CRITICAL AND SYSTEMATIC REVIEW OF POLY(LACTIC ACID) MASS TRANSFER AND EVALUATION OF THE IN - SITU CHANGES OF ITS THERMO - MECHANICAL PROPERTIE S WHEN IMMERSED IN ALCOHOL SOLUTIONS By Uruchaya Sonchaeng A DISSERTATION Submitted to Michigan State University in partial fulfillment of the require ments for the degree of Packaging Doctor of Philosophy 2019 A BSTRACT CRITICAL AND SYSTEMA TIC REVIEW OF POLY(L ACTIC ACID) MASS TRA NSFER AND EVALUATION OF TH E IN - SITU CHANGES OF ITS THER MO - MECHANICAL PROPERTIES WHEN IMMERSED IN ALC OHOL SOLUTIONS By Uruchaya Sonchaeng P roperties of p oly( lactic acid) (PLA) are affected by environmental conditions such as temperature, humidity, and chemical exposure. Mass transport of gases, vapors, and organic compounds in PLA is a concern when designing applications since PLA is permeable to them. Even th ough mass transfer parameters of PLA such as permeability, diffusion, and solubility coefficients have been reported in the literature, the values and units are scattered and inconsistent and most of the analyses only consider PLA as a two - phase structure consisting of a crystalline and an amorphous phase. The recent concept of the three - phase model that separates the amorphous phases into the mobile and rigid amorphous fractions has barely been considered when des gases and vapors, PLA may also contacted with solvents and solutions is scarce. Only a limited number of studies ., in - situ ). Thus, this dissertation aims to: 1 ) provide a comprehensive, systematic, and critical review of mass transfer properties of PLA and PLA - based materials such as blends and composites, along with review of migration of chemical compounds from PL A, and 2 ) evaluate the in - situ changes in thermo - mechanical properties of PLA when in contact with alcohol solutions using a dynamic mechanical analysis technique. The literature review shows that PLA provides moderate barrier to gases, water vapor, and or ganic vapors and that PLA barrier can be enhanced through modification such as blending with other polymers. The in - situ immersion of PLA in alcohol solutions transition temperatures ( T g ) during immersion when compared to t he T g of dry PLA . The T g reductions became smaller as the number of carbon atoms in aliphatic alcohol s C1 C10 increased. Immersion in 50% ( v/v ) 2 - propanol resulted in a T g that was higher than when PLA was immersed in 100% 2 - propanol but lower than when PLA was immersed in water , implying that the concentration s T g . The chemical isomerism in propanol (i.e., 1 - and 2 - propanol) did not affect the T g reduction. The Flory - Huggins interaction parameters and the Hansen solubility parameters were used to explain the reduction in T g of PLA based on the interactions of PLA with the alcohol solutions. The relationship explained the interacti ons between PLA and alcohol s with small molecules (C1 C 8 ), but bigger alcohols (C 9 C10) did not fit the prediction. Overall, the experimental results are not yet sufficient to predict the T g reduction of PLA in other solvents. F urther research on the mass transfer properties of PLA is needed for PLA to reach its full commercial potential. iv To my family and friends , and my mentors, teachers and professors. v ACKNOWLEDGMENTS I would like to thank Dr. Rafael Auras, my major advisor, for his endless support and guidance, and for being a great mentor and role model. I would also like to thank Dr. Susan Selke, Dr. Maria Rubino, and Dr. Loong - Tak Lim for their encouragement and precious comments, and for serving as my committee advisors. I also appreciate the collaborating research projects with Dr. Frank Welle and Dr. Alvin Smucker. I can never thank my parents and family enough for their support and understanding, and for always giving me freedom to choose what I want to do. I am thankful to t he Japanese G overnment S cholarship in support of the study abroad program for the ASEAN youths , and the fellowships from t he School of Packaging and the Col lege of Agriculture and Natural Resources, Michigan State University . I would like to thank the faculty, staff, and students at the School of Packaging, especially students in Dr. Auras research group, for their support. To friends and people I have met d uring my stay in the U.S., thank you so much for making East Lansing feel like home to me. To my old friends, thank you for always keeping in touch no matter how far we are apart. There are people who had passed away and would not be able to read this, but I know if they were alive, they would be very happy for me. Thank you . Y ou will always be remembered. vi TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ........ viii LIST OF FIGURES ................................ ................................ ................................ .......... x KEY TO SYMBOLS AND ABBREVIATIONS ................................ .............................. xvi CHAPTER 1 INTRODUCTION ................................ ................................ ................... 1 1.1 Background and motivation ................................ ................................ ................ 1 1.2 Overall goal and objectives ................................ ................................ ................ 2 1.3 Dissertation overview ................................ ................................ ......................... 3 REFERENCES ................................ ................................ ................................ ............ 4 CHAPTER 2 POLY(LACTIC ACID) MASS TRANSFER PROPERTIES ................... 7 2.1 Abstract ................................ ................................ ................................ .............. 8 2.2 Nomenclature ................................ ................................ ................................ ..... 9 2.3 Introduction ................................ ................................ ................................ ...... 14 2.4 A short background on mass transfer in polymers ................................ ........... 17 2.4.1 Mathematical approach to evaluate mass transfer ................................ .... 17 2.4.2 Experimental methods to quantify mass transfer in polymers .................... 22 2.4.3 Sho rt background of PLA thermal properties and crystal morphology ....... 26 2.4.4 Factors affecting mass transfer in polymers ................................ .............. 37 2.5 Mass transfer of gases in PLA films ................................ ................................ . 40 2.5.1 Permeability ................................ ................................ ............................... 41 2.5.2 Diffusion ................................ ................................ ................................ ..... 53 2.5.3 Solubility ................................ ................................ ................................ .... 56 2.5.4 Effect of modification ................................ ................................ ................. 59 2.5.5 Data gaps and recommendations ................................ .............................. 63 2.6 Mass transfer of oxygen ................................ ................................ ................... 64 2.6.1 Permeability ................................ ................................ ............................... 64 2.6.2 Diffusion ................................ ................................ ................................ ..... 72 2.6.3 Solubility ................................ ................................ ................................ .... 75 2.6.4 Effect of modification ................................ ................................ ................. 78 2.6.5 Data gaps and recommendations ................................ .............................. 80 2.7 Mass transfer of water vapor ................................ ................................ ............ 82 2.7.1 Permeability ................................ ................................ ............................... 83 2.7.2 Diffusion ................................ ................................ ................................ ..... 90 2.7.3 Solubility ................................ ................................ ................................ .... 93 2.7.4 Effect of modification ................................ ................................ ................. 95 2.7.5 Data gaps and recommendations ................................ .............................. 97 2.8 Mass transfer of organic vapors ................................ ................................ ....... 98 2.8.1 Permeability ................................ ................................ ............................... 98 2.8.2 Diffusion ................................ ................................ ................................ ... 101 vii 2.8.3 Solubility ................................ ................................ ................................ .. 104 2.8.4 Effect of modification ................................ ................................ ............... 106 2.8.5 Data gaps and recommendations ................................ ............................ 107 2.9 Comparisons of barrier properties of PLA to common commercial films ........ 107 2.10 Migration of chemical compounds ................................ ............................... 110 2.11 Final remarks ................................ ................................ .............................. 126 2.12 Acknowledgments ................................ ................................ ....................... 128 REFERENCES ................................ ................................ ................................ ........ 129 CHAPTER 3 IN - SITU CHANGES OF THE THERMO - MECHANICAL PROPERTIES OF POLY(LACTIC ACID) FILM IMMERSED IN ALCOHOL SOLUTIONS ................. 156 3.1. Abstract ................................ ................................ ................................ .......... 156 3.2. Introduction ................................ ................................ ................................ .... 157 3.3. Experimental ................................ ................................ ................................ .. 159 3.3.1. Film production ................................ ................................ ........................ 159 3.3.2. Solvents ................................ ................................ ................................ ... 159 3.3.3. Thermal and thermo - mechanical proper ty measurements ...................... 161 3.3.4. Statistical analysis ................................ ................................ ................... 162 3.4. Results and discussion ................................ ................................ .................. 162 3.4.1. Pre - immersion properties ................................ ................................ ........ 162 3.4.2. Effects of solvent sizes ................................ ................................ ............ 165 3.4.3. Effects of branching and concentration ................................ ................... 168 3.4.4. Modelling relationship between the solvent molecules and the changes in the T g of PLA ................................ ................................ ................................ ........ 170 3.5. Conclusion ................................ ................................ ................................ ..... 176 APPENDICES ................................ ................................ ................................ ......... 177 Appendix A: Post - immersion properties ................................ .............................. 178 Appendix B: Detailed calculations ................................ ................................ ....... 184 REFERENCES ................................ ................................ ................................ ........ 188 CHAPTER 4 OVERALL CONCLUSION AND RECOMMENDATIONS FOR FUTURE WORK ............... ................................... ....................................................................... 193 4.1. Overall conclusion ................................ ................................ .......................... 193 4.2. Recommendations for future work ................................ ................................ . 195 REFERENCES ................................ ................................ ................................ ........ 198 viii LIST OF TABLES Table 2 - 1 Crystal forms and systems, chain conformations and cell parameters reported for PLA. ................................ ................................ ................................ ......................... 31 Table 2 - 2 Tentative PLA crystallization model structures for PLA samples below T g . ... 35 Table 2 - 3 Test conditions (temperature and RH) for permeability coefficient ( P ) measurements of neat PLA films for selected pure gases from the literature data used in this review. ................................ ................................ ................................ .................... 43 Table 2 - 4 Activation energy for permeation ( E P ), activation energy of diffusion ( E D ), and heat of sorption ( S ) for selected gases at 0% RH, except fo r ClO 2 at 50% RH. ........ 49 Table 2 - 5 A summary of changes in permeability coefficients ( P ) of gases in PLA as degree of crystallinity ( X c ) increases. ................................ ................................ ............ 52 Table 2 - 6 Average values of E P , E D , and H S of O 2 in PLA below and above T g estimated from literature data [58,59,63,83,143 145,147 188,192 224] as presented in this review, as well as values reported in the literature. ................................ ................ 67 Table 2 - 7 Activation energy of permeation ( E P ), activation energy of diffusion ( E D ), heat of sorption ( S ), heat of condensation ( C ), and heat of melting ( M ) for H 2 O. ...... 86 Table 2 - 8 Mass transfer parameters of organic compounds for modified PLA films, at 25 ° C, 0% RH. ................................ ................................ ................................ ............. 106 Table 2 - 9 Permeability coefficients ( P ) of PLA in kg.m.m - 2 .s - 1 .Pa - 1 to selective gases a nd vapors and a comparison with other commercial available polymers [81]. All P values were measured at 25 °C unless indicated otherwise. ................................ ...... 108 Table 2 - 10 Studies reporting kinetic migration parameters using mathematical models for PLA incorporated with diff erent chemical compounds ................................ ............ 112 Table 2 - 11 D coefficients reported of chemical compounds in different media and temperatures. ................................ ................................ ................................ .............. 121 Table 3 - 1 Solvents used for immersion tests and their properties. .............................. 160 Table 3 - 2 A summary of T g of PLA film sample immersed in different solvents. ......... 171 Table B - 1 The glass transition temperature ( T g ) values of alcohols used for the Fox equation calculations. ................................ ................................ ................................ .. 184 ix Table B - 2 The Hansen solubility parameters (HSP) for PLA [4 0] and solvents used for immersion [25] and the calculated Flory - Huggins interaction parameters ( 12 ). .......... 186 Table B - 3 The adjusted Hansen solubility parameters (HSP) for PLA and solvents used for immersion, the thermal expansion coefficients of liquid ( ), and the calculated Flory - Huggins interaction parameters ( 12 ) at the measured PLA in - situ immersion glass transition temperature ( T g ). ................................ ................................ ......................... 187 x LIST OF FIGURES Figure 2 - 1 Peer - reviewed publications on PLA a and PLA barrier properties b between 1990 and 2016. ( a from Web of Science ® Core Collection search results with keywords b from Web of Science ® ................................ ................................ ............... 16 Figure 2 - 2 Permeation of small molecules from a higher to a lower chemical potential membrane or film side. ................................ ................................ ................................ .. 18 Figure 2 - 3 Common methods to measure per meability, diffusion and solubility in polymer films. Images adapted from: a. [81], b. [81], c. [86], d. [91], e. [91,93], f. [94]. . 25 Figure 2 - 4 a) Chemical structures of a) L (+), D L - lactide, D - lactide and meso - lactide and b) PLA repeatin g unit with an asterisk (*) indicating the chiral carbon atom. ................................ ................................ ................................ ............................. 27 Figure 2 - 5 a) Glass transition ( T g ) and b) melting ( T m ) temperatures of PLA with various combinations of lactide enantiomers versus % D - lactide, adapted from [7,12,98]. References: a [98] , b [97] , c [99], d [60], e [100], f poly( L - co - D - lactides) [98] , g poly( L - co - meso - lactides) [98], h [101]. Each dashed line is based on a linear regression of the overall data in each paper, except for T g of poly( L - co - D , L - lactides) [97] that has two linear re gression lines (% D - lactide = 0 5 and then % D - lactide = 5 or higher). .............. 29 Figure 2 - 6 Glass transition temperature ( T g ) versus number - average molecular weight ( M n ) for PLA with different L : D - lactide ratios, adapted from [7,12]. Dashed lines are predicted lines based on the Flory - Fox equation [105]. References: a [102] , b [103] , c [104], d [60]. ................................ ................................ ................................ ................... 30 Figure 2 - 7 PLA phase model determination. The dashed line represents a two - phase model with crystalline weight fraction ( X c ) and amorphous weight fraction ( X a ), where the sum of these two fractions equals to one. References: a PLA 96 97% L annealed to get semicrystalline samples [119], b PLA 100% L exposed to methanol after drying [121], c PLA 100% L exposed to ethanol after drying [121], d PLA 98% L (4032D) [120]. ........ 33 Figure 2 - 8 A possible schematic representation of the crystalline fraction (CF), the restricted amorphous fraction (RAF), and the mobile amorphous fraction (MAF), adapted from [130]. ................................ ................................ ................................ ....... 34 Figure 2 - 9 a) Crystalline (CF), restricted amorpho us (RAF), and mobile amorphous (MAF) fractions of PLA samples versus annealing time at 100 °C. b) Free volume xi fraction increment of PLA samples versus annealing time at 100 °C. Point Q on the x - axis indicates the as - quenched sample. Figures adapted from [1 34]. ........................... 36 Figure 2 - 10 Specific volume diagram of a polymer as a function of temperature ( T ) , adapted from [142]. An increase in free volume is observed above the glass transition temperature ( T g ). ................................ ................................ ................................ ........... 39 Figure 2 - 11 Arrhenius plot of permeability coefficients ( P ) for O 2 , CO 2 , N 2 , H 2 , He and CH 4 at dry conditions. Data references: O 2 [147], [144] , [133], * [58,59,83,143, 148 188]. CO 2 [63], [144], [172], [58,83,143,156,173,176,177,184 186,189]. N 2 [144], [143], [83,156,167,179,190]. H 2 x [143,144]. He [133,155,161,165,191]. CH 4 + [143]. The vertical dash - dotted line is an arbitrary T g of 58 °C and the dashed lines are from linear regressions of reported experimental data below and above T g . ................................ ................................ ................................ ...... 42 Figure 2 - 12 Plot of average permeability coefficient ( P ) values, with standard error bars, as a function of a) molecular weights ( M w ) and b) kinetic d iameters of different gases for neat PLA at 20 35 °C and 0% RH [58,59,63,83,133,137,143,144,147,149 191,211] , except for P ClO2 which were measured at 50% RH [146] . Each dashed line is a linear trend line, which excludes H 2 and He. The coefficients of deter mination (r 2 ) for the linear trend lines are a) 0.9261 and b) 0.9730. ................................ ................................ ....... 45 Figure 2 - 13 Arrhenius plot of diffusion coefficients ( D ) for O 2 , CO 2 , N 2 , and CH 4 . Data references: O 2 [148], [83], [143], * [133], [163,165,166,168,169,176,180,205]. CO 2 [83,143,176,205]. N 2 [83,143,205]. CH 4 + [143]. Each dashed line represents a least squares linear regression of each gas from the reported experimental data. ....... 54 Figure 2 - 14 Plot of average diffusion coefficient ( D ) values, with standard error bars, as a function of critical volume ( V c ) of different gases for neat PLA at 20 35 °C and 0% RH [83,133,143,148,163,165,166,168,169,176,180,205]. The dashed line is a linear trend line for ln( D ) versus V c . The coefficient of determination (r 2 ) is 0.8286. ........................ 55 Figure 2 - 15 Arrhenius plot of solubility coefficients ( S ) for O 2 , CO 2 , N 2 , and CH 4 . Data references: O 2 [193], [133], * [83,143,148,166,168,176,205]. CO 2 [193], [230], [228], [83,143,176,205]. N 2 [193], [83,143,205,231]. CH 4 + [143]. Each dashed line represents a least squares linear regression of each gas from the reported experimental data. ................................ ................................ ................................ ......... 57 Figure 2 - 16 Plot of average solubility coefficient ( S ) values, with standard error bars, as a function of critical temperature ( T c ) of different gases for neat PLA at 20 35 °C and 0% RH [83,133,143,166,168,176,193,205,228,230,231], except for S ClO 2 which was xii measured at 50% RH [146]. The dashed line is a linear trend line for ln( S ) versus T c . The c oefficient of determination (r 2 ) is 0.8236. ................................ .............................. 58 Figure 2 - 17 Effects of PLA film modifications on a) P CO 2 , b) P N2 , and c) P He . References: a [186], b [173], c [156], d [167], e [179], f [191], g [163], h [168]. The numbers on top and bottom of the bars are P (10 18 kg.m.m 2 .s 1 .Pa 1 ) of neat PLA used in the corresponding experiments. (+) change means increasing P (worse barrier) and ( - ) change means reduction of P (better barrier). Abbreviations: NS = nanosilica, a - Si = amino silane, e - Si = epoxy silane, OMMT = organically - modified montmorillonite , C = Cloisite ® , GO = graphene oxide, GNP = graphene nanoplatelets, PMMA = poly(methyl methacrylate) , PHBV = poly(3 - hydroxybutyrate - co - 3 - hydroxyvalerate), PEG = poly(ethylene glycol), ATBC = acetyl tributyl citrate . ................................ ..................... 62 Figure 2 - 18 Oxygen permeability coefficients ( P O 2 ) of PLA between 5 and 90 °C and 0 and 100% RH. Data references: 0% RH [133], [147], [144], [58,59,83,143,148 167,169 188]. 1 49% RH * [148], [175,202,210]. 50 79% RH [63], [222], [148], [158,175,188,198,200,201,213,214,217,218,221]. 80 100% RH [148], + [153,166,195,206,207,209,223,238]. The vertical dashed line is an arbitrary T g of 58 °C. ................................ ................................ ................................ .................... 65 Figure 2 - 19 a c) P , D , and S values of O 2 measured at 23 C and 0%RH adapted from Guinault et al. [133] with symbols indicating time/temperature conditions of the crystallization treatment for PLA: extruded sample, 85 C , 90 C, 120 C, × 140 C. d f) P , D , and S values of O 2 (shown with symbol) at 20 35 C and 0% RH reported in this review. ................................ ................................ ................................ .. 71 Figure 2 - 20 Oxygen diffusion coefficients ( D O2 ) of PLA between 5 and 40 °C and 0 and 100% RH. Data references: 0% RH [148], [133], [83,143,163,165,166,168,169,176,180,205] . 1 49% RH * [148] . 50 79% RH [148] , [222] , [200] . 80 100% RH [148], + [166,223,238]. ................................ ................. 73 Figure 2 - 21 Oxygen solubility coefficients ( S O 2 ) of PLA between 5 and 40 C and 0 to 90% RH. Data references: 0% RH [148], [133], [83,143,166,168,176,193,205] . 1 49% RH * [148] . 50 79% RH [148] , [222] , [200] . 80 100% RH [148], + [166,223,238]. ................................ ................................ ................................ ............... 76 Figure 2 - 22 Change of PLA P O 2 due to the introduction of additives, blending or compounding with micro - and nanoparticles. The numbers on top and bottom of the bars are P O 2 (10 18 kg.m.m 2 .s 1 .Pa 1 ) of neat PLA used in the corresponding experiments. (+) change means increasing P (wors e barrier) and ( - ) change means xiii reduction of P (better barrier). References: a [153], b [213], c [214] , d [186] , e [165] , f (annealed PLA) [165] , g [182], h [188], i [166], j [168] . Abbreviations: modified MMT = modified montmorillonite, LDH = layered double hydroxide , NS = nanosilica, NS/aminoSi = nanosilica/amino silane, NS/epoxySi = nanosilica/epoxy silane, MMT = montmorillonite, MMT/aminoSi = montmorillonite/amino silane, MMT/epox ySi = montmorillonite/epoxy silane, OMMT = organically - modified montmorillonite, hydroxyl o - LA = hydroxyl lactic acid oligomer, carboxyl o - LA = carboxyl lactic acid oligomer, PBSA = poly(butylene succinate - co - adipate), CA - PBSA = crotonic acid functionalize d poly(lactic acid) coupling poly(butylene succinate - co - adipate), EVOH = ethylene vinyl alcohol, E29+50%AP = blends of EVOH 29% and AP (amylopectin) 50%, ATBC = acetyl tributyl citrate. ................................ ................................ ................................ ...... 80 Figure 2 - 23 Arrhenius plot of water vapor permeability coefficients ( P H 2 O ) of PLA between 6 and 50 °C grouped by different relative humidity (RH) ranges. Each dashed line is from linear regression of the data in each RH range group. References: 1 49% RH [172], [89], x [153,166,175,176,261 263]. 50 79% RH [56], * [63], [42,62,65,89,162,163,172,174,178,182,183,185,197,202,213,252,255,256,258,264 268]. 80 100% RH [56], [147], [254], [269], + [58,89,149,150,152,158,163,171,172,175,184,186,195,214,222,223,247 251,253,257,259,270 276]. ................................ ................................ .......................... 84 Figure 2 - 24 Arrhenius plot of water vapor diffusion coefficients ( D H 2 O ) of PLA between 10 and 50 °C grouped by different relative humidity (RH) ranges. Each da shed line is from linear regression of the data in each RH range group. References: 1 49% RH [21], [89], x [278]. 50 79% RH [56], * [21], [89] . 80 100% RH [56], [21] , [279], [89] , + [222] . ................................ ................................ ................................ .... 91 Figure 2 - 25 Arrhenius plot of water vapor solubility coefficients ( S H 2 O ) of PLA between 10 and 50 °C grouped by different relative humidity (RH) ranges. Each dashed line is from linear regression of the data in each RH range group. References: 1 49% RH [89]. 50 79% RH [56], [89] . 80 100% RH [56], [222], + [89]. ........................ 94 Figure 2 - 26 % P H 2 O change for PLA films and different modifications. The numbers on top of the bars are P H 2 O (10 16 kg.m.m 2 .s 1 .Pa 1 ) of neat PLA used in the corresponding experiments. (+) change means increasing P H 2 O (worse barrier) and ( - ) change means reduction of P H 2 O (better barrier). References: a [272], b [272], c [214], d [56], e [182], f [163], g [262]. Abbreviations: C30B = Cloisite ® 30B, LDH = Mg - Al layered double hydroxide, PEG = poly(ethylene glycol), o - LA = lactic acid oligomer, ATBC = acetyl tributyl citrate, PCL = polycaprolactone. ................................ .............................. 97 xiv Figure 2 - 27 P of different organic compounds grouped by similar test temperatures versus M w at 0% RH. References: methanol [190], ethanol [190], acetaldehyde [184], ethyl acetate [137,159], d - limonene [282], estragole [283], eucalyptol [283]. ................ 99 Figure 2 - 28 Arrhenius plot of permeability coefficients ( P ) of organic compounds in PLA at 0% RH . Ref erences: d - limonene [282], ethyl acetate [137,159], estragole [283], eucalyptol [283], ethanol [190], methanol [190], acetaldehyde [184], trans - 2 - hexanal [184]. ................................ ................................ ................................ ........................... 100 Figure 2 - 29 D of organic compounds grouped by similar test temperature versus their molecul ar volumes ( V ). References: ethyl acetate [137,159], ethyl butanoate [160], ethyl hexanoate [284], d - limonene [282], estragole [283], eucalyptol [283]. ........................ 102 Figure 2 - 30 Arrhenius plot of diffusion coefficients ( D ) of organic compounds in PLA at 0% RH. Refer ences: d - limonene [282], ethyl acetate [137,160,168,284], ethyl butanoate [160], ethyl hexanoate [160,284], estragole [283], eucalyptol [283], linalool [283]. ..... 103 Figure 2 - 31 Arrhenius plot of solubility coefficients ( S ) of organic compounds in PLA at 0% R H. References: 2 - nonanone [285], benzaldehyde [285], ethylene [228], d - limonene [282], ethyl acetate [137,159,168,284,285], ethyl butanoate [285], ethyl hexanoate [284,285], estragole [283], eucalyptol [283]. ................................ ................................ 104 Figure 2 - 32 Comparisons of P of different g ases and vapors at 25 °C in various polymers. (Data adapted from [81] and measured results in this review.) ................... 109 Figure 2 - 33 Migration of organic compounds in ethanol (95% volume by volume). The straight lines shown in the plot are linear least squares regr ession lines at each temperature. References: thymol [316] , butylated hydroxyanisole (BHA) [260], butylated hydroxytoluene (BHT) [195,260], propyl gallate [260], catechin [308], - tocopherol [306], astaxanthin [313]. ................................ ................................ ................................ ........ 125 Figure 3 - 1 Typical pre - immersion (dry sample) test results of PLA from a) dynamic mechanical analysis (DMA) and b) differential scanning calorimetry (DSC). T g , T c , and T m are glass - rubber transition, cold crystallization, and melting temperatures, respectively. ................................ ................................ ................................ ................ 164 Figure 3 - 2 Tan(delta) of PLA films in different aliphatic alcohols as a function of temperature. ................................ ................................ ................................ ................ 166 Figure 3 - 3 Changes in the T g values of the in - situ immersed PLA films in aliphatic alcohols (circle markers, showing average values and standard deviation bars with ) from the T g of dry PLA (horizontal dash line) as a function of number of carbons and molecular volumes of the solvents. Values shown below the circle markers are percent T g reduction in Celsius from the T g of dry PLA. ................................ .............. 167 xv Figure 3 - 4 Tan(d elta) of PLA film in 1 - propanol and 2 - propanol as a function of temperature. Numbers #1 and #2 show replicates of each experiment. ...................... 169 Figure 3 - 5 Tan(delta) of PLA film in water, 2 - propanol and 50% ( v/v ) 2 - propanol aqueous solution as a function of tempera ture. The inset shows T g as a function of 2 - propanol fraction with a linear trendline. ................................ ................................ ...... 170 Figure 3 - 6 Glass transition temperatures ( T g ) prediction of PLA in alcohols. The lines show predicted T g of PLA being plasticized by different alcohols based on the Fox equation [33], compared with the corresponding experimental T g values shown by markers in the same colors as the lines. The numbers C1 C6 indicate the number of carbon atoms in the alcohol ma in chains and C3b denotes 2 - propanol. ..................... 172 Figure 3 - 7 The glass transition temperature ( T g ) of PLA immersed in aliphatic alcohols as a function of the predicted interaction parameters ( 12 ) at T g . The numbers C1 C10 indicate the number of carbon atoms in th e aliphatic alcohol main chains, W denotes water, and C3b denotes 2 - propanol. The fitted exponential decay is shown as a dash line. ................................ ................................ ................................ .............................. 174 Figure 3 - 8 Glass transition temperatures ( T g ) prediction of PLA in 2 - propanol and water. The lines show predicted T g of PLA being plasticized by water and 2 - propanol based on the Fox equation [33], compared with the corresponding experimental T g shown by markers. ................................ ................................ ................................ ...................... 175 Figure A - 1 T g - immersion (wiped - dry) in different alcohols as a function of number of carbon atoms and molecular volume. The immersion temperature and duration were listed in the plot legend, with C representing number of carbo n atoms. ................................ ................................ ............................ 180 Figure A - 2 T g of PLA from DSC post - immersion (wiped - dry) in different alcohols as a function of number of carbon atoms and molecular volume; a) C1, C2, C4, and C9 at various temperatures and 20 min, b) C4 and C9 at 23 °C and various durations. The immersion temperature and duration were listed in the plot legend, with C representing number of carbon atoms . ................................ ................................ ............................ 182 Figure A - 3 Thermograms of the first heating cycle of dry (control) PLA film and PLA films after immersions in different alcohols at 23 ° C for 20 min. C represents number of carbon atoms. ................................ ................................ ................................ ............. 183 Figure A - 4 Thermograms of the second heating cycle of dry (control) PLA film and PLA films after immersions in different alcohols at 23 °C for 20 min. C represents number of carbon atoms. ................................ ................................ ................................ ............. 183 xvi KEY TO SYMBOLS AND A BBREVIATIONS thermal expansion coefficients an empirical factor for interaction parameter calculation d Hansen solubility parameter based on the contributions from the dispersion h Hansen solubility parameter based on the contributions from the hydrogen bond p Hansen solubility parameter based on the contributions from the polar bond C heat of condensation M heat of mixing S heat of sorption 12 interaction parameter AP amylopectin ATBC acetyl tributyl citrate a w water activity BHA butylated hydroxyanisole BHT butylated hydroxytoluene Biophan PLA from Treofan C30B Cloisite® 30B xvii CA - PBSA crotonic acid functionalized poly(lactic acid) coupling poly(butylene succinate - co - adipate ) CF crystalline fraction D diffusion coefficient D ClO 2 diffusion coefficient of chlorine dioxide (gas) D CO 2 diffusion coefficient of carbon dioxide (gas) D H 2 O diffusion coefficient of water vapor DMA dynamic mechanical analysis D O 2 diffusion coefficient of oxygen (gas) DS degree of swelling DSC differential scanning calorimetry E D activation energy of diffusion E D, CO 2 activation energy of diffusion of carbon dioxide (gas) E D, O 2 activation energy of diffusion of oxygen (gas) E P activation energy of permeation E P, CH 4 activation energy of permeation of methane (gas) E P, ClO 2 activation energy of permeation of chlorine dioxide (gas) E P, CO 2 activation energy of permeation of carbon dioxide (gas) E P, H 2 activation energy of permeation of h ydrogen (gas) E P, H 2 O activation energy of permeation of water vapor xviii E P, N 2 activation energy of permeation of nitrogen (gas) E P, O 2 activation energy of permeation of oxygen (gas) EVLON PLA film from BI - AX Internatio nal Inc. EVOH ethylene vinyl alcohol FH Flory - Huggins FTIR - ATR Fourier transform infrared spectroscopy - attenuated total reflection FV free volume HDPE high density polyethylene HSP Hansen solubility parameters IGC inverse gas chromatography K partition coefficient K p,f partition coefficient of solute between polymer p and liquid f LA lactic acid LDH layered double hydroxide LDPE low density polyethylene MAF mobile amorphous fraction MMT montmorillonite M n number - average molecular weight M w weight - average molecular weight MSB magnetic suspension balance NS nanosilica xix o - LA lactic acid oligomer OPLA oriented poly(lactic acid) OMMT organically - modified montmorillonite P permeability coefficient PAN polyacrylonitrile PBSA poly(butylene succinate - co - adipate) PC p olycarbonate P CH 4 permeability coefficient of methane (gas) PCL polycaprolactone P ClO 2 permeability coefficient of chlorine dioxide (gas) P CO 2 permeability coefficient of carbon dioxide (gas) PDLA PLA with isotactic sequences of D - lactide PDLLA PLA formed by meso - lactide ( D , L ) or a mixture of L - and D - lactides , or PLA polymerized from a racemic mixture (50:50) of L - and D - lactides PE polyethylene PEG poly(ethylene glycol ) PEN poly ( ethylene naphthalate) PET poly ( ethylene terephthalate) PG propyl gallate P H 2 permeability coefficient of hydrogen (gas) P H 2 O permeability coefficient of water vapor xx P He permeability coefficient of helium (gas) PHBV poly(3 - hydroxybutyrate - co - 3 - hydroxyvalerate) PLA poly(lactic acid) PLA4030D, PLA4031D, PLA4032D PLA (98% L - lactide) from NatureWorks LLC PLA4040D PLA (94% L - lactide) from NatureWorks LLC PLA5200D PLA (96% L - lactide) from NatureWorks LLC (Not in use according to NatureWorks LLC homepage ) PLLA PLA with isotactic sequences of L - lactide PMMA poly(methyl methacrylate) P N 2 permeability coefficient of nitrogen (gas) P O 2 permeability coefficient of oxygen (gas) PP polypropylene PS polystyrene PTFE polytetrafluoroethylene PVAL poly(vinyl alcohol) PVC poly(vinyl chloride ) PVDC poly(vinylidene chloride ) QCM quartz crystal microbalance QSM quartz spring microbalance RAF r estricted (or rigid) amorphous fraction RH relative humidity RST regular solution theory xxi S solubility coefficient S ClO 2 solubility coefficient of chlorine dioxide (gas) S H 2 O solubility coefficient of water vapor SiOx silicon oxide S O 2 solubility coefficient of oxygen (gas) TBHQ tert - b utylhydroquinone T c critical temperature TEMPO 2,2,6,6 - tetramethylpiperidine - 1 - oxyl T g glass transition temperature TiO 2 titanium dioxide T m melting temperature TMC - 238 N,N - tricyclohexyl - 1,3,5 - benzene - tricarboxylamide TOCN TEMPO - oxidized cellulose nanofib er V molecular volume V c critical volume v/v volume per volume WVTR water vapor transmission rate X c crystallinity degree X MAF mobile amorphous fraction X RAF restricted (rigid) amorphous fraction 1 CHAPTER 1 INTRODUCTION 1.1 Background and motivation Public c oncerns about the negative effects of fossil - based plastics on the environment have result ed in an increase in t he us age of biodegradable and compostable polymers as well as p olymers made from renewable resources [1] . One of the major commercial biodegradable and renewable polymers is poly(lactic acid) (PLA) , which can be made from corn, cassava, or sugar beet. Properties of PLA can be tailored based on the compositions of initial monomers in the production stage [2] . While PLA has suitable properties for applications in the medical, textile, agricultural, and packaging fields [3 6] , its properties are affected by the service and storage environments such as temperature, humidity, and contacted substances. PLA is known to be permeable to gases, vapors, and organic compounds . The values of its mass transfer parameters such as permeability, diffusion, and solubility coefficients have been reported in academic and industrial literature . However, compared to the number of the research reporting other PLA properties, literature on PLA mass transfer properties is scarce and t he reported values and units are inconsistent [7] . Furthermore, a relatively new concept of the three - phase model in semicrystalline polymer that sepa rates the amorphous phase into a rigid a morphous fraction and a mobile amorphous fraction has been prove n to be applicable to PLA [8 11] . T he majority of the reported PLA properties are based on the two phases, namely, 2 the crystalline and amorphous phases. The three - phase model has barely been considered when interpreting PLA properties. Besides mass transfer p roperties of gases and vapors in PL A, there are possible interactions of PLA with solvents and aqueous solutions , which are important to consider ; for example, when PLA is used in applications such as medical implant in the human body or packaging for a liquid medicine or food . In an implan t, PLA is required to dissolve and eventually degrade within the body in a timely manner. On the contrary , when a PLA container is in contact with a liquid medicine or food , dissolution and degradation of PLA is considered a grave failure and possible heal th hazard. A number of articles report ed properties of PLA after being immersed in solvents and solutions ; however, a limited number of articles reported changes in properties of PLA in - situ ) [12] . While the changes after immersion may be used for prediction of interactions during immersion, the in - situ propertie s may lead to valuable insight of what is going on between PLA and the solvents and solutions in contact . With thes e gaps in knowledge for PLA barrier properties, t here is at least a need to reassess the mass transfer of PLA and evaluate the in - situ changes in PLA properties during immersion in solvents and solutions. 1.2 Overall goal and objectives The overall goal of this dissertation is to provide a critical and systematic review o f mass transfer of PLA as well as to evaluate the in - situ changes in P LA properties when in 3 contact with solvents and aqueous solutions. To achieve this goal, this dissertation aims to address two specific objectives , which are : Objective 1 : T o provide a comprehensive, systematic , and critical review of the mass transfer of gases, vapors , and organic compounds in PLA . Objective 2 : To gain an understanding of the in - situ changes in thermo - mechanical properties of PLA when in contact with alcohol solutions. 1.3 Dissertation overview This dissertation is organized as follow s . Th e current chapter (Chapter 1) gives a general idea of the motivation and importance of this study including the overall goal and the specific objectives. Chapter 2 is a version of a published article that provides a comprehensive, systematic , and critical review of the mass transfer properties of PLA , which covers mass transfer background, mass transfer of gases, water vapors, and organic vapors and migration of chemical compounds in PLA, as well as comparisons of PLA barrier properties with other commercia l polymers. C hapter 3 investigates the in - situ thermo - mechanical properties of PLA during immersion in selected alcohol solutions and evaluates the relationship s between the solvent molecules and the changes in the properties of PLA during immersion compar ed to properties before immersion. Finally, Chapter 4 summarize s t he work in this dissertation with overall conclusion and recommendations for future work. 4 REFERENCES 5 R EFERENCES [1] Song JH, Murphy RJ, Narayan R, Davies GBH. Biodegradable and compostable alternatives to conventional plastics. Philos Trans R Soc B Biol Sci 2009;364:2127 39. [2] Lim L - T, Auras R, Rubino M. Processing tech nologies for poly(lactic acid). Prog Polym Sci 2008;33:820 52. [3] Jamshidian M, Tehrany EA, Imran M, Jacquot M, Desobry S. Poly - lactic acid: Production, applications, nanocomposites, and release studies. Compr Rev Food Sci Food Saf 2010;9:552 71. [4] Cast ro - Aguirre E, Iñiguez - Franco F, Samsudin H, Fang X, Auras R. Poly(lactic acid) Mass production, processing, industrial applications, and end of life. Adv Drug Deliv Rev 2016;107:333 66. [5] Auras RA, Singh SP, Singh JJ. Evaluation of oriented poly(lactide) polymers vs. existing PET and oriented PS for fresh food service containers. Packag Technol Sci 2005;18:207 16. [6] Auras R, Harte B, Selke S. An overview of polylactides as packaging materials. Macromol Biosci 2004;4:835 64. [7] Sonchaeng U, Iñiguez - Franco F, Auras R, Selke S, Rubino M, Lim LT. Poly(lactic acid) mass transfer properties. Prog Polym Sci 2018;86:85 121. [8] Nassar SF, Domenek S, Guinault A, Stoclet G, Delpouve N, Sollogoub C. Structural and dynamic heterogeneity in the amorphous phase of poly(L,L - lactide) confined at the nanoscale by the coextrusion process. Macromolecules 2018;51:128 36. [9] Righetti MC, Gazzano M, Delpouve N, Saiter A. Contribution of the rigid amorphous fraction to physical ageing of semi - crystalline PLLA. Pol ymer (Guildf) 2017;125:241 53. [10] Delpouve N, Arnoult M, Saiter A, Dargent E, Saiter J - M. Evidence of two mobile amorphous phases in semicrystalline polylactide observed from calorimetric investigations. Polym Eng Sci 2014;54:1144 50. [11] M. Study of crystalline and amorphous phases of biodegradable poly(lactic acid) by advanced thermal analysis. Polymer (Guildf) 2009;50:3967 73. [12] Iñiguez - Franco F, Auras R, Burgess G, Holmes D, Fang X, Rubino M, et al. Concurrent solvent induced crystal lization and hydrolytic degradation of PLA by 6 water - ethanol solutions. Polymer (Guildf) 2016;99:315 23. 7 CHAPTER 2 POLY(LACTIC ACID) MA SS TRANSFER PROPERTI ES A version of this chapter is published as: Sonchaeng, U., Iñiguez - Franco, F., Auras, R., Selke, S., Rubino, M., & Lim, L. T. (2018). Poly(lactic acid) mass transfer properties. Progress in Polymer Science, 86, 85 121. http://doi.org/ 10.1016/j.progpolymsci.2018.06.008 8 2.1 A bstract Poly(lactic acid) (PLA) , a biodegradable and compostable polymer , is gaining market acceptance and has been extensively investigated. The versatility of PLA has led to its broad and different applications i n medical , agriculture, and food packaging fields. Similar to other polymer s , PLA is permeable to gases, vapors and organic compounds. Thus, the mass transfe r properties of PLA can influence its suitability for end - use application s . Here, we present a comprehensive, systematic , and critical review of more than 300 papers published since 1990 reporting the mass transfer properties of PLA , which include permeabi lity, diffusion and solubility to gases, water vapor and organic vapors, along with migration of chemical compounds from PLA. Overall, PLA provides moderate barrier to gases , water vapor , and organic compounds . B arrier enhancement can be achieved through m odifications such as ble nd ing with other polymers and formation of composite structures . Most of the mass transfer parameters reported in the literature are based on two - phase mobile amor phous and crystalline fractions, omitting the role of the restricted amorphous fraction , which can lead to unclear comprehension of PLA barrier properties as well as what a ffects those properties. A dditional research is needed to address this shortcoming. This review provides an in - depth analysis of PLA mass tran sfer and a foundation for future research and commercial development. 9 2.2 Nomenclature AP amylopectin ATBC acetyl tributyl citrate a w water activity BHA butylated hydroxyanisole BHT butylated hydroxytoluene Biophan PLA from Treofan C30B Cloisite® 30B CA - PBSA crotonic acid functionalized poly(lactic acid) coupling poly(butylene succinate - co - adipate ) CF crystalline fraction D diffusion coefficient D ClO 2 diffusion coefficient of chlorine dioxide (gas) D CO 2 diffusion coefficient of carbon dioxide (gas) D H 2 O diffusion coefficient of water vapor D O 2 diffusion coefficient of oxygen (gas) DS degree of swelling E D activation energy of diffusion E D, CO 2 activation energy of diffusion of carbon dioxide (gas) E D, O 2 activation energy of diffusion of oxygen (gas) E P activation energy of permeation E P, CH 4 activation energy of permeation of methane (gas) 10 E P, ClO 2 activation energy of permeation of chlorine dioxide (gas) E P, CO 2 activation energy of permeation of carbon dioxide (gas) E P, H 2 activation energy of permeation of h ydrogen (gas) E P, H 2 O activation energy of permeation of water vapor E P, N 2 activation energy of permeation of nitrogen (gas) E P, O 2 activation energy of permeation of oxygen (gas) EVLON PLA film from BI - AX International Inc. EVOH ethylene vinyl alcohol FTIR - ATR Fourier transform infrared spectroscopy - attenuated total reflection FV free volume HDPE high density polyethylene C heat of condensation M heat of mixing S heat of sorption IGC inverse gas chromatography K p,f partition coefficient of solute between polymer p and liquid f LA lactic acid LDH layered double hydroxide LDPE low density polyethylene 11 MAF mobile amorphous fraction MMT montmorillonite M n number - average molecular weight M w weight - average molecular weight MSB magnetic suspension balance NS nanosilica o - LA lactic acid oligomer OPLA oriented poly(lactic acid) OMMT organically - modified montmorillonite P permeability coefficient PAN polyacrylonitrile PBSA poly(butylene succinate - co - adipate) PC p olycarbonate P CH 4 permeability coefficient of methane (gas) PCL polycaprolactone P ClO 2 permeability coefficient of chlorine dioxide (gas) P CO 2 permeability coefficient of carbon dioxide (gas) PDLA PLA with isotactic sequences of D - lactide PDLLA PLA formed by meso - lactide ( D , L ) or a mixture of L - and D - lactides , or PLA polymerized from a racemic mixture (50:50) of L - and D - lactides PE polyethylene PEG poly ( ethylene glycol ) 12 PEN poly ( ethylene naphthalate ) PET poly ( ethylene terephthalate ) PG propyl gallate P H 2 permeability coefficient of hydrogen (gas) P H 2 O permeability coefficient of water vapor P He permeability coefficient of helium (gas) PHBV poly(3 - hydroxybutyrate - co - 3 - hydroxyvalerate) PLA poly(lactic acid) PLA4030D, PLA4031D, PLA4032D PLA (98% L - lactide) from NatureWorks LLC PLA4040D PLA (94% L - lactide) from NatureWorks LLC PLA5200D PLA (96% L - lactide) from NatureWorks LLC (Not in use according to NatureWorks LLC homepage ) PLLA PLA with isotactic sequences of L - lactide PMMA poly(methyl methacrylate) P N 2 permeability coefficient of nitrogen (gas) P O 2 permeability coefficient of oxygen (gas) PP polypropylene PS polystyrene PTFE polytetrafluoroethylene PVAL poly(vinyl alcohol) PVC poly ( vinyl chloride ) PVDC poly ( vinylidene chloride ) 13 QCM quartz crystal microbalance QSM quartz spring microbalance RAF r estricted (or rigid) amorphous fraction RH relative humidity RST regular solution theory S solubility coefficient S ClO 2 solubility coefficient of chlorine dioxide (gas) S H 2 O solubility coefficient of water vapor SiOx silicon oxide S O 2 solubility coefficient of oxygen (gas) TBHQ tert - b utylhydroquinone T c critical temperature TEMPO 2,2,6,6 - tetramethylpiperidine - 1 - oxyl T g glass transition temperature TiO 2 titanium dioxide T m melting temperature TMC - 238 - tricyclohexyl - 1,3,5 - benzene - tricarboxylamide TOCN TEMPO - oxidized cellulose nanofib er V molecular volume V c critical volume WVTR water vapor transmission rate X c crystallinity degree 14 2.3 Introduction Mass transfer properties play a crucial role in the research, development, and applications of polymers . Low molecular weight gases, vapors and organic compounds can absorb into and permeate through a polymer matrix. Whether it is determining the pollutant filtration capacity of a plastic membrane [ 1,2] or predicting the shelf life of pharmaceutical and food products [3 6] , polymer mass transfer parameters are critical to know . Nevertheless, sys tematic reviews of the mass transfer properties of most polymers are not readily available. For relatively new polymers such as poly(lactic acid) (PLA) - a biodegradable and bio - based polymer used for medical, agricultural and packaging applications [7,8] - data on its mass transfer properties are either scarce or widely dispersed in the technical and commercial literature. Thus, the goal of this paper is to provide a comprehensive, systematic and critical review of the mass transfer of gases, vapors and organic compounds in PLA. PLA is by far the most researched commercial biodegradable and compostable polymer [9] . It is derived from renewable resources such as corn, cassava, and sugar beets and can be commercially produced from the condensation polymerization of lactic acid (LA) or ring - opening polymerization through lactide [7 10] . NatureWorks LLC is currently the primary producer of PLA while other compani es, including LA manufacturers, are also diversifying into PLA mass production [11] . PLA , like any other polymer, is permeable to gases, vapors and organic compounds, which may impact its end - use performance. From an in trinsic factor standpoint, PLA barrier properties are affected by the enantiomer compositions of LA (i.e., L - LA and D - LA) or lactide (i.e., L - lactide and D - lactide) . D ifferent configurations of LA or lactide can result in PLA with 15 different crystallinity and thermal properties [12 16] . On the other hand, temperature is an extrinsic factor that affects the mass transfer properties of PLA. T he increase in t emperature can enhance the diff usion of gases and vapors, resulting in glass - to - rubber transition, plasticization , and deterioration of a polymer. For m easuring and modeling the mass transfer parameters ( i.e. , permeability ( P) , diffusion ( D) and solubility ( S) coefficients) of PLA, its transition temperatures, as well as the test temperatures , must be considered [1 7 20] . Relative humidity (RH) is another extrinsic factor that can affect the barrier properties of PLA. Exposing PLA to high RH can plasticize the polymer, resulting in non - Fickian mass transfer phenomena [21] . While PLA may have advantages over other polymers because of its biodegradability and origin from renewable resources, it also has significant disadvantages such as brittleness, poor t hermal stability, low toughness , and low elongation at break [13,22] . PLA modifications, such as blending with other polymers , incorporation of additives, formation of composites and nanocomposites , may impact its mass transfer properties in different ways [2 2 59] . So, a comprehensive understanding of the factors affecting PLA mass transfer properties is critical . R esearch activities on PLA have significantly increased over the past few decades. Figure 2 - 1 shows the number of peer - reviewed publications on PLA and PLA mass transfer properties published since 1990. While the number of publicatio ns on PLA has increased exponentially in areas such as material [8,9,13,60] , optical [61] , thermal [57,62] and mechanical [62 65] properties , as well as its end of life [9] , only a small number of the contributions are related to mass transfer. Thus, there is a need for a review of PLA barrier properties and the current data gaps. 16 Figure 2 - 1 Peer - reviewed publications on PLA a and PLA barrier properties b between 1990 and 2016. ( a from Web of Science ® Core Collection search results with keywords b from Web of Science ® The objectives of this review are: To prov ide a comprehensive, systematic and critical review of experimental data for the mass tran sfer parameters ( i.e. , P , D , and S ) of PLA to gases , vapors , and organic compounds . To identify the main factors affecting PLA mass transfer properties , such as temperature, RH, morphology, ratios of L - to D - lactide , and thermal history. 17 To review the modification methods, such as blends, composites and nanocomposites o f PLA, and understand their effects on PLA mass transfer properties. To highlight mass transfer properties that ha ve not been adequately addressed. 2.4 A short background o n mass transfer in polymers 2.4.1 Mathematical approach to evaluate mass transfer [66] on gas diffusion brought to the attention of researchers the mass transfer of small molecules through membranes . Small molecules (i.e., permeants) permeate through a polymer from high to low chemical potential ( ) to maintain thermodynamic equilibrium. This difference in is the fundamental driving force for mass transfer through polymers [67 69] . For a permeant , its chemical potential, , can be expressed in terms of its chemical activity, , as: ( 1 ) where is the chemical potential of the permeant at a standard state, R is the universal gas constant , and T is temperature in Kelvin. At a typical standard state, the chemical activity of permeant is approximately equal to its concentration, [67 69] . In the gas phase, concentration can be expressed as partial pressure, , following the ideal gas law as expressed in Eq. 2 where is the numbe r of the permeant molecules and v is volume. ( 2 ) For a non - ideal gas, the fugacity, , can be used instead of pressure [70] . 18 Permeation in polymer s consist s of three steps [71] : (i) sorption of the permeant into the polymer matrix on the high concentration surface; (ii) diffusion of the permeant through the polymer matrix along a concentration gradient toward s the low concentration side; and (iii) desorption or e vaporation of the permeant from the low concentration surface. Figure 2 - 2 shows a schematic of the permeation process in a homogeneous polymeric film. Deviation from a straight line during the diffusion process can occur when the permeant interacts with the polymer , i.e., when the mass transfer does not follow the Fickian behavior , as may be classified as the diffusion - relaxation model for non - Fickian [21,72,73] . Figure 2 - 2 Permeation of small molecules from a higher to a lower chemical potential membrane or film side. From a mass balance standpoint, assuming that diffusion takes place only in the x - direction in a polymeric membrane, the proc ess can be described by the relationship between the flux ( F ) 19 diffusion [68] : ( 3 ) where D is the diffusion coefficient, c is the concentration, x is the direction of movement of the permeant and dc/dx is the concentration gradient in the direction of the flow. Eq. 3 can be used when the permeant conc entration does not change with time (i.e., steady state). The flux of a permeant through a polymeric film is defined as the amount of permeant passing through a surface of unit area normal to the direc tion of flow per unit time and can be described by Eq. 4 at steady state: ( 4 ) where Q is the amount of permeant, A is the area , and t is time. If the permeant concentrations on both sides of the film, c 1 and c 2 , remain constant, Eq. 3 can then be integrated across the total thickness of the film ( L ), resulting in: ( 5 ) By replacing F using Eq. 4 , Q can be derived as : ( 6 ) When the permeant is a gas, it is more convenient to measure the vapor pressure ( p ), which is at equilibrium with the polymer, rather than the actual concentration within the polymer. At sufficiently low concentration and when the permeant does not interac l aw applies [68] and c can be expressed as: ( 7 ) 20 where S is the solubility coefficient of the permeant in the polymer . At different temperatures, amorphous regions in a polymer exist in either the glassy or rubbery state. Below the T g ), the amorphous regions are in the glassy state where segmental movements of the polymer chains are restricted, and hence the polymer tends to be rigid. Above T g , the polymer chains have more freedom in movement. Existing in the rubbery state, the polymer tends to be flexible. For gases and aroma compounds permeating through a glassy the Henry - Langmuir adsorption equation (Eq. 8 ) can be used. The Henry - Langmuir theory takes into consideration the sorption of the gas in the frozen free volume (FV) of the polymer matrix [74] : ( 8 ) where is the saturated concentratio n of the gas in the FV , and b is the FV affinity constant or ratio of rate constant for adsorption and desorption. I n the case o f high permeant concentration or if the permeant interacts with the polymer, the Flory - Huggins equation (Eq. 9 ) can be used to estimate the permeant concentration within the FV o f the polymer [75] : ( 9 ) w here is the chemical activity of the permeant, is the volume fraction of the permeant in the polymer and is the Flory - Huggins interaction parameter between the permeant and the polymer [76] . Assuming no interaction between the permeant and the polymer, a pplying l aw (Eq. 7 ) to Eq. 6 results in: 21 ( 10 ) which can be rearranged as: ( 11 ) where P is defined as the permeability coefficient of a permeant at steady state and is the partial pressure gradient of the permeant, . P can be determined from the transmission rate data or from the P = DS relationship , where D and S are determined separately [77] . This equation is very simplistic and mostly suitable for rubbery polymers. For g lassy polymers, due to their restricted polymer chain mobility, permeation phenomena may deviate from this relationship. the process of permeation [78] : ( 12 ) where the left side of Eq. 12 is the rate of change of permeant concentration. When there is a strong polymer - permeant interaction, D is dependent on time, position , and concentration [79] . Eq. 12 may be solved using numerical methods. However, if D is time - , position - and concentration - independent, Eq. 12 can be written as: ( 13 ) If the unsteady state and steady state portions of the mass transfer are included and l aw applies, Eq. 13 can be solved as: ( 14 ) 22 where is the flow rate of the permeant through the film at time t du ring the unsteady state portion and is the flow rate at equilibrium (steady state) [80] . Based on the mass transfer profile from Eq. 14 , D can be estimated from: ( 15 ) where is the time when = 0.5. P can be calculated if the value of is known : ( 16 ) 2.4.2 Experimental methods to quantify mass transfer in polymers To determine permeability of gases and vapors in a polymeric film, isostatic and quasi - isostatic methods are often used [81] . In the isostatic method (also known as continuous - flow method ), a film is mounted in a chamber where one side of the film is exposed to a known constant concentration of the permeant and the other side of the film is maintained at near - zero permeant concentration. On the zero - concentration side, the permeant passing t hrough the film is being purged by an inert carrier gas to a detector for quantification. After obtaining the data of permeation rate as a function of time from an experiment, Eqs. 11 16 can be applied to calculate P and D . A schematic diagram and plot of the isostatic method are shown in Figure 2 - 3 a with the unsteady state portion of the experimental data in the shaded area of the plot . In the quasi - isostatic method (also known as the lag - time or constant - volume/variable - pressure method ) , a film is exposed to the permeant on one side and on the other side the concentration is accumulated in general to values below 5 wt% of the concentration on the high concentration side [82] . Samples of permeant from the accumulating side a re 23 taken and quantified at certain time intervals to generate a plot of permeant quantity versus time. The x - axis intercept from the steady - state portion of the plot is the lag time , t , which can be used to estimate D : ( 17 ) From the slope of the linear portion of the permeation plot, i.e., when the permeation is in a steady - state, P can be estimated [82,83] . For both methods, the environmental conditions such as temperature and RH are kept constant throughout the experiment and should be reported together with the results. Using either of the two methods , P and D can be calculated and then S can be estimated from P = DS l aw applies . A schematic diagram and plot of the quasi - isostatic method is shown in Figure 2 - 3 b with the unsteady state portion of the experimental data in the shaded area of the plot. The water vapor transmission rate (WVTR) cup method shown in Figure 2 - 3 c mea sures the weight gain from the amount of moisture transported through a film sample and absorbed by a desiccant. A plot of moisture uptake versus time can be used to calculate WVTR and P . The WVTR cup method has also been used for organic vapors [84,85] . An extensive review of methods to measure permeability for gases and water vapor in polymeric films can be found elsewhere in the literature [81,86] . For vapors and aroma compounds, sorption measurements may be preferable as they have some advantages over permeability measurements. For example, leakage or pinholes in the films will not affect the results [77] . In sorption measurements, the gain or loss of weight of the fil m is measured as a function of time while the film is exposed to a constant concentration or vapor pressure of the permeant. Equipment such as a 24 McBain sorption balance, magnetic suspension balance (MSB), or quartz crystal microbalance (QCM) have often bee n used in sorption measurements. Schematic diagrams of the McBain sorption balance and MSB are shown in Figure 2 - 3 d and QCM in Figure 2 - 3 e. A McBain sorption balance, named after the scientist who inve nted it, is a high vacuum quartz spring adsorption apparatus. Using a quartz spring, the balance can be used for measuring vapor sorption by solid surfaces [87] . For MSB, a magnetic suspension coupling is used to separate the balance from the measuring atmosphere and allows a contact - free weighing method [88] . A QCM measures weight change by measuring the change in resonant frequency of a quartz crystal where the change of mass due to absorption at the crystal surface can be mathematically calculated from the change in frequency [89] . Using microbalance systems with low partial pressure of the solute [90] may reduce mass transfer resistance for the adsorption of the solute onto the polymer surface. However, with this approach, leakages may occur on a long - running experiment . Inverse gas chromatography (IGC) shown in Figure 2 - 3 f has resurged as a method to quantify mass transfer parameters of organic vapors in polymers. It uses a polymer packed i n the IGC column as a stationary phase and a small quantity of the test compound as a mobile phase. Identification and quantification o f the compound can be achieved using its response and retention time [91] . A detailed review of these sorption methods can be found elsewhere in the literature [91,92] . 25 Figure 2 - 3 Common methods to measure permeability, diffusion and solubility in polymer films. Images adapted from: a. [81] , b. [81] , c. [86] , d. [91] , e. [91,93] , f. [94] . 26 2.4.3 Short background of PLA thermal properties and crystal morphology This section summarizes the thermal properties and crystalline structure s relevant to PLA barrier properties. For further detailed information about these topics, the readers are directed to specific review articles [12,95,96] . PLA is synthetized from LA or lactide. Because LA and lactide have a chiral carbon (also known as an asymmetric carbon), they exist in different enantiomers ( Figure 2 - 4 a). Different amounts of LA or lactide enantiomers can be combined to produce the final high molecular weight PLA with a basic repeating unit as shown in Figure 2 - 4 b, which has a molar mass of 72.06 g.mol - 1 . PLA formed by isotactic sequences of L - lactide (or L - LA) and D - lactide (or D - LA) are commonly referred to as PLLA and PDLA, respectively. However, t he term PDLLA may represent PLA formed by meso - lactide or a mixture of L - and D - lactide (or L - and D - LA) , or PLA polymerized from a racemic mixture (50:50) of L - and D - lactides (or L - and D - LA) . To avoid confusion, this review will refer to PLA by its L - or D - enantiomer composition in the final product (e.g. , PLA 92% L ), regardless of its production or processing methods used . Depending on the combination of LA or lactide, final PLA properties can be tailored and changed [8] . Readers may refer to the original papers for details on PLA samples included in this review. 27 Figure 2 - 4 a) Chemical structures of a) L (+), D L - lactide, D - lactide and meso - lactide and b) PLA repeating unit with an asterisk (*) indicating the chiral carbon atom. In general, the T g and melting temperature ( T m ) of PLA are affect ed by the L - and D - lactide contents. Bigg [97] reported a reduction in T g as the amount of D - lactide in creased in ( L - / D , L ) random copolymer s of PLA (i.e., poly( L - co - D , L - lac tides) , made from copolymerization of poly( L - lactides) with copolymers made from a random copolymer of 50% L - and 50% D - lactide ). However, as shown in Figure 2 - 5 a, when D - lactide exceeds 5%, T g no longer changes. Feng et al. [98] recently reported a reduction in T g with increasing amount of D - lactide when various combinations of lactide enantiomers were used as comonomers in PLA copolymers (also shown in Figure 2 - 5 a). Similarly, a reduction in T m was observed as the amount of D - lactide increased [60,97 101] , as shown in Figure 2 - 5 b . Furthermore, Feng et al. [98] highlighted that the differences in T g and T m between poly( L - co - D - lactides) and poly( L - co - meso - lactides) with the same amount of D - lactide were a result of different contributions of D - lactide and meso - lactide to the disruption (i.e., disorder degree) of PLA molecular chain tacticity . Saeidlou et al. 28 [12] conducted an extensive review of PLA crystallization and reported the variation of T g as a function of number - average molecular weight ( M n ) [60,102 104] , as shown in Figure 2 - 6 . Apparently, as M n reaches about 100 kg.mol 1 , T g remains stable regardless of the type of PLA. The effect of the ratio of L : D - lactide is also evident in Figure 2 - 6 that T g tends to decrease as the ra tio of D - lactide increases. 29 Figure 2 - 5 a) Glass transition ( T g ) and b) melting ( T m ) temperatures of PLA with various combinations of lactide enantiomers versus % D - lactide, adapted from [7,12,98] . References: a [98] , b [97] , c [99] , d [60] , e [100] , f poly( L - co - D - lactides) [98] , g poly( L - co - meso - lactides) [98] , h [101] . Each dashed line is based on a linear regression of the overall data in each paper, except for T g of poly( L - co - D , L - lactides) [97] that has two linear regression lines (% D - lactide = 0 5 and then % D - lactide = 5 or higher). 30 Figure 2 - 6 Glass transition temperature ( T g ) versus number - average molecular weight ( M n ) for PLA with different L : D - lactide ratio s , adapted from [7,12] . Dashed lines are predicted lines based on the Flory - Fox equation [105] . References: a [102] , b [103] , c [104] , d [60] . refer [12,14,106] . Their crystal systems, chain con formations and cell parameters are summarized in Table 2 - 1 . Extensive reviews of PLA crystal structures have been published [12,95,96] . 31 Table 2 - 1 Crystal forms and systems, chain conformations and cell parameters reported for PLA. Crystal form Crystal system Helical chain confor - mation Cell parameters References a , nm b , nm c, nm Pseudo - orthorhombic Orthorhombic 10 3 1.03 1.07 1.05 1.07 0.59 0.64 0.60 0.61 2.78 2.88 2.87 2.88 [61,107 112] Orthorhombic Trigonal 3 1 1.03 1.04 1.05 1.77 1.82 1.05 0.90 0.88 [109,113,114 ] Orthorhombic 3 1 0.99 0.62 0.88 [115] (or Pseudo - hexagonal n/a 1.08 0.62 2.88 [14] n/a: not available Polyme r chains of PLA are longer than the th ickness of the crystal lamellae , and therefore they can be entangled on different phases according to the degree of coupling. In the past, a semicrystalline polymer was believed to be composed of an amorphous phase and a crystalline phase . Michaels and Bixler [116] studied the solubility of gases in polyethylene (PE) and proposed that it was sufficient to consider PE as consisting of two phases, amorphous and crystalline, and that the crystalline phase in PE did not sorb gas molecules to a measurable extent. However, later work of Menczel and Wunderlich [117] showed that the amorphous portions in semicrystalline polymers were different from the amorphous portions in fully amorphous polymers . Later on, Wunderlich [118] examined heat capacities of semicrystalline polymers and correlat ed a negative contribution to heat capacity between T g and T m to another phase in a semicrystalline polymer called a rigid amorphous region , which exists due to a 32 strong coupling between the crystal line and the amorphous phases. R ecently Nguyen et al. [119] demonstrated that in PLA and poly(ethylene terephth alate) ( PET ) , the amount of c rystalline fraction ( X c ) and the amount of amorphous fraction ( X a ) did not add up to one, which invalidated the two - phase model. In the case of PET, the deviation from the two - phase model starts at X c < 0.1, but in PLA the deviation occurs at a relatively higher crystallinity ( X c > 0.3). Figure 2 - 7 shows how PLA deviates from the two - phase model as reported by a number of authors [119 121] . A dashed line represents the two - phase model, where X a + X c = 1 . Deviation from the dashed line implies the presence of a third phase. The degrees of deviation vary likely due to the different crystallization m ethodolog ies applied and samples of different L - lactide contents used. Additional evidence of the deviation from the two - phase model in PLA and PET can be found in the literature [122 127] . 33 Figure 2 - 7 PLA phase model determination. The dashed line represents a two - phase model with crystalline weight fraction ( X c ) and amorphous weight fraction ( X a ), where the sum of these two fractions equals to one. References: a PLA 96 97% L annealed to get semicrystalline samples [119] , b PL A 100% L exposed to methanol after drying [121] , c PL A 100% L exposed to ethanol after drying [121] , d PLA 98% L (4032D) [120] . Th e s e findings contributed to the evidence for the assumption of a three - phase model in semicrystalline polymers. The proposed three phases are (1) a crystalline fraction (CF); (2) a mobile amorphous fraction (MAF); and (3) a restricted or rigid amorphous fracti on (RAF). Figure 2 - 8 shows a possible schematic representation of these three domains where the RAFs are constrained by the adjacent CFs . Alternatively, Delpouve et al. [128] used calorimetric methods to investigate the amorphous phase dynamics in semicrystalline PLA . They proposed that beside s the CF, three amorphous phase s with different molecular mobilit ies could coexist, namely, 34 the RAF, the inter - spherulitic MAF and the in tra - spherulitic MAF . Different models such as one with the CF surrounded by a continuum of mobility of the RAF and the MAF [129] were also proposed for other polymers. Figure 2 - 8 A possible schematic representation of the crystalline fraction (CF), the restricted amorphous fraction ( RAF ), and the mobile amorphous fraction ( MAF ) , adapted from [130] . So, t he ten tative structures that can be found in PLA films below T g are shown in Table 2 - 2 . In general, PLA with greater than 8% D - lactide is totally amorphous while that with less than 8% D - lactide is semicrystalline. The highest percentage of D - lactide used in commercial PLAs is approximately 12%. 35 Table 2 - 2 Tentative PLA crystallization model structures for PLA samples below T g . Amount of D - lactide Structure Possible crystallinity model References 8 12% a MAF amorphous One phase [7,131] 2 8% MAF, RAF and CF semicrystalline Three - phase [132] <1% MAF, RAF and CF semicrystalline Three - phase [132] a This applies to commercial PLA where the highest percentage of D - lactide is ~12%. The f ormation of the RAF depends of the polymer and the crystallization processes [132,133] . D el Rio et al. [134] showed that d uring the annealing of PLA, the MAF decrease d while the RAF and the CF increase d . They explained that the formation of new void s of smaller FV increase d t hrough the crystallization process due to the vitrification of the RAF chains . They also attributed t he increase in FV fraction of PLA during annealing to the difference in FV void size and distribution between the RAF and the MAF, which contribut ed to a de - densification of the non - crystalline domain of PLA, as illustrated in Figure 2 - 9 . 36 Figure 2 - 9 a) Crystalline (CF), restricted amorphous (RAF), and mobile amorphous (MAF) fractions of PLA samples versus annealing time at 100 °C. b) Free volume f raction increment of P LA samples versus annealing time at 100 °C . Point Q on the x - axis indicates the as - quenched sample. Figures adapted from [134] . 37 2.4.4 Factors affecting mass transfer in polymers As mentioned in Section 2.3 , mass transfer in polymers is affected by intrinsic and extrinsic factors. The nature of the polymer such as chemical composition, polarity, stiffness of the polymer chains, bulkiness of side - and backbone - chain groups and the degree of crystallinity sign ificantly impact the sorption and diffusion of a permeant [135] . A semicrystalline polymer can hav e varying degrees of crystallinity depending on processing conditions and thermal history. A higher degree of crystallinity, within the same generic class of polymer, usually affords a stronger barrier, since the permeant cannot diffuse through the crystal line domains [116,136] . It is generally accepted that for semicrystalline rubbery polymers, D and S can be expressed as: ( 18 ) ( 19 ) where D sc and S sc are the D and S of a semicrystalline rubbery polymer , respectively ; D a and S a are the D and S of the same polymer in the amorphous phase , respectively; and X c is the degree of crystallinity of the polymer , which can be expressed in terms of mass or volume fraction s of the polymer that is crystalline . Based on the relationship P = DS ( Eq. 11 ) , P for semicrystalline rubbery polymers [137] can be expressed as: ( 20 ) For polymers such as PLA, PET and poly(ethylene naphthalate) (PEN), Eq. 20 may not be applicable w h en the mass transfer measurement is carried out below their T g . The two - phase model is no longer applicable when the semicrystalline polymers are comprised of three phas es (i.e., CF, RAF and MAF ). Theoretically w ith the three - phase model, D sc and S sc may be expressed as: 38 ( 21 ) ( 22 ) where D M AF , S M AF and X M AF are the D , S and the mass or volume fraction of the MAF, respectively; D R AF , S R AF and X R AF are the D , S and the mass or volume fraction of the R AF, respectively. The nature of the permeant also plays an important role in mass transfer. For a series of chemically similar permeants, an increase in the size of a permeant generally results in an increase in S and a decrease in D [138] . Likewise , the polarity of the polymer and the permeant and their affinity affect the extent to which the permeant dissolves in the polymer. As for the environment al effects , t he Arrhenius relationship can be used to describe the temperature dependence of mass transport properties [136] : ( 23 ) ( 24 ) ( 25 ) where P o , D o and S o are the pre - exponential factors for P , D , and S , respectively. E P is the activation energy of permeation, E D is the activation energy of diffusion , and S is the heat of sorption. The Arrhenius relationship is applicable both below and above T g , but as segmental chain movements are dependent on T g due to the change of FV [139] (see Figure 2 - 10 ), the relationship is not applicable across the glass - rubber transition [81,140] . Overall, E P can be expressed as the sum of E D and S (Eq. 26 ) . Furthermore, S can be expressed as the sum of C , the heat of condensation , and 39 M , the heat of mixing (Eq. 27 ): ( 26 ) ( 27 ) Additional factors such as m oisture in the environment c an also play a role in the mass transfer . For example, moisture can swell the polymer and/or act as a plasticizer, resulting in an increased flexibility in segmental chain movement of the polymer and leading to an increase in D . On the other hand, water molecules can cluster on the polymer surface and impede sorption of the permeant [141] . Figure 2 - 10 Specific volume diagram of a polymer as a function of temperature ( T ) , adapted from [142] . A n increase in free volume is observed above the glass transition temperature ( T g ). 40 2.5 Mass transfer of gases in PLA films The P , D , and S of pure gases such as O 2 , CO 2 , N 2 , and CH 4 in PLA films are discussed in this section. W e begin by pre senting an overall summary of PLA mass transfer properties, followed by the specific details of P , D , and S for a set of gases and vapors . Various factors affecting mass transfer properties are discussed in the respective sub - sections. The e ffects of PLA modifications on mass transfer properties are discussed, followed by highlighting data gaps and recommendations for future research. Since P , D , and S for O 2 have been extensively reported, we provide a dedicated section on mass transfer of this specific permanent gas in Section 2.6 . An arbitrary T g of 58 C is used throughout the review when discussing temperature ranges below and above T g . T o compare results from different sources, reported P , D , and S values were converted to the same S.I. units, i.e., kg.m.m 2 .s 1 .Pa 1 for P , m 2 .s 1 for D , and kg.m 3 .Pa 1 for S . Measurements with units that could not be converted into these specific units are discussed separately or were excluded. In summary, under dry conditions, the P values follow the Arrhenius relationship with temperature, with a discontinu ous trend below and above T g . However, limited availability and large dispersion of data for D and S values made it difficult to establish definite trends for the effects of tempe rature. The P values show an increasing trend as the molecular weights of the gases increase, but a decreasing trend as the kinetic diameters of the gases increase. The D values have a decreasing trend as the critical volumes of the gases increase, while t he S values have an increasing trend as the critical temperatures of the gases increase. 41 2.5.1 Permeability A n Arrhenius plot of P (based on Eq. 23 ) of selected pure gases in neat PLA films at 0% RH (i.e., dry conditions) is shown in Figure 2 - 11 . Overall P for O 2 , CO 2 , N 2 , H 2 , He and CH 4 gases are lower than 7×10 17 kg.m.m 2 .s 1 .Pa 1 below T g . Above T g , the overall P of these gases are lower than 3 ×10 16 kg.m.m 2 .s 1 .Pa 1 . The magnitude of P of pure gases at 0% RH observed in Figure 2 - 11 follows this trend: CO 2 > He > H 2 > O 2 > N 2 > CH 4 below T g , with the trend of CO 2 > O 2 > H 2 > N 2 above T g . Arrhenius relationships were observed for all the gases and the changes in slopes of the linear regression lines below and above T g implies discontinu ous barrier properties of PLA across the T g range. C ompared to our reported trend in P of pure gases, Sawada et al. [143] and Komatsuka and Nagai [144] reported the trend being H 2 > CO 2 > O 2 > N 2 > CH 4 , with unit of P in cm 3 (STP).cm.cm 2 .s 1 .cmHg 1 . After converting their values to kg.m.m 2 .s 1 .Pa 1 , which expressed P in mass instead of volume, a similar trend in all three studies was observed. Lehermeier et al. [145] reported P values of O 2 , CO 2 , N 2 and CH 4 that are 1 2 orders of magnitude higher than the average reported P values in this review . They initially attributed this to the processing method. However, the same group of authors reevaluated permeability of O 2 , CO 2 and N 2 and reported new values and determined that the prev ious measurements were out of range due to the measurement method and possible presence of pinhole s in very thin films [83] . Therefore , values from Lehermeier et al. [145] are not included in the plot. Furthermore, P values of ClO 2 ( P ClO 2 ) at 50% RH and different temperatures are available [146] (data not shown) and the values are higher than those of the gases discussed in Figure 2 - 11 . 42 ( P ClO 2 will be discussed further in the next section ) . The test conditions for the measurements of P values o f these gases and the corresponding reference s are summarized in Table 2 - 3 . Figure 2 - 11 Arrhenius plot of permeability coefficients ( P ) for O 2 , CO 2 , N 2 , H 2 , He and CH 4 at dry conditions . Data references: O 2 [147] , [144] , [133] , * [58,59, 83,143, 1 48 188] . CO 2 [63] , [144] , [172] , [58,83,143,156,173,176,177,184 186 ,189 ] . N 2 [144] , [143] , [83,156,167,179,190] . H 2 x [143,144] . He [133,155,161,165,191] . CH 4 + [143] . The vertical dash - dotted line is an arbitrary T g of 58 °C and the dashed lines are from linear regressions of reported experimental data below and above T g . 43 Table 2 - 3 Test conditions ( temperature and RH) for permeability coefficient ( P ) measurements of neat PLA films for selected pure gases from the literature data used in this review . Gas Temperature , °C RH , % References O 2 5 85 0 100 [58,59, 63,83, 1 43 1 45 ,1 47 1 88 , 192 224 ] CO 2 7 85 0 75 [58,63, 83,143,144,156,172,173,176,177, 184 186,189,192,200,204,216] N 2 22 85 0 50 [83,143,144,156,167,179,190,216] H 2 35 85 0 [143,144] CH 4 0 50 0 [143,145] He 20 35 0 [133,143,155,161,163,165,168,191] ClO 2 23 40 50 [146] As shown in Table 2 - 3 , i t is apparent that for most gases, P values were measured only at some test temperature s and RH ranges . This could be due to difficulties in the experiment setups or the limitations of the detector to detect a small amount of the permeated gas . Nevertheless, these missing experimental conditions should be explored to better understand the factors affecting P of gases in PLA. A plot of P versus molecular weight ( M w ) of gases , w ith the y - axis on a logarithmic scale, is shown in Figure 2 - 12 a . To reduce environmental effects such as from temperature and moisture, data were selected from measurements with similar test conditions , 0% RH and 20 35 °C, except for P ClO 2 which w as measured at 50% RH. From Figure 2 - 12 a , low molecular weight gases such as He and H 2 have unusually high permeability compared to other gases that disp lay an increasing trend in P as their 44 molecular weight increases. Plotting average P against kinetic diameters ( Figure 2 - 12 b ) shows unusually low P for He and H 2 , and a linear decreasing trend as kinetic diameters increase for other gases. ClO 2 is not included in Figure 2 - 12 b because the kinetic diameter for ClO 2 is not available. Some pioneer studies [225,226] reported decreasing diffusivities as the molecular weights of gases and vapors increased. For He and H 2 , their high P values in P versus M w plot are indicative of a dominant diffusion effect due to higher diffusivity of the small gas molecules. On the other hand, their low P values in P versus kinetic diameter can be attributed to their low so lubility in the polymer. FV void sizes in PLA are reported to be 86.5 Å 3 in a 60 - min annealed sample and 98.7 Å 3 in an as - quenched sample [134] . Accordingly, the diameter of the FV voids in PLA would range from 5.5 to 5.7 Å. H e and H 2 , with kinetic diameters of 2.6 and 2.9 Å, respectively, should go through the FV voids more easily than other gases with larger kinetic diameters. However, CO 2 , with a kinetic diameter of 3.3 Å, has higher P than He and H 2 despite its larger size. This may imply dominant solubility effect as the kinetic diameters of gases increase. In low barrier polymers with large free volume, such as poly(trimethylsilyl)propy ne, the sizes of these gases (He, H 2 , CO 2 , O 2 , N 2 , CH 4 ) have been reported to directly affect P values [227] . However, the relationship of P and the sizes or molecular weights of the gases should not be concluded without the knowledge of D and S . 45 Figure 2 - 12 Plot of average permeability coefficient ( P ) values , with standard error bars , as a function of a) molecular weight s ( M w ) and b) kinetic diameters of different gases for neat PLA at 20 35 °C and 0% RH [5 8,59, 63, 83,133,137,143,144,147, 1 49 1 91 , 211 ] , except for P ClO2 which were measured at 50% RH [146] . Each dashed line is a linear trend line, which excludes H 2 and He . T he coefficient s of determination (r 2 ) for the linear trend lines are a) 0.9261 and b) 0.9730 . 46 Effect of temperature: Table 2 - 4 shows E P , E D , and S values for selected gases in PLA as reported in individual publication s, as well as the values estimated from linear regressions of the data from Figure 2 - 11 . ( O 2 will be discussed separately in S ection 2.6 ) . S e veral authors [63,83,144,172] examined P CO 2 of PLA films at different temperature ranges and observed an Arrhenius relationship. Auras et al. [63] reported activation energy of permeation of CO 2 ( E P , CO 2 ) for PLA films with 98% L - lactide and PLA films with 94% L - lactide as 15.6 and 19.4 kJ.mol 1 , respectively. Bao et al. [83] reported E P , CO 2 for PLA films with 98 .7 % , 80%, and 50% L - lactide as 18.5, 17.8, and 14.3 kJ.mol 1 , respectively. Komatsuka and Nagai [144] measured P CO 2 at a temperature range below and above T g (35 85 C ) and did not find any evidence of discontinuity across the T g range for films from 96% L - lactide and blends of 96% L - lactide and 88% L - lactide. They reported E P,CO 2 as 48.9 kJ.mol 1 for 96% L - lactide and 41.5 kJ.mol 1 for the blends. E P,CO 2 estimated from data reported by Sansone et al. [172] for neat PLA and neat PLA after a high - pressure pasteuriz ation process between 33 and 48 C were 27.9 and 21.0 kJ.mol 1 , respectively. Linear regressions of Arrhenius relationship for P CO 2 data collected in this review, as shown in Figure 2 - 11 , resulted in E P , CO 2 of 21.6±7.0 kJ.mol 1 below T g and 47.9±13.5 kJ.mol 1 above T g . P values of N 2 ( P N 2 ) also show an Arrhenius relationship [144,190] . Komatsuka and Nagai [144] reported the activation energy of permeation of N 2 ( E P,N 2 ) at dry condition s between 35 and 58 C for PLA 96% L - lactide and blends of 96% L - lactide and 88% L - lactide as 59.0 and 52.8 kJ.mol 1 , respectively. E P,N 2 estimated from P N 2 data from Sato et al. [190] for PLA 4032D from 25 to 45 C is 28.4 kJ.mol 1 . E P,N 2 from data in 47 this review are 31.5±11.6 kJ.mol 1 below T g and 57.0±12.7 kJ.mol 1 above T g . P values of H 2 ( P H 2 ) were also reported at dry condition s by Komatsuka and Nagai [144] with activation energy of permeation of H 2 ( E P,H 2 ) between 35 and 58 C for PLA 96% L - lactide and blends of 96% L - lactide and 88% L - lactide of 33.5 and 27.0 kJ.mol 1 , respectively. From data in this review, E P,H 2 values are 30.7±8.5 and 27.3±10.5 kJ.mol 1 at temperatures below and above T g , respectively. Lehermeier et al. [145] studied the temperature dependence of P CH 4 in 100% linear PLA with L : D ratio of 96:4 and obtained activation energy of permeation of CH 4 ( E P, C H 4 ) of 13.0 kJ.mol 1 at a temperature range from 0 to 50 °C, where P increased from 2.73 × 10 18 to 6.38 × 10 18 kg.m.m 2 .s 1 .Pa 1 . Netramai et al. [146] reported activation energy of permeation of ClO 2 ( E P, ClO 2 ) at 50% RH as 129.03 kJ.mol 1 , where P ClO 2 increased from 5.40 × 10 17 kg.m.m 2 .s 1 .Pa 1 at 23 °C to 9.44 × 10 16 kg.m.m 2 .s 1 .Pa 1 , which is 17 times higher, at 40 °C . To date, there is no research on temperature dependence of PLA barrier propert ies for He. T he reported E P values vary and do not exhibit the same trends, which could be due to different PLA sources, processing methods or treatments . However, knowing an approximate value of E P will help to determine an acceptable temperature range for PLA applications. Furthermore , even though Komatsuka and Nagai [144] reported no transition at the T g region, plotting their data together with d ata from other authors may suggest otherwise. For example, Figure 2 - 11 shows noticeable transitions for P of gases below and above T g , which is to be expected. Higher values of E P for CO 2 , O 2 , and N 2 above T g indicates that at higher temperatures the thermal effect on permeability is highe r, so the change in permeation values is high er . However, the same trend was 48 not found for H 2 . E P values for He ha ve not been reported in the literature although similar behavior w ould be expected . 49 Table 2 - 4 Activation energy for permeation ( E P ) , activation energy of diffusion ( E D ) , and heat of sorption ( S ) for select ed gases at 0% RH, except for ClO 2 at 50% RH. Gas Temperature range, C PLA E P, kJ.mol 1 E D, kJ.mol 1 S , kJ.mol 1 Ref. CO 2 25 45 98% L (4030D a ) 15.6 n/a n/a [63] 25 45 94% L (4040D a ) 19.4 n/a n/a [63] 23 45 98.7% L 18. 5 36.3 18.4 [83] 23 45 80% L 17.8 32.2 13.9 [83] 23 45 50% L 14.3 34.8 25.4 [83] 35 85 96% L 48.9 n/a n/a [144] 35 85 96%:88% L blends 41.5 n/a n/a [144] 33 48 Biophan b 27.9 n/a n/a [172] 33 48 Biophan b after high - pressure 21.0 n/a n/a [172] 20 40 80% L n/a n/a 21.88 [193] 10 40 98% L n/a n/a 23.14 [ 228] 30 50 80% L n/a n/a 22.22 [229] 30 50 98% L n/a n/a 21.58 [229] 5 58 (various) 21.6±7.0 4.6±13.4 22.4±1.4 This review 59 90 (various) 47.9±13.5 n/a n/a This review N 2 35 85 96% L 59.0 n/a n/a [144] 35 85 96%:88% L blends 52.8 n/a n/a [144] 25 45 98% L (4032D a ) 28.4 n/a n/a [190] 23 45 98.7% L 34.6 59.3 25.0 [83] 23 45 80% L 40.9 n/a n/a [83] 23 45 50% L 35.0 n/a n/a [83] 5 58 (various) 31.5±11.6 n/a n/a This review 59 90 (various) 57.0±12.7 n/a n/a This review H 2 35 85 96% L 33.5 n/a n/a [144] 35 85 96%:88% L blends 27.0 n/a n/a [144] 5 58 (various) 30.7±8.5 n/a n/a This review 59 90 (various) 27.3±10.5 n/a n/a This review CH 4 0 50 96% L 13.0 n/a n/a [145] ClO 2 23 40 EVLON c 129.0 n/a n/a [146] n/a: not available, a PLA from NatureWorks LLC , b PLA from Treofan , Germany , c PLA from BI - AX International Inc. , Canada . 50 Effect of relative humidity: Other than the information provided for P values of O 2 ( P O 2 ) as a function of RH, none of the aut hors reported the effect of RH on P of other pure gases. So, it is not possible to determine the effect of RH on P of pure gases from the available data. For N 2 , Samuel et al. [216] reported P N 2 of PLA 4032D at 50% RH and 22 °C of 7.23 × 10 2 4 kg.m.m 2 .s 1 .Pa 1 . H owever, compared with P N 2 of the same type of PLA tested by other authors [167,190] at dry condition s and similar temperature (25 °C) , which average s 3.22 × 10 19 kg.m.m 2 .s 1 .Pa 1 , PLA appears to have much better barrier to N 2 at 50% RH than at dry condition s , which does not seem possible . This could be attributed to different methods o f measurement and sample preparation . A controlled experiment, which varies only the RH where oth er factors are kept constant, is required to verify the effect of RH on P of gases. Effect of crystallinity and L : D ratio : Sawada et al. [143] reported that P H 2 , P CO 2 , P O 2 , P N 2 , and P CH 4 for PLA increased with crystallin i ty from 0 to 9% X c and then decreased from 9 to 40% X c at 35 ° C and 0% RH for PLA with 96% L . Ortenzi et al. [186] reported a slight decrease in P CO 2 as X c increase d from 1.7% to 10.3% at 23 °C , 0% RH. Komatsuka and Nagai [144] studied P CO 2 at 35 85 °C and 0% RH and reported no significant effect of X c on P CO 2 between 35 and 55 °C , but an increase in P CO 2 as X c increased from 7.4 to 24.8% between 65 and 85 °C . Colomines et al. [155] observed no effect of X c on P He at 23 ° C and 0% RH for PLA with 2 39% X c . On the other hand, Guinault et al. [161] reported a decrease in P He at 23 ° C and 0% RH for PLA of 98% L - lactide with 2 40% X c , but an increase in P He from 2 to 40% X c and then a decreas e from 40 to 60% X c at the same test condition s for PLA with 99% L - lactide . Courgneau et al. [168] reported 51 decreasing P He at 23 ° C and 40 60% RH for PLA with 3 43% X c . Guinault et al. [133] tested PLA with different degree s of crystallinity at 23 ° C and 0% RH and found no effect of crystallinity on P He at low crystallinity (2 40% X c for PLA with 99% L - lactide and 1 20% X c for PLA with 96% L - lactide ) . H owever , they reported decreasing P He at higher % X c (50 63% X c for PLA with 99% L - lactide and 30 44% X c for PLA with 96% L - lactide ). These findings on the effect of crystallinity are summarized in Table 2 - 5 . 52 Table 2 - 5 A summary of changes in permeability coefficients ( P ) of gases in PLA as degree of crystallinity ( X c ) increases. PLA Gas(es) Change in P as X c increase d T, °C RH, % X c , % Ref. 96% L H 2 , CO 2 , O 2 , N 2 , CH 4 increase d 3 5 0 0 9 [143] 96% L H 2 , CO 2 , O 2 , N 2 , CH 4 decrease d 3 5 0 9 40 [143] n/a CO 2 decrease d 23 0 2 10 [186] 96% L CO 2 no change 35 55 0 7 25 [144] 96% L CO 2 increase d 65 85 0 7 25 [144] n/a He no change 23 0 2 39 [155] 98% L He decrease d 23 0 2 40 [161] 99% L He increased 23 0 2 40 [161] 99% L He de crease d 23 0 40 60 [161] 92% L He decrease d 23 40 60 3 43 [168] 96% L He no change 23 0 1 20 [133] 99% L He no change 23 0 2 4 0 [133] 96% L He decrease d 23 0 30 44 [133] 99% L He decrease d 23 0 50 63 [133] n/a: not available As for the effect of L : D ratio, Auras et al. [ 63] and Bao et al. [83] reported P CO 2 at 30 °C and 0% RH for PLA with different percentages of L - lactide . Bao et al. [83] show a decrease in P CO 2 when L - lactide increases from 50 to 80% and an increase in P CO 2 as L - lactide increases from 80 to 98.7%. Similarly, Auras et al. [63] report an increase in P CO 2 as L - lactide increases from 94% to 98%. These trends also apply to the effect of 53 crystallinity since for these samples, % X c increases as L - lactide increases with exceptions for PLA 50% L - lactide and 80% L - lactide as both of them have 0% X c . Auras et al. [63] reported a slight increase in P CO 2 at 25 °C , 0% RH as L - lactide goes up from 94 to 98 % ( X c goes up from 25 to 40 % ) . These unusual changes could be explained by the three - phase model discussed in more detail in the sectio n on mass transfer of O 2 (Section 2.6 ). 2.5.2 Diffusion Figure 2 - 13 shows an Arrhenius plot of D (based on Eq. 24 ) for O 2 , CO 2 , CH 4 , and N 2 in PLA at dry condition s . Overall, D of these gases at 0% RH are l ess than 1 × 10 11 m 2 .s 1 below T g . To date, there are no reports on D of these gases above T g . A large dispersion of the data may be due to different methods of testing and film processing as well as film defects . Plotting the average D values of gases against the critical volumes ( V c ) of the gases ( Figure 2 - 14 ) shows that f or gases that do not interact with PLA, the size of the permeant plays an important role in the diffusion proces s , with smaller permeants diffusing faster as expected . While the linear regressions (shown as dashed lines in Figure 2 - 13 ) suggest an increase in D values of O 2 ( D O 2 ) as temperature goes up , individual data such as those reported by Auras et al. [148] did not show the same trend. For the other gases, i.e., N 2 and CH 4 , the test temperatures at 0% RH were not sufficient to estimate an Arrhenius relationship. Figure 2 - 14 shows a plot of average D values at similar test conditions (0% RH and 20 35 °C ) from our review versus V c of the gases. The y - axis ( D values ) is in a logarithmic scale . A trend of decreasing D as V c increases is observed , which agrees 54 with results reported by Sawada et al. [143] . This trend follows an assumption that a permeant with a larger size gen erally has a lower D . S emi - empirical approach es to predict D value s of chemicals in and through polymers using different scaling laws have been summarized by Fang and Vitrac [17] . D v alue of ClO 2 ( D ClO 2 ) was reported as 2.86±0.18 × 10 14 m 2 .s 1 at 23 °C and 50% RH [146] , which is lower than most of the reported D values for other gases, except some of those for O 2 . Since the V c value of ClO 2 is not available, D ClO 2 is not plotted in Figure 2 - 14 . Figure 2 - 13 Arrhenius plot of d iffusion coefficients ( D ) for O 2 , CO 2 , N 2 , and CH 4 . Data references: O 2 [148] , [83] , [143] , * [133] , [163,165,166,168,169,176,180,205] . CO 2 [83,143,176,205] . N 2 [83,143,205] . CH 4 + [143] . Each dashed line represents a least squares linear regression of eac h gas from the reported experimental data. 55 Figure 2 - 14 Plot of average diffusion coefficient ( D ) values , with standard error bars , as a function of critical volume ( V c ) of different gases for neat PLA at 20 35 °C and 0% RH [83,133, 143,148,163, 165,166, 168,169,176 , 180,205] . The dashed line is a linea r trend line for ln( D ) versus V c . The coefficient of determination (r 2 ) is 0.8286. Effect of temperature: T he trend line for CO 2 from the Arrhenius relationship (Eq. 24 ) in Figure 2 - 13 yield s the activation energy of diffusion of CO 2 ( E D , CO 2 ) of 4.6±13.4 kJ.mol - 1 which contradicts the higher E D , CO 2 of 32 36 kJ.mol - 1 reported by Bao et al. [83] (as shown in Table 2 - 4 ). However, D values of CO 2 ( D CO 2 ) at each temperature were reported by d ifferent researchers [83,143,205] using different sources of PLA films resulted in a large variability in the estimated E D , CO 2 value . As a result, the estimated E D , CO 2 in this review may not represent the expected value of E D , CO 2 . Studies to determin e the effect of temperature on D CO 2 by controlling other intrinsic and extrinsic 56 factors are suggested. Effect of relative humidity : The effect of RH on D of gases ha s not been explored , indicating large data gaps in the measurements of D . Effect of crystallinity and L : D ratio : Bao et al. [83] reported a decrease in D CO 2 at 30 °C and 0% RH when L - lactide increases from 50 to 80%, and an increase in D CO 2 at the same test condition s when L - lactide increases from 80 to 98%. Sawada et al. [143] and Komatsuka et al. [ 205] reported D CO 2 at 35 °C , 0% RH and the data show an increasing trend from 0% to 20% X c and a decreasing trend from 20% to 40% X c . Explanation of this behavior is later provided in the o xygen section (Section 2.6 ) . 2.5.3 Solubility F igure 2 - 15 shows an Arrhenius plot of S (based on Eq. 25 ) for O 2 , CO 2 , N 2 , and CH 4 in PLA films at dry condition s . Overall, S of these gases is lower than 4. 9 × 10 4 kg.m 3 .Pa 1 below T g . To date, there are no reports of S of these gases above T g . A large dispersion of the data may be due to different methods of testing and film processing as well as film defects. Sawada et al. [143] reported that S follows the decreasing trend of CO 2 > CH 4 > O 2 > N 2 , in line with the gas critical temperature ( T c ) . Our data show different trends in Figure 2 - 16 . However, as discussed in the permeability section, on ce the unit of S is expressed in mass ( kg.m 3 .Pa 1 ) instead of volume ( cm 3 (STP). cm 3 .cmHg 1 ), the trends become similar. S values of ClO 2 ( S ClO 2 ) was reported as 1.90±0.15 × 10 3 57 kg.m 3 .Pa 1 at 23 °C and 50% RH [146] , which is higher than the reported S of other gases. F igure 2 - 15 Arrhenius plot of s olubility coefficients ( S ) for O 2 , CO 2 , N 2 , and CH 4 . Data references: O 2 [193] , [133] , * [83,143,148,166,168,176,205] . CO 2 [193] , [230] , [228] , [83,143,176,205] . N 2 [193] , [83,143,205,231] . CH 4 + [143] . Each dashed line represents a least squares linear regression of each gas from the reported experimental data. 58 Figure 2 - 16 Plot of average solubility coefficient ( S ) values , with standard error bars , as a function of critical temperature ( T c ) of different gases for neat PLA at 20 35 °C and 0% RH [83,133, 143, 166,168,176,193,205,228,230 ,2 31 ] , except for S ClO 2 which w as measured at 50% RH [146] . The dashed line is a linear trend line for ln( S ) versus T c . The coefficient of determination (r 2 ) is 0.8236. Effect of temperature: As reported in Table 2 - 4 , at 0% RH S of CO 2 estimated from sorption data from Oliveira et al. [193] wa s 21.88 kJ.mol 1 for PLA 80% L with 0% X c at 20 40 °C . Similarly, using sorption data from Oliveira et al. [ 228] , S of CO 2 for PLA 98% L and 20 % X c wa s 23.14 kJ.mol 1 at 10 40 °C . Moreover, other experimental data from the same group of authors [229,230] at 30 50 °C , 0% RH were used to estimate S of CO 2 for PL A 98% L , 20% X c and PLA 80% L , 0% X c , which yielded S values of 21.58 kJ mol 1 and 22.22 kJ mol 1 , respectively. Bao et al. [83] 59 reported S of CO 2 for PLA films with different L - lactide co ntents at 23 45 °C and 0% RH . They reported S values for PLA with 50, 80, and 98.7% L 1 , respectively. To date, the only available S data for PLA are for CO 2 below T g 1 . Data on the effect of temperature on S for other gases below T g , as well as for all gases above T g are still lacking. Effect of relative humidity: No reports of e xperiments on the effect of RH on S were found . At high RH , water vapor from the environment could fil l the FV in PLA and th us reduce the available space for gases to solubilize. Therefore, additional research is needed to further address the effect of RH . Effect of crystallinity and L : D ratio: Data from selected authors [83,193,229,230] suggest that S increases as % L - lactide increases and decreases as crystallinity increases. While the result is contradictory since generally crystallinity increases as L - lactide increases , this can be attribute d to different processing techniques and the presence of RAF as will be discussed in Section 2.6 . 2.5.4 Effect of modification Figures 2 - 17 a), b), and c) show the effects o f PLA film modifications for P CO 2 [173,186] , P N 2 [156,167,179,216,232] , and P He [163,165,168,191] , respectively. The negative % change valu es indicate increased barrier compared to the original unmodified PLA. PLA with nanocomposites and blends showed a decrease in P values while PLA with 60 additives show ed varying results depend ing on the type and the concentration of the additives. Ortenzi et al. [186] stud ied the effects of nanoparticle shape and surface modification on crystallinity , and gas and vapor barrier properties. They used t wo types of nanoparticles : nano silica (NS) and organically - modified montmorillonite (OMMT) , with amino silane (3 - aminopropyltriethoxysilane) or epoxy silane (3 - glycidoxypropyltrimethoxysilane) added to improve polymer - nanoparticle compatibility . They found that the presence of nanoparticles, especially when modified with silanes, greatly enhanced barrier properties to CO 2 w here P CO 2 was r educed up to 50%. For NS , the authors reported an improvement in gas barrier properties with the addition of silane, especially for epoxy silane . They attributed this r esult to a n enhanced crystallization process with the presence of silane. However, in the case of OMMT, the addition of silane did not improve gas barrier properties. Siracusa et al. [173] studied barrier properties of PLA with various surface treatments ( silicon oxide ( SiOx ) coated, anti - UV coated, and varnished) and found that PLA with surface treatments show ed much better barrier properties to O 2 and CO 2 than unmodified PLA. Several auth ors studied PLA modifications with nanoclays [156] , NS [167] , graphene oxide and graphene nanoplatelets [179] and reported improvement in barrier to N 2 . When using additives such as poly ( ethylene glycol ) (PEG) [163] or acetyl tributyl citrate (ATBC) [163,168] , the results varied depending on the amount of additives added. Blending PLA with poly(3 - hydroxybutyrate - co - 3 - hydroxyvalerate) (PHBV) resulted in a better barrier to He, as compared with neat PLA [191] . Besides the data shown in Figure 2 - 17 , Samuel et al. [216] reported enhanced barrier to N 2 for PLA blended with petro - based poly(methyl methacrylate) (PMMA) and Picard et al. [165] , stud ying the effect of OMMT 61 on PLA crystallization and gas barrier properties , reported a n improvement of barrier to He with the presence of OMMT. It is apparent that the barrier properties of PLA can be improved or tailored as needed, with some limitations. This o pens an opportunity to use PLA in wider applications where barrier properties are crucial. 62 Figure 2 - 17 Effects of PLA film modifications on a) P CO 2 , b) P N2 , and c) P He . References: a [186] , b [173] , c [156] , d [167] , e [179] , f [191] , g [163] , h [168] . The numbers on top and bottom of the bars are P (10 18 kg.m.m 2 .s 1 .Pa 1 ) of neat PLA used in the corresponding experiments. (+) change means increasing P (worse barrier) and ( - ) change means reduction of P (better barrier). Abbreviations: NS = nanosilica, a - Si = amino silane, e - Si = epoxy silane, OMMT = organically - modified mon tmorillonite , C = Cloisite ® , GO = graphene oxide, GNP = graphene nanoplatelets, PMMA = poly(methyl methacrylate) , PHBV = poly(3 - hydroxybutyrate - co - 3 - hydroxyvalerate), PEG = poly ( ethylene glycol ) , ATBC = acetyl tributyl citrate . 63 2.5.5 Data gap s and recommendations Extensive research has been conducted regarding the P of O 2 , CO 2 , N 2 , and H 2 below and above T g . However, He and CH 4 have only been investigated below T g . T here is no clear correlation b etween P and M w , D and V c , or S and T c , but the assumption that M w and V c affect D and S , which results in the change in P , should be investigated . E P values have been reported for O 2 , N 2 , CO 2 , and H 2 below and above T g , but E P values for CH 4 and ClO 2 ha ve only been reported below T g . E P average values were estimated for H 2 , CO 2 , and N 2 below and above T g at 0% RH, and for He below T g at 0% RH. No average value has been reported for CH 4 . The effect of RH on P of the gases has not yet been fully investigated . The effect of crystallinity on P of most gases is not clea r, mostly due to lack of consideration of the three - phase morphology (i.e., CF, MAF, and RAF) as later explained for O 2 . For PLA, a linear relation ship between D and V c could be ten tatively established for gases. However, trend s for the relationship between S and T c could not be established. Different scaling laws between D and molecular sizes of gases [17] , as well as between S , T c , and gas condens ability [231,233 236] , in other polymers have been discuss ed in the literature. Additional work under controlled experimental conditions is needed to fully understand these relationships in PLA. Furthermore , the effect of PLA modifications on P , D , and S of gases is not yet totally understood. In the case of nano particles, m ost researchers attributed the improvement of P to the reduction of D caused by increased tortuosity. However, experimental data showed large variability. Therefore, a unified theory about the effect of particle and nanoparticle size, shape and chemistry is still out of reach. 64 2.6 Mass transfer of oxygen Oxygen barrier properties through PLA films have been extensively studied [58,59, 63,83, 133, 1 43 1 45 ,1 47 1 88 ,1 92 224 ] . O 2 is often used as a probe for understanding the impact of modifications on barrier properties. In the following section, we provide a detailed discussion of P, D , and S for O 2 . Interestingly, some studies of O 2 barrier helped to eluci date the crystallin e structure of PLA [132,133,168,232] . Several additional studies have reported the O 2 barrier properties of PLA [145,211,216,237] , but the ir units or information were incomplete or insufficient for compar ison with other reported measurements in this review. 2.6.1 Permeability P O 2 values in PLA films [58,59,83, 133, 1 43,144 ,1 47 1 67 ,1 69 188 ,1 95,198,200 202,206,207,209,210,213,214,217,218, 221 223,238 ] are summarized in Figure 2 - 18 . Most of the values between 5 and 5 8 C are aggregated around 0.5 0.7 × 10 17 kg.m.m 2 .s 1 .Pa 1 . V alues from Lehermeier et al. [145] are ex cluded in this discussion as previo usly explained . In general, there is considerable variability in the measured P O 2 values. All the P values were reported to law s of diffusion. As expected, an increase rate in P O 2 is observed a bove T g implying a dis continuity in the Arrhenius relationship across glass - rubber transition as discussed in Section 2.4.4 . 65 Figure 2 - 18 Oxygen permeability coefficients ( P O 2 ) of PLA between 5 and 90 °C and 0 and 100% RH. Data references: 0% RH [133] , [147] , [144] , [58,59, 83, 143, 1 48 16 7 ,16 9 1 88 ] . 1 49% RH * [148] , [175,202,210] . 50 79% RH [63] , [222] , [148] , [158,175,188,198,200,201,2 13,214,217,218 ,221 ] . 80 100% RH [148] , + [153,166,195,206,207,209,223,238] . The vertical dashed line is an arbitrary T g of 58 °C. Effect of temperature: An increase in P O 2 with temperature is observed for all the values , following the Arrhenius equation ( Eq. 23 ) . Most values were reported at 0% RH. Auras et al. [63,148] studied the effect of temperature for PLA 94% L and 98% L . They report ed activation energ ies of permeation of O 2 ( E P ,O 2 ) at 70% RH and 25 45 C for PLA 98% L - lactide and PLA 94% L - lactide as 41.4 3.5 kJ.mol 1 and 28.4 2.9 kJ.mol 1 , respectively [63] . In a later work, the same group of authors [148] estimated E P , E D and S of O 2 at 5 40 C and 0 9 0% RH for the same PLA polymers and reported t he a verage E P ,O 2 across all the RH conditions (0, 30, 60, and 90% RH) for PLA 98% L of 66 23.39 1.11 kJ.mol 1 , and for PLA 94% L of 20.46 1.57 kJ.mol 1 . Flodberg et al. [222] measured o xygen barrier properties of PLA at 50% RH and 23, 28, 33, and 38 C, and reported E P ,O 2 of 45.1 kJ.mol 1 . Komatsuka and Nagai [144] studied the effect of temperature above T g between 45 and 85 C in PLA with different amount s of L and D content, and reported E P ,O 2 crossing T g of 47.9 kJ.mol 1 . They reported a linear trend of P across T g for two types of PLA : 96% L PLA h o mopolymer and blends of 96% L and 88% L at 8:2 ratio. However, as observed in Figure 2 - 18 , a change in the slope of ln( P ) versus the reciprocal of temperature is observed at T g for the overall experimental determinations. Table 2 - 6 shows the E P , E D , and S of O 2 in PLA below and above T g at different RH for the aggregated values presented in Figure 2 - 18 , as well as from individual values reported in the literature . 67 Table 2 - 6 Average values of E P , E D , and H S of O 2 in PLA below and above T g estimated from literature data [58,59, 63,83, 1 43 1 45 ,1 47 188,19 2 2 24 ] as presented in this review , as well as values reported in the literature. Temperature range, C RH, % PLA E P , kJ.mol 1 E D , kJ.mol 1 S , kJ.mol 1 Ref. 25 45 70 98% L 41.4 3.5 n/a n/a [63] 25 45 70 94% L 28.4 2.9 n/a n/a [6 3] 5 40 0 90 98% L 23.39 1.11 0.9 6 4.97 16.94 22.65 [148] 5 40 0 90 94% L 20.46 1.57 5.05 28 .0 4 n/a [148] 23 38 50 n/a 45.1 45.2 0.074 [222] 23 30 0 98.7% L 24.0 42.7 19.2 [83] 23 30 0 80 % L 24.9 40.8 15.9 [83] 23 30 0 50 % L 26.6 68.8 42.0 [83] 45 85 0 96% L 47.9 n/a n/a [144] 45 85 0 96%:88% L blends 41.4 n/a n/a [144] 5 58 0 (various) 21.8±6.0 28.2±21.2 26.7±16.3 This review 5 58 1 49 (various) 19.6±9.0 7.8±16.3 12.4±20.4 This review 5 58 50 79 (various) 18.9±8.6 58.0±35.6 40.3±40.8 This review 5 58 80 100 (various) 18.8±4.8 14.8±31.9 4.9±39.9 This review 59 90 0 (various) 44.5±13.7 n/a n/a This review n/a: not available 68 Effect of relative humidity : Although PLA is susceptible to hydrolysis when exposed to moisture for a long period of time, overall P O 2 values ( Figure 2 - 18 ) do not show any correlations with RH , implying that short - term exposure to humidity does not affect P O 2 of PLA . However, this may be due to the variability of the reported P O 2 . Auras et al. [148] studied biaxial ly oriented PLA 94% L and 98% L films at 5, 23, and 40 C at 0, 30, 60, an d 90% RH. The authors did not find any effects of RH below room temperature (23 C). However, they reported a decrease in P O 2 as RH increased for both films at 40 C. Cho et al. [158] measured P O 2 of PLA 94% L at 23 C a nd did not report any changes of P O 2 when RH increased from 0 to 50% RH . Fukuzumi et al. [175] also studied oxygen barrier properties of PLA films at 23 C and 0, 35, 50, and 75% RH, and did not find any significant difference s in P O 2 with RH. Yang et al. [188] studied PLA with 96% L ( PLA 5200D , 2% X c ) at 23 C and also did not find any significant difference s in P O 2 between 0 and 50% RH . Therefore , as long as PLA specimens are not being hydrolyze d , RH seems not to influence P O 2 of PLA at low temperature . However, one paper reported a decrease in P O 2 as RH increased at temperature s higher than 40 C [148] . As a result, further studies on the effects of both temperature and RH are recommended for better understanding of PLA barrier properties. Effect of crystallinity and L : D ratio : PLA with L - lactide higher than 92% is a semicrystalline polymer, the crystallinity of which depends on the processing technique. Researchers used different processing methods such as solvent casting, extrusion, quenching, annealing, blowing, and orientation to produc e PLA films of different 69 crystallinity. Although crystalline regions are generally impermeable to gases, Guinault et al. [161] could not determine whether the recrystallization process of PLA with 99% L and PLA with 98% L had any effect on barrier properties. However, they showed that P O 2 decreased when X c increased more than 40% for PLA with 99% L and monotonically decreased for PLA with 98% L . Byun et al. [215] also produced PLA films with different X c (14, 24 and 46%), and found that films greater than 30% X c showed lower P O 2 . On the other hand, other authors [148,205] reported that PLA with high X c (~40%) showed higher P O2 than PLA with lower X c . Komatsuka et al. [205] attributed this behavior to the size and dis tribution of FV in crystalline PLA membrane . The inconsistent trends reported in the literature imply that the relationship of P O2 and X c of PLA cannot be explained simply based on crystalline domains (Eq. 20 ). Bao et al. [83] found differences in P O 2 for different percentages of L - lactide; the higher the L - lactide content, the higher the P value. However, this observation is also inconsistent with the fact that PLA with higher L - lactide tends to have higher X c . Courgneau et al. [168] showed that P O 2 increased slightly with X c . The authors also did not find a change of the D values of O 2 ( D O 2 ) with X c , and they attributed this unusual behavior to the presence of three phases (i.e., CF, RAF, and MAF ) i n PLA . The reason could be the de - densification of the amorphous phase or the f ormation of the RAF created by PLA crystallization. In another study, the same authors [133] reported that D O 2 increased until 40% X c and then decreased ( Figure 2 - 19 b ) . Although there is a monotonic reduction of the S values of O 2 ( S O 2 ) with an increase of crystallinity, a reduction of P O 2 is not observed due to the compensat ing effect of increasing D O 2 ( Figures 2 - 19 a and d). Similar trends are observed in Figures 2 - 19 d , e , and f which 70 show the average P O 2 , D O 2 , and S O 2 at 20 30 o C, 0% RH from this review . Sato et al. [232] recently quantified the RAF and the crystallinity in PLA samples annealed a t different temperature s , and confirmed that the gas diffusivity and permeability in PLA films depended on both the amount of RAF and X c . The density of the RAF is close to that of lamellar crystals, so the high density and low mobility of the RAF might ha ve restricted the gas diffusion and permeation in the PLA films. However, results from these authors showed that RAF had a higher density than MAF, which contradicted recent findings from other studies [128,132 134] . Therefore, in terpretation of results should proceed with caution. A recent study [132] demonstrated that the formation of RAF in the amorphous phase hindered the relaxation of the polymer chains and therefore increased the FV, thereby providing an accelerated pathway for diffusion. Hence, when studying the mass transfer in semicrystalline PLA, the three - phase model must be considered. 71 Figure 2 - 19 a c) P , D , and S values of O 2 measured at 23 C and 0%RH adapted from Guinault et al. [133] with symbols indicat ing time/temperature conditions of the crystallization treatment for P LA: extruded sample, 85 C , 90 C, 120 C, × 140 C. d f ) P , D , and S values of O 2 (shown with symbol) at 20 35 C and 0% RH reported in this review . To improve the barrier properties of PLA u sing a unique approac h , Bai et al. [239] induced parallel - aligned shish - kebab - like crystals with well - interlocked boundaries i n PLA with 9 8 % L by using a highly active nucleating agent. They found that instead of an increase in X c , the type of crystal structure formed , in this case densely packed 72 nanobrick wall structures, was responsible for a reduction of P O 2 from 7.4 × 10 18 kg.m.m 2 .s 1 .Pa 1 for PLA with 50% X c to 2.7 ×1 0 2 0 kg.m.m 2 .s 1 .Pa 1 for PLA with 0.5 wt% of - tricyclohexyl - 1,3,5 - benzene - tricarboxylamide ( TMC - 238) which had 48% X c . Auras et al. [149] studied how the introduction of recycled PLA a ffected P O 2 and found oriented PLA ( OPLA ) with 40% regrind ha d P O 2 twice as high as P O 2 of virgin OPLA. While the results from these studies show that recycled PLA may not be a good choice for applications that require high barrier properties, there are possibilities to enhance barrier of PLA using various viable methods. 2.6.2 Diffusion Consolidated D O2 data from the literature for PLA films [83,143, 148, 163,165,166, 168,176,180,200 , 203,205,222,223] are summarized in Figure 2 - 20 . The plot shows that D O 2 values below T g range from 1.8 × 1 0 14 to 841 × 1 0 14 m 2 .s 1 , indicating a high variation which could be attributed to different measurement and processing methods used between studies , as well as different crystalline and amorphous domains produced in the test specimens that affect the tortuous paths for O 2 molecules to diffuse through the film . Different fi lm processing methods can create PLA films with different amount s of the RAF . An increase in the number of voids with smaller FV in the RAF through annealing can lead to higher diffusion of O 2 molecules . If the solubility remains unchanged, the higher diff usion will result in higher permeation. However, with the observed high variations in the reported D O2 values, further studies with better - controlled experimental condition s are required. 73 Figure 2 - 20 Oxygen diffusion coefficients ( D O2 ) of PLA between 5 and 40 °C and 0 and 100% RH. Data references: 0% RH [148] , [133] , [83,143,163,165,166,168,169,176,180,205] . 1 49% RH * [148] . 50 79% RH [148] , [222] , [200] . 80 100% RH [148] , + [166,223,238] . Effect of temperature: Several authors [83,148,222] show that plots of ln( D O 2 ) versus the reciprocal of temperature follow the Arrhenius equation (Eq. 24 ). However, the values are greatly different between studies. Auras et al. [148] reported activation energy of diffusion of O 2 ( E D , O 2 ) of 1.0 to 5.0 kJ.mol 1 for PLA 9 8 % L (4031D) and 5.0 to 28.0 kJ.mol 1 for PLA 9 4 % L between 5 and 40 C and 0 to 90% RH. Flodberg et al. [222] reported E D , O 2 of 45.2 kJ.mol 1 from the range of 23 38 at 50% RH. Bao et al. [83] reported D O 2 a t 23 and 30 °C for dry condition s and annealed films at which E D , O 2 was estimated as 42.7 kJ.mol 1 for PLA 98.7% L , 40.8 kJ.mol 1 for PLA 80% L , and 68.8 74 kJ.mol 1 for PLA 50% L . Table 2 - 6 shows the reported E D, O 2 values from the literature, as well as the a verage E D , O 2 values estimated from this review . At 0% RH, the average E D , O 2 below T g from this review is 28.2 ±21.2 kJ.mol 1 . Compared to the reported values, the estimated E D , O 2 from this review is fairly low, but it could be a result of high variability of D O 2 , especially at room temperature (23 C ) where the majority of data on D O 2 were reported. Effect of relative humidity: Auras et al. [148] reported an increase in E D , O 2 as RH increase d , indicating that temperature had stronger effect on D O 2 at higher RH. However, this was not reflected i n the P values due to a relative compensation between D and S . A t 24 C , Sanchez - Garcia et al. [166] reported slightly higher D O 2 at 80% RH than at 0% RH. Effect of crystallinity and L : D ratio : Auras et al. [148] reported D O 2 values for PLA films with 98 % L and 94 % L . The results show ed that at dry condition s , PLA with 98 % L ( 25% X c ) ha d lower D O 2 than PLA with 94 % L (40 % X c ) at 5 C . However, as the temperature increase d to 40 C , D O 2 values for both PLA films we re not much different , with PLA 98 % L show ing slightly higher D O 2 than that of PLA 94 % L . Sawada et al. [143] reported an increase in D O 2 when X c increased from 0 % to 20% and a decrease in D O 2 when X c increased from 20% to 40%. Picard et al. [165] reported a decrease of D O 2 from 14.2 × 1 0 1 3 to 7.96 × 1 0 1 3 m 2 .s 1 at 20 C and 0% RH when PLA film was annealed , 75 which resulted in an increase in X c from 0 to 46%. Courgneau et al. [168] reported D O 2 for non - annealed 92 % L PLA films ( 3 % X c ) and 92 % L PLA films (36 43% X c ) recrystallized at different temperatures and found no difference in D O 2 values for these films. Bao et al. [83] reported that at dry condition s , PLA with 98 % L (~40% X c ) and PLA 80 % L (0% X c ) ha d very similar D O 2 values, while PLA with 50 % L (0% X c ) ha d higher D O 2 . Guinault et al. [133] reported an increase in D O 2 for PLA with 96% and 99% L when X c increased from 2% up to 40% and a decrease in D O 2 a t higher X c , show ing the same behavior as their reported P O 2 . Komatsuka and Nagai [ 144] reported D O 2 at dry condition s for homopolymer PLA 98 % L as higher than D O 2 for blends of 96% L and 88% L at 8:2 ratio. Del Río et al. [134] studied the evolution of FV in crystallized PLA with 98 % L and postulated that upon annealing the PLA , there wa s an increase of FV located inside the RAF. Annealing led to a decrease in t he FV void size s while the quantity of the void s increased . The increase in the number of small voids can lead to higher diffusion of O 2 molecules, which results in high permeation . Fernandes Nassar et al. [132] also found an accelerated pathway for diffusion of small molecules such as O 2 due to the occurrence of RAF and thereby the increase of FV. 2.6.3 Solubility Values of S O 2 in PLA films have been reported [83,143, 148,166,168,176,193,200,203,205 , 222,223] and are shown in Figure 2 - 21 . S O 2 values measured or estimated between 5 and 40 C have been r eported with an 76 average of 4.23± 8.04 × 1 0 5 kg.m 3 .Pa 1 . Values of S O 2 reported in the literature differ by many orders of magnitude , ranging between 2.36 × 1 0 7 and 4.88 × 1 0 4 kg.m 3 .Pa 1 at 5 to 40 C and 0 to 90% RH. These variations could be attributed to differences in PLA sources, processing, m easuring meth ods , and different amount s of induce d RAF . Some authors [143,148,166,168,176,205] estimated S from Eq. 7 , while some authors [222] used Eq . 8 , and others [193] measured S dire ctly using QCM. Therefore, dependable S O 2 values with respect to temperature, RH , and X c for PLA are still lacking . Figure 2 - 21 Oxygen solubility coefficients ( S O 2 ) of PLA between 5 and 40 C and 0 to 90% RH. Data references: 0% RH [148] , [133] , [83,143,166,168,176,193,205] . 1 49% RH * [148] . 50 79% RH [148] , [222] , [200] . 80 100% RH [148] , + [166,223,238] . 77 Effect of temperature: Auras et al. [148] reported S O 2 between 5 and 40 C and 0 to 90% RH. The y reported S between 16.9 and 22.6 kJ.mol 1 . Oliveira et al. [193] reported S O 2 between 20 and 40 C and pressure 0.11 to 0.995 bar and fit the experimental values to the Flory - Huggins model , Eq. 9 . Bao et al. [83] reported S between 15.9 to 42 .0 kJ.mol 1 for PLA with 50 98.7% L . On the other hand, Flodberg et al. [222] estimate d a S value of 0. 1 kJ.mol 1 . Furthermore, using Eq. 8 , they decoupled the contribution of S and indicated that the contribution of C is negligible for O 2 mass transfer. So, the main contributor is m, which is always negative for condensable gases. The linear regression lines of the overall S O 2 data ( Figure 2 - 21 ) did not yield reliable values of S due to very high variability in the data ( Table 2 - 6 ). Furthermore, the reported S values from literature also show high variability , which could be due to differen ces in PLA sources and processing conditions , as well as differen ces in measurement methods. Effect of relative humidity: A linear decrease in S O 2 as RH increased to 90% for PLA with 98% L was reported by Auras et al. [148] . They also reported a decrease in S as RH increased from 0 to 90%. Sanchez - Garcia et al. [166] reported slightly lower S at 80% RH than at 0% RH. How ever, no extensive research ha s been conducted on the effect of RH on S O 2 . Effect of crystallinity and L : D ratio : Auras et al. [148] studied S O 2 in PLA with 98% L ( 40% X c ) and PLA with 94% L (25 % X c ) and reported a decrease in S O 2 as X c increased . 78 Sawada et al. [143] reported a slight reduction of S O 2 as X c increased from 0 to 40%. Komatsuka et al. [205] also observe d a decrease in S O 2 from 1.53 × 1 0 6 to 8.5 × 1 0 7 kg.m 3 .Pa 1 when X c was changed from 7.4 to 24.8%. Courgneau et al. [168] also found a decrease in S O 2 from 2.43 × 1 0 6 to 1.00 × 1 0 6 kg.m 3 .Pa 1 when X c increased from 3 to 43%. So, these few reports seem to indicate that S O 2 should decrease when X c increase s , which is expected for most polymers. 2.6.4 Effect of modification T o overcome PLA poor or medium barrier to O 2 , many researcher s have modified PLA by using a number of additives [162,163,168,169,171,176,194,199,221,240] , blend ing with a number of bio - based and fossil - based polymers [149,151,157,166,182,188,203,212,216,223] , or compound ing with fibers, micro - and nanoparticles [58,59, 150,156, 165, 170,174,175, 1 77 18 0 , 183 186 , 198 ,2 00 2 03 , 206 , 207 , 209 , 210 , 212 , 213 , 217 219 , 223 ] . Figure 2 - 22 shows percentage changes of P O 2 due to such modification s. It is clear that properties. The i ncorporation of additives such as ATBC [168] i n to PLA did not improve PLA O 2 barrier . T he addition of 2 wt% talc to PLA slightly improved O 2 barrier ; however , formulated PLA with 1 wt% talc and 17 wt% ATBC did not improve O 2 barrier and showed 95% increase in P O 2 [168] . Plasticization effects of carboxyl and hydroxyl PLA monomers [182] resulted in higher P O 2 of the blends of PLA and the monomers. However, physical aging of neat PLA as well as the blends (data not shown) was reported to improve PLA barrier to O 2 [182] . Blending PLA with poly(butylene succinate - 79 co - adipate) ( PBSA ) increased P O 2 , but coupling PBSA to crotonic acid functionalized PLA resulted in branched plasticized PLA which significantly reduced P O 2 [188] . Sanchez - Garcia et al [166] reported a reduction in P O 2 at 24 C and 0% RH in PLA/ ethylene vinyl alcohol ( EVOH ) blends . Since gas permeability for the blend films is determined by the volume fractions of the polymer components, and since EVOH is hydrophilic, P O2 will be higher in the blends due to swelling of EVOH when exposed to water . However, at 24 C and 80% RH (data not shown), the authors reported that blending EVOH with PLA did not reduce P O 2 and the reason for no O 2 barrier improvement was due to the interaction of the blends with water . The same researcher s also found t he a ddition of amylopectin ( AP) to PLA/EVOH blends slightly decreased P O 2 at 0% RH, but increased P O 2 at 80% RH. The a ddition of nanoparticles such as montmorillonite (MMT) [186] , modified MMT [153] , OMMT [165] , TiO 2 [213] , NS [186] , NS and silanes [186] re duces P O 2 between 15 to 100%. However, the addition of laurate - intercalated Mg - Al layered double hydroxide (LDH - C 12 ) as a nanofiller increa ses P O 2 , which could be due to PLA degradation as will be discussed in Section 2.7.4 . Regarding nanoparticles, m any authors claimed to obtain an intercalate d or exfoliated structure of the layered silicates; however, most of the structure s were predominantly intercalate d [151,154,208,214,241,242] . 80 Figure 2 - 22 Change of PLA P O 2 due to the introduction of additives, blending or compounding with micro - and nanoparticles . The numbers on top and bottom of the bars are P O 2 (10 1 8 kg.m.m 2 .s 1 .Pa 1 ) of neat PLA used in the corresponding experiment s. (+ ) change means increasing P (worse barrier) and ( - ) change means reduction of P (better barrier). References: a [153] , b [213] , c [214] , d [186] , e [165] , f (annealed PLA) [165] , g [182] , h [188] , i [166] , j [168] . Abbreviations : modified MMT = modified montmorillonite, LDH = layered double hydroxide , NS = nanosilica, NS/aminoSi = nanosilica/amino silane, NS/epoxySi = nanosilica/epoxy silane, MMT = montmorillonite, MMT/aminoSi = montmorillonite/amino silane, MMT/epoxySi = montmorillonite/epoxy silane, OMMT = organically - modified montmorillonite , hydroxyl o - LA = hydroxyl lactic acid oligomer , carboxyl o - LA = carboxyl lactic acid oligomer , PBSA = poly(butylene succinate - co - adipate) , CA - PBSA = crotonic acid functionalized poly(lactic acid) coupling poly(butylene succinate - co - adipate) , EVOH = ethylene vinyl alcohol , E29+50%AP = blends of EVOH 29% and AP (amylopectin) 50%, ATBC = acetyl tributyl citrate . 2.6.5 Data gap s and recommendations Robust data to elucidate the effects of RH on PLA film O 2 barrier properties at different temperatures are lacking. A thorough understanding of the simultaneous mass transfer mechanisms of O 2 and H 2 O in PLA films is needed . Further studies are recommended for O 2 barrier properties of PLA from 0 100% RH and 5 50 C . T here are some studies 81 looking at the effect s of crystallinity on O 2 barrier properties and reporting that as L - lactide content increases so does the crystallinity, but the O 2 barrier properties are weakened . This trend is un expect ed since crystalline regions are not permeable to permeant molecules . An initial explanation of this phenomenon was given based on the formation of a t hree - phase structure ( CF, MAF and RAF) and de - densification of the amorphous phase . By controlling morphology and size of spherulites, Fernandes Nassar et al. [132] properties. Similarly , Bai et al. [239] demonstrated ho w specific crystalline architectures could affect PLA barrier properties. However, the effect of the amorphous phase remains elusive; further studies on the amorphous phase structures of PLA are needed. It seems that although researchers can control the am ount of crystallinity in PLA, they cannot fully control the type of crystal regions formed , which seem to have a substantial effect on PLA O 2 barrier properties. So, s pecial attention should be given to fully characteriz e the PLA structure to be able to extrapolate the reported results to other conditions. I t is unclear what are the main effects of crystallinity on P O 2 of PLA. It seems that crystallinity level is not the only factor in controlling P O 2 . Further studies should be conducted to understand whe ther the type of crystals and the amount of MAF and RAF play a significant role . Few D O 2 values in PLA have been reported. Additional work is needed to further understand the diffusion behavior . The e ffect s of the following factors still need to be elucidated : a) L and D - LA content ratio ; b) crystallinity; c) temperature ; d ) orientation . Add itional study of S O 2 should be carried out to ful ly understand the amount of oxygen dissolve d in PLA under different environmental conditions. This information is critical for 82 understanding PLA oxidation at high temperature. Most researchers have modeled solubility of O 2 in PLA following ( Eq. 7 ) . Only one study [222] used the Henry - Langmuir sorption approach ( Eq. 8 ) . Further studies should be conducted to fully understand which model better describes the sorption of O 2 in PLA. T he simultaneous solubility of O 2 with different gases should be assessed to elucidate their synergistic or antagonistic effects on the barrier properties of PLA . 2.7 Mass transfer of water vapor Several researchers have studied P , D , and S of water vapor. In general, P and D increase when temperature increases with some exceptions [63] , while S always decreases with increasing temperature. Due to the antagonistic effect of D and S (i.e., increase in D and decrease in S as temperature increases), P is not affected much by temperature. Overall values of P , D and S do not exhibit increasing or decreasing trend when RH increases. However, water vapor diffusion in PLA and its nanocomposites is reported to follow non - Fickian behavior [21,73] . Regarding how crystallinity affects P , D and S values, experimental results f rom different authors appear to be contradictory, attributable to the two different amorphous regions in PLA, i.e., MAF and RAF. However, at the time of the studies, most researchers did not characterize the CF, MAF and RAF of their PLA specimens. There is also a report [243] that in biaxially drawn PLA films, crystallinity degree was not the main factor affecting barrier property to water vapor , but the increase in the tortuous path from the drawing process reduced water diffusivity and thus improved water vapor barrier. Furthermore, water cluster formation, which is another phenomenon that can further affect the water transport process, has 83 been reported by researchers who investigated S of water in PLA [244 246] . Specific details of the studies and findings for each parameter are discussed below. 2.7.1 Permeability Figure 2 - 23 shows Arrhenius plot s of P values for H 2 O ( P H 2 O ) [5 6,58, 62 ,63,65,147, 149,158,163,171 , 172,174,176,181,184,192,195 197,247 260] in a range from 1.1 × 1 0 1 9 to 1.2 × 1 0 1 3 kg.m.m 2 .s 1 .Pa 1 . The m ajority of the values fall between 10 16 and 10 13 kg.m.m 2 .s 1 .Pa 1 . The l arge dispersion of the data may be due to different film processing and measurement methods. Moreover, s ome values of P H2O were estimated from the relationship P = DS where D and S were measured separately. The low P H2O values in the range of 10 18 to 10 16 kg.m.m 2 .s 1 .Pa 1 reported by Gulati [89] are a result of very low D values, which will be discussed later in Section 2.7.2 . 84 Figure 2 - 23 Arrhenius plot of w ater vapor p ermeability coefficients ( P H 2 O ) of PLA between 6 and 50 °C grouped by different relative humidity (RH) ranges . Each dashed line is from linear regression of the data in each RH range group. References: 1 49% RH [172] , [89] , x [153,166,175,176,261 263] . 50 79% RH [56] , * [63] , [42,62, 65,89 , 162,163,172,174,178,182 , 183,185,197,202,213,252,255,256,258,264 268] . 80 100% RH [56] , [147] , [254] , [269] , + [58,89, 149,150, 152,158,163,171,172,175 , 184,186,195,214,222,223,247 25 1 ,253,2 57,259,270 27 6 ] . Effect of temperature: Auras et al. [63] found that P H 2 O decreased as temperatures increased from 10 to 38 °C , which is counterintuitive. However, this observation is in accordance with the three - phase model, wherein the de - densification of the RAF domain in PLA tends to decrease as temperature increase s . This observatio n implies that the FV does not increase to the same extent as the increase in mobility of the 85 chains with increasing temperature . The decrease in diffusion can result i n a reduction in P H 2 O with in this narrow temperature range. The activation energy of permeation of H 2 O ( E P ,H 2 O ) was studied for two PLA films with differ ent X c at temperatures between 10 and 38 °C and the reported E P ,H 2 O val ues for PLA film with 40 and 25 % X c were 9.8 and 10.1 kJ.mol 1 , respectively [63] . Furthermore, Siparsky et al. [56] and Shogren et al. [247] reported E P ,H 2 O values of amorphous and semicrystalline PLA of 5 and 0.1 kJ.mol 1 , respectively . Since condensation is an exothermic process, the value of C is always negative . F or E P to be negative M must be lower in absolute value than C , and E D must be smaller than the absolute value of the sum of M an d C , according to Eqs. 26 and 27 . Table 2 - 7 shows E P , E D , S , C , and M for different PLA s . T he average s are plotted at different humidit y values in Figure 2 - 23 . Low M for groups of data with different RH values may be attributed to variation in materials and measurements. 86 Table 2 - 7 Activation energy of permeation ( E P ) , activation energy of diffusion ( E D ) , heat of sorption ( S ), heat of condensation ( C ), and heat of melting ( M ) for H 2 O . a Value reported from [148] . b Values estimated from Eq . 26 . c Average values of experimental data reviewed in this article and their standard deviations. d Values estimated from experimental data reviewed in this article. PLA sample E P , kJ.mol 1 E D , kJ.mol 1 S , kJ.mol 1 C , kJ.mol 1 ,a M , kJ.mol 1 ,b Ref. 50% L - lactide 30 62 32 42 10 [56] 70% L - lactide 5 24 19 42 23 [56] 90% L - lactide 7 26 19 42 23 [56] 95% L - lactide, M w 149000 14 41 27 42 15 [56] 95% L - lactide , M w 185000 2 37 39 42 3 [56] 100% L - lactide, quenched 3 37 40 42 2 [56] 100% L - lactide , cooled 9 49 40 42 2 [247] 100% L - lactide , annealed 12 53 41 42 1 [56] crystallined PLA 0.1 n/a n/a 42 [247] 88% L - lactide 5 n/a n/a 42 [247] 88% L - lactide 31.4 39.2 [21] PLA - graft ~ 44 [245] PLA at 1 49% RH 22. 1 ±55.6 c 97. 5 ±19.7 c 71. 6 ±5.5 c 42 ~ 30 d This review PLA at 50 79% RH 10. 7 ±21.9 c 67. 2 ±21.1 c 66. 9 ±9.9 c 42 ~ 2 5 d This review PLA at 80 100% RH 5.3 ±11.2 c 70. 6 ±19.6 c 44.5 ±6. 5 c 42 ~ 3 d This review 87 Effect of relative humidity: While polar materials generally have high affinity to water, PLA, being quite polar due to its ester group s , does not show specific trends in P H 2 O with respect to RH ( Figure 2 - 23 ) . Fukuzumi et al. [175] deter mined P H 2 O at 10 40 °C , 20 80% RH and 23 °C , 30 90% RH conditions , respectively . They found that P H 2 O increased when RH increased. Siparsky et al. [56] estimate d the P H 2 O values of several PLA samples at 50 °C, at 50% RH and 90% RH , but the trends were inconsistent. Auras et al. [63] reported that P H 2 O values at 40 90% RH did not change significantly. Effect of crystallinity and L : D ratio: Results reported in the literature on the effect of PLA crystallinity on P H2O have been inconsistent. Some authors [248,254,269] found that P H 2 O of PLA films decreased as X c increased from 0 to 30% and leveled off when X c was higher than 30%, while one study reported no significant change in P H 2 O at low X c , but a rapid decrease when X c reached 39% [277] . A number of res earchers observed an increase in P H 2 O as X c increased [13,56,63] , and yet another study reported that crystallinity had no effect on P H2O in biaxially drawn PLA films [243] . These variations may be due to different processing metho ds or the existence of more than one amorphous phase in PLA. Further details of these studies are discussed below . Si parsky et al. [56] observed a reduction in P H 2 O when X c of semicrystalline PLA samples increase d . H owever, w hen PLA samples were completely amorphous, a change in L : D ratio did not produce a trend for P H 2 O . Auras et al. [63] studied barrier properties of PLA films at different temperatures (10, 20, 30 and 38 °C) and different L - lactide content s . The a uthors found that higher L - lactide content s resulted in PLA films 88 with higher X c and films with 98% L ( 40 % X c ) h ad approximately 5% higher P H 2 O than films with 94% L ( 25 % X c ) at 10 and 20 °C , and 2.5% higher at 30 and 38 °C. On the contrary, Shogren et al. [247] reported a decrease in P H 2 O when X c increased from 0 to 66 % at 6, 25 and 49 °C , and Duan and Thomas [254] found a monotonic reduction of P H 2 O from 0 to 50% X c at 38 °C . At 25 °C , Tsuji et al. [248] reported a reduction in P H 2 O when X c increased from 0 to 30 % ; however, P H 2 O remained constant above 30 % X c . These different findings of P H 2 O as a function of crystallinity may be explained on the basis of the presence of the three phase s (CF, MAF and RAF) i n the tested PLA films. The effects of M w , D - lactide , and X c were studied by Tsuji et al. [248,269] . They found that change s in M w of PL A films in the range of 9 × 1 0 4 5 × 1 0 5 g.mol 1 and D - lactide content s in the range of 0 50% did not have significant effect s on P H 2 O . Siparsky et al. [56] also evaluated the effect of L : D - lactide ratio and did not find a trend. Tsuji and Tsuruno [269] examin ed the effect of crystallinity on P H 2 O at 25 °C, 90% RH of PLA films with different X c . Their films were synthesized by ring - opening polymerization of L - lactide (PLLA) and D - lactide (PDLA), as well as PLLA/PDLA blend films. For all films, P H 2 O decreased when X c increased. P H 2 O of PLLA/PDLA blend films was 14 23% lower than pure PLLA and PDLA in X c range of 0 30%. Amorphous PLLA/PDLA blend films had lower P H 2 O than pure PLLA and PDLA. Also, they found that P H 2 O was reduced by blending PLLA with PDLA even when the films were amorphous. This study found a dependenc e of P H 2 O on X c for blen d films within a range of 0 30 % X c , but not with X c above 30%. P H 2 O o f pure PLLA, PDLA, and blends decreased rapidly with increasing X c in the range of 0 30% and then slowly between 30 and 100%. They attributed this 89 change to the existence of the restricted amorphous region s at high X c . Their concept of the restricted amorphous region is similar to the RAF in the three - phase model; however, they proposed that at high X c the amorphous regions would be composed of solely restricted amorphous regions while at low X c a small amount of free amorphous regions would coexist. Mathematical models also have been applied to crystallinity results. Duan and Thomas [254] studied the permeability of PLA films with 0 50% X c . The plot of P versus X c showed a good fit to a linear trend line with negative slope . However, the trend line predic ted zero permeability when crystallinity reache d about 78 %, which seemed unlikely (i.e., a polymer with less than 100% X c is unlikely to be totally impermeable ) . They claimed that this phenomenon could be due to the presence of RAF at high X c . The same authors [254] used the tortuous path model to predict that permeability reaches 0 at 100% X c , and they claimed that this model provide d the best explanation of the effect of crystallinity on P H 2 O in PLA. Sansone et al. [172] studied the effect of high - pressure pasteurization on permeability of PLA flexible films. They found that PLA films pasteurized at 700 MPa had slightly lower P H 2 O than untreated PLA films at different temperatures (25 and 30 °C) and RH (30, 50 and 90%) , while both films had similar X c (approximately 25 %) . The reduction in P H 2 O was attributed to a decrease in water solubility due to structural changes in pasteurized films. Early stud ies conducted by Siparsky et al. [56] showed that PLA films after quench ing (11% X c ) and after anneal ing ( 46 % X c ) had different P H 2 O values . A t 30, 40 and 50 °C , 50 and 90 % RH, q uenched films had approximately 50% lower P H 2 O than annealed films . The effect of annealing on P H 2 O was also studied by 90 others [147,254,272,277] and t he results showed a monotonic decrease of P H 2 O as X c decreased. Delpouve et al. [243] studied water barrier of PLA films drawn by different drawing modes resulting in different X c . The authors reported no effect of crystallinity on the P H 2 O . However, the macromolecular reorganization caused by the drawing process increased the tortuous path for water diffusivity, resulting in enhanced barrier. 2.7.2 Diffusion Figure 2 - 24 shows Arrhenius plot of the D values of H 2 O ( D H 2 O ) [56,89,222,278,279] , with repor ted v alues of 2.4 × 1 0 1 5 to 6.6 × 1 0 1 1 m 2 .s 1 . Overall results show that D H 2 O increases as temperature increases. Due to large dispersion of the data, the effects of RH are not clear. However, a study of water diffusion in PLA [21] reported no effect of RH on D H 2 O . Results of D H 2 O reported by Gulati [89] are three orders of magnitude lower than results from other authors which may be attributed to the film thickness inconsistency , the possibility of defects in the films , and the method of measurement . Diffusion behavior of water in PLA films is non - Fickian [21,73] . Davis et al. [21] suggeste d that the initial water uptake is diffusion - driven by concentration gradient, while the later stage is controlled by stress relaxation or swelling due to the non - equilibrium state of PLA (glassy state) . 91 Figure 2 - 24 Arrhenius plot of water vapor diffusion coefficients ( D H 2 O ) of PLA between 10 and 50 °C grouped by different relative humidity (RH) ranges. Each dashed line is from linear regression of the data in each RH range group. Refe rences: 1 49% RH [21] , [89] , x [278] . 50 79% RH [56] , * [21] , [89] . 80 100% RH [56] , [21] , [279] , [89] , + [222] . Effect of temperature: Generally , D H 2 O values increase with increasing temperature. Gulati [89] found that D H 2 O increased by approximately 59 % when temperature increase d from 10 to 40 °C at 20% RH. However, at 40% RH no significant change in D H 2 O was found. When PLA was exposed to 60 and 80% RH, D H 2 O de creased by approximately 4 0% when temperature increase d from 10 to 40 °C . The author attributed inconsistencies in the results to the variability in the film thickness and the possibility of 92 structural defects in the films. Siparsky et al. [56] found that D H 2 O de creased with increasing temperature. Activation energies obtained were in a range from 24 to 62 kJ.mol 1 as shown in Table 2 - 7 . They found a large variation from linearity in Arrhenius plots for PLA with 100% L - lactide between 40 and 50 °C compared to 20 and 40 °C. The authors attributed this variation to cluster formation. High condensation of water at lower temperatures increas es the size and number of water formed clusters. M obilization of water during diffusion is thus hampered at the lower temperatures. The average D H 2 O values measured at 25, 35 and 45 °C reported by Davis et al. [21] also increased with increasing temperature. Effect of relative humidity: With respect to RH, overall data do n ot show a specific trend. When Gulati [89] studied the barrier properties of PLA films, no significant change in D H 2 O values was found in the RH range of 20 80% at 10, 23 and 40 °C . Siparsky et al. [56] reported higher D H 2 O values at 90% RH than at 50% RH for PLA films with different L : D ratios at 5 0 °C . Davis et al. [21] reported that while D H 2 O varied with temperature, the values did not vary with RH in the range of 0 85% RH, whic h could be attributed to the low solubility of water in PLA. Effect of crystallinity and L : D ratio: The effect of L - lactide was studied by Siparsky et al. [56] . The authors found that PLA films with 70% L had higher D H 2 O and lower E D ,H 2 O values than those with 50 and 90% L . Drieskens et al. [147] reported a reduction of D H 2 O as crystallinity increased. However, D H 2 O w as not independently measured, but rather was estimated from Eq. 11 . When PLA films were compression mold ed , Siparsky et al. 93 [56] obtained quenched and annealed films with X c of 11 % and 46%, respectively. The values of D H 2 O calculated were approximately 50% higher when PLA was annealed. Drieskens et al. [147] reported that D H 2 O of PLA (96% L ) annealed at 125 °C was higher than at 100 °C . Samples annealed at higher temperature for the same amount of time have higher X c . The presence of higher RAF i n samples with higher X c could be responsible for this result. 2.7.3 Solubility Figure 2 - 25 shows Arrhenius plot s of the S values for H 2 O ( S H 2 O ) [56,89,193,222,228,231] . Values of S H 2 O reported are 4.8 × 1 0 4 to 1.1 × 1 0 1 kg.m 3 .Pa 1 . The values of S H 2 O decrease as temperature increases. The regression lines for different groups of RH shows higher S H2O at lower RH (1 49%) which is counterintuitive, since the water sorption isotherms showed increasing sorbed water at higher water activities [280] . Th is may be attributed to inconsistency in film thickness and possible film defects, as reported by Gulati [89] . 94 Figure 2 - 25 Arrhenius plot of water vapor solubility coefficients ( S H 2 O ) of PLA between 10 and 50 °C grouped by different relative humidity (RH) ranges. Each dashed line is from linear regression of the data in each RH range group. References: 1 49% RH [89] . 50 79% RH [56] , [89] . 80 100% RH [56] , [222] , + [89] . Effect of temperature: Overall , S H 2 O de creased with increasing temperature , which agree s with reported results [56,89] as shown in Figure 2 - 25 . Effect of relative humidity: A s tudy by Gulati [89] showed that when RH increase d from 20 to 80% , S H 2 O decrease d 72, 33 and 54 % at 10, 23 and 40 °C, respectively. Siparsky et al. [56] showed that S H 2 O was relatively constant at intermediate temperatures between 20 and 50 °C regardless of X c , even for samples with a high amount o f L - lactide ( 50 and 70% ) . They attributed these results to the water cluster formation and 95 defined clustering in a polymer such as PLA as the ordered structure of a body of water within the polymer, which is stabilized by hydrogen bonding between the water molecules. Th e cluster i ng d oes not necessarily affect the degree of crystallinity , but it affects the opacity of the film. Davis et al. [280,281] reported non - equilibrium sorption of water in PLA . They found that at water activities ( a w ) less than 0.65 water is mostly pres ent as dimer s in PLA, but when a w is higher than 0.65 a large hydrogen - bonded water cluster formation was observed. Furthermore, they reported that the water sorption and water - induced PLA relaxation in amorphous PLA ( 4032D ) and in PLA homopolymer are taking place at the same time. Effect of crystallinity and L : D ratio: Amorphous PLA at 20 °C showed smaller S H 2 O than semicrystalline PLA [245] . These results can be attributed to higher S of the semicrystalline PLA. 2.7.4 Effect of modification Different modifications have been performed on PLA to improve its properties. Figure 2 - 26 shows the percentage (%) of change i n P H 2 O when PLA is modified by the incorporation of additives, nanoparticles or blends of PLA with other polymers. The work by Duan et al. [272] showed the highest reduction in P H 2 O with 46% change in P H 2 O for nanocomposite of PLA 98% L and Cloisite ® 30B . Generally, a decrease in P H 2 O was observed with the incorporation of nanoparticles [150,272] . However, Katiyar et al. [214] reported that the addition of LDH - C 12 did not improve PLA barrier against the permeation of water vapor. Higher P H 2 O for PLA melt processed with LDH - C 12 could be 96 due to water in LDH - C 12 causing PLA degradation, or the Mg - Al in LDH - C 12 catalyzing the degradation. W hen additives were present in the PLA matrix , P H 2 O increased [ 56,163] . Courgneau et al. [163] found that the additi on of PEG resulted in approximately a 4 7 0% increase in P H 2 O compar ed with neat PLA. In this case, the presence of PEG could promote sorption of water due to the presence of hydrophilic groups providing hydrogen bonding between water and the ether group of PEG, leading to high S H 2 O and P H 2 O values [56] . 97 Figure 2 - 26 % P H 2 O change for PLA films and different modifications. The numbers on top of the bars are P H 2 O (10 16 kg.m.m 2 .s 1 .Pa 1 ) of neat PLA used in the corresponding experiments. (+) change means i ncreasing P H 2 O (worse barrier) and ( - ) change means reduction of P H 2 O (better barrier) . References: a [272] , b [272] , c [214] , d [56] , e [182] , f [163] , g [262] . Abbreviations: C30B = Cloisite ® 30B, LDH = Mg - Al layered double hydroxide, PEG = poly ( ethylene glycol ) , o - LA = lactic acid oligomer, ATBC = acetyl tributyl citrate, PCL = polycaprolactone . 2.7.5 Data gap s and recommendations As previously described for gases, crystallization of PLA can produce CF and RAF. Due to the de - densification of the amorphous region present in the RAF, studies that properly measure RAF , which are currently lacking, should be conducted . Studies in this a rea could help to elucidate the variation of P , D , and S with CF, RAF and MAF. Large variations of these values have been reported, and most of the inconsistenc ies may be attributed to the lack of assessment of the RAF. A comprehensive understanding o f the effect of RH on mass transfer of water vapor through PLA is missing since data 98 presented in this review do not show any trend of P H 2 O with RH. Additional ly , a few authors reported clustering of water i n PLA. Additional studies should be conducted to f ully quantify clustering of water as a function of temperature. 2.8 Mass transfer of organic vapors Barrier properties of PLA to different organic vapors, such as ethylene (C 2 H 4 ), benzaldehyde, ethyl acetate, eucalyptol, and estragol e have been conducted . In general, P and D increase with temperature, but S decreases when temperature increases. It has been reported that t he higher the X c of PLA the lower the P , D , and S to some organic vapors. Interesting results were found in the study of PLA barrier properties to some organic vapors, such as methanol, ethanol and ethyl acetate. The presence of these compounds induces crystallization in PLA in which P and D decrease as X c increases over time. Specific details of the studies and findings of each parameter are discussed below. 2.8.1 Permeability Figure 2 - 27 shows a plot of P of different organic compounds grouped by similar test temperature s versus M w at 0% RH . A general linear reduction of P is observed as M w increased , as expect ed . However, it is difficult to correlate P with M w since different compounds were not measured at the same temperature, and the functional groups are different. We can observe a large variation of P for the same compound as a function of temperature. 99 Figure 2 - 27 P of different organic compounds grouped by similar test temperature s versus M w at 0% RH . References: methanol [190] , ethanol [190] , acetaldehyde [184] , ethyl acetate [137,159] , d - limonene [282] , estragole [283] , eucalyptol [283] . Effect of temperature: Figure 2 - 28 shows P of PLA exposed to different organic vapors at different temperatures. Different authors have studied the effect s of temperature on P of different organic vapors, such as eucalyptol, estragol e , and ethyl acetate, where P increases with temperature. Eucalyptol and estragol e w ere studied by Leelaphiwat et al. [283] . They reported P of eucalyptol and estragol e increased 2.2 times and 1,258 times, respectively, from 15 to 25 °C. Auras et al. [148] observed that P of ethyl acetate depend ed on temperature where P increased with tempe rature from 30 to 45 °C. 100 Figure 2 - 28 Arrheni us plot of p ermeability coefficients ( P ) of organic compounds in PLA at 0% RH . References: d - limonene [282] , ethyl acetate [137,159] , estragole [283] , eucalyptol [283] , ethanol [190] , methanol [190] , acetaldehyde [184] , trans - 2 - hexanal [184] . Effect of crystallinity and L : D ratio: P of several organic vapors have been investigated in PLA films with different X c and L : D lactide ratios. In general, the higher the X c of PLA the lower the P . Important studies have been carried out where organic vapors induce d the crystalli zation of the test ed PLA. Duan et al. [272] studied permeability of methanol and ethanol at 0% RH, at 25, 35 and 45 °C for up to 1440 min. P values of PLA exposed to methanol at 25 °C and ethanol at 25 and 35 °C w ere constant with increasing exposure time. However, for PLA exposed to methanol at 35 and 45 °C and ethanol at 45 °C, P values decreased with increasing exposure time. This was attributed to the presence of ethanol and methanol that induc ed crystallization in PLA, thereby decreasing the P as 101 X c increased. Amorphous f ilms exposed to methanol for 1440 min at 25 , 35 and 45 °C increased X c to 2.8 , 13.1 and 24.7 %, respectively . In the case of ethanol, PLA exposed at 25 , 35 and 45 °C increased X c to 1.2 , 3.5 and 26.7 % , respectively. 2.8.2 Diffusion Figure 2 - 29 shows D of organic compounds grouped by similar test temperature versus their molecular volumes ( V ). As V increases, we should observe a reduction of D since la rge r molecules diffuse more slo w ly . Figure 2 - 29 shows that there is a reduction in D with respect to V at the same temperature. 102 Figure 2 - 29 D of organic compounds grouped by similar test temperature versus their molecular volumes ( V ). References: ethyl acetate [137,159] , ethyl butanoate [160] , ethyl hexanoate [284] , d - limonene [282] , estragole [283] , eucalyptol [283] . Effect of temperature: Figure 2 - 30 shows D of organic compounds in PLA (in logarithmic scale) as a function of reciprocal of absolute temperature . Leelaphiwat et al . [283] found that D of eucalyptol increased with increasing temperature (15 and 25 °C) almost 4 times , and approximately 14 times for estragol e , w ith eucalyptol having higher D values. For ethyl acetate, D also increase d with increasing temperature [13] . 103 Figure 2 - 30 Arrhenius plot of diffusion coefficients ( D ) of organic compounds in PLA at 0% RH . References: d - limonene [282] , ethyl acetate [137,160,168,284] , ethyl butanoate [160] , ethyl hexanoate [160,284] , estragole [283] , eucalyptol [283] , linalool [283] . Effect of crystallinity and L : D ratio: Ethyl acetate is one of the organic vapors being studied that induces crystallization i n PLA. Diffusion of ethyl acetate at 25 °C was investigated by Courgneau et al. [168] in PLA with different treatments: PLA non - annealed (3% X c ) and PLA with recrystallization temperature 120 °C (43% X c ). After contact with ethyl acetate for 2 weeks, the crystallinity increased to 26 and 44% for PLA non - annealed and PLA recrystallized at 120 °C , respectively. D was lower for PLA with high X c , with values of 2.40 × 1 0 1 7 for 3% X c and 1.60 × 1 0 1 7 kg.m.m 2 .s 1 .Pa 1 for 43% X c . 104 2.8.3 Solubility Figure 2 - 31 shows S of organic compounds in PLA (in logarithmic scale) as a function of reciprocal of absolute temperature at 0% RH. Values are largely reported at 25 °C. The highest S among reported data is from est r agol e at 6.82 kg.m 3 .Pa 1 and the lowest S from ethyl acetate at 4.72 × 1 0 7 kg.m 3 .P a 1 . Figure 2 - 31 Arrhenius plot of solubility coefficients ( S ) of organic compounds in PLA at 0% RH . References: 2 - nonanone [285] , benzaldehyde [285] , ethylene [228] , d - limonene [282] , ethyl acetate [137,159,168,284,285] , ethyl butanoate [285] , ethyl hexanoate [284,285] , estragole [283] , eucalyptol [283] . Effect of temperature: Auras et al. [148] found that S of ethyl acetate in PLA decreased as temperature increased from 30 to 40 °C. The same tendency for S of ethyl acetate was found by Oliveira et al. [228] with PLA cooled down from melt (10% X c ) and PLA 105 annealed at temperature slightly above T g (20% X c ) . W hen exposed to C 2 H 4 , S of PLA from melt decreased 43% from 10 to 40 °C and S of annealed PLA decreased by 50% [ 228] . Effect of crystallinity and L : D ratio: Colomines et al. [159] studied S of ethyl acetate in different PLA films: extruded, quenched, recrystallized and commercial PLA Biophan with X c of 2, 6, 39 and 19%, respectively. They found that ethyl aceta te induces crystallization at 0.5 and 0.7 vapor activities , resulting in lower S value . The authors also found that the higher the X c , the lower the S values . Different processing treatments of PLA affect crystallinity of the materials and thus, S of the organic vapor. Recrystallized samples tend to ab sorb a smal l quantity because the sorption occurs only in the amorphous phase of the polymer ; however, no characterization of RAF was reported in these studies . Solubility of ethyl acetate in PLA has been studied by different authors [159,168,284] . Courgneau et al. [168] studied t he solubility in different PLA treatments: PLA non - annealed (3% X c ) and PLA with a recrystallization temperature of 90 (36% X c ) and 120 °C (43% X c ). They found that the higher the X c the lower the S where S decreased 7.6% from 0 to 36% X c . Furthermore, the S of ethyl acetate was investigated by Domenek et al. [284] along with S of ethyl hexanoate in commercial PLA Biophan, extruded , and cast film PLA. The authors found that S values were the same for different types of films. However, S of ethyl hexanoate was lower than S of ethyl acetate due to the hydrophobicity of ethyl hexanoate. Oliveira et al. [ 228] studied S of C 2 H 4 in PLA cooled down from the melt (10% X c ) and PLA annealed at temperature slightly above T g ( 2 0% X c ) . A t 30 °C , S of C 2 H 4 in annealed PLA was approximately 15% larger 106 than in P LA cooled down from melt , which may be explained by different amount s of FV due to different X c . In addition to S of vapors, Tsuji and Sumida [286] studied the effects of organic solvents on the degree of swelling (DS) and other physical properties, based on solubility parameters of PLA and th e solvents. The authors found that the physical properties of PLA films could be modified due to swelling induced by the solvents with different solubility parameter values, as well as the DS. 2.8.4 Effect of modification Several studies have r eported the effect of modification of PLA on organic vapor barrier properties. Table 2 - 8 shows the incorporation of different additives and the effect on D and S of ethyl acetate, ethyl butyrate, and ethyl hexanoate . In all cases, the values increased with respect to neat PLA . T he addition of talc / ATBC appears to significantly affect both D and S of ethyl acetate, resulting in 4170 00 % and 111 00 % increases, respectively . Table 2 - 8 Mass transfer parameter s of organic compounds for modified PLA films , at 25 ° C, 0% RH. Parameter Type Material Modification % C hange × 10 2 References D Ethyl acetate PLA/talc/ATBC non - annealed Additive 4170 [287] D Ethyl butyrate Plasticized PDLLA (200 µ m) Additive 3 9.2 [160] D Ethyl hexanoate Plasticized PDLLA (200 µ m) Additive 20.8 [160] S Ethyl acetate PLA/talc/ATBC non - annealed Additive 111 [287] 107 2.8.5 Data gap s and recommendations Overall very few P , D , an d S values for organic compounds have been reported for PLA. P and D seem to be reduced as M w and V are increased. Systematic studies of organic family compounds such as n - alkenes, n - alcohols, and esters have not yet been reported. S tudies of the effect of RH on P , D , and S of organic vapors are also lacking. Furthermore, plasticization effects of organic vapors on PLA are not well understood. The effect of modifications is not fully investigated. So, increasing effort should be targeted on measuring these parameters. 2.9 Comparisons of barrier properties of PLA to common commercial films Table 2 - 9 shows P of PLA and other common commercial films at 25 °C and Figure 2 - 32 illustrate s comparisons of these values. PLA shows moderate values of P for gases, higher than PET but mostly lower than PE, PS, and PP. However, PLA has very high P H 2 O when compared with other polymers. It is apparent that PLA by itself is not suitable f or applications that require high barrier against gases and water vapor. However, as mentioned before, there are various methods to enhance PLA barrier properties, such as adjusting the L - and D - lactide ratio s in the manufacturing process, coextruding with a high barrier polymer, or adding clay/nanoclay. Thus, with an advantage of being a bio - based and compostable polymer, PLA can be a candidate for applications that require moderate to high barrier , but may require some modifications. 108 Table 2 - 9 Permeability coefficients ( P ) of PLA in k g.m.m - 2 .s - 1 .Pa - 1 to selective gases and vapors and a comparison with other commercial available polymers [81] . All P values were measured at 2 5 °C unless indicated otherwise. P O 2 ×10 20 P CO 2 ×10 20 P N 2 ×10 20 P H 2 O ×10 1 6 PLA a 315 ±150 2811±842 32.2±2.8 1 61±41 LDPE 3100 18600 914 5.51 HDPE 424 5 38 137 2.02 PP 1790 10500 286 3.12 P S , biaxially oriented 2860 1 5500 742 5 7 . 8 6 7 . 5 PAN 5.87 23.6 n/a 3 9 . 5 PV AL 0.07 9.46 18.2 0.1 3 b n/a EVOH , 32% ethylene 0.09 0.36 0.00 5 n/a EVOH , 44% ethylene 0.41 3.14 0.0 4 n/a PVDC 3.59 31.4 0.5 6 0.56 PTFE 4570 149000 1260 0.49 PVC , unplasticized 49 247 11.1 1 6.5 PET, 40% X c 35.9 179 55.7 7.8 PET, amor phous 62 449 n/a n/a PC (Lexan) 1480 1170 271 83 . 5 Nylon 6 31 171 6.57 n/a Cellophane c 9.46 1 05 7 n/a n/a: not available, a data from this review, reported as average value ± standard deviation , b measured at 14 °C , c measured at 76% RH. Polymer name abbreviations: PLA = poly(lactic acid), LDPE = low density polyethylene, HDPE = high density polyethylene, PP = polypropylene, PS = polystyrene, PAN = polyacrylonitrile, PVAL = poly ( vinyl alcohol ) , EVOH = ethylene vinyl alcohol, PVDC = poly ( vinylidene chl oride ) , PTFE = polytetrafluoroethylene, PVC = poly ( vinyl chloride ) , PET = poly ( ethylene terephthalate ) , PC = polycarbonate. 109 Figure 2 - 32 Comparisons of P of different gases and vapors at 25 °C in various po lymers. (Data adapted from [81] and measured results in this review.) Polymer name abbreviations: PLA = poly(lactic acid), LDPE = low density polyethylene, HDPE = high density polyethylene, PP = polypropylene, PS = polystyrene, PET = poly(ethylene terephthalate), aPET = amorphous PET. 110 2.10 Migration of chemical compounds Migration is a phenomenon result ing from the diffusion and dissolution of low molecular mass compounds (i.e., migr ants) initially present in a polymer that are released into liquid medi a . It is a very crucial process for development of release compounds for medical applications and active packaging. Similar to mass transfer mechanism introduced in Section 2.4.1 , d iffusion of a migrant occurs through the amorphous regions of the polymer towards the interface. When the mass transfer reaches equilibrium, the partition coefficient of the migrants between polymer p and liquid f , K p,f , determines the equilibrium concentration and distribution in the two phases [288 290] . Different studies on migration of chemical compounds from PLA have bee n performed. These studies are focused either on medical application s for drug release or on food packaging applications. In medical application s , PLA has been used as a drug release system due to its biocompatibility, its degradation into non - toxic monome rs, and because the migration of chemical compounds can be controlled by changing the molecular weight and monomer ratio of PLA [291 293] . In food packaging, researchers have focused on positive migration studies , f or example, adding antioxidants/antimicrobials that can migrate from a PLA matrix to food product s to prolong the ir shelf life [49,240,294 297] . Migration phenomena of chemical compounds may be expressed mathematically. Table 2 - 10 shows a list of mathematical models that have been applied to est imate the parameters that describe the migration of different compounds from PLA to different media at different temperatures. Generally, the se models are based on the The Higuchi model is based on the release of high and 111 low water - soluble drugs incorporated in semi - solid and/or solid matrices [298 305] . to migration of chemical compounds from PLA that describe: 1) migration into infinite volume of solution and neglig ible external mass transfer coefficient [260,294,306 312] ; 2) migration into infinite volume of solution and non - negligible external mass transfer coefficient [309] ; and 3) migration into finite volume of solution and negligible external mass transfer coefficient [195,308 310,313] . Other models are based on concepts of dissolution - diffusion or burst effect [314,315] . 112 Table 2 - 10 Studies reporting kinetic migration parameters using mathematical models for PLA incorporated with different chemical compounds Compound Application Medium T, ° C Mathematical models Model Type Parameters estimated Ref . Vancomycin Drug release Phosphate buffer, pH 7.4 37 Higuchi k [300] Acetaminophen Drug release Phosphate buffer, pH 7.4 37 Higuchi k [301] Nonionic hydrophobic dye Agrichemical Water and phosphate buffer, pH 7.4 Room (~23) Higuchi f(t) [302] Lactic acid/lactide Food - contact 8% ethanol 26, 43 n/a D [316] 113 Table 2 - 10 ( cont d ) Compound Application Medium T, ° C Mathematical models Model Type Parameters estimated Ref. Ascorbyl palmitate, - tocopherol, BHA, BHT and TBHQ Food - contact 10, 50 and 95% ethanol 20, 40 law D and K p,f [306] BHA, BHT, PG and TBHQ Food - contact 10, 50 and 95% ethanol 20, 40 D and K p,f [260] 4 - Nitroanisole Drug release Phosphate buffer, pH 7.4 37 D [307] 114 Table 2 - 10 ( cont d ) Compound Application Medium T, ° C Mathematical models Model Type Parameters estimated Ref. Lidocaine Drug release Phosphate buffer, pH 7.4 37 D and K p,f [317] Nimesulide Drug release Phosphate buffer, pH 7.4 37 Higuchi k [304] Nifedipine Drug release Phosphate buffer 37 Dissolution - diffusion D and k [314] Lidoca i ne Drug release Phosphate buffer, pH 7.4 37 Dissolution - diffusion D, k dissolution and K p,f [318] Progesterone Drug release Water 37 n/a D [319] 115 Table 2 - 10 ( cont d ) Compound Application Medium T, ° C Mathematical models Model Type Parameters estimated Ref. S i lver Antimicrobial Water, 3% acetic acid, 95% ethanol 4, 20 n/a D [320] Thyme oil water bath room D [321] Catechin and epicatechin Food - contact 50 and 95% ethanol 20, 30, 40, 50 and K p,f [308] Propolis Food - contact water and ethanol 25 D, Bi, k external mass transfer and K p,f [309] 116 Table 2 - 10 ( cont d ) Compound Application Medium T, ° C Mathematical models Model Type Parameters estimated Ref. - tocopherol Food - contact ethanol 23, 33, 43 and K p,f [310] Resveratrol Food - contact ethanol 9, 23, 33, 43 D and K p,f [312] BHT Food - contact ethanol 23, 31, 43 and K p,f [195] Astaxanthin Food - contact 95% ethanol 30, 40 and K p,f [313] 117 Table 2 - 10 ( cont d ) Compound Application Medium T, ° C Mathematical models Model Type Parameters estimated Ref. Epigallocatechin gallate Food - contact water 37 D [311] Thymol Food - contact 950 and 150 mL/L ethanol/water 30, 40, 50, 60, 65, 75, 83 k, v 0 and D [294] Silver Food - contact 10% ethanol 40 D [252] 118 Table 2 - 10 ( cont d ) Compound Application Medium T, ° C Mathematical models Model Type Parameters estimated Ref. Oligonucleotides Drug release foetal calf serum, phosphate buffer pH 7.4, citrate phosphate pH 5.5 37 D [315] Mitomycin C Drug release Phospate buffer, pH 7.4 37 Higuchi D [303] Phenobarbitone Drug release Buffer pH 2 37 Higuchi k [305] Allyl isothiocyanate Food - contact Head space, 0 and 75% RH 37 n/a k [322] Hydrocortisone Drug release Phospate buffer, pH 7.4 37 n/a B and D [323] 119 Table 2 - 10 ( cont d ) Compound Application Medium T, ° C Mathematical models Model Type Parameters estimated Ref. Thymol Food - contact 10 and 95% ethanol 40 n/a D and K p,f [300] n/a: Model type not specified 120 Table 2 - 11 shows the estimated values of D f or chemical compounds from Table 2 - 10 . Different media have been used i n migration studies of PLA. For drug release systems, generally phosphate buffer solution (pH 7.4) at 37 °C is applied to simulate human body conditions. However, f or food packaging applications , diverse simulants are used in contact with PLA . F or instance , in accordance with the U.S. Food and Drug Administration (U.S. FDA) and the European Union Directives, 95% ethanol, 50% ethanol , and water are commonly used food simulants for fatty, alcoholic and watery liquid products , respectively [324,325] . 121 Table 2 - 11 D coefficient s reported of chemical compounds in different media and temperatures. Compound Media T, °C D × 1 0 14 , m 2 . s 1 Reference Lactic acid/lactide 8% ethanol 26 0.00004 [316] 43 0.002 - tocopherol 95% ethanol 40 14.90 [306] 20 0.20 BHA 95% ethanol 40 43.10 20 0.321 50% ethanol 40 1.12 20 0.074 10% ethanol 40 2.28 20 2.28 BHT 95% ethanol 40 31.20 20 0.16 50% ethanol 40 0.584 20 0.10 Propy l gallate 95% ethanol 40 39.00 20 0.612 50% ethanol 40 1.28 20 0.0756 10% ethanol 40 2.78 20 0.816 BHA 95% ethanol 40 59.90 [260] 20 2.66 50% ethanol 40 0.36 10% ethanol 40 0.19 BHT 95% ethanol 40 24.10 20 0.00297 50% ethanol 40 0.27 20 0.27 Prop y l gallate 95% ethanol 40 176.00 20 30.90 50% ethanol 40 1.02 20 1.50 10% ethanol 40 0.79 122 Table 2 - 11 ( cont d ) Compound Media T, °C D× 1 0 , m 2 .s 1 Reference 4 - nitroanisole p hosphate buffer 37 0.000052 [307] 0.000048 Lidocaine p hosphate buffer 37 0.000008 [317] Nifedipine p hosphate buffer 37 0.005640 [314] Lidoca i ne p hosphate buffer 37 0.000004 [318] Progesterone water 37 0.051 [319] S i lver 95% ethanol 20 0.000450 [320] 10% ethanol 0.0023 water 0.045 3% acetic acid 16.00 Thyme oil water bath room 0.139 [321] Catechin 95% ethanol 20 0.049 [308] 30 1.31 40 4.79 50 3.15 50% ethanol 40 1.12 Epicatechin 95% ethanol 20 0.088 30 1.37 40 5.12 50 3.49 50% ethanol 40 0.941 Pinobanksin water 25 8.90 [309] P inobanksin - 5 - methyl - ether water 25 10.30 p - coumaric acid water 25 7.40 Chrysin water 25 7.55 Pinobanksin ethanol 25 14200.00 - tocopherol ethanol 23 0.316 [310] 33 0.529 43 3.80 123 Table 2 - 11 ( cont d ) Compound Media T, °C D× 1 0 , m 2 .s 1 Reference Resveratrol 1% ethanol 9 0.00347 [312] 23 0.23 33 2.26 43 8.51 Resveratrol 3% ethanol 9 0.00349 23 0.306 33 4.17 43 8.26 BHT 95% ethanol 23 0.295 [195] 31 0.895 43 19.00 Astaxanthin 95% ethanol 30 1.27 [313] 40 2.28 E pigallocatechin gallate water 37 10.10 [311] Thymol 95% et hanol 30 29.00 [294] 40 60.00 50 163.00 60 57.50 15% et hanol 30 17.00 40 24.00 50 66.00 60 262.00 Silver 10% ethanol 40 1.12 [252] Oligonucleotides Phosphate buffer 37 0.00132 [315] foetal calf serum 37 0.00137 Hydrocortisone 4.8% p hosphate buffer 37 0.0417 [323] Hydrocortisone 12.1% 2.45 Hydrocortisone 15.3% 11.40 Hydrocortisone 25.8% 0.000547 124 Table 2 - 11 ( cont d ) Compound Media T, °C D× 1 0 , m 2 .s 1 Reference Thymol 10% ethanol 40 17.00 [3 16] 11.00 8.00 23.00 15.00 13.00 95% ethanol 2700.00 2800.00 2600.00 3500.00 5500.00 7000.00 Diffusion of migrants in PLA is governed by the polymer FV and the size of migrant molecules [17,235,236] . On the other hand, the polarity and the affinity between the migrant compounds and PLA are important in the case of solubility [326] . Moreover, plasticization effects of migrant can also affect the migration by inducing changes in T g or T m of the polymer. A detailed discussion of these factors is out of scope for this review. Additional information can be found elsewhere [306,327,328] . It has been report ed that the high polarity of 10% ethanol acted as a barrier - tocopherol, butylated hydroxyanisole ( BHA ) , butylated hydroxytoluene ( BHT ) and tert - b utylhydroquinone ( TBHQ ) at 20 °C ; however, this barrier was overcome at a higher te mperature ( 40 °C ) for BHA and TBHQ due to increas es in the ir molecular mobility [306] . C ertain organic solvents cause swelling of the PLA matrix, which creat es void spaces in th e polymer structure resulting in the promotion of the diffusion of chemical compounds. This is the case of PLA in contact 125 with ethanol , methanol, propanol, and butanol. For example, as ethanol acts as a plasticizer, - tocopherol, BHA, BHT, prop y l gallate, catechin, epicatechin and thymol increased [260,294,306,308] . Figure 2 - 33 shows migration of various compounds in ethanol (95% volume by v olume M w . The straight lines shown in the plot are linear least squares regression line s at each temperature. The current data did not show any t rends in D as a function of M w or temperature (data not shown) . Figure 2 - 33 Migration of organic compounds in ethanol (95 % v olume by volume ) . The straight lines shown in the plot are linear least squares regression line s at each temperature . References: thymol [316] , butylated hydroxyanisole ( BHA ) [260] , butylated hydroxytoluene ( BHT ) [195,260] , propyl gallate [260] , catechin [308] , - tocopherol [306] , astaxanthin [313] . 126 P olarity and solubility influence release rate of chemical compounds , d ue to the interactions of polymer, chemical compounds and simulants. For instance, faster release was observed for prop y l gallate than for BHT into 95% ethanol at 40 °C , showing the effect of the molecular volume and polarity of the antioxidant s . Also, fa ster release of BHA into 95% ethanol than 50% ethanol at 20 °C show s the influence of simulant polarity and antioxidant solubility [260,306] . No release of flavonoids was observed from PLA films to oil because of the limi ted solubility of the antioxidants in the medium , leaving the molecules trapped in the polymer matrix [308] . However, release of - tocopherol from PLA to oil was observed due to the solubility of this antioxidant in fatty media, which promoted the in teraction [310] . Regular solution theory (RST) was applied to predict the polymer - compound - simulant interac tions considering the absolute distance s between the solubility parameters as dispersion ( D ), polar ( P ), and hydrogen ( H ) bo n ding [195,297] ; The greater the distance between the chemical compound, media and polymer matrix , the lower the affinity. Ortiz - Vazquez et al. [195] calculated the relative distance ( ) of 2.37 MPa 1/2 between PLA and BHT, 20.5 MPa 1/2 between PLA and ethanol, 6.7 MPa 1/2 between PLA and oil, and 26.1 MPa 1/2 between PLA and water, indicating a higher affinity between PLA - BHT than PLA - solvents. The results were in accordance wit h no n - detected release of BHT in water at 13, 23 and 43 °C. 2.11 Final r emarks Until now, more than 2,6 00 unique experimental mass transfer measurements have been recorded for PLA , providing a unique assessment of PLA barrier. Lacking 127 systematic review of th e mass transfer of other polymers makes extensive comparison not possible. The only other polymer with so much reported mass transfer properties is PET , al beit without a systematic review of its mass transfer properties. Although a large number of barrier properties have been reported for PLA, most of these values were not reported with regard to the three - phase structure of semicrystalline PLA ( i.e. , CF, MAF and RAF). This is a large shortcoming of the reported barrier propert y values of PLA. However, it was not until the last decade that we started recognizing the three phases of most glassy polymers including common polymers such as PET. So, it is not the researcher lack of understanding, but the general mass transfer field of glass y polymers that is being challenged with this new finding. I t has been established that the P values of gases in PLA follow this general order ; P CO 2 > P He > P O 2 > P H 2 > P N 2 > P CH 4 . For oxygen, P O 2 follows this order ; LDPE>PS>PP>HDPE>PLA>PET>Nylon 6 . For water vapor, P H 2 O follows this order ; PLA>PS>PET>LDPE>PP>HDPE . Regarding organic vapors, limited values have been reported and additional research is needed. P LA with nanocomposites mostly show s some improvements in barrier properties, but PLA with additives and other treatments such as composites tend to worsen barrier properties , while results from blends vary. Finally, t here is a crucial need to consider the t hree - phase model when reporting th e mass transfer properties of PLA . C haracterization and determination of the free volume will be a strong addition when reporting mass transfer properties of PLA as well as other polymers. In summary, this comprehensive, critical and systematic review prov ides a unique mass transfer assessment of PLA to advance the commercialization and research of this distinctive bio - based and biodegradable polymer. 128 2.12 Acknowledgments The authors thank Violette Ducruet for providing some mass transfer data of PLA for creatin g this review. Author U. S. thanks the College of Agriculture and Natural Resource s (CANR) and the School of Packaging (SOP) for financial support through fellowships. F. I - F. thanks CONACYT, the Mexican Secretariat of Public Education (SEP) and the Mexica n Government for providing financial support through a Ph.D. fellowship. R. A. and M. R. thank the partial support of the USDA NIFA and MI AgBioResearch, Hatch project R. Auras and M. Rubino. 129 REFERENCES 130 R EFERENCES [1] Kraakman NJR, Rocha - Rios J, van Loosdrecht MCM. Review of mass transfer aspects for biological gas treatment. Appl Microbiol Biotechnol 2011;91:873 86. [2] Ahmadi H, Boliniu s DJ, Jahnke A, MacLeod M. Mass transfer of hydrophobic organic chemicals between silicone sheets and through plant leaves and low - density polyethylene. Chemosphere 2016;164:683 90. [3] Liechty WB, Kryscio DR, Slaughter B V., Peppas NA. Polymers for drug d elivery systems. Annu Rev Chem Biomol Eng 2010;1:149 73. [4] Paradiso VM, Summo C, Trani A, Caponio F. An effort to improve the shelf life of breakfast cereals using natural mixed tocopherols. J Cereal Sci 2008;47:322 30. [5] Sharma S, Kaur S, Dar BN, Sing h B. Storage stability and quality assessment of processed cereal brans. J Food Sci Technol 2014;51:583 8. [6] Gavara R, Catalá R, Hernández - Muñoz P. Extending the shelf - life of fresh - cut produce through active packaging. Stewart Postharvest Rev 2009;5:1 5 . [7] Lim L - T, Auras R, Rubino M. Processing technologies for poly(lactic acid). Prog Polym Sci 2008;33:820 52. [8] Jamshidian M, Tehrany EA, Imran M, Jacquot M, Desobry S. Poly - lactic acid: production, applications, nanocomposites, and release studies. Co mpr Rev Food Sci Food Saf 2010;9:552 71. [9] Castro - Aguirre E, Iñiguez - Franco F, Samsudin H, Fang X, Auras R. Poly(lactic acid) Mass production, processing, industrial applications, and end of life. Adv Drug Deliv Rev 2016;107:333 66. [10] Liu, Fishman ML, Hicks KB, Liu C - K. Biodegradable composites from sugar beet pulp and poly(lactic acid). J Agric Food Chem 2005;53:9017 22. [11] Wagner DJ. Sukano and Total Corbion present PLA masterbatch range. Addit Polym 2017;2017:3 4. [12] Saeidlou S, Hunea ult MA, Li H, Park CB. Poly(lactic acid) crystallization. Prog Polym Sci 2012;37:1657 77. [13] Auras R, Harte B, Selke S. An overview of polylactides as packaging materials. Macromol Biosci 2004;4:835 64. [14] Wasanasuk K, Tashiro K. Crystal structure and disorder in poly(L - 131 (Guildf) 2011;52:6097 109. [15] Fukushima K, Kimura Y. Stereocomplexed polylactides (Neo - PLA) a s high - performance bio - based polymers: their formation, properties, and application. Polym Int 2006;55:626 42. [16] Robertson DD, Neppalli R, Van Reenen AJ. Solution crystallization analysis of poly(lactic acid) by Scalls: A facile approach for thermal ana lysis of polymers in solution. Polym Test 2014;40:79 87. [17] Fang X, Vitrac O. Predicting diffusion coefficients of chemicals in and through packaging materials. Crit Rev Food Sci Nutr 2017;57:275 312. [18] Welle F. A new method for the prediction of diff usion coefficients in poly(ethylene terephthalate). J Appl Polym Sci 2013;129:1845 51. [19] Brandsch J, Mercea P, Ruter M, Tosa V, Piringer O. Migration modelling as a tool for quality assurance of food packaging. Food Addit Contam 2002;19:29 41. [20] Fang X, Domenek S, Ducruet V, Réfrégiers M, Vitrac O. Diffusion of aromatic solutes in aliphatic polymers above glass transition temperature. Macromolecules 2013;46:874 88. [21] Davis EM, Minelli M, Giacinti Baschetti M, Elabd YA. Non - Fickian diffusion of wate r in polylactide. Ind Eng Chem Res 2013;52:8664 73. [22] Madhavan Nampoothiri K, Nair NR, John RP. An overview of the recent developments in polylactide (PLA) research. Bioresour Technol 2010;101:8493 501. [23] Yu L, Petinakis E, Dean K, Liu H. Poly(lactic a cid)/starch blends. In: Auras RA, Lim LT, Selke SEM, Tsuji H, editors. Poly(lactic acid), Hoboken, NJ: John Wiley & Sons, Inc.; 2010, p. 217 26. [24] Detyothin S, Kathuria A, Jaruwattanayon W, Selke SEM, Auras R. Poly(lactic a cid) blends. In: Auras RA, L im LT, Selke SEM, Tsuji H, editors. Poly(lactic acid), Hoboken, NJ: John Wiley & Sons, Inc.; 2010, p. 227 71. [25] Matuana LM. Foaming. In: Auras RA, Lim LT, Selke SEM, Tsuji H, editors. Poly( l actic acid), Hoboken, NJ: John Wiley & Sons, Inc.; 2010, p. 273 91. [26] Ghosh SB, Bandyopadhyay - Ghosh S, Sain M. Composites. In: Auras RA, Lim LT, Selke SEM, Tsuji H, editors. Poly( l actic a cid), Hoboken, NJ: John Wiley & Sons, Inc.; 2010, p. 293 310. [27] Ray SS. Nanocomposites. In: Auras RA, Lim LT, Selke SEM, Tsuji H, editors. Poly( l actic a cid), Hoboken, NJ: John Wiley & Sons, Inc.; 2010, p. 311 22. 132 [28] Agrawal AK. Spinning of p oly( l actic a cid) f ibers. In: Auras RA, Lim LT, Selke SEM, Tsuji H, editors. Poly( l actic a cid), Hoboken, NJ: John Wiley & Sons, Inc.; 2010, p. 323 41. [29] Birnin - Yauri A, Ibrahim N, Zainuddin N, Abdan K, Then Y, Chieng B. Effect of m aleic a nhydride - m odified p oly(lactic acid) on the p roperties of i ts h ybrid f iber b iocomposites. Polymers (Basel) 2017;9:165/1 - 16. [30] Dong Q, Chow LC, Wang T, Fr ukhtbeyn SA, Wang F, Yang M, et al. A new bioactive polylactide - based composite with high mechanical strength. Colloids Surfaces A Physicochem Eng Asp 2014;457:256 62. [31] Kang JT, Kim SH. Improvement in the mechanical properties of polylactide and bamboo fiber biocomposites by fiber surface modification. Macromol Res 2011;19:789 96. [32] Mamun AA, Bledzki AK. Micro fibre reinforced PLA and PP composites: Enzyme modification, mechanical and thermal properties. Compos Sci Technol 2013;78:10 7. [33] Cheng Y, Deng S, Chen P, Ruan R. Polylactic acid (PLA) synthesis and modifications: A review. Front Chem China 2009;4:259 64. [34] Raquez JM, Habibi Y, Murariu M, Dubois P. Polylactide (PLA) - based nanocomposites. Prog Polym Sci 2013;38:1504 42. [35] Farah S, Anderson DG, Langer R. Physical and mechanical properties of PLA, and their functions in widespread applications - A comprehensive review. Adv Drug Deliv Rev 2016;107:367 92. [36] Torres A, Ilabaca E, Rojas A, Rodríguez F, Galotto MJ, Guarda A, et al. Effect of processing conditions on the physical, chemical and transport properties of polylactic acid films containing thymol incorporated by supercritical impregnation. Eur Polym J 2017;89:195 210. [37] Wang T, Chow LC, Frukhtbeyn SA, Ting AH, Dong Q, Yan g M, et al. Improve the strength of PLA/HA composite through the use of surface initiated polymerization and phosphonic acid coupling agent. J Res Natl Inst Stand Technol 2011;116:785. [38] Saeidlou S, Huneault MA, Li H, Park CB. Poly(lactic acid) stereoco mplex formation: Application to PLA rheological property modification. J Appl Polym Sci 2014;131:41073/1 - 8. [39] Surin P, Rakkwamsuk P, Wimolmala E, Sombatsompop N. Effects of coir fiber and maleic anhydride modification on the properties of thermoplastic starch/PLA composite laminates. J Nat Fibers 2015;12:108 20. 133 [40] Cacciotti I, Fortunati E, Puglia D, Kenny JM, Nanni F. Effect of silver nanoparticles and cellulose nanocrystals on electrospun poly(lactic) acid mats: Morphology, thermal properties and mec hanical behavior. Carbohydr Polym 2014;103:22 31. [41] Graupner N, Herrmann AS, Müssig J. Natural and man - made cellulose fibre - reinforced poly(lactic acid) (PLA) composites: An overview about mechanical characteristics and application areas. Compos Part A Appl Sci Manuf 2009;40:810 21. [42] Fortunati E, Luzi F, Puglia D, Dominici F, Santulli C, Kenny JM, et al. Investigation of thermo - mechanical, chemical and degradative properties of PLA - limonene films reinforced with cellulose nanocrystals extracted from Phormium tenax leaves. Eur Polym J 2014;56:77 91. [43] Ramos M, Jiménez A, Peltzer M, Garrigós MC. Development of novel nano - biocomposite antioxidant films based on poly(lactic acid) and thymol for active packaging. Food Chem 2014;162:149 55. [44] Zhong J, Li H, Yu J, Tan T. Effects of natural fiber surface modification on mechanical properties of poly(lactic acid) (PLA)/sweet sorghum fiber composites. Polym Plast Technol Eng 2011;50:1583 9. [45] Nuthong W, Uawongsuwan P, Pivsa - Art W, Hamada H. Impact prope rty of flexible epoxy treated natural fiber reinforced PLA composites. Energy Procedia 2013;34:839 47. [46] Yao M, Deng H, Mai F, Wang K, Zhang Q, Chen F, et al. Modification of poly(lactic acid)/poly(propylene carbonate) blends through melt compounding wi th maleic anhydride. Express Polym Lett 2011;5:937 49. [47] Foruzanmehr Mr, Vuillaume PY, Elkoun S, Robert M. Physical and mechanical properties of PLA composites reinforced by TiO 2 grafted flax fibers. Mater Des 2016;106:295 304. [48] Rocca - Smith JR, Karb owiak T, Marcuzzo E, Sensidoni A, Piasente F, Champion D, et al. Impact of corona treatment on PLA film properties. Polym Degrad Stab 2016;132:109 16. [49] Ramos M, Fortunati E, Peltzer M, Dominici F, Jiménez A, Garrigós M del C, et al. Influence of thymol and silver nanoparticles on the degradation of poly(lactic acid) based nanocomposites: Thermal and morphological properties. Polym Degrad Stab 2014;108:158 65. [50] Rhim J - W, Park H - M, Ha C - S. Bio - nanocomposites for food packaging applications. Prog Polym Sci 2013;38:1629 52. [51] Armentano I, Bitinis N, Fortunati E, Mattioli S, Rescignano N, Verdejo R, et al. Multifunctional nanostructured PLA materials for packaging and tissue 134 engineering. Prog Polym Sci 2013;38:1720 47. [52] Reddy MM, Vivekanandhan S, M isra M, Bhatia SK, Mohanty AK. Biobased plastics and bionanocomposites: Current status and future opportunities. Prog Polym Sci 2013;38:1653 89. [53] Faruk O, Bledzki AK, Fink HP, Sain M. Biocomposites reinforced with natural fibers: 2000 - 2010. Prog Polym Sci 2012;37:1552 96. [54] modification of polylactide. J Achiev Mater Manuf Eng 2010;43:192 9. [55] Sujaritjun W, Uawongsuwan P, Pivsa - Art W, Hamada H. Mechanical property of sur face modified natural fiber reinforced PLA biocomposites. Energy Procedia 2013;34:664 72. [56] Siparsky GL, Voorhees KJ, Dorgan JR, Schilling K. Water transport in polylactic acid (PLA), PLA/polycaprolactone copolymers, and PLA/polyethylene glycol blends. J Environ Polym Degrad 1997;5:125 36. [57] Wokadala OC, Ray SS, Bandyopadhyay J, Wesley - Smith J, Emmambux NM. Morphology, thermal properties and crystallization kinetics of ternary blends of the polylactide and starch biopolymers and nanoclay: The role of nanoclay hydrophobicity. Polymer (Guildf) 2015;71:82 92. [58] Elangovan D, Nidoni U, Yuzay IE, Selke SEM, Auras R. Poly (L - lactic acid ) metal organic framework composites. Mass transport properties. Ind Eng Chem Res 2011;50:11136 42. [59] Ali NA, Tariq F, Noori M. Gas barrier properties of biodegradable polymer nanocomposites films. Chem Mater Res 2014;6:44 51. [60] Witzke DR. Introduction to properties, engineering and prospects of polylactide polymers. PhD Thesis. Michigan State University, 1997. [61] Ko bayashi J, Asahi T, Ichiki M, Oikawa A, Suzuki H, Watanabe T, et al. Structural and optical properties of poly lactic acids. J Appl Phys 1995;77:2957 73. [62] Abdulkhani A, Hosseinzadeh J, Dadashi S, Mousavi M. A study of morphological, thermal, mechanical and barrier properties of PLA based biocomposites prepared with micro and nano sized cellulosic fibers. Cellul Chem Technol 2015;49:597 605. [63] Auras RA, Harte B, Selke S, Hernandez R. Mechanical, p hysical, and b arrier p roperties of p oly( l actide) f ilms. J Plast Film Sheeting 2003;19:123 35. [64] Tee YB, Rosnita AT, Khalina A, Chin NL, Roseliza KB, Khairul Faezah MYI. Reinforcing m echanical, w ater a bsorption and b arrier p roperties of p oly( l actic 135 a cid) c omposites with k enaf - d erived c ellulose of t hermally - g rafted a minosilane. Pertanika J Trop Agric Sci 2015;38:563 73. [65] Rhim JW, Lee JH, Ng PKW. Mechanical and barrier properties of biodegradable soy protein isolate - based films coated with polylactic acid. LWT - Food Sci Technol 2007; 40:232 8. [66] Graham T. A s hort a ccount of e xperimental r esearches on the d iffusion of g ases t hrough e ach o ther, and their s eparation by m echanical m eans. Q J Sci Arts 1829;2:74 83. [67] Prausnitz JM, Lichtenthaler RN, Azevedo EG. Molecular thermodynamics of fluid - phase equilibria. Upper Saddle River, NJ: Prentice Hall PTR; 1999. [68] Crank J. The m athematics of d iffusion: 2nd ed. New York: Oxford University Press; 1975. [69] Wijmans JG, Baker RW. The solution - diffusion model: a review. J Memb Sci 1995;107 :1 21. [70] Lewis GN. Outlines of a n ew s ystem of t hermodynamic c hemistry. Proc Am Acad Arts Sci 1907;43:259 93. [71] Stannett V, Yasuda H. The m easurement of g as and v apor p ermeation and d iffusion in p olymer. In: Schmitz JV, editor. Test ing of p olym ers . V ol. 1, New York: Johh Wiley & Sons; 1965, p. 393 418. [72] Berens AR, Hopfenberg HB. Diffusion and relaxation in glassy polymer powders: 2. Separation of diffusion and relaxation parameters. Polymer (Guildf) 1978;19:489 96. [73] Espino - Pérez E, Bras J, Alm eida G, Plessis C, Belgacem N, Perré P, et al. Designed cellulose nanocrystal surface properties for improving barrier properties in polylactide nanocomposites. Carbohydr Polym 2018;183:267 77. [74] Wang L, Corriou JP, Castel C, Favre E. Transport of gases in glassy polymers under transient conditions: Limit - behavior investigations of dual - mode sorption theory. Ind Eng Chem Res 2013;52:1089 101. [75] Flory PJ. Thermodynamics of h igh p olymer s olutions. J Chem Phys 1941;9:660 660. [76] Wolf BA. Making Flory Huggins p ractical: Thermodynamics of p olymer - c ontaining m ixtures. In: Wolf BA, Enders S, editors. Polym er t hermodyn amics: l iq uid p olym er - containing mi x tures vol. 54, Berlin, Heidelberg: Springer Berlin Heidelberg; 2010, p. 1 66. [77] Crank J, Park GS. Diffusion in p olymers. New York: Academic Press; 1968. 136 [78] Meares P. Transient permeation of organic vapors through polymer membranes. J Appl Polym Sci 1965;9:917 32. [79] George SC, Thomas S. Transport phenomena through polymeric system s. Prog Polym Sci 2001;26:985 1017. [80] Pasternak RA, Schimscheimer JF, Heller J. A dynamic approach to diffusion and permeation measurements. J Polym Sci Part A - 2 Polym Phys 1970;8:467 79. [81] Selke SEM, Culter JD. Plastics p ackaging: Properties, p roces sing, a pplications, and r egulations. 3rd ed. Cincinnati, OH: Hanser Publications; 2016. [82] Ewender J, Welle F. Determination and p rediction of the l ag t imes of h ydrocarbons through a p olyethylene t erephthalate f ilm. Packag Technol Sci 2014;27:963 74. [83 ] Bao L, Dorgan JR, Knauss D, Hait S, Oliveira NS, Maruccho IM. Gas permeation properties of poly(lactic acid) revisited. J Memb Sci 2006;285:166 72. [84] Li X, Li L, Song Y, Li H, Zhu Z, Chai Y. A permeation cup method for screening packaging materials fo r fragrance preservation in Chinese medicine. Anal Methods 2016;8:7387 95. [85] Gonçalves JJ, Silva G V., Santos AS, Fernandes EO. Improvement of the m odified d ry - c up m ethod for d etermination of the d iffusion c oefficients of g aseous p ollutants in b uilding m aterials. Defect Diffus Forum 2010;305 306:109 14. [86] Lee DS, Piergiovanni L, Yam KL. Food packaging science and technology. Boca Raton, FL: Taylor & Francis; 2008. [87] McBain JW, Bakr AM. A new sorption balance. J Am Chem Soc 192 6;48:690 5. [88] Dreisbach F, Lösch HW. Magnetic s uspension b alance f or s imultaneous m easurement of a s ample and the d ensity of the m easuring f luid. J Therm Anal Calorim 2000;62:515 21. [89] Gulati N. Use of QCM technology for measuring barrier properties of biodegradable packaging material. MS Thesis. Michigan State University, 2008. [90] Nielsen TJ, Giacin JR. The sorption of limonene/ethyl acetate binary vapour mixtures by a biaxially oriented polypropylene film. Packag Technol Sci 1994;7:247 58. [91] Kadam AA, Karbowiak T, Voilley A, Debeaufort F. Techniques to measure sorption and migration between small molecules and packaging. A critical review. J Sci Food Agric 2015;95:1395 407. [92] Smith AL, Mulligan RB, Shirazi HM. Determinin g the effects of vapor sorption in 137 polymers with the quartz crystal microbalance/heat conduction calorimeter. J Polym Sci Part B Polym Phys 2004;42:3893 906. [93] Laatikainen M, Lindström M. Measurement of sorption in polymer membranes with a quartz crysta l microbalance. J Memb Sci 1986;29:127 41. [94] Milczewska K, Voelkel A. Inverse g as c hromatography in c haracterization of c omposites i nteraction. In: Mohd MA, editor. Adv anced g as c hromatogr aphy - p rog ress in a gric ultural, b iomed ical and industrial applic ations. Rijeka: InTech; 2012, p. 421 36. [95] Fambri L, Migliaresi C. Crystallization and t hermal p roperties. In: Auras RA, Lim LT, Selke SEM, Tsuji H, editors. Poly( l actic a cid), Hoboken, NJ: John Wiley & Sons, Inc.; 2010, p. 113 24. [96] Li H, Huneault M A. Effect of nucleation and plasticization on the crystallization of poly(lactic acid). Polymer (Guildf) 2007;48:6855 66. [97] Bigg DM. Polylactide copolymers: Effect of copolymer ratio and end capping on their properties. Adv Polym Technol 2005;24:69 82. [98] Feng L, Zhang B, Bian X, Li G, Chen Z, Chen X. Thermal p roperties of p olylactides with d ifferent s tereoisomers of l actides u sed as c omonomers. Macromolecules 2017;50:6064 73. [99] Kolstad JJ. Crystallization kinetics of poly(L - lactide - c o - meso - lactide). J Appl Polym Sci 1996;62:1079 91. [100] Hartmann MH. High m olecular w eight p olylactic a cid p olymers. In: Kaplan DL, editor. Biopolym ers from r enew able r esour ces. Berlin, Heidelberg: Springer Berlin Heidelberg; 1998, p. 367 411. [101] Tsuji H, Ikada Y. Stereocomplex formation between enantiomeric poly(lactic acid)s. 6. Binary blends from copolymers. Macromolecules 1992;25:5719 23. [102] Abe H, Kikkawa Y, Inoue Y, Doi Y. Morphological and k inetic a nalyses of r egime t ransition for p oly[(S) - lac tide] c rystal g rowth. Biomacromolecules 2001;2:1007 14. [103] Dorgan JR, Janzen J, Clayton MP, Hait SB, Knauss DM. Melt rheology of variable L - content poly(lactic acid). J Rheol (N Y N Y) 2005;49:607 19. [104] Jamshidi K, Hyon S - H, Ikada Y. Thermal charact erization of polylactides. Polymer (Guildf) 1988;29:2229 34. [105] Fox TG, Flory PJ. Second - order transition temperatures and related properties of polystyrene. I. Influence of molecular weight. J Appl Phys 1950;21:581 91. [106] Zhang J, Tashiro K, Tsuji H , Domb AJ. Disorder - to - order phase transition and 138 multiple melting behavior of poly(L - lactide) investigated by simultaneous measurements of WAXD and DSC. Macromolecules 2008;41:1352 7. [107] De Santis P, Kovacs J. Molecular conformation of poly(S lactic ac id). Biopolymers 1968;6:299 - 306. [108] Kalb B, Pennings AJ. General crystallization behaviour of poly(L - lactic acid). Polymer (Guildf) 1980;21:607 12. [109] Hoogsteen W, Postema A. Crystal structure, conformation and morphology of solution - spun poly(L - lactide) fibers. Macromolecules 1990;23:634 42. [110] Miyata T, Masuko T. Morphology of poly(L - lactide) solution - grown crystals. Polymer (Guildf) 1997;38:4003 9. [111] Miyata T, Masuko T. Crystallization behaviour of poly(L - lactide). P olymer (Guildf) 1998;39:5515 21. [112] Wasanasuk K, Tashiro K, Hanesaka M, Ohhara T, Kurihara K, Kuroki R, et al. Crystal structure analysis of poly(L - - dimensional wide - angle synchrotron X - ray and neutron diffractio n measurements. Macromolecules 2011;44:6441 52. [113] Puiggali J, Ikada Y, Tsuji H, Cartier L, Okihara T, Lotz B. The frustrated structure of poly(L - lactide). Polymer (Guildf) 2000;41:8921 30. [114] Sawai D, Tsugane Y, Tamada M, Kanamoto T, Sungil M, Hyon S - H. Crystal density and heat of fusion for a stereo - complex of poly(L - lactic acid) and poly(D - lactic acid). J Polym Sci Part B Polym Phys 2007;45:2632 9. [115] Cartier L, Okihara T, Ikada Y, Tsuji H, Puiggali J, Lotz B. Epitaxial crystallization and cryst alline polymorphism of polylactides. Polymer (Guildf) 2000;41:8909 19. [116] Michaels AS, Bixler HJ. Solubility of gases in polyethylene. J Polym Sci 1961;50:393 412. [117] Menczel J, Wunderlich B. Heat capacity hysteresis of semicrystalline macromolecular glasses. J Polym Sci Polym Lett Ed 1981;19:261 4. [118] Wunderlich B. Reversible crystallization and the rigid - amorphous phase in semicrystalline macromolecules. Prog Polym Sci 2003;28:383 450. [119] Nguyen TL, Bédoui F, Mazeran PE, Guigon M . Mechanical investigation of confined amorphous phase in semicrystalline polymers: Case of PET and PLA. Polym Eng Sci 2015;55:397 405. [120] poly(lactic acid) by advanced thermal analysis. Polymer (Guildf) 2009;50:3967 139 73. [121] Tsai W - C, Hedenqvist MS, Laiback Å, Melin H, Ngo M, Trollsås M, et al. Physical changes and sorption/desorption behaviour of amorphous and semi - crystalline PLLA exposed to water, methanol and ethanol. Eur P olym J 2016;76:278 93. [122] Delpouve N, Saiter A, Dargent E. Cooperativity length evolution during crystallization of poly(lactic acid). Eur Polym J 2011;47:2414 23. [123] Zuza E, Ugartemendia JM, Lopez A, Meaurio E, Lejardi A, Sarasua J - R. Glass transiti on behavior and dynamic fragility in polylactides containing mobile and rigid amorphous fractions. Polymer (Guildf) 2008;49:4427 32. [124] Arnoult M, Dargent E, Mano JF. Mobile amorphous phase fragility in semi - crystalline polymers: Comparison of PET and P LLA. Polymer (Guildf) 2007;48:1012 9. [125] Righetti MC, Tombari E. Crystalline, mobile amorphous and rigid amorphous fractions in poly(L - lactic acid) by TMDSC. Thermochim Acta 2011;522:118 27. [126] Mano JF, Gómez Ribelles JL, Alves NM, Salmerón Sanchez M . Glass transition dynamics and structural relaxation of PLLA studied by DSC: Influence of crystallinity. Polymer (Guildf) 2005;46:8258 65. [127] Di Lorenzo ML, Righetti MC, Cocca M, Wunderlich B. Coupling between c rystal m elting and r igid a morphous f racti on m obilization in p oly(ethylene terephthalate). Macromolecules 2010;43:7689 94. [128] Delpouve N, Saiter A, Mano JF, Dargent E. Cooperative rearranging region size in semi - crystalline poly(L - lactic acid). Polymer (Guildf) 2008;49:3130 5. [129] Esposito A, Delpouve N, Causin V, Dhotel A, Delbreilh L, Dargent E. From a t hree - p hase m odel to a c ontinuous d escription of m olecular m obility in s emicrystalline p oly(hydroxybutyrate - co - hydroxyvalerate). Macromolecules 2016;49:4850 61. [130] Tsuji H. Hydrolytic d egra dation. In: Auras RA, Lim LT, Selke SEM, Tsuji H, editors. Poly( l actic a cid), Hoboken, NJ: John Wiley & Sons, Inc.; 2010, p. 343 81. [131] Pyda M, Bopp RC, Wunderlich B. Heat capacity of poly(lactic acid). J Chem Thermodyn 2004;36:731 42. [132] Fernandes Nassar S, Guinault A, Delpouve N, Divry V, Ducruet V, Sollogoub C, et al. Multi - scale analysis of the impact of polylactide morphology on gas barrier properties. Polymer (Guildf) 2017;108:163 72. [133] Guinault A, Sollogoub C, Ducruet V, Domenek S. Impact of crystallinity of poly(lactide) on helium and oxygen barrier properties. Eur Polym J 2012;48:779 140 88. [134] Del Río J, Etxeberria A, López - Rodríguez N, Lizundia E, Sarasua JR. A PALS contribution to the supramolecular structure of poly(L - lactide ). Macromolecules 2010;43:4698 707. [135] Hernandez RJ. Plastic in p ackaging. In: Harper CA, editor. Handb ook of p last ics, e lastomers, and c ompos ites. 3rd ed., New York: McGraw - Hill; 1996, p. 627 92. [136] Van Krevelen DW, Te Nijenhuis K. Properties d eterm ining m ass t ransfer i n p olymeric s ystems. In: Van Krevelen DW, Te Nijenhuis K, editors. Prop erties of polymers: Their c orrel ation with c hem ical s truct ure; t heir n umer ical e stim ation and p redict ion from a ddit ive group contributions. 4th ed., Amsterdam, The Netherlands: Elsevier; 2009, p. 655 702. [137] Auras RA. Investigation of polylactide as packaging material. PhD Thesis. Michigan State University, 2004. [138] Rogers CE. Permeation of g ases and v apours in p olymers. In: Comyn J, editor. Polym er Permeabilit y, Dordrecht: Springer Netherlands; 1985, p. 11 73. [139] Danner RP. Measuring and correlating diffusivity in polymer - solvent systems using free - volume theory. Fluid Phase Equilib 2014;362:19 27. [140] Frisch HL, Alexander Stern S. Diffusion of small molec ules in polymers. Crit Rev Solid State Mater Sci 1983;11:123 87. [141] Hernandez R, Giacin J. Factors affecting permeation, sorption, and migration processes in package - product systems. In: Taub IA, Singh RP, editors. Food s torage s tab ility. Boca Raton: CR C Press; 1997, p. 270 331. [142] Vrentas JS, Vrentas CM. Diffusion and m ass t ransfer. Boca Raton, FL: CRC Press; 2012. [143] Sawada H, Takahashi Y, Miyata S, Kanehashi S, Sato S, Nagai K. Gas t ransport p roperties and c rystalline s tructures of p oly(lactic acid) m embranes. Trans Mater Res Soc Japan 2010;35:241 6. [144] Komatsuka T, Nagai K. Temperature d ependence on g as p ermeability and p ermselectivity of p oly(lactic acid) b lend m embranes. Polym J 2009;41:455 8. [145] Lehermeier HJ, Dorgan JR, Way JD. Gas permeation properties of poly(lactic acid). J Memb Sci 2001;190:24 3 51. [146] Netramai S, Rubino M, Auras R, Annous BA. Mass t ransfer s tudy of c hlorine d ioxide g as t hrough p olymeric p ackaging m aterials. J Appl Polym Sci 2009;114:2929 36. 141 [147] Drieskens M, Peeters R, Mullens J, Franco D, Iemstra PJ, Hristova - Bogaerds DG. Structure versus properties relationship of poly(lactic acid). I. effect of crystallinity on barrier properties. J Polym Sci Part B Polym Phys 2009;47:2247 58. [148] Auras R, Harte B, Selke S. Effect of water on the oxygen barrier properties of poly(ethyl ene terephthalate) and polylactide films. J Appl Polym Sci 2004;92:1790 803. [149] Auras RA, Singh SP, Singh JJ. Evaluation of oriented poly(lactide) polymers vs. existing PET and oriented PS for fresh food service containers. Packag Technol Sci 2005;18:20 7 16. [150] Thellen C, Orroth C, Froio D, Ziegler D, Lucciarini J, Farrell R, et al. Influence of montmorillonite layered silicate on plasticized poly(L - lactide) blown films. Polymer (Guildf) 2005;46:11716 27. [151] Cabedo L, Luis Feijoo J, Pilar Villanuev a M, Lagarón JM, Giménez E. Optimization of b iodegradable n anocomposites b ased on aPLA/PCL b lends for f ood p ackaging a pplications. Macromol Symp 2006;233:191 7. [152] Holm VK, Ndoni S, Risbo J. The s tability of p oly(lactic acid) p ackaging f ilms as i nfluenced by h umidity and t emperature. J Food Sci 2006;71:E40 4. [153] Sanchez - Garcia MD, Gimenez E, Lagaron JM. Novel PET Nanocomposites of i nterest in f ood p ackaging a pplications and c omparative b arrier p erformance w ith b iopolyester n anocomposites. J Pla st Film Sheeting 2007;23:133 48. [154] Chowdhury SR. Some important aspects in designing high molecular weight poly(L - lactic acid) - clay nanocomposites with desired properties. Polym Int 2008;57:1326 32. [155] Colomines G, Domenek S, Ducruet V, Guinault A. Influences of the crystallisation rate on thermal and barrier properties of polylactide acid (PLA) food packaging films. Int J Mater Form 2008;1:607 10. [156] Koh HC, Park JS, Jeong MA, Hwang HY, Hong YT, Ha SY, et al. Preparation and gas permeation proper ties of biodegradable polymer/layered silicate nanocomposite membranes. Desalination 2008;233:201 9. [157] Bhatia A, Gupta RK, Bhattacharya SN, Choi HJ. Effect of c lay on t hermal, m echanical and g as b arrier p roperties of b iodegradable p oly(lactic acid)/ p ol y(butylene succinate) (PLA/PBS) n anocomposites. Int Polym Process 2010;25:5 14. [158] Cho S - W, G llstedt M, Hedenqvist MS. Properties of w heat g luten/ p oly(lactic acid) l aminates. J Agric Food Chem 2010;58:7344 50. [159] , Guinault A, Domenek S. Barrier 142 properties of poly(lactic acid) and its morphological changes induced by aroma compound sorption. Polym Int 2010;59:818 26. [160] Ducruet V, Domenek S, Guinault A, Courgneau C, Bernasconi M, Plessis C. Barrier properties of PLA towards oxygen and aroma compounds. Ital J Food Sci 2010;23:59 62. [161] Guinault A, Sollogoub C, Domenek S, Grandmontagne A, Ducruet V. Influence of crystallinity on gas barrier and mechanical properties of PLA food packaging films. Int J Mater Form 2010;3:603 6. [162] Martucci JF, Ruseckaite RA. Three - layer sheets based on gelatin and poly(lactic acid), part 1: Preparation and properties. J Appl Polym Sci 2010;118:3102 10. [163] Courgneau C, Domenek S, Guinault A, Avérous L, Ducruet V. Analysis of th e s tructure - p roperties r elationships of d ifferent m ultiphase s ystems b ased on p lasticized p oly( l actic a cid). J Polym Environ 2011;19:362 71. [164] Guinault A, Menary GH, Courgneau C, Griffith D, Ducruet V, Miri V, et al. The effect of the stretching of PLA extruded films on their crystallinity and gas barrier properties. AIP Conf erence Proc eedings . vol. 1353, 2011, p. 826 31. [165] Picard E, Espuche E, Fulchiron R. Effect of an organo - modified montmorillonite on PLA crystallization and gas barrier properties. Appl Clay Sci 2011;53:58 65. [166] Sanchez - Garcia MD, Nordqvist D, Hedenqvist M, Lagaron JM. Incorporating amylopectin in poly(lactic acid) by melt blending using poly(ethylene - co - vinyl alcohol) as a thermoplastic carrier. II. Physical properties. J Appl Polym Sci 2011;119:3708 16. [167] Wen X, Zhang K, Wang Y, Han L, Han C, Zhang H, et al. Study of the thermal stabilization mechanism of biodegradable poly(L - lactide)/silica nanocomposites. Polym Int 2011;60:202 10. [168] Courgneau C, Dom enek S, Lebossé R, Guinault A, Avérous L, Ducruet V. Effect of crystallization on barrier properties of formulated polylactide. Polym Int 2012;61:180 9. [169] poly(lactide) on barr ier properties. 5th Int ernational Symp osium on Food Packag ing , Berlin, Germany: 2012. [170] Fukushima K, Fina A, Geobaldo F, Venturello A, Camino G. Properties of poly(lactic acid) nanocomposites based on montmorillonite, sepiolite and zirconium phosphonat e. Express Polym Lett 2012;6:914 26. [171] Jamshidian M, Arab Tehrany E, Cleymand F, Leconte S, Falher T, Desobry S. Effects of synthetic phenolic antioxidants on physical, structural, mechanical and 143 barrier properties of poly lactic acid film. Carbohydr Polym 2012;87:1763 73. [172] Sansone L, Aldi A, Musto P, Di Maio E, Amendola E, Mensitieri G. Assessing the suitability of polylactic acid flexible films for high pressure pasteurization and sterilization of packaged foodstuff. J Food Eng 2012;111:34 45. [173] Siracusa V, Blanco I, Romani S, Tylewic z U, Rocculi P, Rosa MD. Poly(lactic acid) - modified films for food packaging application: Physical, mechanical, and barrier behavior. J Appl Polym Sci 2012;125:E390 401. [174] Espino - Pérez E, Bras J, Ducruet V, Guinault A, Dufresne A, Domenek S. Influence of chemical surface modification of cellulose nanowhiskers on thermal, mechanical, and barrier properties of poly(lactide) based bionanocomposites. Eur Polym J 2013;49:3144 54. [175] Fukuzumi H, Saito T, Isogai A. Influence of TEMPO - oxidized cellulose nano fibril length on film properties. Carbohydr Polym 2013;93:172 7. [176] Gonçalves CMB, Tomé LC, Garcia H, Brandão L, Mendes AM, Marrucho IM. Effect of natural and synthetic antioxidants incorporation on the gas permeation properties of poly(lactic acid) fil ms. J Food Eng 2013;116:562 71. [177] Kathuria A, Abiad MG, Auras R. Deterioration of metal - organic framework crystal structure during fabrication of poly(L - lactic acid)mixed - matrix membranes. Polym Int 2013;62:1144 51. [178] Kim HK, Kim SJ, Lee HS, Choi J H, Jeong CM, Sung MH, et al. Mechanical and barrier properties of poly(lactic acid) films coated by nanoclay - ink composition. J Appl Polym Sci 2013;127:3823 9. [179] Pinto AM, Cabral J, Tanaka DAP, Mendes AM, Magalhães FD. Effect of incorporation of graphe ne oxide and graphene nanoplatelets on mechanical and gas permeability properties of poly(lactic acid) films. Polym Int 2013;62:33 40. [180] Di Maio L, Scarfato P, Milana MR, Feliciani R, Denaro M, Padula G, et al. Bionanocomposite polylactic acid/organocl ay films: Functional properties and measurement of total and lactic acid specific migration. Packag Technol Sci 2014;27:535 47. [181] Phupoksakul T, Leuangsukrerk M, Numpiboonmarn P, Somwangthanaroj A, Janjarasskul T. Properties of poly(lactide) - whey prote in isolate laminated films. J Sci Food Agric 2014;95:715 21. [182] Avolio R, Castaldo R, Gentile G, Ambrogi V, Fiori S, Avella M, et al. Plasticization of poly(lactic acid) through blending with oligomers of lactic acid: Effect of the physical aging on pro perties. Eur Polym J 2015;66:533 42. [183] Byun Y, Rodriguez K, Han JH, Kim YT. Improved thermal stability of polylactic 144 acid (PLA) composite film via PLA - cyclodextrin - inclusion complex systems. Int J Biol Macromol 2015;81:591 8. [184] Kathuria A, Al - Gha mdi S, Abiad MG, Auras R. The Influence of Cu 3 (BTC) 2 metal organic framework on the permeability and perm - selectivity of PLLA - MOF mixed matrix membranes. J Appl Polym Sci 2015;132:42764/1 - 10. [185] Moazeni N, Mohamad Z, Dehbari N. Study of silane treatment on poly - lactic acid(PLA)/sepiolite nanocomposite thin films. J Appl Polym Sci 2015;132:41428/1 - 8. [186] Ortenzi MA, Basilissi L, Farina H, Di Silvestro G, Piergiovanni L, Mascheroni E. Evaluation of crystallinity and gas barrier properties of films obtain ed from PLA - lactide with silane - modified nanosilica and montmorillonite. Eur Polym J 2015;66:478 91. [187] Morel F, Espuche E, Eronique Bounor - Legar V, Persynn O, Lacroix M. Impact of c oated c alc ium c arbonate n anofillers and a nnealing t reatments on the m icrostructure and g as b arrier p roperties of p oly(lactide) b ased n anocomposite f ilms. J Polym Sci, Part B Polym Phys 2016;54:649 58. [188] Yang X, Xu H, Odelius K, Hakkarainen M. Poly(lactide) - g - poly(butylene succinate - co - adipate) with h igh c rystallization c apacity and m igration r esistance. Materials (Basel) 2016;9:313/1 - 15. [189] Li Y, Ren P - G, Zhang Q, Shen T - T, Ci J - H, Fang C - Q. Properties of p oly( l actic a cid)/ o rgano - m ontmorillo nite n anocomposites p repared by s olution i ntercalation. J Macromol Sci Part B 2013;52:1041 55. [190] Sato S, Wada T, Ido R, Murakoshi Y, Kanehashi S, Nagai K. Dependence of alcohol vapor - induced crystallization on gas and vapor permeabilities of poly(lacti c acid) films. J Appl Polym Sci 2014;131:40140/1 - 9. [191] Boufarguine M, Guinault A, Miquelard - Garnier G, Sollogoub C. PLA/PHBV films with improved mechanical and gas barrier properties. Macromol Mater Eng 2013;298:1065 73. [192] Drumright RE, Gruber PR, H enton DE. Polylactic a cid t echnology. Adv Mater 2000;12:1841 6. [193] Oliveira NS, Oliveira J, Gomes T, Ferreira A, Dorgan J, Marrucho IM. Gas sorption in poly(lactic acid) and packaging materials. Fluid Phase Equilib 2004;222 223:317 24. [194] Busolo MA, Torres - Giner S, Lagaron JM. Enhancing the g as b arrier p roperties of p olylactic a cid by m eans of e lectrospun u ltrathin z ein f ibers. ANTEC, Proc eedings of the 67th Annu al Tech nical Conf erence , Chicago, IL: Society of Plastics Engineers; 2009, p. 2763 7. 145 [195] Ortiz - Vazquez H, Shin J, Soto - Valdez H, Auras R. Release of butylated hydroxytoluene (BHT) from Poly(lactic acid) films. Polym Test 2011;30:463 71. [196] Pankaj SK, Bueno - Characteriza tion of polylactic acid films for food packaging as affected by dielectric barrier discharge atmospheric plasma. Innov Food Sci Emerg Technol 2014;21:107 13. [197] Mattioli S, Peltzer M, Fortunati E, Armentano I, Jiménez A, Kenny JM. Structure, gas - barrier properties and overall migration of poly(lactic acid) films coated with hydrogenated amorphous carbon layers. Carbon 2013;63:274 82. [198] Aulin C, Karabulut E, Tran A, Wågberg L, Lindström T. Correction to t ransparent n anocellulosic m ultilayer t hin f ilms on p olylactic a cid with t unable g as b arrier p roperties. ACS Appl Mater Interfaces 2013;5:10395 6. [199] Yuniarto K, Welt BA, Purwanto A, Pu rwadaria HK, Abdellatief A. Effect of p lasticizer on o xygen p ermeability of c ast p olylactic acid (PLA) f ilms d etermined using d ynamic a ccumulation m ethod. J Appl Packag Res 2014;6:51 7. [200] Huang HD, Ren PG, Xu JZ, Xu L, Zhong GJ, Hsiao BS, et al. Improv ed barrier properties of poly(lactic acid) with randomly dispersed graphene oxide nanosheets. J Memb Sci 2014;464:110 8. [201] Xie L, Xu H, Chen J - B, Zhang Z - J, Hsiao BS, Zhong G - J, et al. From n anofibrillar to n anolaminar p oly(butylene succinate): Paving the w ay to r obust b arrier and m echanical p roperties for f ull - b iodegradable p oly(lactic acid) f ilms. Appl Mater Interfaces 2015;7:8023 32. [202] Ojijo V, Sinha Ray S, Sadiku R. Concurrent enhancement of multiple properties in reactively processed nanocompo sites of polylactide/poly[(butylene succinate) - co - adipate] blend and organoclay. Macromol Mater Eng 2014;299:596 608. [203] Ambrosio - Martín J, López - Rubio A, José Fabra M, Angel López - Manchado M, Sorrentino A, Gorrasi G, et al. Synergistic effect of lactic acid oligomers and laminar graphene sheets on the barrier properties of polylactide nanocomposites obtained by the in situ polymerization pre - incorporation method. J Appl Polym Sci 2016;133:n/a - n/a. [204] Petersen K, Nielsen P V., Olsen MB. Physical and m echanical p roperties of b iobased m aterials s tarch, p olylactate and p olyhydroxybutyrate. Starch - Stärke 2001;53:356 61. [205] Komatsuka T, Kusakabe A, Nagai K. Characterization and gas transport properties of poly(lactic acid) blend membranes. Desali nation 2008;234:212 20. [206] Sanchez - Garcia MD, Lagaron JM. Novel clay - based nanobiocomposites of biopolyesters with synergistic barrier to UV light, gas, and vapour. J Appl Polym 146 Sci 2010;118:188 99. [207] Sanchez - Garcia MD, Lagaron JM. On the use of pla nt cellulose nanowhiskers to enhance the barrier properties of polylactic acid. Cellulose 2010;17:987 1004. [208] Chang J, An YU, Sur GS. Poly(lactic acid) nanocomposites with various organoclays. I. Thermomechanical properties, morphology, and gas permeab ility. J Polym Sci Part B Polym Phys 2003;41:94 103. [209] Sinha Ray S, Okamoto M. Biodegradable polylactide and its nanocomposites: Opening a new dimension for plastics and composites. Macromol Rapid Commun 2003;24:815 40. [210] Lagaron JM, Cabedo L, Cava D, Feijoo JL, Gavara R, Gimenez E. Improving packaged food quality and safety. Part 2: Nanocomposites. Food Addit Contam 2005;22:994 8. [211] Petersson L, Oksman K. Biopolymer based nanocomposites: Comparing layered silicates and microcrystalline cellulos e as nanoreinforcement. Compos Sci Technol 2006;66:2187 96. [212] Jain S, Reddy MM, Mohanty AK, Misra M, Ghosh AK. A new biodegradable flexible composite sheet from poly(lactic acid)/poly(e - caprolactone) blends and micro - talc. Macromol Mater Eng 2010;295:7 50 62. [213] Zhu Y, Buonocore GG, Lavorgna M, Ambrosio L. Poly(lactic acid)/titanium dioxide nanocomposite films: Influence of processing procedure on dispersion of titanium dioxide and photocatalytic activity. Polym Compos 2011;32:519 28. [214] Katiyar V, Gerds N, Koch CB, Risbo J, Hansen HCB, Plackett D. Melt processing of poly(L - lactic acid) in the presence of organomodified anionic or cationic clays. J Appl Polym Sci 2011;122:112 25. [215] Byun Y, Whiteside S, Thomas R, Dharman M, Hughes J, Kim YT. The effect of solvent mixture on the properties of solvent cast polylactic acid (PLA) film. J Appl Polym Sci 2012;124:3577 82. [216] Samuel C, Raquez J - M, Dubois P. PLLA/PMMA blends: A shear - induced miscibility with tunable morphologies and properties? Polymer (Guildf) 2013;54:3931 9. [217] Battegazzore D, Alongi J, Frache A. Poly(lactic acid) - b ased c omposites c ontaining n atural f illers: Thermal, m echanical and b arrier p roperties. J Polym Environ 2014;22:88 98. [218] Battegazzore D, Bocchini S, Alongi J, Frache A, Marino F. Cellulose extracted from rice husk as filler for poly(lactic acid): Preparation and characterization. Cellulose 2014;21:1813 21. 147 [219] Risse S, Tighzert L, Berzin F, Vergnes B. Microstructure, rheological behavior, and properties of poly(lact ic acid)/poly(butylene succinate)/organoclay nanocomposites. J Appl Polym Sci 2014;131:40364/1 - 8. [220] Cinelli P, Schmid M, Bugnicourt E, Wildner J, Bazzichi A, Anguillesi I, et al. Whey protein layer applied on biodegradable packaging film to improve bar rier properties while maintaining biodegradability. Polym Degrad Stab 2014;108:151 7. [221] Valapa RB, Pugazhenthi G, Katiyar V. Fabrication and characterization of sucrose palmitate reinforced poly(lactic acid) bionanocomposite films. J Appl Polym Sci 201 5;132:41320/1 - 10. [222] Flodberg G, Helland I, Thomsson L, Bodil Fredriksen S. Barrier properties of polypropylene carbonate and poly(lactic acid) cast films. Eur Polym J 2015;63:217 26. [223] Ambrosio - Martín J, Fabra MJ, Lopez - Rubio A, Lagaron JM. Melt po lycondensation to improve the dispersion of bacterial cellulose into polylactide via melt compounding: enhanced barrier and mechanical properties. Cellulose 2015;22:1201 26. [224] Samsudin H. Migration of antioxidants from poly(lactic acid), PLA, films int o food simulants: A parameter estimation approac h. PhD Thesis . Michigan State University, 2015. [225] Berens AR, Huvard GS. Particle s ize d istribution of p olymer p owders by a nalysis of s orption k inetics. J Dispers Sci Technol 1981;2:359 78. [226] Berens AR, Hopfenberg HB. Diffusion of organic vapors at low concentrations in glassy PVC, polystyrene, and PMMA. J Memb Sci 1982;10:283 303. [227] Qiu J, Zheng JM, Peinemann K V. Gas transport properties in a novel poly(trimethylsilylpropyne) composite me mbrane with nanosized organic filler trimethylsilylglucose. Macromolecules 2006;39:4093 100. [228] Oliveira NS, Gonçalves CM, Coutinho JAP, Ferreira A, Dorgan J, Marrucho IM. Carbon dioxide, ethylene and water vapor sorption in poly(lactic acid). Fluid Pha se Equilib 2006;250:116 24. [229] Oliveira NS, Dorgan J, Coutinho JAP, Ferreira A, Daridon JL, Marrucho IM. Gas solubility of carbon dioxide in poly(lactic acid) at high pressures: Thermal treatment effect. J Polym Sci Part B Polym Phys 2007;45:616 25. [23 0] Oliveira NS, Dorgan J, Coutinho JAP, Ferreira A, Daridon JL, Marrucho IM. Gas solubility of carbon dioxide in poly(lactic acid) at high pressures. J Polym Sci Part B Polym Phys 2006;44:1010 9. 148 [231] Auras RA. Solubility of g ases and v apors in p olylactid e p olymers. In: Letcher TM, editor. Thermodyn amics, s olubility and e nviron mental i ssues, London: Elsevier; 2007, p. 343 68. [232] Sato S, Nyuui T, Matsuba G, Nagai K. Correlation between interlamellar amorphous structure and gas permeability in poly(lactic acid) films. J Appl Polym Sci 2014;131:40626/1 - 6. [233] Prabhakar RS, Freeman BD, Roman I. Gas and v apor s orption and p ermeation in p oly(2,2,4 - trifluoro - 5 - trifluoromethoxy - 1,3 - dioxole - c o - tetrafluoroethylene). Macromolecules 2004;37:7688 97. [234] Alentiev A, Yampolskii Y. Correlation of g as p ermeability and d iffusivity with s electivity: Orientations of the c louds of the d ata p oints and the e ffects of t emperature. Ind Eng Chem Res 2013;52:8864 74. [235] Alentiev AY, Yampolskii Y. Free volume model and trade off relations of gas permeability and selectivity in glassy polymers. J Memb Sci 2000;165:201 16. [236] Robeson LM, Smith ZP, Freeman BD, Paul DR. Contributions of diffusion and solubility selectivity to the upper bound analysis for glassy gas separation m embranes. J Memb Sci 2014;453:71 83. [237] Byun Y, Kim YT, Whiteside S. Characterization of an antioxidant polylactic acid - tocopherol, BHT and polyethylene glycol using film cast extruder. J Food Eng 2010;100:239 44. [238] Ambrosio - Martín J, López - Rubio A, José Fabra M, Angel López - Manchado M, Sorrentino A, Gorrasi G, et al. Synergistic effect of lactic acid oligomers and laminar graphene sheets on the barrier properties of polylactide nanocomposites o btained by the in situ polymerization pre - incorporation method. J Appl Polym Sci 2016;133:42661/1 - 11. [239] Bai H, Huang C, Xiu H, Zhang Q, Deng H, Wang K, et al. Significantly improving oxygen barrier properties of polylactide via constructing parallel - al igned shish - kebab - like crystals with well - interlocked boundaries. Biomacromolecules 2014;15:1507 14. [240] Samsudin H, Auras R, Mishra D, Dolan K, Burgess G, Rubino M, et al. Migration of antioxidants from polylactic acid films: A parameter estimation appr oach and an overview of the current mass transfer models. Food Res Int 2018;103:515 28. [241] Leng J, Kang N, Wang D - Y, Wurm A, Schick C, Schönhals A. Crystallization behavior of nanocomposites based on poly(L - lactide) and MgAl layered double hydroxides Unbiased determination of the rigid amorphous phases due to the crystals and the nanofiller. Polymer (Guildf) 2017;108:257 64. [242] Guo Y, Yang K, Zuo X, Xue Y, Marmorat C, Liu Y, et al. Effects of clay platelets 149 and natural nanotubes on mechanical proper ties and gas permeability of Poly (lactic acid) nanocomposites. Polymer (Guildf) 2016;83:246 59. [243] Delpouve N, Stoclet G, Saiter A, Dargent E, Marais S. Water barrier properties in biaxially drawn poly(lactic acid) films. J Phys Chem B 2012;116:4615 25 . [244] transport, and hydrolytic degradation in polylactide. Appl Biochem Biotechnol 2006;131:774 85. [245] Du A, Koo D, Theryo G, Hillmyer MA, Cairncross RA. Water transport and clust ering behavior in homopolymer and graft copolymer polylactide. J Memb Sci 2012;396:50 6. [246] Ebadi - Dehaghani H, Barikani M, Khonakdar HA, Jafari SH, Wagenknecht U, Heinrich G. On O 2 gas permeability of PP/PLA/clay nanocomposites: A molecular dynamic simu lation approach. Polym Test 2015;45:139 51. [247] Shogren R. Water vapor permeability of biodegradable polymers. J Environ Polym Degrad 1997;5:91 5. [248] Tsuji H, Okino R, Daimon H, Fujie K. Water vapor permeability of poly(lactide)s: Effects of molecular characteristics and crystallinity. J Appl Polym Sci 2006;99:2245 52. [249] Chaiwong C, Rachtanapun P, Wongchaiya P, Auras R, Boonyawan D. Effect of plasma treatment on hydrophobicity and barrier property of polylactic acid. Surf Coatings Technol 2010;204: 2933 9. [250] Bhatia A, Gupta RK, Bhattacharya SN, Choi HJ. Analysis of g as p ermeability c haracteristics of p oly( l actic a cid)/ p oly( b utylene s uccinate) n anocomposites. J Nanomater 2012;2012:249094/1 - 11. [251] Svagan AJ, Åkesson A, Cárdenas M, Bulut S, Knuds en JC, Risbo J, et al. Transparent films based on PLA and montmorillonite with tunable oxygen barrier properties. Biomacromolecules 2012;13:397 405. [252] Fortunati E, Peltzer M, Armentano I, Jiménez A, Kenny JM. Combined effects of cellulose nanocrystals and silver nanoparticles on the barrier and migration properties of PLA nano - biocomposites. J Food Eng 2013;118:117 24. [253] Shan Teo P, Shyang Chow W. Water v apour p ermeability of p oly(lactic acid)/ c hitosan b inary and t ernary b lends. KMUTNB Int J Appl Sci Technol 2014;7:23 7. [254] Duan Z, Thomas NL. Water vapour permeability of poly(lactic acid): Crystallinity and the tortuous path model. J Appl Phys 2014;115:64903 9. 150 [255] Qin Y, Wang Y, Yan W, Zhang Y, Hongli L, Yuan M. Effect of h exadecyl l actate as p lasticizer on the p roperties of p oly(L - lactide) f ilms for f ood p ackaging a pplications. J Polym Environ 2015;23:374 82. [256] Lim JH, Kim JA, Ko JA, Park HJ. Preparation and c haracterization of c omposites b ased on p olylactic a cid and b eeswax with i mproved w ater v apor b arrier p roperties. J Food Sci 2015;80:2471 7. [257] Yin Z, Zeng J, Wang C, Pan Z. Preparation and p roperties of c ross - l inked s tarch n anocrystals/ p olylactic a cid n anocomposites. Int J Polym Sci 2015;2015:454708/1 - 5. [258] R him J - W, Hong S - I, Ha C - S. Tensile, water vapor barrier and antimicrobial properties of PLA/nanoclay composite films. LWT - Food Sci Technol 2009;42:612 7. [259] Park S - H, Lee HS, Choi JH, Jeong CM, Sung MH, Park HJ. Improvements in barrier properties of p oly(lactic acid) films coated with chitosan or chitosan/clay nanocomposite. J Appl Polym Sci 2012;125:E675 80. [260] Jamshidian M, Tehrany EA, Desobry S. Release of synthetic phenolic antioxidants from extruded poly lactic acid (PLA) film. Food Control 201 2;28:445 55. [261] Du A, Gelves GA, Koo D, Sundararaj U, Cairncross R. Water t ransport in p olylactide and p olylactide/ m ontmorillonite c omposites. J Polym Environ 2013;21:8 15. [262] Sanchez - Garcia MD, Gimenez E, Lagaron JM. Morphology and barrier propertie s of solvent cast composites of thermoplastic biopolymers and purified cellulose fibers. Carbohydr Polym 2008;71:235 44. [263] Bouakaz BS, Pillin I, Habi A, Grohens Y. Synergy between fillers in organomontmorillonite/graphene - PLA nanocomposites. Appl Clay Sci 2015;116 117:69 77. [264] Rhim J - W, Mohanty AK, Singh SP, Ng PKW. Effect of the processing methods on the performance of polylactide films: Thermocompression versus solvent casting. J Appl Polym Sci 2006;101:3736 42. [265] Boesel LF, De Geus M, Thöny - M eyer L. Effect of PLA crystallization on the structure of biomimetic composites of PLA and clay. J Appl Polym Sci 2013;129:1109 16. [266] Fortunati E, Peltzer M, Armentano I, Torre L, Jiménez A, Kenny JM. Effects of modified cellulose nanocrystals on the b arrier and migration properties of PLA nano - biocomposites. Carbohydr Polym 2012;90:948 56. 151 [267] Wang Y, Qin Y, Zhang Y, Yuan M, Li H, Yuan M. Effects of N - octyl lactate as plasticizer on the thermal and functional properties of extruded PLA - based films. I nt J Biol Macromol 2014;67:58 63. [268] Yu HY, Yang XY, Lu FF, Chen GY, Yao JM. Fabrication of multifunctional cellulose nanocrystals/poly(lactic acid) nanocomposites with silver nanoparticles by spraying method. Carbohydr Polym 2016;140:209 19. [269] Tsuj i H, Tsuruno T. Water v apor p ermeability of p oly(L - lactide)/Poly(D - lactide) s tereocomplexes. Macromol Mater Eng 2010;295:709 15. [270] Jamshidian M, Tehrany EA, Imran M, Akhtar MJ, Cleymand F, Desobry S. Structural, mechanical and barrier properties of act ive PLA - antioxidant films. J Food Eng 2012;110:380 9. [271] Chinsirikul W, Rojsatean J, Hararak B, Kerddonfag N, Aontee A, Jaieau K, et al. Flexible and t ough p oly(lactic acid) f ilms for p ackaging a pplications: p roperty and p rocessability i mprovement by e ffective r eactive b lending. Packag Technol Sci 2015;28:741 59. [272] Duan Z, Thomas NL, Huang W. Water vapour permeability of poly(lactic acid) nanocomposites. J Memb Sci 2013;445:112 8. [273] Suyatma NE, Copinet A, Tighzert L, Coma V. Mecha nical and b arrier p roperties of b iodegradable f ilms m ade from c hitosan and p oly( l actic a cid) b lends. J Polym Environ 2004;12:1 6. [274] Prapruddivongs C, Sombatsompop N, Jayaraman K. Effect of organoclay incorporation on mechanical, barrier and thermal pro perties and anti - bacterial performance of PLA and PLA composites with triclosan and wood flour. Polym Polym Compos 2014;22:643 52. [275] Girdthep S, Worajittiphon P, Molloy R, Leejarkpai T, Punyodom W. Formulation and characterization of compatibilized pol y(lactic acid) - based blends and their nanocomposites with silver - loaded kaolinite. Polym Int 2015;64:203 11. [276] Xu Z, Su L, Jiang S, Gu W, Peng M, Wang P. Crystallization behavior and water vapor permeability of poly(lactic acid) nanocomposite under osc illatory shear. J Appl Polym Sci 2015;132:42321/1 - 8. [277] Cocca M, Lorenzo ML Di, Malinconico M, Frezza V. Influence of crystal polymorphism on mechanical and barrier properties of poly(L - lactic acid). Eur Polym J 2011;47:1073 80. [278] Cava D, Gimenez E, Gavara R, Lagaron JM. Comparative p erformance and b arrier p roperties of b iodegradable t hermoplastics and n anobiocomposites versus PET for f ood p ackaging a pplications. J Plast Film Sheeting 2006;22:265 74. 152 [279] De Santis F, Gorrasi G, Pantani R. A spectroscopic approach to assess transport properties of water vapor in PLA. Polym Test 2015;44:15 22. [280] Davis EM, Minelli M, Baschetti MG, Sarti GC, Elabd YA. Nonequilibrium sorption of water in polylactide. Macromolecules 2012;45:7486 94. [281] Davis EM, Theryo G, Hillmyer MA, Cairncross RA, Elabd YA. Liquid water transport in polylactide homo and graft copolymers. ACS Appl Mater Interfaces 2011;3:3997 4006. [282] Auras R, Harte B, Selke S. Sorption of ethyl acetate and d - limonene in poly(lactide) pol ymers. J Sci Food Agric 2006;86:648 56. [283] Leelaphiwat P, Auras RA, Burgess GJ, Harte JB, Chonhenchob V. Preliminary quantification of the permeability, solubility and diffusion coefficients of major aroma compounds present in herbs through various plas tic packaging materials. J Sci Food Agric 2018;98:1545 53. [284] Domenek S, Pronnier M, Guinault A, Plessis C, Ducruet V. The sorption of aroma compounds influences the thermomechanical properties of PLA. Biopol2007, Alicante, Spain: 2007. [285] Salazar R, Domenek S, Ducruet V. Aroma Sorption in Polylactide. 5th Int ernational Symp osium on Food Packag ing , Berlin, Germany: 2012. [286] Tsuji H, Sumida K. Poly(L - lactide): V . E ffects of storage in swelling solvents on physical properties and structure of poly(L - lactide). J Appl Polym Sci 2001;79:1582 9. [287] Lunt J. Large - scale production, properties and commercial applications of polylactic acid polymers. Polym Degrad Stab 1998;59:145 52. [288] Franz R, Störmer A. Migration of p lastic c onstituents. In: P iringer OG, Baner AL, editors. Plast ic p ackag ing: Interact ion with f ood and p harm aceuticals. New York: John Wiley & Sons; 2008, p. 349 415. [289] Selke SEM, Culter JD, Hernandez RJ. Plastics p ackaging - Properties, p rocessing, a pplications and r egulations. 2nd ed. Cincinnati, OH: Hanser Gardner Publications Inc.; 2004. [290] De Meulenaer B. Predictive m odeling and r isk a ssessment. In: Costa R, Kristbergsson K, editors. Predict ion m odel ing and r isk a ssess ment . vol. 4, Boston, MA: Springer US; 2009, p. 139 51. [291] Manavitehrani I, Fathi A, Badr H, Daly S, Negahi Shirazi A, Dehghani F. Biomedical a pplications of b iodegradable p olyesters. Polymers (Basel) 2016;8:20/1 - 32. 153 [292] Vert M, Schwach G, Engel R, Coudane J. Something new in the field of PLA/GA bioresorb able polymers? J Control Release 1998;53:85 92. [293] Ikada Y, Tsuji H. Biodegradable polyesters for medical and ecological applications. Macromol Rapid Commun 2000;21:117 32. [294] Tawakkal ISMA, Cran MJ, Bigger SW. Release of thymol from poly(lactic acid ) - based antimicrobial films containing kenaf fibres as natural filler. LWT - Food Sci Technol 2016;66:629 37. [295] Erdohan ZÖ, Çam B, Turhan KN. Characterization of antimicrobial polylactic acid based films. J Food Eng 2013;119:308 15. [296] Salmieri S, I slam F, Khan RA, Hossain FM, Ibrahim HMM, Miao C, et al. Antimicrobial nanocomposite films made of poly(lactic acid) cellulose nanocrystals (PLA CNC) in food applications part B: effect of oregano essential oil release on the inactivation of Listeria monoc ytogenes in mixed vegetables. Cellulose 2014;21:4271 85. [297] Gao J. Using Hansen s olubility p arameters (HSP s ) to d evelop a ntioxidant - p ackaging f ilm to a chieve c ontrolled r elease. MS Thesis. Michigan State University, 2014. [298] Higuchi T. Rate of r elease of m edicaments from o intment b ases c ontaining d rugs in s uspension. J Pharm Sci 1961;50:874 5. [299] Higuchi T. Mechanism of sustained action medication. Theoretical analysis of rate of release of solid drugs dispersed in solid ma trices. J Pharm Sci 1963;52:1145 9. [300] Ozalp Y, Ozdemir N, Hasirci V. Vancomycin release from poly(D,L - lactide) and poly(lactide - co - glycolide) disks. J Microencapsul 2002;19:83 94. [301] Lai M - K, Tsiang RC - C. Microencapsulation of acetaminophen into pol y(L - lactide) by three different emulsion solvent - evaporation methods. J Microencapsul 2005;22:261 74. [302] Xiang C, Taylor AG, Hinestroza JP, Frey MW. Controlled release of nonionic compounds from poly(lactic acid)/cellulose nanocrystal nanocomposite fibe rs. J Appl Polym Sci 2013;127:79 86. [303] Tsai DC, Howard SA, Hogan TF, Malanga CJ, Kandzari SJ, Ma JKH. Preparation and in vitro evaluation of polylactic acid - mitomycin C microcapsules. J Microencapsul 1986;3:181 93. [304] Freitas MN, Marchetti JM. Nimes ulide PLA microspheres as a potential sustained release system for the treatment of inflammatory diseases. Int J Pharm 2005;295:201 11. 154 [305] Jalil R, Nixon JR. Microencapsulation using poly(L - lactic Acid) IV: Release properties of microcapsules containing phenobarbitone. J Microencapsul 1990;7:53 66. [306] Jamshidian M, Tehrany EA, Desobry S. Antioxidants r elease from s olvent - c ast PLA f ilm: Investigation of PLA a ntioxidant - a ctive p ackaging. Food Bioprocess Technol 2013;6:1450 63. [307] Romero - Cano M., Vincent B. Controlled release of 4 - nitroanisole from poly(lactic acid) nanoparticles. J Control Release 2002;82:127 35. [308] Iñiguez - Franco F, Soto - Valdez H, Peralta E, Ayala - Zavala JF, Auras R, Gámez - Meza N. Antioxidant activity and diffu sion of catechin and epicatechin from antioxidant active films made of poly(L - lactic acid). J Agric Food Chem 2012;60:6515 23. [309] Mascheroni E, Guillard V, Nalin F, Mora L, Piergiovanni L. Diffusivity of propolis compounds in p olylactic acid polymer for the development of anti - microbial packaging films. J Food Eng 2010;98:294 301. [310] Manzanarez - López F, Soto - - t ocopherol from p oly(lactic acid) films, and its effect on the oxidative stability of soybean oil. J F ood Eng 2011;104:508 17. [311] Moreno - Vásquez MJ, Plascencia - Jatomea M, Ocaño - Higuera VM, Castillo - Yáñez FJ, Rodríguez - Félix F, Rosas - Burgos EC, et al. Engineering and antibacterial properties of low - density polyethylene films with incorporated epigallocatechin gallate. J Plast Film Sheeting 2017;33:413 37. [312] Soto - Valdez H, Auras R, Peralta E. Fabrication of poly(lactic acid) films with resveratrol and the diffusion of resveratrol into ethanol. J Appl Polym Sci 2011;121:970 8. [313] Samsudin H, Soto - Valdez H, Auras R. Poly(lactic acid) film incorporated with marigold flower extract (Tagetes erecta) intended for fatty - food application. Food Control 2014;46:55 66. [314] Zhang L, Long C, Pan J, Qian Y. A Dissolution - d iffusion m odel and q uantitati ve a nalysis of d rug c ontrolled r elease from b iodegradable p olymer m icrospheres. Can J Chem Eng 2006;84:558 66. [315] Lewis KJ, Irwin WJ, Akhtar S. Biodegradable poly(L - lactic acid) matrices for the sustained delivery of antisense oligonucleotides. J Contro l Release 1995;37:173 83. [316] Conn RE, Kolstad JJ, Borzelleca JF, Dixler DS, Filer LJ, Ladu BN, et al. Safety assessment of polylactide (PLA) for use as a food - contact polymer. Food Chem Toxicol 1995;33:273 83. 155 [317] Rouzes C, Leonard M, Durand A, Dellac herie E. Influence of polymeric surfactants on the properties of drug - loaded PLA nanospheres. Colloids Surfaces B Biointerfaces 2003;32:125 35. [318] nanospheres. II. Modelling of drug release. J Control Release 1999;60:169 77. [319] Pitt CG, Gratzl MM, Jeffcoat AR, Zweidinger R, Schindler A. Sustained d rug d elivery s ystems II: Factors a ffecting r elease r ates from p - caprolactone) and r elated b iodegradable p olyesters. J Pharm S ci 1979;68:1534 8. [320] Fernández A, Soriano E, Hernández - Muñoz P, Gavara R. Migration of a ntimicrobial s ilver from c omposites of p olylactide with s ilver z eolites. J Food Sci 2010;75:E186 93. [321] Martins IM, Rodrigues SN, Barreiro MF, Rodrigues AE. Rele ase of thyme oil from polylactide microcapsules. Ind Eng Chem Res 2011;50:13752 61. [322] Vega - Lugo AC, Lim LT. Controlled release of allyl isothiocyanate using soy protein and poly(lactic acid) electrospun fibers. Food Res Int 2009;42:933 40. [323] Leelar asamee N, Howard SA, Malanga CJ, Luzzi LA, Hogan TF, Kandzari SJ, et al. Kinetics of drug release from polylactic acid hydrocortisone microcapsules. J Microencapsul 1986;3:171 9. [324] U.S. Food and Drug Administration/Office of Food Additive Safety. FDA. Guidance for i ndustry: Preparation of p remarket s ubmissions for f ood c ontact s ubstances: Chemistry r ecommendations. Silver Spring, MD: 2017. [325] The European Commission. Commission Regulation (EU) No 10/2011 of 14 January 2011 on plastic materials and articles intended to come into contact with food. Off J Eur Union 2011;12:1 89. [326] Lin H, Freeman BD. Materials selection guidelines for membranes that remove CO2 from gas mixtures. J Mol Struct 2005;739:57 74. [327] Piringer O, Franz R, H uber M, Begley TH, McNeal TP. Migration from f ood p ackaging c ontaining a f unctional b e xperimental e valuation. J Agric Food Chem 1998;46:1532 8. [328] Baner A, Brandsch J, Franz R, Piringer O. The application of a predictive migrat ion model for evaluating the compliance of plastic materials with European food regulations. Food Addit Contam 1996;13:587 601. 156 CHAPTER 3 IN - SITU CHANGES OF THE THERM O - MECHANICAL PROPERT IES OF POLY(LACTIC ACID ) FILM IMMERSED IN A LCOHOL SOLUTIONS 3.1. Abstract Poly(lactic acid) (PLA) has properties suitable for several packaging, medical and agricultural applications . H owever conditions. In this study, the glass - rubber transition temperature s ( T g ) of PLA films were measured during immersion (i.e., in - situ ) in alcoho ls and alcohol aqueous solutions using a dynamic mechanical analysis technique . The T g of PLA decreased when immersed in alcohols . For aliphatic alcohols, the T g reduction became smaller a s the number of carbons (C1 C10) in the alcohol main chains increased . The F ox equation and the Flory - Huggins (FH) model based on the Hansen solubility parameters (HSP) were used to explain the T g reduction of PLA. The relationship explain ed the interactions between PLA and alcohol s with small molecules (C1 C 8 ), but bigger alcohols (C 9 C10) did not fit the prediction. The chemical isomerism in propanol (i.e., 1 - and 2 - propanol) did not affect the T g reduction. The T g reduction in 2 - propanol aqueous solutions was concentration dependent although the partition coeff icients based on the HSP and the FH interaction parameters did not fit t his relationship. The in - situ immersion of PLA in alcohol solutions could be used to evaluate the change in T g from the T g of dry PLA , but more work is needed to correlate the T g reduction with known parameters such as solubility parameters to predict T g in other solvents. 157 3.2. Introduction P oly(lactic acid) (PLA) a biodegradable, compostable, and renewable polymer is the most researched commercial bio - based polymer with properties that can be tailored by controlling its chiral structure composition [1,2] . Although PLA has been produced on a large scale since early 2005 [3] and it has been used in the medical, te xtile, agricultural, and packaging fields [1,4 6] , it is a relatively new polymer and its properties are not yet fully comprehended . One concern for PLA is that it s properties can be affected by the service and storage environments such as temperature, humidity, and contacted substances. Changes in properties of PLA have been observed when PLA is in contact with solvents and solutions [7,8] . Effects of water vapor and liquid water on PLA have been extensively evaluated due to the inherent hydrolytic behavior of the PLA chemical structure [8 14] . PLA may also be exposed to organic solvents and aqueous solutions, resulting in swelling as well as changes in its morphological structure, and therefore, thermal and mechanical properties [7,8,15 19] . E vidence of solvent - induced crystallization in PLA , where permeation of selected solvent s into the amorphous phase of PLA swells the polymer matrix an d promotes crystallization , has also been shown [8,16,20,21] . Despite extensive studies o n the effects of organic solvents and aqueous solutions on PLA morphologies, only a limited number of stud ies addressed the changes happening in - situ , that is when PLA is immersed in solvents. As the glass - rubber transition ( T g ) measurement is one of the experimental methods used to determine plasticization effects of solvents on polymeric materials [22] , s ome initial 158 information about in - situ T g in selected alcohols and aqueous solutions have b een reported by Iñiguez - Franco et al. [8] . Using dynamic mechanical analysis (DMA) immersion equipment, they reported that the measured T g of PLA dropped during immersion in ethanol and ethanol aqueous solutions and dem on strated that T g decreased when the concentration of ethanol in aqueous solutions increased. In contrast , T g of PLA measured after immersion in methanol and ethanol using a differential scanning calorimetry (DSC) technique [7,23] did not show significant reduction s . These findings suggest that PLA undergoes a glass - rubber transition during immersion that is not observed after immersion. A typical commercial grade PLA has a T g around 58 ±2 °C [24] , which is not much higher than room temperature in some regions of the world during summer month s . Therefore, it is likely that PLA will be used at a temperature near its T g , which may cause adverse effects on PLA properties . Additionally, it is possible that the solvents or solutions that PLA is in contact with will affect its T g . To design suitable applications for PLA, it is necessary to understand and evaluate its useful temperature range, exposure to vapors and solvents, and the actual T g during working conditions (i.e., in - situ ). Alcohols are well - known food simulants , common ly present in medicines and alcoholic beverages , and l ikely to be in contact with PLA . Thus, alcohols were chosen as the solvents for this study . T he aim of this study was to determine the in - situ - mechanical properties when immersed in alcoholic solutions , including selected a liphatic alcohols with the number of carbons in the main chain from C1 to C10 (i.e., 159 methanol to decanol) , alcohols with different isomers (i.e., 1 - and 2 - propanol), and aqueous solutions of 2 - propanol at different concentra tions. 3.3. Experimental 3.3.1. Film production MN , USA ) was produced in a Ran d castle Microtruder (Randcastle Extrusion Systems, Inc., Cedar Grove, NJ , USA ) with a 1.5875 cm diameter screw, 34 cc volume and 24/1 L/D ratio at 193 215 °C and 49 rpm. PLA pellets were dried at 60 °C for 24 h prior to film processing. Thickness of the produced amorphous neat PLA film was 20±5 µm. To minimize physical aging, the f ilm was stored in a freezer °C ) and precondition ed at 23 °C and 50% RH for 24 h immediately before use . The number average molecular weight ( M n ) , the weight average molecular weight ( M w ), and the polydispersity index (PDI) of the film measured by size - exclusion chromatography (Waters 1515 Isocratic HPLC pump, Waters 717plus autosampler, and Waters 2414 refractive index detector, Waters Corporation, Milford, MA , USA ) using tetrahydrofuran (Sigma - Aldrich, St. Louis, MO , USA ) as a solvent were 95 kg/mol , 171 kg/mol, and 1.8, respectively. 3.3.2. Solvents The solvents used for immersion tests were a series of selected aliphatic (straight - chain) alcohols with the number of carbon atom s C1 C10 ( i.e., methanol, ethanol, 1 - propanol, 1 - butanol, 1 - hexanol, 1 - octanol, 1 - nonanol, and 1 - decanol), a branched chain 160 alcohol C3 (2 - propanol), and water. A mixture of 50% ( v/v ) 2 - propanol in water was used as an alcoholic aqueous solution. Except for 1 - octanol and 1 - decanol , which were purchased from EMD - Millipore (Burlington, MA , USA ), solvents were purchased from Sigma - Aldrich (St. Louis, MO , USA ). All solvents were used as received. Table 3 - 1 shows the properties of the solvents. Table 3 - 1 S olvents used for immersion tests and their properties. Solvent #C Melting point a , K (°C) Boiling point a , K (°C) Molar mass a , g Molar volume b , mL/mol Purity a water - 273 (0) 373 (100) 18.0 18.0 methanol 1 338 (65) 32.0 40.6 ethanol 2 351 (79) 46.1 58.6 1 - propanol 3 370 (97) 60.1 75.1 2 - propanol 3 355 (82) 60.1 76.9 1 - butanol 4 391 (118) 74.1 92.0 1 - hexanol 6 430 (157) 102.2 125.2 1 - octanol 8 c 368 (195) c 130.2 158.2 1 - nonanol 9 488 (213) 144.3 174.9 1 - decanol 10 279 (6) c 505 (232) c 158.3 191.8 a From manufacturers , unless noted otherwise. b From HSPiP software [25] . c From CRC Handbook of Chemistry and Physics [26] . #C = number of carbon atoms in alc o hol main chain s . 161 3.3.3. Thermal and thermo - mechanical propert y measurements Differential scanning calorimetry (DSC) : A differential scanning calorimeter DSC Q100 (TA Instruments, New Castle, DE, US A ) calibrated with indium standards was used to measure the thermal properties of PLA film s before immersion . F ilm sample s of 5 10 mg w ere weighed and sealed in a TA Instruments hermetic aluminum pan and the thermal analysis was performed under nitrogen atmosphere with a flow rate of 70 mL/min. In the first heating cycle, the sample in a sealed pan was equilibrated at 2 0 °C , °C , °C , and remained isothermal for 1 min . Then , the system continued to the second heating cycle where the sample was cooled °C and finally heated to 200 °C. The temperature ramp rate for all the cycles was 10 °C/min. The s ampl es were tested at least in triplicate . T he results were analyzed with the TA Instruments Universal Analysis 2000 software version 4.5A and the T g values w ere determined from the inflection point at the step change in the DSC thermogram. Dynamic mechanical analysis (DMA) : A n RSA - G2 Solids Analyzer (TA Instruments, New Castle, DE , USA ) DMA unit was used to measure the therm o - mechanical properties of PLA films before and during ( in - situ ) immersion s . The RSA - G2 settings were as follows : loading gap 15 mm, max gap changes up 5 10 mm, max gap changes down 1 mm, preload force 100 g, strain 0.2%, frequency 1 Hz, and temperature ramp rate 5 °C/min. For each DMA experiment, a 10 mm x 50 mm film sample was mounted to the tension clamps. The DMA temperature ramp started at 25 °C for dry film and for the to 10 °C for the pure 162 alcohol immersions. Liquid nitrogen and air , connected to the RSA - orced convection oven , were used for coo ling and heating . The temperature ramp ended at approximately 20 °C above the temperature at peak tan(delta). For the in - situ immersion the desired starting temperature and poured into the immersion cell that contained a mounted sample. Once the required start ing temperature was reached, the temperature ramp started . The s amples were tested at least in triplicate . The storage modulus, loss modulus, and tan(delta) data were ob tained and the results were analyzed with TRIOS software version 4.5.0 (TA Instruments, New Castle, DE , USA ). T he peak tan(delta) value was recorde T g . 3.3.4. Statistical analysis Statistical significance was determined by analysis of variance (AN OVA) and the mean comparisons were determined by tests using SAS analytics software University Edition ( SAS Institute Inc., Cary, NC) at a significan ce level of 0.05. 3.4. Results and d iscussion 3.4.1. Pre - immersion properties Figure 3 - 1 a shows a typical DMA result for a dry PLA sample (pre - immersion) including the storage modulus, loss modulus, and tan(delta) (i.e., the ratio of the loss modulus to the storage modulus ) . The T g for pre - immersion PLA samples measured from peak tan(delta) was 62 .9 ± 1.0 °C . Figure 3 - 1 b shows a typical second heating scan of a pre - 163 immersion PLA sample from the DSC. As determined from the DSC, the T g , t he cold crystallization temperature ( T c ), and the melting temperature ( T m ) were 60. 6 ±0. 1 °C, 126.1 ± 0. 5 °C, and 150.9 ±0.3 °C, respectively. The crystalline fraction ( X C ), the mobile amorphous fraction ( X MAF ), and the rigid amorphous fraction ( X RAF ) determined from the DSC as [27] we re 0.5 ±0.2 %, 78.8 ± 5.1 %, and 20.7 ± 4.9 %, respectively . The low X C indicates that the film wa s practically amorphous and that the X RAF should not affect the evolution of T g [28,29] . The T g values of dry PLA film measured from DMA and DSC are statistically different at a significance level of 0.05 . Since the measured T g c an depend on the instrument settings such as the heating rate s and test frequen cies [30] , a slight difference is expected . 164 Figure 3 - 1 Typical p re - immersion (dry sample) test results of PLA from a) dynamic mechanical analysis ( DMA ) and b) differential scanning calorimetry ( DSC ) . T g , T c , and T m are glass - rubber transition, cold crystallization, and melting temperatures, respectively. 165 3.4.2. Effects of solvent sizes Figure 3 - 2 shows the in - situ immersion test results of PLA in a serie s of aliphatic alcohols where the tan(delta) is plotted against the temperature. Comparing the peak tan(delta) values, the T g of PLA is lower when immersed in alcohol with lower number of carbon atom s in the main chain , implying that smaller straight - chain alcoholic molecules can diffuse faster throu gh the free volume region of PLA. This agrees with a general observation that an increase in the size of the compound in a homologous series (i.e., straight - chain alcohols with increasing number of carbons) results in a decrease in the diffusion coefficien t through a polymer matrix [31] . Interactions of the alcohol molecules with PLA ca use swelling and plasticization of the PLA matrix, leading to an increase in mobility of the PLA chain and thus a decrease in the T g . Trailing peaks are observed at around 64 °C from tan(delta) of alcohols C8 C10 . While the se peaks may be attributed to par tially plasticized PLA films showing another T g value close to the T g of the dry film , they could be due to the sample and instrument limitations . At the temperature range where these peaks appeared, the film was softened and the elongation of the film was beyond the maximum allowed gap between the tension clamps, which was limited by the enclosed oven. Using thicker film samples m ight help to r educe the film elongation to within the restricted gap, but th ick films could result in a non uniform distribution of the solvent molecules within the PLA matrix. Additional tests are needed to validate these peaks and to determine a practical film thickness that best compromises between the film elongation and the solvent distribution issues . Furt hermore, the high temperature tests were not conducted in alcohols C1 C6 since most of the solvents would be evaporated. 166 Figure 3 - 2 Tan(delta) of PLA film s in different aliphatic alcohols as a function of t emperature. The drop in the T g from the T g of dry PLA is plotted in Figure 3 - 3 to show the trend as the number of carbon atom s and molecular volumes of the solvents changed. More discussion on the changes in the T g is provided in Section 3.4.4 . 167 Figure 3 - 3 C hanges in the T g values of the in - situ immersed PLA film s in aliphatic alcohols ( circle marker s , showing average values and standard deviation bars with ) from the T g of dry PLA ( horizontal dash line) as a function of number of carbons and molecular volumes of the solvents. V alues shown below the circle markers are percent T g reduction in Celsius from the T g of dry PLA . The T g of PLA immersed in 1 - decanol is statistically higher than the T g of dry PLA. This could be due to the high viscosity of 1 - decanol as well as different thermal conductivit y of the liquid from that of the air, i.e., t he condit ions when the film was heated in liquid w ere different from when the dry film was directly exposed to the heated air. Since alcohols larger than 1 - decanol are solid at room temperature, they were not tested. However, if the instrument setup for heating sol id alcohols is possible, testing PLA in alcohols with larger molecules is recommended to verify the increase in T g . Additional experiments were conducted to compare the T g of PLA after immersion and detailed discussion is provided in Appendix A . The results show varying values of post - immersion T g (i.e., the film samples were immersed in solvents, wiped 168 dry, and measured for T g ). These findings emphasize that the post - immersion T g measurement s may not capture the actual T g during the time PLA is exposed to the solvents. 3.4.3. Effects of branching and concentration S ome preliminary tests were performed to evaluate the factors affecting the T g reduction of PLA when immersed in branched - chain alcohols and alcohol aqueous solutions . The finding s from these selected alcohols may not be extrapolated to other alcohols, but this section should provide some initial understanding of PLA properties in these alcohols a s well as the factors affecting T g . Effects of branching of propanol : To compare the effects of the solvent chemical structures on T g , 1 - propanol ( a straight - chain alcohol ) and 2 - propanol ( a branched - chain alcohol ) were selected as solvents. These two alcohols are structural isomers with 3 carbon atoms and the same chemical formula. For typical alcohols with small number of carbons, a linear alcohol is more tightly packed than its branched isomer(s) [32] , e.g., the molecular volumes of 1 - and 2 - propanol are 75.1 and 76.9 mL/mol, respectively. However, Figure 3 - 4 shows that the T g value s of PLA when immersed in 1 - propanol and 2 - propanol we re not different. The result suggest s that the branched and shorter chain in 2 - propanol does not affect the solvati on of PLA and the reduction of T g . Also, the difference in mol ecular volumes of 1 - and 2 - propanol may not be large enough to result in different T g values. 169 Figure 3 - 4 Tan(delta) of PLA film in 1 - propanol and 2 - propanol as a function of temperature. Numbers #1 and #2 show replicates of each experiment. Effects of solvent concentrations : The results for PLA in - situ immersion in water, 2 - propanol, and 50% ( v/v ) 2 - propanol aqueous solution are shown in Figure 3 - 5 . Compared to the T g of PLA film before immersion, the reduction in T g is largest in pure (100%) 2 - propanol and the T g reduction becomes smaller as the concentration of 2 - propanol in water decreases. This finding may imply concentration dependency of the mass transfer of 2 - propanol in PLA . These T g reduction trends are in good agreement with the results previously report ed by Iñiguez - Franco et al. [8] that T g of PLA dropped from 60 °C before immersion to 36 °C when immersed in 50% ethanol and that T g decreased when the concentration of ethanol in aqueous solutions increased. However, a linear relationship was not prominent as can be seen from the inset in Figure 3 - 5 . F urther testing is needed to identify the difference in behavior s of PLA film in different alcohol and water solutions. 170 Figure 3 - 5 Tan(delta) of PLA film in water, 2 - propanol and 50% ( v/v ) 2 - propanol aqueous solution as a function of temperature. The inset shows T g as a function of 2 - propanol fraction with a linear trendline. 3.4.4. Modelling relationship between the solvent molecules and the changes in the T g of PLA The T g values measured by DMA pre - immersion and in - situ immersion in alcohols, alcohol solutions, and water are summarized in Table 3 - 2 . Statistical comparisons of the T g values in straight - chain alcohols, mark ed with lowercase letters, show that T g values are different in different alcohols. The T g values in 1 - and 2 - propanol are not statistically different, as marked by the same uppercase letter. Finally, comparisons of the T g values at dry conditions and in 50% 2 - propanol, 100% 2 - propanol, and water, mark ed with Greek letters, show that all T g values we re statistically different. 171 Table 3 - 2 A summary of T g of PLA film sampl e immersed in different solvents. Solvent T g , ° C T g reduction, % none (no immersion) 62. 9 ±1. 0 a, - methanol 14.3±0.9 b 77 ethanol 25.3±0.6 c 60 1 - propanol 29.6±0.8 d , A 53 2 - propanol 30.0± 1 . 5 A, 52 50% ( v/v ) 2 - propanol 33.6±0.3 4 7 1 - butanol 34.3±0.4 e 45 1 - hexanol 41.0±0.1 f 3 5 1 - octanol 45.7±1.9 g 27 1 - nonanol 49.8±1.1 h 21 1 - decanol 64.0±0.5 i - 2 water 53.3± 0.5 15 Note: Uppercase, lowercase, and Greek letters indicate different comparisons based on Tukey HSD tests at a significan ce level of 0.05 . Values with the same letter(s) are not different. A relationship between the T g of a polymer and the plasticization effect of a low molecular weight compound is usually estimated based on the additivity of basic thermo - physical properties such as the Fox, Gordon - Taylor, or Kelley - Bueche equation. The prediction of the T g of PLA in different solvents by the Fox equation [33] is illustrated in Figure 3 - 6 . Detailed calculations are shown in Appendix B . Comparing the experimental and the predicted T g values, only a small fraction of the solvent was absorbed into PLA. The Fox equation prediction shows that none of the weight fractions of solvents in PLA exceed 0.1, with methanol having the highest weight fraction in PLA (0.08) and water having the lowest value (0.02). Other alcohols C2 C6 have weight fractions in PLA in the range of 0.05 0.06. The T g values of alcohols with C>6 are not available; therefore, the estimations for these alcohols were not included. 172 Figure 3 - 6 Glass transition temperatures ( T g ) prediction of PLA in alcohols. The lines show predicted T g of PLA being plasticized by different alcohols based on the Fox equation [33] , compared with the corresponding experimental T g values shown by markers in the same colors as the lines. The numbers C1 C 6 indicate the number of carbon atoms in the al cohol main chains and C3b denotes 2 - propanol. The reduction in the T g of PLA when immersed in alcohols may be explained by the interactions of alcohols with PLA where the small molecular weight alcohols plasticize the PLA matrix resulting in PLA segmental chain movements, and thus a lower value of the measured T g . Lindvig e t al. [34] proposed a Flory - Huggins (FH) model based on the Hansen solubility parameters ( HSP ) to assess the FH interaction parame ters, , of a solvent (denoted by subscript 1) and a polymer (denoted by subscript 2 ) . T he model can be expressed as shown in equation ( 1 ) where * is an empirical factor that needs to b e estimated fr om experimental data , is the molar volume of the solvent, R is the gas constant , T is the temperature in Kelvin, , , an d 173 are the HSP based on the contributions from the dispersion, polar, and hydrogen bonding, respectively. The HSP values for PLA and the solvents used for the immersion tests are listed in Appendix B.2 . The values used for our calculations were adjusted to the measured in - situ T g values. An arbitrary * value of 0.6 was chosen for the current prediction as previously demonstrated as a good estimation [34] . ( 1 ) The predicted interaction parameters between PLA and different solvents were plotted against the experimental T g from the in - situ immersions as shown in Figure 3 - 7 . The values are higher for smaller alcohols and lower for bigger alcohols (i.e., longer chain, higher number of carbons). However, for alcohols larg er than C8, the values do not decrease as the in - situ T g values increase. The trend corresponds well with the prediction from the Fox equation shown in Figure 3 - 6 . However, the parameters used for the calculation were obtained from different sources based on different experimental setups, so they may not directly be correlated to the established trend. Further investigations should be focused on validating the empirical values used for the prediction and confirm whether the pred iction is reliable. Additional tests with 1 - pentanol and 1 - heptanol to confirm this trend will be useful. 174 Figure 3 - 7 The glass transition temperature ( T g ) of PLA immersed in aliphatic alcohols as a function of the predicted interaction parameters ( 12 ) at T g . The numbers C1 C10 indicate the number of carbon atoms in the aliphatic alcohol main chains , W denotes water, and C3b denotes 2 - propanol. The fitted exponential decay is shown as a dash line . S imilar calculations based on the Fox equation for water and 2 - propanol in PLA ( Figure 3 - 8 ) show that water and 2 - propanol reach their corresponding experimental T g values at the solvent weight fractions of 0.02 and 0.06, respectively. 175 Figure 3 - 8 Glass transition temperatures ( T g ) prediction of PLA in 2 - propanol and water. The lines show predicted T g of PLA being plasticized by water and 2 - propanol based on the Fox equation [33] , compared with the corresponding experimental T g shown by markers. The partition coefficients K of solute (denoted by subscript i ) in liquid (denoted L ) and PLA (denoted P ) can be estimated from the interaction parameters (Equation 1 ), as shown in Equation 2 ( 2 ) w here is the ratio of the liquid volume to the solute volume . K of 2 - propanol between water and PLA is 88 at infinite dilution . The reciprocal of this value is K of 2 - propanol in PLA which is 0.0 1 , which is much lower than the predicted weight fraction (~0.06) from the Fox equation. More experiments at different 2 - propanol concentrations as well as in different aqueous solutions are needed to elucidate the effect of solvent concentrations. 176 3.5. Conclusi on PLA T g decreased when immersed in alcohols and alcohol aqueous solutions. The number of carbon atoms and the concentrations of alcohols in aqueous solutions directly affected the T g reduction (i.e., a large r alcohol induce s a smaller T g reduction). However, the chemical structure of alcohol did not affect the reduction in T g for different isomers of propanol. The trend in T g reduction of PLA in alcohols and alcohol aqueous solutions correspond well with the calculations based on the know n values of HSP for low molecular weight alcohol s C 8 , but this relation ship does not apply to higher molecular weight alcohols (C> 8 ) . Further experiments are required to establish the underlying phenomena for the in - situ PLA immersion in alcohol solutions as well as to predict T g when immersed in other solutions . 177 APPENDICES 178 Appendix A : Post - immersion properties T he post - immersion experiments at different temperatures and immersion durations were performed to investigate the influence of s olvent s on the thermal and thermo - mechanical properties of PLA films using dynamic mechanical analysis ( DMA ) and differential scanning calorimetry ( DSC ) . A .1 Sample preparation s For post - immersion at 23 °C, the film samples were cut to a required size of 10 mm x 50 mm for the DMA or a required weight of 5 10 mg for the DSC and immersed in 50 mL glass vials. Each vial contain ed a solvent with a film sample being held in place by a stainless - steel wire and glass beads to ensure contacts with the solvent on both sides of the film [35] . The vial wa s then closed with a plastic cap and stored at room temperature (23±1 °C) for different durations ( e.g., 20 min, 3 d, 6 d, 12 d) . For post - immersion at elevated temperatures, the film samples were immersed in the RSA - G2 immersion cell and the temperature was controlled within ±1 °C by an attached force convection oven for 2 0 min. The 20 min post - immersion tests were performed to simulate the duration of the in - situ immersion tests while the immersion temperatures were selected based on the expected T g of PLA from the in - situ immersion . For each post - immersion sample , o nce the required immersion duration was reached, the film was removed from the solvent and wiped dry with soft, low - lint tissues before testing. To preserve the state of the film after contact with the solvent, the film was tested without further drying or conditioning. The DMA and DSC measurements of post - immersion samples then 179 followed the pre - immersion sample (i.e., dry film) test procedures as described in Section 3.3.3 . A.2 Post - immersion results from DMA Figure A - 1 shows the T g values of the pre - immersion (dry, control film) and the post - immersion (wiped - dry film) from the DMA in the y - ax e s and the number of carbon atoms in alcohol main chains and their molecular volume in the x - axes. The T g for post - immersion in methanol (C1) , ethanol (C2) , butanol (C4), and 1 - nonanol (C9) at 23 °C f or 20 min did not change from the dry T g . For post - immersions in 1 - butanol (C4), T g reductions were observed after 6 d and 12 d . For 1 - nonanol, the size of the solvent might impede the sorption of the solvent into PLA matrix at 23 °C , as can be seen from no change in T g T g regardless of the immersion duration . Immersion e xperiments at elevated temperatures resulte d in crumpled samples which were difficult to wipe dry without damaging the samples, and thus were not tested in the DMA. 180 Figure A - 1 T g - immersion (wiped - dry) in different alcohols as a function of number of carbon atoms and molecular volume. The immersion temperature and duration were listed in the plot legend, with C represen ting number of carbo n atoms . A.3 Post - immersion results from DSC Post - immersion samples were also tested in the DSC and the T g values determined from the first heating cycle are summarized in Figure A - 2 . The results showed scattered data and unexpected high T g values (i.e., higher than T g of dry PLA) of PLA post - immersion in methanol (C1). While anti - plasticization effects have been cited as the c ause of unexpected changes in polymer properties including increases in T g of polymers in contact with low molecular weight species [36] , this is unlikely the case for PLA in methanol in this immersion study. The results a lso show T g reductions of PLA when immersed in alcohols at elevated temperatures , impl ying that both temperature and duration affected the drops in T g after immersion . However, the results from post - immersion are not consistent. 181 The DSC thermograms in Figure A - 3 show a widened transition range for PLA post - immersion in methanol and the evaluated T g values could be due to the unsteady state of the mass transfer period. Scattered results could be attributed to sample handling after immersion. The d ifferent amount s of solvent remaining in the film and on the film surface could affect the measured T g . The T g values from the second heating cycle after immersion shown in Figure A - 4 , however, did not vary by the type of solvent. Research by Sato et. al [7] showed a small variation of T g of post - immersion PLA films in alcohols at 35 °C and 24 h measured by DSC when they dried the film under a vacuum for 48 h at 70 °C prior to the DSC characterizations. They found the post - immersion T g values were not much different from the T g of dry PLA, but solvent - induced crystallization occurred aft er immersion . 182 Figure A - 2 T g of PLA from DSC post - immersion (wiped - dry) in different alcohols as a function of number of carbon atoms and molecular volume ; a) C1, C2, C4, and C9 at various temperatures and 20 min, b) C4 and C9 at 23 °C and various durations. The immersion temperature and duration were listed in the plot legend, with C represen ting number of carbon atoms . 183 Figure A - 3 T hermograms of the first heating cycle of dry (control) PLA film and PLA films after immersions in different alcohols at 23 °C for 20 min . C represents number of carbon atoms . Figure A - 4 Thermograms of the second heating cycle of dry (control) PLA film and PLA films after immersions in different alcohols at 23 °C for 20 min. C represents number of carbon atoms . 184 Appendix B : Detailed calculations B.1 Fox equation The Fox equation (Equation B - 1 ) was used for prediction of T g reduction when PLA film is exposed to different solvents. ( B - 1 ) where T g * is the T g of the mixture, w 1 and w 2 are weight fractions of component 1 and 2, respectively. T g 1 and T g 2 are T g of component 1 and 2, respectively. The T g values of alcohols used for the calculations are listed in Table B - 1 . Table B - 1 The glass transition temperature ( T g ) values of alcohols used for the Fox equation calculations. Solvent #C T g , K (°C) water a - 136 ( - 137) methanol b 1 102.6 ( - 170.4) ethanol b 2 96.9 ( - 176.1) 1 - propanol b 3 99.7 ( - 173.3) 2 - propanol c 3 115 ( - 158) 1 - butanol b 4 111.5 ( - 161.5) 1 - hexanol b 6 129.7 ( - 143.3) 1 - octanol 8 n/a 1 - nonanol 9 n/a 1 - decanol 10 n/a a From [37] , b from [38] , c from [39] , n/a = not available. #C = number of carbon atoms in alcohol main chains. 185 B.2 Hansen solubility parameters The values of , , and of the solvents used in this study are listed in Table B - 2 . These values were recalculated to account for the temperature where the in - situ immersion T g of PLA was observed based on equations B - 2 B - 4 using the HSPiP software [25] and the adjusted values are shown in Table B - 3 , where is the thermal expansion coefficient. The HSP for PLA were based on the values reported by Elangovan et al. [40] . The HSP values of water used for the calculation are from water 1% soluble (i.e., small amount of water) instead of the values for bulk water (with , , and values 15.5, 16.0, and 42.3, respectively) since the very high of bulk water resulted in a high interaction parameter , which was difficult to use to correlate the experimental results with other s olvents. ( B - 2 ) ( B - 3 ) ( B - 4 ) 186 Table B - 2 The Hansen solubility parameters (HSP) for PLA [40] and solvents used for immersion [25] and the calculated Flory - Huggins interaction parameters ( 12 ) . Solvent #C d , MPa 1/2 p , MPa 1/2 h , MPa 1/2 12 at 25 °C PLA - 17.6 5.9 6.5 - water - 15. 1 20.4 16.5 0.37 methanol 1 14.7 12.3 22.3 0.80 ethanol 2 15.8 8.8 19.4 0.68 1 - propanol 3 16.0 6.8 17.4 0.55 2 - propanol 3 15.8 6.1 16.4 0.53 1 - butanol 4 16.0 5.7 15.8 0 .55 1 - hexanol 6 15.9 5.8 12.5 0. 37 1 - octanol 8 16.0 5.0 11.0 0. 33 1 - nonanol 9 16.0 4.8 11.0 0. 35 1 - decanol 10 16.0 4.7 10.5 0. 33 #C = number of carbon atoms in alcohol main chains. 187 Table B - 3 The adjusted Hansen solubility parameters (HSP) for PLA and solvents used for immersion , the thermal expansion coefficients of liquid ( ), and the calculated Flory - Huggins interaction parameters ( 12 ) at the measured PLA in - situ immersion glass transition temperature ( T g ) . Solvent #C 1/ °C d * , MPa 1/2 p * , MPa 1/2 h * , MPa 1/2 12 at T g PLA - 0.00007 a 17.5 5.9 6.2 - water - 0.0009 b 14.6 20.1 15.7 0.32 methanol 1 0.0012 b 14.7 12.3 22.3 0.85 ethanol 2 0.0012 b 15.8 8.8 19.4 0.68 1 - propanol 3 0.0011 b 16.0 6.8 17.4 0.58 2 - propanol 3 0.0011 c 15.8 6.1 16.4 0.53 1 - butanol 4 0.0010 b 16.0 5.7 15.8 0.56 1 - hexanol 6 0.0009 d 15.9 5.8 12.5 0.37 1 - octanol 8 0.0008 d 16.0 5.0 11.0 0.32 1 - nonanol 9 0.0008 d 16.0 4.8 11.0 0.33 1 - decanol 10 0.0008 d 16.0 4.7 10.5 0.34 a From reference [2] , b From r eference [41] , c from reference [39] , d from reference [42] , * values adjusted to the PLA in - situ T g . #C = number of carbon atoms in alcohol main chains. 188 REFERENCE S 189 R EFERENCES [1] Castro - Aguirre E, Iñiguez - Franco F, Samsudin H, Fang X, Auras R. Poly(lactic acid) Mass production, processing, industrial applications, and end of life. Adv Drug Deliv Rev 2016;107:333 66. [2] Lim L - T, Aura s R, Rubino M. Processing technologies for poly(lactic acid). Prog Polym Sci 2008;33:820 52. [3] Datta R, Henry M. Lactic acid: recent advances in products, processes and technologies a review. J Chem Technol Biotechnol 2006;81:1119 29. [4] Jamshidian M, Tehrany EA, Imran M, Jacquot M, Desobry S. Poly - lactic acid: Production, applications, nanocomposites, and release studies. Compr Rev Food Sci Food Saf 2010;9:552 71. [5] Auras RA, Singh SP, Singh JJ. Evaluation of oriented poly(lactide) polymers vs. exis ting PET and oriented PS for fresh food service containers. Packag Technol Sci 2005;18:207 16. [6] Auras R, Harte B, Selke S. An overview of polylactides as packaging materials. Macromol Biosci 2004;4:835 64. [7] Sato S, Gondo D, Wada T, Kanehashi S, Nagai K. Effects of various liquid organic solvents on solvent - induced crystallization of amorphous poly(lactic acid) film. J Appl Polym Sci 2013;129:1607 17. [8] Iñiguez - Franco F, Auras R, Burgess G, Holmes D, Fang X, Rubino M, et al. Concurrent solvent induce d crystallization and hydrolytic degradation of PLA by water - ethanol solutions. Polymer (Guildf) 2016;99:315 23. [9] Iñiguez - Franco F, Auras R, Rubino M, Dolan K, Soto - Valdez H, Selke S. Effect of nanoparticles on the hydrolytic degradation of PLA - nanocomp osites by water - ethanol solutions. Polym Degrad Stab 2017;146:287 97. [10] Beltrán FR, de la Orden MU, Lorenzo V, Pérez E, Cerrada ML, Martínez Urreaga J. Water - induced structural changes in poly(lactic acid) and PLLA - clay nanocomposites. Polymer (Guildf) 2016;107:211 22. [11] Tsuji H, Saeki T, Tsukegi T, Daimon H, Fujie K. Comparative study on hydrolytic degradation and monomer recovery of poly(L - lactic acid) in the solid and in the melt. Polym Degrad Stab 2008;93:1956 63. [12] Tsuji H. Hydrolytic degradat ion. In: Auras RA, Lim L, Selke SEM, Tsuji H, editors. Poly(lactic acid): Synthesis, Structures, Properties, Processing, and Applications, Hoboken, NJ: Wiley & Sons, Inc; 2010, p. 345 81. 190 [13] Cam D, Hyon S, Ikada Y. Degradation of high molecular weight po ly(L - lactide) in alkaline medium. Biomaterials 1995;16:833 43. [14] Rodriguez EJ, Marcos B, Huneault MA. Hydrolysis of polylactide in aqueous media. J Appl Polym Sci 2016;133:1 11. [15] Vayer M, Vital A, Sinturel C. New insights into polymer - solvent affini ty in thin films. Eur Polym J 2017;93:132 9. [16] Naga N, Yoshida Y, Inui M, Noguchi K, Murase S. Crystallization of amorphous poly(lactic acid) induced by organic solvents. J Appl Polym Sci 2011;119:2058 64. [17] Emerson JA, Toolan DTW, Howse JR, Furst EM , Epps TH. Determination of solvent - polymer and polymer - polymer Flory - Huggins interaction parameters for poly(3 - hexylthiophene) via solvent vapor swelling. Macromolecules 2013;46:6533 40. [18] Tsuji H, Sumida K. Poly (L - lactide): V . Effects of storage in swelling solvents on physical properties and structure of poly (L - lactide). Polymer (Guildf) 2001;79:1582 9. [19] Gironi F, Frattari S, Piemonte V. PLA chemical recycling process optimization: PLA solubilization in organic solvents. J Polym Environ 2016;24 :328 33. [20] Gualandi C, Govoni M, Foroni L, Valente S, Bianchi M, Giordano E, et al. Ethanol disinfection affects physical properties and cell response of electrospun poly(L - lactic acid) scaffolds. Eur Polym J 2012;48:2008 18. [21] Tsai W - C, Hedenqvist MS, Laiback Å, Melin H, Ngo M, Trollsås M, et al. Physical changes and sorption/desorption behaviour of amorphous and semi - crystalline PLLA exposed to water, methanol and ethanol. Eur Polym J 2016;76:278 93. [22] Immergut EH, Mark HF. Principles of plasticization. Plasticization and Plasticizer Processes, Washington DC: American Chemical Society; 1965, p. 1 26. [23] Gondo D, Wada T, Kanehashi S, Sato S, Nagai K. Effects of alcohol solvent - induced crystallization on biodegradable poly(lactic acid) film. J Packag Sci Technol Japan 2011;20:501 11. [24] Fambri L, Migliaresi C. Crystallization and thermal properties. In: Auras RA, Lim L, Selke SEM, Tsuji H, editors. Poly(lactic acid): Synthesis, Structures, Properties, Processing, and Applications, Hob oken, NJ: Wiley & Sons, Inc; 2010, p. 113 24. [25] HSPiP [Computer Software] version 5.1.08 2018. [26] Rumble JR. Physical constants of organic compounds. CRC Handbook of Chemistry and Physics, 76th edition. vol. 53, 1996, p. 504 504. 191 [27] Study of crystalline and amorphous phases of biodegradable poly(lactic acid) by advanced thermal analysis. Polymer (Guildf) 2009;50:3967 73. [28] Delpouve N, Arnoult M, Saiter A, Dargent E, Saiter J - M. Evidence of two mobile amorphous phases in semicrysta lline polylactide observed from calorimetric investigations. Polym Eng Sci 2014;54:1144 50. [29] Fernandes Nassar S, Guinault A, Delpouve N, Divry V, Ducruet V, Sollogoub C, et al. Multi - scale analysis of the impact of polylactide morphology on gas barrier properties. Polymer (Guildf) 2017;108:163 72. [30] Yong AXH, Sims GD, Gnaniah SJP, Ogin SL, Smith PA. Heating rate effects on thermal analysis measurement of Tg in composite materials. Adv Manuf Polym Compos Sci 2017;3:43 51. [31] Rogers CE. Permeation of gases and vapours in polymers. In: Comyn J, editor. Polymer Permeability, Dordrecht: Springer Netherlands; 1985, p. 11 73. [32] Walz M - M, Werner J, Ekholm V, Prisle NL, Öhrwall G, Björneholm O. Alcohols at the aqueous surface: chain length and isomer effe cts. Phys Chem Chem Phys 2016;18:6648 56. [33] Fox TG, Flory PJ. Second - order transition temperatures and related properties of polystyrene. I. Influence of molecular weight. J Appl Phys 1950;21:581 91. [34] Lindvig T, Michelsen ML, Kontogeorgis GM. A Flor y - Huggins model based on the Hansen solubility parameters. Fluid Phase Equilib 2002;203:247 60. [35] ASTM. Standard Test Method for Two - Sided Liquid Extraction of Plastic Materials Using FDA Migration Cell. ASTM International 2018;D4754 - 18:1 5. [36] Rahman MS, Roos YH. Glass Transition and Phase Transitions in Food. West Sussex, UK: John Wiley & Sons, Ltd; 2017. [37] Velikov V. The glass transition of water, based on hyperquenching experiments. Science (80 - ) 2001;294:2335 8. [38] Lesikar A V. On the self - a ssociation of the normal alcohols and the glass transition in alcohol - alcohol solutions. J Solution Chem 1977;6:81 93. [39] Ramos MA, Kabtoul B, Hassaine M. Calorimetric and thermodynamic study of glass - forming monohydroxy alcohols. Philos Mag 2011;91:1847 56. [40] Elangovan D, Nidoni U, Yuzay IE, Selke SEM, Auras R. Poly (L - lactic acid ) metal organic framework composites. Mass transport properties. Ind Eng Chem Res 2011;50:11136 42. 192 [41] nd Petrochemical Plants, Volume 1. 4th ed. Oxford, UK: Gulf Professional Publishing; 2006. [42] Matsuo S, Makita T. Volumetric properties of 1 - alkanols at temperatures in the range 298?348 K and pressures up to 40 MPa. Int J Thermophys 1989;10:885 97. 193 CHAPTER 4 O VERALL CONCLUSION AN D RECOMMENDATIONS FO R F UTURE WOR K 4.1. Overall conclusion The amount of research on p oly(lactic acid) (PLA) regarding its properties has been increasing in the past decades ; however, research focusing on its mass transfer properties is scarce . Many of the reported mass transfer properties of PLA in literature were parts of PLA characterizations , and thus lack s ystematic evaluation of the barrier properties . Chapter 2 of this dissertation provides a comprehensive, systematic , and critical review of the experimental data of mass transfer properties of PLA to gases, vapors, and organic compounds. A recent finding of the three - phase structure in semicrystalline polymers , which consists of the crystalline , mobile amorphous, and restricted (or rigid) amorphous fractions contradicts a traditional and simple two - phase structure (i.e., crystalline and amorphous fractions) . This three - phase model helps to explain unexpected barrier properties such as an increas e in gas permeability in PLA when the degree of crystallinity increases [1] , which is counterintuitive since crystalline regions are impermeable to gases and vapors. The de - densification of the restricted amorphous fraction in PLA [2] explains this b ehavior well. Lack of systematic experiments for PLA barrier properties assessment , especially for organic compounds and vapors , was also addressed in C hapter 2 . This finding led us to propose a study for the interaction of PLA with organic solvents in the next chapter. 194 In Chapter 3, the in - situ immersion experiments of PLA in various alcohol solutions were conducted and reported. The glass transition temperature ( T g ) of PLA decreased from the T g of dry PLA when PLA wa s immersed in alcohol and alcohol aqueous solutions . T he highest T g reduction was in methanol, which is the smallest, lowest molecular weight aliphatic alcohol. The changes in T g bec a me smaller for the bigger, higher molecular weight aliphatic alcohols. The Hansen solubility parameters (HSP) and the Flory - Huggins (FH) interaction parameters [3,4] were then used for the prediction of T g reduction and were found to be useful for alcohols with the number of carbon atoms in the main chain up to C 8 . The T g reduction in bigger alcohols did not follow the prediction well. Isomers of propanol with straight and branched chains did not show any difference in T g reduction, but without further experiments th is finding alone may not imply that the location of hydroxyl group in alcohols or the packing of alcohol chain s does not affect the interactio n between PLA and alcohols . The concentration of 2 - propanol in water affected the T g reduction of PLA; the higher the 2 - propanol concentration, the larger the T g reduction. However, while previous work [5] showed a linear trend for concentration dependency of PLA in ethanol aqueous solutions, results for PLA in 2 - propanol solutions from this study deviated from linear ity . The use of the partition coefficient based on the HSP and FH parame ters did not explain the behavior of PLA in aqueous solution either. More work is needed to determine the underlying phenomena when PLA is immersed in a binary mixture. The measurement s of T g post - immersion, where the films were wiped dry before the tests , show ed that the PLA s during in - situ immersion and post - immersion c ould be different and c ould result in much different T g values . Thus, the in - situ immersion test is 195 recommended for the evaluation of the actual T g of PLA when it is in contact with solvents. 4.2. Recommendations for future work Based on the comprehensive literature review in Chapter 2, extensive research is essential to fill data gaps in the mass transfer properties of PLA. Overall, except for oxygen and moisture barriers, mass transfer of other gases and vapors in PLA have not yet been much investigated. E ven for oxygen and moisture that were commonly a ssessed for their barrier properties , not to mention other less researched gases and vapors, systematic studies on factors affecting their mass trans fer in PLA are lacking. Extrinsic f actors such as temperature and relative humid it y and intrinsic factors such as PLA crystallinity and L - and D - lactid e contents, as well as other factors such as modifications of PLA by incorporation of additives, nanoparticles , or blends of PLA with other polymers , must be evaluated in the aspect of how they affect PLA barrier properties to different gases and vapors. Attention should be paid to the three - phase structure of PLA, whether the PLA sample under investigation is affected by the three - phase structure behaviors (e.g. , de - densification of th e restricted amorphous fraction ) and whether the barrier properties should be explained based on the three - phase structure. To avoid the complications from the three - phase structure, amorphous PLA samples were used for the in - situ immersion of PLA in alcohol solutions experiments ( Chapter 3 ) . Further investigation s using semicrystalline PLA samples are recommended to evaluate whether the solvents and PLA interact differently in the 196 presence of crystallinity in PLA. Furthermore, because t he in - situ immersion experiments were conducted in selected solvent s with different melting and boiling points, the starting and ending points for the temperature ramp varied for different solvents. Even though this variation was assumed to have no effects on the test results, more investigation may be needed to validate the assumption. Additionally , effects on the visc osity of the solvents as well as whether the temperature in the immersion cell is uniform (i.e., whether the temperature read by the thermo couple and the tempera ture of the film in the solvent are the same) should be investigated. Our in - situ immersion results from 1 - decanol (C10) imply no further T g reduction in larger aliphatic alcohols . Nevertheless, due to the instrument limitation s , conducting experiments with alcohols that are solid at room temperature such as 1 - dodecanol (C12) w as not feasible. If instrumental setup allows for larger alcohols to be tested, it is recommended. Further investigations on the isomerism effects should be conducted , for example, on 1 - and 2 - butanol and isobutanol . Additional concentrations of 2 - propanol in water are required to fully understand the effect of the solvent concentrations to T g reduction. Aqueous solutions of other alcohols should also be investigated. The in - situ immersion experiments on other families of solvents such as ketones, esters, ethers, aldehydes, aromatic and non - aromatic hydrocarbons should be conducted and compared to the results f rom the alcohol family. In addition to gain ing more understanding of PL A behaviors when immersed in these solvents, correlating the changes in properties of PLA to known properties of solvents such as the molecular volume or the solubility parameters may help to confirm whether our prediction is useful. However, parameters us ed for the prediction should be evaluated carefully since values 197 from different sources or different testing methods can be significantly different and will affect the prediction results. Overall, there are many areas in mass transfer properties of PLA to be explore d . Even for a specific scope such as the in - situ immersion of PLA in alcohol solutions, there are still many questions that have not been answered. With the advance in technology and the knowledge sharing in the polymer science community, hopefully more accurate prediction s of PLA properties based on known parameters could be done to help extend PLA usage to different applications without the need for extensive experiments . 198 REFERENCES 199 R EFERENCES [1] Guinault A, Sollogoub C, Ducruet V, Domenek S. Impact of crystallinity of poly(lactide) on helium and oxygen barrier properties. Eur Polym J 2012;48:779 88. [2] Del Río J, Etxeberria A, López - Rodríguez N, Lizundia E, Sarasua JR. A PALS contribution to the supramolecular structure of poly(L - lactide). Macromolecules 2010;43:4698 707. [3] Lindvig T, Michelsen ML, Kontogeorgis GM. A Flory - Huggins model based on the Hansen solubility parameters. Fluid Phase Equilib 2002;203:247 60. [4] Hansen CM. Hansen Solubility Parameters. 2nd ed. Boca Raton, FL: CRC Press; 2007. [5] Iñiguez - Franco F, Auras R, Burgess G, Holmes D, Fang X, Rubino M, et al. Concurrent solvent induced crysta llization and hydrolytic degradation of PLA by water - ethanol solutions. Polymer (Guildf) 2016;99:315 23.