32-37 Melania.cdr 1. Introduction C o n s u m i n g f o o d s a n d d r i n k s containing benzene for long term, even though only in a low level can cause damage to blood cells, such as anemia and leukemia, immune system disorders, and cancer (Anonym, 2011). Since the discovery of benzene in food products in the early 1990s, a concern about the formation of benzene in food products has been increasing. The presence of benzene in food products can be generated from loss of benzene from the packaging material, water for production process which contains benzene, and the reaction between sodium benzoate and ascorbic acid (Barshick et al., 1995; Aprea et al., 2008). American Beverage Association (ABA) recommends several steps that could be conducted by manufacturers to prevent the formation of benzene in beverages products, one of them is an accelerated testing. Accelerated test can be carried out by c o n d i t i o n i n g p r o d u c t s a t m i n i m u m temperatures of 40 and 60 °C for 24 hours or longer, depending on the product formulation (Anonym, 2006). Accelerated testing is expected the benzene formation reaction. H i g h P e r f o r m a n c e L i q u i d Chromatography (HPLC) is one of several methods that can be used to analyze benzene in various samples (Khan, 2006). Benzene was successfully analyzed using reversed- phase HPLC column by some previous researchers (Gardner and Lawrence, 1993; Zoccolillo et al., 2001). The chromatographic system gave linear response for benzene standards over a concentration range of 0- JURNAL FARMASI SAINS DAN KOMUNITAS, Mei 2014, hlm. 32-37 Vol. 11 No. 1 ISSN : 1693-5683 ANALYTICAL METHOD VALIDATION OF BENZENE USING HIGH PERFORMANCE LIQUID CHROMATOGRAPHY IN BEVERAGE CONTAINING SODIUM BENZOATE AND ASCORBIC ACID 1 2 2 MELANIA PERWITASARI , ENDANG LUKITANINGSIH , SUDIBYO MARTONO 1 Faculty of Pharmacy, Sanata Dharma University 2 Faculty of Pharmacy, Gadjah Mada University Email correspondence: melania@usd.ac.id Abstract: Several countries reported discovering benzene in beverages containing benzoic acid and ascorbic acid. Benzoic acid decarboxylation by ascorbic acid will form benzene. American Beverage Association (ABA) recommended the use of accelerated testing to test benzene in beverages. High Performance Liquid Chromatography (HPLC) is one of several methods to analyze benzene in a wide variety of samples, but there is no much information provided regarding the validation of analysis method of benzene. Therefore, developing analysis method of benzene and its validation becomes a current need. The HPLC system consists of Hitachi L- 2130 pump, sample injector with 20 µL sample loop, and UV detector L-2420 operating at 205 ® nm. The analytical column is a LiChrosorb Phenomenex RP-18 (250 x 4 mm, 10 µm, 100 Ǻ), the mobile phase is acetonitrile:aquabidest (60:40 v/v) and pumped at a flow rate of 0,8 mL/min. Benzene separated from the matrix and follows the validation requirements. The developed analytical method showed that resolution was 8.37, r = 0.995 with LOD and LOQ 6.52 ppb and 19.75 ppb, with a precise of ≤ 11% and recovery of 80-110%. Accelerated testing indicated that benzene levels increased with increasing of the temperature. Beverages containing 400 mg/mL of o ascorbic acid and benzoic acid formed benzene which was detected as 699.38 ppb at 25 C, o o 799.61 ppb at 40 C, and 808.94 ppb at 60 C in 48 hours. In conclusion, the method was fully o validated and can be utilized to analyze benzene in beverages with the accelerated testing at 60 C in 48 hours, so that benefits the producers and consumers in the end. Keywords : benzene, HPLC, validation method, ascorbic acid, benzoic acid 2 100 nM (r > 0.98), with a retention time was approximately 8 minutes, LOD was approximately 1 nM in solution contain benzoic acid, ascorbic acid and transition metal (Gardner and Lawrence, 1993). Benzene was analyzed in gasoline and the validation of analytical method was not been reported (Zoccolillo et al., 2001). The information regarding the development and validation of analytical methods of benzene in beverage using HPLC is not fully being reported furthermore the differences of sample and column will need different analytical method. It is a need to develop analytical method of benzene in simulation solution of drink products containing sodium benzoate and ascorbic acid, and further to validate it. 2. Material and Methods 2.1. Materials Simulated solution containing sodium benzoate (Multi Kimia Raya, Indonesia), ascorbic acid (DSM, USA), and bottled water ® (R , Indonesia). 2.2. Chemicals Reference standard of benzene had purity 97% (Merck, Germany) and stored refrigerated in a capped container. All solvents used (acetonitrile and methanol) were of analytical grade (Merck, Germany). Aquabidest (Ikapharmindo Putramas, Indonesia) was used for mobile phase. 2.3. Sample Preparation Beverage or simulation solution was prepared by dissolved 400 mg ascorbic acid and 400 mg sodium benzoate in 1L of bottled water. Solution represents beverage products in the market (Aprea et al., 2008). The solution were filtered through a 0.45 µm membrane filter (Phenex, NY). 2.4. Calibration 5.70 µL benzene was placed in 50 mL volumetric flask and dissolved in acetonitril. Beginning with this solution, calibration levels were prepared by a serial dilution with beverage. The range of concentrations injected was 200-1100 ppb. PERWITASARI, LUKITANINGSIH, MARTONO Jurnal Farmasi Sains dan Komunitas 33 2.5. Analytical Method M e t h o d d e v e l o p m e n t a n d quantification studies were performed on a LaChrom Elite HPLC system (Merck- Hitachi, Tokyo, Japan), equipped with L- 2420 UV-detector, a sample injector with a 20µL sample loop, and L-2130 pump. An optimum separation of benzene in sample was achieved on Phenomenex LiChrosorb® RP-18 column (250 x 4 mm, 10 μm, 100 Ǻ) and mobile phase comprising a 60:40 mixture of acetonitril and aquabidest. Flow rate and injected sample volume were adjusted to 0,8 mL/min and 20 µL, respectively. Detection was performed at 205 nm. 2.6. Method Validation The HPLC method was validated for accuracy, precision, selectivity, linearity, limit of quantification and detection regarding to USP (Anonym, 2009). Selectivity were determined by optimized the wave length at 254 nm and 205 nm, composition mixture and flow rate of mobile phase (Anonym, 2005). Limit of quantification (LOQ) and limit of detection (LOD) were determined by serial dilution of standard solutions (20-100 ppb) containing benzene and calibration curves were generated by linear regression based on peak height. Accuracy and precision was confirmed by spiking placebo with three concentrations of the quantified standard compounds. 2.7. Accelerated testing 50 mL simulation solution in scoth duran bottle were subjected to the following accelerated testing conditions: 25, 40 and o 60 C for 48 hours. PERWITASARI, LUKITANINGSIH, MARTONO34 Jurnal Farmasi Sains dan Komunitas 3. Results and discussion To select working wavelength of benzene analysis were carefully investigated. Using 25% aqueous acetonitrile (Gardner and Lawrence, 1993) and 100 ppm benzene, a typical chromatogram obtained for benzene is shown in Figure 1. This figure shows chromatogram of benzene at 254 and 205 nm. These two wave length were preferred because benzene has absorption in UV area around 250 and 205 nm (Suzuki, 1967; Zoccolillo et al., 2001). The detector response at 205 nm was greater 74 times than at 254 nm, it shown in Table 1. Wavelength at 205 nm was chosen as working wavelength. The HPLC assay had to be optimized in order to give satisfactory results. Concerning the mobile phase, out of a number of different composition mixture of acetonitrile : aquabidest (60:40, 75:25, 80:20 v/v) and flow rate (0.6; 0.8; 1.0 mL/minutes) tested, the best result were obtained with 60:40 v/v PERWITASARI, LUKITANINGSIH, MARTONO Jurnal Farmasi Sains dan Komunitas 35 of composition mixture and 0.8 mL/minutes of flow rate. Using this optimization, 100 ppb of benzene has the best separated and eluted from the column at 6.27 minutes (Figure 2), ascorbic acid and sodium benzoate had the same retention time around 2.3 minutes. The result of resolution was 8.37. To confirm that an analytical method is suitable for its intended use, it has to be validated. In the present case these investigations were performed according to ICH guidelines (Anonym, 2005). Data presented in Table 2 indicated the linearity of the assay within the tested range (200-1100 ppb), combined with sensitivity and accuracy. The latter was confirmed by spiking placebo at three concentration levels with standard compounds (Table 3). The linearity of benzene obtained from sample preparation was studied by evaluating the calibration curve at ten level of concentrations. The result shown a good linearity with correlation coefficient (r) > 0.98. See Figure 3 for calibration curve. The limit of detection (LOD) of the method was 6.52 ppb and the limit of quantification (LOQ) was 19.75 ppb (Table 2). Accuracy and precision of test results determining levels of benzene in the sample by HPLC, shown at Table 3, represented that PERWITASARI, LUKITANINGSIH, MARTONO36 Jurnal Farmasi Sains dan Komunitas Figure 4. Chromatogram of benzene production in beverages contain sodium benzoate and ascorbic o acid (400 mg/L) at 25 (A), 40 (B) and 60 (C) C storage temperature for 48 hours. Column: LiChrosorb® Phenomenex RP-18 (250 x 4 mm, 10 µm, 100 Ǻ); mobile phase CH CN:H O (60:40 v/v); 3 2 λ 205 nm; flow rate 0.8 mL/min; sample 20 µL. the method used was accurate and precise. The accuracy and precision of three concentration (200, 500, 1100 ppb) of benzene were meet the requirement, 80- 110% recovery for the accuracy and RSD (Relative Standard Deviation) < 11% for the precision. The sample which was stored for 48 o hours at temperature of 25, 40, and 60 C detected benzene (Figure 4) by 699.38 ± 96.30 ppb (RSD 13.8%), 799.61 ± 142.69 ppb (RSD 17.8%), 808.93 ± 65.21 (RSD 8.1%), respectively (Table 4). It showed that higher the storage temperature, higher benzene which was formed. Levels of benzene obtained in this experiment was higher than some previous studies. Aprea et al. (2008) detected 118 ppb benzene in the simulation solution with levels of ascorbic acid and sodium benzoate 400 mg/mL which o were stored in a temperature 45 C for 24 hours. A study conducted by Nyman et al. (2010), stated that cranberry juice containing 0.04% ascorbic acid and benzoate formed 1.3 ppb of benzene that storage at room temperature (25 ° C). In addition, storage at 40 ° C and 60 ° C for 24 hours detected the existence of benzene in a row at 2.2 and 5.8 ppb. Higher benzene level was generated with the increasing of storage temperature. 4. Conclusions HPLC with UV detector has been successfully developed and validated for benzene in beverages. Evaluation of analytical method parameters including selectivity, linearity, accuracy, precision and sensitivity showed acceptable result. The developed method can also be used to quantitative analysis of benzene in beverages samples available in market with accelerated testing. 5. Acknowledgments The authors wish to express their appreciation to Faculty of Pharmacy, Gadjah Mada University and Research and Testing Laboratory, Gadjah Mada University (LPPT- UGM) for the facilities was provided in conducting this study. Reference Anonym, 2005, International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use. Validation of Analytical Procedures: Text and Methodology Q2(R1).1-13. Anonym, 2006, American Beverage Association Guidance Document to Mitigate the Potential for Benzen Formation in Beverage. Access f r o m , access date: 20 August 2011. Anonym, 2009, United State Pharmacopeia, XXXII Edition, Rockville, United States of America, 733. Anonym, 2011, Material Safety Data Sheet: Benzen. A c c e s s f r o m , access date: 11 October 2011. Aprea, E., Biasioli, F., Carlin, S., Mark, T.D., and Gasperi, F., 2008, Monitoring Benzen Formation from Benzoate in Model Systems b y P r o t o n T r a n s f e r R e a c t i o n - M a s s Spectrometry, Int. J. Mass Spectrom, 275: 117- 121. Barshick S.A., Smith S.M., Buchanan M.V., Guerin M.R., 1995, Determination of Benzen Content in Food Using A Novel Blender Purge and Trap GC/MS Method, J. Food Compos. Anal., 8:244-257. Gardner L.K. and Lawrence G.D., 1993, Benzen Production from Decarboxylation of Benzoic Acid in The Presence of Ascorbic Acid and A Transition-Metal Catalyst.J. Agric. Food. Chem., 41:693-5. Khan, H.A., 2006, A Concise Review of Chromatographic Methods for The Analysis of Benzen and Its Metabolites, Croat.Chem. Acta., 79:169-175. Nyman, P.J., Wamer, W.G., Begley, T.H., Diachenko, G.W., and Perfetti G.A., 2010, Evaluation of Accelerated UV and Thermal Testing for Benzen Formation in Beverages Containing Benzoate and Ascorbic Acid, J. Food Sci, Food Chem., 75 (3):263-267. Suzuki, H., 1967, Electronic Absorption Spectra and Geometry of Organic Molecules, Academic Press, New York and London, 196-198. Zoccolillo, L., Alessandrelli, M., and Felli, M., 2001, Simultaneous Determination of Benzen and Total Aromatic Fraction of Gasoline by HPLC- DAD, Chromatographia, 54:659-663. www.ameribev.org/files/aba_benzen.pdf http://www.sciencelab.com/msds.php? msdsId=9927339 PERWITASARI, LUKITANINGSIH, MARTONO Jurnal Farmasi Sains dan Komunitas 37 Page 1 Page 2 Page 3 Page 4 Page 5 Page 6