microstructural changes during osmotic dehydration of

Available online at www.sciencedirect.com

www.elsevier.com/locate/jfoodeng

Journal of Food Engineering 85 (2008) 326–339

Microstructural changes during osmotic dehydrationof parenchymatic pumpkin tissue

L. Mayor a, J. Pissarra b, A.M. Sereno a,*

a REQUIMTE, Department of Chemical Engineering, Faculty of Engineering, University of Porto, Rua Dr. Roberto Frias s/n,

4200-465 Porto, Portugalb Department of Botany, Faculty of Sciences and Functional Plant Biology Unit – Institute for Molecular and Cell Biology,

University of Porto, Rua do Campo Alegre, 823, 4150-180 Porto, Portugal

Received 30 January 2007; received in revised form 17 June 2007; accepted 29 June 2007Available online 25 August 2007

Abstract

Pumpkin (Cucurbita pepo L.) fruit was selected to study the microstructural changes of vegetable tissue during osmotic dehydration.Cylinders (diameter: 1.5 cm; length 2.5 cm) of the parenchymatic tissue were dehydrated with 45% (w/w) sucrose solutions at 25 �C. Themicrostructure of fresh and processed samples was observed using light microscopy techniques. Fresh pumpkin cells showed averagevalues of 0.015 mm2, 0.469 mm and 0.136 mm for cell area, cell perimeter and cell equivalent diameter, respectively; and 1.288, 0.831and 0.871 for cellular elongation, roundness and compactness, respectively. The main modifications observed during osmotic dehydra-tion were shrinkage of cells, plasmolysis and folding of the cell walls. These changes led to the decrease of cellular area, equivalent diam-eter, roundness and compactness; elongation of cells increased whereas the perimeter was maintained essentially constant along theprocess. The observed changes were not homogeneously distributed in the material, and were dependent on cell location in the sampleand on process time. Empirical quadratic functions were used to relate the average cellular shape and size parameters with the dehydra-tion parameters water loss, weight reduction and normalized moisture content. The equations showed a good fit of the experimental data,leading to correlation coefficients ranging 0.93–0.99 and average relative deviations ranging 0.7–2.8%.� 2007 Elsevier Ltd. All rights reserved.

Keywords: Light microscopy; Microstructure; Pumpkin; Vegetable cells; Shape factors; Shrinkage

1. Introduction

Osmotic dehydration is a common water removal pro-cess applied to food materials that consists of placingpieces of a biological tissue, such as a fruit or a vegetable,in a hypertonic solution. Since this solution has higherosmotic pressure and hence lower water activity, a drivingforce for water removal arises between solution and food,while the natural cell membrane acts as a semipermeablemembrane. The diffusion of water is accompanied by thesimultaneous counter diffusion of solutes from the osmoticsolution into the tissue. Since the membrane responsible for

0260-8774/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jfoodeng.2007.06.038

* Corresponding author. Fax: +351 22 508 1449.E-mail address: [email protected] (A.M. Sereno).

osmotic transport is not perfectly selective, other solutespresent in the cells can also be leached in the osmotic solu-tion (Giangiacomo, Torreggiani, & Abbo, 1987).

The technique is often used as a pre-treatment of otherprocesses (Tregunno & Goff, 1996), to produce intermedi-ate moisture foods with improved shelf life characteristics(Monsalve-Gonzalez, Barbosa-Canovas, & Cavalieri,1993), or as a pre-treatment to reduce the energy consump-tion and/or heat damage in other traditional dehydrationprocesses (Hawkes & Flink, 1978; Kim & Toledo, 1987).

When a fruit or vegetable is submitted to a dehydrationprocess, associated heat and mass transfer gradients pro-duce changes in the chemical, physical and structural char-acteristics of the plant tissue, such as changes in volume/porosity (Lozano, Rotstein, & Urbicain, 1983; Mayor &

mailto:[email protected]

Nomenclature

a constant of Eq. (7)A area m2

ARD average relative deviationb constant of Eq. (7)c constant of Eq. (7)C compactnessE elongationm sample mass kgM moisture content, wet basis kgwater/kgsample

NMC normalized moisture contentP perimeter mR roundness

R2 square correlation coefficients solids mass kgSG solids gain kg/kgw water mass kgWL water loss kg/kgWR weight reduction kg/kgX independent variable of Eq. (7)Y dependent variable of Eq. (7)

Subscriptso initial

Vacuole CytoplasmPlasma membrane

CellMiddle lamella

a b

c dFig. 1. Changes of a plant cell at microstructural level during dehydra-tion: (a) fresh cell; (b) shrinkage and plasmolysis; (c) cell to cell debondingand (d) cell rupture and cavity formation.

L. Mayor et al. / Journal of Food Engineering 85 (2008) 326–339 327

Sereno, 2004), changes in mechanical properties (Telis, Tel-is-Romero, & Gabas, 2005) and colour changes (Krokida,Karathanos, & Maroulis, 2000). The knowledge and pre-diction of these changes are important because they arerelated to quality factors (Perera, 2005) and some aspectsof food processing, such as food classification (Rahman,2005), process modelling and design of equipment (Nes-vabda, 2005; Rao & Quintero, 2005).

Most of these changes, although observed at a macro-scopic level, are caused by changes occurred at microstruc-tural/cellular level. In this way, the study of themicrostructural changes during dehydration is an impor-tant task in order to understand and predict the changesoccurred in the physical–chemical properties at higher lev-els of structure.

A fresh plant tissue is composed by cells connected oneto each other by the middle lamella (Fig. 1). These cells arein turgor pressure, defined as the hydrostatic internal pres-sure (normally 1–8 bars) exerted by the vacuole and thecytoplasm against the plasma membrane and cell wall(Aguilera, Cuadros, & del Valle, 1998). The turgor pressuregives elastic mechanical characteristics to the plant tissue.The cellulose of the cell wall gives rigidity and strength tothe tissue, whereas pectin and hemicellulose of the middlelamella give plasticity and dictate the degree which the cellscan be pulled apart during deformations (Lewicki & Paw-lak, 2003).

When plant tissue is placed in a hypertonic solution (asin the case of osmotic dehydration) water will leave the cellby osmosis. As a result the vacuole and the rest of the pro-toplasm will shrink, causing the plasma membrane to pullaway from the cell wall. This phenomenon is known asplasmolysis (Fig. 1b) (Raven, Evert, & Eichhorn, 1999),and it has been observed during osmotic dehydration ofpotato (Mauro, Tavares, & Menegalli, 2002) and straw-berry (Ferrando & Spiess, 2001). Plasmolysis is accompa-nied with a loss in the turgor pressure, shrinkage anddeformation of cells (cell wall and plasma membrane),and concentration of the protoplasmatic liquid phase. Cel-lular shrinkage during dehydration has been observed dur-

ing osmotic dehydration of apple (Lewicki & Porzecka-Pawlak, 2005) and convective drying of grapes (Ramos,Silva, Sereno, & Aguilera, 2004).

Other phenomenon that can be observed during dehy-dration is the detachment of the middle lamella, or cell deb-onding (Fig. 1c). This phenomenon is likely due to thedegradation or denaturation of the components of the mid-dle lamella, as well as to the microstresses produced in thecellular tissue due to water removal. This phenomenon hasan influence on the mechanical properties of the product,as well as in the porosity of the material because intercellu-lar spaces are formed. Cell debonding has been observed

0.5-1 mm

1 2

3 4

65

Fresh orprocessed

sampleDirect

stainingFixation/embedding

before staining

Fig. 2. Sample preparation for direct staining or fixation/embeddingbefore staining.

328 L. Mayor et al. / Journal of Food Engineering 85 (2008) 326–339

during osmotic dehydration of apples (Lewicki & Porze-cka-Pawlak, 2005).

Cell rupture, formation of cavities and shrinkage of cellswere observed during convective drying, puff drying andfreeze drying of apples (Lewicki & Pawlak, 2003). Cell rup-ture (Fig. 1d) is due to cell membrane and cell wall degra-dation and microstresses due to water removal. Cellrupture leads to the formation of cavities of different sizeand shape. This formation of cavities increases the porosityof the product.

Pumpkins (Cucurbita sp.), also known as squashes orgourds, are plants of the family Cucurbitaceae, a wide fam-ily that includes other fruits such as melon, watermelonand cucumber. Pumpkin fruits are consumed both imma-ture and ripe. The flesh of the fruit can be processed inseveral ways, such as boiled, canned, dried and pickled,leading to products of different organoleptic characteristics(Teotia, 1992). The use of an osmotic process can be aninteresting tool in the production of new processed prod-ucts based on fresh pumpkin fruits, improving its alreadygood shelf life.

The aim of this work was to study the microstructuralchanges during osmotic dehydration of the parenchyma(hydrenchyma) of pumpkin fruits. The changes in some cel-lular size and shape parameters were obtained by lightmicroscopy techniques and image analysis. Mathematicalmodels were used in order to predict the changes in micro-structure during osmotic dehydration.

2. Materials and methods

2.1. Preparation of samples

Pumpkin fruits (Cucurbita pepo L.) were purchased froma local producer, and stored at 15–20 �C in a chamber untilprocessing. Fruits with similar initial moisture content (95–97 kg water/100 kg product) and soluble solids (2–4 Brix)were selected for the experiments. Cylinders (25 mm length,15 mm diameter) from the parenchymatic tissue wereobtained employing a metallic cork borer and a cutter. Inorder to obtain a good structural and compositional homo-geneity in the samples, the cylinders were taken from themiddle zone of the mesocarp, parallel to the major axisof the fruit.

2.2. Dehydration experiments

Pumpkin cylinders were dehydrated in 45% (w/w)sucrose solutions at 25 �C. These solutions were preparedwith distilled water and commercial sucrose. The cylinderswere put in baskets, which were introduced in stirred ves-sels containing the osmotic solution. Agitation was con-ducted using a magnetic stirrer; the speed was chosenaccording to the kinematic viscosity so as to have a con-stant Reynolds number (c.a. 3000). The weight ratio ofosmotic solution to pumpkin cylinders was 20:1, allowingsolution to maintain a constant concentration during the

dehydration process. Thermoregulation was obtained bymeans of a thermostatic bath (±0.2 �C).

The samples were removed from the beakers containingthe osmotic solution at different process times (0.5, 1, 3, 6and 9 h), then they were gently blotted with paper toremove the excess of the osmotic solution and kept in plas-tic boxes till experimental determinations. Nine cylinderswere removed at each process time; four were used fordetermination of dehydration kinetic parameters and fivefor microscopic analysis.

2.3. Experimental determinations

2.3.1. Dehydration kinetics parametersAfter the osmotic treatment, each of the four samples

used for the determination of the dehydration kineticparameters was weighed to determine its weight reduction(WR) (Eq. (1)). The dry solids of the sample were deter-mined after vacuum drying at less than 104 Pa at 70 �C tillconstant weight (AOAC, 1984). Then the solids gain (SG)(Eq. (2)) were calculated by

WR ¼ mo � mmo

; ð1Þ

SG ¼ s� so

mo

: ð2Þ

Water loss (WL) and normalized moisture content(NMC) were determined by means of the Eqs. (3) and(4), respectively

WL ¼ SGþWR; ð3Þ

NMC ¼ MMo

¼ w=mwo=mo

: ð4Þ

2.3.2. Light microscopy

The preparation of samples for microscopy analysis wasdone by two different methods. In the first method, a rect-angular slab of ca. 0.5–1 mm of thickness was gently cutparallel to the height of the fresh or processed cylinders


at the maximum section area with a razor blade, as shownin Fig. 2. Then the samples (three at each process time)were stained with a solution of Methylene Blue 0.1% dur-ing 15 s.

In the other method, a slab from the sample (two at eachprocess time) was obtained as explained before. The slabwas divided in four symmetrical cuts. These quarters weredivided in six parts, as shown in Fig. 2. When cylindersshowed considerable shrinkage after dehydration (after 6and 9 h of process), the quarters were divided in four parts.These parts were fixed in 2.5% glutaraldehyde in 1.25%PIPES buffer at pH 7–7.2 during 24 h at room temperature.After that they were dehydrated in a water/ethanol seriesand embedded in LR White resin (London Resin Co.,Basingstoke, UK). After the samples were embedded inresin, semi thin sections (0.6 lm) of the resin blocks wereobtained with a microtome (mod. Reichert-Supernova,Leica, Wien, Austria). The sections were stained with anaqueous solution Azure II 0.5%, Methylene Blue 0.5%,Borax 0.5% (Ramalho-Santos et al., 1997) during 30 s.After that, they were washed with distilled water andmounted in a glass slide.

Microimages of stained samples were obtained under astereomicroscope (Olympus SZ-11, Tokyo, Japan). Thestereomicroscope was calibrated with a stage micrometerof 2 mm length and divisions of 0.01 mm interval (LeitzWetzlar, Germany). The samples were put on a thin glassplate under the objective; a source of light was locatedunder the plate so as to work in transmitted light mode.A digital colour videocamera (SONY SSC-DC50AP,Tokyo, Japan) was attached to the microscope and con-nected to a personal computer. Image acquisition was donewith an interface (PCTV videocard, Pinnacle SystemsGmbH, Munich, Germany).

2.3.3. Image analysis

From a microphotograph of the pumpkin tissue, thecontour of some cells was highlighted with the softwareMicrosoft Photo Editor 3.0 (Microsoft Corporation), ascarried out by Mayor, Silva, and Sereno (2005). This pro-cedure was systematically done in three vertical lines and

Fig. 3. Microscopy images of fresh parenchymatic pumpkin tissue: (a) non-embWhite, sectioned and stained in Azure II/Methylene blue. Horizontal bar is 0.

two horizontal lines of the photograph. The contouringof the cells allowed their isolation and later image analysisof the size and shape factors.

Image analysis of the isolated cell contour was per-formed with the image analysis software ‘‘Image Tool”(free software available from Health Science Centre, Uni-versity of Texas, San Antonio, Texas).

Several geometrical cellular parameters, previously con-sidered by other authors (Lewicki & Pawlak, 2003; Mayoret al., 2005; Reeve, 1953) were analyzed:

(1) Surface area.(2) Perimeter of the contour.(3) Length of the major axis: the length of the longest

line that can be drawn through the cell.(4) Length of the minor axis: the length of the longest

line that can be drawn through the cell perpendicularto the major axis.

(5) Roundness, defined as

Roundness ¼ 4pArea

perimeter2: ð5Þ

(6) Elongation: the ratio of the length of the major axisto the length of the minor axis.

(7) Compactness, defined as

Compactness ¼

ffiffiffiffiffiffiffiffi4 area

p

q

major axis length: ð6Þ

Statistical analysis of the results was performed with thesoftware Microsoft Excel 2003 (Microsoft Corporation).

3. Results and discussion

3.1. Fresh material

The microstructure of pumpkin parenchyma was ana-lyzed by means of the two different preparation techniques:non-embedded samples and embedded samples, as shownin Fig. 3a and b.

edded sample stained with methylene blue and (b) sample embedded in LR2 mm length.


As can be observed, the cells present different shapes:some of them are round, others are more elongated andothers have a polygonal shape. In Fig. 3b is more clearlyobserved the presence of intercellular spaces, which containthe air phase of the tissue and some of them may be filledwith an aqueous solution. In both images the cells show aturgid aspect.

Since the tissue presents a fibre orientation, first of allthe study of the size and shape parameters was done in

0.00 0.01 0.02 0.03 0.04 0.05 0.060

2

4

6

8

10

12

14

16

18

Freq

uenc

y (%

)Fr

eque

ncy

(%)

Freq

uenc

y (%

)

Area (mm2)

Non-embeddedEmbedded

0

2

4

6

8

10

12

14

16

18

Perimeter (mm)

0.00 0.05 0.10 0.15 0.20 0.250

2

4

6

8

10

12

14

16

18

Eq. diameter (mm)

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9



a

c

e

Fig. 4. Frequency histograms for size and shape parameters of the fresh pump(c) perimeter; (d) roundness; (e) equivalent diameter and (f) compactness.

two directions, in order to analyze if some cellular orienta-tion exists. For this purpose, cuts of the tissue parallel (lon-gitudinal fibre orientation) and perpendicular (radial fibreorientation) to the longitudinal axis of the fibres were stud-ied, with fresh non-embedded samples stained with methy-lene blue. Several samples (four for each orientation) of thesame pumpkin were analyzed, totalling 437 cells. No signif-icant differences (t test, p > 0.05 for all the parameters)were observed in the size and shape parameters between

Freq

uenc

y (%

)Fr

eque

ncy

(%)

Freq

uenc

y (%

)

0

2

4

6

8

10

12

14

16

18

Elongation

0

2

4

6

8

10

12

14

16

18

Roundness

0

2

4

6

8

10

12

14

16

18

Compactness

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

0.4 0.5 0.6 0.7 0.8 0.9 1.0

0.4 0.5 0.6 0.7 0.8 0.9 1.0




b

d

f

kin cells for the two sample preparation methods: (a) area; (b) elongation;


both orientations, and it can be concluded that the cells aredisposed in the tissue with no preferential orientation.

After this analysis of the cellular orientation, the effectof the sample preparation method on the size and shapeparameters was studied. For these purpose, several sampleswere obtained and analyzed from different parts of thesame pumpkin, totalling 437 cells for the non-embeddedsamples and 399 cells for the embedded samples. The histo-grams of the size and shape parameters for fresh pumpkinprepared by the two explained methods are shown inFig. 4. It is observed that the size and shape of the curvesare similar for all the parameters, but there is a shift in the

Table 1Average size and shape cellular parameters for fresh and osmotically dehydra

Parameter Fresh samples

Non-embedded Em

Average Interval Av

Area (mm2) 0.015 0.0016–0.047 0.0Perimeter (mm) 0.469 0.155–0.858 0.3Eq. diameter (mm) 0.136 0.046–0.245 0.1Elongation 1.299 1.013–2.104 1.3Roundness 0.831 0.649–0.935 0.8Compactness 0.871 0.659–0.979 0.8

Fig. 5. Cells of pumpkin tissue during osmotic dehydration: (a) fresh material;end of the process. D = Detachment of plasma membrane (plasmolysis). Hori

curves corresponding to the size parameters for embeddedsamples, decreasing their values. This shift can be attrib-uted to the ‘‘cutting effect” produced in the preparationof samples when the method of embedding in resin is used.In this method, when the block of resin is sectioned, a shiftto lower values in the average area obtained is expected asa consequence of the random intersections of the cutthrough the cell (Russ, 2004), assuming cells as spherical.When a fresh sample is cut with a razor blade, it wasobserved that instead of being cut the cells are separatedwhole and kept intact. For this reason, the size valuesobtained are higher than those observed with the samples

ted (OD) pumpkin samples

OD samples (9 h)

bedded Embedded

erage Interval Average Interval

10 0.001–0.037 0.006 0.0013–0.02581 0.064–0.738 0.377 0.087–0.66210 0.016–0.217 0.081 0.014–0.17758 1.024–3.835 1.893 1.171–4.30331 0.462–0.942 0.461 0.211–0.82560 0.510–0.981 0.678 0.423–0.946

(b) beginning of the dehydration; (c) dehydration and plasmolysis and (d)zontal line is 0.2 mm.


embedded in resin and sectioned. The theoretical shift insize values is 67% for cellular area and 78% for perimeterand equivalent diameter. Observing Table 1, the decreasein the average values of the size parameters (fresh samples)is in the order of this theoretical decrease. In this way, thevalues obtained during observation of not-embedded sam-ples can be considered closer to the real values than thoseobtained when sections of the embedded samples are used.

Table 1 shows the average and range values for the cel-lular shape and size parameters obtained from non-embed-ded and embedded fresh samples. For both types ofparameters, the results are similar to those obtained inother works for other vegetables. Area values(0.015 mm2) are in the range of the values found in otherwoks for apple (Lewicki & Pawlak, 2005; Mayor et al.,2005) and carrots (Lewicki & Drzewucka-Bujak, 1998),but higher than those obtained for potatoes (Zdunek &Umeda, 2005). Roundness values (0.83) are similar to those

Fig. 6. Microstructure changes during osmotic dehydration of pumpkin in a qand (e) 9 h. Horizontal bars are 2 mm.

obtained for apples (Bolin & Huxsoll, 1987; Lewicki &Pawlak, 2003; Mayor et al., 2005).

Concerning the shape of the distribution curves of Fig. 4area, elongation compactness and roundness gave an insuf-ficient fit to normal distributions (Kolmogorov–Smirnovtests of normality, p < 0.05), and they seem to follow alog–normal distribution. For equivalent diameter andperimeter the fit to a normal distribution was satisfactory(Kolmogorov–Smirnov test, p > 0.05). Log–normal distri-butions of the cellular area have been observed for apples(Lewicki & Pawlak, 2003; Mayor et al., 2005), and for car-rots (Lewicki & Drzewucka-Bujak, 1998); normal distribu-tions of cellular equivalent diameter have been observedfor apple (Lewicki & Pawlak, 2003). Although all theparameters for the fresh material cannot be considered nor-mally distributed, during the process these curves tended tonormal shapes, and some part of the data analysis was per-formed in terms of average values and standard deviations.

uarter of the slab; non-embedded samples: (a) 0.5 h; (b) 1 h; (c) 3 h; (d) 6 h


3.2. Dehydrated material

Different degrees of dehydration of parenchyma cells areshown in Fig. 5. Initially, in the fresh material, the cellshave its maximum size, and cell walls present a roundand turgid aspect (Fig. 5a). Then, the samples start todehydrate and the cell walls start to loose their initialaspect and become less round (Fig. 5b). In a moreadvanced degree of dehydration, the shrinkage of cells isconsiderable, roundness decreases and plasmolysis isclearly observed in cells (Fig. 5c). Very similar structuralchanges have been observed in light microscopy imagesof osmodehydrated apple in sucrose and glucose solutions(Nieto, Salvatori, Castro, & Alzamora, 2004; Quiles, Perez-Munuera, Hernando, & Lluch, 2003), showing the foldingof the cell wall, plasmolysis and cellular shrinkage. In theend of the process, cells show their maximum shrinkage,maximum decrease in roundness and are highly elongated(Fig. 5d). These different degrees of dehydration are a func-tion of process time and the situation of the cells in thepumpkin samples, as it will be discussed in this section.

It was only possible to observe the cellular changes inthe embedded samples, since in the non-embedded samplesthe cellular structure was not visible after osmotic dehydra-

Fig. 7. Microstructure changes during osmotic dehydration of pumpkin, embshows an amplified zone representative of each one of the six (or four) parts i

tion. According to this, the microphotographs of non-embedded samples were used to observe changes at ‘‘mes-ostructural” level, whereas the fixed-embedded sampleswere used to study the changes at cellular level.

The structural changes observed during osmotic dehy-dration of pumpkin cylinders with 45% sucrose solutionsat 25 �C are depicted in Figs. 6 and 7. Fig. 6 shows thestructural changes observed with the non-embedded sam-ples in the whole quarter of a slab. Fig. 7 shows the changesobserved at cellular level with the embedded samples; inthis case only an amplified zone, representative of each ofthe six (or four) parts in which the quarter was divided,is shown.

Concerning Fig. 6, it can be observed a dehydrationfront which enters in the material as the dehydration pro-cess progresses; at 0.5 h (Fig. 6a) the dehydration front islocated in the external zone of the sample, whereas at 9 hof process (Fig. 6e) practically the dehydration front isnot observed because the entire sample has been dehy-drated in all the zones of the material.

Observing Fig. 7, the changes in the microstructure canbe associated with the changes in the location of the dehy-dration front. The changes in the cellular structure initiallyappear in the external zone of the samples, and in the inner

edded samples: (a) 0.5 h; (b) 1 h; (c) 3 h; (d) 6 h and (e) 9 h. Each imagen which the quarter was divided. Horizontal bars are 0.2 mm.


core the cells present similar characteristics as thoseobserved in the fresh tissue. At 0.5 h and 1 h of process(Fig. 7a and b, respectively), some slight change in cellularshape (folding of the cell walls) is observed in the externalzones of the sample, whereas the inner zones do not showalterations. At 3 h of process (Fig. 7c), shrinkage of cells,plasmolysis and folding of the cell walls are observed inthe external zones of the samples, whereas the inner zonesstill remain unchanged, according to the existence of a non-dehydrated solid core. At 6 h (Fig. 7d) the cells of the exter-

0

2

4

6

8

10

12

14

16

18

Freq

uenc

y (%

)Fr

eque

ncy

(%)

Freq

uenc

y (%

)

Area (mm2)

NMC=1.000NMC=0.915NMC=0.882NMC=0.781NMC=0.719NMC=0.681

0

2

4

6

8

10

12

14

16

18

Perimeter (mm)


0.00 0.05 0.10 0.15 0.20 0.250

2

4

6

8

10

12

14

16

18

Eq. Diameter (mm)


0.000 0.005 0.0150.010 0.020 0.025 0.030 0.035 0.040

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

a b

c

e

Fig. 8. Frequency histograms and average values (normalized) of size paraaverages; (c) perimeter histograms; (d) perimeter averages; (e) equivalent diam

nal zones are more dehydrated and the inner core starts toshow some changes due to the dehydration of its cells. At9 h (Fig. 7e), the cells of the external zones are highly dehy-drated (high degree of shrinkage, cells are wrinkled andelongated), and the internal zone is also dehydrated butin a less accentuated way.

The existence of structural profiles accompanying thecompositional profiles during osmotic dehydration wasalso observed by Salvatori, Andres, Albors, Chiralt, andFito (1998) during osmotic dehydration of apple tissue with

1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.60

1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.60

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

A/A

o

NMC

P/P o

NMC

1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.60

Eq. D

iam

/Eq.

Dia

m. o

NMC

Experimental dataEq.(7)



d

f

meters vs. moisture content (normalized): (a) area histograms; (b) areaeter histograms and (f) equivalent diameter averages.


sucrose solutions. Similarly, during convective drying ofpotatoes, Wang and Brennan (1995), observed that in theinitial stages of the process the changes in cellular structure(shrinkage of cells and damage of the structure) were onlyat the external surface of the samples; then the changesentered in the inner of the tissue with the increase in dryingtime.

The analysis of shape and size cellular parameters wasperformed with the microimages of the embedded samples,

0

5

10

15

20

25

Freq

uenc

y (%

)Fr

eque

ncy

(%)

Freq

uenc

y (%

)

Elongation


0

5

10

15

20

25

Roundness


0

5

10

15

20

25

Compactness


0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5

0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.00.1 0.2

a b

c d

e f

Fig. 9. Frequency histograms and average values (normalized) of shape paraelongation averages, (c) roundness histograms; (d) roundness averages; (e) com

since the microimages of not-embedded samples did notallow a good observation at cellular level. Figs. 8 and 9show the histograms of frequencies and the averagereduced values obtained in the analysis of the size andshape cellular parameters during dehydration, respectively.Table 1 shows the average values of the cellular parametersafter 9 h of osmotic dehydration. These results wereobtained from the analysis of at least 400 cells at each pro-cess time. These cells come from the six zones (four zones

1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.600.4

0.6

0.8

1.0

1.2

1.4

1.6

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0.4

0.6

0.8

1.0

1.2

1.4

1.6

E/E

o

NMC


1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.60

R/R

o

NMC

1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.60

C/C

o

NMC



meters vs. moisture content (normalized): (a) elongation histograms; (b)pactness histograms and (f) compactness averages.

Table 2Fit results of experimental data on cellular size and shape parameters withEq. (7)

a b c R2 ARD (%)

Area

WL 1.00 �0.02 �1.03 0.98 2.0WR 1.00 �0.14 �1.46 0.99 1.6NMC �2.43 6.37 �2.95 0.98 2.6

Perimeter

WL 1.00 �0.09 0.10 0.47 0.7WR 1.00 �0.12 0.17 0.49 0.7NMC 1.35 �0.93 0.58 0.49 0.7

Eq. diameter

WL 1.00 0.03 �0.69 0.98 1.1WR 1.00 �0.03 �0.98 0.99 0.8NMC �1.27 4.36 �2.09 0.98 1.3

Elongation

WL 1.00 �0.59 1.76 0.93 2.8WR 1.00 �0.69 2.73 0.95 2.3NMC 7.07 �13.29 7.24 0.96 2.2

Roundness

WL 1.00 0.14 �1.23 0.98 2.1WR 1.00 0.07 �1.79 0.99 1.7NMC �3.15 8.24 �4.09 0.99 1.7

Compactness

WL 1.00 0.14 �0.70 0.97 1.2WR 1.00 0.14 �1.05 0.98 0.9NMC �1.38 4.92 �2.55 0.98 1.0


in the case of 6 and 9 h of treatment) analyzed in the sam-ple. The same number of cells of each zone was employedin order to obtain the best representative average value ofthe total sample. Reduced values were employed to reducethe variability among fresh samples as well as to minimizethe shift produced by the cutting effect.

Concerning the size parameters (Fig. 8), it can beobserved that the shape of the distribution curves doesnot change in the process. For the area and equivalentdiameter, the curves suffer a shift to lower values, leadingto a decrease in their average values along the process.At the end of the process average area decreased to 55%of the initial value, and equivalent diameter to 75%(Fig. 8b and f; absolute values for embedded samples inTable 1). This is in accordance with cellular shrinkageobserved in the microphotographs (Fig. 5). Lewicki andPawlak (2003) observed a decrease in cross sectional areaand diameter of apple cells during convective drying,whereas Lewicki and Porzecka-Pawlak (2005) observed adecrease in equivalent diameter of apple cells during osmo-tic dehydration. It is interesting to observe that the perim-eter practically does not show change during dehydration(Fig. 8c and d; Table 1). Several authors have observed thatthe plasmatic membrane (plasmalemma) shrinks elasticallyduring osmotic processes without folding (Segui, Fito,Albors, & Fito, 2006), and it is suggested that elasticshrinkage of the cell membrane may be accompanied byan endocytosis phenomenon (Oparka, Prior, & Harris,1990). Observing the results of this work, although the cellmembrane may shrink, the cell wall does not shrink duringthe process, and its response to the cellular shrinkage is thefolding of the cell wall, affecting the shape parameters asdiscussed below.

Concerning the shape parameters (Fig. 9), it can beobserved changes in the shape of the distribution curvesand in the average values during osmotic dehydration.The values of the shape parameters are more broadly dis-tributed when advancing in the dehydration process,changing the shape of the distribution curves to a more nor-mal shape. The average values does not suffer importantchanges at the beginning of the process (up to 3 h,NMC = 0.882); after that, elongation increases, and round-ness and compactness decrease up to the end of the process(Fig. 9b, d and f, absolute values for embedded samples inTable 1). Lewicki and Porzecka-Pawlak (2005) alsoobserved a decrease in the roundness values of cells duringosmotic dehydration of apples. Ramos et al. (2004) andMayor et al. (2005) pointed that cellular shape parameters(elongation, roundness and compactness) did not showchanges during the first stage of convective drying of grapesand apples, respectively; the last authors showed that thechanges of shape factors were observed at the end of thedrying process. Shrinkage of cells, folding of the cell wallsand the transition from a round to an elongated shape(Fig. 5) are the cause of these changes in the shape factors.

During osmotic dehydration of vegetables, it has beenobserved that the gain of solids is only significant in the

external zones of the material (Chenlo, Moreira, Fernan-dez-Herrero, & Vazquez, 2006; Mauro & Menegalli, 2003;Salvatori, Andres, Chiralt, & Fito, 1999), whereas water ini-tially is lost in these external zones but after that attains alsothe inner zones of the vegetable tissue. According to this, itis expected that the changes in size and shape cellularparameters are mainly due to the water loss in the material,leading to internal stresses in the tissue structure as well asto the decrease in the turgor pressure of cells. In other dehy-dration processes, such as convective drying, puff dryingand freeze drying, pressure and temperature gradients aswell as the high degree of dehydration attained lead to tissuedamage and cavities formation (Lewicki & Pawlak, 2003;Lewicki & Pawlak, 2005). During osmotic dehydrationthese gradients are minimized due to the moderate processconditions used and these phenomena are not observed.

Polynomial models were employed to relate some kineticparameters (WL, WR and NMC) with the changes in theaverage cellular size and shape parameters. Since it isexpected that the cutting effect observed in fresh samplesfor embedded samples also occurs in dehydrated samples,the reduced values of size and shape parameters wereemployed in these equations in order to minimize thiseffect. Linear models gave a poor fit, and cubic modelsdid not show significant improvement in the fit comparedwith quadratic models, so these last ones were used for cor-relation purposes. The equations used were of the type

Y ¼ aþ bX þ cX 2; ð7Þ


where Y is the normalized cellular factor (ratio value atdehydration time t to initial value) and X is the kineticparameter. The fit results are shown in Table 2. By usingthe equation parameters shown in Table 2 it is possibleto calculate the average value of any of the studied sizeor shape parameters at any WL, WR or NMC.

In general, Eq. (7) shows good fits of the cellular param-eters with the kinetic parameters WL, WR and NMC. Theexception is the perimeter, but this is very reasonable since

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

A/A

o

t (h)

External zoneInternal zone

Eq. D

iam

./Eq.

Dia

m. o

P/P o

0 2 4 6 8 10

t (h)0 2 4 6 8 10

t (h)0 2 4 6 8 10



a

c

e

Fig. 10. Changes in the cellular size and shape parameters in the external anelongation; (c) perimeter; (d) roundness; (e) equivalent diameter and (f) comp

perimeter remains constant during the process. The qualityof the fits is similar. Correlation coefficients vary from 0.93to 0.99, and average relative deviations from 0.7% to 2.8%.Figs. 8 and 9 show the predicted values of the model, whenthe correlations are done with normalized moisturecontent.

When the analysis of the cellular parameters was done indifferent zones the samples, some interesting results wereobtained. Fig. 10 shows the decrease in the studied param-

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0.4

0.6

0.8

1.0

1.2

1.4

1.6

E/E o

C/C

oR

/Ro

t (h)




0 2 4 6 8 10

t (h)0 2 4 6 8 10

t (h)0 2 4 6 8 10

b

d

f

d internal zones of the samples during osmotic dehydration: (a) area; (b)actness.

A/Ao=1-0.85

A/Ao=0.85-0.69

A/Ao=0.69-0.54

A/Ao=0.54-0.38

a b c

d e f

Fig. 11. Structural profiles in a quarter of the slab for the normalizedcellular area at different process times: (a) fresh material; (b) 0.5 h; (c) 1 h;(d) 3 h; (e) 6 h; (f) 9 h.


eters in the internal and the external (in contact with theosmotic solution) zones. It is observed that in the externalzones (1, 2, 3, and 5 sections of Fig. 2) the changes in all theparameters start at the beginning of the process, whereas inthe internal zones the changes start only after 2 h of pro-cessing, then in both zones the changes are similar up tothe end of the process.

The normalized cellular area, in a quarter of the cylin-drical sample and at different process times is shown inFig. 11. As observed before in Fig. 10a, the change in cel-lular area is initially localized in the external zones of thematerial, and the inner solid core only suffers structuralalterations after 3 h of process. At 9 h, the samples havesuffered changes in the entire sample, although the changesin the external zones are slightly more accentuated. Similarprofiles were observed for all the size and shape parame-ters, with the exception of the cellular perimeter whichremained constant during the process. As commentedbefore, it is believed that the structural changes occursimultaneously with moisture removal. The presence of adehydration front which enters in the material during dehy-dration and a solid core with the initial characteristics ofthe fresh material which decreases with the moistureremoval is in accordance with the structural profilesobserved along the dehydration process.

4. Conclusions

The study of the fresh parenchymatic pumpkin tissueshowed that the cellular size and shape parameters are sim-

ilar to those found in other fruits and vegetables, such asapples, carrots and potatoes. Fresh pumpkin cells showaverage values of 0.015 mm2, 0.469 mm and 0.136 mm forcell area, cell perimeter and cell equivalent diameter,respectively; and 1.288, 0.831 and 0.871 for cellular elonga-tion, roundness and compactness, respectively. The distri-bution curves for area, elongation roundness andcompactness follow a log–normal distribution and thosefor perimeter and equivalent diameter follow a normal dis-tribution; these curves tend to normal shapes during thedehydration process.

Osmotic dehydration causes changes in the cellular sizeand shape parameters of the vegetable tissue. The mainphenomena observed during osmotic dehydration wereshrinkage of cells, plasmolysis and folding of the cell walls.These changes lead to the decrease in cellular area, equiva-lent diameter, roundness and compactness; elongation ofcells increased whereas the perimeter was maintainedconstant along the process. The observed changes are nothomogeneously distributed in the material, and aredependent on the localization of the cells in the samplesand on the process time. Although water loss seems to bethe main cause of the cellular changes during the process,causing microstructural stresses in the vegetable tissue,more work is needed to asses the effect of osmotic solutesintake and process conditions (pressure, temperature) onmicrostructure.

Empirical quadratic functions were used to relate theaverage shape and size parameters with the dehydrationparameters WL, WR and NMC. The equations showed agood fit of the experimental data, leading to correlationcoefficients ranging 0.93–0.99 and average relative devia-tions ranging 0.7–2.8%.

Future work has to be done to create structural modelsaccounting for the microstructural changes during dehy-dration. Finite element approximations (Martins, 2006)or the use of Voronoi Tessellations (Mattea, Urbicain, &Rotstein, 1989; Mebatsion et al., 2006; Wenian, Duprat,& Roudot, 1991) may be some interesting approaches tothis microstructural modelling. The simple empirical mod-els obtained in this work may also be used along with thecommented more complex approximations.

Acknowledgements

The author Luis Mayor wishes to acknowledge SFRH/BD/3414/2000 PhD Grant to Fundac�ao para a Ciencia e aTecnologia, Portugal, as well as the facilities provided bythe Institute for Molecular and Cell Biology at the Univer-sity of Porto, Portugal.

References

Aguilera, J. M., Cuadros, T. R., & del Valle, J. M. (1998). Differentialscanning calorimetry of low-moisture apple products. Carbohydrate

Polymers, 37, 79–86.AOAC (1984). Official methods of analysis (14th ed.). Virginia, USA:

Association of Official Analytical Chemists.


Bolin, H. R., & Huxsoll, C. C. (1987). Scanning electron microscope/image analyzer determination of dimensional postharvest changes infruit cells. Journal of Food Science, 52(6), 1649–1650, 1698.

Chenlo, F., Moreira, R., Fernandez-Herrero, C., & Vazquez, G. (2006).Experimental results and modeling of the osmotic dehydration kineticsof chesnut with glucose solutions. Journal of Food Engineering, 74,324–334.

Ferrando, M., & Spiess, W. E. L. (2001). Cellular response of plant tissueduring the osmotic treatment with sucrose, maltose, and trehalosesolutions. Journal of Food Engineering, 49, 115–127.

Giangiacomo, R., Torreggiani, D., & Abbo, E. (1987). Osmotic dehydra-tion of fruit. Part I: sugar exchange between fruit and extracting syrup.Journal of Food Processing and Preservation, 11, 183–195.

Hawkes, J., & Flink, J. M. (1978). Osmotic dehydration of fruit slices priorto freeze dehydration. Journal of Food Processing and Preservation, 2,265–284.

Kim, M. H., & Toledo, R. T. (1987). Effect of osmotic dehydration andhigh temperature fluidized bed drying on properties of dehydratedrabbiteye blueberries. Journal of Food Science, 52(4), 980–984, 989.

Krokida, M. K., Karathanos, V. T., & Maroulis, Z. B. (2000). Effect ofosmotic dehydration on color and sorption characteristics of apple andbanana. Drying Technology, 18(4–5), 937–950.

Lewicki, P. P., & Drzewucka-Bujak, J. (1998). Effect of drying on tissuestructure of selected fruits and vegetables. In: Proceedings of the 11th

International Drying Symposium (IDS’ 98), Halkidiki, Greece.Lewicki, P. P., & Pawlak, G. (2003). Effect of drying on microstructure of

plant tissue. Drying Technology, 21(4), 657–683.Lewicki, P. P., & Pawlak, G. (2005). Effect of mode of drying on

microstructure of potato. Drying Technology, 23(4), 847–869.Lewicki, P. P., & Porzecka-Pawlak, R. (2005). Effect of osmotic

dewatering on apple tissue structure. Journal of Food Engineering,

66, 43–50.Lozano, J. E., Rotstein, E., & Urbicain, M. J. (1983). Shrinkage, porosity

and bulk density of foodstuffs at changing moisture contents. Journal

of Food Science, 48, 1497–1502, 1553.Martins, R. C. (2006). Simple finite volumes and finite elements

procedures for food quality and safety simulations. Journal of Food

Engineering, 73(4), 327–338.Mattea, M., Urbicain, M. J., & Rotstein, E. (1989). Computer model of

shrinkage and deformation of cellular tissue during dehydration.Chemical Engineering Science, 44(12), 2853–2859.

Mauro, M. A., & Menegalli, F. C. (2003). Evaluation of water and sucrosediffusion coefficients in potato tissue during osmotic concentration.Journal of Food Engineering, 57, 367–374.

Mauro, M. A., Tavares, D. D., & Menegalli, F. C. (2002). Behavior ofplant tissue in osmotic solutions. Journal of Food Engineering, 56,1–15.

Mayor, L., & Sereno, A. M. (2004). Modelling shrinkage duringconvective drying of food materials: a review. Journal of Food

Engineering, 61(3), 373–386.Mayor, L., Silva, M. A., & Sereno, A. M. (2005). Microstructural changes

during drying of apple slices. Drying Technology, 23(9–11), 2261–2276.Mebatsion, H. K., Verboyen, P., Verlinden, B. E., Ho, Q. T., Nguyen, T.

A., & Nicolai, B. M. (2006). Microscale modeling of fruit tissue usingVoronoi tessellations. Computers and Electronics in Agriculture, 52,36–48.

Monsalve-Gonzalez, A., Barbosa-Canovas, G., & Cavalieri, R. P. (1993).Mass transfer and textural changes during processing of apples bycombined methods. Journal of Food Science, 58(5), 1118–1124.

Nesvabda, P. (2005). Thermal properties of unfrozen foods. In M. A. Rao,S. S. H. Rizvi, & A. K. Datta (Eds.), Engineering Properties of Foods

(3rd ed.). Boca Raton, US: Taylor and Francis Group.

Nieto, A. B., Salvatori, D. M., Castro, M. A., & Alzamora, S. M. (2004).Structural changes in apple tissue during glucose and sucrose osmoticdehydration: shrinkage, porosity density and microscopic features.Journal of Food Engineering, 61, 269–278.

Oparka, K. J., Prior, D. A. M., & Harris, N. (1990). Osmotic induction offluid-phase endocytosis in onion epidermal cells. Planta, 180(4),555–561.

Perera, C. O. (2005). Selected quality attributes of dried foods. Drying

Technology, 23, 717–730.Quiles, A., Perez-Munuera, I., Hernando, I., & Lluch, M. A. (2003).

Impact of mass transport on microstructure of Granny Smith appleparenchyma during osmotic dehydration. Journal of the Science of

Food and Agriculture, 83, 425–429.Rahman, M. S. (2005). Mass–volume–area related properties of foods. In

M. A. Rao, S. S. H. Rizvi, & A. K. Datta (Eds.), Engineering

Properties of Foods (3rd ed.). Boca Raton, US: Taylor and FrancisGroup.

Ramalho-Santos, M., Pissarra, J., Verissimo, P., Pereira, S., Salema, R.,Pires, E., et al. (1997). Cardosin A, an abundant aspartic proteinase,accumulates in protein storage vacuoles in the stigmatic papillae ofCynara cardunculus L. Planta, 203, 204–212.

Ramos, I. N., Silva, C. L. M., Sereno, A. M., & Aguilera, J. M. (2004).Quantification of microstructural changes during first stage air dryingof grape tissue. Journal of Food Engineering, 62, 159–164.

Rao, V. N. M., & Quintero, X. (2005). Rheological properties of solidfoods. In M. A. Rao, S. S. H. Rizvi, & A. K. Datta (Eds.), Engineering

Properties of Foods (3rd ed.). Boca raton, US: Taylor and FrancisGroup.

Raven, P. H., Evert, R. F., & Eichhorn, S. E. (1999). Biology of Plants (6thed.). New York, US: W.H. Freeman and Company.

Reeve, R. M. (1953). Histological investigations of texture in apples. IIStructure and intercellular spaces. Food Research, 18(6), 604–617.

Russ, J. C. (2004). Image analysis of food microstructure. Boca Raton, US:CRC Press LLC.

Salvatori, D., Andres, A., Albors, A., Chiralt, A., & Fito, P. (1998).Structural and compositional profiles in osmotically dehydrated apple.Journal of Food Science, 63(4), 606–610.

Salvatori, D., Andres, A., Chiralt, A., & Fito, P. (1999). Osmoticdehydration progression in apple tissue I: spatial distribution of solutesand moisture content. Journal of Food Engineering, 42, 125–132.

Segui, L., Fito, P. J., Albors, A., & Fito, P. (2006). Mass transferphenomena during osmotic dehydration of apple isolated protoplasts(Malus domestica var. Fuji). Journal of Food Engineering, 77(1),179–187.

Telis, V. R. N., Telis-Romero, J., & Gabas, A. L. (2005). Solids rheologyfor dehydrated food and biological materials. Drying Technology, 23,759–780.

Teotia, M. S., (1992). Advances in chemistry and technology of pumpkins.Indian Food Packer, January–February, pp. 9–31.

Tregunno, N. B., & Goff, H. D. (1996). Osmodehydrofreezing of apples:structural and textural effects. Food Research International, 29(5–6),471–479.

Wang, N., & Brennan, J. G. (1995). Changes in structure, density andporosity of potato during dehydration. Journal of Food Engineering,

24(1), 61–76.Wenian, C., Duprat, F., & Roudot, A. C. (1991). Evaluation of the

importance of the cellular tissue geometry on the strains observed onapples after a compression or an impact. Sciences des Aliments, 11,99–110.

Zdunek, A., & Umeda, M. (2005). Influence of cell size and cell wallvolume fraction on failure properties of potato and carrot tissue.Journal of Texture Studies, 36, 25–43.

microstructural changes during osmotic dehydration of

Documents