distillate, n.d. – not detected, n.a. – not analyzed.

4.35±0.15 24.3±0.28 11.7±0.18 9.49±0.07 0.14±0.01 4.41±0.32 4.07±0.14 0.38±0.08 0.34±0.02 1.61±0.02 0.46±0.01 1.19±0.04 6.80±0.18 4.65±0.06 0.28±0.11 11.4±0.40 2.74±0.18 2.23±0.06 3.05±0.13

26

4.55±0.06 20.9±0.12 12.7±0.07 7.33±0.05 0.17±0.02 3.94 ±0.08 3.76±0.08 0.38±0.18 0.40±0.03 1.62±0.02 0.52±0.01 1.43±0.18 6.74±0.10 4.72±0.08 0.19±0.02 9.33±0.38 2.43±0.21 2.60±0.06 3.72±0.11

3.80±0.16 26.0±0.59 13.9±2.97 8.07±2.10 0.15±0.01 4.42±0.03 4.04±0.01 0.41±0.07 0.35±0.01 1.59±0.05 0.38±0.07 1.53±0.20 4.89±0.43 4.69±0.35 0.26±0.03 7.90±0.69 2.27±0.62 2.17±0.15 2.92±0.20

n.a. 29.0±1.5 n.a. n.a. n.a. n.a. 4.01±0.24 n.a. n.a. n.a. n.a. 0.96±0.14 n.a. 3.68±0.32 n.a. 9.48±0.47 1.80±0.23 n.a. 2.41±0.24

115 117 110 102 100 102 97 99 99 88 95 106 112 92 94 89 101 107 103

Fig. 1.

27

Fig. 2.

28

Fig. 3.

29

Research highlights

•

Cu is quantitated by HR-CS-GFAAS via directly dispensing microliters of samples.

•

Alternative Cu line (249.2146 nm) & adjacent lines of internal standards were used.

•

Fast & regular GF-heating programs without & with Pd modifier were optimized.

•

Pyrolysis & atomization curves were recorded for Au, Cu, Cr, Fe, Na, Os and Rh.

•

External & internal calibration was used for analyzing ~40 distilled beverages.

30