Characterization of Poly(ε-caprolactone)-Based NanocompositesContaining Hydroxytyrosol for Active Food PackagingAna Beltran,† Artur J. M. Valente,‡ Alfonso Jimenez,† and Marıa Carmen Garrigos*,†

†Analytical Chemistry, Nutrition & Food Sciences Department, University of Alicante, 03080 Alicante, Spain‡Department of Chemistry, University of Coimbra, 3004-535 Coimbra, Portugal

ABSTRACT: Antioxidant nanobiocomposites based on poly(ε-caprolactone) (PCL) were prepared by incorporatinghydroxytyrosol (HT) and a commercial montmorillonite, Cloisite30B (C30B), at different concentrations. A full structural,thermal, mechanical, and functional characterization of the developed nanobiocomposites was carried out. The presence of thenanoclay and HT increased PCL crystallinity, whereas some decrease in thermal stability was observed. TEM analysescorroborated the good dispersion of C30B into the PCL macromolecular structure as already asserted by XRD tests, because nolarge aggregates were observed. A reduction in oxygen permeability and an increase in elastic modulus were obtained for filmscontaining the nanoclay. Finally, the presence of the nanoclay produced a decrease in the HT release from films due to someinteraction between HT and C30B. Results proved that these nanobiocomposites can be an interesting and environmentallyfriendly alternative for active food packaging applications with antioxidant performance.

KEYWORDS: poly(ε-caprolactone), hydroxytyrosol, nanobiocomposites, characterization, active packaging

■ INTRODUCTION

Biodegradable and/or biobased polymers show a number ofproperties adequate to different applications, including foodpackaging, automotive, and biomedical fields.1 Most of thesematerials have properties comparable to many petroleum-basedplastics and are readily biodegradable, making them anattractive potential alternative to reduce the environmentalproblems induced by the accumulation of plastic waste.2

Among biodegradable polymers, aliphatic polyesters, such aspoly(ε-caprolactone) (PCL), are now commercially availableoffering an interesting alternative to conventional thermo-plastics. PCL can be synthesized either by ring-openingpolymerization (ROP) of the monomer, ε-caprolactone, witha variety of anionic, cationic and coordination catalysts or viafree radical ROP of 2-methylene-1-3-dioxepane.3 PCL is asemicrystalline polymer with a high degree of crystallinity,reaching 69%,4 but with this value decreasing at higher molarmasses. The good solubility of PCL in some common solvents,low melting point (59−64 °C), and exceptional blend-compatibility has raised some interest for the extensive researchon potential applications of PCL.3 However, some drawbacks inusing PCL as polymer matrix should be taken into account,particularly its poor thermal and mechanical resistance andlimited gas barrier properties. In this sense, PCL commercialuses are currently tempered by its high water solubility, highhydrophilicity, brittleness, low heat distortion temperature, highgas permeability, and low melt viscosity.5

The use of PCL formulations in food packaging applicationshas been recently evaluated by several authors. In fact, the maincurrent commercial application of PCL is in the manufacture ofbiodegradable bottles and compostable bags.6 Martinez-Abad etal. suggested that the combination of cold storage with PCLincorporating ciannamaldehyde, as a natural biocide, could besuitable for the controlled diffusion of this agent extending theshelf life of packaged food products.7 Antimicrobial nano-

composites of PCL with thymol were also developed bySanchez-Garcia et al.8 On the other hand, Perez-Masia et al.9used PCL to encapsulate dodecane developing coatingmaterials with energy storage capacity in refrigerationconditions. Blends of chitosan and poly-(ε-caprolactone) forfood packaging applications with good tensile strength and lowwater vapor permeability were studied by Swapna et al.,10

concluding that fruits and vegetables packaged in PCL filmswere expected to extend their storage life.In order to improve PCL properties, the incorporation of

nanoclays into this matrix is attracting some interest. It isknown that the addition of montmorillonites (MMT) incontents lower than 10 wt % to polymer matrices leads toremarkable increases in rigidity (elastic modulus), thermalstability, and barrier to gases and vapors.1 This strategy will beexplored in this study to limit the current PCL disadvantages infood packaging applications.In the last years, several authors have worked on the

preparation and characterization of PCL-based nanocompo-sites.11−13 Pantoustier et al.14 used the in situ intercalativepolymerization method and compared the properties ofnanocomposites prepared with both pristine MMT and aftermodification with amino-dodecanoic acid. Fukushima et al.developed nanocomposites of PCL with MMT and sepioliteshowing a good dispersion level of clays within the polymermatrix as well as thermomechanical improvement in theresulting nanocomposites.15

An additional functionality recently proposed for nano-composites is the controlled release of active substancesembedded in food packaging materials.1 Active packaging is

Received: November 14, 2013Revised: February 17, 2014Accepted: February 19, 2014Published: February 19, 2014

Article

pubs.acs.org/JAFC

© 2014 American Chemical Society 2244 dx.doi.org/10.1021/jf405111a | J. Agric. Food Chem. 2014, 62, 2244−2252

based on the release of specific compounds present in thepolymer formulation with controlled kinetics in particularenvironments.16 It is known that the addition of active agentsto polymer matrices avoids food degradation processesimproving the quality, safety, and health properties offoodstuff.17 In this sense, the combination of activetechnologies, such as the addition of antioxidant and/orantimicrobial agents to packaging materials, with the use ofnanocomposites can synergistically lead to formulations withbalanced properties and functionalities.8,18 Regarding thedevelopment of active films based on PCL, Salmieri et al.19

incorporated oregano, savory, and cinnamon essential oils(EOs) at 1% (w/w) as natural antioxidants to alginate/PCL-based films. Their results confirmed the plasticizing effect ofEOs in addition to their bioactive role, with the oregano-basedfilms demonstrating the most effective antiradical systems. β-carotene is another widely available natural antioxidant, and ithas been reported that the addition of low amounts of thiscompound to PCL resulted in a significant plasticization of thematrix improving its ductile properties.20 In this sense, the useof hydroxytyrosol (3,4-dihydroxyphenylethanol, HT), a com-pound with the highest antioxidant effect among the variouspolyphenols present in olives,21 could be a promising approachfrom an economical point of view, because it would turnagricultural residues into higher added-value active foodpackaging additives. In previous studies, certain decrease inthermal properties of natural antioxidants has been observed forcompression molded packaging materials at high temper-atures.22 However, hydroxytyrosol showed a good performancein polypropylene materials acting as antioxidant, and it mightbe considered as a promising alternative to use in activepackaging formulations.The novelty of this work lies on the use of HT as natural

antioxidant combined with a commercial nanoclay to developfilms for active packaging based on PCL. The addition of thenanoclay could help to improve the intrinsically poor PCLbarrier properties. Also, the use of HT is justified to improveshelf life of foodstuff by the antioxidant performance of thiscompound. At the best of our knowledge, there is no evidencein scientific literature of PCL-based packaging systems wherethe poor thermal resistance and barrier properties of PCL areovercome. In this way, this study is focused on the developmentand full characterization of new nanocomposites based on PCL,HT, and a commercial MMT (Cloisite30B, C30B), with highpotential to improve thermal and barrier properties inbiopolymers.23 Active nanobiocomposite films were obtainedby melt-blending followed by compression molding, whichwere further characterized on their thermal, structural,mechanical, and functional properties. In addition, the releaseof HT from these nanobiocomposites at different times wasalso evaluated.

■ MATERIALS AND METHODSMaterials. Poly(ε-caprolactone) (PCL, CAPA6800) com-

mercial grade (pellets) was kindly supplied by Perstorp HoldingAB (Sweden). Hydroxytyrosol (98% purity) (HT) was kindlysupplied by Fine & Performance Chemicals Ltd. (Middles-brough, U.K.). The commercial organo-modified montmor-illonite used in this study was Cloisite30B (C30B), and it wassupplied by Southern Clay Products (Austin, TX). Thisnanoclay was synthesized by replacing sodium ions in differentsilicate layers by methyl bis(2-hydroxyethyl) tallow alkylammonium cations by ion exchange.

Nanobiocomposites Preparation. PCL nanocompositeswere processed by melt blending in a Haake Polylab QC mixer(ThermoFischer Scientific, Waltham, MA) at 80 °C for 5 minat 100 rpm. Before processing, PCL and C30B were left in anoven at 50 °C for 20 h and 100 °C for 24 h, respectively, toeliminate moisture. Both additives were introduced in the mixeronce the polymer was in the melt state in order to avoidunnecessary losses. The 50 cm3 mixing chamber was filled with50 g of total mass. Eight different formulations were obtainedby adding different amounts of HT and C30B in concentrationsranges of 5−10 wt % and 2.5−5 wt %, respectively (Table 1).An additional sample without any additive was also preparedand used as control (PCL0).

Films were obtained by compression-molding at 120 °C in ahot-plate hydraulic press from Carver, Inc., model 3850(Wabash, IN). Materials were kept between the plates atatmospheric pressure for 5 min until melting, and then theywere successively pressed under 2 MPa (1 min), 3 MPa (1min), and lastly 5 MPa (5 min) to liberate the trapped airbubbles. The average thickness of the obtained films wasaround 200 μm, as measured with a 293 MDC-Lite DigimaticMicrometer (Mitutoyo, Japan) at five random positions (Table1), after 48 h of conditioning at 50% relative humidity (RH)and 23 °C. The final appearance of the films was completelyhomogeneous.

Nanobiocomposites Characterization. DifferentialScanning Calorimetry (DSC). Tests were conducted intriplicate by using a TA DSC Q-2000 instrument (New Castle,DE) under inert N2 atmosphere. Samples (3 mg) wereintroduced in aluminum pans (40 μL) and were submitted tothe following thermal program: heating from −90 to 160 °C (3min hold), cooling to −90 °C (3 min hold), and heating to 160°C, all steps at 10 °C min−1. The crystallization and meltingparameters were determined from the second heating scan. Thecrystallinity at room temperature (χc) of each material wasevaluated by using eq 1:

χ =Δ

Δ −×

⎛

⎝

⎜⎜⎜

⎞

⎠

⎟⎟⎟( )H

H 1100c

m

m0 % wt

100clay

(1)

where ΔHm is the specific melting enthalpy of the sample, andΔHm

o is the melting enthalpy of the 100% crystalline PCLmatrix (142.0 J g−1).10

Thermogravimetric Analysis (TGA). Thermogravimetricanalysis was performed in a TGA/SDTA 851e Mettler Toledothermal analyzer (Schwarzenbach, Switzerland). Samples (5mg) were weighed in alumina pans (70 μL) and were heated

Table 1. PCL Nanobiocomposites and Thickness (Mean ±SD, n = 5)

formulation HT (wt %) C30B (wt %) thickness (μm)

PCL0 204 ± 3PCL/HT5 5 202 ± 6PCL/HT10 10 206 ± 6PCL/C30B2.5 2.5 191 ± 4PCL/C30B5 5 198 ± 3PCL/HT5/C30B2.5 5 2.5 196 ± 5PCL/HT10/C30B2.5 10 2.5 201 ± 3PCL/HT5/C30B5 5 5 191 ± 2PCL/HT10/C30B5 10 5 198 ± 4

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf405111a | J. Agric. Food Chem. 2014, 62, 2244−22522245

from 30 to 700 °C at 10 °C min−1 under N2 atmosphere (flowrate of 50 mL min−1). Analyses were repeated three times foreach sample.Tensile Tests. Tensile tests were carried out by using a 3340

Series Single Column System Instron Instrument, LR30Kmodel (Fareham Hants, U.K.) equipped with a 2 kN load cell.Tests were performed in rectangular probes (100 × 10 mm2),an initial grip separation of 60 mm, and a crosshead speed of 25mm min−1. Before testing, all samples were equilibrated for 48h at 50 RH %. Tensile strength, elongation at break, and elasticmodulus were calculated from the resulting stress−strain curvesby following the ASTM D882-09 standard.24 Tests were carriedout at room temperature. Five repetitions were performed foreach film composition, and mean values were reported.Oxygen Transmission Rate (OTR). Oxygen transmission rate

was determined with an oxygen permeation analyzer (8500model Systech, Metrotec S.A, Spain). Tests were carried out byintroducing O2 (99.9% purity) into the upper half of thediffusion chamber while N2 was injected into the lower half,where an oxygen sensor was located. Films were cut into 14 cmdiameter circles for each formulation, and they were clamped inthe diffusion chamber at 25 °C. Tests were performed intriplicate and were expressed as oxygen transmission rate perfilm thickness (OTR·e).Scanning Electronic Microscopy (SEM). Micrographs were

obtained for surfaces and cross sections of cryo-fractured filmswith a JEOL JSM-840 microscope (Peabody, MA) running at12 kV. Samples were coated with gold layers prior to analysis toincrease their electrical conductivity. Images were registered atmagnifications of 100× and 500×.Transmission Electron Microscopy (TEM). Transmission

electron microscopy tests were performed by using a JEOLJEM-2010 microscope (Tokyo, Japan) operated at 100 kV.Films were previously ultramicrotomed to obtain slices of 100nm thickness (RMC, model MTXL).X-ray Diffraction (XRD). X-ray diffraction patterns were

recorded at room temperature in the 2.1−40° (2θ) range (stepsize = 0.01°, scanning rate = 8 s step−1) by using filtered Cu Kαradiation (λ = 1.54 Å). A Bruker D8-Advance diffractometer(Billerica, MA) was used to determine the interlayer spacing ofclay platelets.Color. Color changes on the films’ surfaces due to the

additives were analyzed with a KONICA CM-3600d COLOR-FLEX-DIFF2 colorimeter, HunterLab (Reston, VA). The L*,a*, b* system (CIELab) was employed; the L* axis representsthe lightness from black (L* = 0) to absolute white (L* = 100),the a* axis varies from green (−) to red (+), and the b* axisvaries from blue (−) to yellow (+). These parameters weremeasured at five different locations on each specimen, and

average values were calculated. The total color difference, ΔE*,between the control PCL film and nanocomposite films wascalculated by using eq 2:

Δ ∗ = Δ ∗ + Δ ∗ + Δ ∗E L a b[( ) ( ) ( ) ]2 2 2 1/2(2)

where ΔL*, Δa*, and Δb* are the difference between initialand final values (before and after the additives addition) of L*,a*, and b*, respectively.

HT Release from Films. The evaluation of the release rateof HT from each film at different times was carried out for 4cm2 samples immersed in 100 mL of methanol. The extractswere periodically taken and further analyzed by UV−visspectroscopy with a Shimadzu Spectrophotometer 2450(Kyoto, Japan). Spectra were obtained on extracts directlyinserted in the cell compartment, and they were studied overthe 200−300 nm wavelength range, which includes themaximum absorption wavelength of HT at 281 nm. Spectrawere acquired each minute for 1 h to ensure that the release ofthe active agent to the methanol solution had reached amaximum value. In order to quantify the amount of HTreleased, the specific extinction coefficient, ε (mg−1 L cm−1),for HT in methanol was calculated to be 0.0124 mg−1 L cm−1,through the measurement of the absorbance A at 281 nm of aset of HT standard solutions with concentration, c (mg L−1),and by using the Lambert−Beer law equation:

ε=A cl (3)

where l = 1 cm. In all experiments, temperature was keptconstant at 25.0 °C with a Multistirrer 6 thermostatic bath fromVelp Scientifica (Usmate, Italy). During HT release experi-ments, composite-containing solutions were stirred at ∼220rpm.

Statistical Analysis. Statistical analysis of results wasperformed with SPSS commercial software (Chicago, IL). Aone-way analysis of variance (ANOVA) was carried out.Differences between average values were assessed on thebasis of confidence intervals by using the Tukey test at a p ≤0.05 significance level.

■ RESULTS AND DISCUSSIONThermal Characterization. DSC tests were performed to

elucidate the effect of the nanoclay and HT addition on thethermal properties of the PCL matrix. Four parameters weredetermined: crystallization temperature, Tc; melting temper-ature, Tm; crystallization enthalpy, ΔHc; and melting enthalpy,ΔHm. These results are summarized in Table 2. It should behighlighted that the melting temperature of neat PCL waslower than that obtained for the rest of the materials formulatedin this study. This result gives an indication of the lower

Table 2. DSC Results Obtained for All Formulations in Nitrogen Atmosphere (Mean ± SD, n = 3)a

sample TC (°C) Tm (°C) ΔHC (J g−1) ΔHm (J g−1) χ (%)

PCL0 28 ± 2a 51 ± 2a 48 ± 1a 66 ± 4a 46 ± 1a

PCL/HT5 32 ± 2ab 56 ± 1b 53 ± 3b 70 ± 2a 49 ± 1b

PCL/HT10 27 ± 3a 55 ± 2ab 55 ± 1b 72 ± 2a 51 ± 2b

PCL/C30B2.5 35 ± 1b 59 ± 2b 58 ± 3b 72 ± 1a 52 ± 2b

PCL/C30B5 34 ± 4ab 58 ± 1b 53 ± 2b 69 ± 2a 51 ± 1b

PCL/HT5/C30B2.5 31 ± 1a 57 ± 3b 52 ± 1b 69 ± 3a 50 ± 1b

PCL/HT10/C30B2.5 28 ± 2a 56 ± 1b 49 ± 2ab 71 ± 2a 51 ± 2b

PCL/HT5/C30B5 28 ± 2a 56 ± 2b 51 ± 2ab 63 ± 2a 47 ± 2b

PCL/HT10/C30B5 27 ± 3a 55 ± 1a 49 ± 1ab 60 ± 4a 48 ± 1b

aDifferent superscripts within the same column indicate statistically significant different values (p < 0.05).

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf405111a | J. Agric. Food Chem. 2014, 62, 2244−22522246

crystallinity of neat PCL when compared to nanobiocompo-sites, as it has been reported by other authors.11

Regarding crystallinity values, this parameter increased withthe addition of HT (p < 0.05) due to the interaction betweenthe polymer matrix and the additive molecules, which can showsome plasticizing effects. On the other hand, the addition of thepolar C30B to the polyester matrix also resulted in a significantincrease (p < 0.05) of the crystallization rate and could modifythe PCL crystalline structure due to the ability of the nanoclayparticles to heterogeneously nucleate the crystallization of thepolymer matrix, as it has been well-documented for a widerange of nanocomposites.25 Similar results were obtained byPersico et al. in LDPE nanocomposite films containingcarvacrol, where differences in crystallinity values wereconsidered a consequence of two main factors: the presenceof heterogeneous nucleation sites and changes in chainmobility. In this sense, they reported that clay platelets couldact as nuclei for the initial heterogeneous nucleation andsubsequent growth of crystallites. Furthermore, they hypothe-sized that the increase in chain mobility promoted by carvacrolenhanced the ability of the polymer to crystallize.26

However, no significant differences (p > 0.05) were observedin crystallinity values for films containing 2.5 and 5 wt % ofnanoclay (Table 2). Similar results were found for the ternarynanobiocomposites containing the same amount of C30B (p >0.05), and a decrease was noted for those with 5 wt % ofnanoclay, although without significant differences (p > 0.05). Inthis sense, it was reported that the variations in the crystallinityby the addition of nanoclays can be accounted for by twofactors: nucleation that increases crystallinity and reduction inthe flexibility of polymer macromolecular chains that impedestheir rearrangement into ordered crystalline structures,reducing crystallinity. Both factors are related to nanoclaydispersion and content, and different results should be obtaineddepending on the prevailing effect.11 Furthermore, the additionof HT enhanced polymer chain mobility, promoting its self-nucleation and crystallinity. In conclusion, the neat crystallinityobserved for materials with both additives (HT and C30B) canbe attributed to the superposition of all these contributions.The reduction in chain mobility as the result of the nanoclayaddition can be considered an important factor in controllingthe final properties and structure developed by the material.TGA results indicated that the thermal degradation of PCL

in inert atmosphere took place through the rupture of thepolyester chains via ester pyrolysis reaction with the release ofCO2, H2O, and carboxylic acids. In the case of polyesters, suchas PCL, pyrolysis results in chain cleavages randomlydistributed along the macromolecular chains. It was reportedthat when two pyrolysis reactions occur with neighboring ester

functions, 5-hexenoic acid is one of the reaction products.13

TGA curves obtained for all the studied PCL nano-biocomposites just showed one main degradation step. Theinitial degradation temperatures (Tini) determined at 5% weightloss and maximum degradation temperatures (Tmax) obtainedfor all formulations are shown in Table 3. The incorporation ofthe active compound and the nanoclay brings about asignificant effect on the thermal stability of the obtainednanobiocomposites (p < 0.05). In this sense, the addition of 10wt % HT to PCL resulted in a 18 °C reduction in Tini. Thisresult could be explained by the plasticizing effect of HT toPCL, as already stated in the DSC study. In this sense, it wasreported that the addition of plasticizers to biopolyesters, suchas PLA, results in some decrease in the polymer thermalstability.27 On the other hand, the addition of the nanoclay alsoproduced some significant decrease (p < 0.05) in the PCLthermal stability, but this decrease was not enough to result inPCL degradation at processing temperatures. It was reportedthat layered silicates could catalyze PCL pyrolysis by a catalyticaction played by the nanoclay due to the presence of Lewisacidic sites, created upon organic modifier degradation.13

Mechanical Properties. The addition of low nanoclaycontent (less than 10 wt %) to polymers is expected to improvemechanical properties, particularly when effective nanoclayexfoliation occurs.22,1 The enhancement in mechanical proper-ties of polymer nanocomposites can be attributed to the highrigidity and aspect ratio of nanoclays together with the goodaffinity through interfacial interaction between polymer matrixand dispersed nanoclay.28

In this study, tensile tests were performed to evaluate theinfluence of the addition of the nanoclay and/or the activeadditive on the mechanical properties of PCL-based nano-biocomposites. Results are shown in Table 3, where the elasticmodulus E (MPa), elongation at break εB (%), and tensilestrength TS (MPa) were determined in all materials. Theaddition of HT to PCL resulted in some modification in E andε values (p < 0.05), whereas no significant differences wereobserved in TS (p > 0.05). In this sense, an increase in the HTconcentration caused an increase in ε and a decrease in E valuescompared to the neat polymer. This behavior could beexplained by the plasticizing effect of HT resulting in theincrease in ductility of the polymer, and it is consistent withresults obtained by other authors with the addition of similaradditives to biopolymer matrices.20,29 In this sense, it should benoted that the physical state of HT is a viscous liquid, with theconsequence of the intimate mixture of HT molecules withinthe macromolecular chains, permitting the enhancement on theinternal movement and consequently on ductile properties.

Table 3. Thermal (Mean ± SD, n = 3) and Mechanical Properties (Mean ± SD, n = 5) of the Studied Filmsa

sample Tini 5wt% (°C) Tmax (°C) εB (%) E (MPa) TS (MPa)

PCL0 348 ± 2a 415 ± 3a 20 ± 1a 507 ± 21a 12 ± 2a

PCL/HT5 350 ± 2a 412 ± 4a 29 ± 2b 453 ± 25b 10 ± 2ab

PCL/HT10 330 ± 3b 416 ± 1a 38 ± 2c 413 ± 25b 10 ± 1a

PCL/C30B2.5 300 ± 5c 409 ± 4a 40 ± 2c 687 ± 31c 9 ± 1a

PCL/C30B5 280 ± 3d 390 ± 2b 49 ± 1d 794 ± 24d 7 ± 1b

PCL/HT5/C30B2.5 330 ± 1b 405 ± 4a 36 ± 3c 491 ± 33a 9 ± 1ab

PCL/HT10/C30B2.5 330 ± 3b 411 ± 1a 41 ± 2cef 463 ± 18b 8 ± 2ab

PCL/HT5/C30B5 328 ± 3b 403 ± 3a 43 ± 2e 622 ± 29c 9 ± 2ab

PCL/HT10/C30B5 327 ± 4b 407 ± 3a 46 ± 1ef 732 ± 45d 7 ± 3ab

aDifferent superscripts within the same column indicate statistically significant different values (p < 0.05).

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf405111a | J. Agric. Food Chem. 2014, 62, 2244−22522247

On the other hand, the addition of the nanoclay caused asignificant increase in both the elastic modulus and elongationat the break of nanobiocomposites (p < 0.05). Concomitantly, adecrease (∼40%) of tensile strength (p < 0.05) was observedwhen the amount of the nanoclay in the film increased fromzero to 5 wt %. The constrained polymer model developed byBeall can be used to explain some of the general phenomenaobserved in polymer/clay nanocomposites, including changesin mechanical properties, such as modulus improvement andincreased elongation at break. In this sense, the constrainedpolymer region could be viewed as temporary cross-links, whichcertainly contribute to increase moduli but when strained canbreak and reform as the stress and strain increase. In examiningthe crazes formed in composites that exhibit this increase inelongation at break, it has been observed that fibrils form wherethe clay plate align along the direction of the fibril length. Thisgives a convenient platform for the polymer to break thetemporary cross-links and then reattach along the clayplatelets.30 For the ternary nanobiocomposites and regardingthe E values, it was observed that the effect of the nanoclay waspredominant for C30B5 ternary composites, resulting in a clearincrease in E values (p < 0.05), although a superposition of theinfluence of both additives (i.e., nanoclay and HT) should betaken into account to explain results obtained for C30B2.5materials. In fact, the ultimate behavior observed in mechanicalproperties can be considered as the result of several factors,such as crystallinity, polymer plasticization, filler−matrixinterface, all of which affect the stress−strain properties in acomplex way.26 Although there are difficulties in consideringantagonist effects by HT and C30B in ternary composites, itcould be concluded that the addition of the nanoclay inducedreinforcement effects provided by the high aspect ratio andsurface area of silicate layers. The improvement in theseproperties could be considered a clear indication of the highdispersion of the nanoclay through the polymer matrix andtheir high compatibility.31

Morphological Study. SEM micrographs of films (notshown) showed a featureless and nonporous morphology,indicating a good dispersion of C30B and HT in the PCLmatrix; that is, no phase separation was observed.The morphological characterization of these materials was

completed by transmission electron microscopy (TEM),allowing a qualitative evaluation of the dispersion degree ofthe nanoclay in the PCL matrix. TEM micrographs of PCLnanocomposites containing C30B showed partial exfoliation,because swollen and single dispersed clay layers were observed(Figure 1). Complete exfoliation of nanoclays into individualplatelets is not often achieved in nanocomposites, and

consequently, a mixed morphology consisting of intercalatedand exfoliated structures is usually obtained.1

In general terms, TEM tests suggested the good dispersion ofC30B in the PCL matrix, because no aggregates were observed.These results could be attributed to the strong interactionsbetween polymer and nanoclay, which are caused by thehydrogen bonding between the carbonyl group in the polymerstructure and the hydroxyl groups of the organo-modifiednanoclay.1

XRD Analysis. The dispersion of clays in nanocompositefilms was studied by using X-ray diffraction (XRD). The XRDpattern of PCL was characterized by the presence of twodistinct peaks at 2θ = 24.9° and 27.5° corresponding to the(110) and (200) planes, respectively, suggesting a semicrystal-line structure.14 No noticeable differences were found from theXRD patterns for all formulations in this angle range,suggesting that the polymer matrix structure was not influencedby the presence of the filler and/or HT.The most important variations in XRD patterns of these

PCL-based nanobiocomposites were found in the low anglerange (below 10°), which gives an indication of the claydispersion into the polymer matrix (Figure 2). C30B wascharacterized by a single diffraction peak at 2θ = 4.8°,corresponding to the (001) basal reflection,14 accounting forthe 18.6 Å interlayer distance. A shift of the clay diffraction peakto lower angles in these nanobiocomposites was observed,indicating a good interaction of the nanoclay with the polymermatrix, showing an increase in the interlayer distance up to 32.7Å for PCL/C30B5 due to the intercalation of PCL macro-molecular chains into the clay galleries. Luduena et al. reportedsimilar results for PCL/C30B nanocomposites showing a finald001 spacing value increased to 33.1 Å (79% from the initialvalue).32 These results could be correlated as well with thoseobtained for oxygen barrier properties, because an inversecorrelation between d-spacing and oxygen transmission rate wasapparent.28 However, there was no shift in the interlayerdistance in nanobiocomposites with different clay content,suggesting that clay load did not affect the clay platelets’intercalation. Similar results were found by Ahmed et al. forPCL/C30B films with nanoclay compositions ranging from 2.5to 10 wt %.12

In addition, some decrease in the C30B peak intensity wasobserved, in particular when the active additive was introducedin the formulation. Because C30B is characterized by thepresence of free hydroxyl groups, the short alkyl chains of theC30B organic modifier would make these groups available forinteractions with the polymer macromolecules and HT reactivefunctional groups, resulting in higher dispersion degree of thenanoclay in the polymer matrix.14 On the other hand, the basaldiffraction around 5.2° in nanobiocomposites’ XRD patternsmay correspond to a fraction characterized by a differentalkylammonium chain arrangement in the interlayer space or toa small amount of unmodified C30B.26

Barrier Properties. Barrier to oxygen is one of the mainissues in the design of materials for food packaging applications.The high sensitivity of many food products to oxygendegradation, microbial growth stimulated by moisture, andaroma retention needed for keeping the food quality requiresthe improvement in biopolymers barrier properties to gases,vapors, and aromas.1 In general, permeability of a polymer tooxygen or water vapor is dependent on some inter-relatedfactors, including polarity, hydrogen bonding between side

Figure 1. TEM images of neat PCL (A) and PCL/HT5/C30B2.5 (B)nanobiocomposite.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf405111a | J. Agric. Food Chem. 2014, 62, 2244−22522248

chains, molar mass and polydispersity, cross-linking, processingmethodology, and crystallinity.33

Results obtained for the oxygen transmission rate per filmthickness (e), OTR·e, for all the studied materials are shown inTable 4. Films containing nanoclay showed a slight decrease inOTR·e (p > 0.05) due to the nanoclay intercalation into thePCL structure. On the other hand, some increase in OTR·e wasobserved for samples with HT (p > 0.05). This effect could bedue to the plasticizing effect of HT resulting in the increase inthe mobility of PCL chains.34 A significant improvement inoxygen barrier (p < 0.05) was achieved for ternary nano-biocomposites containing C30B and 10 wt % HT, renderingmore competitive materials for oxygen-sensitive products. Thisresult also suggests that the incorporation of the active additivecould promote the intercalation of the nanofiller into the PCLmatrix by improving the dispersion of the nanoclay in the PCLmatrix and finally improving oxygen barrier properties. Asimilar effect was observed in nanobiocomposites based onplasticized PLA containing C30B.23 This improvement inbarrier properties may be explained by the theory developed byNielsen,35 based on the tortuous pathway that gas moleculesshould follow to diffuse through the polymer. This tortuosity isproduced by the good dispersion in the matrix of the layeredsilicates platelets and the consequent longer diffusion pathwayincreasing the diffusion time and decreasing permeability.33−36

According to Duncan,35 barrier properties are not onlyinfluenced by tortuosity but also by changes in the polymermatrix at the interfacial regions. In the case of favorableinteractions between polymer and nanofiller, polymer strands

close to each nanoparticle can be partially immobilized,working against the gases’ diffusion. In this sense, the Beall30

model states that the polymer region around the clay affectingdiffusion is the constrained region, characterized by lower freevolume and diffusion coefficient.Regarding the nanofiller content, no significant differences

were observed by comparing OTR·e results for films containing2.5 and 5 wt % of nanoclay (p > 0.05), suggesting a possiblefiller agglomeration in films with higher amounts of C30B. Inthis sense, Sanchez et al. found that composites containing 5 wt% of different nanofillers exhibited the highest oxygen barrierperformance, and that the addition of larger amounts ofnanoclays (10 wt %) to various biopolyesters did not result infurther oxygen barrier improvements.1 These results suggestthat there should be a balance between the filler content, thenanocomposite morphology, and the possibility of permeabilitydeterioration caused by filler agglomeration.22

Colorimetric Analysis. Color is an important factor to beconsidered for some industrial applications. Colorimetricanalysis of films was carried out to evaluate the effect of thepresence of the studied additives (Table 4) in this importantproperty. Significant differences in color were observed as theresult of the HT and/or C30B addition to PCL (p < 0.05). Inthis sense, neat PCL showed the highest L* value, with asignificant decrease in the presence of additives, especially forthe ternary nanobiocomposites, indicating a significantdarkening of these films. ΔE* values increased in ternarynanobiocomposites. A significant improvement in a* and b*values was also found for these films (p < 0.05), resulting in a

Figure 2. XRD patterns of C30B, PCL, and the obtained PCL bionanocomposites in the (2−10°) angle range.

Table 4. Color Coordinates (Mean ± SD, n = 5) and OTR·e Values from Active Films (e: Thickness, mm; Mean ± SD, n = 3)a

colorimetric parameters

sample OTR·e (cm3 mm m−2 day) L* a* b* ΔE

PCL0 51 ± 2a 83.6 ± 0.6a −0.9 ± 0.4a 2.4 ± 0.6a

PCL/HT5 54 ± 1a 79.1 ± 1.2b −0.6 ± 0.6a 6.1 ± 0.3b 2.5 ± 0.2a

PCL/HT10 56 ± 3a 75.5 ± 1.7c −0.7 ± 0.2a 9.8 ± 0.3c 3.1 ± 0.1b

PCL/C30B2.5 49 ± 4a 76.0 ± 2.2c −0.9 ± 0.2a 4.1 ± 0.2d 2.1 ± 0.2a

PCL/C30B5 49 ± 5a 75.4 ± 1.2c −1.3 ± 0.5a 6.9 ± 0.4b 2.6 ± 0.2a

PCL/HT5/C30B2.5 50 ± 4a 52.9 ± 0.8d 2.3 ± 0.2b 10.4 ± 0.4c 3.2 ± 0.2bc

PCL/HT10/C30B2.5 31 ± 1b 51.6 ± 2.9d 3.6 ± 0.8c 12.9 ± 04e 3.6 ± 0.3c

PCL/HT5/C30B5 40 ± 3c 47.5 ± 0.9e 3.1 ± 0.6c 11.9 ± 0.6e 3.4 ± 0.2bc

PCL/HT10/C30B5 32 ± 2b 40.6 ± 2.8f 2.9 ± 0.5c 10.1 ± 0.3c 3.2 ± 0.2bc

aDifferent superscripts within the same column indicate statistically significant different values (p < 0.05).

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf405111a | J. Agric. Food Chem. 2014, 62, 2244−22522249

slightly amber color. Similar color changes were reported forpolypropylene stabilized with hydroxytyrosol,22 contributing tostrengthen the color of the obtained films. Rhim37 also reporteda decrease in L* values and an increase in b* and ΔE* values infilms produced by blending agar with C20A and C30B,suggesting that differences in color might be attributed to thedispersion of the nanoclays in the polymer matrix.Release of HT from PCL Nanobiocomposites. The

combination of active additives and nanocomposites cansynergistically lead to materials with balanced properties andfunctionalities for food packaging applications. The formulationof novel antioxidant nanobiocomposites based on PCL and HTcan be considered as a way to control solubility and release ofthe antioxidant agent to foodstuff. In this sense, phenoliccompounds from virgin olive oil, such as HT, are known to playan antioxidant role in preventing damage to biomolecules. Oneof the most abundant components of olive oil is an oleuropeinderivative named oleacein (dialdehydic form of decarbox-ymethyl elenolic acid linked to HT).38 The evaluation ofequilibrium and kinetic properties of hydroxytirosol in thesenanobiocomposite films was carried out with methanol asrelease media. It is known that the tendency of HT (and otherphenols) to be solubilized, transferred, or diffused into a givensolvent is governed by thermodynamics. One of the primarythermodynamic factors describing this tendency is the activitycoefficient,39 which increases as solubility decreases. Althoughmethanol is not commonly used as food simulant, the activitycoefficient of HT at 25 °C in this solvent (0.20) is betweenthose shown in ethanol (0.027) and water (0.91),40 therebyallowing the exploration of broader applications of thesenanobiocomposites.Figure 3a shows the HT release for the three obtained

nanobiocomposite films containing HT 10 wt %. As indicated,

desorption equilibrium was reached after ∼1 h from thebeginning of the process. It is important to note that the releaseexperiment was very aggressive, because materials werecompletely immersed in methanol. The presence of thenanoclay leads to the decrease in the HT release, withcumulative release values of 7.9 and 7.7% for C30B2.5 andC30B5, respectively (Table 5). This behavior could be probablyrelated to interactions between C30B and HT. In this sense,

Sanchez et al. found an enhancement in thymol solubility inPCL nanocomposites due to the presence of nanoplatelets,possibly due to the retention of the active agent over theirsurface.1 They also found a decrease in the thymol diffusioncoefficient with the addition of nanoclays, as the result of thetortuous path imposed to the diffusion of the active agentthrough the nanocomposite bulk. These results suggest that it ispossible to control the release of natural agents with interest inthe design of novel active films and coatings throughincorporation of laminar nanoclays into bioplastics, such asPCL. A similar behavior was found for Pereira et al. for urea−montmorillonite nanocomposites, showing slower release whencompared to pure urea.41 This profile was probably related tothe fact that, as a consequence of some interaction betweenmontmorillonite and urea, this compound would need toadsorb in montmorillonite platelets after its dissolution; thisalso shows the importance of exfoliation in the controlledrelease.The assessment on the release mechanism was performed by

using the power law Korsemeyer−Peppas equation42

=∞C C kt/tn

(4)

where Ct and C∞ are the cumulative concentrations of HTreleased at time t and at infinite time, respectively, and k and nare fitting parameters, giving the later useful information on therelease mechanism. The validity of eq 4 is restricted to Ct/C∞ <0.60.43

Figure 3b shows the fitting of eq 4 to the experimental data,and the computed fitting parameters are summarized in Table5. It was observed that the desorption mechanism of HT inPCL/HT10 is non-Fickian (0.5 < n < 1), which is indicative ofcoupling diffusion and relaxation mechanisms. However, fornanobiocomposites, n values were approximately 0.5, showingthat HT desorption is diffusion-controlled (so-called Fickian).In this case, the apparent diffusion coefficients, Dap, can becomputed directly from k, by using the following equation

π= ∗k l D(4/ ) ( / )ap0.5

(5)

where l is the film thickness (cm). The calculated diffusioncoefficients for HT desorption from PCL/HT10/C30B2.5 andPCL/HT10/C30B5 were 2.9 (±0.5) × 10−10 cm2 s−1 and 4.5(±0.6) × 10−10 cm2 s−1, respectively. These values show thatthe release of HT under these conditions is very slow and alsosuggests that the diffusion follows the trend of the O2permeation; that is, an increase in the nanoclay content leadsto an increase in O2 permeation and HT diffusion. Taking intoaccount these data, it is evident that the transport mechanism israther complex and the reasons behind that are not yet wellunderstood.In conclusion, the effect of the addition of HT and C30B onproperties of PCL-based nanobiocomposites was evaluated byusing several analytical techniques. In this sense, some decreasein elastic modulus was observed for films containing HT,suggesting some plasticizing effect in the polymer matrix. Onthe other hand, the incorporation of C30B resulted in somedecrease in thermal stability and a significant enhancement inoxygen barrier and tensile properties, due to the successfulintercalation of the nanofiller into the matrix. Slight differencesin color with the addition of the additives for all films were alsoobserved. Nanobiocomposites showed slow release for HT,which is an important result for its potential application asPCL-based active films and coating systems, with eventual

Figure 3. (a) Desorption release profiles (a) and short-time range (Ct/C∞ < 0.6) desorption kinetics (b) of HT from PCL/HT10 (□), PCL/HT10/C30B2.5 (○), and PCL/HT10/C30B5 (△) films; solid lines(b) represent the fitting of eq 4 to desorption kinetic experimentaldata.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf405111a | J. Agric. Food Chem. 2014, 62, 2244−22522250

substitution of common synthetic antioxidants used inpackaging materials. The obtained PCL-based nanocompositeshave shown improved functional properties and can beregarded as potentially interesting materials for active packagingapplications within the food manufacturing and agriculturalsectors. To complete these investigations, further tests are to becarried out in order to evaluate the migration and antioxidantperformance of these nanobiocomposites by contact with food.

■ AUTHOR INFORMATIONCorresponding Author*E-mail: mc.garrigos@ua.es. Fax: +34 965903527. Tel.: +34965903400, Ext 1242.FundingFinancial support by the Spanish Ministry of Economy andCompetitiveness (no. MAT2011-28468-C02-01) and Univer-sity of Alicante (UAUSTI12-04) is also acknowledged.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSAuthors would like to thank Fine & Performance ChemicalsLtd. and Perstorp for kindly providing hydroxytyrosol and PCL,respectively.

■ REFERENCES(1) Sanchez, M. D.; Lagaron, J. M. Novel clay-based nano-biocomposites of biopolyesters with synergistic barrier to UV light,gas and vapour. J. Appl. Polym. Sci. 2010, 118, 188−199.(2) Luduena, L.; Vazquez, A.; Alvarez, A. Effect of lignocellulosicfiller type and content on the behavior of polycaprolactone based eco-composites for packaging applications. Carbohydr. Polym. 2012, 87,411−421.(3) Woodruff, M. A.; Hutmacher, D. W. The return of a forgottenpolymer: Polycaprolactone in the 21st century. Prog. Polym. Sci. 2010,35, 1217−1256.(4) Labet, M.; Thielemans, W. Synthesis of polycaprolactone: Areview. Chem. Soc. Rev. 2009, 38, 3484−3504.(5) Lee, J. E.; Kim, K. M. Characteristics of soy protein isolate−montmorillonite composite films. J. Appl. Polym. Sci. 2010, 118, 2257−2263.(6) Khan, R.; Beck, S.; Dussault, D.; Salmieri, S.; Bouchard, J.;Lacroix, M. Mechanical and barrier properties of nanocrystallinecellulose reinforced poly(caprolactone) composites: Effect of gammaradiation. J. Appl. Polym. Sci. 2013, 129, 3038−3046.(7) Martínez-Abad, A.; Sanchez, G.; Fuster, V.; Lagaron, J. M.; Ocio,M. J. Antibacterial performance of solvent cast polycaprolactone(PCL) films containing essential oils. Food Control 2013, 34, 214−220.(8) Sanchez-García, M. D.; Ocio, M. J.; Gimenez, E.; Lagaron, J. M.Novel polycaprolactone nanocomposites containing thymol of interestin antimicrobial film and coating applications. J. Plast. Film Sheeting2008, 24, 239−251.(9) Perez-Masia, R.; Lopez-Rubio, A.; Fabra, M. J.; Lagaron, J. M.Biodegradable polyester-based heat management materials of interestin refrigeration and smart packaging coatings. J. Appl. Polym. Sci. 2013,130, 3251−3262.

(10) Swapna, J. C.; Prashanth, H.; Rastogi, N. K.; Indiramma, A. R.;Reddy, Y.; Raghavarao, K. S. M. S. Optimum blend of chitosan andpoly-(ε-caprolactone) for fabrication of films for food packagingapplications. Food Bioprocess Tech. 2011, 4, 1179−1185.(11) Azeredo, H. Nanocomposites for food packaging applications.Food Res. Int. 2009, 42, 1240−1253.(12) Ahmed, J.; Auras, R.; Kijchavengkul, T.; Varshney, S. K.Rheological, thermal and structural behavior of poly(ε-caprolactone)and nanoclay blended films. J. Food Eng. 2012, 111, 580−589.(13) Oana, M. I.; Biqiong, C. Porous exfoliated poly(ε-caprolactone)/clay nanocomposites: Preparation, structure, andproperties. J. Appl. Polym. Sci. 2012, 125, 102−112.(14) Pantoustier, N.; Lepoittevin, B.; Alexandre, M.; Dubois, P.;Kubies, D.; Calberg, C.; Jerome, R. Biodegradable polyester layeredsilicate nanocomposites based on poly(ε-caprolactone). Polym. Eng.Sci. 2002, 42, 1928−1937.(15) Fukushima, K.; Tabuani, D.; Camino, G. Nanocomposites ofPLA and PCL based on montmorillonite and sepiolite. Mater. Sci. Eng.2009, 29, 1433−1441.(16) Costantino, U.; Bugatti, V.; Gorrasi, G.; Montanari, F.;Nocchetti, M.; Tammaro, L.; Vittoria, V. New polymeric compositesbased on poly(ε-caprolactone) and layered double hydroxidescontaining antimicrobial species. ACS Appl. Mater. Interfaces 2009, 1,668−677.(17) Fabra, M. J.; Hambleton, A.; Talens, P.; Debeaufort, F.; Chiralt,A. Effect of ferulic acid and α-tocopherol antioxidants on properties ofsodium caseinate edible films. Food Hydrocolloids 2011, 25, 1441−1447.(18) Abdollahi, M.; Rezaei, M.; Farzi, G. A novel activebionanocomposite film incorporating rosemary essential oil andnanoclay into chitosan. J. Food Eng. 2012, 111, 343−350.(19) Salmieri, S.; Lacroix, M. Physicochemical properties of alginate/polycaprolactone-based films containing essential oils. J. Agric. FoodChem. 2006, 54, 10205−10214.(20) Lopez, A.; Lagaron, J. M. Improvement of UV stability andmechanical properties of biopolyesters through the addition of β-carotene. Polym. Degrad. Stab. 2010, 95, 2162−2168.(21) Bubonja, M.; Giacometti, J.; Abram, J. Antioxidant andantilisterial activity of olive oil, cocoa and rosemary extractpolyphenols. Food Chem. 2011, 127, 1821−1827.(22) Peltzer, M.; Jimenez, A. Determination of oxidation parametersby DSC for polypropylene stabilized with hydroxytyrosol (3,4-dihydroxy-phenylethanol). J Therm. Anal. Calorim. 2009, 96, 243−248.(23) Martino, V. P.; Jimenez, A.; Ruseckaite, R. A.; Averous, L.Structure and properties of clay nano-biocomposites based onpoly(lactic acid) plasticized with polyadipates. Polym. Adv. Technol.2011, 22, 2206−2213.(24) ASTM D882-09. Standard test method for tensile properties ofthin plastic sheeting. 468 Annual Book of ASTM Standards. AmericanSociety for Testing and Materials: Philadelphia, PA, 2009.(25) Miltner, H. E.; Watzeels, N.; Gotzen, N. A.; Goffin, A. L.;Duquesne, E.; Benali, S.; Ruelle, B.; Peeterbroeck, S.; Dubois, P.;Goderis, B.; Assche, G.; Rahier, H.; Van Mele, B. The effect of nano-sized filler particles on the crystalline-amorphous interphase andthermal properties in polyester nanocomposites. Polymer 2012, 53,1494−1506.(26) Persico, P.; Ambrogi, V.; Carfagna, C.; Cerruti, P.; Ferrocino, I.;Mauriello, G. Nanocomposite polymer films containing carvacrol forantimicrobial active packaging. Polym. Eng. Sci. 2009, 49, 1447−1455.

Table 5. Concentration of HT Released in the HT 10 wt % Formulations

eq 4

sample HT released (%) n k / s−n R2

PCL/HT10 9.2 0.65 (±0.02) 0.008 (±0.001) 0.9942PCL/HT10/C30B2.5 7.9 0.56 (±0.01) 0.013 (±0.001) 0.9952PCL/HT10/C30B5 7.7 0.57 (±0.02) 0.016 (±0.002) 0.9914

R2: correlation coefficient.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf405111a | J. Agric. Food Chem. 2014, 62, 2244−22522251

(27) Martino, V. P.; Jimenez, A.; Ruseckaite, R. A. Processing andcharacterization of poly(lactic acid) films plasticized with commercialpolyadipates. J. Appl. Polym. Sci. 2009, 112, 2010−2018.(28) Rhim, J. W.; Park, H. M.; Ha, C. S. Bio-nanocomposites for foodpackaging applications. Prog. Polym. Sci. 2013, 38, 1629−1652.(29) Jamshidian, M.; Tehrany, E. A.; Imran, M.; Akhtar, M. J.;Cleymand, F.; Desobry, S. Structural, mechanical and barrierproperties of active PLA−antioxidant films. J. Food Eng. 2012, 110,380−389.(30) Beall, G. W. New conceptual model for interpretingnanocomposite behavior. In Polymer-Clay Nanocomposites; Pinnavaia,T. J., Beall, G. W., Eds.; John Wiley & Sons, Inc.: New York, 2000; pp267−279.(31) Quilaqueo-Gutierrez, M.; Echeverría, I.; Ihl, M.; Bifani, V.;Mauri, A. N. Carboxymethylcellulose−montmorillonite nanocompo-site films activated with murta (Ugni molinae Turcz) leaves extract.Carbohydr. Polym. 2012, 87, 1495−1502.(32) Luduena, L. N.; Kenny, J. M.; Vazquez, A.; Alvarez, V. A. Effectof clay organic modifier on the final performance of PCL/claynanocomposites. Mater. Sci. Eng. A. 2011, 529, 215−223.(33) Duncan, T. V. Applications of nanotechnology in foodpackaging and food safety: barrier materials, antimicrobials andsensors. J. Colloid Interface Sci. 2011, 363, 1−24.(34) Sothornvit, R.; Krochta, J. M. Oxygen permeability andmechanical properties of films from hydrolyzed whey protein. J.Agric. Food Chem. 2000, 48, 3913−3916.(35) Nielsen, L. E. Models for the permeability of filled polymersystems. J. Macromol. Sci. A. 1967, 1, 929−942.(36) Molinaro, S.; Romero, M. C.; Boaro, M.; Sensidoni, A.; Lagazio,C.; Morris, M.; Kerry, J. Effect of nanoclay-type and PLA optical purityon the characteristics of PLA-based nanocomposite films. J. Food Eng.2013, 117, 113−123.(37) Rhim, J. W.; Lee, S. B.; Hong, S. I. Preparation andcharacterization of agar/clay nanocomposite films: The effect of claytype. J. Food Sci. 2011, 76, 40−48.(38) Czerwinska, M.; Kiss, A. K.; Naruszewicz, M. A comparison ofantioxidant activities of oleuropein and its dialdehydic derivative fromolive oil, oleacein. Food Chem. 2012, 131, 940−947.(39) Queimada, A. J.; Mota, F. L.; Pinho, S. P.; Macedo, E. A.Solubility of biologically active phenolic compounds: Measurementsand modelling. J. Phys. Chem. B 2009, 113, 3469−3476.(40) Galanakis, C. M.; Goulas, V.; Tsakona, S.; Manganaris, G. A.;Gekas, V. A knowledge base for the recovery of natural phenols withdifferent solvents. Int. J. Food Prop. 2013, 16, 382−396.(41) Pereira, E. I.; Minussi, F. B.; Da Cruz, C. C. T.; Bernardi, A. C.C.; Ribeiro, C. Urea-montmorillonite-extruded nanocomposites: Anovel slow-release material. J. Agric. Food Chem. 2012, 60, 5267−5272.(42) Korsmeyer, R. W.; Gurny, R.; Docler, E.; Buri, P.; Peppas, N. A.Mechanisms of solute release from porous hydrophilic polymers. Int. J.Pharm. 1983, 15, 25−35.(43) Crank, J. The Mathematics of Diffusion, 2nd ed.; OxfordUniversity Press, Inc.: New York, 1975.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf405111a | J. Agric. Food Chem. 2014, 62, 2244−22522252