Simple and Fast Method for Iron Determination in White and Red

Jul 29, 2014 - ... St. Louis, MO) was prepared by dissolving an appropriate amount of this reagent ...... Pereira , E. R.; Soares , B. M.; Maciel , J...
0 downloads 0 Views 350KB Size
Article pubs.acs.org/JAFC

Simple and Fast Method for Iron Determination in White and Red Wines Using Dispersive Liquid−Liquid Microextraction and Ultraviolet−Visible Spectrophotometry Juliana V. Maciel,† Bruno M. Soares,† Jaime S. Mandlate,† Rochele S. Picoloto,‡ Cezar A. Bizzi,‡ Erico M. M. Flores,‡ and Fabio A. Duarte*,‡ †

Escola de Química e Alimentos, Universidade Federal do Rio Grande, Rio Grande, Rio Grande do Sul 96203-900, Brazil Departamento de Química, Universidade Federal de Santa Maria, Santa Maria, Rio Grande do Sul 97105-900, Brazil



ABSTRACT: This work reports the development of a method for Fe extraction in white and red wines using dispersive liquid− liquid microextraction (DLLME) and determination by ultraviolet−visible spectrophotometry. For optimization of the DLLME method, the following parameters were evaluated: type and volume of dispersive (1300 μL of acetonitrile) and extraction (80 μL of C2Cl4) solvents, pH (3.0), concentration of ammonium pyrrolidinedithiocarbamate (APDC, 500 μL of 1% m/v APDC solution), NaCl concentration (not added), and extraction time. The calibration curve was performed using the analyte addition method, and the limit of detection and relative standard deviation were 0.2 mg L−1 and below 7%, respectively. The accuracy was evaluated by comparison of results obtained after Fe determination by graphite furnace atomic absorption spectrometry, with agreement ranging from 94 to 105%. The proposed method was applied for Fe determination in white and red wines with concentrations ranging from 1.3 to 4.7 mg L−1. KEYWORDS: dispersive liquid−liquid microextraction, iron, wine, UV−vis spectrophotometry, food analysis, sample preparation, GFAAS



INTRODUCTION Wine is defined as a beverage obtained exclusively by alcoholic fermentation of grapes.1 With regard to its chemical composition, wine has a great concern considering nutritional and toxicological aspects, because it is a worldwide consumed beverage.2 Among the main components present in wine, some of them can be cited as volatile organic compounds (ethanol), non-volatile organic compounds (superior alcohols, sugars, organic acids/salts, and others) and major and trace elements.3 The presence of metals in wine can be originated from natural and/or anthropogenic sources.2 The range and level of inorganic compounds in wine depends upon several factors, such as the type of grape, soil characteristics, environmental conditions, and contamination during the manufacturing process.4 Iron is an essential element, which plays an important role in biological systems, participating in the maintenance of cellular homeostasis and several metabolic and fermentation processes, such as enzymatic activator, stabilizer, or as a functional component of proteins. Iron deficiency leads manly to anemia, causing changes in muscle metabolism and dysfunctions in the immunological system.5 Otherwise, excess of Fe cause hemochromatosis, characterized by pigmentation in the skin, pancreatic injury, such as diabetes, cirrhosis, and others.6 The presence of iron (ferric salts) in wine may cause cloudiness and precipitation, which modify organoleptic characteristics.7 In general, detection techniques, such as flame atomic absorption spectrometry (FAAS),7 graphite furnace atomic absorption spectrometry (GFAAS),8 and inductively coupled plasma−optical emission spectrometry (ICP−OES),9 have been applied for Fe determination in wine samples. Although © 2014 American Chemical Society

there are a few works reporting ultraviolet−visible (UV−vis) spectrophotometry for the determination of Fe in wine, it is a useful technique and stands out mainly by its availability, simplicity, and relatively low cost when compared to other spectrometric techniques already cited.10 However, the direct element determination at low concentrations by UV−vis is often difficult because of its insufficient sensitivity and selectivity. Besides, some wines (especially the red wines) show an intense color, which causes several interferences for UV−vis detection.11 Therefore, a prior extraction and preconcentration step is convenient, especially for matrix removal. Thus, many sample preparation methods, such as solid-phase extraction (SPE), 7 cloud point extraction (CPE),12,13 and liquid−liquid extraction (LLE),14,15 have been evaluated for subsequent element determination in wine. However, some of these methods demand excessive solvent consumption, high waste generation, and long extraction times. Recently, dispersive liquid−liquid microextraction (DLLME) was developed16 as an alternative for conventional LLE. This method shows some advantages, such as simplicity, quickness, low waste generation, and good enrichment factors.17 Initially, DLLME was employed for the extraction of organic analytes from aqueous samples.16 However, it has recently been applied for element extraction18 and selected chemical species.19 It is important to mention that, when DLLME is applied for Received: Revised: Accepted: Published: 8340

May 2, 2014 July 18, 2014 July 29, 2014 July 29, 2014 dx.doi.org/10.1021/jf5019774 | J. Agric. Food Chem. 2014, 62, 8340−8345

Journal of Agricultural and Food Chemistry

Article

2.3. Samples. Samples used in this work (two samples of white wine and five samples of red wine) were produced and purchased in Rio Grande do Sul State (Brazil). All samples were stored in glass bottles in the dark at room temperature and were named as WW1 and WW2 (white wine) and RW1, RW2, RW3, RW4, and RW5 (red wine). The alcohol content of samples ranged from 10 to 13.4%. Sample WW1 (1 mL of sample, pH adjustment and dilution up to 5 mL with water) was used for method optimization. 2.4. DLLME Optimization. Initially, experiments were performed to evaluate the wavelength with a maximum absorbance signal and less susceptible to interferences on Fe determination by UV−vis. The formation of the sedimented phase using all possible combinations between extraction and dispersive solvents was evaluated. Some parameters of DLLME, such as type of dispersive solvent (acetone, acetonitrile, ethanol, methanol, and THF), volume of dispersive solvent (100−1500 μL), type of extraction solvent (C6H4Cl2, CCl4, C6H5Cl, and C2Cl4), volume of extraction solvent (20−160 μL), pH (2−4), APDC concentration (0.1−3%, m/v), extraction time (0.5−5 min), and NaCl concentration (0−0.5 mol L−1), were evaluated. For these experiments, samples were spiked with a Fe(III) solution up to a final Fe concentration of 2 mg L−1, and results were shown as Fe recovery (%). Even considering the natural Fe concentration, samples were spiked to ensure the detectability of analyte and to allow for a suitable comparison of extraction conditions. 2.5. DLLME Procedure. Aliquots of 1 mL of wine and 500 μL of APDC solution were placed in a 15 mL glass tube with a conic bottom and diluted up to 5 mL with ultrapure water. The pH was adjusted with HCl or NaOH solutions. A mixture containing the dispersive and extraction solvent was quickly injected into the sample solution with aid of a 2.5 mL microsyringe. After injection, a cloudy solution was formed, which was centrifuged at 3000 rpm for 3 min, and small droplets of extraction solvent containing the complexed analyte [Fe(APDC)3] were deposited at the bottom of the glass tube. To reduce some interference during the measurements, the sedimented phase (extract) was washed with ultrapure water (3 mL and centrifugation at 3000 rpm for 3 min). The aqueous phase was removed, and extract was further diluted up to 3 mL with acetonitrile for subsequent Fe determination by UV−vis. All statistical calculations were performed using GraphPad InStat software (GraphPad InStat Software, Inc., version 3.06). A significance level of 95% was adopted for all comparisons.

extraction of elements (or their species), it is necessary to perform a chemical modification (addition of a complexant agent), which results in a complex with affinity by extraction solvent, which can be easily extracted.20 In this sense, the aim of this work was to develop a method for Fe extraction in white and red wine samples using DLLME to allow for the determination by UV−vis. Some parameters, such as type and volume of dispersive and extraction solvents, pH, complexant agent [ammonium pyrrolidinedithiocarbamate (APDC)] concentration, NaCl concentration, and extraction time, were exhaustively investigated. The optimized method was applied for Fe determination in commercial white and red wine samples. For accuracy evaluation, Fe was also determined by GFAAS in all samples after sample digestion.

2. EXPERIMENTAL SECTION 2.1. Instrumentation. Iron determination was performed in a Shimadzu UV−vis spectrophotometer (model UV-2550, Japan) with 1.0 cm quartz microcells. Iron was determined [as a complex of Fe(APDC)3] at 587 nm using peak height.18 For results comparison, a graphite furnace atomic absorption spectrometer (model AAnalyst 800, PerkinElmer, Waltham, MA) equipped with a Zeeman-effect background correction system and an auto sampler (model AS 800, PerkinElmer) were used for the analysis of wine samples previously digested with nitric acid. Transversely heated pyrolytically coated graphite tubes using standard platforms were used throughout. A Fe hollow cathode lamp (Hamamatsu Photonics, Japan) was (248.3 nm). The spectral bandpass was set at 0.2 nm. The main conditions of the heating program were optimized, and the follow temperatures were chosen: drying 1, ramp of 15 °C s−1 up to 90 °C and hold for 15 s; drying 2, ramp of 5 °C s−1 up to 130 °C and hold for 10 s; pyrolysis, ramp of 15 °C s−1 up to 1300 °C and hold for 20 s; atomization, ramp of 2000 °C s−1 up to 2400 °C and hold for 6 s; and clean out, ramp of 500 °C s−1 up to 2500 °C and hold for 3 s. Results were obtained using integrated absorbance (peak area), with 20 μL of reference solutions or sample extract and 5 μL of 1000 mg L−1 Pd solution (Merck, Germany), which were injected into the graphite tube. Argon (99.999%, White Martins, Brazil) was used as the purge gas. Wine samples were digested in a heating block with open vessels. About 2 mL of sample were transferred to the vessels and submitted to a step for alcohol removal (80 °C for 2 h). After cooling, 4 mL of concentrated HNO3 was added, followed by heating at 80 °C for 1 h and 120 °C for 1 h. After cooling, the digests were diluted with ultrapure water up to 50 mL for further Fe determination by GFAAS. A centrifuge (model 80-2B, Centribio, Brazil) with 15 mL calibrated centrifuge tubes was used for separation of phases after the DLLME procedure. The pH was adjusted using HCl or NaOH solutions with a Hanna pH meter (model pH21, Brazil). 2.2. Reagents. All reagents used in this work were of analytical grade. Ultrapure water was purified by a Direct-Q UV3 purification system (resistivity of 18.2 MΩ cm, Millipore, Billerica, MA). Stock reference solutions of 100 mg L−1 Fe were prepared by dissolving an appropriate amount of FeNH4(SO4)2·12H2O (Vetec, Brazil) in 1.0 mol L−1 HCl (Merck). Other reference solutions were prepared daily by serial dilution of stock solution in ultrapure water. The complexant agent (APDC, purity of 99%, Sigma-Aldrich, St. Louis, MO) was prepared by dissolving an appropriate amount of this reagent in ultrapure water. Dichlorobenzene (C6H4Cl2, d = 1.30 g mL−1), carbon tetrachloride (CCl4, d = 1.59 g mL−1), monochlorobenzene (C6H5Cl, d = 1.11 g mL−1), and tetrachloroethylene (C2Cl4, d = 1.62 g mL−1) were evaluated as extraction solvents and purchased from Vetec. Acetone, acetonitrile, ethanol, methanol, and tetrahydrofuran (THF) were evaluated as dispersive solvents and obtained from J.T.Baker (Center Valley, PA). Solutions of 0.1 mol L−1 HCl and 0.1 mol L−1 NaOH (both Merck) were used to adjust the pH of wine samples. All materials used in this work were previously cleaned by immersion in 20% (v/v) HNO3 (Merck) solution for 24 h and washed with ultrapure water.

3. RESULTS AND DISCUSSION Initially, experiments were performed to find the spectral region with maximum absorbance for the Fe(APDC)3 complex. In a previous study of our group,18 APDC was used for complexation of Fe(II) and Fe(III) species. It was observed that Fe(III) is instantaneously complexed by APDC, while Fe(II) is complexed by APDC and quickly oxidized to Fe(III), making it difficult to perform Fe speciation analysis.18 It was observed [using an aqueous solution containing Fe(III) and APDC] that Fe(APDC)3 complex shows one absorption band in UV region (wavelength of maximum absorbance at 354.0 nm) and two absorption bands in visible region (wavelengths of maximum absorbance at 499.0 and 587.0 nm). Although the most intense absorption band was in the UV region (data not shown), it was highly influenced by the absorption band of APDC, making the use of this wavelength difficult for analytical purposes. Thus, to minimize possible interferences, the wavelength of 587.0 nm was selected for Fe determination by UV−vis. 3.1. Effect of the Type of Extraction and Dispersive Solvents. The choice of dispersive and extraction solvents is essential to obtain suitable analyte recoveries using DLLME. The extraction solvent should have the ability to extract the analytes and present low solubility in water.21 Initially, several combinations between extraction and dispersive solvents were qualitatively evaluated (results not showed) for red and white 8341

dx.doi.org/10.1021/jf5019774 | J. Agric. Food Chem. 2014, 62, 8340−8345

Journal of Agricultural and Food Chemistry

Article

wines. The main parameters to choose the solvents that were used for subsequent experiments were taken into account, with the minimum amount of solids around the drop and the minimum undesirable effects (e.g., changes in baseline) during measurements by UV−vis. In this case, solvents with higher density than water, such as dichlorobenzene, carbon tetrachloride, monochlorobenzene, and tetrachloroethylene, were evaluated as the extraction solvent. These solvents were investigated using 50 μL of extraction solvent combined with 500 μL of acetonitrile as the dispersive solvent. As showed in Figure 1, the higher recovery (48%) was found using

Figure 2. Effect of the type of dispersive solvent on Fe recovery. Conditions: 5 mL of white wine sample (5 times diluted), addition of 240 μL of 25 mg L−1 Fe(III) solution, 500 μL of 1% APDC, 500 μL of dispersive solvent, 80 μL of extraction solvent (C2Cl4), and pH 3.3. The error bars represent the standard deviation (n = 5).

Figure 1. Effect of the type of extraction solvent on Fe recovery. Conditions: 5 mL of white wine sample (5 times diluted), addition of 240 μL of 25 mg L−1 Fe(III) solution, 500 μL of 1% APDC, 50 μL of extraction solvent, 500 μL of dispersive solvent (acetonitrile), and pH 3.3. The error bars represent the standard deviation (n = 5). Figure 3. Effect of the dispersive solvent volume (acetonitrile) on Fe recovery. Conditions: 5 mL of white wine sample (5 times diluted), addition of 240 μL of 25 mg L−1 Fe(III) solution, 500 μL of 1% APDC, 80 μL of extraction solvent (C2Cl4), and pH 3.3. The error bars represent the standard deviation (n = 5).

tetrachloroethylene, whereas the recoveries for other solvents were lower than 33%. Additionally, tetrachloroethylene presented the lowest relative standard deviation (RSD, about 3%), which was also considered for choosing this solvent for further experiments. For the selection of the dispersive solvent, the key factor is the miscibility of dispersive solvent into the extraction solvent and aqueous sample.22 The effect of dispersive solvent (acetone, acetonitrile, ethanol, methanol, and THF) on Fe extraction was evaluated using 500 μL of each solvent and 50 μL of tetrachloroethylene (as the extraction solvent). As can be seen in Figure 2, the highest recovery (about 48%) was obtained using acetonitrile, with RSD below 2%. Thus, acetonitrile was used as the dispersive solvent for subsequent experiments. 3.2. Effect of Volumes of Dispersive and Extraction Solvents. The volume of the dispersive solvent is an important parameter in DLLME because, at small volumes, the dispersion of the extraction solvent does not occur and the cloudy solution is not formed. On the other hand, the use of large volumes increases the solubility of the extraction solvent (and analyte) in the aqueous phase, decreasing the extraction efficiency.23 After dispersive (acetonitrile) and extraction (tetrachloroethylene) solvents have been chosen, experiments were performed to evaluate the volume of these solvents. The volume of acetonitrile ranged from 200 to 1500 μL in combination with 50 μL of C2Cl4 and 500 μL of 1% (m/v) APDC at pH 3.3. As can be seen in Figure 3, Fe recovery increased according to the increase of acetonitrile volume up to 1300 μL. A significant difference (Tukey−Kramer test) between 1300 and 1500 μL of acetonitrile was not observed,

and the condition using 1300 μL was selected for further experiments. After optimization of the volume of acetonitrile (1300 μL), the volume of tetrachloroethylene was evaluated from 20 and 160 μL. Results are shown in Figure 4, and a significant difference (Tukey−Kramer test) was not observed among 40 and 160 μL of CCl4. Considering the use of lower volumes of

Figure 4. Effect of the extraction solvent volume (C2Cl4) on Fe recovery. Conditions: 5 mL of white wine sample (5 times diluted), addition of 240 μL of 25 mg L−1 Fe(III) solution, 500 μL of 1% APDC, 1300 μL of dispersive solvent (acetonitrile), and pH 3.3. The error bars represent the standard deviation (n = 5). 8342

dx.doi.org/10.1021/jf5019774 | J. Agric. Food Chem. 2014, 62, 8340−8345

Journal of Agricultural and Food Chemistry

Article

tetrachloroethylene (from 20 to 60 μL), it was difficult to handle the extract because of the low volume of solvent. Thus, 80 μL was chosen for the subsequent experiments, which presented the lowest RSD values (about 2%) and also assured a suitable amount of extraction solvent. 3.3. Effect of pH. The effect of pH for element extraction using DLLME has an important role on the formation of the complex and its subsequent extraction. The influence of pH in the Fe recovery was evaluated from 2 to 4. Results are shown in Figure 5. Using pH 3, Fe recovery of about 80% was obtained

3.6. Effect of the Extraction Time. In general, the extraction time for DLLME is defined as the time between the injection of the mixture of dispersive and extraction solvents before the centrifugation.27 The extraction time was evaluated from 0.5 to 5 min, but results (not shown) did not presented significant influence (Tukey−Kramer test) on Fe recovery. Because of the large surface area between the extraction solvent and aqueous phase after the formation of a cloudy solution, the complex quickly diffuses into the extraction solvent.28 Therefore, the DLLME procedure was considered time-independent, which is an important advantage of the proposed method. 3.7. Analytical Performance. The optimized DLLME method was evaluated considering some parameters, such as linearity, limit of detection (LOD), limit of quantification (LOQ), accuracy, and precision. A complementary study for evaluating the matrix effect on Fe determination by UV−vis was also performed comparing the slopes of calibration curves, through the DLLME method, in aqueous reference solution and wine samples. When these slopes were compared, suppression in the analytical curve of about 16 and 26% was found for white and red wines, respectively. This suppression was also observed during method optimization, once recoveries were not better than 84%. Thus, to minimize this effect, the instrument calibration was carried out using a calibration curve with analyte addition. It is important to point out that the behavior (slope) for the same type of wine (white or red) was similar for all samples. In this sense, for Fe determination in white wine, sample WW1 was used for instrument calibration. Similarly, for red wine, sample RW1 was used for instrument calibration for further Fe determination. The calibration range was linear from 0.75 to 2.5 mg L−1 Fe(III) with coefficients of determination (R2) of 0.9985 and 0.9988, for white and red wines, respectively. It is important to mention that the effect of coexisting ions was also investigated. As expected, it was observed that transition metals, such as Cd2+, Cr3+, Mn2+, Ni2+, Pb2+, Sn2+, and Zn2+, with concentration below 1 mg L−1, did not affect Fe extraction/determination by UV−vis spectrophotometry. For major elements, such as Ca2+, K+, Mg2+, and Na+, the concentration limit is about 5000 mg L−1. It is important to mention that Co2+ and Cu2+ have a strong influence on Fe extraction/determination (concentration limit of about 0.4 mg L−1). The LOD and LOQ were estimated using a calibration curve with analyte addition, defined as 3 and 10 times the standard deviation of 10 measurements of blanks divided by the slope of the calibration curve.29 The LOD and LOQ were 0.2 and 0.8 mg L−1, which are significantly lower than the maximum Fe concentration established by Brazilian legislation. Considering that the proposed DLLME method is relatively fast and low cost, it can be an important alternative for Fe monitoring in a wine sample easily applied in routine analysis. Because of the lack of a certified reference material presenting certified values for Fe in wine samples, accuracy was evaluated by comparison to Fe determination by GFAAS. It is important to mention that Fe determination by GFAAS was performed using samples decomposed in a heating block with open vessels. Results are summarized in Table 1. By comparison to results obtained by GFAAS, the proposed method showed an agreement ranging from 93 to 105% for all samples, with RSD lower than 7%. The proposed method was also applied for Fe determination in commercial white and red wine samples (Table 1), and the concentration ranged from 1.3 to 4.7 mg L−1 and from 2.5 to 4.2 mg L−1, respectively. It is

Figure 5. Effect of pH on Fe recovery. Conditions: 5 mL of white wine sample (5 times diluted), addition of 240 μL of 25 mg L−1 Fe(III) solution, 500 μL of 1% APDC, 80 μL of extraction solvent (C2Cl4), and 1300 μL of dispersive solvent (acetonitrile). The error bars represent the standard deviation (n = 5).

(RSD of 3%). For pH values below 3, low recovery values could be attributed to the protonation of APDC (complexant agent). Similarly, low recoveries were obtained at pH values higher than 3, which can be attributed to the formation of Fe− hydroxyl complexes. It can directly affect the transference of analyte to the extraction solvent because the formation of the Fe(APDC) complex is avoided.24 Therefore, pH 3 was selected for further experiments. This pH (3) was in agreement with those found for Fe extraction using APDC as the complexant agent.25 3.4. Effect of the APDC Concentration. The effect of the APDC concentration was investigated using 500 μL of solution ranging from 0.1 to 3% (m/v). The maximum recoveries (about 80%) were found using 1% APDC solution, which remained constant up to 3%. Once significant differences (p > 0.05) among the values ranging from 1 to 3% were not observed, 1% APDC solution was selected for further studies. For the APDC concentration below of 1%, the lower recoveries can be explained because of the low APDC concentration, which was not enough for complete complexation of the analyte, especially because of the presence of other ions that could compete with the analyte by complexation with APDC. 3.5. Effect of the NaCl Addition. Generally, the addition of salt in the DLLME procedure increases the extraction efficiency because of the salting out effect that promotes the transfer of analyte to the organic phase by decreasing its solubility in the aqueous phase.26 To investigate the influence of ionic strength, experiments were performed by adding variable amounts of NaCl (from 0 to 25 g L−1). Other experimental conditions were kept constant, but at this range of the NaCl concentration, no statistical difference (Tukey− Kramer test) on Fe recovery was observed. Thereby, for further studies, NaCl was not added. 8343

dx.doi.org/10.1021/jf5019774 | J. Agric. Food Chem. 2014, 62, 8340−8345

Journal of Agricultural and Food Chemistry

Article

beverages by atomic absorption spectrometry. J. AOAC Int. 2000, 83, 189−195. (9) Lara, R.; Cerutti, S.; Salonia, J. A.; Olsina, R. A.; Martinez, L. D. Trace element determination of Argentine wines using ETAAS and USN−ICP−OES. Food Chem. Toxicol. 2005, 43, 293−297. (10) Abadi, M. D. M.; Ashraf, N.; Chamsaz, M.; Shemirani, F. An overview of liquid phase microextraction approaches combined with UV−vis spectrophotometry. Talanta 2012, 99, 1−12. (11) Andruch, V.; Kocúrová, L.; Balogh, J. S.; Škrlíková, J. Recent advances in coupling single-drop and dispersive liquid−liquid microextraction with UV−vis spectrophotometry and related detection techniques. Microchem. J. 2012, 102, 1−10. (12) Zheng, C.; Hongtao, Y.; Jia-Jian, H. Determination of trace iron in water and wine samples by spectrophotometry with cloud point extraction. Acta Chim. Sin. 2010, 68, 717−721. (13) Paleologos, E. K.; Giokas, D. L.; Tzouwara-Karayanni, S. M.; Karayannis, M. I. Micelle mediated methodology for the determination of free and bound iron in wines by flame atomic absorption spectrometry. Anal. Chim. Acta 2002, 458, 241−248. (14) Costa, R. C. C.; Araújo, A. N. Determination of Fe(III) and total Fe in wines by sequential injection analysis and flame atomic absorption spectrometry. Anal. Chim. Acta 2001, 438, 227−233. (15) Tasev, K.; Karadjova, I.; Arpadjan, S.; Cvetković, J.; Stafilov, T. Liquid/liquid extraction and column solid phase extraction procedures for iron species determination in wines. Food Control 2006, 17, 484− 488. (16) Rezaee, M.; Assadi, Y.; Hosseini, M. R. M.; Aghaee, E.; Ahmadia, F.; Berijani, S. Determination of organic compounds in water using dispersive liquid−liquid microextraction. J. Chromatogr. A 2006, 1116 (1−2), 1−9. (17) Yan, H.; Wang, H. Recent development and applications of dispersive liquid−liquid microextraction. J. Chromatogr. A 2013, 1295, 1−15. (18) Pereira, E. R.; Soares, B. M.; Maciel, J. V.; Caldas, S. S.; Andrade, C. F. F.; Primel, E. G.; Duarte, F. A. Development of a dispersive liquid−liquid microextraction method for iron extraction and preconcentration in water samples with different salinities. Anal. Methods 2013, 5, 2273−2280. (19) Soares, B. M.; Pereira, E. R.; Maciel, J. V.; Vieira, A. A.; Duarte, F. A. Assessment of dispersive liquid−liquid microextraction for the simultaneous extraction, preconcentration, and derivatization of Hg2+ and CH3Hg+ for further determination by GC−MS. J. Sep. Sci. 2013, 36, 3411−3418. (20) Skrlíková, J.; Andruch, V.; Balogh, I. S.; Kocúrová, L.; Nagy, L.; Bazel, Y. A novel, environmentally friendly dispersive liquid−liquid microextraction procedure for the determination of copper. Microchem. J. 2011, 99, 40−45. (21) Anthemidis, A. N.; Ioannou, K. G. Development of a sequential injection dispersive liquid−liquid microextraction system for electrothermal atomic absorption spectrometry by using a hydrophobic sorbent material: Determination of lead and cadmium in natural waters. Anal. Chim. Acta 2010, 668, 35−40. (22) Anthemidis, A. N.; Ioannou, K. G. On-line sequential injection dispersive liquid−liquid microextraction system for flame atomic absorption spectrometric determination of copper and lead in water samples. Talanta 2009, 79, 86−91. (23) Ma, J.; Lu, W.; Chen, L. Recent advances in dispersive liquid−liquid microextraction for organic compounds analysis in environmental water: A review. Curr. Anal. Chem. 2012, 8, 78−90. (24) Hulanick, A. Complexation reactions of dithiocarbamates. Talanta 1967, 14, 1371−1392. (25) Sitko, R.; Kocot, K.; Zawisza, B.; Feist, B.; Pytlakowska, K. Liquid-phase microextraction as an attractive tool for multielement trace analysis in combination with X-ray fluorescence spectrometry: An example of simultaneous determination of Fe, Co, Zn, Ga, Se and Pb in water samples. J. Anal. At. Spectrom. 2011, 26, 1979−1985. (26) Mirzaeia, M.; Behzadi, M.; Abadi, N. M.; Beizaei, A. Simultaneous separation/preconcentration of ultra-trace heavy metals in industrial wastewaters by dispersive liquid−liquid microextraction

Table 1. Iron Determination in White and Red Wines by UV−vis and GFAAS (Results in mg L−1 ± Standard Deviation; n = 5) sample white wine 1 white wine 2 red wine 1 red wine 2 red wine 3 red wine 4 red wine 5

UV−vis 1.3 4.7 4.2 3.3 3.9 2.5 3.0

± ± ± ± ± ± ±

0.1 0.1 0.2 0.1 0.2 0.1 0.2

GFAAS 1.4 5.0 4.4 3.2 4.1 2.5 3.0

± ± ± ± ± ± ±

0.1 0.2 0.2 0.1 0.3 0.1 0.2

worth highlighting that results are below the maximum limit of Fe in Brazilian legislation (maximum value of Fe in wine is 15 mg L−1), making the proposed extraction procedure appropriate for Fe control in wines. In summary, the proposed DLLME method combined with UV−vis spectrophotometry was successfully developed for total Fe determination in wine samples. This method was considered as environmentally friendly because it uses low volumes of organic solvents and showed the same suitable features, such as simplicity, quickness, and relatively low cost, when compared to AAS- or ICP-based techniques. In addition, the proposed method showed good results in terms of linearity, accuracy, and precision, providing evidence of the feasibility of UV−vis spectrophotometry as an alternative to routine quality control for Fe concentration in white and red wines.



AUTHOR INFORMATION

Corresponding Author

*Fax: +55-55-3220-9445. E-mail: [email protected]. Funding

The authors are grateful to the Conselho Nacional de ́ Desenvolvimento Cientifico e Tecnológico (CNPq), the ́ Superior Coordenaçaõ de Aperfeiçoamento de Pessoal de Nivel (CAPES), and the Fundaçaõ de Amparo à Pesquisa do Estado do Rio Grande do Sul (FAPERGS) for supporting this study. Notes

The authors declare no competing financial interest.



REFERENCES

(1) Pohl, P. What do metals tell us about wine? TrAC, Trends Anal. Chem. 2007, 26, 941−949. (2) Tariba, B. Metals in wineImpact on wine quality and health outcomes. Biol. Trace Elem. Res. 2011, 144, 143−156. (3) Grindlay, G.; Mora, J.; Gras, L.; de Loos-Vollebregt, M. T. C. Atomic spectrometry methods for wine analysis: A critical evaluation and discussion of recent applications. Anal. Chim. Acta 2011, 691, 18− 32. (4) Catarino, S.; Curvelo-Garcia, A. S.; Sousa, R. B. Measurements of contaminant elements of wines by inductively coupled plasma−mass spectrometry: A comparison of two calibration approaches. Talanta 2006, 70, 1073−1080. (5) Bleackley, M. R.; MacGillivray, R. T. A. Transition metal homeostasis: From yeast to human disease. BioMetals 2011, 24, 785− 809. (6) World Health Organization (WHO). Guidelines for DrinkingWater Quality; WHO: Geneva, Switzerland, 1996. (7) Pohl, P.; Prusisz, B. Application of tandem column solid phase extraction and flame atomic absorption spectrometry for the determination of inorganic and organically bound forms of iron in wine. Talanta 2009, 77, 1732−1738. (8) Olalla, M.; González, M. C.; Cabrera, C.; López, M. C. Optimized determination of iron in grape juice, wines, and other alcoholic 8344

dx.doi.org/10.1021/jf5019774 | J. Agric. Food Chem. 2014, 62, 8340−8345

Journal of Agricultural and Food Chemistry

Article

based on solidification of floating organic drop prior to determination by graphite furnace atomic absorption spectrometry. J. Hazard. Mater. 2011, 186, 1739−1743. (27) Wu, Q.; Wu, C.; Wang, C.; Lu, X.; Li, X.; Wang, Z. Sensitive determination of cadmium in water, beverage and cereal samples by a novel liquid-phase microextraction coupled with flame atomic absorption spectrometry. Anal. Methods 2011, 3, 210−216. (28) Tabrizi, A. B. Development of a dispersive liquid−liquid microextraction method for iron speciation and determination in different water samples. J. Hazard. Mater. 2010, 183, 688−693. (29) Ribani, M.; Bottoli, C. B. G.; Collins, C. H.; Jardim, I. C. S. F.; Melo, L. F. C. Validation for chromatographic and electrophoretic methods. Quim. Nova 2004, 27, 771−780.

8345

dx.doi.org/10.1021/jf5019774 | J. Agric. Food Chem. 2014, 62, 8340−8345