2.1. Chemicals and reagents Sodium acetate 97%, HPLC-grade chloroform, n-hexane, and acetonitrile were obtained from Merck (Darmstadt, Germany). Glacial acetic acid was procured from Sigma Chemical Co. (Darmstadt, Germany). Diethyl acetate from RCI Lifescan Ltd. was used. Deionised water (18.2 MΩ) used for chromatography processing was procured from a Barnstead Nanopure water purification system (Barnstead, USA). Sunset Yellow FCF (E110) was purchased from Rayner, Co. Ltd. (London, England).
2.2. Samples A total of 54 samples from six brands of orange jellies (research raw material) were purchased from three stores in Tangail city, Tangail, Bangladesh after getting verbal consent from the shopkeepers. All samples were considered valid based on their expiry dates. The weights of the samples were 200 gm to 500 gm. The collected samples were preserved in refrigerator at 4°C in the laboratory of the Department of Food Technology and Nutritional Science, Mawlana Bhashani Science and Technology University (MBSTU). Ethical approval was taken from the research cell of MBSTU.
2.3. Mobile phase The mobile phase was prepared using the modified method of PYLYPIW and GRETHER (2000). It consisted of a mixture of 3.0 mM acetate buffer (pH ~4.00) and HPLC-grade acetonitrile with a ratio of 17:3. Acetate buffer was prepared by mixing of 1.0 ml of glacial acetic acid and 1.0 g of sodium acetate trihydrate in 1.0 L deionized water and mixed well. The mixture was filtered with a filter membrane (Pore size 0.2 µm).
2.4. Preparation of standard solution Approximately100 mg of anhydrous sunset yellow was taken in a 25 ml volumetric flask. Ten ml 85% aqueous acetonitrile was added to the volumetric flask and was shaken well. Finally, 85% aqueous acetonitrile was added up to mark. The solution was filtered with syringe filter. The standard stock solution-1 was labeled as 4 mg/ml. Approximately 5 ml of stock solution-1 was taken in 50 ml volumetric flask and mobile phase was added up to the mark. The standard solution-2 was labeled as 0.4 mg/ml standard solution. From the stock solutions, working standard solutions 0.0, 1.0, 5.0, 10.0, 20.0, 40.0 µg/ml were prepared by dilution of aliquots. The solution was filtered through sample filters (Pore size 0.2 µm) prior to inject into the column.
2.5. Preparation of sample solution Approximately 1.0 gm of orange jelly was weighted accurately and placed in a conical flask and it was made to 10 ml by adding aqueous 85% acetonitrile solution and mixed well by vigorous shaking for 10 minutes. About 2.0 ml of the solution was filtered through sample filter (Pore size 0.2 µm) and the filtrate was then diluted 5 times and placed in an eppendorf tube. Finally, 20 µl was injected into the HPLC column. The concentration of injected sample solution was 20 µg/ml.
2.6. Sample solution preparation for Spiked/Recovery assay Approximately 2.0 gm of orange jelly and 1.0 mg of SY was weighed accurately and placed in a conical flask and 85% aqueous acetonitrile solution was added to it to make 20 ml solution and mixed well by vigorous shaking for 10 minutes. About 2.0 ml of the solution was filtered through sample filter (Pore size 0.2 µm) and the filtrate was then diluted 5 times and placed in an Eppendorf tube. Finally, 20 µl was injected into the HPLC column. The concentration of injected sample solution was 20 µg/ml.
2.7. Chromatographic analysis The chromatographic system consisted of a Shimadzu isocratic pump, a degasser, column, oven, a UV-Vis detector, and a LC Workstation Class-VP for data acquisition and analysis. Each of orange jelly samples of 1.0 g was diluted 1:10 with mobile phase and then the sample was again diluted 1:5 with mobile phase. After that the solution was transferred into dry eppendorf tube. The clear aqueous solution was filtered through a PTFE syringe filter. Then the solution was transferred to the dry HPLC vials. 20 µl of the sample was injected into the injector. For the chromatographic analysis, a Luna 5µ C18 (2) 100A column (250 × 4.6 mm) was used and the column temperature was set at 33 °C. The sunset yellow analysis was performed with isocratic solvent system using sodium acetate and acetic acid buffer (pH ~4.0)/acetonitrile- 17:3 with a flow rate of 1.0 ml/min.
2.8. SY identification and quantification Optimum absorption wavelength for SY color was evaluated before and using standard solutions with UV-spectrophotometer. The determined wavelength for the analysis of SY was 480 nm. Several runs were made to determine the retention time for the analysis. The retention time for this color was used for the identification of the color present in different brands of orange jellies. Quantification of the studied colors was done by external standard calibration. Five level analytical curves (0.00, 1.0, 5.0, 10.0, 20.0, 40.0 µg/ml) were used and the mean of 3 injections of each standard was used to represent each calibration point. By plotting analytes (y) against the concentration (µg/ml) of the color the peak areas were measured. To determine the slope, y-intercept and the correlation coefficients of the standards plots, least square linear regression analysis was used. Limit of detection (DL) and limit of quantification (QL) were determined by considering 3 and 10 times the signal to noise ratios respectively estimated by the regression lines as mentioned in the previous report (MACDOUGALL et al. 1980). For HPLC method validation the performance parameters, i.e., precision, linearity, the limit of detection, the limit of quantification, the expanded uncertainty were calculated. By spiking known amounts of the studied colors to the unprocessed sample and comparing the output with the same sample without spiking, recovery evaluations were carried out. Recoveries were calculated by differences of concentrations and were expressed as percentages.
2.9. Statistical analysis Each test was performed in triplicate. The descriptive analyses (means, median, standard errors coefficient of variation) were summarized. Data were expressed as mean±standard error (SE). One-way analysis of variance (ANOVA) was carried out using SPSS software version 20 at a significance level of 5%. The Least Significant Difference (LSD) test was used to detect differences in means.