Carbohydrate Polymers 92 (2013) 2018–2026
Contents lists available at SciVer ScienceDirect
Carbohydrate
Polymers
j o u r n a l h o m e p a g e :w w w.e l s e v i e r.c o m /l o c a t e /c a r b p o
l
Modeling and optimization of ultrasound-assisted extraction of polysaccharide from Cucurbita moschata
J.Prakash Maran a ,∗,V.Mekala b ,S.Manikandan c
indicatesa
Department of Food Technology,Kongu Engineering College,Perundurai,Erode 638052,TN,India b
School of Food Science and Nutrition,University of Leeds,Leeds,England,United Kingdom c
Department of Food and Process Engineering,SRM University,SRM Nagar,Kattankulathur,Chennai 603203,TN,India
a r t i c l e
i n f o
Article history:
Received 26October 2012Received in revid form 10November 2012
Accepted 26November 2012
Available online 1 December 2012
Keywords:Ultrasound Extraction
Polysaccharide Pumpkin
revlon
Central composite design
a b s t r a c t
Polysaccharides from pumpkin were extracted by ultrasound-assisted extraction technology using four factors at five levels central composite rotatable respon surface design (CCRD).On using single factor analysis,process variables such as extraction temperature (50–70◦C),power of ultrasou
nd (50–70W),time (15–25min)and solid–liquid ratio (1:10–1:20g/ml)were lected.Experiments were conducted to evaluate the effects of four independent variables on the maximum extraction yield of polysaccharides.From the experimental data,cond order polynomial mathematical model were developed with high coefficient of determination values (R 2>0.96).From respon surface plots,temperature and ultrasound power exhibited independent and interactive effects on the extraction yields.Extraction temperature of 70◦C,ultrasound power of 70W,time of 23min and solid–liquid ratio of 1:10g/ml were determined as optimal conditions with a maximum polysaccharides yield of 16.21%,which was confirmed through the validation of the experiments.
© 2012 Elvier Ltd. All rights rerved.
1.Introduction
Polysaccharides from fruits and vegetables have drawn atten-tion of both food producers and consumers due to their physical properties,health-promoting and dia preventing potential and hence have been employed in cosmetic and pharmaceutical prod-ucts (Willats,Knox,&Mikkeln,2006).In recent years,the isolated polysaccharides have been found to play an important role in the biomedical field due to their antioxidant (Guo et al.,2010),immunostimul
atory (Sun &Liu,2009)and antitumor (Zhou,Song,Feng,&Tan,2011)effects.In the past few years,the pumpkin has received considerable attention becau of its nutritional and health protective value,and also its potential in medicinal us have been explored.
Pumpkin (Cucurbita moschata )is an annual herbaceous plant of the family Cucurbitaceae.The fruit of pumpkin is one of the most important vegetables in traditional agricultural sys-tems in the world.The flesh and peel of the fruit reprent rich sources of pectin-type dietary fiber and antioxidants (Caili,Huan,&Quanhong,2006).So far,veral beneficial physiological effects,immunological activity and other pharmacological activi-ties such as lipid-lowering,hepatoprotective (Makni et al.,2008),
∗Corresponding author.Tel.:+914294226606;fax:+914294220087.E-mail address: (J.Prakash Maran),
(V.Mekala), (S.Manikandan).
anti-carcinogenic,anti-microbial (Caili,Haijun,Tongyi,Yi,&Quanhong,2007;Park,Lee,&Kim,2010)and anti-diabetic proper-ties (Adams et al.,2011;Caili et al.,2006;Xia &Wang,2006;Yadav,Jain,Tomar,Prasad,&Yadav,2010)of various pumpkin extracts have been publish
ed and its antioxidant activity was reported by Nara,Yamaguchi,Maeda,and Koga (2009)for a water-soluble polysaccharide from the fruit of pumpkin.Pumpkin polysaccha-rides (PP)are compod of galacto,gluco,arabino,xylo and glucuronic acid and are water insoluble but organic solvents soluble macromolecular compounds with important biological functions.PP has the biological effects of detoxification,anti-oxidation,reduc-ing blood pressure,reducing blood lipids,lowering cholesterol levels (Yong,Ning,&Liu,2006),promote the biosynthesis of nucleic acids and proteins,control cell division and differentiation,regulat-ing cell growth and aging,especially for the treatment of diabetes (Zhang,Shen,&Zhu,2002).
The traditional extraction methods of polysaccharides from plant tissues are maceration,mechanical rabbling and heat reflux.The extraction methods depend largely on energy input and agitation to improve the solubility and mass transfer efficiency of polysaccharides.Usually,the conventional extraction method requires long extraction time and high extraction temperature with low extraction yield,but high-energy consumption (Chen,Li,Liu,Yang,&Li,2012).Ultrasound in combination with conventional extraction is a potential technique,which is a fully reproducible food process,completed in a shorter time with high reproducibil-ity,reduced processing cost,simplified manipulation and work-up.
0144-8617/$–e front matter © 2012 Elvier Ltd. All rights rerved.dx.doi/10.1016/j.carbpol.2012.11.086
vanquishJ.Prakash Maran et al./Carbohydrate Polymers 92 (2013) 2018–2026
2019
This process gives a higher purity of the final product,eliminates post-treatment of waste water and consumes only a fraction of the time and energy,normally needed for conventional process (Kim,Chi,&Hong,2009;Li,Wei,You,&Lydy,2010;Maran,Manikandan,Thirugnanasambandham,Nivetha,&Dinesh,2012;Sun,Liu,Chen,Ye,&Yu,2011).Ultrasound-assisted extraction (UAE)is an ideal extraction method capable of producing high quantities of polysaccharides and is non-destructive with a shorter extraction time.
Previous findings have reported the influence of many inde-pendent variables,such as solvent composition,pH,temperature,extraction time,and solid to liquid ratio,on the yields of bioactive compounds which can be extracted from diver natural prod-ucts (Buji´c-Koji´c,Planinic,Srécko,Jakguarantee是什么意思
abek,&Seruga,2009;Cacace &Mazza,2002,2003;Pinelo,Rubilar,Sineiro,&Nunez,2005;Santos,Veggi,&Meireles,2010;Wettasinghe &Shahidi,1999).Respon Surface Methodology (RSM)is a collection of statistical and mathematical techniques uful for developing,improving and optimizing process,in which a respon of interest is influenced by independent variables,and it generates a mathematical model that describes the chemical process (Prakash Maran,Sivakumar,Sridhar,&Prince Immanuel,2013)ud to find out their opti-mal values (Triveni,Shamala,&Rastogi,2001).Several studies on the optimized conditions for the extraction of phenolic com-pounds from different sources using RSM have been published (Hayouni,Abedrabba,Bouix,&Hamdi,2007;Pinelo et al.,2005;Pompeu,Silva,&Rogez,2009;Yang &Zhai,2010).Hence the objective of the prent study is to investigate the individual and interactive effect of UAE process variables such as extrac-tion temperature,power of ultrasound,time and solid–liquid ratio on the extraction yield of polysaccharide from pumpkin and to optimize the processing variables of UAE for the highest yield of pumpkin polysaccharides using central composite rotatable
respon surface design coupled with Derringer’s desired function methodology.
2.Materials and methods
2.1.Pumpkin powder preparation
Freshly harvested pumpkins (C.moschata )with similar maturity and weight were purchad from a local fruit orchard near Leeds,UK.The thick layer of the skin and eds were peeled from the fruits manually and the fruits were washed thoroughly in running tap water to remove any impurities adhered to the surface of the fruit.The washed fruits were dried in the hot air oven (NSW 143,India)at 40◦C until it attains the constant weight.Then the dried samples were pulverized and sifted through a 40-mesh sieve to obtain the powdered samples.The powder (moisture content 8–12%in dry basis)was stored in dark bags and kept in dry environment prior to conduct the experiments.
2.2.UAE of polysaccharide
For the UAE experiments,10g of dried pumpkin powder was mixed with an appropriate amount of distilled water in a 500ml beaker.Experiments were performed using a 20kHz ultrasonic device (VCX 400,Sonics and Materials,USA and 0–400W)with a 2.00cm flat tip probe in a beaker with provisions to t required output power,temperature and time.Ultrasonic generator probe was directly submerged into the suspension and the samples were extracted with continuous ultrasound waves at frequencies of 20kHz with different levels of power output.An amplitude controller was ud to t the desired level of ultrasonic power.Ultrasonic output powers were determined calorimetrically and r
anged from 50to 70W according to the method described by Li,
Table 1
Central composite rotatable experimental design with results for extraction of polysaccharide yield.
Run order
Blocks
Temperature (◦C)Power of ultrasound (W)
Time (min)SL ratio (g/ml)
Polysaccharide yield (%)
Residual error
%Error
Actual error
Experimental
Predicted 115070151:109.849.590.25 2.530.25216060201:1513.5814.09−0.51−3.780.51315070251:1011.9912.18−0.19−1.580.19415070251:209.429.75−0.33−3.540.33517050151:2016.0216.03−0.01−0.050.01617070151:1013.6713.84−0.17−1.250.177********:1514.4514.090.36 2.470.36817070151:2013.4313.79−0.36−2.680.36916060201:1514.2514.090.16 1.100.161016060201:1514.314.090.21 1.440.211117070251:2013.1513.17−0.02−0.140.021*********:1013.6113.060.55 4.010.551317070251:1015.5216.05−0.53−3.380.531416060201:1513.9314.09−0.16−1.180.161515050251:1010.7210.150.57 5.330.571615050151:2013.5112.770.74 5.450.741717050251:1013.4913.470.020.160.021*********:1514.0514.09−0.04−0.310.0419********:2010.7110.74−0.03−0.250.032017050251:2013.5713.61−0.04−0.270.0421********:109.189.36−0.18−1.960.182215070151:209.779.99−0.22−2.250.222324060201:159.419.71−0.30−3.170.302426040201:1512.6313.44−0.81−6.380.812526060101:159.9110.20−0.29−2.920.292628060201:1517.1116.830.28 1.650.282726080201:1514.0213.230.79 5.640.792826060301:1510.6410.370.27 2.570.272926060201:0512.0812.23−0.15−1.280.1530
2
keepalive60
60
20
1:25
12.91
12.77
0.14
1.07
0.14
2020J.Prakash Maran et al./Carbohydrate Polymers92 (2013) 2018–2026 Pordesimo,and Weiss(2004).During the extraction period,tem-
perature was controlled at a desired level within±1◦C and the
experiments were carried out according to Table1.All the experi-
ments were performed in triplicates and the reported result is the
mean of the triplicate measurements.
2.3.Determination of polysaccharide yield
After extraction,the extracts were centrifuged at2600×g for
15min(Remi R-24Centrifuge,India)andfiltered through afil-
ter paper(Whatman no.1,England).The obtained extracts were
concentrated with a rotary evaporator(Büchi,UK)at60◦C under
vacuum.The remaining solution was mixed with four volumes of
95%(v/v)ethanol for48h at4◦C and centrifuged again to collect
the precipitate as the crude extract,which was freeze dried at
−40◦C under vacuum and ground to powder.The percentage yield
of polysaccharide(Y)was calculated by the following equation:
Y(%)=w t
w i
×100(1)
where w t is the weight of the crude extract and w i is the weight of the pumpkin powder.
2.4.Experimental design
Central composite rotatable respon surface design(CCRD) was employed to study and optimize the effect of process variables such as extraction temperature(50–70◦C),power of ultrasound(50–70W),time(15–25min)and solid–liquid ratio (1:10–1:20g/ml)on the extraction yield of polysaccharide from pumpkin.The application of statistical experimental design tech-niques in bioprocess development and its optimization can result in enhanced product yields,clor conformance of the process output or respon to target requirements and reduced process variability,development time and cost(Maran,Sivakumar,Sridhar, &Thirgananasambandham,2012).On single factor analysis,pro-cess variables and their ranges were lected and independent variables were coded atfive levels between−2and2.The cod-ing of the variables was done by the following equation(Prakash Maran et al.,2013):
x i=X i−X z
X i
i=1,2,3,...,k(2)
where x i is the dimensionless coded value of an independent vari-able;X i,the real value of an independent variable;X z,the real value of an independent variable at the center point;and X i,step change of the real value of the variable i.
In this study,CCRD consists of16factorial points,8axial points, 6center points and two blocks.Totally30experiments were per-formed to optimize and study the effect of process variables on the respon.The Center point is replicated tofind and allow the estimation of experimental error.So the replication of the entire experimental design is not required.It is recommended that six center points have taken in a CCRD with four factors and the total number of experiments(N)was calculated by the following equa-tion:
N=2K+2K+C p(3) where K is the number of process variable,2K is the number of factorial points,2K is the number of the axial points on the axis of each design factor at a distance of±˛(˛=2K/4=2for K=4)and C p is the replicate number of the central point.
In this study,the experimental run was randomized in order to reduce the error arising from the experimental process due to the extraneous factors.A nonlinear regression method was ud tofit the cond order polynomial(Eq.(4))to the experimental data and express the mathematical relationship between process variables (X1,X2,X3and X4)and the respon(Y).The generalized form of the cond order polynomial equation is shown in Eq.(4).
Y=ˇ0+
k
˙
j=1
ˇj x j+
k
˙垃圾桶 英文
j=1
ˇjj x2j+˙
i
k
˙
<j=2
ˇij x i x j+e i(4)
4月英文where Y is the respon;x i and x j are variables(i and j range from1 to k);ˇ0is the model intercept coefficient;ˇj,ˇjj andˇij are inter-action coefficients of linear,quadratic and the cond-order terms, respectively;k is the number of independent parameters(k=4in this study);and e i is the error(Prakash Maran&Manikandan,2012). Thefinal mathematical cond order polynomial model includes4 linear terms,6two factor interaction terms,4squared terms and1 intercept term.
2.5.Statistical analysis
Design expert8.0.7.1statistical software package(Stat-Ea Inc.,USA)was ud to analyze the experimental data.Multiple regression analysis and Pareto analysis of variance(ANOVA)were ud to evaluate the experimental data and the ANOVA table was generated.Significant terms in the model(linear,interactive and quadratic)for the respon were found by analysis of variance (ANOVA)and significance was judged by the F-statistic value cal-culated from the data.The experimental data was evaluated with various descriptive statistical analysis such as p value,F value, degrees of freedom(DF),sum of squares(SS),coefficient of vari-ation(CV),determination coefficient(R2),adjusted determination of coefficient(R2a)and predicted determination of coefficient(R2p)to reflect the statistical significance of the developed quadratic math-ematical model.Afterfitting the data to the models,the model was ud for the construction of three dimensional respon sur-face plots to predict the relationships between independent and dependent variables.
2.6.Total percentage contributions of process variables
The total percentage contributions(P i)of each individual pro-cess variables were calculated bad on the regression coefficients obtained from the ANOVA analysis.The following equations were ud tofind out the P i of individual process variables as described by Khataee,Fathinia,Aber,and Zarei(20
10).
P i=
earlˇ2
i
ˇ2
i
×100(i/=0)(5) whereˇi is the regression coefficient of individual process variable.
2.7.Determination of optimal conditions
After analyzing the polynomial equation depicting the depend-ent and independent variables,optimization process was carried out by Derringer’s desired function methodology(Derringer& Suich,1980).This numerical optimization technique will optimize any combination of one or more goals;the may be either process variables or respons.The possible
goals are:maximize,minimize, target,within range,none(for respons only)and t to an exact value(factors only).In this study,goals of the process variables were lected as in a range and the respon goals were lected as maxi-mize.A weight factor of1was chon for the respon,which can be ud to adjust the shape of its particular desirability function.The default value of1creates a linear ramp function between the low value and the goal or the high value and the goal.Incread weight (up to10)moves the result towards the goal and the reduced weight (down to0.1)creates the opposite effect.Default importance of3
J.Prakash Maran et al./Carbohydrate Polymers92 (2013) 2018–20262021 Table2
Sequential modelfitting for the yield of polysaccharide.
Source Sum of squares DF Mean square F value Prob>F Remarks Sequential model sum of squares
Mean4835.4014835.40
Linear76.58419.149.230.0001
2FI20.966 3.49 2.150.0948
Quadratic26.864 6.7225.02<0.0001Suggested Cubic 3.3780.42 4.500.0312Aliad Residual0.6670.09
Total4963.8230165.46
Lack offit tests
Linear51.3620 2.5726.390.0009
2FI30.4014 2.1722.320.0015Suggested Quadratic 3.54100.35 3.640.0834Aliad Cubic0.1720.080.870.4750
Pure error 4.87E−0159.73E−02
Source Std.dev.R2Adjusted R2Predicted R2Press Remarks Model summary statistics
Linear 1.440.5960.5320.41675.00
2FI 1.270.7590.6330.55357.37
Quadratic0.520.9690.9390.83621.08Suggested Cubic0.310.9950.9790.80525.00Aliad
was chon for the respon,which can reprent the goals to be equally important.
2.8.Verification of the predicted optimized conditions
After optimization,in order to determine the validity of the optimized conditions,triplicate verification experiments were per-formed under the optimal conditions as predicted by the model.The average value of the experiments was compared with the predicted values of the developed model in order tofind out the accuracy and suitability of the optimized conditions.
3.Results and discussion
3.1.Experimental design(CCRD)analysis
The total number of30statistically designed batch experiments were performed for different combinations of the process variables in order to optimize and study the combined effect of indepen-dent variables(extraction temperature,power of ultrasound,time and solid–liquid ratio)on the extraction yield of polysaccharides and the results are shown in Table1,that includes the experimen-tal design,experimental and predicted values of the respon.The experimental data wasfitted to the various models(linear,interac-tive(2FI),quadratic and cubic)to obtain regression equation.Three differ
女孩儿英文名ent tests namely the quential model sum of squares,lack of fit tests and model summary statistics were carried out in this study to decide about the adequacy of models among various models to reprent the maximum yield of polysaccharide and the results are listed in Table2.
From Table2,linear and interactive(2FI)models were exhibited lower R2,adjusted R2,predicted R2and also high p-values,when compared with quadratic model.Cubic model was found to be aliad.Therefore the quadratic model incor-porating linear,interactive and quadratic terms was chon to describe the effects of process variables on the extraction of
Table3
Analysis of variance for the extraction of polysaccharide yield.
Source Coefficient estimate Sum of squares Degree of freedom Standard error Mean square F value p-Value Model14.09124.40140.218.8933.11<0.0001 X1 1.7876.0410.1176.04283.36<0.0001 X2−0.050.0610.110.060.240.6322 X30.040.0410.110.040.160.6991 X40.130.4310.110.43 1.610.2239 X120.140.3010.130.30 1.110.3094 X13−0.100.1510.130.150.550.4688 X14−0.110.2010.130.200.750.3987 X230.45 3.2410.13 3.2412.070.0034 X24−0.759.0910.139.0933.87<0.0001 X34−0.717.9810.137.9829.74<0.0001 X12−0.21 1.1710.10
1.17 4.360.0543 X22−0.190.9910.100.99 3.700.0737 X32−0.9524.9010.1024.9092.77<0.0001 X42−0.40 4.3410.10 4.3416.170.0011 Residual 4.03150.27
Lack offit 3.54100.35 3.64
Pure error0.48750.10
Cor total128.4229
Std.dev.0.52R20.969
Mean12.70Adjusted R20.939
C.V.% 4.08Predicted R20.836
Press21.08Adeq.precision20.39
2022J.Prakash Maran et al./Carbohydrate Polymers92 (2013) 2018–2026 polysaccharide from pumpkin.Furthermore,analysis of variance
(ANOVA)was also ud to check the adequacy of quadratic
model.
3.2.Fitting of cond order polynomial equation
By applying multiple regression analysis on the experimental
data,the Design-Expert software generated a cond-order poly-
nomial equation that can express the relationship between process
多伦多大学世界排名
variables and the respon.Thefinal equation obtained in terms of
coded factors is given below:
Yield(%)=14.09+1.78X1−0.052X2+0.042X3+0.13X4
+0.14X1X2−0.096X1X3−0.11X1X4+0.45X2X3
−0.75X2X4−0.71X3X4−0.21X2
1−0.19X2
2
−0.95X2
3
−0.40X2
4
(6) where Yield(%)is the polysaccharide yield;X1,X2,X3and X4are the coded values of extraction temperature,power of ultrasound, time and solid–liquid ratio,respectively.3.3.Statistical analysis
Pareto analysis of variance(ANOVA)and multiple regression analysis were ud to analyze the experimental data.The statistical significance of the regression equation was evaluated by the corre-sponding F and p-values and it is prented in Table3.The model F and p-value was found to be33.11and<0.0001,which indicated that the model was highly statistically significant.Thefitness of the model was studied through the lack offit test.The lack offit F-value of3.64and the associated p-value of0.0834was indicated the suit-ability of the model to predict the variations.The goodness of thefit of the model was evaluated by the determination co-efficient(R2), adjusted determination co-ef
ficient(R2a),predicted determination co-efficient(R2p)and co-efficient of variance(CV)and it was listed in Table3.The R2value of the predicted model was0.969,while R2a value was0.939,which exhibited the high degree of correla-tion between the experimental and predicted values.If there are many terms in the models and the sample size is not large enough, R2a may be noticeably smaller than R2(Yetilmezsoya,Demirelb, &Vanderbeic,2009).In our study,the R2a value was found to be smaller and very clo to the R2.The values of R2a and R2p should be approximately within0.20of each other,to be in
reasonable Fig.1.Diagnostic plots for the model adequacy.
