Characterization of the Degradation of Polylactic Acid Polymer in aSolid Substrate Environment

Mukul Agarwal, Kurt W. Koelling,* and Jeffrey J. Chalmers*

Department of Chemical Engineering, The Ohio State University, 140 West 19th Avenue, Columbus, Ohio 43210

Polylactic acid (PLA) polymer film was degraded in abiotic and biotic environmentsto understand the role of microbes in the degradation process of lactic acid basedpolymers. The degradation studies were conducted in a well-characterized bioticsystem, an abiotic system, a sterile aqueous system, and a desiccated environmentmaintained at 40, 50, and 60 °C. The combination of experiments in differentenvironments isolated the distinct effect of microbes, water, and temperature on themorphological changes in the polymer during degradation. Due to lack of availabilityof radiolabeled PLA, various analytical techniques were applied to observe changesin the rate and/or mechanism of degradation. CO2 evolved, weight loss, and molecularweights were measured to evaluate the extent of degradation. X-ray diffraction anddifferential scanning calorimetry techniques monitored the morphological changes inthe polymer. FTIR was used as a semiquantitative tool to gather information aboutthe chemistry of the degradative process. Neither of the above analytical techniquesindicated any difference in the rate or mechanism of degradation attributable to thepresence of microorganisms. The extent of degradation increased at higher processtemperatures. FTIR data were evaluated for significant statistical difference by t-testhypothesis. The results confirmed hydrolysis of ester linkage as the primarymechanism of degradation of PLA. On the basis of these data, a probable path ofPLA degradation has been suggested.

Introduction

The aliphatic polyesters of poly(2-hydroxy acid) havebeen synthesized for more than four decades, but theirsensitivity to heat and water precluded them from anyserious attention by polymer scientists for consumerproducts in an era when durability was the valued asset.The polymers of the polyactic acid (PLA) family receiveda second consideration when biocompatible and biore-sorbable materials were needed for temporary therapeu-tic applications in pharmacology and surgery. Increasingenvironmental concerns and advances in fermentation,recovery, and purification have allowed PLA to receivea third consideration. The combination of the “solidwaste crisis” and the ability to produce PLA for less than$1.00 per pound (1) has led to the commitment of at leastone commercial enterprise to produce PLA on a largescale expressly for consumer biodegradable plastics. Itis possible that the reduced costs of PLA, problems ofsolid waste management, and environmental conscious-ness will make application of lactic acid based polymersas degradable substitutes for commodity plastics eco-nomically sensible soon.Solid waste management professionals and suppliers

of biodegradable polymers agree that large-scale com-posting will provide the ideal environment for spent,biodegradable plastics (2, 3). A managed compost unithas the potential to be the most effective disposalinfrastructure for biodegradable plastics. A typical com-

post system supports a diverse microbial population ina moist aerobic environment in a temperature range of40-70 °C.Despite this new interest in biodegradable commodity

polymers, until recently, most of the work on degradationof these aliphatic polyesters has focused on PLA/GAcopolymers in vivo or in vitro (water, enzyme, or physi-ological buffers). Kulkarni et al. (4) showed high molec-ular weight lactic acid based polymers to be degradablein vitro and in vivo. Williams et al. (5, 6) investigatedthe degradation of PGA and PLA polymers in aqueousand enzymatic environments by monitoring weight lossand lactic acid production. Recent publications on deg-radation of PLA/GA copolymers in phosphate buffer (pH) 7.4, isomolar media, 37 °C) that mimic the physiologi-cal conditions of biological fluids provide an insight intothe complex mechanism of hydrolytic degradation andmorphological changes in the polymer (7-10). On thebasis of these and other studies, the mechanism ofdegradation of PLA is accepted to be primarily hydrolyticcleavage of ester bonds in aqueous media (11). Pitt etal. (12) showed that PLA degradation is autocatalyticwith respect to the number of carboxylic chain ends whichincreases with the scission of bonds.From these previous studies, there is no conclusive

demonstration of the ability of microorganisms to en-hance the degradation of high molecular weight PLAsince most, if not all, of the previous studies were limitedto aqueous media in water or sterile physiological buffersand enzymes. The primary goal of the research reportedbelow was to determine if microbes can enhance degra-dation of high molecular weight PLA. Secondary goals

* Corresponding authors. Telephone: (614) 292-2727. Fax: (614)292-3769. E-mail: Chalmers.1@osu.edu, Koelling.1@osu.edu.

517Biotechnol. Prog. 1998, 14, 517−526

S8756-7938(98)00015-0 CCC: $15.00 © 1998 American Chemical Society and American Institute of Chemical EngineersPublished on Web 03/13/1998

were to determine, or begin to determine, the optimumconditions for degradation of PLA and whether PLAdegradation affects the overall biodegradation process incompost media. A third goal, which is necessary toachieve the first two goals, was to develop analyticaltechniques to quantify changes in PLA during degrada-tion.An obvious method to characterize and quantify the

degradation of PLA is the use of a radiolabeled form.However, as a result of prohibitive expenses, this wasnot an option. Other methods of characterization includeweight loss, but as will be discussed in the results section,this also is not effective in “real world” applications.Consequently, a number of other analytical techniqueswere used. Each technique provides a significant pieceof information and, when combined, presents a reason-able characterization of the overall process of the deg-radation of the polymer.To correctly address the goals of this research, the use

of proper controls is vital. To that end, four types ofenvironments were used for this study: (1) a bioticsystem environment, (2) an abiotic environment, (3) asterile water environment, and (4) a dry oven environ-ment. In each of these environments, three constanttemperatures were used: 40, 50, and 60 °C. The use ofthree constant temperatures was chosen since materialin a “typical” compost system will experience a range oftemperatures including these three. In addition, sinceit is known that the rate of PLA degradation in anaqueous environment is a function of temperature (13),it is important that these studies be conducted at aconstant temperature so that the rate of degradation, ateach temperature, can be accurately compared. Toquantitatively determine changes in PLA after beingsubjected to these four different environments, samplesbefore and after degradation were subjected to analysisusing gel permeation chromatography (GPC), X-ray dif-fraction(XRD), differential scanning calorimetry (DSC),and Fourier transform infrared spectrophotometry (FTIR).A flowchart outlining the overall procedures is presentedin Figure 1.

Materials and Methods

Biotic Reactor. The typical solid substrate biodeg-radation (composting) process will experience a number

of different process variables as a function of time. Thistransitory behavior is most strongly driven by the heatgenerated by the microbes during the degradation pro-cess. To study the effect that temperature, moisturelevel, and aeration have on biodegradation, an apparatuswas developed (15) which allows sensitive temperaturecontrol at a desired set point (Figure 2) (14). With thissystem, oxygen concentrations, CO2 concentrations, andmoisture levels were monitored on-line, and oxygen levelswere maintained above specified levels through theincreased air flow rates to the system. Mass balanceswere performed around the reactor to determine the gasevolution and uptake rate for CO2 and O2, respectively.More details of the system and method of calculationsare reported by Tseng et al.(15).Abiotic Reactor. Since it was not possible to sterilize

the biotic reactor, a smaller vessel was developed whichcould be sterilized. The vessel consisted of a 2-L, jacketedPyrex beaker (Sigma-Aldrich Laboratory EquipmentSupplies). The temperature was monitored by a ther-mocouple and maintained by passing water, set at thespecific temperature, through the jacket. Since thesystem was sterile, temperature control was much easierthan the biotic reactor because no self-heating existed.Oxygen and CO2 concentrations were measured manuallyevery 48 h by gas analyzers. All material transfers forthe abiotic experiments were conducted in a sterilelaminar flow hood.Sterile Water and Dry Oven Studies. The sterile

water experiments were conducted in a 500-mL sealedglass bottle. Approximately 1.5 g of PLA was added to150 mL of sterile water. The unagitated, sealed bottlewas then placed in a constant-temperature bath, main-tained at the desired set point for the specified time. Forthe dry oven studies, PLA samples, contained in alumi-num pans, were placed in an oven and kept dry throughthe use of a desiccant. All four environments, including

Figure 1. Flowchart of the experiments performed.

Figure 2. Schematic of the biotic system.

518 Biotechnol. Prog., 1998, Vol. 14, No. 3

the sterile water and dry oven experiments, were con-ducted at 40, 50, and 60 °C.PLA. The PLA polymer film used in all the studies

was obtained from Eco-PLA unit of Cargill, Inc. The filmused in this study is a clear, transparent film with anaverage thickness of 75 µm. The composition of thepolymer is 95% L-lactide and 5% mesolactide.Biodegradable Substrate. To obtain reasonably

reproducible results, a semidefined feed was used as abiodegradable substrate. The composition of the sub-strate was 14% rabbit chow, 12% corn cobs, 6.0% sand,0.4% shredded newspapers, 0.6% manure, 8.0% inocu-lum, 58% distilled water, 1.0% PLA polymer strips, and0.01 M buffer (NaH2PO4, KH2PO4). This formula wasused and discussed previously by Tseng et al. (16). Thecompost (microbial) inoculum was a mixture of compostfrom a commercial yard waste facility (Kurtz Bros.,Columbus, OH) and a municipal solid waste facility (Cityof Columbus Composting Facility, Lockborune, OH).However, it should be noted that numerous runs havebeen conducted in the biotic system previous to thesestudies and the system is well inoculated purely on thebasis of the those prior runs (the system is not sterilizedbetween runs). Approximately 500 g of semidefined feedwas loaded on each tray in 1-cm layers. PLA [1%(wt/wt)], in the form of 2.5× 4 cm strips, was mixed withthis semidefined feed. For the abiotic studies, thesemidefined feed was sterilized for 1.5 h in an autoclave.Sterilization of PLA. Since PLA is sensitive to

thermal degradation, especially in aqueous and humidenvironments, all PLA samples, whether used in bioticor abiotic operation, were sterilized by γ irradiation. Theactual radiation exposure consisted of 0.4 Mrad/hr, froma Cobalt-60 γ source (Nuclear Reactor Laboratory, OhioState University, Columbus, OH), for 3 h (total dose of1.2 Mrad) at room temperature and 1 atm. Sterility waschecked in the abiotic systems by inoculating soy agarbroth with feed before and after the experiments, and itwas then compared for microbial growth or turbidity witha positive control.Operational Parameters (CO2, pH, and Temper-

ature). The operational parameters of temperature,CO2, and pH for the biotic system were monitored. Theinstantaneous temperature and CO2 concentrations wererecorded by an on-line data acquisition system. The pHof the bulk feed material (biotic and abiotic) was mea-sured by dissolving 2 g of compost in 100 mL of distilledwater. The mixture was stirred for 5 min before beingallowed to settle for half an hour. The pH of the feedmaterial and water matrix was determined before andafter degradation of the polymer.Method of Weight Loss Determination. The initial

and final weights of the plastic were measured aftervacuum-drying for 48 h in a Fisher Scientific A-seriesbalance with an error of 0.1 mg. The maximum waterabsorbed by the polymer in 15 days at room temperaturewas determined (modified ASTM method D570-81: Stan-dard Test Method for Water Absorption of Plastics) tofind if the water absorbed was substantial enough toinfluence the changes in weight measurements.Gel Permeation Chromatography. PLA samples

before and after degradation were analyzed on a Waters410 Differential Refractometer and model 6000A solventdelivery system. The solvent delivery system and detec-tor are connected to three ultra-styragel columns of 103,104, and 105 Å connected in series. Each column is 30cm in length and 7.8 mm i.d. packed with highly cross-linked styrene divinylbenzene porous copolymer. Theinternal temperature of the instrument was maintained

at 32 °C. The flow rate of the eluent (mobile phase, THF)was 1.0 mL/min for all experiments. The polymersolution (0.25% w/v) in THF was prepared and keptovernight before analysis. The injection volume of thepolymer solution was 0.2 cm3 for each run. The molec-ular weight of the polymer was calibrated againstpolystyrene standards. The trapezoidal rule (17) wasused to analyze the molecular weight distribution (MWD)curve to obtain weight average molecular weight (Mw),number average molecular weight (Mn), and polydisper-sity index (PDI).X-ray Diffraction. A PAD V Scintag instrument was

used to obtain X-ray diffraction patterns to monitorchanges in the structural order and crystallinity ofpolymers due to degradation. The X-ray beams (sourceCu KR, wavelength ) 1.54 Å) traverse a scattering anglefrom 2θ ) 5-40° at 1 deg/min. The polymer films wereset to a single-crystal silicon substrate with silicon grease.When the polymer became brittle after degradation, thepolymer was ground into a powder in a mortar-pestlebefore analysis.Differential Scanning Calorimetry. The DSC scans

were obtained on a Thermal Analyst 2000 (of Du PontInstruments) with N2 purge and calibrated with indium.The scan rate was 10 °C/min for the first scan, and 5°C/min of the second scan (residual scan). Sampleweights of 4-8 mg were placed in hermetically sealedaluminum pans for analysis.Changes in glass transition temperature (Tg), melting

point (Tm), and enthalpy of fusion (Hm) of polymersamples before and after degradation were calculated bysoftware packages in the instrument. The area underthe melting peak (representative of crystallinity of poly-mer) was normalized by the sample weight to get theenthalpy of fusion in J/g. Percent crystallinity of thesample was based on eq 1, where the enthalpy of a 100%crystalline sample (Hc) and enthalpy of an amorphoussample (Ha) were assumed to be 93.6 J/g and zero,respectively (18). Hm is the enthalpy of the degradedpolymer sample based on the area of the melting peak.

Fourier Transform Infrared Spectroscopy. Quan-titative Analysis. A solvent casting method was usedfor analysis of polymer samples of PLA before and afterdegradation. Approximately 20 µL of 4% (w/v) polymersolution in chloroform was pressed onto the IR transmit-ting windows (25 × 4 mm SpectraTech NaCl disks) toform a uniform layer. The disks were dried undervacuum for 24 h to remove any traces of residual volatilesolvent before spectra acquisition. The infrared spectrawere obtained on Magna IR 550 FTIR spectrometersseries II model from Nicolet. For analysis, Omnic andGrams/386 software were used. The optical bench con-sisted of a DTGS-KBr detector, an IR source, and a KBrbeam splitter with an aperture of 95, velocity of 0.6329cm/s, and range of 4000-400 cm-1. Spectra were col-lected with a resolution of 4 cm-1, average of 32 scans,and Happ Genzel apodization.Internal Standard. The percent change of normal-

ized peak areas, which corresponds to the decrease inspecific chemical bonds, was calculated on the basis ofthe average area for virgin PLA peaks and the area ofthe degraded polymer peak positions

where Av and As are the area under peak bands for virgin

% crystallinity ) (Hm - Ha)/(Hc - Ha) (1)

% change in area ) 100 × [Av - As]/Av (2)

Biotechnol. Prog., 1998, Vol. 14, No. 3 519

and degraded sample. To obtain quantitative analysisof the bands and peaks in the spectra, it was necessaryto use an internal standard for normalization (19). The-C-H- stretch band peak was chosen as the internalstandard peak on the basis of the chemistry of bondstrength, strong absorption by the band in the spectra,and lack of presumed degradation. Liquid sample cellswere used to verify this assumption (further discussedin the Results). The liquid sample cell contained KCl asa IR transparent window and constant path length (b)of 0.05 mm. With a fixed concentration of sample (c), itwas possible to determine absorbance intensity, A, in theregion of interest (-CH- stretch peak area) which wasa direct measure of the absolute amount of the -CH-bonds using the relationship

where a is an absorptivity constant.Statistical Analysis. The values of the integrated

area under the peak regions of interest of the polymer,before and after the treatment (degradation), were scat-tered. To determine whether the final values for areaswere statistically different from the original data set, atwo-tailed Welch’s approximation t-test hypothesis wasperformed on the spectral areas. The FTIR results havebeen reported as a percentage increase or decrease in thepeak area (attributed to the specific bonds) after thetreatment process.

Results

Biotic System. Operating Variables. There wasless than half a degree difference between the tempera-ture inside the solid organic layer and that in the reactor.The dry bulb and wet bulb temperatures were within 1°of each other, which indicates near-saturated conditionsin the system.An initial surge in the CO2 levels was observed which

was attributed to the degradation of readily degradablecompounds. The cumulative CO2 produced with PLA inthe feed material at 40, 50, and 60 °C in 15 days is givenin Figure 3. The pH of the liquid extract from the feedmaterial in the biotic system increased at all degradationtemperatures (Figure 4).Abiotic System, Sterile Water System, and Dry

Oven. Operating Variables. Since microbial activitywas not present, there was no change in the operatingparameters (temperature, CO2, and O2), except for pH.The temperature profiles for the above systems wereuniform because the environments were maintained ata set point temperature. The CO2 levels in sterile feedwere always between 0.5 and 0.7% CO2 (v/v). The pH ofthe matrix of abiotic feed and water decreased at alltemperatures (Figure 4) due to degradation products ofPLA. Williams (20) and Therin et al. (21) have attributedthe decrease in pH of the matrix to desorbtion of lacticacid liberated from PLA during degradation.Polymer Characterization. Initial PLA Polymer

Properties. The virgin PLA polymer film was amor-phous with a melting point (Tm) of 139 °C and a glasstransition temperature (Tg) of 62 °C as determined byXRD and DSC, respectively. Weight average molecularweight (Mw) of the polymer was 8.3 × 104 g/mol with aPDI of 2.04 as evaluated by GPC. The polymer after γsterilization was the same except for a 10% decrease inaverage molecular weight. The polymer film had anonuniform thickness varying from 60 to 80 µm (end toend along the width).

Visual and Physical Characteristics. The PLAsamples maintained their transparent and flexible formin all environments at 40 °C, whereas PLA polymer filmlost its mechanical strength and became brittle andopaque when subjected to 50 and 60 °C in all environ-ments (except dry oven). Clumping of PLA samplestogether and with feed was observed at all temperatures,which made retrieval and consequent weight loss analy-sis difficult.Weight Loss. Figure 5 shows the changes in weight

of PLA polymer immersed in water at room temperaturefor 15 days. No appreciable weight loss occurred at 40and 50 °C experiments, while there was an approxi-mately 35% loss in weight at 60 °C. There was no weightloss in PLA samples in dry oven environments. Weightloss in biotic and abiotic feed could not be determineddue to fusing of PLA with particles of the organic feed.Gel Permeation Chromatography. Figure 6 is a

graph presenting the final molecular weight distributionof PLA degraded at different temperatures in the bioticsystem (360 h). As can be observed, there is a significantdecrease in molecular weight with increasing tempera-ture, indicating a significant decrease in polymer chainlength. However, a dramatic decrease takes place at 60°C, with a significant fraction of the polymer having achain length in the 1050-4000 molecular weight range.At 60 °C, the molecular weight distribution curve wasbimodal and narrower than at 50 and 40 °C.Figure 7 presents the weight average molecular weight

as a function of temperature for PLA samples in the bioticsystem, the abiotic system, and sterile water system.There is an almost exponential decrease in molecularweight with increasing temperature. As can be observed,there is no detectable difference between the threesystems.Figure 8 presents the PDIs of the degraded polymer

at 40, 50, and 60 °C in the biotic system, the abioticsystem, and the sterile water system. The PDI (ameasure of the spread of the molecular weight distribu-tion curve) of the PLA increases at 40 and 50 °C anddecreases at 60 °C in all the degrading environments.X-ray Diffraction. A representative X-ray diffraction

pattern is shown in Figure 9 for PLA samples retrievedfrom biotic feed at different temperatures. The retrievedpolymer samples show sharp crystalline peaks at 50 and60 °C and an amorphous broad hump for polymersdegraded at 40 °C. There is no difference in the diffrac-tion patterns on the basis of degrading environments.Differential Scanning Calorimetry. Figure 10, a

representative differential scanning calorimetry scan ofvirgin PLA, has two endothermic peaks: a sharp pre-dominant peak at 60 °C and a subdued peak at 140 °C.To verify the melting point of the polymer, the polymersamples were kept at 50, 70, 120, and 150 °C in an ovenfor 6 h. The polymer becomes a liquid melt at 150 °C;thus the second peak in Figure 10 represents changesdue to melting. The sharp endothermic peak at 60 °C inFigure 10 represents a superimposition of stresses ofenthalpy (induced during film manufacturing) and glasstransition temperature of the polymer. The Tg of thepolymer was determined by repeated heating and coolingof the polymer from room temperature to 70 °C to relaxthe stresses.Figure 11 is a scan of PLA polymer film after degrada-

tion in sterile water at 60 °C and is a representative scanof the degradation process in the biotic and abiotic systemas well. The exotherm just before the melting endothermis hypothesized to be due to recrystallization of oligomerchains. This recrystallization exotherm was not visible

A ) abc (3)

520 Biotechnol. Prog., 1998, Vol. 14, No. 3

Figure 3. Cumulative CO2 (mol) evolved during the experiments in biotic systems maintained at 40 (a), 50 (b), and 60 °C (c),respectively. For comparison, two runs without PLA, one from Tseng et al. (15) (control A), are presented along with biotic systemcontaining PLA. The calculations are based on dynamic mass and rate balances (15).

Biotechnol. Prog., 1998, Vol. 14, No. 3 521

after heating the polymer sample to 120 °C (the ap-proximate range up to which that exotherm spans),cooling it back to room temperature in the DSC chamber(∼2 °C/min) and heating it again to 160 °C. Theshoulders in the melting endotherm are attributed to themelting of crystalline structure domains of different sizes.As the set point of degradation temperature increases,

the Tg reduces and the area under the melting peak,which corresponds to the percent crystallinity, increases.The percent increase in crystallinity and the changes in

melting points/glass transition temperatures of thesamples in the three systems (biotic, abiotic, sterilewater) were almost the same within the experimentalrange of variation (∼5%). The Tg decreases sharply dueto degradation at higher temperatures, whereas Tm isaffected only slightly. Table 1 summarizes the decreasein Tg and Tm and an increase in percent crystallinity at50 and 60 °C.Fourier Transform-Infrared Spectroscopy. A typi-

cal FTIR spectrum of virgin PLA, solvent cast on a NaCl

Figure 4. pH of the liquid extract from the feed material inthe abiotic (b) and biotic (c) system at the end of the runs. Thefinal pH of the liquid used in the sterile liquid system is alsopresented (a).

Figure 5. Percentage change in weight of PLA in sterile waterat different degradation temperatures. Samples were vacuum-dried for 48 h previous to weight determination.

Figure 6. Molecular weight distributions of virgin PLA (a) andPLA degraded in the biotic system at 40 (b), 50 (c), and 60 °C(d).

Figure 7. Average molecular weight of PLA degraded at 40,50, and 60 °C in the biotic, abiotic, and sterile water systems.Note the logarithmic y-axis.

Figure 8. Polydispersity index of degraded polymer at 40, 50,and 60 °C in biotic, abiotic, and sterile water system. The PDIfirst increases (50 °C), indicating the increase in the spread ofthe MWD curve, then decreases (60 °C), indicating the narrowspread of the MWD.

Figure 9. X-ray diffraction of PLA polymer film after degrada-tion in the biotic system at 40 (a), 50 (b), and 60 °C (c).

522 Biotechnol. Prog., 1998, Vol. 14, No. 3

halide window disk, is shown in Figure 12. The spectrumwas classified into five regions, which correspond to thefollowing bands: -CH- stretch, -CdO- carbonyl, -CH-deformation, -C-O- stretch, and -C-C stretch. Thesepeak assignments and corresponding positions are givenin Table 2.Figure 13 presents the normalized area under the four

peak areas chosen from final PLA samples from the bioticsystem, sterile system, and sterile water systems at thethree different temperatures. Peaks at 2995 and 2944cm-1 were used for the internal normalization, whichcorrespond to a -CH- bond in a PLA monomer. As canbe observed, because of the apparent randomness of thedata, it is difficult to make clear conclusions. Conse-quently, a two-tailed Welch approximation t-test with

95% confidence was used (22). Peak areas that werefound to be statistically different from virgin PLA areindicated in Table 3.To determine the accuracy of the internal standard

assumption, namely that the -CH- bond does notdegrade, the absorbances at 2995 and 2944 cm-1 fordissolved liquid samples of PLA were determined.Samples of virgin PLA and of material after each type ofdegradation were tested. As long as the same conditionsare used for each test, i.e. concentration of sample, samesolvent, same sample holder and calibration settings, theabsorbance is directly proportional to the concentrationof the band of interest. There was no change, withinexperimental error, in the peak area at 2995-2944 cm-1

between virgin and degraded samples.

DiscussionDo Microbes Enhance Degradation of PLA? The

primary goal of this work was to determine if microbessignificantly enhance the degradation of PLA. Upon thebasis of the various analytical techniques used, nosignificant evidence was obtained to indicate that anymicrobial enhancement of degradation exists at 40, 50,and 60 °C. The results from the various analyticaltechniques that support this conclusion will be discussedbelow.The most obvious method, besides radiolabeled PLA,

to measure degradation of PLA is to determine weight

Figure 10. DSC scan of virgin, irradiated PLA after variousheating histories (the irradiation had a very minor effect on thePLA). Line a corresponds to the first heating in the DSCappartus. The strong endothermic peak at 60 °C is a super-imposed Tg and enthalpy of stress relaxation induced in thepolymer film during processing. The melting point of thepolymer is at ca. 140 °C. Lines b, c, and d correspond to thesecond, third, and fourth heating of the PLA sample to relievethe stresses in the material. Line e is a full scan of the PLApolymer sample from the curve d. The melting point, the glasstransition temperature, and enthalpy of fusion are determinedon the basis of this curve.

Figure 11. DSC scans of PLA samples obtained from sterilewater at 60 °C. The samples were heated at 10 °C/min (a, d)and 5 °C/min (b, c). There was no evident glass transitiontemperature value obtained from the scans. (a) PLA sampledegraded at 60 °C in sterile water system. Note the clear broadexothermic peak superimposed just before the melting peak. (b)Polymer sample heated until 80 °C in an attempt to isolate Tg.The sample is cooled to room temperature in the DuPontInstrument chamber. (c) Sample of b heated until 110 °C tocomplete the exothermic reaction. This sample is cooled in thechamber to room temperature again. (d) Same sample from partc again heated to 170 °C. This curve is analyzed for meltingpoint and heat of fusion values determination.

Table 1. Values of Glass Transition Temperature (Tg),Melting Temperature (Tm), Enthalpy of Melting (∆Hm),and Percent Crystallinity (%C) for PLA Polymer Samplesin the Virgin Form, after Sterilization, and after 15 daysin the Biotic, Abiotic, Sterile Water, and Dry Oven at 50and 60 °C

sample Tg (°C) Tm (°C) ∆Hm (J/g) %C

virgin PLA 62.21 138.88 0.596 0.63irradiated PLA 60.79 140.47 0.637 0.68biotic reactor, 50 °C 52.45 136.9 31.14 33.26biotic compost, 60 °C 136.7 46.65 49.83abiotic reactor, 50 °C 47.71 138.13 36.04 38.50abiotic reactor, 60 °C 128.67 58.49 62.48sterile water, 50 °C 53.04 141.01 30.94 33.05sterile water, 60 °C 135.49 49.96 53.37

Table 2. Peak Band Assignments for PLA Spectra

assignment peak position, cm-1

-CH- stretch 2995, 2944-CdO- carbonyl 1759-CH- deformation(including sym and asym bend)

1453, 1382, 1362

-C-O- stretch 1268, 1194, 1130, 1093, 1047-C-C- stretch 868

Table 3. Results of Statistical Analysis of Degraded PLAPolymersa

samplesdegraded at -CdO-

-CH-deformation

-C-O-ester

-C-C-stretch

40 °Cbiotic compost X X Xsterile compost Xsterile water X X X

50 °Cbiotic compost Xsterile compost X Xsterile water X

60 °Cbiotic compost X X Xsterile compost X Xsterile water X X X Xa The Bands that were found to be statistically different due to

degradation have been identified by a check (X).

Biotechnol. Prog., 1998, Vol. 14, No. 3 523

change. This method showed a 35% weight loss at 60°C in sterile H2O. However, as discussed above, it is notpossible to measure weight loss when feed was present.The combination of fusing of the PLA to the feed material,

the brittle nature of the PLA after exposure to degrada-tion, and the variability in the thickness of the PLAsheets prevented accurate weight loss measurements.After weight loss, the next best measure of degradation

is the use of gel permeation chromatography. Thistechnique determines the distribution of the molecularweight of the polymer chains in a sample. As can beobserved in Figure 6, significant and unique changes inmolecular weight distributions of the PLA chains areobserved at each degradation temperature. The keyquestion, with respect to this paper, is whether, at a giventemperature, this distribution significantly changes inthe biotic, abiotic, sterile water, and dry oven environ-ments. Two quantitative methods can be used to com-pare the results in each of the degradation systems: theaverage molecular weight and the polydispersity index(Figures 7 and 8).The nearly exponential decrease in average molecular

weight as a function of temperature was expected sincethe generally accepted mechanism for PLA degradationis the chemical hydrolysis of the ester bond in PLA. Ascan be observed in Figure 7, while there is a smalldifference in PLA between the biotic and abiotic systems,we do not believe that the difference is significant. Inaddition, except for the sterile water sample at 50 °C,there is no significant difference in the polydispersityindex between the three aqueous degradation systems.The sharp peak in chromatographs at 60 °C is believedto be related to the change in PLAmorphology (discussedbelow) and the increased rate of degradation at thattemperature.A decrease of CO2 production due to degradation of

PLA was observed at 40 and 50 °C when compared tocontrols without PLA. However, due to the variabilityinherent in these degradation studies (and common inany composting environment) and fusing of the PLA tothe feed, care should be taken in drawing any conclu-sions. In addition, the byproduct of PLA, lactic acid,should be readily degraded by most microbes and wouldnot be expected to be inhibitory (23). It should be notedthat the lactic acid used in the production of PLA is afermentation product.Changes in Morphology and Mechanism of PLA

Degradation. X-ray diffraction and differential scan-

Figure 12. Spectra of virgin PLA cast onto an IR transparent alkali halide (KBr) disk from a 4% (w/v) solution in chloroform.

Figure 13. Results of % change in normalized areas of polymerpeak bands. The positive values represent decreases in bondconcentration due to treatment. The negative values indicatean increase in band over control, virgin polymer.

524 Biotechnol. Prog., 1998, Vol. 14, No. 3

ning calorimetry techniques were used to monitor thechanges in morphology and percent crystallinity of PLAafter degradation. As is evident from Figure 9, theposition of the crystalline peaks is the same in all theenvironments, indicating that there was no difference inthe diffraction patterns of degraded PLA in differentdegrading environments. Consequently, the presence orabsence of microorganisms did not affect the morphologyor the thermal characteristics of the degraded polymer.The increase in percent crystallinity is part temperaturedriven and part degradation induced. Quantitative de-termination of the percent crystallinity, evaluated byDSC, indicated that the Tg decreases as the numberaverage molecular weight of the polymer decreases.These results demonstrate that polymer degradation andtemperature increase simultaneously reorganize andcrystallize the residual polymer chains. These resultsconfirm previous results of crystallization of amorphouspolymers on degradation (24).FTIR analysis also indicated that there was no differ-

ence in the mechanism of degradation in the presence orabsence of microorganisms. The -C-O- ester linkage,the site of hydrolysis, was found to be statisticallydifferent from virgin polymer after degradation in bothabiotic and biotic environments at all temperatures(Table 3). This confirms that the degradation mechanismis predominantly hydrolysis of ester linkage in thepresence or absence of microorganisms. This is consistentwith the work of Toores et al. (25). They identifiedstrains of microorganisms which assimilated low molec-ular weight PLA oligomers but were unable to demon-strate that these microorganisms could break down highmolecular weight forms of PLA. At 60 °C the -C-C-stretch was found to be statistically different (area underthe curve increased) after treatment in both the bioticand abiotic environments. It has been reported that thispeak position is sensitive to the crystallinity of thepolymer (26). As will be discussed below, the degree ofcrystallinity increased by ∼30% at 50 °C and ∼50% at60 °C; consequently, it is hypothesized that this changein the -C-C- is the result of a change in the degree ofcrystallinity of the polymer.Proposed Model of the Mechanism of PLA Deg-

radation in Compost. While the results of the variouscharacterization methods did not show any difference inthe mechanism and/or rate of degradation in the presenceof microorganisms, when combined together, a model ofthe mechanism of degradation can be proposed whichexplains the various analytical changes observed atdifferent temperatures and environments. The followingparagraphs outline this model.At all temperatures studied in aqueous conditions, the

PLA polymer absorbed water, resulting in the hydrolysisof ester linkages, breaking down long macromolecularchains. The rate of hydrolysis was observed to increasewith temperature. This decreases the average molecularweight, leading to a reduction in both the glass transitiontemperature and mechanical properties of the sample.If the temperature of the sample is above the Tg, lowmolecular weight oligomers created by the hydrolysisreaction can diffuse away, resulting in a decrease inweight of the sample. The reduction in mechanicalproperties due to degradation may also lead to erosionof the polymer matrix, resulting in further weight lossof the sample. Previous studies by Pitt et al. (27) haveshown that weight loss of PLA starts to occur at anumber-average molecular weight,Mn, of approximately15 000. In addition, if the temperature of the sample isabove the Tg, amorphous regions of the polymer will

slowly become more crystalline (28). It has been shownin biomedical studies that highly crystalline samplesdegrade more slowly than less crystalline samples (29).In our 40 and 50 °C studies, where Tg > 50 °C andMn

> 20 000, there was no weight loss observed. However,in our 60 °C studies, where Tg , 60 °C andMn < 10 000,the weight of the sample decreased significantly (35%).These results are consistent with the degradation anderosion mechanisms mentioned above.This mechanism, and the supporting data, has signifi-

cant impact on both the production of a PLA for consumeruse as well as the management of a composting process.In particular, polymers with as low a Tg and molecularweight as possible would be best for composting systemsand the actual composting processes should be operatedat temperatures significantly above the Tg for maximumdegradation. However, decreasing the molecular weightand Tg of the polymers would yield a product with thelowest mechanical properties. Thus a trade-off arisesbetween two competing interests. Further work isneeded to clarify this conflict.

Acknowledgment

We thank Dr. Mary Jo Zidwick of Cargill Inc. forhelpful discussion of this work. We also thank Cargill,Inc., the Consortium for Plant Biotechnology, and theNational Science Foundation (BCS #9258004) for finan-cial assistance.

References and Notes(1) Datta, R.; Tsai, S. P.; Bonsignore, P.; Moon, S. H.; Frank,J. R. Technological and economic potential of poly (lactic acid)and lactic acid derivatives. FEMS Microbiol. Rev. 1995, 16,221-231.

(2) Ogando, J. Biodegradable polymers crop up all over again.Plastics Technol. 1992, 8, 60-62.

(3) Thayer, A. M. Degradable plastics generate controversy insolid waste issues. Chem. Eng. News 1990 (6), 7-14.

(4) Kulkarni, R. K.; Moore, E. G.; Hegyeli, A. F.; Leonard, F.Biodegradable poly lactic acid polymers. J. Biomed. Mater.Res. 1971, 5, 169-181.

(5) Williams, D. F.; Mort, E. Enzyme accelerated hydrolysis ofpolyglycolic acid. J. Bioeng. 1977, 1, 231-238.

(6) Williams, D. F. Enzymic hydrolysis of poly lactic acid.Mater.Eng. Prog. 1981, 10, 5-7.

(7) Li, S. M.; Garreau, H.; Vert, M. Structure-property rela-tionships in the case of the degradation of massive aliphaticpoly-(R-hydroxy acids) in aqueous media. J.Mater. Sci. Mater.Med. 1990, 1, 123-139, 198-206.

(8) Li, S. M.; Vert, M. Crystalline oligomeric stereocomplex asan intermediate compound in racemic poly (DL-lactic acid)degradation. Polym. Int. 1994, 33, 37-41.

(9) Vert, M.; Li, S.; Garreau, H. Attempts to map the structureand degradation characterisitics of aliphatic polyesters de-rived from lactic and glycolic acids. J. Biomater. Sci., Polym.Ed. 1994, 6, 639-649.

(10) Vert, M.; Mauduit, J.; Li, S. M. Biodegradation of PLA/GA polymers: increasing complexity. Biomaterials 1994, 15,1209-1213.

(11) Li, S. M.; Vert, M. Morphological changes resulting fromthe hydrolytic degradation of stereo-copolymers derived fromL- and DL-lactides. Macromolecules 1994, 27, 3107-3110.

(12) Pitt, C. G.; Chasalow, F. I.; Hibionada, Y. M.; Klimas, D.M.; Schindler, A. J. Appl. Polym. Sci. 1981, 26, 3779.

(13) Mason, N. S.; Miles, C. S.; Sparks, R. E. Hydrolyticdegradation of poly DL-lactide. Polym. Sci. Technol. 1981, 14,279-301.

(14) Tseng, D. Y.; Chalmers J. J.; Tuovinen, O. H.; Hoitink, H.A. J. Characterization of a bench-scale system for studyingthe biodegradation of organic solid wastes. Biotechnol. Prog.1995, 11, 443-451.

Biotechnol. Prog., 1998, Vol. 14, No. 3 525

(15) Tseng, D. Y. Application of spectroscopic analysis (FTIR)to a bench scale solid susbtrate fermentation (composting)system. Ph.D. Thesis, Department of Chemical Engineering,The Ohio State University, Columbus, OH, 1995.

(16) Tseng, D. Y.; Chalmers J. J.; Tuovinen, O. H.; Hoitink, H.A. J. Characterization of a bench-scale system for studyingthe biodegradation of organic solid wastes. Biotechnol. Prog.1995, 11, 443-451.

(17) Carnahan, B.; Luther, H. A.; Wilkes, J. O. Applied Nu-merical Methods; John Wiley and Sons: New York, 1969.

(18) Migliaresi, C.; De Lollis, A.; Fambri, L.; Cohn, D. The effectof thermal history on the crystallinity of different molecularweight PLA biodegradable polymers. Clin. Mater. 1991, 8,111-118.

(19) Tseng, D. Y.; Vir, R.; Traina, S. J.; Chalmers, J. J. A FourieTransform Infrared Spectcroscopic Analysis of Organic MatterDegradation in a Bench-Scale Solid Substrate Fermentation(Composting) System. Biotechnol. Bioeng. 1996, 52, 661-671.

(20) Williams, D. F. Enzymic hydrolysis of poly lactic acid.Mater. Eng. Prog. 1981, 10, 5-7.

(21) Therin, M.; Christel, P.; Li, S.; Garreau, H.; Vert., M. Invivo degradation of massive poly (R-hydroxy acids): validationof in vitro findings. Biomaterials 1992, 13, 594-600.

(22) Perry, R. H.; Green, D.; Maloney, J. O. Chemical Engineers’Handbook, 6th ed.; McGraw-Hill Book Co.: New York, 1984.

(23) Ozeki, E., Fujii, Y. Biodegradable Plastics and Polymers.In Proceedings of the Third International Scientific Workshop

on Biodegradable Plastics and Polymers, Osaka, Japan,November 1994; Doi, Y., Fukuda, K., Eds.; Elsevier: NewYork, 1994.

(24) Vert, M.; Li, S.; Garreau, H. More about the degradationof LA/GA-derived matrixes in aqueous media. J. ControlledRel. 1991, 16, 15-25.

(25) Torres, A.; Li, M. S.; Roussos, S.; Verti, M. Screening ofMicroorganisms for Biodegradation of Poly(Lactic Acid) andLactic Acid-Containing Polymers. Appl. Environ. Microbiol.1996, 62, 2393-2397.

(26) Kister, G.; Cassanas, G.; Vert, M.; Pauvert, B.; Terol, A.Vibrational analysis of poly (l-lactic acid). J. Raman Spectrosc.1995, 26, 307-311.

(27) Pitt, C. G.; Gratzl, M. M.; Kummel, G. L.; Surles, J.;Schindler, A. Biomaterials 1981, 2, 215.

(28) Billmeyer, F. W., Jr. Textbook of Polymer Science; JohnWiley and Sons: New York, 1984.

(29) Vert, M.; Mauduit, J.; Li, S. M. Biodegradation of PLA/GA polymers: increasing complexity. Biomaterials 1994, 15,1209-1213.

Accepted February 5, 1998.

BP980015P

526 Biotechnol. Prog., 1998, Vol. 14, No. 3