Development of ultra high performance liquid chromatography methods using chemometric optimization methodology for quantitative analysis of ingredients in commercial energy drinks

Today, commercial energy drinks have been dramatically consumed by people, especially teenagers. For this reason, their quality, quality control, and routine analysis are very interesting in terms of human health and the food industry. In addition, the content of these energy drinks has changed day by day. This situation forces scientists and analysts working in the food industry to develop new analytical methods or update old analytical methods to correctly detect the content of commercial energy drinks. In this doctoral thesis, new UPLC methods were developed for the simultaneous quantitative analysis of ingredients, which can be detected by the photodiode arrays (PDA) system, in two different types of commercial energy drinks (commercial energy drink (I) and commercial energy drink (II)) using experimental design and optimization procedures. In the development of the UPLC methods with experimental design and optimization methodology, full factorial design (FFD) with three factors three levels was used. In the preliminary experiments for the analysis of caffeine (CAF) and potassium sorbate (PS) in the first commercial energy drink (I), the significant factors and their interactions, which affect on the chromatographic response, were found as column temperature, flow rate, and percentage of buffer in the mobile phase. In the next step, factors' levels were determined, and then the chromatograms of the 33 full factorial design matrix were recorded with the PDA detection at the wavelength of 284 nm to obtain ICRF values. A second-order equation was computed from the relationship between the design matrix and ICRF variables and then was solved to get optimal chromatographic conditions. For the UPLC analysis of the commercial energy drink (I), the optimal chromatographic conditions were found to be the column temperature of 58.9 °C, the percentage of the buffer system in the mobile phase of 59.3%, and the flow rate of 0.24 mL/min. Under the above conditions, the separation of CAF and PS in samples was carried out on the Waters Acquity ® BEH C18 column (100 mm x 2.1 mm i.d., 1.7 µm) using a mobile phase containing MeOH and 0.1 M H3PO4 buffer, pH 2.15 (59.3:40.7, v/v) in the presence of the 0.4 mL triethylamine (TEA)/L. As in the analysis of the first commercial energy drink (I), a new UPLC method was improved for the analysis of four compounds (niacin (B3), sodium benzoate (SB), PS, and CAF) in the second commercial energy drink (II). The same chromatographic factors (column temperature, percentage of buffer in the mobile phase, and flow rate) were optimized using the full factorial design with three factors at three levels for the UPLC method development for the quantitation of four compounds in the second commercial energy drink (II). In the quantitative estimation of the second commercial energy drink, optimal chromatographic conditions were determined as 52.7 °C for the column temperature, %24 for the percentage of the buffer content in the mobile phase and 0.25 mL/min for the flow rate. The chromatographic separation of B3, SB, PS, and CAF in energy drink sample (II) was performed by Waters Acquity ® BEH C18 column (100 mm x 2.1 mm i.d., 1.7 µm) using a mobile phase system containing 0.08 M (NH4)2CO3, H2O, and methanol (24: 24.8: 51.2, v/v/v) in the presence of 0.5 mL TEA /L. Under the above chromatographic conditions, the chromatograms of the full factorial design samples were obtained by using the PDA detection at the wavelength of 230 nm. ICRF values were calculated from the recorded chromatograms. Then, the second-order equation was calculated using the regression analysis between the ICRF and design matrix variables. After solving the second-order equation, optimal chromatographic conditions were obtained. Under the above-mentioned experimental UPLC conditions, the chromatograms of the calibration samples containing CAF and PS in the linear concentration range of 2.0-26.0 µg/mL and 4.0-28.0 µg/mL were recorded with the PDA detection at 272.4 and 262.9 nm, which correspond to maxima of wavelength, respectively. A linear regression equation for each compound was obtained using the relationship between concentration and peak areas with a high correlation coefficient (1.0 for both compounds). From the standard deviation of the intercept of the calibration curve, the limit of detection (LOD) and limit of quantification (LOQ) values were calculated as 0.18 µg/mL and 0.59 µg/mL for CAF and 0.20 µg/mL and 0.65 µg/mL for PS, respectively. Then, the optimized UPLC method was validated by analyzing test samples (independent test samples, inter-day and intra-day samples, and standard addition samples). Satisfactory percent recovery results were obtained (100.7% for CAF and 100.5% for PS). The results obtained from the validation studies revealed that the newly developed method is a reproducible, precise, accurate, selective, and applicable method. Finally, the developed method was applied to the analysis of the commercial energy drink (I) sample consisting of CAF and PS. The obtained results were compatible with the amount specified on the label of product. For the simultaneous analysis of B3, SB, PS and CAF in second energy drink sample (II), under the above-mentioned chromatographic UPLC conditions, the chromatograms of the calibration samples containing B3, SB, PS and CAF in the linear concentration range of 5.0-25.0 µg/mL, 4.0-20.0 µg/mL, 6.0-30.0 µg/mL, and 5.0-25.0 µg/mL were recorded with the PDA detection at 215.7, 223.9, 253.5, and 273.6 nm, which correspond to maxima of wavelength, respectively. Using the relationship between concentration and peak areas of the B3, SB, PS and CAF, linear regression equations with high correlation coefficients (0.9994 for B3, 0.9982 for SB, 0.9996 for PS and 0.9996 for CAF) were generated for each compound. The LOD and LOQ values of B3, SB, PS and CAF were calculated using the standard deviation and intercept of the calibration curve and were found to be 1.00 µg/mL and 3.34 µg/mL for B3, 1.37 µg/m L and 4.55 µg/mL for SB, 0.84 µg/mL and 2.80 µg/mL for PS,and 1.01 µg/mL and 3.37 µg/mL for CAF, respectively. Validation of the newly developed UPLC method for the simultaneous analysis of B3, SB, PS, and CAF was performed by using test samples (independent test samples, inter-day and intra-day samples, and standard addition samples). The percent recovery results for B3, SB, PS, and CAF were obtained as 101.2%, 99.9%, 99.2% and 99.6%, respectively. These results showed that this method is a selective, accurate, and reliable method for the simultaneous analysis of B3, SB, PS, and CAF. Finally, the developed method was applied to the simultaneous analysis of four compounds in commercial energy drink (II) sample and satisfactory results were obtained. A good agreement was observed between experimental analysis results and the claimed label of the commercial energy drink (II). This doctoral thesis revealed that the optimized and validated UPLC methods can be used as alternative methods for the quality control and routine analysis of the subject matter compounds in the analyzed energy drinks.