Optimization of continuous and intermittent microwave extraction of pectin from banana peels

Food Chemistry 220 (2017) 108–114

Contents lists available at ScienceDirect

Food Chemistry journal homepage: www.elsevier.com/locate/foodchem

Optimization of continuous and intermittent microwave extraction of pectin from banana peels Gabriela John Swamy ⇑, Kasiviswanathan Muthukumarappan Department of Agriculture and Biosystems Engineering, South Dakota State University, Brookings, SD, United States

a r t i c l e

i n f o

Article history: Received 31 May 2016 Received in revised form 24 September 2016 Accepted 29 September 2016 Available online 29 September 2016 Keywords: Banana peel Pectin Continuous and intermittent microwave extraction Box–Behnken design Optimization

a b s t r a c t Continuous and intermittent microwave-assisted extractions were used to extract pectin from banana peels. Extraction parameters which were employed in the continuous process were microwave power (300–900 W), time (100–300 s), pH (1–3) and in the intermittent process were microwave power (300–900 W), pulse ratio (0.5–1), pH (1–3). The independent factors were optimized with the Box– Behnken response surface design (BBD) (three factor three level) with the desirability function methodology. Results indicate that the independent factors have substantial effect on the pectin yield. Optimized solutions for highest pectin yield (2.18%) from banana peels were obtained with microwave power of 900 W, time 100 s and pH 3.00 in the continuous method while the intermittent process yielded the highest pectin content (2.58%) at microwave power of 900 W, pulse ratio of 0.5 and pH of 3.00. The optimized conditions were validated and close agreement was observed with the validation experiment and predicted value. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction Pectin is a polysaccharide that holds the cell wall and middle lamellae of plant cells. It is the methylated ester of polygalacturonic acid (Sriamornsak, 2003). It contains 1,4-a linked galacturonic acid units (Wang et al., 2015). Commercial natural pectin is derived from fruit peels, the by-products of juice manufacturing process (Kratchanova, Pavlova, & Panchev, 2004). Pectin is used as an emulsifier, texturizer, thickener, and stabilizer in the Food and Bioprocessing industry (Kermani, Shpigelman, Pham, Van Loey, & Hendrickx, 2015; Yuliarti et al., 2015). It is largely used as a gelling agent in the preparation of jellies and jam. Health benefits of pectin include lowering blood cholesterol and reducing low density lipoprotein (LDP) that cause heart diseases (Fraser, 1994). Banana peels are trashed after banana fruit juice and puree processing. A major portion goes to landfills while it is also used as cattle feed. Approximately 780 million pounds of banana peels are annually sent to the landfill in the U.S. Industries can invest in creating a value added product from the peels. Pectin production is a profitable value added product (Oliveira et al., 2016). Several methods are used to extract pectin. In recent times, continuous microwave extraction has been researched extensively (Bagherian, Ashtiani, Fouladitajar, & Mohtashamy, 2011). A ⇑ Corresponding author. E-mail address: [email protected] (G.J. Swamy). http://dx.doi.org/10.1016/j.foodchem.2016.09.197 0308-8146/Ó 2016 Elsevier Ltd. All rights reserved.

simultaneous increase in temperature coupled with molecular rotation has shifted attention towards microwave extraction from other conventional techniques. Moreover, the intermittent microwave extraction process is more advantageous than the continuous process as it employs pulsed heat supply while increasing efficiency (Kumar, Joardder, Farrell, Millar, & Karim, 2016). Intermittent extraction avoids overheating of samples while balancing heat and mass transfer processes (Kumar, Joardder, Karim, Millar, & Amin, 2014). Intermittent extraction of bioactive compounds also improves extraction efficiency (Chumnanpaisont, Niamnuy, & Devahastin, 2014; Hiranvarachat & Devahastin, 2014). Pectin has been extracted using a continuous microwave extraction process from the passion fruit peels (Seixas et al., 2014), orange peels (Guo et al., 2012; Kratchanova et al., 2004; Zhongdong, Guohua, Yunchang, & Kennedy, 2006), apple pomace (Wang et al., 2007), and lime (Fishman, Chau, Hoagland, & Hotchkiss, 2006). Literature analysis reveals that no research has been found in the intermittent microwave-assisted extraction (IMAE) of pectin from banana peels. This research aims at comparing continuous and intermittent extraction processes for pectin extraction from banana peels. The empirical statistical modeling technique, Response surface methodology (RSM) that performs multiple regression analysis has been used to solve simultaneous multivariate equations (Swamy, Sangamithra, & Chandrasekar, 2014). It is a highly efficient statistical tool that carries out optimization for multifaceted processes. The experimental trials needed to assess independent

109

G.J. Swamy, K. Muthukumarappan / Food Chemistry 220 (2017) 108–114

factors and their interactions can be minimized. In this research, RSM was used to optimize the CMAE (microwave power, pH and time) and IMAE (microwave power, pH and pulse ratio) extraction of pectin from banana peels. The study focuses on optimizing the extraction factors to enhance pectin yield using the BBD design.

Based on this equation, the design suggested 17 experiments and 3 center points (determine experimental error) (Asokapandian, Venkatachalam, Swamy, & Kuppusamy, 2015). The methodology was assessed by examining the responses (Y), a function of z1, z2, z3. . .. zc, as described by

Y ¼ f ðz1 ; z2 ; z3 ; . . . . . . zc Þ þ e 2. Materials and methods 2.1. Sample preparation Ripe bananas (stage 6) were purchased from the local store Walmart, Brookings (South Dakota, USA). The banana peels were used for the study. The banana skins were cut into fine slices (1 mm). They were washed in flowing water and blanched in hot water for 2 min. They were dried in a hot air oven at a temperature of 60 °C. After constant weight was obtained, the dried banana peel was powdered in a mill. The powdered peels were stored in an airtight pouch before conducting the experiments.

ð4Þ

where f – response function; e – error. Response surface behavior was analyzed for Y using the secondorder polynomial equation (Quanhong & Caili, 2005). The general equation for a response surface model is

Y ¼ b0 þ

c c c X X XX ba za þ baa z2a þ bab za zb þ ea a¼1

a¼1

ð5Þ

a b62

where Y is the response; za and zb – variables, where a and b range from 1 to c; b0 – intercept; ba, baa and bab – interaction coefficients; c – independent factors (c = 3 in this research); ea – error. 2.4. Statistical analysis

2.2. Microwave assisted extraction of pectin Pectin extraction was conducted with a microwave oven (GE Advantium 120 Oven). Roughly, 2 g of the powdered banana peels was added to a 100 ml beaker containing 20 cc distilled water. The pH (1–3) of the water was adjusted using HCl. The beaker was placed at the middle of the rotating table and microwave power levels (300–900 W) were changed. In continuous microwave process, the processing time varied between 100 and 200 s while the pulse ratio was varied between 0.5–1 for the intermittent extraction process. The pulse ratio was calculated as (Kumar, Karim, & Joardder, 2014):

Pulse ratio ¼

ton toff

ð1Þ

Upon completion of the extraction process, the beaker was cooled to room temperature. The solution was filtered and subjected to centrifugation at 1000 RPM for 10 min. The supernatant was precipitated with 95% (v/v) ethanol. The coagulated mass was washed with 95% ethanol thrice. The sample was then dried at 50 °C in the hot air oven. After constant weight was obtained, the yield was calculated as

Yield ð%Þ ¼

Weight of dry sample ðgÞ  100 Weight of initial sample ðgÞ

ð2Þ

To ensure the accuracy of the experiment, the enzymatic process was also conducted according to the method described by (Qiu et al., 2010) and the yield was calculated. 2.3. Experimental design Using Box–Behnken design, experiments were designed to contain three factors and three levels. The independent factors were coded as 1, 0 and +1 denoting the low, mid and high level. Based on the following equation, the factors were coded:

zi ¼

Zi  Zc ; DZ i

i  1; 2; 3; . . . ::z

ð3Þ

where zi is the dimensionless value of the independent factor; Zi is the real value of the independent factor; Zc is the real value of the independent factor at the center point; DZi is the change of real value with respect to changes in dimensionless value. Experimental runs (N) needed to frame a Box–Behnken design (BBD) were N = 2p (p  1) + Cp (Swamy et al., 2014), where p – number of factors; Cp – number of center points.

Design Expert Statistical Software 8.0.7.1 (Stat Ease Inc., Minneapolis, USA) was employed to run statistical tests. Results were analyzed with multiple regressions. The significance of regression coefficients was assessed by F-test. A quadratic model was developed and model adequacies were determined by R2, adjusted R2 and prediction error sum of squares (PRESS) values. Further, the significant terms in the model were deduced for the response using ANOVA. The coefficients from the regression models were used to plot response surface graphs. 2.5. Validation of optimized solutions and predicted models Optimal conditions for CMAE (microwave power, pH and time) and IMAE (microwave power, pH and pulse ratio) on the extraction of pectin from banana peels were derived from Derringer’s desired function methodology. Experiments were conducted for the predicted optimal conditions and to confirm the validity of the models, comparison of experimental and predicted values was carried out. 3. Results and discussions 3.1. Experimental design and analysis As stated earlier, RSM is an empirical modeling technique. It calculates the correlation between the actual and forecast outcomes. To attain a suitable model to optimize the CMAE (microwave power, pH and time) and IMAE (microwave power, pH and pulse ratio) process, the BBD design was used. To understand the effect of the independent factors on the pectin yield from banana peels and to determine the optimum solutions, the experimental design was followed. The model fitting technique in the software enabled to achieve the anticipated values. The predicted pectin yield correlated well with the experimental numbers. Table 1 presents the experimental data and anticipated values. The effects of the enzymatic process are slightly higher as the procedure is more specific however it is extensively time consuming. To get regression equations, the data were fitted to linear, interactive, quadratic and cubic models (Swamy et al., 2014). To check model adequacy, the sequential model sum of squares and test statistics were done. The results of the statistical tests are tabulated in Table 2. Analysis of Table 2, shows that the quadratic model is highly significant to extract pectin from banana peels. Maximum R- Squared, Adjusted R-Squared and Predicted R-Squared values were observed in the model. The BBD matrix had enough data to comprehend the results

110

G.J. Swamy, K. Muthukumarappan / Food Chemistry 220 (2017) 108–114 Table 1 Pectin yield from the extraction processes. Run

Pectin (%)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Continuous microwave extraction

Intermittent microwave extraction

Yexp

Ypre

Yexp

Ypre

1.68 1.58 1.79 1.09 1.18 1.38 1.31 1.02 1.21 1.95 1.24 1.38 1.58 1.38 2.17

1.67 1.61 1.82 1.09 1.15 1.38 1.32 1.04 1.18 1.95 1.25 1.38 1.57 1.38 2.15

1.85 1.85 2.07 2.31 1.57 2.07 1.69 1.54 2.65 1.85 1.78 1.99 1.65 1.49 2.31

1.85 1.85 2.08 2.32 1.56 2.09 1.67 1.53 2.62 1.85 1.76 2.01 1.64 1.52 2.32

Enzymatic process pectin yield = 3.96%

Table 2 Model adequacy test results. Source

Sum of squares

DF

Mean square

F value

Prob > F

Remarks

66.3 1.0 16.2 63660000.0

<0.0001 0.4249 0.0052 <0.0001

Suggested Aliased

62.5 4.4 13.4 63660000.0

<0.0001 0.0421 0.0080 <0.0001

Suggested Aliased

Predicted R-squared

PRESS

0.933 0.934 0.990 1.000

0.902 0.857 0.944

0.147 0.213 0.084 +

Suggested Aliased

0.930 0.963 0.994 1.000

0.889 0.920 0.963

0.166 0.120 0.056 +

Suggested Aliased

Sequential model sum of squares for pectin (CMAE) Mean 32.091 1 Linear 1.414 3 2FI 0.022 3 Quadratic 0.051 3 Cubic 0.005 3 Residual 0.000 2 Total 33.583 15

32.091 0.471 0.007 0.017 0.002 0.000 2.239

Sequential model sum of squares for pectin (IMAE) Mean 54.798 1 Linear 1.415 3 2FI 0.052 3 Quadratic 0.028 3 Cubic 0.003 3 Residual 0.000 2 Total 56.296 15 Source Std. Dev. R-squared

54.798 0.472 0.017 0.009 0.001 0.000 3.753 Adjusted R-squared

Model summary statistics for pectin (CMAE) Linear 0.084 2FI 0.084 Quadratic 0.032 Cubic 0.000

0.948 0.962 0.996 1.000

Model summary statistics for pectin (IMAE) Linear 0.087 2FI 0.063 Quadratic 0.026 Cubic 0.000

0.945 0.979 0.998 1.000

+ case(s) with leverage of 1.0000: PRESS statistic not defined.

of the test and hence the cubic model was not suggested. Since the design proved that the quadratic model was sufficient, further analysis of data was carried out. 3.2. Data fitting and statistical analysis From the BBD model, the obtained second-order polynomial equations are given below

Pectin content ðCMAEÞ ¼ 1:38 þ 0:38 A þ 0:17 B þ 0:044 C þ 0:068 AB þ 0:018 AC  0:025 BC þ 0:12 A2 þ 0:033 B2 þ 0:0075 C2

ð6Þ

Pectin content ðIMAEÞ ¼ 1:85 þ 0:39 A þ 0:17 B  0:005 C þ 0:11 AB þ 0:0075 AC þ 0:042 BC þ 0:083 A2 þ 0:033 B2

ð7Þ

To avoid poor or disingenuous response surface results, the adequacy and fitness of the model were checked using regression analysis and ANOVA (Sangamithra, Sivakumar, Kannan, & John, 2015). Table 3 confirms that the equation can efficiently depict the relationship between the input and output factors. ANOVA subdivides the total variation of the results into smaller parts coupled with sources of variation to test hypotheses about the independent factors (Swamy et al., 2014). Further, the Fisher’s statistical test (F-test) analyzed the significance of the individual independent

111

G.J. Swamy, K. Muthukumarappan / Food Chemistry 220 (2017) 108–114 Table 3 ANOVA analysis of the experimental data. Source Pectin (CMAE) Model A-MW B-pH C-time AB AC BC A2 B2 C2 Residual Lack of Fit Mean C.V.% Adeq Precision Pectin (IMAE) Model A-MW B-pH C-PR AB AC BC A2 B2 C2 Residual Lack of Fit Mean C.V.% Adeq Precision

Coefficient estimate

Sum of squares

DF

Standard error

Mean square

F-value

p-value

1.380 0.383 0.169 0.044 0.068 0.018 0.025 0.115 0.033 0.008

1.486 1.170 0.228 0.015 0.018 0.001 0.002 0.049 0.004 0.000 0.005 0.005

9 1 1 1 1 1 1 1 1 1 5 3

0.019 0.011 0.011 0.011 0.016 0.016 0.016 0.017 0.017 0.017

0.165 1.170 0.228 0.015 0.018 0.001 0.002 0.049 0.004 0.000 0.001 0.002

158.1 1120.0 218.0 14.7 17.4 1.2 2.4 46.7 3.7 0.2

<0.0001 <0.0001 <0.0001 0.0123 0.0087 0.3284 0.1826 0.0010 0.1112 0.6744

1.494 1.194 0.221 0.000 0.044 0.000 0.007 0.025 0.004 0.000 0.003 0.003

9 1 1 1 1 1 1 1 1 1 5 3

0.015 0.009 0.009 0.009 0.013 0.013 0.013 0.014 0.014 0.014

0.166 1.194 0.221 0.000 0.044 0.000 0.007 0.025 0.004 0.000 0.001 0.001

238.9 1717.3 318.1 0.3 63.5 0.3 10.4 36.2 5.6 0.0

<0.0001 <0.0001 <0.0001 0.6147 0.0005 0.5940 0.0234 0.0018 0.0900 1.0000

1.5 2.2 41.8 1.850 0.386 0.166 0.005 0.105 0.008 0.043 0.083 0.033 0.000

1.9 1.4 51.3

factor. The F-value, ratio of the regression mean square and real error mean, indicates the effect of the individual controlled factor on the model. From Table 3, it is observed that the F-value for CMAE pectin as 158.1 and IMAE pectin is 238.9. The large F figures show that the deviation in the responses could be clarified by the regression equation, indicating that the model is highly significant. Also, the related p values are less than 0.05, confirming that the model is statistically significant. The R2 and adjusted-R2 confirm the adequacy and suitability of the model. The R2 values were 0.996 for CMAE and 0.998 for IMAE pectin respectively. This shows that at least 95% of

experimental data suits the model. The adjusted-R2, 0.990 for CMAE pectin and 0.994 for IMAE pectin, denotes high correlation of the actual and predicted values. Higher coefficient of determination in combination with a very small p-value (<0.0001) ensures that the quadratic model is adequate to illustrate the liaison between the output and input factors. The coefficient of variation (CV) represents the deviation of the actual points from the predicted ones and a low coefficient of variation shows the least variation in the mean value (Swamy et al., 2014). The CV value of 2.2 (CMAE) and 1.4 (IMAE) indisputably denotes high precision and reliability.

Fig. 1a. Three dimensional RSM plots - Effect of microwave power and pH on pectin content (CMAE).

112

G.J. Swamy, K. Muthukumarappan / Food Chemistry 220 (2017) 108–114

3.3. Investigating model adequacy Diagnostic plots ensures model adequacy and exhibits the correlation between actual and predicted results. The data points lie close to the straight line, signifying close agreement between the two data. It also suggests that the model is capable to select suitable operating conditions to extract pectin from banana peels. Further, the normality assumption was verified by plotting a normal probability plot of the residuals. The data was found to lie near the straight line, it is concluded that the data was normally distributed. 3.4. Influence of the independent factors in the process Three dimensional response surface plots generated by the software were analyzed to interpret the impact of the independent factors on the procedure. The plots show the impact of two factors on the process while holding the third variable constant. From Figs. 1a–1c, we would be able to comprehend the effects of microwave power levels, pH and time on the

continuous extraction of pectin. The figure shows that the pH and time significantly (p < 0.0001) influenced the pectin yield in a linear fashion. Microwave power levels influenced the pectin yield in both linear and quadratic positions. All independent factors revealed positive effects in linear terms. From Fig. 1a, it is understood that the increase in pectin content corresponded to the rise in microwave power and pH. At increased microwave power levels, the plant tissue softens and cuts down the phenolic compound and protein/carbohydrates interface. Thus, the solubility of the phenolic compounds increases. This results in an enhanced diffusion rate, thereby prominent accelerating extraction rate. Plot Fig. 1b illustrates that pH and time have a strong sway on the process. Table 2 shows the quadratic model has been selected for the process and higher R2 and adjusted-R2 confirming the suitability of the model. A curvilinear increase in pectin content was observed due to the positive linear effect (p < 0.0001). In Fig. 2a, the effect microwave power and the pH of the intermittent extraction of pectin is depicted. Accelerating the microwave power level at regular time intervals enhanced pectin yield. Pulse ratio, regular

Fig. 1b. Three dimensional RSM plots - Effect of pH and time on pectin content (CMAE).

Fig. 1c. Three dimensional RSM plots - Effect of microwave power and time on pectin content (CMAE).

G.J. Swamy, K. Muthukumarappan / Food Chemistry 220 (2017) 108–114

Fig. 2a. Three dimensional RSM plots - Effect of microwave power and pH on pectin content (IMAE).

Fig. 2b. Three dimensional RSM plots - Effect of pH and pulse ratio on pectin content (IMAE).

Fig. 2c. Three dimensional RSM plots - Effects of microwave power and pulse ratio on pectin content (IMAE).

113

114

G.J. Swamy, K. Muthukumarappan / Food Chemistry 220 (2017) 108–114

Table 4 Validation results at optimum conditions. Optimized parameters

a

Pectin (%) Optimized data (predicted)

Experimental dataa

Microwave power (W) = 900 pH = 3 Extraction time (s) = 100

2.18

2.18 ± 0.06

Microwave power (W) = 900 pH = 3 Pulse ratio = 0.5

2.58

2.58 ± 0.03

Mean ± standard deviation of triplicate determinations from experiments.

on–off periods increase swelling of the plant cells, enhancing pectin content (Fig. 2b). Likewise, under the lower pulse ratio and higher microwave power, pectin can dissociate into the solvent easily (Fig. 2c). The lower the pulse ratio, the higher is the extraction efficiency. 3.5. Optimal solutions for CMAE and IMAE of pectin For the optimization of the input factors, the Derringer’s desirability function method was employed. The desirability function identifies a combination of the variable levels that in combination optimizes a set of responses. The desirability (d) for each reaction is accomplished by declaring the goals and boundaries important for each answer. The importance of the goals can vary from 1 (least important) to 5 (most important) and using the methodology of the desired function, the optimized input variables were attained. It directed that for the CMAE process, microwave power of 900, pH of 3 and time of 100 s will yield 2.18% of pectin content with overall desirability value of 0.971. Likewise, for the IMAE process, microwave power of 900, pH of 3 and pulse ratio of 0.5 will yield 2.58% of pectin content with overall desirability value of 0.960. 3.6. Authentication of optimized solutions and predictive model Under the predicted conditions, the fitness of the model equations was examined and to verify the validity of the optimized solutions, experiments were carried out. The results are tabulated in Table 4. The efficiency of the extraction of pectin by the experimental operation was 2.18 ± 0.06% and 2.58 ± 0.03% for the CMAE and IMAE respectively. The mean experimental values were compared with the anticipated numbers. The validation results are within 95% of forecast values. Thus, it is apparent that the quadratic models are well-suited and the optimal values are valid inside the selected scope of autonomous agents. 4. Conclusion The work puts forward that the BBD design demonstrates to be effective and competent in narrowing down the optimal results of the CMAE and IMAE extraction of pectin from banana peels. The observational data indicate that the independent factors have a marked effect on the extraction procedure and response surface plots assessed the interactive effect input factors on the reaction. Second order polynomial models predicted the pectin yield. ANOVA results indicated a high coefficient of determination values (R2) of 0.990 for CMAE and 0.994 for IMAE. Therefore, the fit of the regression model with the experimental values was assessed and

established to be fit. Optimized solutions for the CMAE and IMAE were 2.18 ± 0.06% and 2.58 ± 0.03% yield. Validation experiments suggested a very good correlation between experimental and predicted pectin values. References Asokapandian, S., Venkatachalam, S., Swamy, G. J., & Kuppusamy, K. (2015). Optimization of foaming properties and foam mat drying of muskmelon using soy protein. Journal of Food Process Engineering. Bagherian, H., Ashtiani, F. Z., Fouladitajar, A., & Mohtashamy, M. (2011). Comparisons between conventional, microwave-and ultrasound-assisted methods for extraction of pectin from grapefruit. Chemical Engineering and Processing: Process Intensification, 50(11), 1237–1243. Chumnanpaisont, N., Niamnuy, C., & Devahastin, S. (2014). Mathematical model for continuous and intermittent microwave-assisted extraction of bioactive compound from plant material: Extraction of b-carotene from carrot peels. Chemical Engineering Science, 116, 442–451. Fishman, M. L., Chau, H. K., Hoagland, P. D., & Hotchkiss, A. T. (2006). Microwaveassisted extraction of lime pectin. Food Hydrocolloids, 20(8), 1170–1177. Fraser, G. E. (1994). Diet and coronary heart disease: Beyond dietary fats and lowdensity-lipoprotein cholesterol. The American Journal of Clinical Nutrition, 59(5), 1117S–1123S. Guo, X., Han, D., Xi, H., Rao, L., Liao, X., Hu, X., & Wu, J. (2012). Extraction of pectin from navel orange peel assisted by ultra-high pressure, microwave or traditional heating: A comparison. Carbohydrate Polymers, 88(2), 441–448. Hiranvarachat, B., & Devahastin, S. (2014). Enhancement of microwave-assisted extraction via intermittent radiation: Extraction of carotenoids from carrot peels. Journal of Food Engineering, 126, 17–26. Kermani, Z. J., Shpigelman, A., Pham, H. T. T., Van Loey, A. M., & Hendrickx, M. E. (2015). Functional properties of citric acid extracted mango peel pectin as related to its chemical structure. Food Hydrocolloids, 44, 424–434. Kratchanova, M., Pavlova, E., & Panchev, I. (2004). The effect of microwave heating of fresh orange peels on the fruit tissue and quality of extracted pectin. Carbohydrate Polymers, 56(2), 181–185. Kumar, C., Joardder, M., Farrell, T., Millar, G. J., & Karim, M. (2016). A mathematical model for intermittent microwave convective (IMCD) drying of food materials. Drying Technology, 34(8), 962–973. Kumar, C., Joardder, M. U. H., Karim, A., Millar, G. J., & Amin, Z. (2014). Temperature redistribution modelling during intermittent microwave convective heating. Procedia Engineering, 90, 544–549. Kumar, C., Karim, M., & Joardder, M. U. (2014). Intermittent drying of food products: A critical review. Journal of Food Engineering, 121, 48–57. Oliveira, T. Í. S., Rosa, M. F., Cavalcante, F. L., Pereira, P. H. F., Moates, G. K., Wellner, N., & Azeredo, H. M. (2016). Optimization of pectin extraction from banana peels with citric acid by using response surface methodology. Food Chemistry, 198, 113–118. Qiu, L.-P., Zhao, G.-L., Wu, H., Jiang, L., Li, X.-F., & Liu, J.-J. (2010). Investigation of combined effects of independent variables on extraction of pectin from banana peel using response surface methodology. Carbohydrate Polymers, 80(2), 326–331. Quanhong, L., & Caili, F. (2005). Application of response surface methodology for extraction optimization of germinant pumpkin seeds protein. Food Chemistry, 92(4), 701–706. Sangamithra, A., Sivakumar, V., Kannan, K., & John, S. G. (2015). Foam-Mat Drying of Muskmelon. International Journal of Food Engineering, 11(1), 127–137. Seixas, F. L., Fukuda, D. L., Turbiani, F. R., Garcia, P. S., Carmen, L. D. O., Jagadevan, S., & Gimenes, M. L. (2014). Extraction of pectin from passion fruit peel (Passiflora edulis f. flavicarpa) by microwave-induced heating. Food Hydrocolloids, 38, 186–192. Sriamornsak, P. (2003). Chemistry of pectin and its pharmaceutical uses: A review. Silpakorn University International Journal, 3(1–2), 206–228. Swamy, G. J., Sangamithra, A., & Chandrasekar, V. (2014). Response surface modeling and process optimization of aqueous extraction of natural pigments from Beta vulgaris using Box–Behnken design of experiments. Dyes and Pigments, 111, 64–74. Wang, S., Chen, F., Wu, J., Wang, Z., Liao, X., & Hu, X. (2007). Optimization of pectin extraction assisted by microwave from apple pomace using response surface methodology. Journal of Food Engineering, 78(2), 693–700. Wang, W., Ma, X., Xu, Y., Cao, Y., Jiang, Z., Ding, T., ... Liu, D. (2015). Ultrasoundassisted heating extraction of pectin from grapefruit peel: Optimization and comparison with the conventional method. Food Chemistry, 178, 106–114. Yuliarti, O., Matia-Merino, L., Goh, K. K., Mawson, J., Williams, M. A., & Brennan, C. (2015). Characterization of gold kiwifruit pectin from fruit of different maturities and extraction methods. Food Chemistry, 166, 479–485. Zhongdong, L., Guohua, W., Yunchang, G., & Kennedy, J. F. (2006). Image study of pectin extraction from orange skin assisted by microwave. Carbohydrate Polymers, 64(4), 548–552.