osmotic dehydration of apricot

8/10/2019 Osmotic Dehydration of Apricot

1/15

chemical engineering research and design 8 7 ( 2 0 0 9 ) 166180

Contents lists available at ScienceDirect

Chemical Engineering Research and Design

j o u rn a l h om epa ge : www. e l s ev i e r . c om / lo ca t e / ch e rd

Osmotic dehydration of apricot: Kinetics and the effect of process parameters

Ayse Ispir a , , Inci Trk To grul ba Firat University, Engineering Faculty, Department of Chemical Engineering, 23279 Elaz g, Turkeyb Afyon Kocatepe University, Engineering Faculty, Department of Chemical Engineering, 03200 Afyonkarahisar, Turkey

a b s t r a c t

The effect of different parameters on the osmotic dehydration of apricot in terms of water loss and solid gain,such as the different osmotic matter the concentration of solution (4070%, w/w), temperature (2545 C), the ratioof sample/solution (1/41/25), time, and geometry of sample were investigated. The increasing of temperature andconcentration of osmotic medium caused increased water loss and solid gain. The decreasing of the ratio of sampleto solution avoids signicant dilution of the medium by water removal and subsequent decrease of osmotic driving force during the process. The water loss and solid gain was increased when the dimension of apricot was decreased.

Effective diffusion and mass transfer coefcients of water as well as solid were estimated. The transport coef-cients for water loss and solid gain ( De and k) increases with an increase in osmotic solution concentration andincrease in temperature. Non-linear analysis of the estimated De and k of water and solute reveal that these valuesdepend on temperature and concentration of the osmotic solution as well as the combined effect of temperature andconcentration. In addition, the effect of the ratio of sample to solution on these transport coefcients was modeled.The statistical comparison methods such as 2, MBE and RMSE were used to explore the condence level of themodels.

2008 Published by Elsevier B.V. on behalf of The Institution of Chemical Engineers.

Keywords: Apricot; Osmotic dehydration; Effective diffusivity; Mass transfer coefcient; Empirical models

1. Introduction

Osmotic dehydration involves the partial removal of waterfrom food, such as fruits and vegetables, by immersing it in ahypertonic solution. A driving force for the diffusion of waterfrom the food into the solution is set up because the food cel-

lular surface structure acts as a semi permeable membrane.The diffusion of water is accompanied by the simultaneouscounter diffusion of solute from the osmotic solution intothe food. The existence of those simultaneous and oppositeuxes is one of the main difculties in modeling of osmoticdehydrationkinetics ( Spiazzi andMascheroni, 1997 ).Sincethemembrane responsible for osmotic transport is not perfectlyselective, other solutes present in the food arealso leakedintothe osmotic solution ( Torreggiani et al., 1988; Rastogi et al.,1997). The rate of diffusion of water from any material madeup of such tissues depend on factors such as; temperature andconcentration of the material, the solution to material mass

Corresponding author .E-mail addresses: a [email protected] (A. Ispir), [email protected] , [email protected] (I.T. To grul).Received30 October 2007; Accepted21July2008

ratioand the size andshape of food, thelevel ofagitation in thesolution, and the vacuum level, if applied. A number of recentpublicationshave describedthe inuenceof these variablesonmass transfer rates during osmotic dehydration ( Torreggiani,1993; Rastogi and Raghavarao, 1994, 1995, 1997; Rastogi et al.,2002; Azoubel and Murr, 2004; Kaymak-Ertekin and Cakaloz,

1996a, 1996b; Kaymak-Ertekin and Sultano glu, 2000).Much work has been done in developing models to pre-

dict the mass transfer kinetics of OD at atmospheric pressure.Nonetheless, it is very difcult to develop a mathematicalmodel capable of including all of the factors involved in theprocess.Mechanisticand empiricalapproaches have beenpro-posed by many authors as mentioned by Panagiotou et al.(1998) and Shi and Le Maguer (2002) .

Mechanistic approaches describe the underlying phenom-enaby means ofvariousmechanisms:some authorshave usedFicks law of diffusion (i.e. Kaymak-Ertekin and Sultano glu,2000; Rastogi et al., 1997; Salvatori et al., 1999 ) and some other

0263-8762/$ see front matter 2008 Published by Elsevier B.V. on behalf of The Institution of Chemical Engineers.doi:10.1016/j.cherd.2008.07.011
http://www.sciencedirect.com/science/journal/02638762mailto:[email protected]:[email protected]:[email protected]://localhost/var/www/apps/conversion/tmp/scratch_6/dx.doi.org/10.1016/j.cherd.2008.07.011http://localhost/var/www/apps/conversion/tmp/scratch_6/dx.doi.org/10.1016/j.cherd.2008.07.011mailto:[email protected]:[email protected]:[email protected]://www.sciencedirect.com/science/journal/02638762


2/15

chemical engineering research and design 8 7 ( 2 0 0 9 ) 166180 167

Nomenclature

D the relative % deviationDes effective diffusivity of solute (m 2 s 1)Dew effective diffusivity of water (m 2 s 1)M the moisture content (kg [H 2O]/kg [dry matter])

MBE mean bias errorML mass loss (kg [H2O]/kg [fresh fruit])Mr the moisture ratioRMSE root mean square errorS solid content (dry basis kg [dry matter]/kg [dry

matter])SG solid gain (kg [H2O]/kg [fresh fruit])SG solid gain in the equilibrium (kg [H 2O]/kg [fresh

fruit])Sr solute ratioWL water loss (kg [H2O]/kg [fresh fruit])WL waterloss in theequilibrium (kg[H 2O]/kg [fresh

fruit])

Greek letter2 reduced chi-square

authors have proposed models based on the knowledge aboutcellular physiology of tissues ( Spiazzi and Mascheroni, 1997;Toupin et al., 1989; Yao and Le Maguer, 1996 ).

On the other hand, empirical and semi-empirical mod-els have been proposed. These models correlate processing variables with water loss (WL) or solid gain (SG) without tak-ing into account the underlying phenomena and they includemulti-variable regressions, response surface analysis, modelsderived from mass balances, etc. Relevant examples of theseare reviewed by Shi and Le Maguer (2002) .

Although mechanistic models give a description of themass transfer mechanism, diffusion approach has a numberof assumptions which are difcult to fulll ( Kaymak-Ertekinand Sultano glu, 2000), and the effective diffusivity becomesan adjustable kinetic parameter that strongly depends on theexperimental conditions and the physical properties of thefruit (Salvatori et al., 1999 ). Also, cellular physiology approachdepends on a large number of biophysical properties, suchas elastic modulus of the cell wall, cell wall void fraction,cell wall tortuosity and membrane permeabilities, which arenot always available ( Kaymak-Ertekin and Sultano glu, 2000;Spiazzi and Mascheroni, 1997 ).

On the other hand, even though the empirical and semi-empirical models that have been proposed in the literaturegive a reasonable t to experimental data, their use is limitedbecause they are only capable of representing data at condi-tions similar to those on which such models were developed,and they cannot take into account the process complexity(Trelea et al., 1997 ).

Osmotic dehydration is used as a pretreatment to manyprocesses and improves nutritional, sensorial and functionalproperties of food without changing its integrity ( Torreggiani,1993). Osmotic dehydration is, generally, used as an upstreamstep for the dehydration of food before they are subjected tofurther processingsuch as freezing ( Ponting, 1973 ), freeze dry-ing (Hawkes and Flink, 1978 ), vacuum drying ( Dixon and Jen,1977) and air drying ( El-Aouar et al., 2003; Piotrowski et al.,2004; Mandala et al., 2005 ).

The traditional method, which is often used in Turkey forapricot drying, is to spread it on the ground to subject it tothe direct sunlight. As we know, in traditional drying methodsserious decreases of nutritive and sensorial values are pos-sible, damaging mainly the avor, color, and nutrients of theproduct ( Lenart, 1996; Linet al., 1998 ).Due to the large amountof dry apricot production, it is essential to nd better ways fordrying apricots. One of these methods is to pre-dry the apri-cots, before applying warm air or other drying methods, using osmotic dehydration method. A combined process, consist-ing of osmotic dehydration, followed by air dehydration hasbeen proposed to obtain dry apricot ingredients, having a nat-ural color, without sulphur dioxide, which could be suitablefor different applications ( Forni et al., 1997 ).

A continuing interest in dried apricot products has openedavenues for the dehydration of apricot by different methods,such as tray drying ( To grul and Pehlivan, 2003; Doymaz, 2004 ),andsolar drying ( To gruland Pehlivan, 2002, 2004 ).A good qual-itydried apricot, whole or in slices, commands greater marketpotential than apricot powder because of its versatile culinaryutility.

The purpose of the present work is (i) to study the effect of temperature and concentration of the osmotic solution, var-ious osmotic agents, the ratio of sample to solution and thegeometry of sample on the osmotic dehydration of apricotand (ii) to determine the effective diffusion and mass transfercoefcients of water and solute during osmotic dehydrationfor the whole range of experimental conditions and (iii) tomodel the effects of procedure parameters such as concen-tration and temperature of osmotic solution and the ratioof sample to solution, on this calculated effective diffusionsand mass transfer coefcients by using non-linear regres-sion and to select the best models by using statistical testindicators.

2. Materials and methods

2.1. Materials

Fresh apricots were directly collected from tree and broughtto the laboratory in wooden boxes. The apricots were refrig-erated at 5 C and 8090% relative humidity until they wereused in the experiments. The apricots were sorted visually formaturity and size (average weight of 25g (range 23.1826.11g),average diameter of 3 cm (range 2.823.21 cm)). The dimen-sions of apricots were measured by a digital micrometer. Theapricots with above mentioned values were selected. Theaverageinitialmoisturecontentwas 80.86%in wetbasis,gravi-metrically measured using an oven at 75 C for 24h. Since thisis a temperature at which no structural changes occur dur-ing drying process this is one of the preferred values in bothinfrared and oven methods ( To grul and Pehlivan, 2002, 2003,2004; Doymaz, 2004).

2.2. Experiments

Theosmoticdehydrationexperimentswerecarriedout in FiratUniversity Chemical Engineering Research Laboratories.

Various osmotic agents such as sucrose, glucose, fructose,maltodextrin and sorbitol have been used for osmotic dehy-drationof apricot. The initial concentrationof solutions variedfrom 40% to 70% (w/w) and temperatures varied from 25 C to45 C. The ratio of fruit to solution was kept at 1/25 to avoidsignicant dilution of the medium by water removal, which


3/15

168 chemical engineering research and design 8 7 ( 2 0 0 9 ) 166180

Table 1 % Relative deviation ( D ) of measured values and water loss and solid gain at equilibrium

Solution Concentration(w/w)

Temperature( C)

% Relativedeviation ( D)

WL (kg[H2O]/kg[freshfruit])

SG (kg[H2O]/kg[freshfruit])

Fructose 40 25 3.30 0.512 0.0343450 25 3.94 0.667 0.0596860 25 5.35 0.735 0.10740

70 25 4.06 0.746 0.1128070 35 6.49 0.753 0.126270 45 1.67 0.762 0.1370

Glucose 40 25 4.73 0.486 0.0008850 25 2.71 0.585 0.0009260 25 5.62 0.647 0.0011170 25 5.43 0.686 0.0026870 35 5.52 0.6992 0.0258770 45 5.93 0.7136 0.04963

Sucrose 40 25 7.88 0.600 0.0140150 25 6.11 0.611 0.0150860 25 6.03 0.6523 0.0161370 25 6.41 0.720 0.0171770 35 8.60 0.7289 0.0199

70 45 5.79 0.7613 0.01395Maltodextrin 40 25 5.90 0.5683 0.0831

50 25 6.06 0.5873 0.085060 25 4.12 0.6833 0.098670 25 4.10 0.7304 0.150370 35 4.26 0.7532 0.162370 45 3.81 0.7622 0.1688

Sorbitol 40 25 2.65 0.3977 0.0117850 25 4.19 0.4699 0.0128960 25 3.17 0.4884 0.016370 25 5.66 0.5572 0.02026

Solution The ratio of sampleto solution

% Relativedeviation ( D)



Glucose 1/4 2.55 0.733 0.001921/8 3.49 0.736 0.02261/12 2.21 0.738 0.04171/16 5.37 0.741 0.06031/20 6.55 0.743 0.0813

Maltodextrin 1/4 7.45 0.711 0.09721/8 8.32 0.714 0.1041/12 8.32 0.717 0.1181/16 3.15 0.722 0.1231/20 7.31 0.726 0.1237

Table 2 The effect of the geometry of sample on weight loss (ML), water loss (WL) and solid gain (SG)

Osmotic agents The geometry of apricot

ML (kg[H2O]/kg[freshfruit])



Fructose Cube 0.463 0.563 0.100Half 0.296 0.365 0.069

Whole 0.113 0.168 0.055

Glucose Cube 0.503 0.568 0.065Half 0.238 0.301 0.063

Whole 0.099 0.133 0.034

Sucrose Cube 0.385 0.498 0.113Half 0.175 0.259 0.084

Whole 0.085 0.109 0.024

Maltodextrin Cube 0.260 0.302 0.042Half 0.101 0.139 0.038

Whole 0.044 0.079 0.035

Sorbitol Cube 0.332 0.531 0.199Half 0.244 0.344 0.100

Whole 0.139 0.196 0.057


4/15


5/15


would lead to local reduction of osmotic driving force during the process.

Apricots were immersed in the osmotic solution in 850 mLglass jars. The glass jars were placed in a water bath tomaintain required temperaturesof thesyrup during theexper-iment. (NML-WB020 model, Double Walled, inside made of stainless steel 304 grade and outside mild sheet painted withattractive stove enamel. Temperature was controlled by ahydraulic thermostat from 5 C above ambient to 100 C withsensitivity of 0.5 C tted with ISI mark Tublar Element.)

The jar was covered with a plastic plate to reduce moistureloss from syrup during experiments.

All experiments were conducted within 8 days. Apricotswere withdrawn at periodic intervals of 2 h, quickly rinsedandgently blotted with tissue paper to remove excesssolutionfrom the surface, then weighed and returned to the osmoticsolution to continue the drying process. After 8 days, osmoticdehydration was nished and the nal moisture content of apricots was determined in an oven at 75 C for 24h. All theexperiments were repeatedve timesand averagevalueshavebeen reported. The water loss and solid gain at any time andat equilibrium were determined by the continuous method asdened by Azuara et al. (1998) .

The experiments divided into four groups were carried outas follows.

Group IThe effects of osmotic agent and concentration were inves-tigated using sucrose, fructose, glucose, maltodextrin andsorbitol, using sugar concentrations of 40, 50, 60 and 70%(w/w).Group II

The effect of process temperature was investigated using solution temperatures of 25,35 and45 C using sucrose, fruc-tose, glucose and maltodextrin of 70% (w/w).Group IIITo study the effect of the ratio of sample to solution, theosmotic process experiments were repeated at various ratiosof 1/4, 1/8, 1/12, 1/16, and 1/20 using glucose and maltodex-trin of 70% (w/w).Group IVTo study the effect of sample geometry, apricots were driedin various shapes as whole, half and cube. Apricots dividedinto two pieces pips were used for the half shape. Apricotscut into cube shaped pieces (10mm 10mm 10mm) were

used for cube shape. Whole apricots were used with pips.The whole apricots were osmotically dehydrated (24 h) using

sucrose, fructose, glucose, maltodextrin and sorbitol at the70% (w/w) concentration.

2.3. Continuous method in osmotic dehydration

Azuara et al. (1992) calculated water loss and solid gain dur-ingosmotic dehydrationusing equations withtwoparametersobtained from mass balances.

WL =s1t WL1 + s1t

(1)

SG = s2t SG1 + s2t (2)

where t is time, s1 a constant related to water loss, s2 a con-stant related to solid gain, WL the amount of water lost by thefoodstuff at time t (fraction, percent, g, or kg), SG the amountof solids gained by the foodstuff at time t (fraction, percent, g or kg), WL the amount of water lost at equilibrium, and SG is the amount of solids gained at equilibrium.

Massloss (ML)duringosmoticdehydrationis equal to waterlost (WL) minus solids gained (SG) at the same time.

ML= WL SG (3)

Subtracting Eq. (1) f rom Eq. (2) and rearranging, we obtain:

ML= s1t WL [1 (SG/ WL)]

1 + s1t (4)

Eq. (4) can be written in the following form:

ML= s2t SG [(WL/ SG) 1]

1 + s2t (5)

Linearizing Eqs. (4) and (5):

tML

=1

s1 WL [1 (SG/ WL)] +

tWL [1 (SG/ WL)]

(6)

tML

=1

s2 SG [(WL/ SG) 1] +

tSG [(WL/ SG) 1]

(7)

If we plot t /ML vs. t we can dene b to be the intercept and p the slope of the resulting straight line, then the following equations arise:

WL =(1/p 1)

[1 (SG/ WL)m] (8)

Fig. 2 The comparison of various osmotic agents for 70% (w/w) concentration: ( ) glucose, ( ) fructose, ( ) maltodextrin,( ) sucrose, ( ) sorbitol.


6/15


7/15


mass loss (ML)= water loss (WL) solid gain (SG)

The moisture content kg [H 2O]/kg [dry matter] at any timecan be calculated as follows:

=(initial water present-water loss)/initial dry solidsSolid content kg [dry matter]/kg [dry matter] at any time ondry basis=(initial dry solids+ solid gain)/initial dry solids

Considering the apricots as a sphere with initially uniformwater and solid contents, the solution for Ficks equation forconstant process conditions is ( Crank, 1975 )

Mr =Mt MeM0 Me

=62

i= 0

exp i2 2Dewt

r2 (12)

Sr =St Se

S0 Se=

62

i= 0

exp i2 2Dest

r2 (13)

where Mr and Sr are the moisture and solute ratio; thesubscripts 0, e and t represent the relevant concentrations ini-tially, at equilibrium, andat any time; Dew and Des are effectivediffusivities of water and solute, respectively, m 2 s 1; i is thenumber of series in the time; r is the sphere radius, m; t is thetime, s.

Considering the equilibrium approach mass transfer thefollowing equations for moisture and solid mass transfer canbe written:

dM

dt = kw(M Me) (14)

dSdt

= ks (S Se) (15)

where kw and ks are the moisture and solid mass transfercoefcients.

There are two methods, frequently used for determining transport parameters in osmotic dehydration research. Oneof them is by using diffusion approach, obtained from solu-tion of Ficks IInd law for explaining water and soluble solidmovement, the other is by using equilibrium approach equa-tion for explaining equilibrium time of osmotic dehydration.Peleg equation was used in some research instead of equilib-rium approach equation ( Azoubel and Murr, 2004; El-Aouar etal., 2003; Park et al., 2002; Palou et al., 1993; Khoyi and Hesari,2007).

The effective diffusivity and mass transfer coefcientswere determined, and then the effects of process parame-ters on these coefcients were modeled by using non-linearregression analyses. Regression analysis was done by using the statistical package Statistica 5.0 ( Statistica, 1995 ).

2.5. Data analysis

Therelative% deviation ( D) between themeanand theindivid-ualvalueswasdetermined by themeandescribedby Yanniotisand Zarmboutis (1996) .

D =100

5

5

i= 1

|xi x|xi

(16)

where xi is the normalized individual value of mass transfers(weight loss, water loss or solid gain) of the ve samples ineach time and x is their arithmetic mean.

The correlation coefcient ( R) was one of the primary crite-riafor selecting thebestequationto denea suitable model.Inaddition to R, various statistical parameters such as; reducedchi-square ( 2), mean bias error (MBE) and root mean squareerror (RMSE) were used to determine the quality of the t.

Fig. 4 (a) The effect of the ratio of sample to solution on the water loss and the solid gain by using (a) glucose, (b)maltodextrin solutions ( 1/4, 1 /8, 1/12, 1/16, 1/20, 1/25).


8/15


Fig. 5 The effect of concentration of solution on thetransport coefcients for sucrose: ( ) D ew , ( ) D es , ( ) k w , ( )k s .

Fig. 6 The effect of concentration of solution on thetransport coefcients for glucose: ( ) D ew , ( ) D es , ( ) k w , ( )k s .

These parameters can be calculated as follows:

2 =

Ni= 1(Xexp ,i Xpre ,i)

2

N n (17)

MBE= 1N

N

i= 1

(Xpre ,i Xexp ,i) (18)

Fig. 7 The effect of concentration of solution on thetransport coefcients for fructose: ( ) D ew , ( ) D es , ( ) k w , ( )k s .

Fig. 8 The effect of concentration of solution on thetransport coefcients for maltodextrin: ( ) D ew , ( ) D es , ( )k w , ( ) k s .

Fig. 9 The effect of concentration of solution on thetransport coefcients for sorbitol: ( ) D ew , ( ) D es , ( ) k w , ( )k s .

Fig. 10 The effect of temperature of solution on thetransport coefcients for sucrose: ( ) D ew , ( ) D es , ( ) k w , ( )k s .

RMSE= 1N

N

i= 1

(Xpre ,i Xexp ,i)2

1/ 2

(19)

where Xexp, i stands for the experimental values and Xpre, i arethe predicted values by calculating from the model for thesemeasurements. N and n are the number of observations andconstants, respectively ( Yaldiz and Ertekin, 2001; Yaldiz et al.,2001; Ertekin and Yaldiz, 2004; Menges and Ertekin, 2006a,2006b).


9/15


Fig. 11 The effect of temperature of solution on thetransport coefcients for glucose: ( ) D ew , ( ) D es , ( ) k w , ( )k s .

Fig. 12 The effect of temperature of solution on thetransport coefcients for fructose: ( ) D ew , ( ) D es , ( ) k w , ( )k s .

3. Results and discussion

3.1. The effect of process parameters on osmoticdehydration of apricot

The relative % deviation ( D) for experimental values and thedetermined values of the water loss and solid gain at equilib-rium by using Azuara et al. model (1998) are given in Table 1 .

Fig. 13 The effect of temperature of solution on thetransport coefcients for maltodextrin: ( ) D ew , ( ) D es , ( )k w , ( ) k s .

Fig. 14 The effect of the ratio of sample to solution ontransport parameters for glucose: ( ) D ew , ( ) D es , ( ) k w , ( )k s .

Fig. 15 The effect of the ratio of sample to solution ontransport parameters for maltodextrin: ( ) D ew , ( ) D es , ( )k w , ( ) k s .

It was seen that deviation from experimental ML of predictedML for each value are 23% approximately.

Theosmotic process was studiedin terms of waterloss andsolid gain ( Fig. 1ae). An initial high rate of water removal (andsolid uptake) followed by slower removal (and uptake) in thelatter stageswasobserved. Rapid loss of water (and solid gain)in thebeginning is apparently due to the large osmotic driving force between the dilute sap of the fresh fruit and the sur-rounding hypertonic medium. Several research groups havepublished similar curves for osmotic dehydration of foods(Azoubel andMurr, 2004; Kowalska andLenart, 2001; Lazaridesetal., 1997;Park etal., 2002 ).Increasein solutionconcentrationresulted in an increase in the osmotic pressure gradients and,hence, higher water loss (and solid uptake) values throughoutthe osmosis period were obtained. These results indicate thatby choosing a higher concentration medium, some benets interms of faster water loss could be achieved. However, a muchgreater gain of solid is observed.

The comparison of various osmotic agents for the 70%(w/w) concentrationis given in Fig. 2. Thehighest andthe low-estwater loss were obtained bysucrose andsorbitol solutions,respectively. The highest and the lowest solid gain wereobtained by maltodextrin and fructose solutions, respectively.


10/15


Table 3 The non-linear regression models (where C is % concentration, kg [solute] 100kg 1 [H2O] and T is temperaturein degree centigrade)

Model 1: Dw , Ds , kw , ks =a + bC + cT + dCT

Glucose Fructose

Dw Ds kw ks Dw Ds kw ks

a 4.554734 1.87473 1.87471 4.554743 4.554734 1.87473 4.554761 4.554761b 0.065068 0.026782 0.026785 0.065066 0.065068 0.026782 0.065059 0.065059c 0.182189 0.074989 0.074987 0.182189 0.182189 0.074989 0.182199 0.182199d 0.002603 0.001071 0.001071 0.002603 0.002603 0.001071 0.002603 0.002603R 0.98862 0.99811 0.99459 0.99725 0.99848 0.97717 0.98839 0.98728

Sucrose Maltodextrin


a 6.096846 1.87473 4.554755 4.554754 4.554734 1.87473 4.554749 4.554745b 0.243874 0.074989 0.065063 0.065064 0.065068 0.026782 0.065064 0.065067c 0.087098 0.026782 0.182203 0.182202 0.182189 0.074989 0.182182 0.182185d 0.003484 0.001071 0.002603 0.002603 0.002603 0.001071 0.002603 0.002603R 0.99545 0.99343 0.97152 0.98485 0.99481 0.98647 0.74960 0.99537

Model 2: Dw , Ds , kw , ks =a + bC +cT +dCT + eC2 + fT 2 +gC2T 2

Glucose Fructose


b 1880.35 1880.32 1880.46 1880.44 1878.04 1880.32 1880.67 1880.77c 5264.99 5264.90 5265.29 5265.22 5258.51 5264.89 5265.87 5266.14d 75.2142 75.2128 75.2185 75.2175 75.1216 75.2127 75.2267 75.2306e 26.9115 26.9110 26.9130 26.9127 26.8785 26.9109 26.9159 26.9174 f 210.986 210.982 210.998 210.995 210.727 210.982 211.021 211.032g 0.04306 0.0431 0.04306 0.04306 0.04301 0.04306 0.04307 0.04307R 0.21765 0.11954 0.99961 0.99994 0.05781 0.16491 0.98881 0.98737



a 241.460 241.462 174.749 174.677 242.688 241.482 241.479 241.505b 1880.33 1880.35 3785.74 3784.45 1879.40 1880.34 1880.49 1880.48c 5264.92 5264.99 1352.05 1351.59 5262.31 5264.96 5265.38 5265.36d 75.2132 75.2142 54.08202 54.06359 75.1758 75.2138 75.2197 75.2194e 26.9111 26.9115 151.7093 151.6575 26.8981 26.9113 26.9135 26.9134 f 210.983 210.986 19.35067 19.34407 210.881 210.985 211.001 211.001g 0.04306 0.04306 0.030961 0.030951 0.04304 0.04306 0.04306 0.04306R 0.13789 0.08731 0.99995 1.0000 0.15872 0.18469 0.85033 0.99951

Model 3: Dw , Ds , kw , ks =a +exp(bC) +cexp(dT )

Glucose Fructose

Dw Ds kw ks Dw Ds kw ksa 8.68 10 11 7.77 10 11 0.000207 0.00018 1 10 10 1.352 10 10 0.000559 0.000555b 6.06 1010 5.89 1010 0.227975 1.02867 5.846 1010 5.903 1010 0.522547 0.517988c 0.01226 0.00996 0.201269 0.158659 0.00881 0.007933 0.438190 0.434685d 1.89454 1.87576 1.15970 1.16509 1.86615 1.87605 1.07646 1.07758R 8.68 10 7 2.60 10 6 0.00050 0.00438 1.382 10 6 1.07 10 6 0.12054 0.01057



a 1.336 10 10 1.321 10 10 0.000420 0.000400 1.315 10 10 9.29 10 10 0.00031 0.00022b 5.857 1010 6.026 1010 0.976196 0.970776 5.92 1010 5.809 1010 1.01104 0.985508c 0.009535 0.009448 0.156578 0.178109 0.008217 0.009935 0.159253 0.207968d 1.86888 1.889054 1.16755 1.16204 1.881576 1.86843 1.16625 1.15909

R 9.71 10 7

7.62 10 7

0.00013 0.00278 1.481 10 6

1.839 10 6

0.00031 0.00489


11/15


Table 3 ( Continued )

Model 4: Dw , Ds , kw , ks =a exp(bC)cexp(dT )

Glucose Fructose


a 0.324905 0.485584 0.004508 0.000306 0.338990 0.256057 0.226624 0.239277b 0.034607 0.004546 0.020407 0.015075 0.009048 0.010543 0.025333 0.025331c 0.118861 0.160613 0.013225 0.160924 0.136141 0.222972 0.000495 0.000473d 0.884157 0.841696 0.003036 0.011977 0.805784 0.822495 0.001545 0.001022R 0.12018 0.18175 0.12574 0.05470 0.03187 0.15687 0.98297 0.98182



a 0.482414 0.400968 0.000475 0.000358 0.497008 0.212239 0.310109 0.141862b 0.071111 0.020714 0.031236 0.032127 0.007011 0.139044 0.005892 0.006265c 0.034345 0.082276 0.089728 0.094024 0.133728 0.075537 0.000541 0.000571d 0.930431 0.825776 0.011283 0.015280 0.819473 1.14207 0.008763 0.020638R 0.12187 0.04268 0.94609 0.96818 0.17250 0.14587 0.74658 0.99797

Model 5: Dw , Ds , kw , ks =a + bCc

+dT e

Glucose Fructose


a 0.229554 0.229554 0.238670 0.213074 0.229554 0.229554 0.225696 0.225537b 0.108972 0.108972 0.237827 0.113634 0.108972 0.108972 0.109642 0.109695c 0.027150 0.027150 0.000901 0.006037 0.027150 0.027150 0.006002 0.005984d 0.099765 0.099765 0.000092 0.099696 0.099765 0.099765 0.113636 0.113472e 0.022948 0.022948 0.225901 0.008833 0.022948 0.022948 0.000688 0.000569R 0.01785 0.1867 0.98719 0.17936 0.18362 0.05387 0.98665 0.98522



a 0.229554 0.229554 0.245619 0.246360 0.229554 0.229554 0.258097 0.227564b 0.108972 0.108972 0.109520 0.110305 0.108972 0.108972 0.136737 0.226777c 0.027150 0.027150 0.005783 0.005256 0.027150 0.027150 0.000776 0.000270d 0.099765 0.099765 0.133206 0.133117 0.099765 0.099765 0.120929 0.000352e 0.022948 0.022948 0.001559 0.002110 0.022948 0.022948 0.000764 0.228423R 0.02143 0.04763 0.98202 0.99043 0.24387 0.05478 0.75619 0.99123

Model 6: Dw , Ds , kw , ks =a(Cb)c(T d)

Glucose Fructose


a 0.003338 0.004103 0.000001 0.001210 0.006686 0.000887 0.004469 0.000007b 2.61050 2.45789 1.120001 0.824896 2.49732 2.52833 0.863366 1.428801c 0.011771 0.009971 1.117997 0.001210 0.006926 0.001433 0.000805 0.188723d 0.998816 1.14593 0.127908 0.421062 0.953646 0.180748 0.435582 0.051186R 0.01254 0.031465 0.99551 0.99778 0.03147 0.04281 0.92232 0.98670



a 0.001238 0.001238 0.000001 0.000001 0.002062 0.007544 0.111819 0.000005b 2.71674 2.71674 1.800353 1.835633 2.73828 2.76921 0.348929 0.321983c 0.008913 0.008913 0.140126 0.040570 0.012830 0.016999 0.000260 0.996897d 0.313957 0.313957 0.380431 0.522013 0.525229 1.10821 0.280907 0.707941R 0.14263 0.13054 0.95874 0.97595 0.08472 0.04235 0.75362 0.99408

Model 7: Dw , Ds , kw , ks =a +b(Cc)d(T e)

Glucose Fructose


a 1.404 10 9 3.124 10 10 0.012427 0.000628 9.284 10 10 1.175 10 9 0.012711 0.009556b 0.579510 0.5795339 0.635885 0.694261 0.579505 0.57949 0.421029 0.421647


12/15


13/15


Table 4 Statistical analyses results and the best models expressed the relationship between transport properties andsolution concentration and temperature (where C is % concentration, kg[solute] 100kg 1 [H2O] and T is temperature indegree centigrade)

R MBE RMSE 2

GlucoseDw = 4.5547 0.06507C 0.1822T +0.0026CK 0.9886 1.71 10 4 6.26 10 2 2.20 10 23

Ds = 1.87473+ 0.02678C+ 0.07499T 0.001071CK 0.9981 1.45 10 4

4.40 10 3

1.12 10 25

kw = 241.486 1880.46C 5265.29T +75.2185TC+26.913C2

+ 210.998T 2 0.04306C2T 20.9996 5.53 10 6 7.73 10 3 2.21 10 12

ks = 24.486 1880.44C 5265.22T +75.2175TC+ 26.9127C2

+ 210.995T 2 0.04306C2T 20.9999 3.29 10 5 2.95 10 3 2.08 10 13

FructoseDw = 4.5547 0.06507C 0.1822T +0.002.6TC 0.9985 9.54 10 4 6.82 10 3 1.04 10 24

Ds =1.87473+ 0.02678C+0.07499T 0.001071TC 0.9772 1.86 10 3 4.32 10 2 2.72 10 23

kw = 241.5 1880.67C 5265.87T + 75.2267TC+26.916C2

+ 211.021T 2 0.04307C2T 20.9888 1.87 10 3 5.10 10 2 5.33 10 10

ks = 241.520 1880.77C 5266.14T +75.2306TC+26.9174C2

+ 211.032T 2 0.04307C2T 20.9874 2.23 10 3 5.38 10 2 5.85 10 10

SucroseDw = 6.09685 0.24387C 0.087098T +0.00348TC 0.9955 9.55 10 4 3.54 10 2 1.15 10 23

Ds = 1.87473+ 0.07499C+ 0.026782T 0.001071TC 0.9934 1.65 10 3 4.09 10 2 1.58 10 23

kw = 174.749 3785.74C 1352.05T + 54.08202TC+ 151.7093C2+ 19.35067T 2 0.030961C2T 2

0.9999 8.57 10

4 5.86 10

3 3.04 10

12

kw = 174.67 3784.45C 1351.59T + 54.06359TC+ 151.6575C2

+ 19.34407T 2 0.030951C2T 21.0000 3.01 10 4 5.20 10 4 9.34 10 6

MaltodextrinDw = 4.554734 0.06507C 0.1822T + 0.0026TC 0.9948 1.65 10 3 9.35 10 3 1.55 10 24

Ds = 1.87473+ 0.02678C+ 0.07499T 0.001071TC 0.9865 6.17 10 4 2.44 10 2 4.50 10 24

kw = 241.5 1880.67C 5265.87T + 75.2267KC+26.916C2

+ 211.021T 2 0.04307C2T 20.8503 6.16 10 2 1.14 10 1 9.14 10 10

kw = 241.505 1880.48C 5265.36T + 75.21939TC+ 26.91336C2

+ 211.0007T 2 0.043061C2T 20.9995 2.79 10 4 7.17 10 3 1.93 10 12

SorbitolDw = 9.24 10 11 + 1.356 10 12C+ 7 10 15 exp(0.107075 C) 0.9941 1.55 10 3 7.80 10 3 1.53 10 24

Ds = 1.47 10 10 6.85 10 13C+ 1.2510 14C2 0.9950 3.37 10 4 3.37 10 4 2.50 10 27

kw = 2.21 10 4 1295.96C 4.3718 0.9944 2.39 10 3 2.56 10 2 2.40 10 11

kw = 1.42 10 4 31.8994C 3.5337 0.9875 1.93 10 3 3.20 10 2 1.47 10 11

Typical temperature effect on osmotic dehydration of apri-cots is presented in Fig. 3ad for the 70% (w/w) solutions. Itwas observed that temperature has increasing effect on theosmotic dehydration of apricot. The increasing of osmoticmedium temperature caused increased water loss and solidgain. The temperature effect is similar in all osmotic agents.This increasing temperature inuencehasbeenclearly shownin previous works ( Sereno et al., 2001; Lazarides et al., 1995;Lenart and Flink, 1984 ).

The effect of ratio of sample to solution is shown in Fig. 4aand b using glucose and maltodextrin of 70% (w/w). The waterloss and solid gain are considerably increased when the ratioof sample to solution is decreased. The decreasing of the ratioof sample to solutionavoidssignicantdilutionof themediumby water removal and subsequent decrease of osmotic driving force during the process.

Theeffect of sample geometry is given in Table2 . Thewaterloss andsolidgain was increasedwhen the dimension of apri-cot is decreased becauseof increasing contactsurfaceareaanddeforming of the cell.

The results obtained in sucrose medium for determining the effect of osmotic dehydration medium and temperature,the ratio of sample to solution, realized by Khoyi and Hesari(2007), are similar with the results of this study.

3.2. Osmotic dehydration kinetics

The effective diffusivitiesand the masstransfer coefcients of water loss andsolid gain werecalculatedby Eqs. (12)(15)using

experimental values. The effect of concentration and temper-ature on transport parameters (diffusivities and masstransfercoefcients) is given in Figs. 59 and Figs. 1013, respectively,for all of the osmotic agents.

Osmotic pressure gradient is the driving force for osmoticmass transfer. This driving force in turn depends on concen-tration and temperature of the osmotic solution. An increasein osmotic solution concentration increases this gradient andin turn the driving force. The transport coefcients for waterloss and solid gain ( De and k) increases with an increase inosmotic solution concentration due to change in the physicalproperties of the food (such as porosity and cell permeabiliza-tion). On the other hand, the transport coefcient increaseswith increase in temperature. The diffusivity dependence onthe temperature can be represented by the Arrhenius equa-tion (De = D0 exp( E /RT )), where Ea is the activation energy(kJmol 1), D0 the Arrhenius factor (m 2 s 1), and T is the abso-lute temperature (K).

The effective diffusion and the mass transfer coefcientsfor water ( Dew , kw) and solid ( Des , ks ) which are obtainedfor different combinations of concentration and temperatureof osmotic solution were modeled by using nine non-linearregression models. The model results are given in Table 3 . Thebest models obtained for transport parameters and statisti-cal analysis results are given in Table 4 . It can be seen fromTable 4 that the value of effective diffusion and mass trans-fer coefcients for water loss and solid gain were found tobe dependent on the concentration and temperature of theosmotic solution in addition to the combined effect of both of


14/15


Table 5 Statistical analyses results and the models expressed the relationship between transport properties and theratio of sample to solution

Dw , Ds , kw , ks =a(Ratio)b

Glucose

a b R MBE RMSE 2

Dw 6 10 11 0.1749 0.9545 1.27 10 2 0.0362 9.945 10 24Ds 3 10 12 1.0525 0.9944 1.23 10 1 0.1269 4.447 10 23

kw 7 10 5 0.4204 0.9944 4.51 10 2 0.0536 1.020 10 10

ks 4 10 5 0.4382 0.9483 6.42 10 3 0.0889 1.454 10 10

Maltodextrin

a b R MBE RMSE 2

Dw 4 10 11 0.3158 0.9556 0.0377 0.0913 5.28 10 23

Ds 1 10 11 0.5466 0.9013 0.1188 0.4084 1.05 10 22

kw 6 10 5 0.5001 0.9966 0.0839 0.1022 5.06 10 10

ks 4 10 5 0.4778 0.9101 0.0968 0.1396 2.51 10 10

these parameters. The given equations in Table 4 can be con-dently used for explaining the effect of concentrations andtemperatures used in these experiments.

The effect of the ratio of sample to solution on transportparameters (diffusivities and mass transfer coefcient), forglucose and maltodextrin solutions is given in Figs. 14 and 15.The models and statistical analysis results explained the rela-tionshipbetweenthe ratio of sample to solutionandtransportproperties are given in Table 5 .

4. Conclusions

The osmotic dehydration of apricot at various conditions hasbeen studied.The effects of concentrationand temperature of solution, the ratio of sample to solution and the geometry of sample on the water loss and solid gain were investigated.

The rate of water loss and solid gain in the osmotic dehy-dration of apricot was directly related to the concentrationand temperature of solution, the ratio of sample to solu-tion and the geometry of sample. The water loss and solidgain were increased with increasing concentration and tem-perature, decreasing the ratio of sample to solution and thegeometry of sample.

The osmotic dehydration kinetics by using effective diffu-sivitiesandmasstransfer coefcientswere determined. Thesetransport coefcients have been empirically correlated withthe concentration, temperature and the ratio of sample tosolution. The moisture and solid content at any immersiontimeduring thecourse of osmotic dehydrationof apricot couldbe predicted with sufcient accuracy by using diffusivities ormasstransfer coefcientscalculated fromthe proposedmodelequations.

Acknowledgement

This study was supported by the Research Foundation of FiratUniversity (project no.: FUBAP-1145).

References

Azoubel, P.M and Murr, F.E.X., 2004, Mass transfer kinetics of osmotic dehydration of cherry tomato. J Food Eng, 61: 291295.

Azuara, E., Beristain, C.I. and Gutierrez, G.F., 1998, A method forcontinuous kinetics evaluation of osmotic dehydration.Lebensmittel- Wissenschaft und- Technologie, 31: 317321.

Azuara, E., Cortes, R., Garcia, H.S. and Beristain, C.I., 1992, Kineticmodel for osmotic dehydration and its relationship with FicksSecond Law. Int J Food Sci Technol, 27: 409418.

Beristain, C.I., Azuara, E., Cortes, R. and Garcia, H.S., 1990, Masstransfer during osmotic dehydration of pine-apple rings. IntFood Sci Technol, 25: 576582.

Crank, J., (1975). The Mathematics of Diffusion (2nd ed.). (ClarendonPress, Oxford).

Dixon, G.M. and Jen, J.J., 1977, Changes of sugar and acid of osmovac dried apple slices. J Food Sci, 42: 11261131.

Doymaz, I., 2004, Effect of pre-treatments using potassiummetabisulphide and ethyl oleate on drying kinetics of apricots. Biosyst Eng, 89(3): 281287.

El-Aouar, A.A., Azoubel, P.M. and Murr, F.E.X., 2003, Drying kinetics of fresh and osmotically pre-treated papaya (Carica

papaya L.). J Food Eng, 59(1): 8591.Ertekin, C. and Yaldiz, O., 2004, Drying of eggplant and selectionof a suitable thin layer drying model. J Food Eng, 63(3):349359.

Forni, E., Sormani, A., Scalise, S. and Torreggiani, D., 1997, Theinuence of sugar composition on the colour stability of osmodehydrofrozen intermediate moisture apricots. Food ResInt, 30(2): 8794.

Hawkes, J. and Flink, J.M., 1978, Osmotic concentration of fruitsslices prior to freeze dehydration. J Food Process Preserv, 2:265284.

Kaymak-Ertekin, F. and Cakaloz, T., 1996, Osmotic dehydration of peas: I. Inuence of process variables on mass transfer. J FoodProcess Preserv, 20: 87104.

Kaymak-Ertekin, F. and Cakaloz, T., 1996, Osmotic dehydration of

peas: II. Inuence of osmosis on drying behaviour andproduct quality. J Food Process Preserv, 20: 105119.Kaymak-Ertekin, F. and Sultano glu, M., 2000, Modelling of mass

transfer during osmotic dehydration of apples. J Food Eng, 46:243250.

Khoyi, M.R. and Hesari, J., 2007, Osmotic dehydration kinetics of apricot using sucrose solution. J Food Eng, 78: 13551360.

Kowalska, H. and Lenart, A., 2001, Mass exchange during osmoticpretreatment of vegetables. J Food Eng, 49(23):137140.

Lazarides, H.N., Katsanidis, E. and Nickolaidis, A., 1995, Masstransfer kinetics during osmotic preconcentration aiming atminimal solid uptake. J Food Eng, 25(2): 151166.

Lazarides, H.N., Gekas, V. and Mavroudis, N., 1997, Apparentmass diffusivities in fruit and vegetable tissues undergoing

osmotic processing. J Food Eng, 31(3): 315324.Lenart, A., 1996, Osmo-convective drying of fruits and vegetables:technology and Application. Drying Technol, 14: 391413.

Lenart, A. and Flink, J.M., 1984, Osmotic dehydration of potato. II.Spatial distribution of the osmotic agent. J Food Technol, 19:6589.


15/15


Lin, T.M., Durance, T.D. and Scaman, C.H., 1998, Characterizationof vacuum microwave air and freeze dried carrot slices. FoodRes Int, 4: 111117.

Mandala, I.G., Anagnostaras, E.F. and Oikonomou, C.K., 2005,Inuence of osmotic dehydration conditions on appleair-drying kinetics and their quality characteristics. J FoodEng, 69(3): 307316.

Menges, H.O. and Ertekin, C., 2006, Thin layer drying model fortreated and untreated Stanley plums. Energy ConversManage, 47(1516): 23372348.

Menges, H.O. and Ertekin, C., 2006, Mathematical modeling of thin layer drying of Golden apples. J Food Eng, 77(1): 119125.

Palou, E., Lopez-Malo, A., Argaiz, A. and Welti, J., 1993, Osmoticdehydration of papaya. Effect of syrup concentration. RevistaEspasimnola de Ciencia Y Technologia de Alimentos, 33(6):621630.

Panagiotou, N., Karathanos, V. and Maroulis, Z., 1998, Masstransfer modeling of the osmotic dehydration of some fruits.Int J Food Sci Technol, 33: 267284.

Park, K.J., Bin, A., Brod, F.P.R. and Park, T.H.K.B., 2002, Osmoticdehydration kinetics of pear Danjou (Pyrus communis L.). JFood Eng, 52: 293298.

Piotrowski, D., Lenart, A. and Wardzynski, A., 2004, Inuence of osmotic dehydration on microwave-convective drying of frozen strawberries. J Food Eng, 65(4): 519525.

Ponting, J.D., 1973, Osmotic dehydration of fruits: recentmodications and applications. Process Biochem, 8: 1820.

Rastogi, N.K. and Raghavarao, K.S.M.S., 1994, Effect of temperature and concentration of osmotic dehydration of coconut. Lebensmittel- Wissenschaft und- Technologie, 27:264267.

Rastogi, N.K. and Raghavarao, K.S.M.S., 1995, Kinetics of osmoticdehydration of coconut. J Food Process Eng, 18: 187197.

Rastogi, N.K. and Raghavarao, K.S.M.S., 1997, Water and solutediffusion coefcients of carrot as a function of temperatureand concentration. J Food Eng, 34: 429440.

Rastogi, N.K., Raghavarao, K.S.M.S. and Niranjan, K., 1997, Masstransfer during osmotic dehydration of banana: Fickiandiffusion in cylindrical conguration. J Food Eng, 31: 423432.

Rastogi, N.K., Raghavarao, K.S.M.S., Niranjan, K. and Knorr, D.,2002, Recent developments in osmotic dehydration: methodsto enhance mass transfer. Trends Food Sci Technol, 13(2):5869.

Riva, M., Campolongo, S., Leva, A.A., Maestrelli, A. andTorreggiani, D., 2005, Structureproperty relationships inosmo-air-dehydrated apricot cubes. Food Res Int, 38(5):533542.

Salvatori, D., Andrs, A., Chiralt, A. and Fito, P., 1999, Osmoticdehydration progression in apple tissue I: spatial distributionof solutes and moisture content. J Food Eng, 42(3): 125132.

Sereno, A.M., Moreira, R. and Martinez, E., 2001, Mass transfercoefcients during osmotic dehydration of apple in single andcombined aqueous solutions of sugar and salt. J Food Eng,47(1): 4349.

Shi, J. and Le Maguer, M., 2002, Osmotic dehydration of foods:mass transfer and modeling aspects. Food Rev Int, 18(4):305335.

Spiazzi, E. and Mascheroni, R., 1997, Mass transfer model forosmotic dehydration of fruits and vegetables. I. Developmentof the simulation model. J Food Eng, 34: 387410.

Statistica for Windows 5.0, 1995. Computer program manual.StatSoft Inc., Tulsa.

To grul, I.T. and Pehlivan, D., 2002, Mathematical modeling of solardrying of apricots in thin layers. J Food Eng, 55(3): 209216.

To grul, I.T. and Pehlivan, D., 2003, Modelling of drying kinetics of single apricot. J Food Eng, 58(1): 2332.

To grul, I.T. and Pehlivan, D., 2004, Modelling of thin layer drying kinetics of some fruits under open-air sun drying process. JFood Eng, 65(3): 413425.

Torreggiani, D., 1993, Osmotic dehydration in fruits and vegetableprocessing. Food Res Int, 26: 5968.

Torreggiani, D., Maltini, E., Bertolo, G. and Mingardo, F., 1988,Frozen intermediate moisture fruit: studied on techniquesand product properties, In Proceedings of Symposium on Progressin Food Preservation Process , pp. 7178.

Toupin, C.J., Marcotte, M. and Le Maguer, M., 1989, Osmoticallyinduced mass transfer in plant storage tissues: amathematical model. Part 1. J Food Eng, 10: 1338.

Trelea, I.C., Raoult-Wack, A.L. and Trystram, G., 1997, Note:Application of neural network modeling for the control of dewatering and impregnation soaking process (osmoticdehydration). Food Sci Technol Int, 3: 459465.

Yaldiz, O. and Ertekin, C., 2001, Thin layer solar drying of somedifferent vegetables. Drying Technol-An Int J, 19: 583596.

Yaldiz, O., Ertekin, C. and Uzun, H.I., 2001, Mathematicalmodeling of thin layer solar drying of sultana grapes. Energy,26(5): 457465.

Yanniotis, S. and Zarmboutis, I., 1996, Water sorption isotherm of pistachio nuts. Lebensmittel- Wissenschaft und- Technologie,29: 372375.

Yao, Z. and Le Maguer, M., 1996, Mathematical modeling andsimulation of mass transfer in osmotic dehydrationprocesses. Part I: Conceptual and mathematical model. J FoodEng, 29: 349360.

osmotic dehydration of apricot

Documents