Desarrollo de nuevos métodos de análisis cuantitativo de mercaptanos de alto impacto aromático en vino. Aplicación de novedosas estrategias analíticas de extracción, aislamiento y derivatización

Resumen

SUMMARY DEVELOPMENT OF NEW METHODS FOR THE QUANTITATIVE ANALYSIS OF MERCAPTANS WITH HIGH AROMATIC IMPACT IN WINE. APPLICATION OF NEW ANALYTICAL STRATEGIES OF EXTRACTION, ISOLATION AND DERIVATIZATION 1.-INTRODUCTION Volatile thiols such as 2-methyl-3-furanthiol -2M3F-, 2-furfurylthiol -FFT-, 4-mercapto-4-methyl-2-pentanone -4M4M2P-, 3-mercaptohexylacetate -MHA-, 3-mercaptohexanol -3MH- and benzylmercaptan -BM-, are very strong-smelling molecules that play an outstanding role on the aroma of numerous food such as coffee (Semmelroch y Grosch 1995), meat (Kerscher y Grosch 1998), onions (Granvogl, Christlbauer et al. 2004), green tea (Kumazawa, Kubota et al. 2005) and wine (Tominaga y Dubourdieu 2006). Although the contribution of polyfunctional mercaptans to the sensory characteristics of some wines is very important, there are just few reports on methods for their analytical determination, surely because of the difficulties found in their quantitative determination. A first problem is their poor detectability, and a second one derives from their instability. Described methods up to the beginning of the present work are mainly based on the selective complexing properties of the thiol function, particularly to certain forms of organic mercury. Some authors make use of a selective extraction of thiols from an organic solution with an aqueous solution of p-hydroxymercurybenzoate (p-OHHgB) (Darriet, Tominaga et al. 1993; Darriet, Tominaga et al. 1995; Tominaga, Murat et al. 1998; Tominaga y Dubourdieu 2006; Ferreira, Ortin et al. 2007); other authors use mercury covalently bonded to a certain resin or sorbent (Full y Schreier 1994; Full y Schreier 1995; Schneider, Kotseridis et al. 2003). The use of aqueous solutions of p-OHHgB in order to obtain a selective enrichment of mercaptans is an appealing idea but all these procedures are tedious, labor-intensive, difficult and slow and involve numerous phase-transferences of analytes, which not only makes them unattractive, but seriously compromises their analytical performance due to the instability and reactivity of these compounds. Any possible shortcut should limit the number and volume of phases to which mercaptans are transferred and in this sense it would be advantageous performing all the analytical operations in a single and small device. Because of the fact that p-OHHgB salt is strongly retained in some neutral polymeric resins, a SPE method based on the extraction-isolation of mercaptans by forming complex with p-OHHgB inside the cartridge has been developed. An important strategy to improve the detectability of these compounds makes use of the formation of a derivative of the thiol group with an adequate derivatization reagent. Up to this date, only 4-vinylpyridine (Hofmann, Schieberle et al. 1996; Ortín, Ferreira et al. 2005; Ortín 2006) has been presented as derivatization reagent. Throughout this PhD thesis, pentafluorobenzyl bromide (PFBBr) has been presented and widely studied as derivatization reagent for these compounds. Pentafluorobenzyl derivatives can be selectively and sensitively detected by GC-MS analysis with negative ion chemical ionization (NCI). The objective of this work has been the development of different derivatization strategies with PFBBr (including reaction matrix, reaction media and conditions) allowing the quantitative determination of mercaptans at ultratrace level in any complex matrix. 2.- EXPERIMENTAL 2.1. SOLID PHASE MICROEXTRACTION WITH IN FIBER DERIVATIZATION (SPME) (Mateo-Vivaracho, Ferreira et al. 2006). A 65 µm PDMS-DVB fiber and 20 mL sealed standard HS-SPME vials were used. The fiber was exposed 5 min to the headspace of 10 mL of tributylamine, 5 min to the headspace of 10 mL of a 200 mg/L PFBBr water solution with 10 % of acetone and finally, 10 min to the headspace of 10 mL of wine containing 0.05 g of EDTA. The extraction processes took place at 55 ºC and 250 rpm agitation. The fiber containing the derivatives was desorbed in splitless mode during 2 min at 275 ºC. Analytes were determined by GC-MS-NCI in a gas chromatograph Shimadzu QP-2010. 2.2. EXTRACTION-DERIVATIZATION IN LIQUID MEDIA (Mateo-Vivaracho, Ferreira et al. 2007) In a 13 mL screw capped test tube, 6 mL of wine were extracted with 1.5 mL of benzene containing 4 internal standards. In this extract the pentafluorobenzyl derivatives were formed by adding small amounts of PFBBr (100 mg/L) and a strong alkali: 1,8-diazabicyclo[5.4.0.]undec-7-ene (DBU). After washing the reaction product with 1 mL of a 0.5 M in HCl water:methanol solution (5:1), 20 µL of the extract were directly injected in the gas chromatograph. Derivatives were detected by GC-MS-NICI. 2.3. SOLID PHASE EXTRACTION (SPE) WITH IN SORBENT DERIVATIZATION (Mateo-Vivaracho, Ferreira et al. 2008). In a 50 mg cartridge of BondElut-ENV resins from Varian, 6 mL of wine were loaded. Then, 4 mL of a 0.2 M phosphate solution with 40 % MeOH and pH 7.7 were passed through the cartridge to remove the major volatile compounds from wine. Derivatives were formed in the cartridge by adding 1 mL of a 6.7 % DBU solution in water and 50 µL of a 2000 mg/L PFBBr solution in hexane. Reaction took place during 20 min at room temperature. 100 µL of a 2000 mg/L 1-thioglycerol solution in 6.7 % DBU were then added to remove the excess of reagent. The reaction media was, again, kept 20 min at room temperature. Before elution, the cartridge was washed with 4 mL of a 0.2 M phosphoric acid solution containing 40 % MeOH. The PFB-derivatives were eluted with 600 µL of hexane with 25 % of diethyl-ether. 8 µL of the extract were directly injected in the GC-MS-NICI system in splitless mode. 2.4. SELECTIVE ISOLATION IN A MICRO SOLID PHASE EXTRACTION CARTRIDGE CONTAINING A MERCURY SALT (publication under review). 20 mg of LiChrolut-EN resins were packed in 1 mL standard reservoirs to form a small solid phase extraction cartridge. The cartridges were conditioned and rinsed with 1 mL of a 0.1 M TRIS solution, pH 7.2. Then, 1 mL of a 2 mM p-OHHgB solution in 0.1 M TRIS buffered at pH 7.2, is poured into the cartridge. Once the p-OHHgB is fixed in the bed, 100 mL of wine are percolated through the cartridge. After this, the cartridge is rinsed with 5 mL of an aqueous solution containing 40 % methanol (v/v) and 5 g/L of tartaric acid buffered at pH 3. Non polar impurities were then removed from the cartridge by rinsing with 6 mL of pentane. Before elution, the bed was dried with a N2 stream. Complexes formed in the cartridge were eluted with 600 µL of dichloromethane 100 mM in 1,4-dithioerythritol. The extract was washed with 5 mL water and, after concentration, was transferred to a 2 mL vial containing a small amount of anhydrous Na2SO4. 2 µL of this extract were injected and analyzed by GC-MS-EI in a Shimadzu QP-2010 instrument. 2.5. QUANTIFICATION OF POLYFUNCTIONAL MERCAPTANS IN WINES BY ISOLATION AND IN-SORBENT PENTAFLUOROBENZYL ALKYLATION OF THESE COMPOUNDS. STUDY OF THE QUANTITATIVE RESULTS AND CONCLUSIONS (publication under preparation). The method described in (Mateo-Vivaracho, Ferreira et al. 2008) has been improved in the control and long term performance and has been applied to the analysis of 111 wines from different parts of the world. The results have been studied and some groups of wines have been established attending to their content in polyfunctional mercaptans. Furthermore, a sensory study in which different concentration levels of polyfunctional mercaptans have been added to different model wines has been carried out. 3.- RESULTS 3.1. SPME. A fully automated method to analyze FFT and MHA was developed; these are two of the target compounds in the aroma of numerous products. Although this approach is appealing and suggests that automated in SPME fiber derivatizations may be a competitive alternative for the analysis of these types of analytes, the restrictions and limitations of the approach suggest that more studies are still necessary to understand the reaction mechanisms and the behaviour of the fiber, since it was not possible to obtain satisfactory results for all the analytes. Method repeatability, expressed as relative standard deviation, was between 10 and 20 % for high and low concentration levels, respectively. These values were satisfactory at ng/L levels. Detection and quantification limits were 0.3 and 1 ng/L, respectively, for FFT, and 3 and 10 ng/L for MHA. The linear range of the method included little more than a magnitude order but fortunately, covered the expected concentrations of these analytes in wine. Calibration was based on the use of different internal standards using fortified wines. 3.2. IN LIQUID MEDIA REACTION. The amount of derivative finally formed in the media was found to differ in different solvents or solvent systems. The presence of a strong alkali was extremely important to get the derivatives in a reaction media with a small amount of PFBBr. The best results were obtained with DBU, an organic alkali that could not be substituted by an inorganic strong base. The method allowed us to simultaneously determine FFT, 4M4M2P, MHA and 3MH. Inconsistent results were obtained for 2M3F. Detection limits were 0.5 (FFT), 0.1 (4M4M2P), 0.6 (MHA) and 7 (3MH) ng/L, all of them below the olfactory thresholds. As for the repeatability of the method (10 %menor RSD menor17 %) and the linearity (0.98 menor R2 menor 0.999), they could be considered satisfactory. The linear range was greater of 2 magnitude orders. The slopes of standard addition curves for different wines were very similar. Nevertheless, the application of this method was problematic in the midterm, due to the stress suffered by the chromatographic system as a result of the presence of relatively small amounts of the underivatized reagent. 3.3. SPE. A fast method for the determination of aroma-powerful polyfunctional thiols at ng/L was developed. Carrying the reaction in the cartridge makes it possible to use simultaneously water and non-polar reagents, to avoid large volumes of toxic solvent, and to eliminate the excess of reagent. The method makes it possible to simultaneously determine 2M3F, FFT, 4M4M2P as its methoxime, MHA, 3MH and BM. Absolute limits of detection were 1.0, 0.5-1, 0.6, 3-5, 8-10 and 0.5-1 ng/L respectively. Repeatability (1 %menor RSD menor 25 %) and linearity (0.998 menorR2 menor0.9999) were satisfactory. Problems with matrix effects were solved by the use of internal standards. Advantages of this method are that it makes it possible to stabilize the matrix and to eliminate the excess of reagent, which in the midterm guarantees a greater stability and robustness of the system. 3.4. SELECTIVE ISOLATION WITH p-OHHgB. LiChrolut-EN resins were particularly efficient at retaining p-OHHgB at neutral pH, and the beds containing the Hg-salt can efficiently work as affinity-microcolumns to mercaptans, retaining some of the advantages of these kinds of polymeric sorbents. At the natural acid pH of wines, it was demonstrated that the breakthrough volume of mercaptans was higher than 100 mL for a micro-SPE cartridge containing only 20 mg of sorbent. The versatility of this technique has been demonstrated both concentrating light mercaptans present in aqueous solutions and developing a specific isolation method for polyfunctional mercaptans in wine. As a demonstration, the method has been applied to the determination of key polyfunctional mercaptans in wine, reaching detection limits in the ng/L level and acceptable repeatability and linearity (0.990 menor R2 menor0.999). 3.5. STUDY OF THE CONCENTRATION RESULTS OBTAINED WITH THE ANALYSIS OF 111 SAMPLES OF DIFFERENT WINES. An adequate analytical method to quantify polyfunctional mercaptans in wine has been developed and, for the first time, results of the concentrations of 6 polyfunctional mercaptans (2M3F, FFT, 4M4M2P, MHA, 3MH and BM) in 111 wines from different kinds and geographical origins are shown. Signals obtained (relative areas) were interpolated in the calibration curves and an exhaustive data analysis with all the results was carried out. Results are in satisfactory agreement with those previously reported and corroborate previous hypothesis about the sensory relevance of these compounds in some wines based on GC-olfactometric and sensory observations. The study of the quantitative results makes it possible to establish different groups or aromatic profiles according to the levels of polyfunctional mercaptans. Furthermore, the study has made it possible to define the natural ranges of occurrence for these compounds revealing that levels can vary by most than 2 magnitude orders in all cases except 2M3F. The sensory study corroborates the importance and contribution of these compounds to wine aroma even when they are at extremely low levels which adds practical value to the development of analytical methods and the results make it possible to conclude that even if the concentration of these compounds is most often very low, their contribution to the aroma of wine is most often highly significant or even determinant in nearly every kind of wine. 4.- CONCLUSIONS Apart from a competitive strategy for the fast and simple qualitative isolation of mercaptans from liquid matrixes, the use of a micro-SPE cartridge containing p-OHHgB, can also be successfully applied to the development of quantitative procedures. Furthermore, and leaving aside Hg strategies, the use of PFBBr as derivatization reagent makes it possible the formation of derivatives that can be selectively and sensitively detected by means of GC-MS-NICI. According to this, a robust, simple and straightforward method, based on solid-phase extraction and in-sorbent pentafluorobenzyl alkylation has been developed. The method makes it possible to reduce the sample used in the analysis by factors as high as 80 in comparison with previous reports. The method has been applied to quantify 6 polyfunctional mercaptans in a relatively large sample of wines and quantitative and sensory results reveal that the analysis of polyfunctional mercaptans in wine is critical because their presence has a deep impact on the aroma of wine and that even trace amounts of these compounds can completely modify the aroma of wines. 5.- REFERENCES Darriet, P., T. Tominaga, et al. (1993). "Evidence of the Presence of a 4-Mercapto-4-Methylpentan-2-One Precursor in Vitis-Vinifera Sauvignon Blanc Grape Variety." Comptes Rendus de L'Academie des Sciences Serie III-Sciences de La Vie-Life Sciences 316(11): 1332-1335. Darriet, P., T. Tominaga, et al. (1995). "Identification of a Powerful Aromatic Component of Vitis vinifera L. var Sauvignon Wines: 4-Mercapto-4-methylpentan-2-one." Flavour and Fragrance Journal 10: 385-392. Ferreira, V., N. Ortin, et al. (2007). "Optimization of a procedure for the selective isolation of some powerful aroma thiols: Development and validation of a quantitative method for their determination in wine." Journal of Chromatography A 1143(1-2): 190-198. Full, G. and P. Schreier (1994). "[Covalent chromatography: a valuable aid in analysis of thiol aroma compounds in the trace concentration range.] Kovalente Chromatographie: Ein wertvolles Hilfsmittel fuer die Aromastoffanalytik von Thiolen im Spurenbereich." Lebensmittelchemie 48: 1-4. Full, G. and P. Schreier (1995). Covalent chromatography as efficient ''clean-up'' technique in the analysis of 8-mercapto-p-menthan-3-one. Bioflavour 95 Symposium, Dijon, France. Hofmann, T., P. Schieberle, et al. (1996). "Model studies on the oxidative stability of odor-active thiols occurring in food flavors." Journal of Agricultural and Food Chemistry 44(1): 251-255. Mateo-Vivaracho, L., J. Cacho, et al. (2007). "Quantitative determination of wine polyfunctional mercaptans at nanogram per liter level by gas chromatography-negative ion mass spectrometric analysis of their pentafluorobenzyl derivatives." Journal of Chromatography A 1146(2): 242-250. Mateo-Vivaracho, L., J. Cacho, et al. (2008). "Improved solid-phase extraction procedure for the isolation and in-sorbent pentafluorobenzyl alkylation of polyfunctional mercaptans Optimized procedure and analytical applications." Journal of Chromatography A 1185(1): 9-18. Mateo-Vivaracho, L., V. Ferreira, et al. (2006). "Automated analysis of 2-methyl-3-furanthiol and 3-mercaptohexyl acetate at ng L-1 level by headspace solid-phase microextracion with on-fibre derivatisation and gas chromatography-negative chemical ionization mass spectrometric determination." Journal of Chromatography A 1121(1): 1-9. Ortín, N. (2006). Caracterización química del aroma de vinos de alta calidad. Contribución al análisis y caracterización de importantes aromas tiólicos del vino. Zaragoza, University of Zaragoza. Ph Doctor. Ortín, N., V. Ferreira, et al. (2005). Thiol derivatization in wine with 4-vinylpyridine. 11 Jornadas de Análisis Instrumental (JAI), Barcelona (Spain). Schneider, R., Y. Kotseridis, et al. (2003). "Quantitative determination of sulfur-containing wine odorants at sub parts per billion levels. 2. Development and application of a stable isotope dilution assay." Journal of Agricultural and Food Chemistry 51(11): 3243-8. Tominaga, T. and D. Dubourdieu (2006). "A novel method for quantification of 2-methyl-3-furanthiol and 2-furanmethanethiol in wines made from Vitis vinifera grape varieties." Journal of Agricultural and Food Chemistry 54(1): 29-33. Tominaga, T., M. L. Murat, et al. (1998). "Development of a Method for Analyzing the Volatile Thiols Involved in the Characteristic Aroma of Wines Made from Vitis-Vinifera L. CV Sauvignon Blanc." Journal of Agricultural and Food Chemistry 46(3): 1044-1048.