Journal of Food Engineering 128 (2014) 79–87

Contents lists available at ScienceDirect

Journal of Food Engineering

journal homepage: www.elsevier .com/ locate / j foodeng

Peeling mechanism of tomato under infrared heating:Peel loosening and cracking

0260-8774/$ - see front matter Published by Elsevier Ltd.http://dx.doi.org/10.1016/j.jfoodeng.2013.12.020

⇑ Corresponding author. Address: Processed Foods Research Unit, USDA-ARS-WRRC, Albany, California 94710, USA. Tel.: +1 510 559 5861; fax: +1 510 559 5851.

E-mail addresses: zhongli.pan@ars.usda.gov, zlpan@ucdavis.edu (Z. Pan).1 Dr. Atungulu’s present affiliation is the Department of Food Science & Division of

Agriculture, University of Arkansas, Fayetteville, AR 72704, USA.

Xuan Li a, Zhongli Pan a,b,⇑, Griffiths G. Atungulu a,1, Delilah Wood c, Tara McHugh b

a Department of Biological and Agricultural Engineering, University of California, Davis, Davis, CA 95616, USAb Processed Food Research Unit, Western Regional Research Center, USDA Agricultural Research Service, Albany, CA 94710, USAc Bioproduct Chemistry and Engineering Research Unit, Western Regional Research Center, USDA Agricultural Research Service, Albany, CA 94710, USA

a r t i c l e i n f o

Article history:Received 18 July 2013Received in revised form 15 December 2013Accepted 22 December 2013Available online 28 December 2013

Keywords:Infrared radiationTomato peelingPuncture testScanning electron microscopyStress analysisModeling

a b s t r a c t

Critical behaviors of peeling tomatoes using infrared radiation heating are thermally induced peel loos-ening and subsequent cracking. Fundamental understanding of the two critical behaviors, peel looseningand cracking, remains unclear. This study aimed at investigating the mechanisms of peel separation fortomatoes subjected to a newly developed infrared dry-peeling process. Microstructural changes intomato epidermal tissues under infrared heating were compared with those of fresh, hot lye and steamtreated samples. Theoretical stress analyses coupled with the experimentally measured failure stress oftomato skin were combined to interpret the occurrence of peel cracking within a framework of elasticthin shell theory. With the use of light microscopy and scanning electron microscopy, it was observedthat peel loosening due to infrared heating appeared to result from reorganization of extracellular cuti-cles, thermal expansion of cell walls, and collapse of several cellular layers, differing from samples heatedby hot lye and steam. Crack behaviors of tomato skin were attributed to the rapid rate of infrared surfaceheating which caused the pressure build-up under the skin and strength decrease of the skin. In order toachieve a sufficient skin separation for effective peeling using infrared, promoting rapid and uniformheating on the tomato surface is essential. The findings gained from this study provide new insightsfor developing the sustainable infrared dry-peeling technology.

Published by Elsevier Ltd.

1. Introduction

Peeling is a particularly important unit operation in the produc-tion of canned fruits and vegetables. The process can affect the pal-atability and nutritive values of final canned products (Li, 2012).From a processing standpoint, the currently used lye and steampeeling methods are water and energy intensive, and pose serioussalinity issues and wastewater disposal problems (Barringer, 2003;Masanet et al., 2007; Pan et al., 2009; Rock et al., 2011; Li et al.,2013). To address these challenges, a sustainable alternative ofpeeling tomatoes using infrared radiation heat without relyingon water, steam, and chemicals has been developed. This peelingmethod is named as infrared dry-peeling (Pan et al., 2009). Theinfrared dry-peeling technology has been successfully tested bothat the bench scale and pilot scale using tomatoes from multipleharvesting seasons. Currently, onsite demonstrations to comparethe performance of the new method with conventional lye and

steam peeling methods are being conducted at various tomato pro-cessing plants in California. To further develop the technology andmake it commercially applicable, clear elucidation of the mecha-nism underlying infrared dry-peeling of tomatoes is crucial.Although several experimental and modeling aspects have beenaddressed in our previous investigations (Pan et al., 2009; Liet al., 2011; Li, 2012; Wang et al., 2013), the thermally inducedphysical and biochemical changes of tomato peel, in particularthe peel loosening and subsequent cracking phenomena, appeardifferent from traditional wet-peeling methods and have not befully understood. Study of the behavior of peel loosening andcracking should provide insight into the mechanism of dry-peelingof tomatoes using infrared.

Limited studies have been conducted to determine the peelingmechanisms (Floros and Chinnan, 1988). Most previous researchconcentrated on prediction of peeling performance or optimizationof various peeling processes mainly for the widely used lye andsteam peeling (Barreiro et al., 1995; Das and Barringer, 2005;Milczarek and McCarthy, 2011; Garcia and Barrett, 2006a,b;Matthews and Bryan, 1969; Schlimme et al., 1984; Toker andBayndrl, 2003; Wongsa-Ngasri, 2004). Possible mechanisms ofsteam and lye peeling of pimento pepper and tomato wereproposed based on examination of the skin microstructural

80 X. Li et al. / Journal of Food Engineering 128 (2014) 79–87

changes under different peeling conditions (Floros and Chinnan,1988, 1990; Floros et al., 1987). In the steam peeling process, themain cause of skin separation is a combination of biochemicaland physical changes due to the effects of high temperature steam.In lye peeling, chemical diffusion of hot lye solution into the tissuewith subsequent dissolving of the cell wall materials is the primarycause of skin release. Light Microscopy (LM) and Scanning ElectronMicroscopy (SEM) have proved to be useful tools for observing themicrostructural changes in skin morphology and anatomy occur-ring during lye and steam peeling of several vegetables, includingtomatoes (Floros and Chinnan, 1990; Mohr, 1990). These micro-scopic techniques can be used to determine whether the loosenedmicrostructure of infrared heated tomatoes is different from thatresulting from lye and steam treatments. Both of the above men-tioned peeling mechanisms may not directly apply to infrareddry-peeling because neither steam, water, nor chemicals are used.Instead, radiative thermal effects resulting in substantial changesin strength and biomechanical properties of tomato skin are pre-sumably the main cause for infrared induced peel loosening andcracking.

Several techniques have been attempted to experimentallydetermine skin strength and membrane biomechanical failure,including tensile, puncture, and bursting diaphragm methods (Cal-vin and Oyen, 2007; Haman and Burgess, 1986; Miles et al., 1969).The puncture-based method is a widely accepted approach forobtaining skin strength and failure stress. In this test, a force is ap-plied uniformly on the skin membrane by using a smooth round-ended probe or a uniform pressure loading so that the skin deformsin response to membrane biaxial tension. This technique enablesdetection of the increase in pressure on the skin membranes sur-rounding the fruit (Haman and Burgess, 1986; Henry and Allen,1974). In light of former mechanical studies, the present study esti-mated the rupture stress of tomato skin during infrared peeling bydetermining the force–displacement relationship of the skin mem-brane. Because the tomato skin is much thinner than the overallfruit diameter, tomato skin is considered as a thin-walled shell(Considine and Brown, 1981; Henry and Allen, 1974). The stresson a thin spherical shell by an internal pressure loading underconstant temperature can be estimated by using the membranetheory for spherical shells (Timoshenko et al., 1959; Upadhyayaet al., 1986, 1985). In this study, shell mechanics were appliedfor the analysis of the transient stress changes within the skin.The results were further analyzed to quantitatively evaluate therelationship between skin mechanical behavior and peel crackingsusceptibility.

The specific objectives of this study were to (1) compare themorphologies of epidermal cells of tomatoes subjected to infrared,lye, and steam treatments and fresh tomatoes; (2) use puncturetest to determine tomato skin rupture stress after infrared heating;and (3) investigate the correlations between transient skin stressand increasing temperature during infrared heating by using anintegrated approach of experimental measurements and theoreti-cal analysis.

2. Materials and methods

2.1. Experimental setup and sample preparation

Tomatoes of cultivars CXD179 and AB2 with uniform ripenessand size were subjected to infrared heating from two sides for60 s. Tomatoes were collected at red-matured stage according tothe USDA standard (i.e., USDA tomato classification 6) (Li et al.,2013). Only defect-free tomatoes at a size level ranging from42 mm to 54 mm were used for peeling and subsequent measure-ments. During infrared heating, a tomato was rotated continuously

at a speed of 1 rpm by means of a motor driven turntable to receiveuniform heating. A custom-designed metal holder was used toplace the tomato between the vertically aligned emitters (Liet al., 2013). The specific infrared heating setup and procedureare described in our previous publications (Pan et al., 2009; Li,2012). During infrared heating, initial peel cracking was visuallynoted and the time was recorded by a stopwatch. Peel crackingwas normally accompanied by a sudden sound due to skin raptureduring infrared heating. After infrared heating, each treated tomatowas sealed in a plastic bag to prevent further moisture loss, and itwas allowed to cool to ambient temperature in the laboratory forabout half an hour. Peels from these tomatoes were then usedfor microstructural studies and puncture tests that are describedlater.

Light microscopy (LM) imaging was used to observe the layerseparation in pericarp tissue of tomato treated by 60 s infraredheating. Pericarp cubes (approximately 1 cm3) of tomato with theskin attached were cut from the equatorial region of the tomatoand prepared by fixation and critical point drying methodspreviously described (Li, 2012). Specimens were then viewed andphotographed by using a Leica MZ16F stereoscope (LeicaMicrosystems, Wetzlar, Germany). Digital images were obtainedwith a QImaging Retiga 2000R FAST color camera (QImaging,Surrey, B.C., Canada). Representative images were presented.

2.2. Low temperature high resolution scanning electron microscopy

Tomato pericarp tissue (approximately 3 � 3 � 4 mm) was ob-tained from the middle region of the tomato immediately afterinfrared heating. Each tissue specimen was trimmed to a wedgeshape and mounted onto a copper sample holder with Tissue-Tek adhesive (Sakura Finetek USA Inc., Torrance, Cal., USA) andthen prepared for low temperature SEM that was conducted withan Alto 2500 cryo system (Gatan Inc., Pleasanton, Cal., USA). Thesample holder was attached to the rod of a vacuum transfer de-vice and plunged into super-chilled liquid nitrogen. The specimenwas evacuated, pulled up into the vacuum transfer device, andtransferred to a cryo preparation chamber. The specimen wasfractured in the cryo preparation chamber at approximately�180 �C, warmed to �85 �C and held at that temperature for15 min to remove excess surface water. The specimen was thencooled by shutting off the heater of the cryo system to less than�135 �C, sputter coated with gold palladium, and transferred tothe cryo stage specimen chamber in the SEM. All samples wereobserved and photographed below �135 �C at 2.0 kV by meansof a Hitachi S-4700 field emission SEM (Hitachi High-Technolo-gies Corp., Tokyo, Japan). Digital images were collected at2180 � 960 pixels and were viewed under the scanning electronmicroscope at 400�magnification for the outer surface of tomatoskin and at 200� and 400� magnification for the cross-sectionalimages of tomato dermal system. Fresh tomatoes prepared by thesame method were used as a control.

To better understand infrared thermal effects on changing thetomato microstructure, a set of experiments was conducted tocompare the differences of tomato samples treated with infrared,lye and steam heating for the same time period (60 s). Lye treatedsamples were prepared by using sodium hydroxide as described inPan et al. (2009). Steam treated tomatoes were obtained by usingsaturated steam from boiling water under atmospheric pressureas described in Li (2012). A minimum of three replicates wereobtained for each treatment method.

2.3. Measurement of skin rupture

A small segment of skin membrane was carefully dissectedfrom the peeled skin by using a cork borer with a diameter of

Displacement (mm)0 1 2 3 4 5

Forc

e (N

)

0

1

2

3

4

5

6

Peak force

Rupture distance

AB

Fig. 1. Measurement of tomato skin rupture: (A) experimental setup for the puncture test; (B) a representative force–displacement diagram.

X. Li et al. / Journal of Food Engineering 128 (2014) 79–87 81

22 mm. As shown in Fig. 1A, the edges of the skin membrane werethen clamped between two steel plates that had a central circularopening with a diameter of 12.8 mm. The sample was placed skinsurface side down on this platform and centered over the opening.To provide adequate support to the sample, the diameter ratio ofhole to probe was in the range of 1.5:1–3:1, as recommended byBourne (2002). A force was applied onto the inner surface of theskin membrane using a Magness–Taylor type probe with a convexindenter. The probe measured 6.9 mm in diameter and had a smallcurvature so that the natural curvature of the tomato skin closelyfollowed the curvature of the probe. Measurement of skin ruptureresistance was carried out on a fruit texture analyzer TA-XT2i(Texture Technologies Corp., New York, N.Y., USA) equipped witha load cell of 19.6 N. The force–displacement relationship wasrecorded for each sample at a loading rate of 0.10 mm/s. To mini-mize variability, all samples were taken from approximately thesame latitude around the periphery of the tomatoes. Each testwas completed when the rupture of skin membrane occurred.The puncture test produces a biaxial state of stress on the skinmembrane, which closely represents infrared heating conditionswhere internal pressure acts on the tomato inner skin.

As shown in Fig. 1B, the rupture force is defined as the peakforce, and the corresponding displacement was recorded as therupture distance. After the test, skin thickness was measured atthree locations on each dissected skin membrane by using a dialcaliper with 0.001 mm accuracy. Average values of rupture force,rupture distance, and skin thickness were reported from 10 repli-cates and were used later for calculating the skin rupture stress.All puncture measurements were completed within approximately2 h after infrared heating.

3. Theoretical analysis

3.1. Skin rupture stress

In the puncture test, the skin membrane can be considered as athin membrane under pressure. The stress within each skin speci-men increased as the probe was pressed harder onto the skinmembrane. Frictionless slip was assumed between the skinmembrane and the round end probe. At the moment skin ruptureoccurred, the rupture force was equal to the peak puncture force(Fp). The equilibrium of forces in the vertical direction yieldsFp =

PFy (Fig. 2A), and this relationship then leads to

Fp ¼ 2pRptrr sinðhÞ ð1Þ

where, Fp is the total downward puncture force, N; Rp is the radiusof puncture probe, mm; t is the thickness of skin membrane, mm; rr

is the stress in skin membrane, MPa; h is the contact angle, degree.

Thus, the normal stress in the skin membrane can beobtained as

rr ¼Fp

2pRpt sinðhÞ ð2Þ

Based on Eq. (2), skin rupture stress is expressed in terms ofseveral measurable parameters (i.e., Fp, Rp, t, and h) and thus canbe experimentally determined through a puncture test, whichprovides a measure of the skin strength changes before and afterinfrared heating.

From the right angle ABC in Fig. 2B, the trigonometric identitycan be used to determine the contact angle

h ¼ arctanab

� �ð3Þ

where, ‘‘a’’ is the vertical distance of segment AB, mm; and ‘‘b’’ is thehorizontal distance of segment AC, mm.

Based on the geometry of a spherical cross section, parameter‘‘a’’ can be expressed in terms of other geometric parameters D,d, and R, to give

a ¼ D� d ¼ D�ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiR2 �

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiR2 � R2

s

qrð4Þ

where, D is the rupture distance, mm; d is the height of the convextip end of the Magness–Taylor probe, mm; R represents the radiusof the convex tip of the Magness–Taylor probe, mm; and Rs repre-sents the radius of the hole of steel plate, mm.

According to Fig. 2B, parameter ‘‘b’’ is the difference betweenthe plate hole diameter and the probe diameter

b ¼ Rs � Rp ð5Þ

3.2. Estimation of stress in tomato skin membrane

During infrared heating, the stress in the skin increases as accu-mulation of internal pressure with the rising temperature. In orderto estimate how much stress may arise within the loosened skin,tomato skin can be modeled as a thin membrane that is subjectedto biaxial tensile stresses in two perpendicular directions, longitu-dinal, and circumferential, respectively (Timoshenko et al., 1959).When a tomato is considered as a spheroid, the biaxial membranestresses in the skin are equal in the two perpendicular directions,according to the theory of thin-wall shell (Considine and Brown,1981). Because the thickness of skin membrane is much less thanthe remaining radius of tomato fruit, the tomato can be consideredas a thin-walled pressure vessel. Accordingly, the intensity of stressin a thin spherical membrane can be expressed as (Timoshenkoet al., 1959)

Fig. 2. Schematic diagram of the skin rupture test: (A) free body diagram when rupture occurs; (B) cross sectional view of the skin membrane being loaded by the Magness–Taylor probe.

82 X. Li et al. / Journal of Food Engineering 128 (2014) 79–87

r ¼ pr2d

ð6Þ

where, r is the membrane stress in tomato skin, MPa; p is the inten-sity of uniform internal pressure, MPa; r is the radius of skin mem-brane, mm; and d is the thickness of skin membrane, mm.

In Eq. (6), three parameters (p, r, and d) can affect the membranestress. The radius and thickness of the skin are descriptive of eachindividual tomato (Considine and Brown, 1981). For simplicity, thisstudy explored an idealized case of great interest. It was assumedthat the tomatoes were spherical and had uniform shape and size;the skin membrane had uniform thickness and physical properties;and the curvature of the tomato skin did not change abruptly. Theequivalent radius of skin membrane (r) was justified to be that of aspheroid with the same surface area as a medium size tomato asdefined in a previous publication (Li et al., 2011). The averagethickness of skin membrane (d) can be obtained from our experi-mental measurements, which was 0.59 ± 0.038 mm in average fora 60 s infrared heating (Li, 2012).

The overall moisture loss of tomatoes was found to be less than2–3% during a 60 s infrared heating, and small bubbles were occa-sionally observed at the stem scar. It was concluded that limitedmoisture could pass through the tomato hydrophobic waxy surfacedue to the absence of stomata and non-liquefaction of the cuticularmembrane covering the outside of the tomato. The model was sim-plified by assuming the skin membrane was impervious to mois-ture and was thus subjected to water vapor pressure underinfrared heating. The total internal pressure on the skin membranewas almost entirely from the buildup of vapor pressure. The inter-cellular atmosphere of most horticultural commodities has beencommonly assumed to be saturated (Ferrua and Singh, 2009a,b).The high moisture content of tomatoes (�95% water) allows thereasonable assumption that the vapor pressure below the skin sur-face can be approximated by the saturated pressure of pure waterusing Antoine equation (Poling et al., 2001).

Dp � pv ¼ exp 16:3872� 3885:70230:170þ Tðd; tÞ

� �� 10�3 ð7Þ

where Dp represents the pressure difference between skin mem-brane, MPa; pv is the vapor pressure, MPa, T is the temperatureunder a skin membrane with a thickness of d, �C, and t stands forthe heating time, s.

The estimated vapor pressure beneath the skin is modeled as afunction of temperature. The range of temperature increase duringheating depended on heating time and specific region, but it was

typically 23 �C to 100 �C in our test conditions. Normally, peelcracking occurred at a location that was directly exposed to theinfrared emitter, where the strongest irradiation occurred. Basedon the results from the heat transfer model we developed (Li andPan, 2013a,b), this location was identified as in the region nearestthe emitter. The skin membrane thickness (d) can be experimen-tally determined from the peeled skin thickness. For simplicity,the d value was assumed to be uniform, and the average value ofpeeled skin thickness measured from more than ten peeled toma-toes was used to represent d. Once the location of cracking and skinmembrane thickness were identified, the predicted temperaturehistory at that location beneath the skin was obtained from theconstructed heat transfer model and used for calculating theinternal pressure. Relationship between internal pressure andmembrane stress given in Eq. (6) can be used to determinetransient membranes stress within the tomato skin during infraredheating. Note that because the skin thickness values are very small(less than 1 mm), the temperature gradient within the skinmembrane was assumed negligible in the study. This assumptionsimplified the theoretical calculation. As a case study, Eqs. (6)and (7) were evaluated for a typical peeling condition: 60 s infraredheating of a medium size tomato.

4. Results

4.1. Microstructure changes of tomato skin

Fig. 3 shows the microstructure changes on tomato outermostsurface after various treatments. The outermost tomato skin fea-tures a very thin hydrophobic waxy cuticular membrane (Domín-guez et al., 2011). On the fresh tomato surface (Fig. 3A), clearlydefined contours of cell wall structures can be observed. Afterinfrared heating, the contour and overall shape of epidermal cellsbecome difficult to discern and a knoblike protuberance arisesfrom each cell surface (Fig. 3B). For the lye treated samples(Fig. 3C), although the knoblike protuberances are formed, con-tours of epidermal cells are more readily visible than those forinfrared treated samples (Fig. 3A). This observation confirms previ-ous reports that the action of lye can increase cell visibility due tothe dissolution of waxes (Floros et al., 1987). It is suspected thatthe cell protuberance is caused by phase transition of the waxycuticular membrane, and melting and re-distribution of cuticularwax results in less visible contours of cell walls due to distortionby cuticular wax. It is interesting that tomato skin treated by steam

100 m 100 m

100 m 100 m

A B

C D

clearly defined contours knoblike protuberance

concave surfaces

increased visibility of cell contours

less visible cell contours

least damaged surface

Fig. 3. Scanning electron microscopic images of tomato outermost surface: (A) fresh control skin; (B) infrared heated skin; (C) hot lye heated skin; (D) steam heated skin.

X. Li et al. / Journal of Food Engineering 128 (2014) 79–87 83

was less damaged than the skin of lye and infrared treated samples(Fig. 3D). This observation supports the remark from Garcia andBarrett (2006b) that lye peeling of tomatoes is more efficient thansteam peeling. It is probable that steam causes less damage to theepidermal layers than the hot lye solution causes within the sametimeframe. Therefore, in order to achieve a sufficient degree of peelloosening, pressurized steam and a longer exposure to heat aretypically adopted by commercial steam peeling operations. Fromthe heat transfer perspective, infrared radiative heating has ahigher heat delivery rate and capability compared to convectiveheating of using steam. Within a very short duration (60 s), rapidradiative heating of infrared with a limited penetration depth(<1 mm) created a ‘‘heat shock’’ at tomato surface, causing thermaldamage of several layers of tomato epidermal cells and adjacentflesh tissues. This evidences the effectiveness of infrared heatingin promoting disruption of tomato epidermal layers towards peelloosening.

Cross-sectional images of the outer pericarp tissues of the fresh,infrared-, lye-, and steam-heated tomatoes are presented in Fig. 4.Fig. 4A shows that the fresh tomato dermal system consists of a

cuticle layer

A200µm

200µmC

epidermal cellhypodermal cell

parenchymatous cell

enlarged intercellular spaces

Fig. 4. Scanning electron microscopic images of cross-sectional images of tomato dermalheated skin.

cuticle layer, one-cell thick layer of epidermal cells, and a two tofour cell thick layer of thick-walled hypodermal cells. Adjacent tothe hypodermal layer, cells become larger and tend to be roundshaped (Fig. 4A). These round cells are parenchymatous cells andrepresent the edible flesh portion of tomato. In contrast to the freshcontrol, the infrared treated sample shows thermal expansion ofcell walls and separation of the cytoplasm from cell membrane(Fig. 4B). These anatomical differences indicate that the thermaleffect of infrared heating dramatically disrupts the tomatooutermost skin layer and adjacent pericarp cells. Unlike in theinfrared heated sample, cell wall expansion is not observed in lyetreated tomatoes (Fig. 4C). Instead, cytoplasm separation andenlarged intercellular spaces are observed, which indicate severedegradation of pectin substances in the middle lamella by lye thathas diffused into the pericarp. Thermal expansion of cell walls insteam treated samples was insignificant due to inefficient deliveryand transfer of heat by steam treatment as compared to infraredtreatment (Fig. 4D).

To provide more detailed views of the cytoplasm inside cellwalls, higher magnification cryo-SEM images were obtained, as

B200µm

200µmD

thermal expansion of cell walls

system: (A) fresh control; (B) infrared heated skin; (C) hot lye heated skin; (D) steam

84 X. Li et al. / Journal of Food Engineering 128 (2014) 79–87

shown in Fig. 5. Even with this higher magnification, it is impossi-ble to distinguish cell walls, middle lamella, and cytoplasm for theinfrared treated tomato tissue (Fig. 5B), which is in contrast to thefresh tomato tissue (Fig. 5A). Loss of cell integrity indicated infra-red heating caused severe thermal damage to certain cell layers,and resulted in mechanical failure of those cells and possible sub-sequent layer separation. Compared to infrared treated samples,lye treated tomato tissue exhibited thickened cell walls and icycrystals in the cytoplasm (Fig. 5C). These results are attributed tothe diffusion and penetration of lye solution into the interior of to-mato cells. Because no chemical or water was used in the infraredpeeling process, such changes did not occur in infrared treatedsamples (Fig. 5B). Radiation heat transfer is the dominant causeof skin loosening during the infrared dry-peeling process.

In theory, infrared irradiation first impinges on the skin of toma-to, then the heat penetrates inside the tissue through conductiveheat transfer. Thermal energy causes a sudden temperature increasein cell walls and inside cell fluids, and decreases the local resistanceor adhesion of cells. The tissue damage is the result of a combinationof vaporization of cell fluid, breakdown of pectin substances in themiddle lamella, and degradation of cell wall polysaccharides. Theresulting biochemical reactions and enzyme inactivation dependon the heating duration, penetration depth, and radiation intensity.From macroscopic appearance, it was inferred that heat treatmentcauses degradation of certain skin boundaries.

To validate this speculation, a magnified image was obtainedwith LM; it confirmed the layer separation (Fig. 6). The averagemeasured thickness of the separated layer obtained at multiplelocations was 0.69 ± 0.09 mm, which was in good agreement withmeasured peeled skin thickness (0.59 ± 0.35 mm) after 60 sinfrared heating. In contrast to the infrared and lye treated tissues,cytoplasmic content in steam treated samples were less damaged,which might be because the relatively low heat delivery capabilityof the steam used in this study (Fig. 5D). Similar to results obtainedwith infrared heating, the heat and diffusive mass transferproduced by steam treatment resulted in biophysical changes thatled to loss of cell rigidity and reduced turgor pressure (Floros andChinnan, 1988).

The observed layer separation can be correlated with peelloosening and the ease of peel removal. The contributions of cellstructure characteristics of tomato to the ease of peeling have beensummarized previously by Mohr (1990). The microstructural

A 100 m

100 m C

thicken cell walls

icy crystals

Fig. 5. Scanning electron microscopic images of cross-sectional high magnification imagheated skin; (D) steam heated skin.

changes noted in the images from this study confirmed that earlierreport (Mohr, 1990). For example, tomato cell size increased fromsmaller dermal cells towards the inner larger pericarp cells (Fig. 4).This abrupt cell size gradient coupled with thermal expansionresulting from infrared heating contributes to the tendency forlayer separation. Water vapor resulting from infrared heatingbuilds up at the interface where steep cell size transition exists,and enlarges the intercellular space. Thus, thermal effects presum-ably cause mechanical failure of the cells.

4.2. Transient membrane stress and peel cracking

Theoretical calculations were performed for a typical case of 60 sinfrared heating of a medium size tomato with a spherical radius of28 mm. Fig. 7A illustrates the predicted temperature profile and therise of internal vapor pressure at a place 0.6 mm under the skin withinfrared irradiation of 5000 W/m2 during infrared heating. The tem-perature profile was obtained from a computer simulation study oftomatoes subjected to the infrared dry-peeling process, and readersinterested in more details are referred to Li and Pan (2013a,b). With-in the 60 s infrared heating time, temperature increases markedlyfrom 23 �C to higher than 80 �C. Due to the increasing temperature,the internal vapor pressure gradually builds up (Fig. 7A), and eventu-ally causes mechanical failure of cells and skin layer separation orloosening. The effects of temperature and pressure on the resultingchanges in skin’s intrinsic strength are presented in Fig. 7B. Theoret-ically predicted skin membrane stress increases exponentially withtemperature. Analyzing all of the measured skin rupture data usingEq. (2) yields an average rupture stress of 1.18 MPa after 60 s infraredheating. The mean experimentally determined skin rupture stressand its corresponding 95% confidence intervals are also graphed inFig. 7B. At about 80 �C the predicted skin membrane stress valuewas virtually identical to the mean experimental value of skinrupture stress. Therefore, the results suggest that peel crackingmay begin when the skin membrane stress exceeds the rupturestress of tomato skin.

The 95% confidence interval ranging from 0.93 to 1.44 MPaestimates the overall uncertainty in the experimentally measuredmean value of skin rupture stress. The predicted skin membranestress intersected with the lower confidence limit value at about75 �C but did not intersect with the upper limit value (Fig. 7B). Thisresult indicates that occurrences of peel cracking may vary at

B 100 m

100 m D

layer separations

loss of cell integrity

es of tomato exocarp tissue: (A) fresh control; (B) infrared heated skin; (C) hot lye

layer separation

thickness

1 mm 1 mm

A B

Fig. 6. Cross-sectional light microscopic image of outer portion of tomato pericarp tissue: (A) tomato skin after infrared heating; (B) fresh tomato skin.

Time (s)

Tem

pera

ture

(o C)

0

20

40

60

80Pr

essu

re D

iffer

ence

(MPa

)

0.00

0.01

0.02

0.03

0.04

0.05

0.06

TemperatureInternal vapor pressure

Temperature (oC)

0 10 20 30 40 50 60

20 30 40 50 60 70 80 90

Stre

ss (M

Pa)

0.0

0.4

0.8

1.2

1.6

Mean of estimated rupture stress95% confidence interval

A

B

Fig. 7. Temperature, internal pressure, and stress changes during infrared heating:(A) temperature and pressure profiles during infrared heating; (B) membrane stresschanges with temperature.

Time Interval (s)35 - 40 40 - 45 45 - 50 50 - 55 55 - 60 60 - 65 65 - 70 70 - 75

Cou

nts

0

5

10

15

20

Cum

ulat

ive

Freq

uenc

y (%

)

0

30

60

90

120

Fig. 8. Frequency distribution of the recorded initial cracking time.

X. Li et al. / Journal of Food Engineering 128 (2014) 79–87 85

higher than certain temperature levels (�80 �C in the studied case)but do not always happen under the selected condition, which isbecause the skin membrane stress is less than the rupture stress.The measured cracking time supported this argument. First, 90%of the tomatoes in all the infrared peeling tests cracked within

75 s of heating. Second, the distribution of the cracking time, asshown in Fig. 8, reveals that 70% of cracks occurred within 60 sinfrared of heating and a majority of the cracking happened afterheating for 50–60 s. After 50–60 s heating, the temperature0.6 mm under the tomato skin was between 76 �C and 82 �C(Fig. 7A), which is consistent with the temperature range we foundin our previous Dynamic Mechanical Analysis.

5. Discussion and conclusion

The above experimental data and theoretical analysiselucidated the relationship between the increase in temperatureand increase in skin stress. Quantifying the dynamic tempera-ture-dependent stress developed in tomato skin is vital foraccurate analysis of peel cracking. From an engineering point ofview, the surface layer of tomato pericarp can be modeled as aporous media since it consists of solid cellular tissues with voidsin between. The total stress in the dermal layer can be decomposeinto effective stress assigned on the cellular skeleton and porepressure formed in the void space (Ho et al., 2013; Datta, 2007).Further in-depth study of stress generation could be coupled withtransport phenomena in the porous tissue of tomato surface layer,including the diffusion of cytosolic fluid and its capillary effects.Prediction of the time and temperature at which peel cracking willoccur is affected by several other factors in the practical infraredpeeling process, such as variable tomato shape and size, skinthickness, and skin permeability. For simplicity, in this study thepeeling mechanism was modeled using a spherical tomato withuniform temperature and thickness. For the actual elongated shapeof processing-tomato, the horizontal stress in the hoop direction issupposed to be greater than the vertical stress in the meridionaldirection due to the non-axial symmetry (Considine and Brown,1981; Timoshenko et al., 1959). According to the theory ofthin-walled shell behavior, cracking of a prolate spheroid type ofshape is most likely to occur in the longitudinal direction (stem-blossom axis direction) at a 45� incline. This theoretical rupturepattern agrees well with the pattern of observed peel cracking.Hence, the effect of different tomato shapes should be givenparticular attention in further investigations.

Skin permeability is another important factor to consider for acomplete explanation of the infrared peeling mechanism. Smallbubbles formed by water vapor were experimentally observed atthe stem scar in the last stage of infrared heating. These bubblesindicate that the tomato skin still functions as a vapor barrier be-fore rupture. If the skin was completely permeable, the skin wouldnot be under any stress at all and bubbles would not form. It isproposed that although the skin is slightly permeable to vapor

86 X. Li et al. / Journal of Food Engineering 128 (2014) 79–87

generated by infrared heating of leaking cytosolic fluids, an inter-nal pressure under the skin membrane is generated when the rateof the pressure build-up is larger than the rate of vapor leakagethrough skin. Impermeable skin was assumed for the theoreticalmodels in this study. There is a need to carry out further researchto determine more accurately the vapor permeability of tomatoskin.

From this study, we learnt that when the internal vapor pres-sure reaches a high enough level, the skin cracks because of thepressure-generated stress and reduced skin strength caused byinfrared radiation. In the design of a commercial infrared dry-peel-ing system, a vacuum chamber could be included to increase thepressure difference across the skin membrane and thus enhancethe occurrence of skin cracking. This idea was implemented andtested in a devised pilot scale infrared dry-peeling system fortomatoes. Nearly all tomatoes cracked after a sequential infraredand vacuum treatments, which validated the vacuum effect (Panet al., 2012). According to the theoretical model used in this study,peel cracking depends on pressure difference, the skin thicknessand the indenter tip radius. In practice, biological factors, such astomato maturity and cultivar characteristic also affects the occur-rence of peel-loosening and cracking.

It must note that the evolution of stress with pressure and tem-perature is a major but not the sole factor determining the skinseparation. Responses of tomato skin to temperature increasethrough other multi-physical and biochemical phenomena mayalso contribute to the skin rupture and degradation in skin innertissues. For example, thermal softening due to temperature in-crease at tomato surface may reduce the overall skin strengthand lower the critical rupture stress of tomato skin, leading to aneasier skin rupture. Various biochemical reactions along with therupture of cell walls and changes in moisture and temperature oc-cur at tomato dermal systems during infrared heating, contributingto the pronounced changes in skin’s dynamic biomechanical prop-erties. The heterogeneity and anisotropy nature of biomechanicalproperties of fruit skin may play an important role in the formationof skin separation resulting from infrared heating, particularly forpeel cracking. This speculation is true when extending the infrareddry-peeling technology to other fruits and vegetables with differ-ent and complex skin characteristics such as clingstone peach orBartlett pear (Li et al., 2014). To optimize the peeling performanceover different fruits and vegetables, the in-depth and comprehen-sive elucidation of infrared dry-peeling peeling mechanism is vitaland warrants further study.

In conclusion, the analysis presented in this study provides afundamental understanding of tomato peel loosening and crackingphenomena that occurred during infrared heating. Thermally in-duced skin separation was identified from LM images. SEM imagesrevealed that microstructure changes in tomato skin resulting frominfrared heating differ from changes induced by other peelingmethods using hot lye or steam. SEM images showed that infraredheating altered the organization of skin extracellular cuticles,caused expansion of cell walls, and damaged middle lamella. Itwas concluded that the mechanism underlying infrared dry-peeling was fundamentally different from the mechanisms of con-ventional hot lye and steam peeling. By combining experimentaldata with a theoretical model based on a thin-walled shell, thisstudy quantitatively analyzed the relationship between skinmicrostructure and the mechanical behavior of skin membrane.Infrared heating caused vaporization of escaped cytosolic fluidand thereby caused an increase of internal vapor pressure underthe loosened skin membrane. This increase in internal vapor pres-sure increased stress within the skin membrane. Peel crackingoccurred when skin membrane stress exceeded the critical rupturestress. For a typical infrared dry-peeling condition, it was deter-mined that peel cracking may occur when skin temperature

reaches to at least 80 �C. These results enhance understanding ofthe link between processing conditions and biomechanical changesof skin membrane. Although the experimental and theoreticalanalyses were limited to a typical scenario and were evaluated un-der certain assumptions, this integrated approach increases under-standing of infrared dry-peeling technology and provides insightinto ways to improve the design of infrared heating systems forpeeling tomatoes.

Acknowledgement

We would like to express our gratitude to Patrick Rooney fromthe Campbell’s Seed Inc. for his provision of tomato samplesthroughout this research. We appreciate Professor Kathryn McCar-thy and Tina William for providing facility support, and also thankProfessors David Hills Shrini Upadhyaya, and Mikal Saltveit of theUniversity of California at Davis for their technical advice.

References

Barreiro, J.A., Caraballo, V., Sandoval, A.J., 1995. Mathematical model for thechemical peeling of spherical foods. J. Food Eng. 25 (4), 483–496.

Barringer, S., 2003. Canned tomatoes: production and storage. In: Hui, Y.H., Ghazala,S., Graham, D.M., Murrel, K.D., Nip, W.K. (Eds.), Handbook of VegetablePreservation and Processing. CRC press, pp. 109–120.

Bourne, M.C., 2002. Food Texture and Viscosity: Concept and Measurement.Academic Press.

Calvin, S.E., Oyen, M.L., 2007. Microstructure and mechanics of the chorioamnionmembrane with an emphasis on fracture properties. Ann. N. Y. Acad. Sci. 1101(1), 166–185.

Considine, J., Brown, K., 1981. Physical aspects of fruit growth: theoretical analysisof distribution of surface growth forces in fruit in relation to cracking andsplitting. Plant Physiol. 68 (2), 371–376.

Das, D.J., Barringer, S.A., 2005. Potassium hydroxide replacement for lye (sodiumhydroxide) in tomato peeling. J. Food Process. Preserv. 30, 15–19.

Datta, A., 2007. Porous media approaches to studying simultaneous heat and masstransfer in food processes. I: Problem formulations. J. Food Eng. 80 (1), 80–95.

Domínguez, E., Cuartero, J., Heredia, A., 2011. An overview on plant cuticlebiomechanics. Plant Sci. 181 (2), 77–84.

Ferrua, M., Singh, R., 2009a. Modeling the forced-air cooling process of freshstrawberry packages, Part I: numerical model. Int. J. Refrig 32 (2), 335–348.

Ferrua, M., Singh, R., 2009b. Modeling the forced-air cooling process of freshstrawberry packages, Part III: experimental validation of the energy model. Int.J. Refrig 32 (2), 359–368.

Floros, J.D., Chinnan, M.S., 1990. Diffusion phenomena during chemical (NaOH)peeling of tomatoes. J. Food Sci. 55 (2), 552–553.

Floros, J.D., Chinnan, M.S., 1988. Microstructural changes during steam peeling offruits and vegetables. J. Food Sci. 53 (3), 849–853.

Floros, J.D., Wetzstein, H.Y., Chinnan, M.S., 1987. Chemical (NaOH) peeling asviewed by scanning electron microscopy: pimiento peppers as a case study. J.Food Sci. 52 (5), 1312–1316.

Garcia, E., Barrett, D.M., 2006a. Evaluation of processing tomatoes from twoconsecutive growing seasons: quality attributes, peelability and yield. J. FoodProcess. Preserv. 30 (1), 20–36.

Garcia, E., Barrett, D.M., 2006b. Peelability and yield of processing tomatoes bysteam or lye. J. Food Process. Preserv. 30 (1), 3–14.

Haman, D.Z., Burgess, G.J., 1986. Theoretical development for measuring the elasticproperties of spherical cuticular membranes. Trans. ASABE 29, 1470–1476.

Henry, Z., Allen, W., 1974. Use of hydrostatic pressure in development of stressstrain information for tomato skins. Trans. ASABE 99, 787–792.

Ho, Q.T., Carmeliet, J., Datta, A.K., Defraeye, T., Delele, M.A., Herremans, E., Opara, L.,Ramon, H., Tijskens, E., van der Sman, R., Van Liedekerke, P., Verboven, P.,Nicolaï, B.M., 2013. Multiscale modeling in food engineering. J. Food Eng. 114(3), 279–291.

Li, X., 2012. A study of infrared heating technology for tomato peeling: processcharacterization and modeling, PhD Dissertation. Department of Biological andAgricultural Engineering. University of California at Davis, Davis.

Li, X., Pan, Z., 2013a. Dry-peeling of tomato by infrared radiative heating: Part I.Model development. Food Bioprocess Technol. http://dx.doi.org/10.1007/s11947-013-1203-8.

Li, X., Pan, Z., 2013b. Dry-peeling of tomato by infrared radiative heating: Part II.Model validation and sensitivity analysis. Food Bioprocess Technol. http://dx.doi.org/10.1007/s11947-013-1188-3.

Li, X., Pan, Z., Atungulu, G.G., Zheng, X., Wood, D., Delwiche, M., McHugh, T.H., 2013.Peeling of tomatoes using novel infrared radiation heating technology. Innovat.Food Sci. Emerg. Technol.. http://dx.doi.org/10.1016/j.ifset.2013.10.011.

Li, Xuan, Zhang, Ang, Atungulu, Griffiths G., Delwiche, Michael, Milczarek, Rebecca,Wood, Delilah, Williams, Tina, McHugh, Tara, Pan, Zhongli, 2014. Effects ofinfrared radiation heating on peeling performance and quality attributes ofclingstone peaches. LWT-Food Sci. Technol. 55 (1), 34–42.

X. Li et al. / Journal of Food Engineering 128 (2014) 79–87 87

Li, X., Pan, Z., Upadhyaya, S.K., Atungulu, G.G., Delwiche, M., 2011. Three-dimensional geometric modeling of processing tomatoes. Trans. ASABE 54 (6),2287–2296.

Masanet, E., Worrel, E., Graus, W., Galitsky, C., 2007. Energy efficiency improvementand cost saving opportunities for the fruit and vegetable processing industry.Energy Analysis Department, Environmental Energy Technologies Division,Ernest Orlando Lawrence Berkeley National laboratory, University of California,Berkeley.

Matthews, R.F., Bryan, H.H., 1969. Lye peeling of Florida tomatoes – effects of time,temperature and concentration. Florida State Hortic. Soc. Proc. 81, 178–181.

Milczarek, R.R., McCarthy, M.J., 2011. Prediction of processing tomato peelingoutcomes. J. Food Process. Preserv. 35, 631–638.

Miles, J., Fridley, R., Lorenzen, C., 1969. Strength characteristics of tomatoessubjected to quasi-static loading. Trans. ASAE 12, 627–630.

Mohr, W.P., 1990. The influence of fruit anatomy on ease of peeling of tomatoes forcanning. Int. J. Food Sci. Technol. 25 (4), 449–457.

Pan, Z., Li, X., Bingol, G., McHugh, T., Atungulu, G., 2009. Development of infraredradiation heating method for sustainable tomato peeling. Appl. Eng. Agric. 25(6), 935–941.

Pan, Z., Li, X., Atungulu, G., Khir, R., McHugh, T., Delwiche, M., 2012. Development ofa infrared dry-peeling technology for tomatoes. Progress Report for theCalifornia Energy Commission (CEC) Project.

Poling, B.E., Prausnitz, J.M., John Paul, O.C., 2001. The Properties of Gases andLiquids. McGraw-Hill, New York.

Rock, C., Yang, W., Goodrich-Schneider, R., Feng, H., 2011. Conventional andalternative methods for tomato peeling. Food Eng. Rev., 1–15.

Schlimme, D.V., Corey, K.A., Frey, B.C., 1984. Evaluation of lye and steam peelingusing four processing tomato cultivars. J. Food Sci. 49 (6), 1415–1418.

Timoshenko, S., Woinowsky-Krieger, S., Woinowsky, S., 1959. Theory of Plates andShells. McGraw-hill, New York.

Toker, I., Bayndrl, A., 2003. Enzymatic peeling of apricots, nectarines and peaches.Lebensmittel-Wissenschaft und-Technologie 36 (2), 215–221.

Upadhyaya, S., Cooke, J., Gates, R., Rand, R., 1986. A finite element analysis of themechanical and thermal strength of avian eggs. J. Agric. Eng. Res. 33 (1), 57–78.

Upadhyaya, S., Cooke, J., Rand, R., 1985. A fluid-filled spherical shell model of thethermo-elastic behaviour of avian eggs. J. Agric. Eng. Res. 32 (2), 95–109.

Wang, Y., Li, X., Sun, G., Li, D., Pan, Z., 2013. A comparison of dynamic mechanicalproperties of processing-tomato peel as affected by hot lye and infraredradiation heating for peeling. J. Food Eng. 126, 27–34.

Wongsa-Ngasri, P., 2004. Ohmic heating of biomaterials: peeling and effects ofrotating electric field. PhD Dissertation. Department of Food, Agricultural, andBiological Engineering. The Ohio State University, Columbus, Ohio.