The Determination of Salicylic Acid in Wines Using a Ligand-Exchange Reaction

Advanced technologies 3(1) (2014), 11-15

THE DETERMINATION OF SALICYLIC ACID IN WINES USING A LIGAND-EXCHANGE REACTION

Aleksandra Pavlović1*, Snežana Mitić1, Snežana Tošić1, Ružica Micić2, Ivana Rašić1, Milan Mitić1, Vojkan Miljković3 (ORIGINAL SCIENTIFIC PAPER) UDC 547.587.11+663.2:543.48 1University of Niš, Faculty of Sciences and Mathematics, Department of Chemistry, Niš, Serbia 2University of Priština, Faculty of Sciences and Mathematics, Kosovska Mitrovica, Serbia 3University of Niš, Faculty of Technology, Leskovac, Serbia

The analytical method based on a ligand-exchange reaction is proposed for the determination of salicylic acid (SA) in wines. The reaction was followed spec- trophotometrically by monitoring the rate of appearance of the manganese(II)- salicylate complex in alkaline medium at 370 nm. The initial-rate method is adopted for constructing the calibration curve, which was found to be linear Keywords: salicylic acid; kinetic-spectropho- over the concentration range of 0.69-4.14 µg mL-1. The optimized conditions tometry; solid-phase extraction; wine; validation -1 yielded a theoretical detection limit of 0.11 µg mL based on the 3-3S0 crite- -1 rion and quantification limit of 0.33 µg mL based on the 10S0 criterion. The results obtained by the proposed method were compared with those obtained by HPLC using UV-DAD detection and showed a good agreement.

Introduction

The analysis of wine additives is of considerable co- fluorimetric [6,11] and mass spectrometry [4,16] de- mmercial importance. Among different additives, antise- tection system has been used to a lesser extent. Most ptics and antioxidants are commonly used in the wine fluorimetric methods proposed have lower detection li- treatment, and their addition is regulated. Salicylic acid mits than photometric methods, but these methods have is a powerful antioxidant that helps in lowering choleste- not been widely used in practice because the species rol and prevents blood clots in clogged arteries. In large that can be detected are limited [6,11]. Also, some HPLC quantities it is a toxic, but in lower quantities it is used as methods for SA determination using postcolumn deri- a food preservative and antiseptic. vatization reaction with teribium(III) [6]. LC-MS is a very For some people with salicylate sensitivity, chronic urti- expensive technique not widely used in routine laborato- caria patients may react adversely to some food additives ries in the wine industry. Spectrophotometry is the tech- [1] and even these small doses can be harmful [2,3]. Be- nique of choice even today due to its inherent simplicity. cause of this, it is important to develop a better unde- It is frequently used in the laboratories of the developing rstanding of the role of this substance and the adequate countries to overcome a variety of analytical problems. levels of intake for good health. Also, developing a sensitive and rapid analytical method for the determination of Experimental SA is a key aspect of analytical chemistry. The large number of batch and automated published Apparatus methods for the determination of SA is indicative of the The reaction rate was monitored spectrophotometri- great interest in the determination of this compound, but cally. The absorbance of the solution was measured at also of the problem encountered with its determination the wavelength of 370 nm. The readings were performed with respect to simplicity, selectivity, sensitivity, rapidity, on a Perkin-Elmer Lambda 15 UV/Vis spectrophotometer etc. The development of a simple, rapid to be performed, connected to a thermostatic bath (20.00-60.00±0.02 ºC). reliable, selective method for the determination of SA, A model 1200 Agilent Technologies was used for HPLC as an alternative to chromatographic techniques, would analysis. A J. T. Baker model SPE-12 with the vacuum be highly advantageous. The analysis of free phenolic pump was used for the solid phase extraction. The so- acids in wines has been extensively studied by liquid lutions were thermostated at 22.00±0.02 ºC before the chromatographic methods [4-16]. Diode array detection beginning of the reaction. has been extensively used for the development of liquid Reagents chromatography methods [5,8,9,10,12,14,15], whereas Acetonitrile (LC gradient grade) was used from J.T.

*Author address: Aleksandra Pavlović, University of Niš, Faculty of Sciences and Mathematics, Department of Chemistry, Višegradska 33, P.O.Box 224, 18000 Niš, Serbia E-mail: [email protected] The manuscript received: March, 24, 2014. Paper accepted: April, 03, 2014.

11 Advanced technologies 3(1) (2014), 11-15

Baker and o-phosphoric acid 85 % (analytical reagent tion at different concentrations of each of the reactants grade) was from Merck® (KGaA, Darmstadt, Germany) was obtained by measuring the slope of the linear part respectively. The 1.0×10-3 mol L-1 stock solution of sali- of the kinetic curves to the absorbance-time. The calibra- cylic acid (Merck®) was prepared in ethanol (95 %, V/V). tion graph was constructed by plotting the slope of the The salicylic acid solution was stored at 4 ºC. The 1.0 mol linear part of the kinetic curve, slope=dA/dt, the versus -1 -1 L sodium hydroxide solution (Merck®) was prepared in concentration of SA (cSA, μg mL ). deionised water. The 1.0×10-3 mol L-1 1-nitroso-2-naph- Chromatographic conditions thol solution (Merck®) was prepared by dissolving a Salicylic acid was detected and quantified on a 250 known amount in ethanol (95 %, V/V). The manganese(II) x 4.6 mm Eclipse XDB-C18 (5μm) analytical column ope- solution (1.8×10-3 mol L-1) was prepared by dissolving rating at room temperature. The mobile phase was a mix MnCl2×6H2O (Merck®) in water. The ionic strength of the of phosphoric acid-acetonitrile-water, 2:400:600 (by vol). reaction mixture was kept constant at 0.1 by adding the The eluate was monitored at 237 nm. The injection of appropriate amount of NaCl solution (1.0 mol L-1). Analytical the samples (10 μL) was performed by using an autosa- grade chemicals and deionised water with 0.05 μS cm-1 mpler. The flow rate was 1 mL min-1. conductivity (MicroMed high purity water system, TKA Analysis of white wine samples Wasseraufbereitungssysteme GmbH) were used for the White wine samples were diluted (1+1) with deio- preparation of all solutions. nized water and directly injected into the chromatograph- Wines ic system. For kinetic application, white wine samples Six white (vintage 2007) and six red (vintage 2007) were also directly analyzed following the procedure for commercial wines produced in Serbia (Table 1) in 200 the kinetic measurements. mL bottles were opened and analyzed within 24 h. In the Analysis of red wine samples case that samples required storage, they were kept at 4 When the both methods were applied to the red wine °C in a refrigerator. samples, a solid-phase extraction was necessary to re- All analyzed wines were from the West Morava move the coloured compounds and decrease the inter- vineyard area. The wines were obtained from the most ference of the matrix. The sample preparation scheme is widespread white grape cultivars (Smederevka, Riesling, given in Table 2. Chardonnay and Sauvignon Blanc) and red cultivars (Ca- bernet Sauvignon, Vranac, Kratošija, Merlot and Game). Table 2. Set up parameters for red wine samples preparation

Table 1. Wines and grape varieties of the analyzed wines

Results and discussion

Mechanism of the reaction SA shows the complexing ability with Mn(II) [17]. The complex agrees with the empiric formula Mn2(HSal)4(H2O)4. Manganese(II) salicylate complex is more stable than that formed with R(NO)OH. The aW1-W6, white wines; W7-W12, red wines reaction moves to the right (ligand-exchange reaction) and SA was determined by monitoring the rate of appe- Kinetic-spectrophotometric procedure arance of the manganese(II)-1-nitroso-2-naphthol com- The reaction was carried out in the following way. In a plex at 370 nm. special four-compartment vessel, the solution of 1-nitro- 2Mn[R(NO)O]2•2H2O + 4SA → [Mn2(HSal)4(H2O)4] + so-2-naphthol was placed in the first, salicylic acid and 4R(NO)OH sodium hydroxide in the second, manganese(II) in the Optimization of chemical variables third, and electrolyte for the ionic strength and ethanol In order to achieve the best sensitivity, the working (95 %, V/V) (total volume 10 mL) in the fourth compart- conditions needed to be optimized. ment. The vessel was thermostated at 22.00±0.02 ºC. The The effect of the concentration of sodium hydroxide content was mixed well and then immediately transferred was studied in the range of 0.02-0.10 mol L-1. The rea- to the spectrophotometric cell. The change in absorbance ction rate was increased with increasing the concentra- was recorded at 370 nm as a function of time every 30 s tion of NaOH up to 0.06 mol L-1; beyond this concentra- for the first 5 min of the reaction. The rate of the reaction, the rate of the reaction was decreased. For further

12 Advanced technologies 3(1) (2014), 11-15

-1 work the concentration of 0.06 mol L was used. rate=kcNaOHcSA...... (2) The dependence of the reaction rate on the conce- k- constant proportional to the rate constant of the reac- ntration of 1-nitroso-2-naphthol was studied in the range tion of 0.1-1.0×10-5 mol L-1. The reaction rate increased with The limit of the detection was determined as the increasing the 1-nitroso-2-naphthol concentration from signal-to-noise ratio (3.3σ/S) and it was 0.11 µg mL-1 0.1-0.6×10-5 mol L-1 and became constant at 0.6×10-5 and the limit of quantification (10σ/S ) was 0.33 µg mL-1, mol L-1. Thus, the concentration of 0.8×10-5 mol L-1 was where σ is the standrad deviation of the blank measure chosen as the optimum concentration. (standard deviation of y-intercept of regression equation) The influence of the concentration of Mn(II) on the rate and is the slope of the calibration curve [19-21]. of the reaction was examined in the range of 0.5-3.6×10-5 The precision and accuracy of the above system mol L-1. The reaction rate increased with increasing the were studied by performing the experiment five times for concentration of Mn(II) from 0.5-1.8×10-5 mol L-1 and different concentrations of salicylic acid. The results of became constant at 1.8×10-5 mol L-1. Thus the concen- accuracy and precision of the recommended procedure tration of 2.3×10-5 mol L-1 in the final solution was used are given in Table 4. throughout the experiment. The effect of the temperature on the reaction rate Table 4. The accuracy and precision of the determination of was studied in the range of 18-28 °C. The rate for dif- salicylic acid ferent concentrations of SA at each temperature was calculated and utilized for plotting the calibration curve. At the temperature > 22 °C, the linear dynamic range of the determination decreased. It was found that the calibration graph obtained at 22 °C possessed good linearity (r = 0.9989) and it is recommended that the determination can be carried a out at 22 °C. Mean and standard deviation of five determinations at 95 % confidence interval, brelative standard deviation, caccuracy of the method Validation of the proposed method The least squares' equation [18] ( y=bx+a, where b To assess the selectivity of the method, the interfer- and a are its slope and intercept, respectively) for the ence of some species was studied. The tolerance limits calibration graph and the correlation coefficient (r) for the de- (expressed as w/w ratio) for the species studied on the -1 termination of salicylic acid in the interval 0.69 to 4.14 µg mL determination of 2.07 µg mL-1 of SA are given in Table 5. -5 -1 under the optimal reaction conditions (cR(NO)OH = 0.8×10 mol L , -1 -5 -1 -1 cNaOH= 0.06 mol L ,cMn2+ = 2.3×10 mol L , cNaCl= 0.1 mol L , Table 5. The effect of foreign species on the determina- t = 22.00±0.02 °C) were calculated: tion of 2.07 µg mL-1 salicylic acid

2 slope×10 =0.29649×cSA+0.63702 r = 0.9997...... (1) where slope is the slope of the linear part of the kinetic curve to the absorbance-time plot (slope=dA/dt=ε·l·dc/dt, Beer`s law) and cSA is the salicylic acid concentration expressed in µg mL-1. Quantitative parameters of the analysis are given in Table 3.

Table 3. Quantitative parameters of the analysis

a Interference coefficient, I = (cºSA – cSA) / cºSA cºSA and cSA are measured concentrations of salicylic acid without and with the interfering substance

Applicability of the method The proposed method was applied for the determi- The following kinetic equation for the reaction was de- nation of salicylic acid in white and red wines using the duced on the basis of the graphic correlations obtained. direct calibration curve. They were treated as described

13 Advanced technologies 3(1) (2014), 11-15

in the Experimental section. As it can be seen in Table statistical analysis of the results (Table 6) showed that 6, the results obtained by this method are in accordance calculated F- and t- values at 95 % confidence levels are with the HPLC method. The results of the proposed less than the theoretical ones, confirming no significant method were statistically compared with those of the off- differences between the performance of the proposed icial method using a point hypothesis test [22,23]. The and the HPLC method.

Table 6 The determination of salicylic acid in wine samples by the kinetic and HPLC method

aData are based on the average obtained from five determinations bRelative standard deviation c Theoretical F-value ( ν1=4, ν2=4) and t-value (ν=8) at 95 % confidence level are 6.39 and 2.306, respectively

Conclusion [2] A. Korolkovas, J. H. Burckhalter, Essential of Medicinal Chemistry, John Wiley & Sons, New York, 1988, p.193. Despite the great number of methods for the analysis [3] H. Chi-Tang, Y. L. Chang, H. Mou-Tuan, Phenolic of SA described in literature, the proposed kinetic-spec- Compounds in Food and their Effects on Health I. Analysis, Occurrence, and Chemistry, American Chemical Society, trophotometric method for the determination of salicylic Washington, 1992, p. 35. acid in wine samples reported in this paper is simple, ra- [4] G. L. La Torre, M. Saitta, F.Vilasi, T. Pellicano, G. Dugo, pid, inexpensive, and thus very appropriate for the rou- Direct determination of phenolic copmounds in Sicilian tine quality control analyses of SA in laboratories in the wines by liquid chromatograohy with PDA an MS detection, wine industry. Statistical comparison of the results with Food Chemistry, 94(4) (2006) 640-650. the HPLC method showed a good agreement and indi- [5] M. Calull, R. M. Marce, G. Sanchez, F. Borull, cates no significant difference in accuracy and precision. Determination of additives in wine by high-performance So, the detection limit of 0.11 µg mL-1 can be achieved, liquid chromatography, Journal of Chromatography A, much better than those provided by fluorimetric [11] or 607(2) (1992) 339-347. [6] M. A. Rodriguez-Delgado, S. Malovana, J. P. Perez, T. even by HPLC [5] methods. Also, the precision of the Borges, F. J. Garcia-Montelongo, Separation of phenolic proposed method of analysis was found to be good compounds by high-performance liquid chromatography (RSD < 5 %). with absorbance and fluorimetric detection, Journal of Chromatography A, 912(2) (2001) 249-257. Acknowledgement [7] M. A. Rodriguez-Delgado, G. Gonzalez-Hernandez, J. E. Conde-Gonzalet, J. P. Perez-Trujillo, Principal component This research was supported by grant number 172047 analysis of the polyphenol content in young red wines, from the Ministry of Education and Sciences, the Repu- Food Chemistry, 78(4) (2002) 523-532. blic of Serbia. The authors are grateful for the financial [8] P. Ho, T. A. Hogg, M. C. M. Silva, Application of a liquid chromatographic method for the determination of phenolic support provided by this Ministry. compounds and furans in fortified wines, Food Chemistry, 64(1) (1999) 115-122. References [9] Q. Zhang, H. Cui, A. Myint, M. Lian, L. Liu, Sensitive determination of phenolic compounds using high- [1] C. E. H. Grattan, Aspirin sensitivity and urticaria, Clinical performance liquid chromatography with cerium(IV)- and Experimental Dermatology, 28(2) (2003) 123-127. rhodamine 6G-phenolic compund chemiluminiscence

14 Advanced technologies 3(1) (2014), 11-15

detection, Journal of Chromatography A, 1095(1-2) [16] M. N. Bravo, S. Silva, A. V. Coelho, L. Vilas-Boas, M. R. (2005) 94-101. Bronze, Analysis of phenolic compounds in Muscatel [10] R. Sladkovsky, P. Solich, M. Urbanek, High-performance wines produced in Portugal, Analytica Chimica Acta, liquid chromatography determination of phenolic 563(1-2) (2006) 84-92. components in wine using off-line isotachophoretic pre- [17] R. A. Lal, A. Lemtur, S. Choudhury, M. Chakrabarty, treatment, Journal of Chromatography A, 1040(2) (2004) D. Basumatary, M. K. Singh, S. Bhaumik, A. K. 179-184. De, A. Kumar, Synthesis and crystal structure of [11] R. C. Rodriguez-Diaz, M. P. Aguilar-Caballos, A. Gomez- [Mn2(H2Sal)2(HSal)2(H2O)4]. First example of the Hens, Determination of some hydroxybenzoic acids and reductive synthesis of a binuclear manganese(I) salicylate catechins in white wine samples by liquid chromatography complex, Transition Metal Chemistry, 31(4) (2006) 423- with luminescence detection, Journal of Separation 428. Science, 29(18) (2006) 2772-2779. [18] J. N. Miller, Basic Statistical Methods for Analytical [12] R. J. Robins, S. R. Bean, Development of a quantitative Chemistry. Part 2. Calibration and Regression Methods, high-performance liquid chromatography-photodiode Analyst, 116 (1991) 3-14. array detection measurement system for phenolic acids, [19] D. Perez-Bendito, M. Silva, Kinetic Methods in Analytical Journal of Chromatography, 1038(1-2) (2004) 97-105. Chemistry, Ellis Horwood, Chichester, 1988, p. 251. [13] R. C. Minussi, M. Rossi, L. Bologna, L. Cordi, D. Rotilio, [20] H. A. Mottola, Kinetic Aspects of Analytical Chemistry, G.M. Pastore, N. Duran, Phenolic compounds and total John Wiley & Sons, New York, 1988, p. 40. antioxidant potential of commercial wines, Food Chemistry, [21] V. Thomsen, D. Schatzlein, D. Mercuro, Limits of Detection 82(3) (2003), 409-416. in Spectroscopy, Spectroscopy, 18(12) (2003) 112-114. [14] V. F. Samanidou, C. V. Antoniou, I. N. Papadoyannis, [22] C. Hartmann, J. Smeyers-Verbeke, W. Penninckx, Y. V. Gradient RP-HPLC determination of free phenolic Heyden, P. Vankeerberghein, D. L. Massart, Reappraisal acids in wines and wine vinegar samples after SPE, of Hypothesis Testing for Method Validation: Detection with photodoide array identification, Journal of Liquid of Systematic Error by Comparing the Means of Two Chromatography and Related Technology, 24(14) (2001) Methods or of Two Laboratories, Analytical Chemistry, 2161-2176. 67(24) (1995) 4491- 4499 [15] Z. Kerem, B. Bravdo, O. Shoseyov, Y. Tugendhaft, Rapid [23] D. A. Skoog, D. M. West, F. J. Holler, Fundamentals liquid-chromatography-ultraviolet determination of of Analytical Chemistry, Saunders College Publishing, organic acids and phenolic compounds in red wines and Philadelphia, 1996, p. 42. must, Journal of Chromatography A, 1052(1-2) (2004) 211-215.

Izvod ODREĐIVANJE SALICILNE KISELINE U VINU NA OSNOVU REAKCIJE IZMENE LIGANADA

Aleksandra Pavlović1, Snežana Mitić1, Snežana Tošić1, Ružica Micić2, Ivana Rašić1, Milan Mitić1, Vojkan Miljković3 (ORIGINALNI NAUČNI RAD) UDC 547.587.11+663.2:543.48 1Univerzitet u Nišu, Prirodno-matematički fakultet, Departman za hemiju, Niš, Srbija 2Univerzitet u Prištini, Prirodno-matematički fakultet, Kosovska Mitrovica, Srbija 3Univerzitet u Nišu, Tehnološki fakultet, Leskovac, Srbija U radu je predložena metoda za određivanje salicilne kiseline (SA) u vinu koja se zasniva na reakciji izmene liganada. Reakcija je praćena spektrofotometrijski, merenjem brzine obrazovanja kompleksa između mangana i salicilne kiseline u baznoj sredini na 370 nm. Korišćenjem metode početnih brzina kostruisana je kalibraciona prava, koja je linearna u opsegu koncentracija 0,69-4,14 µg mL-1. Optimizovanjem uslova metode, dobijene su vrednosti za granicu detekcije ( 10S0 Ključne reči: salicilna kiselina, kinetička- -1 -1 spektrofotometrija, ekstrakcija u čvrstoj fazi, kriterijum) i granicu određivanja ( 3,3S0 kriterijum) od 0,11 µg mL i 0,33 µg mL , respektivno. Razvijena metoda je primenjena za određivanje salicilne kiseline vino, validacija u uzorcima vina. Dobijeni rezultati su poređeni sa rezultatima dobijenim HPLC hromatografijom, pri čemu je pokazano da postoji dobro slaganje između dva niza rezultata.