Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

15
Migration of a-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol Sung Wook Hwang a , Jin Kie Shim a,, Susan Selke b , Herlinda Soto-Valdez c , Laurent Matuana b , Maria Rubino b , Rafael Auras b,a Korea Packaging Center, The Korea Institute of Industrial Technology, Bucheon, South Korea b The School of Packaging, Michigan State University, East Lansing, MI 48823, USA c Centro de Investigación en Alimentación y Desarrollo, A.C., CTAOV, Hermosillo, Sonora 83304, Mexico article info Article history: Received 7 August 2012 Received in revised form 22 January 2013 Accepted 27 January 2013 Available online 8 February 2013 Keywords: PLLA Starch a-Tocopherol Resveratrol Blends abstract Poly(L-lactic acid) (PLLA)/starch blends with various concentrations of two natural antioxidants, a- tocopherol (a-TOC) and resveratrol, were fabricated by a melt blending and compression molding pro- cesses. The effects of the two antioxidants on the optical (color), thermal and mechanical properties of PLLA/starch blends with antioxidants were assessed. PLLA/starch blend films with a-TOC and resveratrol showed a yellowish color influenced by the combined effect of white starch and the brown color of the antioxidants. The glass transition and melting temperatures were significantly reduced with the addition of antioxidants while enhanced thermal stability was observed, which could be a benefit and important for processing and production. The enhanced mechanical properties could be attributed to not only a compatibilization effect based on the chemical linkage between PLLA and starch chains, but also restriction of the chain mobility by antioxidants. The release of resveratrol from PLLA and PLLA/starch blend films into ethanol followed Fickian behavior. The D values of a-TOC were in the range of 0.47–3.95 10 11 cm 2 s 1 for PLLA films and 0.70–6.83 10 11 cm 2 s 1 for PLLA/starch blend films at 13 °C, 5.67–13.0 10 11 cm 2 s 1 for PLLA films and 4.10–24.2 10 11 cm 2 s 1 for PLLA/starch blend films at 23 °C, and 89.0–118.0 10 11 cm 2 s 1 for PLLA films and 123–282 10 11 cm 2 s 1 for PLLA/starch blend films at 43 °C. The D values of resveratrol were in the range of 0.073–0.54 10 10 cm 2 s 1 for PLLA films and 1.42–6.93 10 10 cm 2 s 1 for PLLA/starch blend films at 13 °C, 0.90–3.44 10 10 cm 2 s 1 for PLLA films and 4.16–22.3 10 10 cm 2 s 1 for PLLA/starch blend films at 23 °C, and 24.8–74.1 10 10 cm 2 s 1 for PLLA films and 40.1–309 10 10 cm 2 s 1 for PLLA/starch blend films at 43 °C. Ó 2013 Elsevier Ltd. All rights reserved. 1. Introduction The development of new types of polymers synthesized from natural renewable resources has been widely investigated. Poly(lac- tic acid) (PLA) is one of the most promising polymers due to its bio- compatibility and biodegradability (Auras et al., 2004; Lim et al., 2008). The global lactic acid and PLA market is expected to grow by about US $3.8 billion by 2016 (Markets and Markets, 2011). PLA is currently used for single or multilayer films, trays, cups, and bot- tles, and can be manufactured by extrusion, thermoforming, injec- tion and blow molding processes for packaging applications. PLA is also suitable for the production and use of functional membranes (Auras et al., 2004; Manzanarez-López et al., 2011; Soto-Valdez et al., 2011). Although PLA shows good physical properties similar to polystyrene (PS) and polyethylene terephthalate (PET), the relatively high oxygen permeability of PLA and its brittleness are main drawbacks for flexible packaging applications, especially for foods and/or pharmaceuticals products susceptible to oxidation. Controlled release systems have been developed and are exten- sively being used in pharmaceutical, food and packaging applica- tions (Brayden, 2003). Specifically, these have been used in drug delivery systems. Application of these systems to food packaging has increased since they enable controlled release of active com- pounds such as antioxidants and antimicrobials from the packag- ing system at an appropriate rate during the storage of products, allowing protection and extension of the product’s shelf life. One of the first applications of this technology was reported to extend the shelf life of oatmeal cereal packaged in high density polyethyl- ene (HDPE) film with a high concentration of added butylated hydroxytoluene (BHT) (Miltz et al., 1988). Wessling et al. studied the migration and sorption behavior of a-Tocopherol (a-TOC) and BHT in low density polyethylene (LDPE) in contact with a fatty food simulant, and found that the migration of a-TOC into the food simulant was slower than that of BHT (Wessling et al., 1998). Byun 0260-8774/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jfoodeng.2013.01.032 Corresponding authors. Tel.: +82 032 624 4758; fax: +82 032 624 4770 (J.K. Shim), tel.: +1 517 432 3254, +1 517 355 0172; fax: +1 517 353 8999 (R. Auras). E-mail addresses: [email protected] (J.K. Shim), [email protected] (R. Auras). Journal of Food Engineering 116 (2013) 814–828 Contents lists available at SciVerse ScienceDirect Journal of Food Engineering journal homepage: www.elsevier.com/locate/jfoodeng

Transcript of Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

Page 1: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

Journal of Food Engineering 116 (2013) 814–828

Contents lists available at SciVerse ScienceDirect

Journal of Food Engineering

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

Migration of a-tocopherol and resveratrol from poly(L-lactic acid)/starchblends films into ethanol

Sung Wook Hwang a, Jin Kie Shim a,⇑, Susan Selke b, Herlinda Soto-Valdez c, Laurent Matuana b,Maria Rubino b, Rafael Auras b,⇑a Korea Packaging Center, The Korea Institute of Industrial Technology, Bucheon, South Koreab The School of Packaging, Michigan State University, East Lansing, MI 48823, USAc Centro de Investigación en Alimentación y Desarrollo, A.C., CTAOV, Hermosillo, Sonora 83304, Mexico

a r t i c l e i n f o

Article history:Received 7 August 2012Received in revised form 22 January 2013Accepted 27 January 2013Available online 8 February 2013

Keywords:PLLAStarcha-TocopherolResveratrolBlends

0260-8774/$ - see front matter � 2013 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.jfoodeng.2013.01.032

⇑ Corresponding authors. Tel.: +82 032 624 4758;Shim), tel.: +1 517 432 3254, +1 517 355 0172; fax: +

E-mail addresses: [email protected] (J.K. Shim), a

a b s t r a c t

Poly(L-lactic acid) (PLLA)/starch blends with various concentrations of two natural antioxidants, a-tocopherol (a-TOC) and resveratrol, were fabricated by a melt blending and compression molding pro-cesses. The effects of the two antioxidants on the optical (color), thermal and mechanical properties ofPLLA/starch blends with antioxidants were assessed. PLLA/starch blend films with a-TOC and resveratrolshowed a yellowish color influenced by the combined effect of white starch and the brown color of theantioxidants. The glass transition and melting temperatures were significantly reduced with the additionof antioxidants while enhanced thermal stability was observed, which could be a benefit and importantfor processing and production. The enhanced mechanical properties could be attributed to not only acompatibilization effect based on the chemical linkage between PLLA and starch chains, but alsorestriction of the chain mobility by antioxidants. The release of resveratrol from PLLA and PLLA/starchblend films into ethanol followed Fickian behavior. The D values of a-TOC were in the range of0.47–3.95� 10�11 cm2 s�1 for PLLA films and 0.70–6.83� 10�11 cm2 s�1 for PLLA/starch blend films at 13 �C,5.67–13.0 � 10�11 cm2 s�1 for PLLA films and 4.10–24.2 � 10�11 cm2 s�1 for PLLA/starch blend films at23 �C, and 89.0–118.0 � 10�11 cm2 s�1 for PLLA films and 123–282 � 10�11 cm2 s�1 for PLLA/starch blendfilms at 43 �C. The D values of resveratrol were in the range of 0.073–0.54 � 10�10 cm2 s�1 for PLLA filmsand 1.42–6.93 � 10�10 cm2 s�1 for PLLA/starch blend films at 13 �C, 0.90–3.44 � 10�10 cm2 s�1 for PLLAfilms and 4.16–22.3 � 10�10 cm2 s�1 for PLLA/starch blend films at 23 �C, and 24.8–74.1 � 10�10 cm2 s�1

for PLLA films and 40.1–309 � 10�10 cm2 s�1 for PLLA/starch blend films at 43 �C.� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

The development of new types of polymers synthesized fromnatural renewable resources has been widely investigated. Poly(lac-tic acid) (PLA) is one of the most promising polymers due to its bio-compatibility and biodegradability (Auras et al., 2004; Lim et al.,2008). The global lactic acid and PLA market is expected to growby about US $3.8 billion by 2016 (Markets and Markets, 2011). PLAis currently used for single or multilayer films, trays, cups, and bot-tles, and can be manufactured by extrusion, thermoforming, injec-tion and blow molding processes for packaging applications. PLAis also suitable for the production and use of functional membranes(Auras et al., 2004; Manzanarez-López et al., 2011; Soto-Valdezet al., 2011). Although PLA shows good physical properties similarto polystyrene (PS) and polyethylene terephthalate (PET), the

ll rights reserved.

fax: +82 032 624 4770 (J.K.1 517 353 8999 (R. Auras)[email protected] (R. Auras).

relatively high oxygen permeability of PLA and its brittleness aremain drawbacks for flexible packaging applications, especially forfoods and/or pharmaceuticals products susceptible to oxidation.

Controlled release systems have been developed and are exten-sively being used in pharmaceutical, food and packaging applica-tions (Brayden, 2003). Specifically, these have been used in drugdelivery systems. Application of these systems to food packaginghas increased since they enable controlled release of active com-pounds such as antioxidants and antimicrobials from the packag-ing system at an appropriate rate during the storage of products,allowing protection and extension of the product’s shelf life. Oneof the first applications of this technology was reported to extendthe shelf life of oatmeal cereal packaged in high density polyethyl-ene (HDPE) film with a high concentration of added butylatedhydroxytoluene (BHT) (Miltz et al., 1988). Wessling et al. studiedthe migration and sorption behavior of a-Tocopherol (a-TOC)and BHT in low density polyethylene (LDPE) in contact with a fattyfood simulant, and found that the migration of a-TOC into the foodsimulant was slower than that of BHT (Wessling et al., 1998). Byun

Page 2: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

S.W. Hwang et al. / Journal of Food Engineering 116 (2013) 814–828 815

et al. prepared PLA antioxidant film with a-TOC, BHT, andpolyethylene glycol 400 (PEG 400) by a film casting technique.PLA film with BHT and PEG 400 (BP-PLA) had antioxidant activitydue to the presence of BHT; however, it had less antioxidant activ-ity than PLA film with a-TOC, BHT and PEG 400 (ABP-PLA). Theantioxidant activity of the ABP-PLA film was significantly increasedby the addition of a-TOC into the BP-PLA film (Byun et al., 2010).

a-TOC is one of four tocopherols having antioxidant activity,and is the most abundant form in nature. It is a natural antioxidantpresent in grains like soybeans, cottonseed, and sunflowers. It hasbeen widely known to be effective in preventing lipid peroxidationand other radical oxidative reactions (Esterbauer et al., 1991;Tappel, 1962). Min and Boff reported that singlet oxygen oxidationproduced undesirable compounds in foods during processing andstorage, and carotenoids and tocopherols in foods can minimizesinglet oxygen oxidation (Min and Boff, 2002). Wessling et al. re-ported that the addition of above 360 ppm of a-TOC in LDPE de-layed the oxidation of linoleic acid at 6 �C. In addition, they alsosuggested possible change in the mechanical properties, color,and the oxygen permeability of LDPE film when a-TOC is added(Wessling et al., 2000). Manzanarez-López et al. reported that thediffusion of a-TOC from poly(L-lactic acid) (PLLA) films to ethanolsimulant showed Fick’s law behavior with diffusion coefficients (D)at levels between 3.16 ± 0.19� 10�11 and 5.29 ± 0.71� 10�11 cm2

s�1 at 23 and 33 �C, respectively. Diffusion to fractionated coconutoil was much lower than to ethanol. They also suggested the po-tential production of PLLA packages with added a-TOC for protec-tion of oily foods (Manzanarez-López et al., 2011). Soto-Valdezet al. also suggested that films produced with resveratrol couldbe used as antioxidant release membranes for a variety of pharma-ceutical, medical, and food applications (Soto-Valdez et al., 2011).

Resveratrol is a polyphenolic compound, which is a naturalantioxidant mainly present in grapes (Gambuti et al., 2004), grapejuice (Meng et al., 2004), wine (Mattivi et al., 1995), peanuts (Sobo-lev and Cole, 1999), and a number of other plant species (Kinget al., 2006). Resveratrol has relatively good thermal stability witha melting temperature of around 255 �C. The two aromatic rings inresveratrol showed a higher radical-scavenging capacity than pro-pyl gallate, ascorbic acid and a-TOC (see Fig. 1) (Soares et al., 2003).Mielink et al. found that the oxygen scavenging efficiency of grapeseed extracts with a high concentration of resveratrol increased

Fig. 1. Chemical structures of a-toco

with increasing antioxidant concentration from 0.4 to 1.6 g kg�1,suggesting potential application of resveratrol in active packaging(Mielnik et al., 2006). Soto-Valdez et al. recently prepared PLLAfilms containing 1 and 3 wt% resveratrol, and found that the diffu-sion of resveratrol from PLLA film into ethanol showed Fickianbehavior. The D values were found to be between 3.47 ± 0.10 �10�13 and 8.51 ± 0.38 � 10�10 cm2 s�1 between 9 and 43 �C (Soto-Valdez et al., 2011).

Many researchers have evaluated the release of antioxidantcompounds from homopolymers such as LDPE, HDPE and polypro-pylene (PP) to improve shelf life of food products (Miltz et al.,1988; Wessling et al., 2000). However, little research has been di-rected towards evaluating the dual release of antioxidants frompolymer blends, which may allow a more complex and targeted re-lease of antioxidants. The design of these novel functional mem-branes can create new opportunities for extending food products’shelf life and the creation of new pharmaceutical applications.

Thus, the overall goal of this work was to develop a PLLA/starchblend film containing two natural antioxidants, a-TOC and resvera-trol, for creating new functional membranes. In a previous series ofpapers, we have presented the properties of PLLA films with addeda-TOC and resveratrol and developed a PLLA/starch blend graftedwith maleic anhydride (MA) (Hwang et al., 2012a, 2012b, 2012c).In the work presented in this paper, the effect of adding both anti-oxidants into the PLLA/starch blend films on the optical, thermal,and thermo-mechanical properties was assessed. Specifically, thesimultaneous release kinetics of a-TOC and resveratrol fromPLLA/starch blend films to ethanol was evaluated and comparedwith the release kinetics of a-TOC and resveratrol from PLLA films.

2. Materials and methods

2.1. Materials

Poly(94% L-lactic acid), PLLA (4042D), resin was provided byNatureWorks LLC, Minnetonka, MN. a-TOC (97+ % purity) and res-veratrol (99% purity) were purchased from Alfa Aesar (Ward Hill,MA, USA) and ChromaDex Inc. (Santa Ana, CA, USA), respectively.The chemical structures of both antioxidants are given in Fig. 1.Dicumyl peroxide (DCP) and maleic anhydride (MA – briquettes,99%) were purchased from Sigma–Aldrich (St. Louis, MO, USA).

pherol (a-TOC) and resveratrol.

Page 3: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

816 S.W. Hwang et al. / Journal of Food Engineering 116 (2013) 814–828

Prior to processing, PLLA was dried in a vacuum oven at 60 �C for24 h, and starch was dried in a vacuum oven at 120 �C for 3 h to re-move residual moisture. All solvents used for quantification wereHPLC grade and obtained from Sigma–Aldrich (St. Louis, MO, USA).

2.2. Preparation of PLLA/starch blends with antioxidants

PLLA/starch blends with various concentrations of both a-TOCand resveratrol were prepared in an intermixer, Brabender Plasti-coder, PLE 331 (Brabender, Duisburg, Germany) at 190 �C with arotor speed of 60 rpm, and the total mixing time was fixed at12 min. An 80/20 (weight percent, wt%) ratio of PLLA/starch wasused in this study, and 0.1 phr of DCP and 2 phr of MA were fixedfor compatibilization of PLLA and starch. It was previously demon-strated that this blend provides the best film properties (Hwanget al., 2012a). The nominal and after processing composition andnomenclature of the films produced with the two antioxidantsare provided in Table 1. The PLLA/antioxidant mixtures werecooled to room temperature and then compression molded to afilm with a thickness of 220–270 lm by a hydraulic laboratorypress (Carver, model 3925, Wabash, IN) at 200 �C and 50 kg cm�2.All samples were kept at freezing conditions (�15 �C) to preventpossible migration of antioxidants from the films. PLLA films withsimilar composition of antioxidants were previously prepared. De-tails of the preparation technique for these films and their proper-ties can be found elsewhere (Hwang et al., 2012c).

2.3. Optical properties

The total color difference (DE�) of PLLA/starch blend films withtwo antioxidants were determined as described by the Commis-sion Internationale de l’Eclairage (CIE) L�a�b� color scale with aspectro-colorimeter (JS555, Color Techno System Corporation,Tokyo, Japan). The L�, a�, b�, and DE� can be determined accordingto the following equation:

DE� ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiDL�ð Þ2 þ Da�ð Þ2 þ Db�ð Þ2

q

DL� ¼ L�sample � L�standard

Da� ¼ a�sample � a�standard

Db� ¼ b�sample � b�standard

ð1Þ

The standard values were the initial value for PLLA/starch film.

2.4. Thermal properties

The glass transition temperature (Tg), melting temperature (Tm),and percent crystallinity (Xc) were obtained with a differential scan-ning calorimeter (DSC) (Diamond™ DSC, Perkin–Elmer, Waltham,MA, USA). The measurements were carried out under nitrogenatmosphere with a temperature range from 20 to 220 �C at a heating

Table 1Nominal and after processing composition of antioxidants in the PLLA and PLLA/starch bl

Sample Nominal After processing

a-TOC (mg/g resin) Resveratrol (mg/g resin) PLLA

a-TOC (mg/g resin)

T0R5 0 50 n/aT1R4 10 40 7.8 ± 0.2T2R3 20 30 13.2 ± 0.1T2R2 25 25 14.8 ± 0.2T3R2 30 20 20.8 ± 0.5T4R1 40 10 29.6 ± 0.9T5R0 50 0 n/a

n/a: Not available – samples were not produced.

rate of 10 �C/min, and were analyzed by Diamond Analysis software(Perkin–Elmer). The percent crystallinity was determined by the fol-lowing equation:

Xcð%Þ ¼DHm � DHc

DHcmð1� xÞ

� 100 ð2Þ

where DHm is the enthalpy of fusion, DHc is the enthalpy of coldcrystallization, DHc

m is the enthalpy of fusion of pure crystallinePLLA (DHc

m = 93.1 J g�1 (Fischer et al., 1973)), and x is the sum ofthe fractions of the antioxidants and starch.

A thermo-gravimetric analyzer (TGA Q600, TA instruments,New Castle, DE) was used to measure the change of sample weight.The analysis was performed up to 800 �C from room temperaturewith a rate of 10 �C min�1 under a nitrogen atmosphere with a flowrate of 20 sccm.

2.5. Thermomechanical properties

Dynamic mechanical analysis (DMA) of the PLLA/starch blendswith antioxidants was performed on a Perkin–Elmer dynamicmechanical analyzer, N535 (Perkin–Elmer). Samples with a widthof 6.0–6.5 mm and a length of 8–10 mm were cut for testing in ten-sion mode. The sample dimensions were carefully measured andthe mean values of 5 measurements of the samples were insertedinto the instrument. The experiments were carried out at a heatingrate of 3 �C min�1 and a temperature range of �20 to 170 �C with afrequency of 1 Hz. The storage modulus (G0) and loss tangent (tand)were measured for each blend sample in this temperature range.

2.6. Mechanical properties

Tensile strength, Young’s modulus, and elongation at breakwere obtained by a Universal Test Machine (UTM) (Instron 4465,Instron, Canton, MA) according to ASTM D 882-02. Samples werecut into strips with dimensions of 1 � 8 cm and conditioned at23 �C and 50% RH for 24 h before testing. The specimens weretested at a crosshead speed of 10 mm min�1 with a 4 cm initialgap separation.

2.7. Molecular weight measurement

The combined molecular weight of the PLLA/starch blends with-out and with antioxidants was determined with a gel permeationchromatograph (GPC) (Alliance GPCV 2000 system, Waters, Mil-ford, MA, US). Sample films (20 mg) were dissolved in 10 mL ofchloroform (CHCl3) (Merck, Darmstadt, Germany), and then100 lL of each sample solution were injected into the GPC. TheGPC was equipped with an isocratic pump, an autosampler, a seriesof 2 columns (Waters Styragel� HR5E and HR4E), and a refractiveindex detector. A flow rate of 1 mL min�1, a runtime of 40 min,

end films.

PLLA/starch

Resveratrol (mg/g resin) a-TOC (mg/g resin) Resveratrol (mg/g resin)

n/a n/a 48.2 ± 0.333.0 ± 0.8 7.6 ± 0.2 35.6 ± 0.125.4 ± 0.5 n/a n/a19.9 ± 0.1 19.7 ± 0.7 23.1 ± 0.518.1 ± 0.5 n/a n/a

6.9 ± 0.1 22.8 ± 0.4 9.1 ± 0.1n/a 36.7 ± 1.7 n/a

Page 4: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

S.W. Hwang et al. / Journal of Food Engineering 116 (2013) 814–828 817

and a temperature of 35 �C were used. The Mark-Houwink cor-rected constants K = 0.0131 (mL g�1) and a = 0.759 for dilute PLLAsolution in chloroform at 30 �C were used (Dorgan, 2010).

2.8. Quantification of a-TOC and resveratrol after processing film

To determine the actual concentration of a-TOC and resveratrolin the PLLA and PLLA/starch blend films after processing, the antiox-idants were extracted from the films (0.25 cm2 pieces) in methanolat 40 �C for 24 h in dark conditions. BHT (100 lg/mL) was intro-duced to protect the antioxidants from degradation during theextraction period. Quantification was performed with an ultra-per-formance liquid chromatograph (UPLC) (Waters ACQUITY UPLC,Milford, USA) equipped with a photodiode array detector. AnACQUITY BEH C18 Column (50 � 2.1 mm) (Waters) with an isocraticelution of methanol:water (98:02) at a flow rate of 0.15 mL/min and5 lL of injection volume was used. Calibration curves were gener-ated by using solutions with both a-TOC and resveratrol in metha-nol (0.50–50 lg/mL) with triplicates for both standard solutions(R2 = 0.9997 for a-TOC, R2 = 0.9993 for resveratrol.)

2.9. Diffusion of a-TOC and resveratrol from PLLA and PLLA/starchblend film

Diffusion of the antioxidants from the PLLA and PLLA/starchblend films to 100% ethanol (a food simulant) was performed usinga migration cell in accordance with ASTM D4754-98 (2003). Etha-nol was used as an alcoholic food simulant. The test cells consistedof 40 mL glass amber vials with screw tops with hole caps andPTFE/silicon sealing septa. Round test specimens (approximatediameter 18 mm) were cut from the PLLA and PLLA/starch films,and were immersed in 30 mL of ethanol. Specimens were sepa-rated by means of glass beads to avoid contact between films. Vialswere stored at each of the three temperatures (13, 23, and 43 �C)for a period of time long enough to determine the release of theantioxidants. Simulant samples were taken periodically during 24and/or 48 h for quantification.

The one dimensional diffusion solution equation of Fick’ssecond law for the two antioxidants, a-TOC and resveratrol, fromthe films to a limited volume solution was initially considered todetermine the diffusion coefficient (D) (Crank, 1975):

Mt

M1¼ 1�

X1n¼1

2að1þ aÞ1þ aþ a2q2

nexp �Dq2

nt=l2n o

ð3Þ

where l is thickness of the plastic film, qn are the non zero positiveroots of tanqn = �aqn, and a is expressed as

a ¼ VS

VP � KP;Sð4Þ

where VS and VP are molar volume of simulant and the plastic filmor package, KP,S is the partition coefficient of a-TOC and resveratrolbetween plastic film and the simulant.

The partition coefficient can be calculated from the ratio of theconcentration of a-TOC and resveratrol in PLLA and PLLA/starchblends (Cp,1) and the simulant (Cs,1) at equilibrium according tothe following equation:

KP;S ¼CP;1

CS;1ð5Þ

when the amount of solvent can be considered very large and a� 1since VS� VP and/or KP,S < 1, Eq. (3) can be simplified as:

Mt

M1¼ 1� 8

p2

X1m¼0

1

ð2mþ 1Þ2exp �Dð2mþ 1Þ2p2t=l2

n oð6Þ

where l is the thickness of the plastic film.

The mass of antioxidants diffused at time t, divided by the massof antioxidants diffused at equilibrium (Mt/M1) were plotted asfunction of the square root of time t (sec1/2) and the diffusion coef-ficients (D) were determined at each different temperature.

To determine the fit of the experimental data to Eqs. (3) and (6),the non-linear regression function in MATLAB� R2010a (Math-Works, Natick, MA, USA) was applied to the data (Araújo et al.,2005; Dhoot et al., 2009; Manzanarez-López et al., 2011; Masche-roni et al., 2010; Ortiz-Vazquez et al., 2011).

2.10. Determination of activation energy

To assess the effect of temperature on the diffusion behavior ofa-TOC and resveratrol from PLLA and PLLA/starch blend films intoethanol, the activation energy (Ea) was determined using theArrhenius equation for diffusion:

D ¼ D0 exp �Ea=RT½ � ð7Þ

where D is the diffusion coefficient, D0 is pre-exponential factor ofdiffusion, Ea is activation energy, R is gas constant (8.3145 J mol�1),T is temperature (K). Ea was obtained from the slope of a plot ofreciprocal temperature (1/T) vs. the logarithm of D (Ea = �slope �2.303R).

2.11. Statistical analysis

Statistical analyses of the results from the PLLA/starch blendswithout and with antioxidants were performed with the SPSS soft-ware (SPSS, Inc., Chicago, IL). One-way analysis of variance usingthe General Linear Model procedure and Tukey’s honestly signifi-cant difference (HSD) tests were used to determine significant dif-ferences (a < 0.05).

3. Results and discussion

The study about the production, and the characterization of theoptical, physical and mechanical properties of PLLA film addedwith two antioxidants, a-TOC and resveratrol has been previouslyreported (Hwang et al., 2012c). Those previously reported resultsare compared with PLLA/starch blend film added with those twoantioxidants in the following section.

The total amount of antioxidants in the PLLA and PLLA/starchfilms after processing is shown in Table 1. Around 72.9–81.5% oftotal initial antioxidant remained in the five PLLA films, and around73.5–96.4% of both antioxidants remained in the five PLLA/starchblends after processing, except for T4R1 where only 57% ofa-TOC remained. Soto-Valdez et al. reported that the loss of resve-ratrol from PLA films with 1–3% of resveratrol was 16.7% and25.5%, and Manzanarez-López et al. also found 15.7% loss ofa-TOC from 3% of a-TOC added to PLLA films (Soto-Valdez et al.,2011; Manzanarez-López et al., 2011). As reported, the loss ofa-TOC and resveratrol could be attributed to thermal degradationduring the mixing process in this study.

It is interesting to note that a higher amount of resveratrol re-mained in both PLLA and PLLA/starch blend films after processingthan did a-TOC. This could be related to the better thermal stabilityof resveratrol than that of a-TOC. Resveratrol is known to have rel-atively good thermal stability with a melting temperature of 255 �C(ChromaDex, 2009).

3.1. Optical properties

PLLA/starch blends without and with antioxidants displayed ayellowish color with various intensities as shown in Fig. 2. ThePLLA/starch blend film without antioxidant (T0R0) was used as a

Page 5: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

Fig. 2. The appearance of PLLA/starch blends without and with antioxidants.

Table 2L�, a�, b� and DE� of PLLA/starch blend films added with a-TOC and resveratrol.

Sample L� a� b� DE�

T0R0 55.3a �1.3a 3.6a –T0R5 49.5 ± 0.9b �4.3 ± 0.1b 19.4 ± 0.9b 17.0 ± 0.6a

T1R4 48.9 ± 0.7b �3.7 ± 0.0b 24.3 ± 1.8cd 21.8 ± 1.0b

T2R2 50.2 ± 0.8b �1.5 ± 0.2a 28.5 ± 0.6d 25.4 ± 0.4c

T4R1 49.8 ± 0.4b �1.3 ± 0.0a 22.0 ± 0.2bcd 19.1 ± 0.2d

T5R0 49.6 ± 0.8b �1.6 ± 0.1a 19.1 ± 0.7bc 16.5 ± 0.4a

Values in the same column with different superscript letters are significantly dif-ferent at a = 0.05; all of the values are expressed as average values ± standard error.

Table 3Thermal properties and % crystallinity of PLLA/starch blend films without and withantioxidants.

Sample Tg (�C) Tm (�C) Xc (%)

T0R0 53.1 ± 0.0a 153.5 ± 0.2a 3.1 ± 0.0a

T0R5 51.6 ± 0.4b 148.0 ± 0.2b 2.4 ± 0.5ab

T1R4 49.8 ± 0.1c 147.5 ± 0.2b 2.5 ± 0.1ab

T2R2 49.0 ± 0.1d 149.8 ± 0.1c 2.3 ± 0.5b

T4R1 48.4 ± 0.4e 150.2 ± 0.4cd 2.2 ± 0.3b

T5R0 47.4 ± 0.0f 150.4 ± 0.2d 2.1 ± 0.4b

Values in the same column with different superscript letters were significantlydifferent at a = 0.05; all of the values are expressed as average values and standarderror.

Fig. 3. DSC thermogram of PLLA/starch without and with antioxidants.

(a)

(b)

Fig. 4. TGA thermograms of PLLA/starch without and with antioxidants (a) wt% vs.temperature �C; and derivative weigth %/�C vs. �C.

818 S.W. Hwang et al. / Journal of Food Engineering 116 (2013) 814–828

reference film. The addition of the two antioxidants significantlyaffected the lightness (L�) and yellow (b�) color of the PLLA/starchblend films while minor differences in the green (a�) color of theblend films were observed (Table 2). The total color difference(DE) of the blend films was between 16.5 and 25.4 indicating thatthe addition of a-TOC and resveratrol had a great effect on colordifference, and this color difference was perceptible to the nakedeye, and T0R5 shows significant difference from T5R0. But, colorchanges among the antioxidant added films could not be differen-tiated (Fig. 2). The yellow color observed in the films could beattributed to the presence of quinones, organic compounds pro-duced due to the oxidization of the aromatic rings during process-ing of a-TOC and resveratrol.

In the case of PLLA films with these two antioxidants added,there was a significant difference between neat PLLA and PLLAfilms with antioxidants due to a-TOC having a clear yellowishbrown color and resveratrol having a dark brown color, indicatedby the higher a� (redness) and b� (yellowness) values.

3.2. Thermal properties

The results of the thermal analysis for PLLA/starch blends with-out and with antioxidants are presented in Table 3 and Fig. 3. Theaddition of the two antioxidants to PLLA/starch blends significantlydecreased Tg and Tm. This could be due to the plasticizing effect ofthe antioxidants on the PLLA/starch blends. In our previous work,3–5 �C reduction of Tg and Tm was found in PLLA/antioxidant filmwith a similar concentration of a-TOC and resveratrol (Hwang

Page 6: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

Fig. 5. DMA thermograms of G0 and G00 of PLLA/starch blends without and withantioxidants.

Fig. 6. Tand of PLLA/starch blends without and with antioxidants.

Table 4Mechanical Properties of PLLA/starch blends without and with antioxidants.

Sample Tensile strength (MPa) Modulus (GPa) Elongation at break (%)

T0R0 34.2 ± 1.9a 2.3 ± 0.1a 1.8 ± 0.3a

T0R5 48.2 ± 1.1b 2.3 ± 0.1a 2.3 ± 0.2b

T1R4 47.7 ± 5.9b 2.6 ± 0.3ab 2.2 ± 0.4ab

T2R2 51.5 ± 3.7b 2.5 ± 0.1ab 2.4 ± 0.0b

T4R1 48.2 ± 3.3b 2.4 ± 0.2ab 2.2 ± 0.2ab

T5R0 49.3 ± 3.4b 2.6 ± 0.0b 2.1 ± 0.2ab

Values in the same column with different superscript letters were significantlydifferent at a = 0.05; all of the values are expressed as average values and standarderror.

Fig. 7. Number average molecular weight (Mn), weight average molecular weight(Mw), and polydispersity index (PI) of PLLA/starch blends without and with twoantioxidants.

S.W. Hwang et al. / Journal of Food Engineering 116 (2013) 814–828 819

et al., 2012c). Similar results were recently reported by Soto et al.when adding resveratrol (Soto-Valdez et al., 2011). The additionof 1–3% of resveratrol also decreased Tm of PLA film (about 1 �C).Manzanarez et al. also found a small reduction of Tg and Tm withaddition of a-TOC in PLA film (about 1 �C) (Manzanarez-Lópezet al., 2011). In this study, Tg and Tm were reduced about 2–6 �C;this could be due to a higher concentration of both antioxidantsresulting in much greater plasticization effect. As seen in Fig. 3,the addition of the two antioxidants broadened the cold crystalli-zation temperature (Tcc) peak of PLLA/starch blends indicating that

the crystal growth of the PLLA chain was hindered by the antioxi-dants. This was confirmed by the significant decrease of Xc for thePLLA/starch blends with antioxidants.

Fig. 4 shows the thermal stability of the PLLA/starch blendswithout and with antioxidants. T0R5 displayed the highest thermalstability with an onset degradation temperature of 265–267 �C.The onset decomposition temperature decreased as the a-TOC con-tent increased. Arora et al. reported that a-TOC is thermally stableup to a temperature of about 240 �C; thereafter, it undergoes asharp mass loss, losing about 95% of its original mass at around460 �C (Arora et al., 2010). T5R0 showed a similar thermogram pat-tern. However, it can be seen that the addition of both antioxidantswith resveratrol amount larger than 20 mg/g enhances the thermalstability of PLLA/starch blends, providing additional benefit forindustrial production applications. Enhanced thermal stabilitydue to the addition of resveratrol was also found in PLLA film withantioxidants in our previous study (Hwang et al., 2012c).

3.3. Thermomechanical properties

Fig. 5a and b shows the storage modulus (G0) and loss modulus(G00) of the PLLA/starch blends without and with antioxidants. Theonset temperature in the transition region decreased as a-TOCcontent increased due to plasticization. The blends with only res-veratrol showed slight higher onset temperature, indicating theresveratrol did not have a critical role in changing Tg. In addition,the recrystallization temperature increased with addition of anti-oxidants. T5R0 had the lowest recrystallization temperature, indi-cating that addition of a-TOC induced slightly faster crystal

Page 7: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

Table 5Partition Coefficient (Kp,s) of a-TOC from PLLA and PLLA/starch blend films into ethanol at 13, 23, 43 �C.

Sample Parameter Temperature (�C)

13 23 43

PLLA PLLA/Starch PLLA PLLA/Starch PLLA PLLA/starch

T1R4 Kp,s 0.33 ± 0.02a⁄ 0.25 ± 0.02a⁄ 2.48 ± 0.09a⁄ 0.0029 ± 0.0007a 0.011 ± 0.001a 0.0018 ± 0.0005a

T2R3 Kp,s 3.66 ± 0.13b⁄ n/a 4.02 ± 0.20b⁄ n/a 0.008 ± 0.002b n/aT2R2 Kp,s 11.62 ± 1.06c⁄ 0.86 ± 0.06b⁄ 4.27 ± 0.08b⁄ 0.0113 ± 0.0027b 0.004 ± 0.000c 0.0007 ± 0.0001b

T3R2 Kp,s 6.52 ± 0.48d⁄ n/a 3.61 ± 0.24b⁄ n/a 0.007 ± 0.002b n/aT4R1 Kp,s 40.37 ± 2.13e⁄ 2.17 ± 0.24c⁄ 7.50 ± 0.77c⁄ 0.672 ± 0.026c⁄ 0.018 ± 0.001d 0.0015 ± 0.0003ab

T5R0 Kp,s n/a 4.00 ± 0.11d⁄ n/a 1.297 ± 0.173d⁄ n/a 0.0024 ± 0.0005c

Values in the same column with different superscript letters were significantly different at a = 0.05; All of the values are expressed as average values and standard error.n/a: Not available – samples were not produced.⁄Values were estimated since equilibrium was not obtained by the end of the test.

Table 6Partition Coefficient (Kp,s) of resveratrol from PLLA and PLLA/starch blend films into ethanol at 13, 23, 43 �C.

Sample Parameter Temperature (�C)

13 23 43

PLLA PLLA/Starch PLLA PLLA/Starch PLLA PLLA/Starch

T0R5 Kp,s n/a 0.13 ± 0.01a n/a 0.075 ± 0.007a n/a 0.043 ± 0.005a

T1R4 Kp,s 1.11 ± 0.03a⁄ 0.31 ± 0.02b 1.22 ± 0.01a 0.039 ± 0.006b 0.016 ± 0.003a 0.038 ± 0.007a

T2R3 Kp,s 0.85 ± 0.03b⁄ n/a 0.93 ± 0.08b n/a 0.025 ± 0.006a n/aT2R2 Kp,s 2.13 ± 0.02c⁄ 0.51 ± 0.07c 0.92 ± 0.18b 0.050 ± 0.004b 0.030 ± 0.011a 0.038 ± 0.008a

T3R2 Kp,s 1.46 ± 0.07d⁄ n/a 0.37 ± 0.02c n/a 0.025 ± 0.002a n/aT4R1 Kp,s 4.62 ± 0.29e⁄ 1.10 ± 0.23d 0.80 ± 0.04b⁄ 0.147 ± 0.021c 0.028 ± 0.005a 0.063 ± 0.005b

Values in the same column with different superscript letters were significantly different at a = 0.05; All of the values are expressed as average values and standard error.n/a: Not available – samples were not produced.⁄Values were estimated since equilibrium was not obtained by the end of the test.

820 S.W. Hwang et al. / Journal of Food Engineering 116 (2013) 814–828

growth, while the molecular hindrance from resveratrol resulted inslower crystal growth. In the PLLA film with antioxidants, elasticityalso increased with a higher concentration of resveratrol, while alarger plasticizing effect was observed for samples with higherconcentration of a-TOC (Hwang et al., 2012c).

The damping factor, tand (Fig. 6) was slightly affected by theaddition of the two antioxidants. T0R5 showed better dampingability in the temperature range of 100–120 �C compared to othersamples, due to better thermal stability.

3.4. Mechanical properties

The effects of both antioxidants on the tensile strength, Young’smodulus, and elongation at break of PLLA/starch blends withoutand with antioxidants are listed in Table 4. The tensile strengthand the Young’s modulus significantly increased with the additionof the two antioxidants. This could be attributed to not only thecompatibilization effect based on the chemical linkage betweenPLLA and starch chains, but also to the restriction of chain mobilityby the antioxidants. Elongation at break also slightly increased ascompared to PLLA/starch blends without antioxidants. Similarly,our previous work on PLLA films with both antioxidants showedincreased tensile strength and modulus, except for film with a highconcentration of a-TOC. The amount of a-TOC beyond some criticalpoint may weaken intermolecular forces between adjacent PLLAchains and cause disentanglement, resulting in poor mechanicalproperties in films with higher concentrations of a-TOC (Hwanget al., 2012c).

3.5. Molecular weight measurement

The weight average molecular weight of PLLA/starch blend filmswithout and with antioxidants before migration is given in Fig. 7.No significant change was found for samples with added resvera-trol and a-TOC. Jang et al. reported that extruded PLA/starch

blends at ratios of 50:50 and 70:30 showed decrease of Mw andbroadening of PI; this was attributed to existing moisture in starchleading to hydrolysis of PLA (Jang et al., 2007). In addition, reactivecompatibilization of MA onto PLLA and starch in PLLA/starchblends showed a significant effect on decreasing both Mn and Mw

due to the PLLA chain scission induced by hydrolysis of the PLLA/starch blends, as previously shown by Hwang et al. (2012b). In thisstudy, no significant decrease in molecular weight was found, andit appears that the addition of the two antioxidants reduced chainscission of the PLLA in the PLLA/starch blends. This is also consis-tent with the positive effect of the antioxidants on the thermal sta-bility of the films. Furthermore, the use of DCP as initiator resultedin no reduction of molecular weight, positively affecting the finalblend film properties.

3.6. Partition coefficients of a-TOC and resveratrol from PLLA andPLLA/starch blend films

Tables 5 and 6 summarize the partition coefficients (Kp,s) valueof a-TOC and resveratrol as indicated by Eqs. (5) and (4), respec-tively, for the PLLA and PLLA/starch films. The partition coefficientindicates the relative solubility of the migrants, a-TOC and resve-ratrol, between the polymers and ethanol at equilibrium. Kp,s < 1expresses a higher concentration of a-TOC or resveratrol in theethanol compared to the PLLA or PLLA/starch blend films. As seenin Tables 5 and 6, the Kp,s values for both a-TOC and resveratrolfor 43 �C were found to be the lowest, having the faster migrationbehavior from all PLLA and PLLA/starch blend films, and indicatingthat most of the antioxidant migrated to ethanol at equilibrium.The Kp,s values of PLLA/starch blend films for a-TOC or resveratrolshowed lower values at all three temperature as compared to thePLLA films. This could indicate that the amount of both antioxi-dants dispersed in starch could be higher than that of PLLA sincethe molecular interaction and affinity between the antioxidantsand starch could be greater than that of PLLA due to the larger

Page 8: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

Fig. 8. Diffusion of resveratrol (a) and a-tocopherol (b) from PLLA films to ethanol at 13, 23, 43 �C according to the Fick’s second law. The y-axis is the ratio of theconcentration of resveratrol and a-tocopherol in solution at time t to the concentration of resveratrol and a-tocopherol in solution at equilibrium (Mt/Meq), and the x-axis istime (t) in s. The central line (red) shows the best fit to the experimental data, and the outer lines (blue) are the predicted intervals for the experimental values. (Forinterpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

S.W. Hwang et al. / Journal of Food Engineering 116 (2013) 814–828 821

Page 9: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

Fig. 8. (continued)

822 S.W. Hwang et al. / Journal of Food Engineering 116 (2013) 814–828

Page 10: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

Fig. 9. Diffusion of resveratrol (a) and a-tocopherol (b) from PLLA/starch blend films to ethanol at 13, 23, 43 �C according to the Fick’s second law. The y-axis is the ratio of theconcentration of resveratrol and a-tocopherol in solution at time t to the concentration of resveratrol and a-tocopherol in solution at equilibrium (Mt/Meq), and the x-axis istime (t) in s. The central line (red) shows the best fit to the experimental data, and the outer lines (blue) are the predicted intervals for the experimental values. (Forinterpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

S.W. Hwang et al. / Journal of Food Engineering 116 (2013) 814–828 823

number of hydroxyl groups. Therefore, the antioxidants in thestarch phase would be expected to have faster migration than inPLLA resulting in lower Kp,s values in PLLA/starch blend systems.Partially, this could also be attributed to preferential swelling ofthe starch phase, which needs further research. Soto-Valdezet al. reported the Kp,s values of 1143.1 and 506.1 for PLLA filmwith 1 and 3 wt% of resveratrol at 9 �C, and lowest values werefound at 43 �C (Soto-Valdez et al., 2011). Kp,s values of 796.6,115.2, and 2.69 at 23, 33, and 43 �C were calculated from thestudy of PLLA with 2.58 wt% of a-TOC (Manzanarez-López et al.,2011). In this study, the Kp,s values for each antioxidant were

found to be much lower at low temperature and vice versa. ThePLLA and PLLA/starch blend films with two antioxidants were pre-pared by a compression molding process, and the molded filmswere thicker than sample films produced by the extrusion blow-ing process (Manzanarez-López et al., 2011; Ortiz-Vazquez et al.,2011; Soto-Valdez et al., 2011). Therefore, the migration behaviorof both antioxidants in films having differing morphology are dif-ferent from other studies resulting in different Kp,s values. Furtherinvestigation would be required to understand the effect of theprocessing and the preferential dissolution of the antioxidants inPLLA and starch.

Page 11: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

Fig. 9. (continued)

824 S.W. Hwang et al. / Journal of Food Engineering 116 (2013) 814–828

3.7. Diffusion of a-TOC and resveratrol from PLLA and PLLA/starchblend films

Figs. 8a and b and 9a and b show the diffusion of resveratrol anda-TOC from PLLA and PLLA/starch blend films into ethanol at 13,23, and 43 �C. The equilibration time decreased with increasingtemperature between 13 and 43 �C due to the temperature depen-dence of diffusion. In PLLA films, equilibrium for resveratrol wasreached after 167 and 14 h for all samples at 23 and 43 �C, respec-tively, except for T4R1 at 23 �C which did not reach equilibrium.Equilibrium was not reached for any samples at 13 �C. It was esti-mated using Eq. (6) that the equilibrium would be reached at 49(T1R4), 463 (T2R3), 509 (T2R2), 498 (T3R2), and 414 (T4R1) days,respectively, at 13 �C. For a-TOC, equilibrium was reached after

500 and 69 h at 23 and 43 �C, respectively. At 13 �C, equilibriumwas not reached after 833 h; it was estimated that equilibriumwould be reached after about 490 days except for T4R1 whichwould reach equilibrium at 89 days. For the PLLA/starch blendfilms, equilibrium for resveratrol was reached after approximately416, 111, and 13 h at 13, 23 and 43 �C, respectively, except forT4R1, which reached equilibrium after about 555, 222, and 20 h,respectively. It can be seen that the movement of resveratrol withrelatively low concentration could be hindered by a-TOC havinghigher concentration in PLLA/starch matrix. For a-TOC, equilibriumwas reached after 388 and 50 h at 23 and 43 �C, respectively; at13 �C, equilibrium was not reached after 777 h.

Diffusion coefficients (D) were estimated from Figs. 8a and band 9a and b according to Eq. (6) since a was�1 for all the samples

Page 12: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

Table 7Diffusion coefficient (D) of a-TOC from PLLA and PLLA/starch blend films into Ethanol at 13, 23, 43 �C.

Sample Parameter Temperature (�C)

13 23 43

PLLA PLLA/Starch PLLA PLLA/Starch PLLA PLLA/Starch

T1R4 D � 10�11 (cm2 s�1) 0.47 ± 0.00a 6.83 ± 0.05a 5.67 ± 0.05a 24.2 ± 0.5a 104.3 ± 3.0a 282 ± 4.0a

95% CI (0.36, 0.59) (6.08, 7.58) (5.18, 6.15) (22.3, 26.1) (99.3, 109.2) (264.4, 300.4)

T2R3 D � 10�11 (cm2 s�1) 0.78 ± 0.00b n/a 11.6 ± 0.4b n/a 96.7 ± 0.2b n/a95% CI (0.66, 0.90) n/a (10.8, 12.4) n/a (93.5, 99.9) n/a

T2R2 D � 10�11 (cm2 s�1) 0.76 ± 0.00b 4.63 ± 0.03b 9.73 ± 0.03c 22.2 ± 0.4b 107.3 ± 3.0a 239 ± 4.0b

95% CI (0.69, 0.82) (4.30, 4.95) (9.30, 10.17) (20.8, 23.6) (101.8, 112.0) (221.2, 256.1)

T3R2 D � 10�11 (cm2 s�1) 0.70 ± 0.00b n/a 13.0 ± 0.4d n/a 118.0 ± 3.0c n/a95% CI (0.61, 0.79) n/a (12.2, 13.8) n/a (112.1, 124.4) n/a

T4R1 D � 10�11 (cm2 s�1) 3.95 ± 0.54c 0.82 ± 0.00c 9.40 ± 0.04c 7.41 ± 0.03c 89.0 ± 0.3d 237 ± 4.0b

95% CI (3.47, 4.43) (0.72, 0.92) (8.75, 10.06) (7.01, 7.80) (84.8, 93.1) (221.4, 252.1)

T5R0 D � 10�11 (cm2 s�1) n/a 0.70 ± 0.00d n/a 4.10 ± 0.03d n/a 123 ± 3.0c

95% CI n/a (0.58, 0.83) n/a (3.84, 4.37) n/a (117.9, 128.8)

Values in the same column with different superscript letters were significantly different at a = 0.05; all of the values are expressed as average values and standard error.n/a: Not available – samples were not produced.

Table 8Diffusion coefficient (D) of resveratrol from PLLA and PLLA/starch blend films into ethanol at 13, 23, 43 �C.

Sample Parameter Temperature (�C)

13 23 43

PLLA PLLA/Starch PLLA PLLA/Starch PLLA PLLA/Starch

T0R5 D � 10�10 (cm2 s�1) n/a 6.93 ± 0.06a n/a 22.3 ± 0.4a n/a 309 ± 3.0a

95% CI n/a (5.88, 7.98) n/a (19.9, 24.9) n/a (284.9, 334.0)

T1R4 D � 10�10 (cm2 s�1) 0.54 ± 0.00a 4.16 ± 0.05b 2.30 ± 0.03a 16.5 ± 0.6b 74.3 ± 0.4a 263 ± 5.0b

95% CI (0.49, 0.58) (3.67, 4.64) (2.20, 2.40) (13.9, 19.1) (68.1, 80.4) (232.9, 293.6)

T2R3 D � 10�10 (cm2 s�1) 0.090 ± 0.00b n/a 3.44 ± 0.03b n/a 74.1 ± 0.4a n/a95% CI (0.081, 0.098) n/a (3.29, 3.59) n/a (67.6, 80.5) n/a

T2R2 D � 10�10 (cm2 s�1) 0.073 ± 0.00c 1.42 ± 0.12c 2.14 ± 0.04c 14.8 ± 0.7c 55.6 ± 0.3b 124 ± 6.0c

95% CI (0.066, 0.080) (1.04, 1.80) (2.00, 2.27) (12.5, 17.21) (52.0, 59.2) (106.7, 140.9)

T3R2 D � 10�10 (cm2 s�1) 0.073 ± 0.00c n/a 2.32 ± 0.03a n/a 48.4 ± 0.4c n/a95% CI (0.066, 0.080) n/a (2.19, 2.45) n/a (45.1, 51.8) n/a

T4R1 D � 10�10 (cm2 s�1) 0.085 ± 0.00b 1.66 ± 0.08d 0.90 ± 0.03d 4.16 ± 0.05d 24.8 ± 0.3d 40.1 ± 0.5d

95% CI (0.079, 0.092) (1.39, 1.94) (0.85, 0.94) (3.82, 4.50) (23.6, 26.0) (36.3, 44.0)

Values in the same column with different superscript letters were significantly different at a = 0.05; all of the values are expressed as average values and standard error.n/a: Not available – samples were not produced.

S.W. Hwang et al. / Journal of Food Engineering 116 (2013) 814–828 825

(see Tables 5 and 6). The D values are shown in Tables 7 and 8. Forresveratrol and a-TOC, the D values of all PLLA samples at 43 �Cwere one order of magnitude higher than the D values at 23 �C,which also were one order of magnitude higher than the D valuesat 13 �C. Diffusion of resveratrol and a-TOC from PLLA/starchblends at 13, 23 and 43 �C were at least one order of magnitudehigher than the D values for PLLA films. It was also found thatthe D values for resveratrol show a higher dependence on resvera-trol concentration in PLLA and PLLA/starch blend films than the Dvalues for a-TOC did for a-TOC concentration. PLLA and PLLA/starch blend films with two antioxidants are new types of func-tional release membranes, so literature data was not available forone to one comparison.

Therefore, comparison with films with one type of antioxidantwas conducted. Soto-Valdez et al. evaluated resveratrol diffusionfrom PLA films with 3 wt% of resveratrol into 100% ethanol andreported D values of 3.49 � 10�13, 3.06 � 10�11, 4.17 � 10�10, and8.26 � 10�10 cm2 s�1 at 9, 23, 33, and 43 �C, respectively(Soto-Valdez et al., 2011). The resveratrol D values in this study are oneorder of magnitude higher than in Soto-Valdez’s study. Co-existenceof a-TOC which acts as a plasticizer in the PLLA matrix could formmore void volume, and then results in faster movement of resvera-

trol. This also could be attributed to the processing technique usedto produce the PLLA films and the type of their final structure.Soto-Valdez’s films were produced by extrusion blowing whereasthe films in this work were produced by compression molding.Heirlings et al. studied D values for a-TOC from polymers such asLDPE and ethylene vinyl acetate copolymer (EVA) in 95% ethanol,and reported D values of 2.64 � 10�11 and 4.23 � 10�11 cm2 s�1,respectively, at 7 �C (Heirlings et al., 2004). Granda-Restrepoet al. reported D values of 2.34 � 10�11 cm2 s�1 for a multilayerfilm including a LDPE layer with 4 wt% a-TOC in contact withwhole milk powder at 20 �C (Granda-Restrepo et al., 2009). Man-zanarez-López et al. recently prepared PLA films containing2.58 wt% of a-TOC, and evaluated the kinetics of release of a-TOCfrom the PLLA films to ethanol at temperatures between 23 and43 �C (Manzanarez-López et al., 2011). D values of 3.16 � 10�11,5.29 � 10�11, and 38.0 � 10�11 cm2 s�1 at 23, 33, and 43 �C, respec-tively. The D value for a-TOC in this study also was one order ofmagnitude higher in value, which can be partially attributed tothe co-existence of a-TOC and resveratrol having a completely dif-ferent final morphology and structure. From this study, migrationfrom the dual antioxidant system appears to be faster than thatfrom the single antioxidant system. The D value for resveratrol is

Page 13: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

Fig. 10. The activation energy of the diffusion of resveratrol (a) and a-tocopherol and (b) from PLLA/starch blend films to ethanol, and the slope of each line was equal to�Ea/2.303R.

Table 9Activation energy of the diffusion (Ea) of resveratrol and a-TOC into Ethanol fromPLLA and PLLA/starch blend films.

Sample Activation energy (kJ mol�1)

PLLA PLLA/starch

Antioxidant a-TOC Resveratrol a-TOC Resveratrol

T0R5 n/a n/a n/a 96.2 ± 0.1a

T1R4 125.1 ± 0.2a 132.6 ± 0.1a 92.4 ± 0.0a 104.5 ± 0.0b

T2R3 162.5 ± 0.3b 116.3 ± 0.4b n/a n/aT2R2 161.8 ± 0.3b 120.4 ± 0.4c 96.0 ± 0.1b 108.5 ± 0.3c

T3R2 157.5 ± 0.4c 123.5 ± 0.4d n/a n/aT4R1 140.6 ± 0.1d 79.2 ± 0.1e 135.8 ± 0.1c 80.8 ± 0.1d

T5R0 n/a n/a 121.4 ± 0.0d n/a

Values in the same column with different superscript letters were significantlydifferent at a = 0.05; all of the values are expressed as average values and standarderror.

826 S.W. Hwang et al. / Journal of Food Engineering 116 (2013) 814–828

higher than that for a-TOC due to its smaller molecular size givinghigher mobility. In addition, although improved interfacial adhe-sion between PLLA/starch blends was found in our previous study(Hwang et al., 2012b), the morphological instability of the interfa-cial region in the PLLA/starch blends could result in much higher Dvalues as compared to PLA-resveratrol and PLA- a-TOC single sys-tems. In addition, the preferential dispersion of two antioxidants inthe starch phase due to more hydroxyl groups could result in dif-ferent diffusion behavior compared to that in the PLLA phase inthe PLLA/starch blends. The chain softening at 43 �C, which is nearthe Tg of PLLA, could accelerate the diffusion of resveratrol anda-TOC from PLLA and PLLA/starch blend films as well. Release ofa-TOC and resveratrol for PLLA and PLLA/starch blends in othersimulants such as water and mygliol should be much slower thanin ethanol as previously reported (Manzanarez-López et al., 2011;Soto-Valdez et al., 2011).

n/a: Not available – samples were not produced.

3.8. Activation energy of diffusion (Ea) a-TOC and resveratrol for PLLAand PLLA/starch blend films into ethanol

A plot of log(D) vs. T�1 for resveratrol and a-TOC (Fig. 10a andb) produced a straight line for all PLLA (R2 = 0.9597–0.9955) andPLLA/starch blend samples (R2 = 0.9627–0.9993) For resveratrol,the PLLA films had an Ea of 132.6 ± 0.1, 116.3 ± 0.4, 120.4 ± 0.4,123.5 ± 0.4, and 79.2 ± 0.1 kJ mol�1for T1R4, T2R3, T2R2, T3R2,and T4R1, respectively, and Ea values of 125.1 ± 0.2, 162.5 ± 0.3,161.8 ± 0.3, 157.5 ± 0.4 and 140.6 ± 0.1 kJ mol�1 for T1R4, T2R3,T2R2, T3R2, and T4R1 kJ mol�1, respectively, were found fora-TOC. For PLLA/starch blend films, for resveratrol the Ea of

96.2 ± 0.1, 104.5 ± 0.0, 108.5 ± 0.3, and 80.8 ± 0.1 kJ mol�1 werefound for T0R5, T1R4, T2R2, and T4R1, respectively, and Ea ofT1R4, T2R2, T4R1, and T5R0 kJ mol�1 for a-TOC were found to be92.4 ± 0.0, 96.0 ± 0.1, 135.8 ± 0.1, and 121.4 ± 0.0 kJ mol�1, respec-tively (R2 = 0.8628–0.9971) (see Table 9). The Ea values for resvera-trol obtained from both PLLA and PLLA/starch blend were lowerthan the reported value for PLA-resveratrol (176 kJ mol�1) whilethe Ea of a-TOC for PLLA and PLLA/starch blend films were higherthan reported values from PLA-a-TOC (96.2 kJ mol�1) except forT1R4 and T2R2 (Manzanarez-López et al., 2011; Soto-Valdezet al., 2011). It was reported that less energy is needed for diffusion

Page 14: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

S.W. Hwang et al. / Journal of Food Engineering 116 (2013) 814–828 827

of a-TOC from PLA to ethanol compared to diffusion of resveratrolsince resveratrol has two aromatic rings and three hydroxylgroups, allowing the molecule to interact with the polar groupsin PLLA and thus requiring more energy to diffuse among the PLAchains (Manzanarez-López et al., 2011). Thus, a-TOC requiredmuch higher energy to diffuse from PLLA in the presence of resve-ratrol as compared to when it is alone. This can be attributed tocoexistence of two antioxidants, a-TOC and resveratrol, havingmolecular interaction within PLLA and PLLA/starch blends result-ing in restriction of molecule movements among them. Even lowerenergy was required for PLLA/starch blends as compared to PLLAfilms. PLLA/starch blends can be used to increase and to tailorthe release of a-TOC and resveratrol since higher diffusion andlower energy barrier is present than in PLLA films as shown bythe lower slope of the Arrhenius plots (Fig. 10). MA was used asa compatibilizer between PLLA and starch, reacting with the hydro-xyl groups of PLLA and starch to improve the interfacial adhesionof the PLLA/starch blend. This high interaction between MA andhydroxyl group of PLLA and starch would give less chance for res-veratrol to have molecular interaction with PLLA and starch, result-ing in the lower Ea.

4. Conclusions

PLLA and PLLA/starch blend films with added a-TOC and resve-ratrol were produced. The Tg and Tm of the PLLA/starch blend filmswere significantly reduced with the addition of antioxidants,while the thermal stability was improved. The enhanced mechan-ical properties can be attributed to not only a compatibilizationeffect based on the chemical linkage between PLLA and starchchains, but also to restriction of the chain mobility by the antiox-idants. The release of resveratrol from PLLA and PLLA/starch blendfilms into ethanol followed Fickian behavior. The D values ofa-TOC were in the range of 0.47–3.95 � 10�11 cm2 s�1 for PLLAfilms and 0.70–6.83 � 10�11 cm2 s�1 for PLLA/starch blend filmsat 13 �C, 5.67–13.0 � 10�11 cm2 s�1 for PLLA films and 4.10–24.2 � 10�11 cm2 s�1 for PLLA/starch blend films at 23 �C,and 89.0–118.0 � 10�11 cm2 s�1 for PLLA films and 123–282 �10�11 cm2 s�1 for PLLA/starch blend films at 43 �C. The D valuesof resveratrol were in the range of 0.073–0.54 � 10�10 cm2 s�1

for PLLA films and 1.42–6.93 � 10�10 cm2 s�1 for PLLA/starchblend films at 13 �C, 0.90–3.44 � 10�10 cm2 s�1 for PLLA filmsand 4.16–22.3 � 10�10 cm2 s�1 for PLLA/starch blend filmsat 23 �C, and 24.8–74.1 � 10�10 cm2 s�1 for PLLA films and40.1–309 � 10�10 cm2 s�1 for PLLA/starch blend films at 43 �C. Re-lease of antioxidants from PLLA and PLLA/starch blends with dualantioxidant systems differed from that of PLLA with a single anti-oxidant. Morphological instability such as cavities can result inhigher D values and the molecular interaction between the anti-oxidants and the polymers can have a critical effect on in the val-ues of D and Ea in PLLA and PLLA/starch blend systems. Hence,PLLA/starch blends could permit more efficient release systemsof antioxidants for applications where the released antioxidantneeds protection or in the case of improvement of release of anti-oxidant activity by synergistic effects.

Acknowledgments

The authors would like to thank the Korea Packaging Center(KOPACK) at the Korea Institute of Industrial Technology (KITECH)for providing funding to conduct this research and financially sup-porting the authors S.W.H.

References

Araújo, M.A., Ferreira, E.C., Cunha, A.M., Mota, M., 2005. Determination of diffusioncoefficients of glycerol and glucose from starch based thermoplasticcompounds on simulated physiological solution. Journal of Materials Science.Materials in Medicine 16 (3), 239–246.

Arora, S., Bagoria, R., Kumar, M., 2010. Effect of alpha-tocopherol (vitamin E) on thethermal degradation behavior of edible oils. Journal of Thermal Analysis andCalorimetry 102 (1), 375–381.

Auras, R., Harte, B., Selke, S., 2004. An overview of polylactides as packagingmaterials. Macromolecular Bioscience 4 (9), 835–864.

Brayden, D.J., 2003. Controlled release technologies for drug delivery. DrugDiscovery Today 8 (21), 976–978.

Byun, Y., Kim, Y.T., Whiteside, S., 2010. Characterization of an antioxidant polylacticacid (PLA) film prepared with [alpha]-tocopherol, BHT and polyethylene glycolusing film cast extruder. Journal of Food Engineering 100 (2), 239–244.

ChromaDex, 2009. Cerificate of Analysis for Resveratrol. ChromaDex Inc., Santa Ana,CA, USA.

Crank, J., 1975. The Mathematics of Diffusion. Oxford Science Publication, Oxford,UK.

Dhoot, G., Auras, R., Rubino, M., Dolan, K., Soto-Valdez, H., 2009. Determination ofeugenol diffusion through LLDPE using FTIR-ATR flow cell and HPLC techniques.Polymer 50 (6), 1470–1482.

Dorgan, J.R., 2010. Rheology of poly(lactic acid). In: Auras, R., Lim, L.-T., Selke, S.E.M.,Tsuji, H. (Eds.), Poly(lactic acid): Synthesis, Structures, Properties, Processing,and Applications. John Wiley & Sons, New York, NY.

Esterbauer, H., Dieber-Rotheneder, M., Striegl, G., Waeg, G., 1991. Role of vitamin Ein preventing the oxidation of low-density lipoprotein. The American journal ofclinical nutrition 53 (1 Suppl), 314S–321S.

Fischer, E.W., Sterzel, H.J., Wegner, G., 1973. Investigation of the structure ofsolution grown crystals of lactide copolymers by means of chemical reactions.Colloid & Polymer Science 251 (11), 980–990.

Gambuti, A., Strollo, D., Ugliano, M., Lecce, L., Moio, L., 2004. Trans-resveratrol,quercetin, (+)-catechin, and (-)-epicatechin content in South Italianmonovarietal wines: relationship with maceration time and marc pressingduring winemaking. Journal of Agriculture and Food Chemistry 52 (18), 5747–5751.

Granda-Restrepo, D.M., Soto-Valdez, H., Peralta, E., Troncoso-Rojas, R., Vallejo-Córdoba, B., Gámez-Meza, N., Graciano-Verdugo, A.Z., 2009. Migration of[alpha]-tocopherol from an active multilayer film into whole milk powder.Food Research International 42 (10), 1396–1402.

Heirlings, L., Siro, I., Devlieghere, F., Van Bavel, E., Cool, P., De Meulenaer, B., Vansant,E.F., Debevere, J., 2004. Influence of polymer matrix and adsorption onto silicamaterials on the migration of a-tocopherol into 95% ethanol from activepackaging. Food Additives and Contaminants 21 (11), 1125–1136.

Hwang, S.W., Lee, S.B., Lee, C.K., Lee, J.Y., Shim, J.K., Selke, S.E.M., Soto-Valdez, H.,Matuana, L., Rubino, M., Auras, R., 2012a. Grafting of maleic anhydride onpoly(L-lactic acid). Effects on physical and mechanical properties. PolymerTesting 31 (2), 333–344.

Hwang, S.W., Shim, J.K., Selke, S., Soto-Valdez, H., Rubino, M., Auras, R., 2012b. Effectof maleic anhydride grafting reaction on the physical and mechanical propertiesof poly(L-lactic acid)/starch blends. Macromolecular Materials and Engineering297. http://dx.doi.org/10.1002/mame.201200111.

Hwang, S.W., Shim, J.K., Selke, S.E.M., Soto-Valdez, H., Matuana, L., Rubino, M.,Auras, R., 2012c. Poly(L-lactic acid) with added a-tocopherol and resveratrol:optical, physical, thermal and mechanical properties. Polymer International 61(3), 418–425.

Jang, W.Y., Shin, B.Y., Lee, T.J., Narayan, R., 2007. Thermal properties andmorphology of biodegradable PLA/starch compatibilized blends. Journal ofIndustrial and Engineering Chemistry 13 (3), 457–464.

King, R.E., Bomser, J.A., Min, D.B., 2006. Bioreactivity of resveratrol. ComprehensiveReviews in Food Science and Food Safety 5, 65–70.

Lim, L.T., Auras, R., Rubino, M., 2008. Processing technologies for poly(lactic acid).Progress in Polymer Science 33 (8), 820–852.

Manzanarez-López, F., Soto-Valdez, H., Auras, R., Peralta, E., 2011. Release of[alpha]-tocopherol from poly(lactic acid) films, and its effect on the oxidativestability of soybean oil. Journal of Food Engineering 104 (4), 508–517.

Markets and Markets, Global Lactic Acid & Poly Lactic Acid (PLA) Market is Growingat a CAGR of 18.7% and is Expected to Reach $3831.3 Million by 2016, 2011.Markets and Markets, Dallas.

Mascheroni, E., Guillard, V., Nalin, F., Mora, L., Piergiovanni, L., 2010. Diffusivity ofpropolis compounds in polylactic acid polymer for the development of anti-microbial packaging films. Journal of Food Engineering 98 (3), 294–301.

Mattivi, F., Reniero, F., Korhammer, S., 1995. Isolation, characterization, andevolution in red wine vinification of resveratrol monomers. Journal ofAgricultural and Food Chemistry 43 (7), 1820–1823.

Meng, X., Maliakal, P., Lu, H., Lee, M.-J., Yang, C.S., 2004. Urinary and plasma levels ofresveratrol and quercetin in humans, mice, and rats after ingestion of purecompounds and grape juice. Journal of Agricultural and Food Chemistry 52 (4),935–942.

Mielnik, M.B., Olsen, E., Vogt, G., Adeline, D., Skrede, G., 2006. Grape seed extract asantioxidant in cooked, cold stored turkey meat. LWT – Food Science andTechnology 39 (3), 191–198.

Page 15: Migration of α-tocopherol and resveratrol from poly(L-lactic acid)/starch blends films into ethanol

828 S.W. Hwang et al. / Journal of Food Engineering 116 (2013) 814–828

Miltz, J., Hoojjat, P., Han, J.K., Giacin, J.R., Harte, B.R., Gray, I.J., 1988. Loss ofantioxidants from high-density polyethylene, food and packaging interactions.American Chemical Society 23 (1), 83–93.

Min, D.B., Boff, J.M., 2002. Chemistry and reaction of singlet oxygen in foods.Comprehensive Reviews in Food Science and Food Safety 1 (2), 58–72.

Ortiz-Vazquez, H., Shin, J., Soto-Valdez, H., Auras, R., 2011. Release of butylatedhydroxytoluene (BHT) from Poly(lactic acid) films. Polymer Testing 30 (5), 463–471.

Soares, D.G., Andreazza, A.C., Salvador, M., 2003. Sequestering ability of butylatedhydroxytoluene, propyl gallate, resveratrol, and vitamins C and E against ABTS,DPPH, and hydroxyl free radicals in chemical and biological systems. Journal ofAgricultural and Food Chemistry 51 (4), 1077–1080.

Sobolev, V.S., Cole, R.J., 1999. Trans-resveratrol content in commercial peanuts andpeanut products. Journal of Agricultural and Food Chemistry 47 (4), 1435–1439.

Soto-Valdez, H., Auras, R., Peralta, E., 2011. Fabrication of poly(lactic acid) films withresveratrol and the diffusion of resveratrol into ethanol. Journal of AppliedPolymer Science 121 (2), 970–978.

Tappel, A.L., 1962. Vitamin E as the biological lipid antioxidant. In: Robert, S., Harris,I.G.W.G.F.M., Kenneth, V.T. (Eds.), Vitamins & Hormones. Academic Press, pp.493–510.

Wessling, C., Nielsen, T., Leufvén, A., Jägerstad, M., 1998. Mobility of a-tocopheroland BHT in LDPE in contact with fatty food simulants. Food Additives andContaminants 15 (6), 709–715.

Wessling, C., Nielsen, T., Leufvén, A., 2000. The influence of a-tocopherolconcentration on the stability of linoleic acid and the properties oflow-density polyethylene. Packaging Technology and Science 13 (1),19–28.