123-effects of aging on the thermomechanical properties of pla

Effects of Aging on the Thermomechanical Properties ofPoly(lactic acid)

Michael Niaounakis, Evangelia Kontou, Menelaos Xanthis

Section of Mechanics, Department of Applied Mathematical and Physical Sciences, National TechnicalUniversity of Athens, 5 Heroes of Polytechnion, Athens GR-15773, Greece

Received 8 February 2010; accepted 15 April 2010DOI 10.1002/app.32644Published online 27 July 2010 in Wiley Online Library (wileyonlinelibrary.com).

ABSTRACT: In this study, we examined the role of envi-ronmental parameters and physical structure in the agingprocess of poly(lactic acid) (PLA). The role of heating his-tory on the aging behavior of the material was also investi-gated. PLA samples with a D-content of 4.25% wereexposed to a relative humidity of 80% at three differenttemperatures, 20, 40, and 50C (below the glass-transitiontemperature of the material), at various aging periods of30, 60, 80, 100, and 130 days. Selected samples were sub-jected to two consecutive heating runs. The stability ofPLA was monitored by a number of techniques, includingsize exclusion chromatography, differential scanning calo-rimetry, dynamic mechanical analysis, and tensile meas-urements. The initial thermal processing (150C) of the

material resulted in an overall molecular weight reduction.A substantial lowering of properties was observed forPLA samples aged at 20C for 30 days. No further loss ofproperties was observed for samples aged up to 40C forseveral time intervals. A major portion (8090%) of theinduced changes in the tensile properties could be reversedafter drying. At 50C and 100 days of aging, a sharpdecrease in the overall properties was noticed. The resultsseem to confirm the earlier finding that PLA degradationdriven by hydrolysis needs a higher temperature (>50C) incombination with ample time to take place. VC 2010 WileyPeriodicals, Inc. J Appl Polym Sci 119: 472481, 2011

Key words: ageing; biodegradable; mechanical properties

INTRODUCTION

Poly(lactic acid) (PLA) is a (bio)degradable materialwith a wide range of medical, textile and packagingapplications. The (bio)degradability of PLA, though,has both negative and positive aspects. The positiveaspect is that (bio)degradation forms nonhazardousproducts when PLA polymers or articles are dis-carded or composted after their useful life is com-plete. The negative aspect is the degradation of PLApolymers during processing. Thus, the same proper-ties that make PLA polymers desirable as replace-ments for nondegradable petrochemical polymersalso create undesirable effects that must be over-come. For several applications, the degradation rateof PLA is still low compared to the waste accumula-tion rate. On the other hand, the fact that PLAdegrades slowly (over a period of several weeks upto ca. 1 year) is advantageous for some other appli-cations, as it leads to a relatively good shelf life.PLA degrades during thermal processing or under

hydrolytic conditions; this causes a reduction in themolecular weight that affects the final properties ofthe material, such as its mechanical strength.15 Thedegradation behavior depends strongly on the mo-

lecular weight and the crystallinity of the PLA.6,7

Previous studies have shown that degradation ofPLA is enhanced by an increase in the temperatureand relative humidity (RH).8 The temperature, in thepresence of oxygen, triggers thermal oxidation,whereas moisture promotes hydrolytic degradation.PLA degradation driven by hydrolysis needs ahigher temperature (>50C) to take place.Chemical hydrolysis of the hydrolytically unstable

backbone is the primary mechanism for the degrada-tion of the PLA polymer. Degradation occurs first bywater penetrating the bulk of the polymer andhydrolyzing the ester bonds, preferentially those inthe amorphous phase, and converting long chainsinto shorter water-soluble fragments.9

PLA has been known to degrade slowly becauseof its hydrophobic and semicrystalline structure,which does not allow fast water penetration.Although PLA is a hydrophobic polymer with greatwater resistance, differences in the RH affect themoisture content of the material.10 Water absorbedduring aging at low temperatures [less than theglass-transition temperature (Tg)] act as an effectiveplasticizer, increasing the molecular mobility ofPLA, which facilitates the aggregation of PLA chainsand results in faster nucleation and crystallization ofthe PLA molecules.10,11 At high temperatures (>Tg),the absorbed water causes hydrolysis of the esterbonds, preferentially in the amorphous phase andconverts long polymer chains into shorter ones,

Correspondence to: E. Kontou ([email protected]).

Journal of Applied Polymer Science, Vol. 119, 472481 (2011)VC 2010 Wiley Periodicals, Inc.

including oligomers and monomers. This explainsthe much faster hydrolysis rate of amorphous PLA(D,L) compared to semicrystalline PLA (L,L). The lac-tic acid oligomers, which result from hydrolysis, areknown to catalyze the degradation reaction.12 Thedegradation mechanisms can be affected by variousfactors, such as the chemical structure, molar massand distribution, purity (e.g., water content, low-mo-lecular-weight impurities, catalyst residues), morphol-ogy, shape, size, history of the specimen, and condi-tions under which the hydrolysis is conducted (e.g.,aging time, temperature, presence of water).13,14

The majority of previous art studies on the stabil-ity of PLA have been carried out in buffer solutions,primarily under physiological conditions, that is, attemperatures of 37C with a pH of 7.4.7,9,12,1526 Asmaller number of articles has been dedicated to theeffects of environmental conditions, such as humid-ity, pressure, UV radiation, temperature, agingtime,8,10,2730 and compost microorganisms.3133 Onlya fraction of the most recent studies have reportedon the effects of environmental conditions on themechanical properties of PLA.27,30,34,35 Only few ofthem have contained tensile measurements.7,27,35

PLA has a Tg value of 5060C. Depending on its

D-content and molecular weight, its cold crystalliza-tion temperature ranges between 95 and 110C,whereas its melting temperature (Tm) ranges from150 to 190C.36 The crystallizability of PLA is deter-mined by its chain regularity and mobility. As lacticacid is chiral, the regularity of PLA chains is signifi-cantly affected by the ratio of L components over Dcomponents;37 this leads to the suppression of thecrystallizability when a small fraction of D is presentin most L-chains. The PLA has insufficient flexibility,and plasticizers are used to increase its flexibility foruse as, for example, packaging wrap films, stretchfilms, and agricultural mulch films. However, thereis evidence that the mechanical properties of theplasticized material are eroded by aging.34,35 PLA isa slow-crystallizing polymer, like PET, and it haslong mold cycle time. It can be quenched into aquasi-amorphous state or crystallized upon anneal-ing and/or orientation. When crystallized quies-cently from unnucleated state, PLA generallybecomes opaque as a result of light scattering due tothe formation of large spherulites.36

PLA is, therefore, a thermosensitive material proneto morphological and chemical changes by variationsin the temperature, time, and water content, whichcan have a great impact on its thermomechanicalproperties. The inadequate thermal stability of PLAduring service puts restrictions on the range of possi-ble large-scale applications. In view of the scarcity ofavailable experimental results on the effects of envi-ronmental conditions on the mechanical propertiesand, especially, the tensile properties of PLA, the aim

of this study was to supplement the existing literaturewith extra measurements and analysis on the effectsof RH, aging time, and temperature (where the chosentemperature was below the Tg of the material) on themolar mass and thermomechanical properties of PLAhaving a D-content of 4.25%, as evidenced by sizeexclusion chromatography (SEC), differential scan-ning calorimetry (DSC), dynamic mechanical analysis(DMA), and tensile measurements.

EXPERIMENTAL

The polylactide resin (PLA) was supplied by Nature-Works LLC, Minnetonka, MN. The selected grade2002D had a D content of 4.25%, a residual monomercontent of 0.3%, and a density of 1.24 g/cm3. Thematerial in pellet form was dried at 45C for a mini-mum of 8 h before use in a desiccating dryer. Themelt processing of PLA was performed in a Bra-bender mixer at a temperature of 150C, whereas therotation speed of the screws was 40 rpm. The tem-perature for melt mixing was kept as low as possibleto minimize the effect of degradation. Hereafter, thematerial was compression-molded at 130C with athermopress and a special mold 2 mm thick. Thematerial was then cooled slowly to ambient tempera-ture. The PLA samples were put in an oven andexposed at 80 6 2% RH in an ambient saturated so-lution of NH4Cl, at three different temperatures, 20,40, and 50C (60.5C), and at time intervals of 0, 30,60, 80, 100, and 130 days. The aging temperaturerange was chosen to be below the Tg of PLA. Thevarious aged samples were annotated as PLA x/y,where x is the number of days and y is the tempera-ture (e.g., PLA 30/50); the unaged material, whichwas processed in the Brabender mixer and the ther-mopress, was denoted PLAproc. The raw material inpellet form was designated PLApellet.The evolution of the molecular weight of the PLA

and its crystallinity contents at different tempera-tures and aging time intervals were recorded bySEC and DSC, respectively. All aged and unagedsamples were dissolved in tetrahydrofuran withoutany insoluble residues being visible to the nakedeye; this suggested that no crosslinking took place.Calorimetric measurements were carried out with

a Setaram DSC 141 (Caluire, France) instrument,calibrated with an indium standard. Each PLA sam-ple (aged and unaged) was heated at a constantheating rate of 10C/min from 20 to 170C, and thethermogram was recorded. A supplementary proce-dure was performed on selected samples, namely,those that were exposed for 130 days at 20 and40C. After they were heated to 170C, the PLA sam-ples were isothermally held for 2 min, subsequentlycooled to 0C, held there for 5 min, and heated againat a rate of 10C/min to 170C. Thermograms of

THERMOMECHANICAL PROPERTIES OF POLY(LACTIC ACID) 473

Journal of Applied Polymer Science DOI 10.1002/app

both the first and second heating runs wererecorded. We calculated the degree of crystallinityby considering a melting enthalpy of 93.1 J/g for100% crystalline PLA.38 Three sheet specimens 2 mmthick were tested for every aging procedure. DSCsamples were taken from both the center and the tipof every PLA sheet, and we found that the DSCresults were repeatable, with an average scatterlower than 5%.The molecular weights and molecular weight distri-

butions of the PLA samples were determined by SEC(Milford, MA) with a Waters system composed of aWaters 1515 isocratic pump, a set of three l-Styragelmixed bed columns with a porosity range of 102106

A, and a Waters 2414 refractive-index (RI) detector (at40C), operated/controlled through Breeze software.Tetrahydrofuran, containing 3% (v/v) triethylamine,was the mobile phase and was used at a flow rate of1.0 mL/min at 30C. The setup was calibrated withlinear polystyrene standards having weight-averagemolecular weights in the range 1250 to 900,000 g/mol.The PLA samples were dissolved directly in the car-rier solvent overnight at concentrations of about 0.1%w/v and characterized the next day.DMA experiments were performed with a Perkin-

Elmer (Norwalk, CT) DMA 7e instrument. The modeof deformation applied was the three-point bendingsystem, and the mean dimensions of the sample pla-ques were 2 4 20 mm2. The temperature variedfrom 50 to 130C. We studied the temperature-de-pendent behavior by monitoring changes in the forceand phase angle, keeping the amplitude of oscillationconstant. The experiments were performed at a con-stant frequency of 1 Hz, and the heating rate was5C/min. The storage and loss moduli and loss tan-gent curves versus temperature were evaluated.Tensile measurements were performed with an Ins-

tron (Buckinghamshire, UK) 1121 type tester at roomtemperature. The dumbbell-type specimens were at agauge length of 30 mm, and the applied crossheadspeed was 0.2 mm/min. This value corresponded toan effective strain rate of 1.1 104 s1. The deforma-tion could be measured very accurately with an ex-perimental procedure, which was based on a noncon-tact method with a laser extensometer and wasdescribed in detail in a previous article.39 Therefore,the true stressstrain curves were then obtained up tothe breaking point. Five specimens were tested foreach aging procedure, and the standard deviationbetween the experimental data was lower than 65%.

RESULTS AND DISCUSSION

SEC

The raw material in pellet form had a number-aver-age molecular weight (Mn) of 197,000 and a polydis-

persity index (I) of 1.35. After melt processing, Mnwas reduced to 169,000 and I was equal to 1.25. Thedrop in the molecular weight was attributed to thethermal oxidation of the PLAproc sample duringprocessing in the Brabender at a relatively low tem-perature (150C).1,3,4

The degradation of PLA was tested by examina-tion of the molecular weight distribution after expo-sure to 80% RH at various temperatures, namely, 20,40, and 50C, for different time periods. The corre-sponding results of the molecular weight variationversus time are presented in Figure 1 as elutioncurves, and the Mn and I (Mw/Mn) values of thePLA samples measured by SEC are given in Table I.As shown in Table I, after aging at 20C for 130

days, PLA underwent little change in its molecularweight. Significantly lower molecular weights weremeasured for samples subjected to more intensedegradation treatments. This is exemplified in Figure2, which shows the elution curves of the PLA sam-ples aged at 40 and 50C for different time intervals.For comparison, the corresponding curve of theunaged sample (PLAproc) is also presented. Asshown in Figure 2, the elution peak shifted to alower molecular weight with increasing temperature.Furthermore, the elution peaks of the aged samplesdisplayed a higher contribution of low-molecular-weight fractions, which was attributed mainly tochain scission due to hydrolysis.As shown in Table I, the Mn of PLA decreased

dramatically when it was exposed for at least 80days to 50C at 80% RH; with a decrement on theorder of 80%; after 130 days at the same tempera-ture, the material was completely depolymerized.The reduction was much less pronounced at lowertemperatures, namely, 4.8% at 20C and 32% at 40Cafter 130 days. As shown in Table I, the molecular

Figure 1 Variation of the PLA molecular weight withtime at three different temperatures and 80% RH. Experi-mental data were obtained by SEC.

474 NIAOUNAKIS, KONTOU, AND XANTHIS


weight of PLA was reduced with greater aging timeat a given temperature. The effect of aging timebecame more pronounced at higher temperatures.With regard to I, no dependence on the environ-

mental conditions (aging time and temperature)was noticed, except for the temperature of 50C,where I 5.59 after 80 days.The degradation process was very slow at temper-

atures equal or below 40C, but the speed wasgreatly enhanced at 50C. The degradation rate wasquantified in terms of the average hydrolytic degra-dation rate constant (kt). The kt values were eval-uated with the assumption of an exponentialdecrease of Mn with the following equation:

7

lnMnt2 lnMnt1 ktt

where t is the degradation time, Mn(t2) and Mn(t1)are the number-average molecular weights at thehydrolytic degradation times t2 and t1, respectively.According to Figure 1, the Mn decrease rate was dif-ferent depending on the time period of aging. There-fore, two different values for the constant kt wereevaluated. Apart from this, an average value of ktfor this constant was calculated over the entire pe-riod of aging. The estimated kt values for the differ-

ent time intervals were 3.67 104 and 8.67 105days1 at 20C and 6.126 103 and 4.07 103days1 at 40C. The corresponding average values ofkt were 3.73 104, 2.96 103, and 20 103days1 at 20, 40, and 50C, respectively. The resultsshow that the degradation rate was higher at the be-ginning (first 30 days) for every temperature anddecreased significantly after 30 days. This findingwas in agreement with previous results.1,15,22

DSC

The DSC thermograms of the raw material (PLApellet),the unaged PLA samples (PLAproc), and the agedPLA samples under various environmental conditionsare presented in Figures 35. The DSC results, interms of Tg, Tm, heat of fusion, and correspondingcrystallinity content are summarized in Table II. Asshown in Figures 35, the Tg region appeared atabout 55C, and in most cases, it was followed by anendothermic enthalpy relaxation, which was attrib-uted to a secondary molecular reordering in theamorphous phase of the semicrystalline polymers.The Tg was defined from the endothermic peakposition.

TABLE IMn, Mw, and I Values of PLA Samples Measured by SEC Under Various

Aging Conditions

Time (days)

20C 40C 50C

Mn I Mw/Mn Mn I Mw/Mn Mn I Mw/Mn0 169,000 1.25 169,000 1.25 169,000 1.2530 165,000 1.25 145,000 1.55 68,000 1.5680 161,700 1.26 122,000 1.56 33,100 5.59130 161,000 1.26 115,000 1.56

The raw material in pellet form had an Mn value of 197,000 and an I value of 1.35.

Figure 2 SEC chromatograms of PLA aged for 30 days at80% RH and different temperatures.

Figure 3 DSC results for PLA aged at 20C and 80% RHfor various time periods.



As shown in Figures 3 and 4, at temperatures of20 and 40C, respectively, Tg shifted slightly to lowertemperatures with aging time. The dropping of Tgwas attributed to the plasticization of the amorphousphase by the absorbed water, which penetratedbetween the polymer chains and increased their mo-bility. Two mechanisms were thought to be involvedwith the shift of Tg. The first was attributed to theplasticizing effect of the absorbed water, and the sec-ond was attributed to the degradation effect of wateron the amorphous and crystalline fractions of thematerial. At 50C, the shorter chains that arose bythe degradation of PLA were reorganized in crystal-lites of lower melting points, which restricted thesegmental mobility in the amorphous region.As shown in Figures 35, the unaged sample

(PLAproc) showed a broad cold crystallization (exo-therm) area extending from Tg up to Tm. When coldcrystallization occurred, less perfect crystallites wereformed, which melted during the DSC heating run.The exothermic area was attributed to the reorgan-ization of the amorphous region into a crystallineone on account of the increased polymer chain flexi-bility and mobility as result of the thermal oxidationduring processing. However, the total cold crystalli-zation area was reduced by aging.The PLAproc sample also exhibited a bimodal

melting peak, in which the first melting peak (Tm1)was lower than the second one (Tm2). With time, theTm1 peak diminished, with this effect being morepronounced at 20 and 40C (Figs. 3 and 4), andeventually disappeared at 50C (Fig. 5), whereas themagnitude of the Tm2 peak increased, and the peaktemperature shifted slightly to a lower temperature.The diminishing of the Tm1 peak with time wasattributed to the reorganization of the crystals,which arose from the cold crystallization duringaging. Water absorbed during aging at temperatures

equal or lower than 40C and acting mainly as aplasticizer promoted either the growth or perfectionof already existing crystallites.10,40 The Tm2 peak wasattributed to the inherent crystallinity of the mate-rial. Alternatively, the Tm2 peak resulted from therecrystallization of the crystallites of the Tm1 peakaccording to the melt recrystallization model.41 Thismodel suggests that the small and/or imperfectcrystals changed successively to more stable crystalsthrough the melt-crystallization process. The Tm1and Tm2 peaks suppressed recrystallization, whichwas not revealed on the DSC thermograms. Theslight shift to lower temperatures of the Tm2 meltingpeak was attributed to the incurred degradation dur-ing the physical aging of the PLA samples at roomtemperature.Samples aged at 50C for 30 days showed an

increase in crystallinity (see Table II). The lower mo-lecular weight (Mn 68,000) and effect of interac-tions with water were likely to have enhanced themobility and, in combination with the formation of a

Figure 4 DSC results for PLA aged at 40C and 80% RHfor various time periods. Figure 5 DSC results for PLA aged at 50

C and 80% RHfor various time periods.

TABLE IIDSC Results

Sampletype

Tg(C)

Tm1(C)

Tm2(C)

DH*(J/g)

Crystallinity(%)

PLApellet 60.0 153.4 26.5 28.5PLAproc 55.4 156.9 165.9 33.8 36.3PLA 30/20 49.0 153.5 163.7 34.1 36.6PLA 60/20 49.2 153 162.2 32.0 34.3PLA 100/20 51.5 161.3 27.8 29.9PLA 30/40 58.0 156.0 165.7 28.7 30.8PLA 60/40 56.4 153.1 162.6 39.9 42.8PLA 100/40 54.0 150.0 159.8 38.7 41.6PLA 130/40 55.7 155.1 164.6 30.8 33.0PLA 30/50 54.0 154.0 165.1 35.4 38.0PLA 100/50 140.0 12.70 13.6

* Heat of fusion.



higher number of crystallization centers that couldhave originated from the degradation products,could have led to the formation of crystallites withlower melting points.15,42

The role of heating history on the aging behaviorof the material was also investigated and is shownin Figures 6 and 7, which present the DSC thermo-grams of the first and second heating runs, respec-tively, of PLA in pellet form, PLAproc, and theselected aged samples PLA/130/20 and PLA/130/40.The Tg of the second run was lower than the Tg of

the first run in all of the samples. This was attrib-uted to the difference in the crystallinity betweenthe two runs of the PLAproc sample. The higher crys-tallinity of the first run was considered to restrictthe segmental mobility in the amorphous region.

Tm and the crystallinity content of the second runwas always lower than the Tm of the first run in allof the tested PLA samples (cf. Figs. 6 and 7). Thefirst heating run of the aged samples showed twomelting peaks, which were reduced to one peak inthe second run, which corresponded to the lowermelting peak of the first run. The broad meltingpeak of the PLAproc sample of the first run shifted toa lower temperature and decreased in size. Particu-larly, the first heating run of the PLA pellets showedan endothermic peak (Tm 158C), which disap-peared in the second heating run, where no endo-thermic or exothermic peak was detected at all. Thisbehavior was in agreement with the reported slowcrystallization rate of high-molecular-weight PLA,which did not allow the development of crystallinedomains upon cooling.1,17 On the other hand, theappearance of an endothermic peak in the secondrun of the processed and aged samples was attrib-uted to the increased molecular flexibility as resultof the lower molecular weight, as observed by SEC,in combination with the enhanced mobility as resultof the absorbed water, which favored the develop-ment of crystallites. The slow crystallization of PLAwas a disadvantage for the processing of the mate-rial; very long cooling times of up to several minuteswere necessary to obtain partially crystalline mold-ings having heat deflection temperatures above theTg.The induced morphological changes would signifi-

cantly influence the materials performance duringaging, and any application (e.g., packaging) neededto take into account such morphological changesand their effect on the material properties.15

DMA

The dynamic mechanical relaxation behavior of theaged and PLAproc samples is presented by the tem-perature dependence of loss modulus in Figures 810. In these plots, the effect of aging time and tem-perature (80% RH) can be examined. In Figure 8, theeffect of aging time (30 days) is presented for threetemperatures, 20, 40, and 50C. Figure 9 exhibits theeffect of time for the same temperature of 20C,whereas Figure 10 reveals the same effect for thetemperature of 40C. A single glass transition wasobserved, which was about 10C higher than the Tgmeasured by DSC.As shown in the loss modulus plot of Figure 9, a

single peak at about 64C was observed for samplesaged at 20C and 80% RH for 30 and 130 days. Aslight shift of the peak to a lower temperature and areduction of its intensity was observed only for thesample aged for 130 days. The decrease in loss mod-ulus was correlated to the plasticization effect ofwater absorbed after 130 days of aging.

Figure 6 DSC results for the first heating scan of PLAaged for 130 days at 80% RH and two differenttemperatures.

Figure 7 DSC results for the second heating scan for PLAaged for 130 days at 80% RH and two differenttemperatures.



The samples aged at 40C and 80% RH alsoshowed a slight shift of a few degrees of the peak tolower temperatures with a more pronounced reduc-tion of its intensity after 130 days (Fig. 10).As shown in Figure 8, the position and magnitude

of the peak remained almost invariable for samplesaged for 30 days at temperatures up to 40C. Onlyat 50C was a shift of the Tg to lower temperaturesobserved accompanied by a broadening of the peak.This was an indication that at up to 40C, no sub-stantial structural changes took place in the material.

Tensile results

The tensile results of the unaged sample and PLAsamples exposed to 80% RH under different condi-tions of time and temperature are summarized in

Table III. The corresponding stressstrain curves areplotted in Figures 1113. In Figure 11, the tensilestressstrain curves are plotted to compare thePLAproc sample with the aged samples exposed to80% RH at three different temperatures for the sametime period of 30 days. Figures 12 and 13 show thevariation of the tensile curves at temperatures of 20and 50C, respectively, for various time periods. Allof the tensile results confirmed the general knowl-edge43 that PLA is a brittle material with a low pos-sible elongation.The unaged PLA sample displayed a yield stress

manifested as a gradual slope change, which fol-lowed the initial linear part of the stressstraincurve. Yielding occurred at a strain about 2.6%.Thereafter, PLAproc was driven to fracture at a strainequal to 3.5%. The tensile behavior of the PLA sam-ples exposed to various conditions of degradationwas gradually turned to lower values of Youngmodulus, tensile strength, and strain at break. After100 days of aging at a temperature of 50C, the ten-sile property decrement appeared to be the highest.

Figure 8 Loss modulus versus the temperature at the fre-quency of 1 Hz for PLA aged for 30 days at 80% RH anddifferent temperatures.

Figure 9 Loss modulus versus the temperature at the fre-quency of 1 Hz for PLA aged at 20C and 80% RH for var-ious time periods.

Figure 10 Loss modulus versus the temperature at thefrequency of 1 Hz for PLA aged at 40C and 80% RH forvarious time periods.

TABLE IIITensile Results

Sampletype

Youngsmodulus(GPa)

Tensilestrength(MPa)

Strain atbreak (%)

PLAproc 3.2 57.0 0.035PLA 30/20 3.0 49.0 0.018PLA 130/20 2.6 42.3 0.016PLA 30/40 2.6 40.5 0.016PLA 30/50 1.9 39.1 0.032PLA 100/50 1.5 20.0 0.013

Five specimens were tested for each aging procedure,and the standard deviation between experimental datawas lower than 65%.



Namely, decrements of 51.4% for the Youngs modu-lus, 65% for the tensile strength, and 67.5% for thestrain at break were observed. With regard to expo-sure at 40C for 30 days, the Youngs modulusdecreased 8%, whereas the tensile strengthdecreased 30%, which was a sufficiently lower deg-radation compared to that at 50C. Lower values, onthe order of a 20% decrement, were observed after130 days at 20C, whereas PLA embrittlement wasmore pronounced even at this ambient temperature.With regard to the strain at break, after an initial

decrease at 20C, no further loss was observed forthe aged samples, except for the sample exposed to50C for 130 days and 80% RH, where the strain atbreak decreased by 67.5%. Taking into account thatthe molecular weight did not change substantiallywhen the PLA samples were aged at 20 and 40C

for up to 130 days (see Table I), we deduced that atthis temperature range, the elongation at break wasnot affected by the hydrolytic degradation and/orplasticizing effect of the absorbed moisture. Theplasticizing effect of water was somewhat analogousto the reported effects of other plasticizers. Thesharp decrease in the elongation at break (67.5%) forsamples aged at 50C for 130 days and 80% RH wasattributed to the irreversible change in the molecularstructure toward a lower degree of polymerizationcaused by the hydrolysis of the PLA molecules inaccordance with the SEC results shown in Table Iand Figure 2. The PLA sample aged at 50C for 30days appeared to retain its ductility. This unusualbehavior could be explained in terms of the extraplasticizing effect of the lactic oligomers, whichresulted from the hydrolysis of the ester bonds ofthe backbone, on the remaining high-molecular-weight (Mn 68,000) polymer chains.The Youngs modulus (rigidity) decreased in gen-

eral with aging time and temperature. Between 20and 40C, the Young modulus decreased onlyslightly during the various aging periods. TheYoungs modulus of a semicrystalline material isrelated to the crystallinity content and its molecularweight. As the molecular weight was not substan-tially affected by the aforementioned aging condi-tions, it appeared that the rigidity of the materialwas slightly affected by the absorbed moisture. Thisbehavior was not in full agreement with the typicalbehavior of a plasticized material, where a decreasein Youngs modulus is expected with increasingmoisture content.30 The sharp decrease in theYoungs modulus at 50C was attributed to theinduced hydrolysis of the PLA molecules in theamorphous and crystalline regions, in accordancewith the SEC results shown in Table I.

Figure 11 Tensile stressstrain results at a strain rate of1.1 104 s1 for PLA aged for 30 days at 80% RH andthree different temperatures.

Figure 12 Tensile stressstrain results at a strain rate of1.1 104 s1 for PLA aged at 20C and 80% RH for vari-ous time periods.

Figure 13 Tensile stressstrain results at a strain rate of1.1 104 s1 for PLA aged at 50C and 80% RH for vari-ous time periods.



The tensile strength decreased steadily with agingtime and temperature. To distinguish the contribu-tion of reversible versus irreversible changes towardthe loss of the tensile strength, selected aged PLAsamples were exposed to a low 25% RH for 80 days.Samples aged at 20 and 40C regained almost 90and 80% of their initial tensile strength, respectively,after exposure to dry conditions. This finding was inagreement with other results carried out on agedsamples of PLA copolymers under similar condi-tions.30 The mechanism proposed is that theabsorbed moisture was removed during the drying,and the plasticization of samples aged at 20 and40C was nearly reversed. Samples aged at 50C didnot regain their initial strength, probably as result ofthe induced hydrolytic degradation of PLA.

CONCLUSIONS

The role of environmental parameters and physicalstructure in the aging process of PLA was investi-gated on a number of PLA samples with a D-contentof 4.25% aged at temperatures below the Tg of thematerial for various aging periods. The initial ther-mal processing (150C) of the material resulted in anoverall molecular weight reduction. A substantialdecrease in the properties was observed for PLAsamples aged at 20C for 30 days. No further loss ofproperties was observed for samples aged up to40C for several time intervals. A major portion (8090%) of the induced changes in the tensile propertieswere reversed after drying. At 50C and 100 days ofaging, a sharp decrease in the overall properties wasnoticed. The results seem to confirm the earlier find-ing that PLA degradation driven by hydrolysisneeds a higher temperature in combination withample time to take place (>50C).The degradation process was very slow at temper-

atures equal or below 40C, but the speed wasgreatly enhanced at 50C. The results showed thatthe rate of degradation was higher at the beginning(first 30 days) and then decreased significantly overtime. Three mechanisms were proposed for theinterpretation of the experimental results. The initialdrop in the molecular weight was attributed to thethermal oxidation of the material during processingin the Brabender. The decrease in the propertiesobserved for the PLA samples aged at 40C or lowerfor different time intervals was attributed to theplasticizing effect of the absorbed water, whichincreased the molecular mobility of PLA andresulted in faster nucleation and crystallization ofthe PLA molecules. The time required by water topermeate and plasticize the polymer structure wasat least 1 month. At 50C, water acted mainly as adegradation (hydrolysis) agent, and the shorterchains that arose by the degradation of PLA were

reorganized in crystallites with lower melting points.The hydrolysis of PLA at higher temperatures(50C) needed at least 3 months to complete. Thecalculated values were indicative because extra pa-rameters, such as the size and shape of the sample,could influence the calculations.

References

1. Signori, F.; Coltelli, M.-B.; Bronco, S. Polym Degrad Stab 2009,94, 74.

2. Kopinke, F. D.; Remmler, M.; Mackenzie, K.; Moder, M.;Wachsen, O. Polym Degrad Stab 1996, 53, 329.

3. Shishoo, R.; Taubner, V. J Appl Polym Sci 2001, 79, 2128.4. Lim, L. T.; Auras, R.; Rubino, M. Prog Polym Sci 2008, 33,

820.5. Pillin, I.; Montrelay, N.; Bourmaud, A.; Grohens, Y. Polym

Degrad Stab 2008, 93, 321.6. Auras, R.; Harte, B.; Selke, S. Macromol Biosci 2004, 4, 835.7. Saha, S. K.; Tsuji, H. Polym Degrad Stab 2006, 91, 1665.8. Ho, K. L. G.; Pometto, A. L.; Hinz, P. N. J Polym Environ

1999, 7, 83.9. Middleton, J. C.; Tipton, A. J. Biomaterials 2000, 21, 2335.10. Acioli-Moura, R.; Sun, X. S. Polym Eng Sci 2008, 48, 829.11. Wang, H.; Sun, X.; Seib, P. J Appl Polym Sci 2003, 90,

3683.12. Tsuji, H.; Ikada, Y. Polym Degrad Stab 2000, 67, 179.13. Vert, M.; Schwarch, G.; Coudane, J. J Macromol Sci Pure Appl

Chem 1995, 32, 787.14. Paul, M. A.; Delcourt, C.; Alexandre, M.; Degee, P.; Monte-

verde, F.; Dubois, P. Polym Degrad Stab 2005, 87, 535.15. Zhang, X.; Espiritu, M.; Bilyk, A.; Kurniawan, L. Polym

Degrad Stab 2008, 93, 1964.16. Proikakis, C. S.; Mamouzelos, N. J.; Tarantili, P. A.; Andreo-

poulos, A. G. Polym Degrad Stab 2006, 91, 614.17. Tsuji, H.; Ikarashi, K.; Fukuda, N. Polym Degrad Stab 2004,

84, 515.18. Vasanthan, N.; Ly, O. Polym Degrad Stab 2009, 94, 1364.19. Zhou, Q.; Xanthos, M. Polym Degrad Stab 2008, 93, 1450.20. Grizzi, I.; Garreau, H.; Li, S.; Vert, M. Biomaterials 1995, 16,

305.21. Tsuji, H.; Del Carpio, C. A. Biomacromolecules 2003, 4, 7.22. Li, S.; McCarthy, S. Biomaterials 1999, 20, 35.23. Duek, E. A. R.; Zavaglia, C. A. C.; Belangero, W. D. Polymer

1999, 40, 6465.24. Wiggins, J. S.; Hassan, M. K.; Mauritz, K. A.; Storey, R. F.

Polymer 2006, 47, 1960.25. Laitinen, O.; Tormala, P.; Taurio, R.; Skutnabb, K.; Saarelai-

nen, K.; Iivonen, T.; Vainionpaa, S. Biomaterials 1992, 13,1012.

26. Tsuji, H.; Suzuyoshi, K. Polym Degrad Stab 2002, 75, 357.27. Copinet, A.; Bertrand, C.; Govindin, S.; Coma, V.; Couturier,

Y. Chemosphere 2004, 55, 763.28. Henry, F.; Costa, L. C.; Devassine, M. Eur Polym J 2005, 41,

2122.29. Solarski, S.; Ferreira, M.; Devaux, E. Polym Degrad Stab 2008,

93, 707.30. Holm, V. K.; Ndoni, S.; Risbo, J. J Food Sci 2006, 71, 40.31. Kale, G.; Auras, R.; Singh, S. P. Packaging Technol Sci 2007,

20, 49.32. Hakkarainen, M.; Karlsson, S.; Albertsson, A. C. Polymer

2000, 41, 2331.33. Torres, A.; Li, S. M.; Roussos, S.; Vert, M. J Appl Polym Sci

1996, 62, 2295.34. Hu, Y.; Hu, Y. S.; Topolkaraev, V.; Hiltner, A.; Baer, E. Poly-

mer 2003, 44, 5711.



35. Martino, V. P.; Ruseckaite, R. A.; Jimenez, A. Polym Int 2009,58, 437.

36. Ou, X.; Cakmak, M. Polymer 2008, 49, 5344.37. Hoogsteen, W.; Postema, A. R.; Pennings, A. J.; ten Brinke, G.;

Zugenmaier, P. Macromolecules 1990, 23, 634.38. Fischer, E. W.; Sterzel, H. J.; Wegner, G. Colloid Polym Sci

1973, 251, 980.39. Kontou, E.; Farasoglou, P. J Mater Sci 1998, 33, 147.

40. Pluta, M.; Murariu, M.; Alexandre, M.; Galeski, A.; Dubois, P.Polym Degrad Stab 2008, 93, 925.

41. Yasuniwa, M.; Sakamo, K.; Ono, Y.; Kawahara, W. Polymer2008, 49, 1943.

42. Zenkiewicz, M.; Richert, J.; Rytlewski, P.; Moraczewski, K.;Stepczyska, M.; Karasiewicz, T. Polym Test 2009, 28, 412.

43. Niemela, T.; Niiranen, H.; Kellomaki, M.; Tormala, P. ActaBiomater 2005, 1, 235.



123-effects of aging on the thermomechanical properties of pla

Documents