Food Chemistry 148 (2014) 138–146

Contents lists available at ScienceDirect

Food Chemistry

journal homepage: www.elsevier .com/locate / foodchem

Interactions of flavoured oil in-water emulsions with polylactide

0308-8146/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.foodchem.2013.10.018

⇑ Corresponding author. Tel.: +33 169935047.E-mail address: violette.ducruet@agroparistech.fr (V. Ducruet).

Rómulo Salazar a,b, Sandra Domenek b, Violette Ducruet a,⇑a INRA, UMR1145 Ingénierie Procédés Aliments, 1 Avenue des Olympiades, 91300 Massy, Franceb AgroParisTech, UMR1145 Ingénierie Procédés Aliments, 91300 Massy, France

a r t i c l e i n f o

Article history:Received 13 June 2013Received in revised form 29 September 2013Accepted 3 October 2013Available online 12 October 2013

Keywords:Poly(lactic acid)PLAPackagingAroma compoundSorptionPartition coefficientSolubility coefficient

a b s t r a c t

Polylactide (PLA), a biobased polymer, might prove suitable as eco-friendly packaging, if it provesefficient at maintaining food quality. To assess interactions between PLA and food, an oïl in-water modelemulsion was formulated containing aroma compounds representing different chemical structure classes(ethyl esters, 2-nonanone, benzaldehyde) at a concentration typically found in foodstuff (100 ppm). Tostudy non-equilibrium effects during food shelf life, the emulsions were stored in a PLA pack (tray andlid). To assess equilibrium effects, PLA was conditioned in vapour contact with the aroma compoundsat concentrations comparable to headspace conditions of real foods. PLA/emulsion interactions showedminor oil and aroma compound sorption in the packaging. Among tested aroma compounds, benzalde-hyde and ethyl acetate were most sorbed and preferentially into the lid through the emulsion headspace.Equilibrium effects showed synergy of ethyl acetate and benzaldehyde, favouring sorption of additionalaroma compounds in PLA. This should be anticipated during the formulation of food products.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Food packaging is one of the main agents in preservation of foodsafety and quality during shelf life. This justifies the large researcheffort undertaken by industry and academia in proposing novelstrategies, such as modified atmosphere packaging, activepackaging, nanocomposites, etc. Eco-design of system packaging/food is now also a driver for innovation and within this frameworknovel polymers based on renewable resources have been proposed.Poly(lactide) (PLA) forms part of this new generation of polymers,being a possible solution for reducing the ecological footprint(Hermann, Blok, & Patel, 2010).

During storage (up to several years as a function of the consid-ered food) a large number of molecules are exchanged betweenfood and packaging. Additives, polymer stabiliser, volatile com-pounds from degradation products of additives or polymers maymigrate into food and conversely constituents of food could besorbed by the packaging (Gillet, Vitrac, & Desobry, 2010). Migrationand sorption may impact on the sensorial quality of the foodstuff.In particular, sorption of volatile compounds in packaging poly-mers is of interest as it might cause imbalance of flavour profiles(Dury-Brun, Chalier, Desobry, & Voilley, 2007). This subject ismainly concerned with polyolefins and was recently reviewed(Caner, 2011) but little literature on this topic exists on newbiobased polymers, and most published literature concerns

polysaccharides and proteins (Kurek, Descours, Galic, Voilley, &Debeaufort, 2012; Pau Balaguer, Gavara, & Hernández-Muñoz,2012).

Polyolefins are hydrophobic materials and rubbery at roomtemperature, both properties being a priori favourable for sorptionof hydrophobic aroma compounds. The biobased polymer films areon the contrary much more hydrophilic. For example, films basedon gliadins or chitosan presented low sorption capacities for aromacompounds (Pau Balaguer et al., 2012). PLA is a synthetic polymerbut nevertheless more polar than polyolefins. Furthermore it isglassy at room temperature, where room temperature correspondsin most cases to its service temperature. Much less knowledge ex-ists on aroma sorption in glassy polymers. Literature data on ethylvinyl alcohol (Gavara, Catalá, Aucejo, Cabedo, & Hernandez, 1996),poly(ethylene terephthalate) (PET) and poly(vinyl chloride)(Ducruet et al., 2007) showed that interactions of foodstuff andglassy polymers are low. Main explanations are the higher polarityof the polymers reducing partitioning of hydrophobic compoundsand lower diffusion coefficients in glassy matrices. As an example,the comparatively low sorption of hydrophobic compounds such aslimonene in PLA has already been shown (Auras, Harte, & Selke,2006). On the contrary, high sorption of more hydrophilicmolecules sharing structure similarities with PLA, such as ethylacetate was observed (Auras et al., 2006; Colomines, Ducruet,Courgneau, Guinault, & Domenek, 2010; Courgneau et al., 2012).The investigation of the aroma sorption in glassy materials holdsmethodological difficulty, because it needs the analysis of traceamounts of different molecules in the polymer. This challenges

R. Salazar et al. / Food Chemistry 148 (2014) 138–146 139

not only analytical chemistry but also extraction methods. Due tothese methodological constraints, in most published work onlysingle aroma compounds were tested, mostly at high concentrationor high partial pressure. These conditions are however far fromservice conditions of the considered packaging polymers. In realfoods aroma compounds are usually present in a mixture and atlow concentration, being typically in the range of 100 ppm in thefoodstuff and of 1–10 ppm in the headspace. Foodstuff is moreovera complex system including compounds interacting with aromacompounds, the most important of which are lipids and proteins(Caner, 2011). Furthermore polysaccharides, such as pectin orcarboxymethylcellulose, change sorption behaviour by increasingviscosity of foodstuff and slowing down diffusion of flavour com-pounds from the food matrix to packaging (Dury-Brun et al., 2007).

The objective of this work is the investigation of aroma com-pound sorption in a glassy, biobased polymer, PLA. In order tostudy the sorption phenomena, a two-step approach was under-taken. A model food emulsion was formulated and flavoured witha mixture of molecules being representative of different chemicalstructure classes at a concentration comparable to their occurrencein foodstuff. The emulsions were then packed in PLA trays, liddedwith a PLA/aluminium foil. This first study gives insight into non-equilibrium interaction effects, which occur during the lifetime ofthe food. In parallel, to gain understanding of equilibrium interac-tion effects, PLA samples were conditioned in the headspace of thesame aroma compound mixture and the sorption and mixtureeffects were studied. To be able to provide mechanistic explana-tions, changes in the polymer morphology were analysed.

2. Materials and methods

2.1. Materials

2.1.1. ChemicalsTween 80 (Merck, Darmstadt, Germany) and xanthan gum

(Keltrol�, Monsanto, Cambridge, UK) were used in this study.Potassium sorbate, triheptadecanoin standard, sodium sulphateanhydrous, potassium hydroxide, b-carotene (>95%), borontrifluoride methanol complex, pentane (98%), hexadecane (99%)were purchased from Sigma–Aldrich (St Louis, MO). Methanoland n-heptane (99%) were purchased from Carlo Erba Reagents(Val de Reuil, France). Rapeseed oil of food grade (Lesieur,Asnieres-sur-Seine, France) was used throughout this work.

Ethyl acetate (EA), ethyl butanoate (EB), ethyl 2-methylbutano-ate (E2MB), ethyl hexanoate (EH) and 2-nonanone (2NNE) werepurchased from Sigma–Aldrich and benzaldehyde (BZA) waspurchased from Fluka. Their purities were higher than 99%.

2.1.2. Preparation of flavoured food emulsionsThe emulsions consisted (in weight proportions) in 40% rape-

seed oil with or without 0.01% b-carotene as a colour marker, 2%emulsifier (Tween 80), 0.1% potassium sorbate as antimicrobialstabiliser, 0.3% of xanthan gum (being 0.5% in the aqueous phase)as thickener, and 57.6% distilled water. To prepare the emulsion,xanthan gum was dissolved in warm water (60 �C) inside a her-metic flask by stirring with a magnetic mixer for 15 min. The solu-tion was cooled to room temperature and Tween 80 and potassiumsorbate were added. The solutions were stirred for 15 min by mag-netic stirrer. Rapeseed oil and b-carotene were added subsequentlyunder vigorous stirring using a Polytron� PT3000 homogeniser(Kinematica AG, Lucerne, Switzerland) at 10,000 rpm for 10 min.The flasks were placed in an ice bath to prevent heating duringthe homogenising phase. Finally, the emulsions were placed in

an ice bath and treated by ultrasound (Ultrasonic Processor;Bioblock Scientific, Illkirch, France) for 20 min in pulses of 10 s.

A solution of the six aroma compounds (AC) mentioned abovewas prepared in ethanol (5000 ppm of each AC). Then, a volumeof 2 mL was added to 100 mL of emulsion to give a final concentra-tion of 100 ppm of each AC in the emulsion and stirred by amagnetic mixer for 15 min.

2.1.3. Polylactide films and traysTwo extruded PLA4032D (NatureWorks LLC, Blair, NE) and one

PLA2002D (NatureWorks LLC) with thickness of 235 lm, 55 lmand 65 lm, respectively, were studied. The extrusion conditionsof PLAs such as the content of D-lactic acid present in materialswere described in our previous work (2% for PLA4032D and be-tween 4% and 4.5% for PLA2002D) (Salazar, Domenek, Courgneau,& Ducruet, 2012). PLA4032D trays for conditioning of emulsionswere produced by thermoforming at 180 �C for 24 s (ILLIG Maschi-nenbau GmbH, Heilbronn, Germany) and had a thickness of150 lm. Trays were heat-sealed with a NATIVIA™ NZSS (alumin-ium/PLA) lid of 35 lm thickness, where the PLA face was in contactwith the headspace inside the tray.

In order to obtain very thin films, three samples (30 ± 1 mg eachone) of PLA4032D and PLA2002D with a thickness of 10 lm wereprepared using a microtome (LKB Bromma Histo-Range 2218).

2.2. Methods

2.2.1. Characterisation of emulsionsEmulsion flow curves were measured at 20 ± 1 �C over a shear

rate range of 0.001–10 s�1 using a dynamic rheometer (MCR301;Anton Paar GmbH, Graz, Austria) equipped with cone-plate geom-etry (CP 50/4�). All measurements were performed within 24 hafter emulsion preparation. The oil droplet size was characterisedwith the help of a Mastersizer 2000 (Malvern Instruments, GreatMalvern, UK) equipment. The stability of the different emulsionswas studied by multi-angle light scattering measurement using aTurbiscan MA 2000 (Formulaction, L’Union, France). Light trans-mission and backscattering profiles were used to evaluate thephysical stability of emulsions.

2.2.2. Conditioning of PLA samples in oil and emulsionsFor immersion tests, PLA4032D film (thickness 55 lm) was cut

into rectangular pieces measuring 50 by 20 mm using a paper-cut-ter. Ten strips (1 dm2) were immersed in 100 mL of rapeseed oilwith or without b-carotene in a 250-mL flask with stopper for28 days at 40 �C. The conditions were adapted according to thenorm NF-EN-1186-2 of overall migration into olive oil by totalimmersion. Five flasks were prepared in this way.

For storage tests of emulsions in PLA 100 mL of each type ofemulsion were filled in three PLA trays and heat sealed with aNATIVIA™ NZSS (aluminium/PLA) lid. Trays were stored in an ovenat 40 �C for 30 days.

2.2.3. Quantification of oil sorbed in PLA samplesThe rapeseed oil sorbed in PLA samples was extracted and

determined by gas chromatography after conversion to methylesters according to the norm NF EN 1186-2 (January 2003). Afterconditioning time, PLA strips were removed from the rapeseedoil and rinsed with ethanol in order to remove the rest of the oil.The same procedure was carried out for PLA trays after removalof O/W emulsion. Then strips and trays (1 dm2 each) were wipedusing KimTech precision wipes, cut in small squares and placedin a 25-mL vial. The extraction of oil was carried out by adding10 mL of n-pentane into vials with stoppers under magneticstirring at 50 �C for 30 min. The operation was renewed two timeswith fresh solvent. Five millilitres of a solution of 2000 ppm of

140 R. Salazar et al. / Food Chemistry 148 (2014) 138–146

triheptadecanoin in heptane were added into each vial as internalstandard prior to the first extraction.

The resulting extracts were collected and evaporated to drynessusing a rotary evaporator (Büchi R-114, Flawil, Switzerland) andthen dissolved in 10 mL n-pentane. Fatty acids were formed byhydrolysing the extracts with 10 mL of a solution of potassiumhydroxide (11 g/L) in methanol under reflux. Then methylationwas carried out by reacting the extracts with 5 mL of borontrifluoride/methanol. Finally, the upper phase of the extracts wasinjected into a gas chromatograph (CE Instruments, ModelGC8000Top, GC/FID) equipped with a DB-FFAP capillary column(30 m length � 0.32 mm, 1 lm film thickness, J & W Scientific).The temperature of the split/splitless injector was 250 �C with apurge time of 0.5 min. The flame ionisation detector was set at250 �C with hydrogen as carrier gas (50 kPa). The oven tempera-ture was programmed to increase from 60 to 250 �C at 7 �C/min.Calibration was carried out using different quantities of rapeseedoil (5, 10, 20, 50 and 150 mg) which were subjected to the sameprocedure as the samples. For each quantity of oil, the ratio ofthe combined area of C16 and C18 peaks to area of internal stan-dard was calculated and then ratios were plotted against the oilquantities.

2.2.4. Moisture loss and water vapour transfer rate (WVTR)The moisture loss of the emulsions packaged in PLA trays was

measured in following the weight loss during storage at 40 �C for20 days, which implied the hypothesis that water loss was theprincipal cause of the weight loss of the emulsion. Weight losswas measured using an analytical balance (Sartorius E5500S,10�4 g). The water vapour transfer rate (WVTR) of PLA trays wascalculated from the slope of the plot of weight loss over time.Given values are averages of three experiments.

2.2.5. Conditioning of PLA films in aroma atmosphereThe experimental system used to condition microtomed PLA

samples has already been described in Salazar et al., 2012. Theconditioning was carried out in a hermetic flask and placed in aclimatic chamber for 20 days at 25 ± 0.3 �C. Three microtomedsamples (30 mg each one) of PLA4032D and PLA2002D wereconditioned separately. Specific partial pressure of the ACs in theatmosphere was generated by a solution of ACs in hexadecane.Four AC solutions were tested individually and in mixture. Thecompositions are presented in Table 1.

2.2.6. Quantification of ACs in PLA samples by headspace methodsAfter contact time, PLA samples from trays, lids and

microtomed PLA sample were analysed using multiple headspaceextraction (MHE) or multiple headspace solid-phase micro extrac-tion (MHS-SPME). The two applied procedures were described inSalazar et al. (2012).

The apparent solubility coefficient (S) of each AC was calculatedaccording to:

S ¼ M1

M0� d

Pp; ð2Þ

where M1 is the mass sorption after 20 days of conditioning (kg),M0 is the initial mass of PLA sample (kg), d is the film density(kg/m3) and Pp is the partial pressure of each AC.

The partition coefficient is defined as the mass concentrationrelation of a molecule between two phases. The partitioncoefficients calculated in this study were:

Kpolym=hs ¼Cpolymer

Cheadspace; ð3Þ

Kpolym=liq ¼Cpolymer

Cliquid; ð4Þ

Khs=liq ¼Cheadspace

Cliquid; ð5Þ

where Cpolymer is the concentration of the compound in the polymer(lg/g), Cheadspace is the concentration of the compound in theheadspace (lg/g) and Cliquid is the concentration of the compoundin the liquid or emulsion (lg/g).

2.2.6.1. Quantification of sorbed ACs in trays and lids byMHS-SPME. The ACs sorption in PLA trays was quantified byMHS-SPME. After conditioning time, the trays and lids were re-moved, rinsed with ethanol and wiped. Approximately 80 mg ofPLA tray and 40 mg of lid were sealed separately in 20-ml glassvials with silicon/PTFE septa. Three vials were prepared and eachvial was sampled four times using an Agilent Technologies 6890GC/FID coupled with a 5975 mass spectrometer and a GerstelMPS2 autosampler. The sample vials were incubated at 60 �C for30 min and agitated at 250 rpm. MHS-SPME was performed for15 min at 60 �C using a 75 lm carboxen–polydimethylsiloxane(CAR-PDMS) fibre. Desorption time was 5 min. The oven tempera-ture programme began with an initial temperature of 30 �C for5 min, and then temperature increased at a rate of 5 �C/min upto 230 �C, and held for 5 min. Split/splitless injector and detectortemperatures were 250 �C. The column was a J & W ScientificDB5-MS capillary column (30 m length � 0.32 mm inner diameter;0.5 lm film thickness). The carrier gas was He at 1.4 mL/min.

2.2.6.2. Quantification of ACs in microtomed PLA samples by MHE.The aroma sorption in PLA samples, after conditioning in an aromaatmosphere, was quantified according to the MHE method. In brief,after conditioning time, microtomed PLA samples (30 mg) weresealed in 20-ml headspace glass vials with silicon/PTFE septa andan aluminium cap. Each vial was sampled five times at equal timeintervals (55 min) at 60 �C using an automatic headspace sampler(Perkin Elmer HS40 XL) coupled with GC/FID. The oven tempera-ture was programmed at 35 �C for 2 min, 6 �C/min up to 230 �C.The column was a SupelcoWax-10 capillary column (30 mlength � 0.32 mm, 0.5 lm film thickness). Injector temperaturewas 250 �C and its split was 15.6 mL/min. Detector temperaturewas 250 �C. The carrier gas was He at 1.8 mL/min.

The limit of detection (LOD) was calculated as the quantity ofcompound producing a signal exceeding the average backgroundsignal by three standard deviations of the signal-to-noise ratio inthe lowest concentration standard, divided by the slope of the cor-responding analytical curve. The limit of quantification (LOQ) wasobtained as the quantity of compound that is 10 standard devia-tions above the average background signal, divided by the slopeof the corresponding calibration curve.

2.2.7. Size exclusion chromatography (SEC)The average molecular weight and the dispersity of PLA samples

were measured by SEC using a Gilson 305 apparatus equipped witha refractive index detector (RID-101, Shodex). PLA samples weredissolved in chloroform (5 mg/mL) and analysed at 25 �C and1.0 mL/min in THF stabilised with BHT on a PL Gel Mixed-C column(5 lm, 100 Å, 600 � 7.5 mm). The calibration was performed withPS standards (EasiCal) from 580 to 377400 g/mol.

2.2.8. Thermal properties: differential scanning calorimetry (DSC)Before and after conditioning, the thermal properties of PLA

samples were determined in triplicate using a QSC 100 differentialscanning calorimeter (TA Instruments Ltd, Crawley, UK). PLA sam-ples were analysed following the modulated temperature pro-gramme used by Delpouve, Saiter, and Dargent (2011). Sampleswere heated from 10 to 190 �C, with oscillation amplitude of0.318 �C, an oscillation period of 60 s and a heating rate of 2 �C/

Table 1Data on conditioning of the PLA samples: concentrations of aroma compound in solution and in headspace, partition coefficient headspace/aroma solution (khs/liq), partial pressure(Pp) and activity.

Aroma compound Aroma solution in hexadecane(mol/L) � 10�2

Concentration of aroma in headspace(mol/L)

Khs/liq Pp (Pa) Activity

Ethyl acetate 2.0 1.7 � 10�4 8.50 � 10�3 455.36 0.037Benzaldehyde 3.0 4.5 � 10�6 1.50 � 10�4 12.45 0.104Aroma compound mixture 1 (M1) EB 1.5 1.9 � 10�5 1.27 � 10�3 51.08 0.025

E2MB 2.0 1.3 � 10�5 6.50 � 10�4 33.61 0.031BZA 3.0 4.5 � 10�6 1.50 � 10�4 12.45 0.104EH 1.8 2.0 � 10�6 1.11 � 10�4 6.13 0.0042NNE 1.7 1.4 � 10�6 8.24 � 10�5 4.74 0.006

Aroma compound mixture 2 (M2) EB 1.5 1.9 � 10�5 1.27 � 10�3 51.08 0.025E2MB 2.0 1.3 � 10�5 6.50 � 10�4 33.61 0.031EH 1.8 2.0 � 10�6 1.11 � 10�4 6.13 0.0042NNE 1.7 1.4 � 10�6 8.24 � 10�5 4.74 0.006

K = concentration in headspace (mol/L)/concentration in solution (mol/L) at 25 �C after 20 days of conditioning.

R. Salazar et al. / Food Chemistry 148 (2014) 138–146 141

min. The degree of crystallinity (Xc) of materials was calculatedaccording to:

Xc ¼DHm � DHcc

DHom

; ð6Þ

where DHm is the enthalpy of fusion, DHcc is the enthalpy of coldcrystallisation and DHo

m is the heat of melting crystalline PLA(93 J/g).

2.2.9. Statistical analysisThe statistical analysis of data was performed through one-way

analysis of variance (ANOVA) using XLSTAT-Pro 7.0 software(Addinsoft, Paris, France). When differences were significant(p < 0.05), Duncan’s test was used to evaluate the differences be-tween pairs of groups.

3. Results and discussion

3.1. Characterisation of interactions between PLA and food emulsions

The storage conditions were adapted from the norm NF-EN-1186-2 representing conservation for 30 days at 40 �C. These areunfavourable conditions for PLA, because 40 �C is very near tothe glass transition of the polymer (glass transition temperatureTg = 60 �C). Moreover the stability of the emulsion needed to beguaranteed during the storage. For that, small droplet size wasaimed for and xanthan gum used as thickener. Obtained dropletsize was 1.2 ± 0.2 lm for all the three types of emulsions: (i) blank,(ii) containing b-carotene, and (iii) containing b-carotene and ACs.All emulsions presented shear thinning behaviour with a smallyield stress (0.1 Pa), due to depletion flocculation. No changes inbackscattering and clearing zones due to creaming were evidencedfor all emulsion formulations. PLA trays after storage were notstained by the colour marker b-carotene. This is an advantage overpolyolefins, where coloration occurs (Nielsen & Olafsson, 1995). Inorder to assess moisture loss of emulsions under storage, the trayswere weighed for 30 days and the WVTR was calculated. We foundin PLA trays (thickness 150 lm) a WVTR of 73.73 ± 0.92 g/m2 daywhich is in accordance with the WVTR value (30.6 g/m2 day) re-ported by Holm, Mortensen, and Risbo (2006), who studied theWVTR in PLA trays of 500 lm thickness at 38 �C and 90% RH. Ourresult is furthermore in accordance with Courgneau, Domenek,Guinault, Averous, and Ducruet (2011) who reported a WVTR of61.9 ± 6.7 g/m2 day at 38 �C and 90% RH in PLA films.

In order to check the oil stability of PLA, the quantities of rape-seed oil sorbed in PLA4032D film after immersion in rapeseed oiland in PLA 4032D tray samples after storage of the emulsions weremeasured (Table 2). To the best of our knowledge, there are no

studies dealing with oil sorption in PLA and only few results existon interactions between oil and conventional polymer packagingmaterials. The amount of oil sorbed by polymer films after 40 daysof contact at room temperature was about 27.5 mg/dm2 for HDPE,14.5 mg/dm2 for APET and 5 mg/dm2 for PP (Johansson & Leufvén,1994). Arora and Halek (1994) reported that sorption of triglycer-ides was higher in PP than in PET after 10 days at 23 �C and de-pended on the nature of the triglyceride. In PP, sorption was4.4 mg/dm2 for tripalmitin, 2.93 mg/dm2 for triolein and 1.3 mg/dm2 for trilinolein, while in PET it was 1.2 mg/dm2, 0.6 mg/dm2

and 0.3 mg/dm2, respectively. Using the same protocol as ours,Kassouf (personal communication, 2012) measured 33 ± 2 mg/dm2 of olive oil sorbed by PET samples after 10 days of immersionat 40 �C. In conclusion, a large discrepancy exists among the quan-titative results depending on the experimental procedure but theorder of the oil sorption propensity of polymers seems to beexplainable by simple physicochemical considerations. Glassyand more polar polymers sorb generally less oil than hydrophobicand rubbery polyolefins. When we compared our results with theresults of Kassouf who applied the same experimental approach,PLA showed an oil sorption lower than PET.

From Table 2 on the oil sorption of PLA from the emulsions, itappeared that the amount of oil sorbed in PLA4032D increasedwith the oil content of the contact medium. It was approximatelytwo to three times higher in pure oil compared to the emulsionswith 40% oil content. Sorption of lipids contained in foodstuff intopolyolefins has been measured (vom Bruck, Figge, & Rudolph,1981). After 5 days of contact, fat sorptions were 1760 lg/g,960 lg/g, 255 lg/g and 20 lg/g, respectively, when LDPE sampleswere put in contact with food such as olive oil, mayonnaise, coffeecream and yogurt, having a fat content of 100%, 80%, 10% and 3.7%respectively. They concluded that fat sorption in LDPE was gov-erned by the global fat content of foodstuffs, more than by the tex-ture. This validates the present result, which shows that the oilsorption of PLA was proportional to the amount of oil in the condi-tioned food.

Interactions of foodstuff with the packaging polymer are depen-dent on the microstructure of the polymer and can even inducechanges in its microstructure. Table 2 showed the evolution ofthe PLA microstructure and the changes in the molar weight aver-age were measured before and after storage of the food emulsion.PLA is a polyester, therefore one of the main degradation routes ofthe polymer chain is hydrolysis. Storage of a water-containing foodat temperature close to the glass transition might therefore inducepolymer degradation. However, results in Table 2 show only smalldecrease in molecular weight average. PLA is in conclusion stableunder the tested conditions. Glass transition temperature (Tg)and crystallinity degree (Xc) of PLA films before and after oil con-tact showed no major differences, even though oil was sorbed in

Table 2PLA/food emulsion interactions: oil sorbed, PLA glass transition temperature (Tg), degree of crystallinity (Xc), number and weight average molecular mass (Mn and Mw) anddispersity (I) of PLA4032D samples before and after oil sorption.

Packaging Medium in contact Oil sorbed (mg of oil/dm2 of material) Tg (�C) Xc Mn (g/mol) Mw (g/mol) I

PLA film Before contact 56.8 ± 0.9a 12 ± 1a 75,200 160,900 2.1After contact Rapeseed oil 10 ± 1a 56.0 ± 0.5a 11 ± 1a nd nd

Rapeseed oil + b carotene 15 ± 1b 55.2 ± 0.1b 10 ± 1a nd ndPLA tray Before contact 55.8 ± 0.9a 10 ± 2a 91,400 173,200 1.9

After contact EmulsionTween80 <5 61.3 ± 0.4c 36 ± 1c 83,800 149,200 1.8Emulsion Tween80 + b-carotene <5 59.6 ± 1.1b 33 ± 4b 74,200 140,800 1.9Flavoured emulsion Tween 80 <5 55.8 ± 0.9a 28 ± 2b 72,700 138,200 1.9

a,b,c Different letters indicate significant differences at p < 0.05 (Duncan). Statistical analysis was carried out separately for each PLA type.

142 R. Salazar et al. / Food Chemistry 148 (2014) 138–146

the structure. Apparently, oil had no plasticising or crystallising ef-fect on the polymer. In the case of emulsions, Xc and Tg increased,which shows interaction most probably with water. This interac-tion seemed to induce crystallisation of the polymer. The increasein Tg might be related to differences in Xc and physical aging, asboth phenomena generally shift the Tg of polymers to higher tem-perature (Delpouve et al., 2011). The effects of time, temperatureand humidity on physical properties of PLA were studied by Auras,Harte, and Selke (2004). They showed that humidity had a signifi-cant effect on Tg of PLA (98% L-lactide) after 7 days of contact at40 �C and 0.9 activity, but after 30 days of contact, no significant ef-fect was found.

High water vapour permeability and high oxygen permeabilityare known drawbacks of PLA in food packaging applications. PLAis roughly one order of magnitude more permeable for water thanPS and PET and two orders than HDPE (0.024 � 10�14 kg m/m2 s Pa,at 38 �C) (Domenek, Courgneau, & Ducruet, 2011). NeverthelessPLA shows only low oil sorption. PLA could therefore be suitablefor the storage of food products such as salad dressings, oils andshortening, peanut butter, which are ‘‘oxygen tolerant’’ (Domeneket al., 2011) or which need moderate to high water vapour trans-mission rates, like fruits, vegetables or cheeses, in order to preventwater condensation inside the pack. Moreover PLA might be analternative candidate to PET in the conditioning of non-aqueousfoods.

3.2. Non-equilibrium effects of AC/PLA interactions during theemulsion shelf life

PLA is generally claimed to provide a good aroma barrier andsome literature exists on this (Auras et al., 2006; Salazar et al.,2012). Indeed, the specific sorption of given ACs with packagingcan modulate significantly the sensorial attributes and thereby de-crease food quality. The sorption depends on the chemical natureof the polymer and the considered AC, but also on the compositionof the foodstuff. The effects of these factors on the aroma sorptioninto packaging materials have already been reported (Dury-Brunet al., 2007) and the relation can be expressed as an apparent par-tition coefficient (Kapp).

In order to investigate non-equilibrium interactions betweenACs belonging to different chemical structure classes and PLA,the emulsion has been flavoured by a model aroma solution. After30 days storage, aroma sorption in PLA trays (thickness = 150 lm)and lids (thickness = 35 lm) was determined by MHS-SPME. PLAtrays were in liquid contact with a flavoured emulsion whereasPLA lids were in contact with the vapour atmosphere generatedby the emulsions. The results (Table 3A) show that the AC uptakeof PLA trays was in any case lower than in PLA lids; the differencebetween uptake of tray and lid was smallest in the case of EA andhighest in the case of 2NNE. The AC uptake decreased with increaseof its molecular weight. Only a few diffusion coefficients of ACs inPLA are available today. Courgneau et al. (2012) gave a diffusioncoefficient of ethyl acetate in the order of 10�17 m2/s. As diffusion

coefficients are negatively correlated to molecular mass (Fang,Domenek, Ducruet, Réfrégiers, & Vitrac, 2013) even lower valuesmight be expected for the other ACs used in this work. Taking intoaccount the thickness of the PLA tray, the sorption equilibriumwith the ACs contained in the emulsion cannot therefore bereached within the timeframe of the experiment. Aroma uptakeseemed to be correlated to molecular mass. To account for non-equilibrium effects during shelf life of food, apparent partitioncoefficients were calculated and values are given in Table 3A. Theyare low, due to the fact of being taken far from thermodynamicequilibrium. Their order shows that in formulations containingethyl esters, more loss of the smallest molecules is to be expected.Partitioning is also influenced by the nature of the food matrix.Proteins, carbohydrates, oil are able to decrease significantly thesorption of ACs in LLDPE (van Willige, Linssen, & Voragen, 2000).In the present case, retention effects of the more hydrophobicACs inside the food matrix due to favourable partitioning in theoil phase seemed only of secondary importance, though.

The quantity of ACs sorbed from the headspace in the lidshowed a different picture (Table 3A). The sorbed quantities werehigher and no longer correlated to the molecular mass, becausesorption was most probably closer to the equilibrium value. Theorder was nevertheless not fully concordant with the order of par-tition coefficients obtained by Salazar et al. (2012) under equilib-rium contitions (data given in Table 3). For comparison of bothvalues normalisation of quantities needs to be carried out by com-puting the apparent sorption coefficient. For that, the partitioningbetween emulsion and headspace must be known. It was, however,not accessible in our experimental setup, therefore literature val-ues (Philippe et al., 2003) available for the series of ethyl estersin an aqueous system were compared in Table 3A. The use of par-tition coefficients of an aqueous system leads most probably tounderestimation of the apparent partition coefficient given inTable 3A. What was interesting in comparison with the data of Sal-azar et al., the order of the partition coefficients corresponds againto the order of molecular weight. A second interesting result wasthe fact that the Kapp(lid/headspace) was 10 to 100 times higherthan the Kapp(tray/emulsion). Scarce literature observations of par-titioning between an AC in liquid contact or in vapour contact anda polymer show qualitatively similar data (Johansson & Leufven,1997). These authors studied the sorption of ethyl butanoate inpolypropylene from vapour and from 40% ethanol solutions after14 days of storage in screw-capped glass bottles at 23 �C in thedark. They observed that the polymer/headspace partition coeffi-cient of ethyl butanoate and polypropylene amounted to 562 whenthe headspace concentration of this compound was around 2.8 lg/cm3, while the polymer/liquid partition coefficient was only 2.4 inthe case of a solution of 5 lg/cm3 of ethyl butanoate in ethanol/water. So, in vapour phase the partitioning is related mainly tomolecular mass of compound, whereas in liquid phase the parti-tioning depends of the molecular mass and also on the chemicalproperties of compounds, as the affinity compound/polymer andthe logP.

Table 3(A) Non equilibrium effects of PLA/aroma compounds interaction: sorption of aroma compounds in trays and lids determined by MHS-SPME, apparent partition coefficients (Kapp),and partition coefficients from literature; (B) Equilibrium effects of PLA/aroma compounds interactions: sorption (Cp) of aroma compounds in PLA samples (thickness = 10 lm)determined by MHE and partition coefficient (Kpolym/hs) polymer/headspace.

Aroma compound Literature data PLA4032D tray Aluminium/PLA lid

K(PLA/hs)a K(water/hs)b Cp (lg/g) Cliq (lg/g) Kapptray=liqC(lid) (lg/g) Chs (lg/g)c Kapplid=hs

Non equilibrium effects (A)Ethyl acetate EA 197 10 1.682 ± 0.082 118.3 0.0142 6.638 ± 1.056 1.916 3.46Benzaldehyde BZA 18,750 – 0.142 ± 0.023 118.3 0.0012 8.374 ± 0.610 – –Ethyl butanoate EB 157 25 0.252 ± 0.009 118.3 0.0021 2.476 ± 0.224 2.165 1.11Ethyl 2-methylbutanoate E2MB 116 – 0.196 ± 0.023 118.3 0.0017 3.33 ± 0.427 – –Ethyl hexanoate EH 414 41 0.073 ± 0.009 118.3 0.0006 1.914 ± 0.203 3.572 0.542-nonanone 2NNE 447 – 0.015 ± 0.001 118.3 0.0001 0.265 ± 0.054 – –

Equilibrium effects (B)Aroma compound LOD (lg) LOQ (lg) PLA4032D sample PLA2002D sample

Cp (lg/g) Chs (lg/g) Kpolym/hs Cp (lg/g) Chs (lg/g) Kpolym/hs

Ethyl acetate EA 0.01 0.069 3263.27 ± 189.63 15.23 214 3125.46 ± 184.99 15.23 205Benzaldehyde BZA 0.028 0.22 3133.82 ± 97.59 0.48 6156 4913.15 ± 181.01 0.48 10,299Aroma compound mixture 1 EB 0.007 0.051 245.83 ± 44.45 2.23 110 372.36 ± 79.42 2.23 167

E2MB 0.009 0.072 170.29 ± 22.78 1.73 98 315.93 ± 78.11 1.73 183BZA 0.028 0.22 6368.0 ± 273.12 0.48 8829 7457.75 ± 249.74 0.48 13,605EH 0.011 0.085 54.67 ± 11.17 0.29 187 84.87 ± 9.05 0.29 2912NNE 0.038 0.328 <LOD 0.19 <LOD 0.19

Aroma compound mixture 2 EB 0.007 0.051 235.80 ± 27.67 2.23 106 290.38 ± 44.59 2.23 130E2MB 0.009 0.072 122.95 ± 26.88 1.73 71 234.41 ± 40.51 1.73 135EH 0.011 0.085 <LOD 0.29 <LOD 0.292NNE 0.038 0.328 <LOD 0.19 <LOD 0.19

a From a mixture of 6 aroma compounds at low concentration in hexadecane at 25 �C after 20 days for PLA2002D, calculated from Salazar et al. (2012).b Estimated from a aqueous solution of aroma compounds at 50 ppm after 48 h at 37 �C, in Philippe et al. (2003).c Estimated from the partition coefficients headspace/aroma solution in water at 30 �C.

R. Salazar et al. / Food Chemistry 148 (2014) 138–146 143

The practical implication of those data was that only 0.072% ofEA, 0.015% of EB, 0.017% of E2MB, 0.027 % of BZA, 0.008% of EH and0.001% of 2NNE present in the emulsion were sorbed by the wholePLA packaging (tray and lid). This shows the good aroma barrierproperties of PLA in the present case. The higher partition coeffi-cient and faster sorption kinetics of the PLA lid showed moreoverthe preponderant importance of transfers passing by the head-space into the polymer. The technique of lidding and the propertiesof the lid are therefore of primary importance in the design of thepackaging container.

3.3. Equilibrium effects of AC/PLA interactions

The limit of our investigation with the model emulsion was thatthe thermodynamic sorption equilibrium was not attained be-tween the headspace and the lid and that the partitioning betweenemulsion and headspace was not known, despite the high practicalimportance of transfers in the gaseous phase.

In order to attain thermodynamic equilibrium between head-space and PLA samples, a model system was used. The characteris-tic film thickness of PLA was reduced to 10 lm to allow thesorption equilibrium to be reached in 20 days based on the hypoth-esis of a diffusion coefficient of the order of 10�17 m2/s (Auras et al.,2006; Courgneau et al., 2012). The influence of the composition ofthe headspace on the sorption of compounds in PLA and the effectof sorption on thermal properties were investigated. This enabledconclusions to be drawn not only on single molecules, but alsoon the synergistic or antagonistic effects between molecules.

The measured data of sorption were normalised by the calcula-tion of the solubility coefficient (S) after Eq. (2) and the results aregiven in Fig. 1A and B. The sorption profiles were qualitatively sim-ilar between both commercial PLA grades tested, althoughPLA2002D showed a tendency to higher sorption. This differencewas however small and not significant in all cases.

As expected according to literature (Auras et al., 2006;Colomines et al., 2010; Salazar et al., 2012), the sorption of ethylacetate and benzaldehyde in PLA was high. In the case of ethyl

acetate, this affinity can be sought in its quite similar molecularstructure to PLA, being the most hydrophilic of a homologousseries of ethyl esters. The benzaldehyde case may be explainedby the miscibility of PLA with aromatic solvents such as tolueneand benzene (Jin, Tian, & Wang, 2010).

The partition coefficients were calculated and are given inTable 3B. They are several orders of magnitude higher than theapparent partition coefficient of the headspace of the model foodemulsion. This shows that Kapp in Table 3 was clearly underesti-mated and that the system headspace of emulsion/lid was stillfar from thermodynamic equilibrium.

The synergy effect between ACs was tested with the help of dif-ferent AC mixtures and in comparison to values reported by Sala-zar et al. (2012), using a mixture of EA, EB, E2MB, BZA, EH, and2NNE. In the previous work, the sorption of 2NNE in PLA wasproved, while in the present case it was lower than the MHE detec-tion limit (Table 3B). The presence of EA in the mixture of the pre-vious work had thus most probably had a synergistic effect infacilitating the sorption of other ACs. Furthermore, S values of EHhere were smaller than previous ones (1.66 ± 0.18� 10�2 kg m�3 Pa�1

in this work for PLA2002D and 2.55 ± 0.45 � 10�2 kg m�3 Pa�1 inthe previous one). The comparison of sorption from aromasolutions M1 and M2 gives information on mixing effects ofbenzaldehyde, because M1 contains and M2 does not containbenzaldehyde. In the case of M2 only EB and E2MB can be detectedafter sorption in PLA by MHE. This shows that benzaldehyde hadthe same synergistic effect as ethyl acetate on the sorption of theother ACs. This synergy was of importance primarily for the ACspresent in the headspace at very low activity.

Sorption of volatile organic compounds follows Flory–Hugginslaw, meaning that the solubility coefficient increases strongly withthe activity, while in some case the sorption isotherm of volatileorganic compounds in the low activity (concentration) region canbe approximated by Henry’s law, though. Synergy of sorption be-tween ACs is expected in the higher concentration range (ap-prox. > 100 ppm) where sorption causes swelling of the polymerwhile antagonistic effects can be expected in the low concentration

Fig. 1. Solubility coefficients (S) of aroma compounds in PLA samples (thickness = 10 lm): PLA4032D (A), PLA2002D (B). The ordinate is discontinuous as indicated to presentall small values, n = 3. Partial pressures used are given in Table 1; DSC results and total mass uptake for PLA4032D (C) and PLA2002D (D) before conditioning and afterconditioning with different aroma solutions. Tg (j).Xc ( ). M1Aroma compound mixture 1, M2Aroma compound mixture 2. a,b,c Different letters indicate significant differencesat p < 0.05 (Duncan). Statistical analysis was carried out separately for each PLA type and the aroma compounds.

144 R. Salazar et al. / Food Chemistry 148 (2014) 138–146

range where polymer morphology remains unchanged. In the lat-ter case the permeation is predicted to follow the laws of perma-nent gases, which are based on competitive sorption anddiffusion (Story & Koros, 1989). Table 4 presents a literature reviewover different studies carried out on the sorption of molecules inpolymers from AC mixtures. The method most often used was per-

meation, which only allows for an indirect determination of solu-bility coefficients and the most prominent polymer LDPE. Theoverview shows that the hypothesis of antagonistic effects in thelow concentration range and synergy in the high concentrationrange holds for many cases, but there are exceptions. Exceptionsare found in particular, when limonene, a very strongly interacting

Table 4Literature review of synergy and antagonistic effects of aroma compounds mixtures on sorption.

Reference Polymer Molecules Methoda Concentrationrange

Effect

Theodorou and Paik(1992)

LDPE LinaloolCitralEthyl butanoateLimoneneOctanal

Permeation Low (1 ppm) Antagonist

Nielsen, Jagerstad, Oste,and Wesslen (1992)

LDPE Ethyl butanoateHexyl acetateHexanalIsopentanol

Permeation Low (1 ppm) Antagonist for ethyl acetate and hexyl acetate, noeffect for hexanal and isopentanol

Johansson and Leufven(1997)

LDPE Ethyl butanoateHexyl acetate

Sorption inliquid contact

Low (50 ppm) Antagonist

DeLassus et al. (1988) LDPE trans-2-hexenalHexanalEthyl-2-methylbutanoate

Permeation High(144 ppm)

Synergy

Johansson and Leufven(1997)

PP Hexyl acetate in mixture with ethylbutanoate

Sorption inliquid contact

High(1000 ppm)

Synergy for hexyl acetate. No effect for ethylbutanoate

Dury-Brun et al. (2007) HDPE Limonene (21 ppm) in mixture withethyl butanoate (100 ppm)

Permeation Low Synergy

Dury-Brun et al. (2007) BiOPP Ethyl acetate in mixture withlimonene

Permeation High(activity P 0.3)

Antagonist for ethyl acetate. Synergy for limonene

a Permeation refers to studies where solubility was calculated from the permeation experiment with the time lag method. Sorption refers to studies where sorption wasmeasured directly by mass uptake.

R. Salazar et al. / Food Chemistry 148 (2014) 138–146 145

molecule with LDPE is present. This is comparable to our case,where synergy is observed when the very strongly interactingmolecule benzaldehyde is part of the mixture.

One of the main drawbacks of many studies dealing withsorption phenomena in packaging materials is that the thermalproperties of the polymer were not measured. Therefore possiblechanges in the polymer due to plasticisation or induced crystallisa-tion were not detected, although they provide insight into thesorption mechanism. Solvent-induced crystallisation is a knownphenomenon in particular in glassy polymers and Colomineset al. (2010) and Courgneau et al. (2012) showed plasticising andinduced crystallisation due the sorption of ethyl acetate into PLAat high activity, 0.5 and 0.2 respectively. Salazar et al. (2012) re-ported that at very low activities of six ACs in a mixture, such as3.45 � 10�2 for ethyl acetate, a slight decrease in Tg was observedin PLA2002D and PLA4032D. These authors reported a small plas-ticisation of PLAs film samples by ACs at low concentration(<200 lg/g) but they did not observe any solvent-induced crystal-lisation of the samples. The effect of AC sorption on the glass tran-sition temperature and the crystallinity degree of both PLA gradesis plotted in Fig. 1C and D. It is shown that PLA4032D was morecrystalline than PLA2002D, which might explain the small differ-ence in sorption of ACs observed in Fig. 1A and B. Crystallites areexcluded volume where sorption cannot take place, therefore high-er crystallinity leads to lower sorption, which was verified in Fig. 1.Moreover, solvent-induced crystallisation was observed for bothPLA types when benzaldehyde was present, but the effect wasstronger for the initially less crystalline material PLA2002D.

Glass transition of PLA4032D was almost unaffected by thesorption of ACs, whereas PLA2002D showed a slight decrease.The temperature of the glass transition seemed to be correlatedto the total quantity of sorbed AC, and it was thus smaller for M1compared to M2 and benzaldehyde alone. The synergy of AC sorp-tion in our case might therefore be explained by small but measur-able changes in the PLA morphology when benzaldehyde waspresent.

4. Conclusion

PLA showed good compatibility with oil and the flavoured oïlin-water emulsion. Oil sorption was lower than oil sorption in

PET. The colour marker contained in the food emulsions did notstain the polymer. The apparent partition coefficients of the differ-ent ACs were evaluated and showed that sorption passed preferen-tially by the headspace of the food emulsion. In order to investigatethe sorption behaviour of the ACs and the effects of mixtures, amodel study of gaseous contact was carried out in parallel, ensur-ing the reaching of thermodynamic equilibrium. It showed that themixture of ACs had a synergistic effect if the highly interactingmolecule benzaldehyde was present. The main conclusion oninteractions between flavoured foodstuff and PLA is that aromaticmolecules seem to favour interaction and sorption of ACs in thepolymer and that the interactions occur preferentially throughthe headspace. This has to be anticipated in the aroma formulationof the concerned food product and in the design of the packagingcontainer.

Acknowledgements

The authors would like to acknowledge Alain Guinault, DanielaRusu and Aurélien Piroëlle in preparation of PLA films and Véro-nique Bosc for fruitful discussion. Furthermore the authors wouldlike to thank Bahhar Yesilkaya and Laura Rowenczyk for precioussupport in sample preparation and analysis. We also acknowledgeSENESCYT of Ecuador for funding.

References

Arora, A. P., & Halek, G. W. (1994). Structure and cohesive energy density of fats andtheir sorption by polymer-films. Journal of Food Science, 59(6), 1325–1327.

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

Auras, R., Harte, B., & Selke, S. (2006). Sorption of ethyl acetate and d-limonene inpoly(lactide) polymers. Journal of the Science of Food and Agriculture, 86(4),648–656.

Caner, C. (2011). Sorption phenomena in packaged foods: Factors affecting sorptionprocesses in package-product systems. Packaging Technology and Science, 24(5),259–270.

Colomines, G., Ducruet, V., Courgneau, C., Guinault, A., & Domenek, S. (2010). Barrierproperties of poly(lactic acid) and its morphological changes induced by aromacompound sorption. Polymer International, 59(6), 818–826.

Courgneau, C., Domenek, S., Guinault, A., Averous, L., & Ducruet, V. (2011). Analysisof the structure-properties relationships of different multiphase systems basedon plasticized poly(lactic acid). Journal of Polymers and the Environment, 19(2),362–371.

146 R. Salazar et al. / Food Chemistry 148 (2014) 138–146

Courgneau, C., Domenek, S., Lebossé, R., Guinault, A., Averous, L., & Ducruet, V.(2012). Effect of crystallization on barrier properties of formulated polylactide.Polymer International, 61(2), 180–189.

DeLassus, P. T., Tou, J. C., Babinec, M. A., Rulf, D. C., Karp, B. K., & Howell, B. A. (1988).Transport of Apple Aromas in Polymer Films. In J. H. Hotchkiss (Ed.). Food andPackaging Interactions (Vol. 365, pp. 11–27). American Chemical Society.

Delpouve, N., Saiter, A., & Dargent, E. (2011). Cooperativity length evolutionduring crystallization of poly(lactic acid). European Polymer Journal, 47(12),2414–2423.

Domenek, S., Courgneau, C., & Ducruet, V. (2011). Characteristics and Applicationsof Poly(lactide). In S. Kalia & L. Avérous (Eds.), Biopolymers: Biomedical andEnvironmental Applications (pp. 183–223). John Wiley & Sons, Inc.

Ducruet, V., Vitrac, O., Saillard, P., Guichard, E., Feigenbaum, A., & Fournier, N.(2007). Sorption of aroma compounds in PET and PVC during the storage of astrawberry syrup. Food Additives and Contaminants Part A—Chemistry AnalysisControl Exposure & Risk Assessment, 24(11), 1306–1317.

Dury-Brun, C., Chalier, P., Desobry, S., & Voilley, A. (2007). Multiple mass transfers ofsmall volatile molecules through flexible food packaging. Food ReviewsInternational, 23(3), 199–255.

Fang, X., Domenek, S., Ducruet, V., Réfrégiers, M., & Vitrac, O. (2013). Diffusion ofaromatic solutes in aliphatic polymers above glass transition temperature.Macromolecules, 46, 874–888.

Gavara, R., Catalá, R., Aucejo, S., Cabedo, D., & Hernandez, R. (1996). Solubility ofalcohols in ethylene–vinyl alcohol copolymers by inverse gas chromatography.Journal of Polymer Science Part B: Polymer Physics, 34(11), 1907–1915.

Gillet, G., Vitrac, O., & Desobry, S. (2010). Prediction of partition coefficients ofplastic additives between packaging materials and food simulants. Industrial &Engineering Chemistry Research, 49(16), 7263–7280.

Hermann, B. G., Blok, K., & Patel, M. K. (2010). Twisting biomaterials around yourlittle finger: Environmental impacts of bio-based wrappings. InternationalJournal of Life Cycle Assessment, 15(4), 346–358.

Holm, V. K., Mortensen, G., & Risbo, J. (2006). Quality changes in semi-hard cheesepackaged in a poly(lactic acid) material. Food Chemistry, 97(3), 401–410.

Jin, Z., Tian, Y., & Wang, Y. (2010). Chemistry and Thermodynamics Properties ofLactic Acid and Lactide and Solvent Miscibility. In R. A. Auras, L. T. Lim, S. E. M.Selke, & H. Tsuji (Eds.), Poly(lactic Acid): Synthesis, Structures, Properties,Processing, and Applications (pp. 19–25). John Wiley & Sons.

Johansson, F., & Leufvén, A. (1994). Influence of sorbed vegetable oil and relativehumidity on the oxygen transmission rate through various polymer packagingfilms. Packaging Technology and Science, 7(6), 275–281.

Johansson, F., & Leufven, A. (1997). Concentration and interactive effects on thesorption of aroma liquids and vapors into polypropylene. Journal of Food Science,62(2), 355–358.

Kurek, M., Descours, E., Galic, K., Voilley, A., & Debeaufort, F. (2012). Howcomposition and process parameters affect volatile active compounds inbiopolymer films. Carbohydrate Polymers, 88(2), 646–656.

Nielsen, T. J., Jagerstad, I. M., Oste, R. E., & Wesslen, B. O. (1992). Comparativeabsorption of low-molecular aroma compounds into commonly used foodpackaging polymer films. Journal of Food Science, 57(2), 490–492.

Nielsen, T. J., & Olafsson, G. E. (1995). Sorption of beta-carotene from solutions of afood colorant powder into low-density polyethylene and its effect on theadhesion between layers in laminated packaging material. Food Chemistry,54(3), 255–260.

Pau Balaguer, M., Gavara, R., & Hernández-Muñoz, P. (2012). Food aroma masstransport properties in renewable hydrophilic polymers. Food Chemistry, 130(4),814–820.

Philippe, E., Seuvre, A.-M., Colas, B., Langendorff, V., Schippa, C., & Voilley, A. (2003).Behavior of flavor compounds in model food systems: A thermodynamic study.Journal of Agricultural and Food Chemistry, 51(5), 1393–1398.

Salazar, R., Domenek, S., Courgneau, C., & Ducruet, V. (2012). Plasticization ofpoly(lactide) by sorption of volatile organic compounds at low concentration.Polymer Degradation and Stability, 97(10), 1871–1880.

Story, B. J., & Koros, W. J. (1989). Comparison of three models for permeation of CO2/CH4 mixtures in poly(phenylene oxide). Journal of Polymer Science Part B:Polymer Physics, 27(9), 1927–1948.

Theodorou, E., & Paik, J. S. (1992). Effect of organic vapour interaction on permeationrate through polymer films. Packaging Technology and Science, 5(1), 21–25.

van Willige, R. W. G., Linssen, J. P. H., & Voragen, A. G. J. (2000). Influence of foodmatrix on absorption of flavour compounds by linear low-density polyethylene:Oil and real food products. Journal of the Science of Food and Agriculture, 80(12),1790–1797.

vom Bruck, C., Figge, K., & Rudolph, F. (1981). Interaction of fat- containing foodwith plastics packaging. Journal of the American Oil Chemists’ Society, 58(8),811–815.