SUPPLEMENTARY MATERIAL Ultrasonic extraction of anthocyanin from Clitoria ternatea flowers using response surface methodology Fui Chin Chong* and Xian Fu Gwee Faculty of Chemical and Natural Resources Engineering, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Kuantan, Pahang, Malaysia *Corresponding author E-mail: fcchong@ump.edu.my The ultrasonic extraction (UE) method of anthocyanin from Clitoria ternatea flowers using response surface methodology was performed in this study. By using response surface methodology (RSM), the objective is to optimise the extraction yield of anthocyanin from Clitoria ternatea which is influenced by various factors, including the extraction temperature, time, ratio of solvent to solid and ultrasonic power. The empirical model was investigated by performing first level optimization in two-level factorial design with Design Expert 7 software. In comparison with the conventional solvent extraction (SE), ultrasonic extraction (UE) showed a 246.48 % better extraction yield and produced an anthocyanin extract with a radical scavenging activity of 68.48% at the optimized factors of 50°C, 150 min, 15 ml/g and 240 W. Keywords: Clitoria ternatea; ultrasonic extraction; anthocyanin; antioxidant activity; response surface methodology A Experimental A.1 Plant material and chemicals Fresh Clitoria ternatea flowers were obtained through self-planting from Balik Pulau, Penang, Malaysia. Only individual Clitoria ternatea flower that was free of defects and diseases was chosen for this study. Fresh collection of blue pea flowers was washed with distilled water and stored in -25°C freezer and used for extraction within a month. Only the petals of fresh Clitoria ternatea flowers were used in extraction process. They were cut into smaller pieces (average size of 0.5 cm x 0.3 cm) using a blender (MX-800S, Panasonics, Selangor, Malaysia) before usage. Only a single batch collection of flowers was used for extraction purpose to eliminate inconsistency of flowers batches. In addition, the wet flowers sample without any oven drying was used for extraction as it was interesting from technological point of view. This was due to that the sample did not require previous drying of the source. In addition, wet flowers sample was easier to separate the extracts from exhausted vegetal source compared to fine ground particles. This provided an edge against the long and conventional method of oven-drying and maceration of vegetal source sample before extraction process. Potassium chloride (KCl) and sodium acetate (NaC2H3O2) were purchased from Merck (Selangor, Malaysia). Acetic acid glacial, L-Ascorbic acid of analytical reagent grade and hydrochloric acid were purchased from Fisher Scientific (Selangor, Malaysia). 2,2-diphenyl-2picrylhydrazyl (DPPH) was purchased from Sigma-Aldrich (Kuala Lumpur, Malaysia). 95% ethanol was purchased from Copens Scientific (Selangor, Malaysia). A.2 Ultrasonic Extraction Ultrasonic extraction (UE) was done according to the modified method of Chen et al. (2007). Distilled water was used as solvent. Ultrasonic bench top cleaner machine (CP1800, Crest Ultrasonics, New York, USA) with power from 96 to 240 W was used for ultrasonic extraction. The UE variables were temperature (30-50°C), time (30-150 min), ratio of solvent to solid (2-15 ml/g) and ultrasonic power (96-240 W). The temperature was controlled and maintained at targeted temperature by periodically adding ice into and removing water from the ultrasonic machine. Anthocyanin extract was filtered through Whatman No. 1 paper twice where the residue was gathered and extracted again under the same conditions. The anthocyanin extract was determined using UV-VIS spectrophotometer (U1800, Hitachi HighTech, Kuala Lumpur, Malaysia). The colour change of sample during the ultrasonic extraction was observed in Figure S1. A.3 Conventional Solvent Extraction (SE) This was done using the modified version of the process described by Fuleki and Francis (1968). One gram of sample was added with 15 mL of 1.5 M HCl–95% ethanol. The SE variables were set to be equivalent to the optimised UE variables, which were temperature (50°C), time (150 min), ratio of solvent to solid (15 ml/g). Solvent extraction (SE) was carried out in a water bath machine (ONE 14, Memmert, Schwabach, Germany). A.4 Experimental design for UE The design of experiment (DOE) for UE was done according to two-level factorial design in Design Expert 7 software. However, only the first level known as screening test (ST) of the twolevel factorial design was performed in this study. Extraction variables were extraction temperature (°C), extraction time (min), ratio of liquor to solid (ml/g) and sonication power (W). Extraction variables were input according to their upper and lower limit and numerical or categorical factor for the ST as shown in Table S1. Range of parameters values was entered into Design Expert 7 software which was obtained from literature review for the trial runs for ST. To generate the DOE for ST, two level factorial design was chosen where 4 factors were chosen with 8 runs and 3 replicates. 8 runs (half design) were chosen instead of 16 runs (full design) due to the time and resource constraint in conducting 64 runs to 24 runs for 3 replicates each. Then, the extraction variables were entered according to their label, unit, type and upper and lower limit. All were numerical factor except for sonication power at categorical factor as explained earlier. Next, the number of response was selected which was only anthocyanins content with its label and unit. Lastly, the DOE for ST was generated and presented in Table S2. A.5 Determination of total anthocyanin content (TAC) TAC determination was performed according to the modified method of Kerio et al. (2012). As shown in Figure S2, 200 µl of ultrasonic extraction (anthocyanin extracts) sample was dissolved in 1.8 ml of 0.025 M potassium chloride (KCl) buffer at pH 1.0 and 0.4 M sodium acetate (NaC2H3O2) buffer at pH 4.5 each with a pre-determined dilution factor of 10. This is known as the pH differential method. Blank reading for measurement was taken using deionized water. Then, measurement by UV-VIS spectrophotometer on absorbance at 520 nm and 700 nm was done in triplicate. The absorbance (A) of the diluted sample was calculated as in Eq. (3) and Eq. (4). A = (A520 – A700)pH1.0 – (A520-A700)pH4.5 Anthocyanin content (mg/L) = (3) A×MW×DF×1000 (4) e ×1 where MW (molecular weight) of anthocyanin equivalent, delphindin-3-glucoside = 500.8, DF (dilution factor) = 10, e = 29,000 molar absorbance of delphinidin-3-glucoside. The anthocyanin content was calculated from delphindin-3-glucoside equivalent. A.6 Comparison of anthocyanin content between UE and SE The comparison of yield between ultrasonic extraction and conventional solvent extraction with reference to anthocyanin content of both extractions was calculated by Eq. (5). Yield comparison, % = ACOUE -ACOCSE ACOCSE × 100% (5) where ACOUE (mg/g) is the anthocyanin content of ultrasonic extraction and ACOCSE (mg/g) is the anthocyanin content of conventional solvent extraction. A.7 Comparison of ST and validation test (VT) of anthocyanin content The comparison was done in term of difference between both tests. Validation test (VT) was performed according to the optimized experimental variables value. The difference was calculated by Eq. (6). It was considered accurate shall the difference was less than 10%. Difference, % = ST - VT VT × 100% (6) A.8 DPPH radical-scavenging activity assay The DPPH radical-scavenging activity was estimated by the modified method of Bondet et al. (1997). Aliquots (0.5 ml) of 0 (control), 25, 50, 100, 142.86, 333.33 and 1000 µg/ml of ascorbic acid dissolved in 95% ethanol were added to 2.5 ml of 0.2 mM DPPH solution in 95 % ethanol. Anthocyanin extracts conducted in the DPPH assay were obtained from the validation test of anthocyanin extracts. Concentrations of anthocyanin extracted of 5.45, 9.23, 14.01, 23.41 and 73.62 µg/ml were used to replace ascorbic acid used earlier for DPPH assay. The absorbance at 517 nm of samples was measured after 30 min of incubation at room temperature in the dark. Measurements were done in triplicates. The percentage of DPPH radical-scavenging activity was calculated as in Eq. (7). DPPH, % = [1- (A-B) A0 ] ×100% (7) Where A = absorbance of sample with DPPH solution in ethanol with DPPH solution in ethanol as blank, B = absorbance of 0.5 ml of extracts + 2.5 ml of ethanol with ethanol as blank, A0 = absorbance of control. References Bondet, V., Brand-Williams, W., & Berset, C. (1997). Kinetics and mechanism of antioxidant activity using the DPPH free radical method. LWT-Food Science and Technology, 30, 609– 615. Chen, F., Sun, Y.Z., Zhao, G.H., Liao, X.J., Hu, X.S., Wu, J.H., & Wang, Z.F. (2007). Optimization of ultrasonic-assisted extraction of anthocyanin in red raspberries and identification in extract using high-performance liquid chromatography-mass spectrometry. Ultrasonics Sonochemistry, 14, 767-778. Fuleki, T., & Francis, F.J. (1968). Quantitative methods for anthocyanin. 3. Purification of cranberry anthocyanin. Journal of Food Science, 33, 266–274. Kerio, L.C., Wachira, F.N., Wanyoko, J.K., & Rotich, M.K. (2012). Characterization of anthocyanin in Kenyan teas: Extraction and identification. Food Chemistry, 131, 31-38. Table S1. Screening test results Run A B C D Anthocyanins content, Y (mg/g) 1 50 150 2 96 0.028551 2 50 150 15 240 0.935978 3 30 30 15 240 0.144196 4 50 150 15 240 1.125937 5 30 30 15 240 0.147650 6 30 30 15 240 0.147650 7 30 150 2 240 0.092792 8 50 30 2 240 0.124567 9 30 150 2 240 0.077941 10 50 150 15 240 1.124210 11 30 150 15 96 0.138152 12 30 150 2 240 0.107413 13 50 30 2 240 0.124567 14 50 30 2 240 0.133777 15 50 150 2 96 0.031430 16 30 30 2 96 0.107528 17 50 30 15 96 0.141606 18 50 150 2 96 0.025328 19 50 30 15 96 0.145923 20 30 30 2 96 0.101311 21 50 30 15 96 0.149377 22 30 150 15 96 0.138152 23 30 150 15 96 0.143332 24 30 30 2 96 0.108794 A: extraction temperature, B: extraction time, C: ratio of liquor to solid, D: sonication power Table S2. ANOVA results for screening test S SSb DFc MSd F-value Prob-F Me 2.40 7 0.34 224.44 < 0.0001 A 0.29 1 0.29 189.61 < 0.0001 B 0.24 1 0.24 156.13 < 0.0001 C 0.49 1 0.49 318.75 < 0.0001 D 0.38 1 0.38 250.00 < 0.0001 AB 0.26 1 0.26 172.00 < 0.0001 AC 0.35 1 0.35 228.08 < 0.0001 AD 0.39 1 0.39 256.52 < 0.0001 f 0.02 16 0.00 PE S: Source, SS: Sum of square, DF: Dilution factor, MS: Mean square, M: Model, PE: Pure error Figure S1. Color change (left to right) of water solvent during ultrasonic extraction. Figure S2. Anthocyanins extracts in KCl buffer (pink solution) and NaC2H3O2 buffer (blue solution).