lable at ScienceDirect

Polymer Degradation and Stability 95 (2010) 2334e2342

Contents lists avai

Polymer Degradation and Stability

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

Thermodegradation of medium-chain-length poly(3-hydroxyalkanoates)produced by Pseudomonas putida from oleic acid

Mei Chan Sin a, Seng Neon Gan b,*, Mohd Suffian Mohd Annuar a, Irene Kit Ping Tan a

a Institute of Biological Sciences, Faculty of Science, University of Malaya, 50603 Kuala Lumpur, MalaysiabDepartment of Chemistry, Faculty of Science, University of Malaya, 50603 Kuala Lumpur, Malaysia

a r t i c l e i n f o

Article history:Received 26 May 2010Received in revised form16 August 2010Accepted 28 August 2010Available online 18 September 2010

Keywords:Thermal degradationMedium-chain-lengthpoly(3-hydroxyalkanoates)Thermo-kinetic parametersTGAThermal decomposition mechanism

* Corresponding author. Tel.: þ60 3 79674241; fax:E-mail address: sngan@um.edu.my (S.N. Gan).

0141-3910/$ e see front matter � 2010 Elsevier Ltd.doi:10.1016/j.polymdegradstab.2010.08.027

a b s t r a c t

Medium-chain-length poly(3-hydoxyalkanoates) (mcl-PHA), comprising six to fourteen carbon-chain-length monomers, are natural thermoplastic polyesters synthesized by fluorescent pseudomonades. Inthis study, mcl-PHA was produced by Pseudomonas putida from oleic acid in aerobic shake flask fermen-tation. Thermal degradation of mcl-PHA was performed at temperatures in the range of 160e180 �C.Thermodynamic parameters of mcl-PHA thermal degradation were determined where degradation acti-vation energy, Ed and pre-exponential factor, A equal to 85.3 kJ mol�1 and 6.07� 105 s�1, respectively; andexhibited a negative activation entropy (ΔS) of �139.4 J K�1 mol�1. Titration was carried out to determinethe carboxylic terminal concentration and used to correlate number-averagemolecular weight (Mn) of thepolymers. Thermally-degraded PHA contained higher amount of carboxylic terminals and lower Mn

compared to the initial PHA and these results coincide with the decreased Mn in GPC analysis. Thermalproperties of initial and degraded mcl-PHA were characterized by thermogravimetric analysis (TGA) anddifferential scanning calorimetry (DSC). The thermal decomposition mechanism was investigatedfollowing the analyses of the degradation products using 400-MHz 1H NMR, FTIR spectroscopy and GCanalysis. The overall decomposition reaction is the hydrolysis of ester linkages to produce hydroxyl andcarboxylic terminals. A small proportion of unsaturated side chain fragments would undergo oxidativecleavage at C]C linkages, producing minor amount of low-molecular weight esters and acids. At highertemperatures, the hydroxyl terminal can undergo dehydration to form an alkenyl terminal.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Petroleum based plastics are considered one of the biggest envi-ronmental pollutants, and this had led to intense research activitiesto develop alternative materials which are more environmentalfriendly. The desirable alternatives should be biodegradable intoharmless intermediates and end products. Polyhydroxyalkanoates(PHAs) are polyesters synthesized by some microorganisms, usuallyin response to unfavorable growth conditions e.g.when an essentialnutrient such as nitrogen, phosphorus, oxygen or sulfur becomeslimiting [1e4]. PHAs are accumulated as insoluble granules in thecells, and they are believed to be the carbon and energy storagecompounds [1,5,6].

In these microbial polyesters, the carboxyl group of onemonomer forms an ester bond with the hydroxyl group of theneighboring monomer. Each monomer contains the chiral carbonand has the (R) stereochemical configuration, thus PHA are

þ60 3 79674193.

All rights reserved.

optically active and isotactic [7,8]. The alkyl side chain can besaturated, unsaturated, aromatic, halogenated and deoxidized[9e14]. The family of PHAs consists of 2 major classes: short-chain-length (scl) and medium-chain-length (mcl) PHAs, and they exhibitphysical properties ranging from hard and crystalline to elastic gel.Scl-PHAs are comprised of monomers having 3e5 carbon atomswhereas mcl-PHA is comprised of monomers having 6e14 carbonatoms. Different combinations of monomers yield polymers witha wide variety of chemical and physical properties, and thusbroaden the potential applications of PHAs. Compared to scl-PHAs,mcl-PHAs have lower crystallinity and glass transition temperature.They are more elastic, less stiff and brittle [6,15e17]. Hence mcl-PHAs are easier to process and have potentially wider applications.This class of PHA is primarily produced by the fluorescent pseu-domonades belonging to rRNA homology group I [16e18], and themonomeric composition of mcl-PHA is closely related to thestructure of the carbon substrate that is fed to the bacteria[16,19e21]. Mcl-PHAs derived from oleic acid (C18:1Δ

9 ) and otherunsaturated oils and fatty acids contain unsaturated bonds in theside chains [6,22,23], and these are useful sites for chemicalmodifications.

Mei Chan Sin et al. / Polymer Degradation and Stability 95 (2010) 2334e2342 2335

For PHA to be commercialized, there has to be some degree ofcost-competitiveness. Although the cost of production for PHAs isstill higher than petrochemical plastics, it has gradually decreaseddue to several important factors: strain improvement, identificationof lower-priced fermentation substrate, better production tech-niques and improvement of downstream processes. On the otherhand, the prices of petrochemical plastics have gone up significantlydue to a surge in prices of crude oil and fossil fuel in the world.Currently, only the scl-PHA has been successfully commercialized.The major producers being Metabolix which markets poly(hydrox-ybutyrate-co-hydroxyvalerate) (PHBV); Procter and Gamble whichdeveloped Nodax poly(hydroxybutyrate-co-hydroxyhexanoate)(PHBH) [24]. However, mcl-PHAs need to be targeted for specificapplications. In fact mcl-PHA has wide and versatile chemical andphysical properties. These have to be well-understood in order todesign modifications to develop desired material properties such asthermal degradation and stability, fire retardancy, mechanical andbarrier properties. Many authors have devoted concentration to thestudy on thermal degradation of scl-PHA such as poly-hydroxybutyrate (PHB) and the copolymer, poly(hydroxybutyrate-co-hydroxyvalerate) (PHBV) [25e31] and they reached at the sameconclusion that thermodegradation of scl-PHA occurs almost exclu-sively via a random chain scission mechanism involving a six-member ring transition state and bhydrogen elimination reaction.Nevertheless the thermal decompositionmechanismofmcl-PHAhasnot been investigated. This paper describes the thermal degradationof mcl-PHAs biosynthesized by Pseudomonas putida from oleic acid.The changes in some thermal, physical and chemical properties ofthe thermally-degraded mcl-PHA were analyzed and the data wereused to elucidate the mechanism of thermal decomposition.

2. Materials and methods

2.1. PHA production by P. putida

Oleic acid is a derivative of palm oil, and was a kind gift fromSouthern Acids. In this study, P. putida PGA1 was used to producethe mcl-PHA, utilizing 0.5% (v/v) oleic acid as the renewable carbonsource. The strain was maintained at 4 �C on nutrient agar plate forperiodic subculture. The bacteria were cultured by shake flaskfermentation at 30 �C, 200 rpm (HOTECH orbital shaker incubator).

PHA production was performed in a two-stage culture systemwhich consisted of a cell-growth phase and a PHA-accumulationphase. In the first phase, the bacteria were grown in nutrient broth(8 g L�1), a nutritionally rich medium, to produce a large amount ofcells. After 20 h of incubation, the cells were harvested by centrifu-gation, washed with 0.85% saline, and transferred to a nitrogen-limitingM9mediumto inducePHAbiosynthesis by the cells. TheM9mediumcontained the following ingredients (g L�1):Na2HPO4.7H2O12.8, KH2PO4 3, NH4Cl 1, NaCl 0.5; trace elements (2 ml of a 1 MMg2SO4$7H2O stock solution and 1 ml of a 0.1 M CaCl2 stock solu-tion). The carbon source was 0.5% v/v oleic acid. All solutions weresterilized at 121 �C at 1 atm for 15 min using TOMY autoclavemachine. The carbon substrate and trace element solutions ofmagnesium salt and calcium salt were sterilized separately beforeadding to the rest of the ingredients to avoid precipitation duringautoclaving.

2.2. Removal of oleic acid residue from the cells bybiomass pretreatment

After 72 h of incubation in the M9 medium, the cells wereharvested by centrifugation (Continent R large capacity refrigeratedcentrifuge), washedwith 0.85% saline and dried at 70 �C to constantweight. The dried cells were washed by suspending them in 95%

ethanol and shaking for 30 min at 22 �C and 160 rpm. Oleic acidresidues and other polar lipids attached to the cells would dissolvein the alcohol which was then decanted off. The washed cells weredried in oven at 70 �C to constant weight.

2.3. PHA extraction and purification

Intracellular PHA was extracted by suspending the cells inchloroform and refluxed for 6 h at 70 �C. The chloroform solutionwas filtered throughWhatman number 1 filter paper to remove thecellular debris, and the filtrate was concentrated by rotary evapo-ration (EYELA N-1000 rotary evaporator). The concentratedchloroform was added drop-wise into rapidly-stirred cold meth-anol to precipitate the PHA. Further purification was carried out byre-dissolving the PHA in a small amount of chloroform and re-precipitating it in excess methanol. It was then dried in a vacuumoven (JEIOTECH) at 37 �C, 0.6 atm for 24 h.

2.4. Determination of thermal properties of mcl-PHA

Thermal behavior of control and degraded mcl-PHA werecharacterized using Thermogravimetric Analysis (TGA) (Per-kineElmer) and Differential Scanning Calorimetry (DSC) (MettlerToledo) to determine the initial degradation temperatures (Td) andglass transition temperatures (Tg) respectively. For TGA analysis,about 10 mg sample was heated from 50 �C to 900 �C at a rate of10 K min�1 under an atmospheric nitrogen flow of 20 ml min�1.This was done to establish the temperature in which degradationstarts to occur. DSC analysis was carried out in the temperaturerange of �100e180 �C at a heating rate of 20 K min�1 undernitrogen atmosphere. Tg of the sample was determined as themidpoint of steep change in energy.

2.5. Determination of thermodynamic parameters forthermal degradation of mcl-PHA

Thermodegradation kinetics of mcl-PHA was studied usingKissinger’s method through Thermogravimetric Analysis [32e35].PHA samples were analyzed by TGA non-isothermal procedures atmultiple heating rates of 10, 15, 20, 25 and 30 Kmin�1. The nitrogenflow rate was fixed at 20 ml min�1. The degradation activationenergy and pre-exponential factor of the process were estimatedusing Kissinger expression as shown in Equation (1).

�ln�q=T2p

�¼ Ed=RTp � lnðAR=EdÞ (1)

where q ¼ heating rate (K min�1), Tp ¼ maximum degradationtemperature (K), R ¼ universal gas constant (8.3143 J K�1 mol�1),Ed ¼ degradation activation energy (J mol�1), A ¼ pre-exponentialfactor (s�1)

Kissinger’s method for calculating degradation activationenergy uses the temperature at maximum degradation rate. Fromthe linear plot of eln (q/Tp2) versus 1/Tp, activation energy and pre-exponential factor could be calculated from the slope and theintersection at y-axis respectively [32e34]. The entropy of activa-tion (ΔS) for PHA thermal degradationwas also calculated followingEquation (2) [36].

A ¼ �kTp=h

�eDS=R (2)

where A ¼ Arrhenius parameter (s�1), k ¼ Boltzman constant,Tp ¼ peak temperature (K), h ¼ Planck constant (J s), ΔS ¼ entropyof activation (J K�1 mol�1), R ¼ universal gas constant(8.3143 J K�1 mol�1).

Mei Chan Sin et al. / Polymer Degradation and Stability 95 (2010) 2334e23422336

2.6. Thermal degradation of the purified PHA

Thermal degradation of the mcl-PHA was performed at 160 �C,170 �C and 180 �C.PHA samples (known weight) were heated at160 �C, 170 �C and 180 �C respectively in a silicon oil bath. Eachsample was heated from ambient temperature to the set temper-ature and maintained for 30 min. After that, the sample was cooleddown to room temperature, and a small amount of chloroformwasadded to dissolve the degradation products. The solution waspoured into a pre-weighed glass petri dish and dried in vacuum toevaporate off the chloroform.

2.7. Characterizations of mcl-PHA

2.7.1. End-group analysisTitration was carried out to determine the concentration of

carboxylic acid end groups in the polymer. It was performed inaccordance with the procedure of ASTM D 1980-87 Standard TestMethod. The PHA samples (undegraded control and thermally-degraded ones) were dissolved in a 2:1 mixture of toluene andethanol. The acid values were determined by titrating the solutionswith standardized potassium hydroxide/ethanol solution, usingphenolphthalein as indicator. All tests were performed in duplicate.

2.7.2. Molecular weight measurements and distributionThe average molecular weight and distribution of control and

degraded PHA were investigated by gel permeation chromatog-raphy (GPC) using a Waters 600-GPC (USA) instrument equippedwith a Waters Styrogel HR columns (7.8 mm internal diame-ter � 300 mm) (USA) connected in series (HR1, HR2, HR5E andHR5E) and aWaters 2414 refractive index detector. Tetrahydrofuranwas used as the eluent at 40 �C at a flow rate of 1 ml min�1; 100 mlof a 2 mg ml�1 solution was filtered through a 0.45 mm filter andinjected for each polymer sample. The instrument was calibratedusing monodisperse polystyrene standards.

2.7.3. Fourier transform infrared (FTIR) spectroscopyFTIR analysis was conducted with a PerkineElmer FTIR RX

Spectrometer. PHA samples were cast on the NaCl FTIR cells as thinfilms. The spectrawere recorded between 4000 and 650 cm�1, with4 cm�1 resolution and 16 scans at room temperature.

Fig. 1. TGA thermogram of mcl-PHA. Drastic weight loss of about 99% w

2.7.4. Proton Nuclear Magnetic resonance (1H NMR) spectroscopy1H NMR analysis of the control (not subject to heat treatment)

and thermally-degraded PHA samples was carried out on a JEOLJNM-LA 400 FTNMR spectrometer. The 400 MHz 1H NMR spectrawere recorded at ambient temperature in a 2% (w/v) PHA-CDCl3solution. Chemical shifts in ppm were relative to TMS at 0 ppm asinternal reference.

2.7.5. Gas chromatography analysis (GC)Analysis of PHA monomeric compositions before and after

thermal treatment was performed using a GC 2014 Shimadzu(Japan) equipped with an SGE forte GC capillary column BP20(30 m � 0.25 mm internal diameter � 0.25 mm) (Australia) anda flame ionization detector (FID). Approximately 8 mg of purifiedPHA was subjected to methanolysis by heating at 100 �C for140 min in the mixture of 1.0 ml chloroform, 0.85 ml methanol and0.15 ml concentrated sulfuric acid in a screw-cap tube sealed withPTFE tape. The mixture was shaken occasionally throughout theheating process and subsequently cooled to room temperature.1 ml of distilled water was added to the reaction mixture. Themixture was then vortexed for 1 min and allowed to stand for10min to induce phase separation. The organic phase at the bottomlayer was recovered by glass Pasteur pipette and analyzed by gaschromatography. 1 ml of sample was injected by split injection witha split ratio of 10:1 using an SGE 10 ml syringe (Australia). Nitrogenwas used as the carrier gas at a flow rate of 3 ml min�1. The columnoven temperature was programmed from 120 �C for 2 min,increased at a rate of 20 �C min�1 to 230 �C, and held at thistemperature for 10 min. The temperature of injector and detectorwas 225 �C and 230 �C respectively. 3-hydroxyalkanoic acid methylester standards (Larodan) were used to determine respectiveretention times for peak identification.

3. Results and discussion

3.1. Thermo-kinetic analysis of control PHA thermodegradation

Fig. 1 shows the TGA thermogram of the oleic acid derived mcl-PHA, which appeared to proceed by a one-step process between200 and 300 �C, and this was represented by a single peak in thecorresponding derivative thermogravimetric (DTG) curve. During

as observed at 180 �Ce520 �C, due to the pyrolysis of the polyester.

Table 1Thermodynamic parameters for thermal degradation of mcl-PHA.

q (/Kmin�1)

Tp(/K) 1/Tp(�103 K)

eln(q/Tp2)

Ea, activationenergy(kJ�1 mol)

A, pre-exponentialfactor(�10�5 s)

ΔS, Entropychangeof activation(J �1 K mol)

10 552 1.81 10.33 85.3 6.07 �139.415 571 1.75 9.9920 576 1.74 9.7225 578 1.73 9.5030 582 1.72 9.33

Mei Chan Sin et al. / Polymer Degradation and Stability 95 (2010) 2334e2342 2337

the degradation step, a weight loss of 99% was observed with onlytraces of carbonaceous residues (carbon black) remained.

Fig. 2a and b shows TG and DTG thermograms for mcl-PHAat different heating rates. All thermograms in Fig. 2a shows onlyone-step degradation process between 200 and 400 �C, andincreasing heating rate has shifted the onset temperaturehigher. From Fig. 2b, temperature at the lowest point of theDTG curve indicates the temperature at the highest degradationrate.

To obtain thermodynamic parameters by Kissinger equation, therelationship between heating rates and temperature at most rapiddegradation is shown in Table 1, and the plot of eln(q/Tp2) versus 1/Tp is displayed in Fig. 3.

The degradation activation energy, Ed and pre-exponentialfactor, A for mcl-PHA were 85.3 kJ mol�1 and 6.07 � 105 s�1

respectively. Table 1 also shows that the change of activationentropy (ΔS) has a negative value of �139.4 J K�1 mol�1, and therelatively high Ed indicates that thermal degradation can only occurat high temperature.

3.2. Visual observation of heat-treated PHA

During the heat treatment at 160 �C, 170 �C and 180 �C, the colorof the mcl-PHA changed from an initial yellow to brown and darkbrownwith black precipitates, while viscosity of the elastomer hasdecreased. PHA which had been heat treated at 160 �C and 170 �Cproduced compounds which did not vaporize. Evolution of gaseousproducts with pleasant odors accompanied by bubbles formation inthe sample was observed when PHA was heated at 180 �C.According to Gonzalez et al. (2005) when PHA was thermolyzed,monomeric and oligomeric (dimeric, trimeric or tetrameric) vola-tile products might be generated. However, oligomers larger thantetramers were not volatile enough and would remain within thesample [37]. The vapors changed the color of moist blue litmuspaper to red indicating an acidic property. It is believed that thevapors were composed of low-molecular weight esters andhydroxyl acids.

Fig. 2. (a) TG (b) DTG curves of mcl-PHA at different heating rates.

3.3. Thermal properties of mcl-PHA beforeand after thermal treatment

We have employed thermogravimetric analysis (TGA) tocompare the thermal stability of control PHA and those which hadbeen heated at 160 �C, 170 �C and 180 �C respectively. Fig. 4 showsthat the PHA (the control samples and those that had been subjectto heat treatment at 160 �C, 170 �C and 180 �C) was pyrolysed ina one-step process. The control and 160 �C and 170 �C-treated PHAstarted to decompose at around 200e250 �C, with the controlregistering a lower initial decomposition temperature. The lowestinitial decomposition temperature was however registered by the180 �C-treated PHAwhich started to decompose at around 100 �C. Apossible explanation could be that the 160 �C and 170 �C-treatedPHAs were only slightly degraded, leaving behind polymer chainswhich were more thermal stable, thus they showed higher thermalstability. The 180 �C-treated PHA sample contained lower-molec-ular weight species which could be more easily disintegrated byheating.

The glass transition temperature (Tg) for both control anddegraded PHA were shown in Table 2. For mcl-PHA derived fromoleic acid, low Tg around �44.1 �C was observed and there wasa decrease in Tg from �44.6 �C to �48.1 �C when PHA were heattreated at 160 �Ce180 �C. The decrease was primarily due to a lessordered arrangement of the polymer chains where the weakintermolecular forces that hold the polymer chains together hadbroken down at high temperature. However no melting tempera-ture (Tm) and enthalpy changes were observed in both controland degraded PHA, indicating the polyester was an amorphouselastomer.

Fig. 3. Kissinger plot for mcl-PHA.

Fig. 4. Relative thermal stability of control PHA and the degraded PHA as indicated bydifferent TG curves at 10 �C min�1.

Fig. 5. (a)Variation of acid numbers and Mn; (b) Carboxylic acid terminal concentra-tions for the mcl-PHA degraded at temperatures of 160 �C, 170 �C and 180 �C.

Mei Chan Sin et al. / Polymer Degradation and Stability 95 (2010) 2334e23422338

3.4. mcl-PHA before and after thermal treatment

3.4.1. End-group analysisThe control PHA had an acid number of 4.04 � 0.04 mg KOH g�1.

This value increased in proportion to increasing treatmenttemperature: 5.27 � 0.01 mg KOH g�1 (160 �C), 14.55 �0.02 mg KOH g�1 (170 �C) and 33.11 � 0.3 mg KOH g�1 (180 �C). APHA polymer chain contains one unit of hydroxyl and one unit ofcarboxylic end group. Thus the number of the carboxylic terminalsis equal to the number of the polymer chain. The concentration ofcarboxylic acid terminals and the number-average molecularweight (Mn) were then determined. The control PHA had thehighest Mn of 13890 g mol�1, while the Mn of the heat-treated PHAdecreased with increasing treatment temperatures. The Mn of the160 �C, 170 �C and 180 �C-treated PHAwas 77%, 28%, and 12% of theMn of the control PHA respectively. The decrease in Mn wasprimarily caused by random chain scission during thermaldecomposition of the PHA. The concentration of terminal carbox-ylic acid also increased with higher temperature, presumably dueto faster rate, as shown in Fig. 5. The data support the view thatmcl-PHAwere thermally decomposed at 160 �C, 170 �C and 180 �C.

3.4.2. GPC analysisThe average molecular weight and molecular weight distribu-

tion of control and degraded polymers were determined by GPCanalysis and the results are summarized in Table 2. It should bepointed out here that the GPC column was calibrated using PSstandards, and hence the Mn and Mw values are relative to PS.Results showed that number-averagemolecular weight (Mn) of PHAdecreased with increase in polydispersity index (PDI) when the

Table 2Average molecular weight, molecular weight distributions and glass transitiontemperatures of control and degraded PHA. Mn, number-average molecular weight;Mw, weight-average molecular weight; PDI, polydispersity index, defined byMw/Mn.

Sample Molecular weight Thermal property

Mn (Da) Mw (Da) PDI Tg (�C)

Control PHA 30600 60900 2.0 �44.1PHA degraded at 160 �C 29300 71400 2.4 �44.6PHA degraded at 170 �C 3500 16200 4.7 �45PHA degraded at 180 �C 2200 9000 4.0 �48.1

polymers were heat treated from 160 �C to 180 �C. The decrease inMn values is in the same trend as the end-group analysis. IncreasedPDI in degraded PHA indicated the polymers had broader molecularweight distribution with more lower-molecular weight speciesgenerated by the random scission of the ester linkages as a result ofthermal decomposition at high temperature.

Fig. 6. FTIR spectra of mcl-PHA. a) Control PHA b) PHA degraded at 160 �C c) PHAdegraded at 170 �C d) PHA degraded at 180 �C. The two arrows indicate distincteC¼Ceabsorption peak corresponding to the unsaturated terminal group in 180 �Cdegraded PHA.

Fig. 7. 400-MHz 1H NMR spectrum of oleic acid derived mcl-PHA. The protons in themcl-PHA structure were denoted by the corresponding letters in the spectrum.

Mei Chan Sin et al. / Polymer Degradation and Stability 95 (2010) 2334e2342 2339

3.5. Chemical structure analysis of control andthermally-degraded mcl-PHA

3.5.1. FTIR spectroscopyFTIR spectra of the control, 160 �C and 170 �C-treated PHA given

in Fig. 6a, b, c showed the carbonyl stretching band (C]O) ataround 1736 cm�1 and the other peaks were quite similar. Thepartial degradation at these two temperatures has producedshorter chains of basically similar structures. However the spec-trum for the 180 �C-treated PHA (Fig. 6d) showed two additional

Fig. 8. 1H-NMR spectra of mcl-PHA samples: (a) control, (b) degraded at 160 �C, (c) degraded(methyl group) has decreased, as hydrolysis of ester linkages would producing more hydro

absorption peaks, at 1653 cm�1 and 1414 cm�1. The band centeredat 1653 cm�1 is assignable to non-conjugated C]C stretchingvibration, and the one at 1414 cm�1 is due to ¼ CH-bending. Thesesignals indicate the presence of terminal vinyl groups.

3.5.2. 1H NMR spectroscopyFig. 7 shows the proton NMR spectrum for the control PHA. The

peak e at 0.8 ppm and peak d at 1.2 ppm were assigned to themethyl andmethylene group in the side chain respectively. The twopeaks b and a around 2.5 ppm and 5.1 ppm represented themethylene group at the a-position and methine group at the b-position of the ester respectively. The a-hydrogens in esters aredeshielded by the adjacent carbonyl group whereas b-hydrogenson the carbon attached to the single-bonded oxygen are deshieldeddirectly by the electronegative oxygen. While the peak shown at3.7 ppm was assigned to the proton of the hydroxyl groups andpeak shown at 2.1 ppm was assigned to the proton attached to thecarbon next to carboxyl groups. The signals for the olefinic protonsin the side chain appeared in the expected region between 5.2 and5.3 ppm. Another characteristic signal was at 1.9 and 2.4 ppmwhichbelonged to protons adjacent to the double bonds. Remainingsignal at 7.27 ppm was due to the CHCl3 in CDCl3.

The 1H NMR spectrum of oleic acid derived mcl-PHA producedby P. putida showed peaks at almost identical chemical shift in thespectrum studied by Huijbert et al. (1992) and Tan et al. (1997).According to Huijbert et al., the chemical shifts of peak j, k, g and hindicated the presence of unsaturated groups in some of themonomers’ side chains [23,38]. The peak f and lwith chemical shift1.9 ppm and 2.8 ppmwere assigned to allylic methylene group anddiallylic methylene group. Chemical shift of peak i was assigned tothe methylene group between the double bond in the side chainand the hydroxyl group of ester.

at 170 �C and (d) degraded at 180 �C. The ratio of proton a (methine group) to proton exyl acids with the eOH end group at 3.7 ppm.

Fig. 9. Expanded 1H NMR spectrum of PHA degraded at 180 �C, where the olefinicprotons m at 5.8 and n at 6.9 ppm.

Mei Chan Sin et al. / Polymer Degradation and Stability 95 (2010) 2334e23422340

Fig. 8 compares the 1H NMR spectra of control mcl-PHA andthose which had been treated at 160 �C, 170 �C and 180 �C. Theintensity ratio of signal a (proton originated from b methine groupadjacent to esterified alcohol group) to signal e (proton originatedfrom side chain methyl group) decreased in proportion toincreasing temperature. This showed that hydrolysis of ester bondhad occurred, generating free alcohol end groups as the b hydrogenwas no longer deshielded by the electronegative oxygen in the estergroup. This corresponds to an increase of intensity of the alcoholend groups, shown by the proton resonance at 3.7 ppm chemicalshift.

The 1H NMR spectra of 160 �C and 170 �C-treated PHA werealmost identical to the spectrum of control PHA. Nevertheless, twoextra peaks were observed at position between 5.0 and 6.0 ppm inthe 180 �C-treated PHA as shown in Fig. 9. These could be attributedto the olefinic protons in the degradation product. These observa-tions agree with the presence of monounsaturated terminalstructure reported by Kunioka and Doi (1990) [26].

The elucidation of intensity ratio of hydroxyl terminals tocarboxyl terminals in 160 �C, 170 �C and 180 �C-treated PHA wereshown in Fig. 10. From the 1H NMR spectra of degraded PHA

Fig. 10. 1H-NMR spectra of the thermolyzed PHA samples: (a) degraded at 160 �C, (b) deCH*eCOOH group is almost similar in the 160 �C and 170 �C degraded samples. Howeve

samples, amount of OH terminal groups and COOH terminal groupswere almost similar in the 160 �C- and 170 �C-treated PHA. Theseobservations suggested that hydrolysis at ester linkages occurred,leading to the production of a mixture of oligomeric hydroxyl acidsmethyl ester with almost equal concentration of hydroxyl andcarboxylic groups. However the presence of more carboxylicterminals than hydroxyl terminals was observed in spectrum of180 �C-treated PHA. One plausible reason for this phenomenon isthat the hydroxyl acids tend to dehydrate at high temperature.

3.5.3. GC analysisGC and 1H NMR analyses showed that oleic acid derived mcl-

PHA contained four saturated monomers: 3-hydroxyhexanoate(C6), 3-hydroxyoctanoate (C8), 3-hydroxydecanoate (C10) and 3-hydroxydodecanoate (C12) and three unsaturated monomers: 3-hydroxy-5-dodecenoate (C12:1Δ

5 ), 3-hydroxy-7-tetradecenoate(C14:1Δ7 ) and 3-hydroxy-5,8-tetradecenoate (C14:2Δ5,8 ). The polyestermainly consisted of the saturated units, C8 (42.3%) and C10 (30%),which seems to be a common feature for homology Group I Pseu-domonads [6,16]. The unsaturated monomer content in controlPHA was 13.6% of the total units. The change in monomericcomposition of PHA before and after thermal treatments wasshown in Fig. 11. The trend of variation in monomer compositionswas almost similar in 160 �C and 170 �C-treated PHA. The unsatu-rated monomer content of the polymer decreased when PHA washeat treated at 160 �C (9.1%) and 170 �C (8.9%). A decrease in thenumber of unsaturated monomers leads to a contention thatoxidative cleavage at unsaturated linkages in the side chain wouldoccurred at high temperature. Indeed number of shorter saturatedmonomers (C6, C8, C10 and C12) had increased in degraded PHA.However, the unsaturated monomer content in 180 �C-treated PHA(12.7%) was higher than both 160 �C- and 170 �C- treated PHA.Small amounts of C4:1 (1%) and C6:1 (1%) monomers were detectedin 180 �C-treated PHAwhich could be confirmed with the presence

egraded at 170 �C, and (c) degraded at 180 �C. The amount of eCH*eOH group andr, a lower level of eCH*eOH was observed in the 180 �C degraded PHA.

Fig. 11. Relative monomer composition of control and degraded PHA analyzed by GC.

Mei Chan Sin et al. / Polymer Degradation and Stability 95 (2010) 2334e2342 2341

of olefinic signals in the 1H NMR spectrum as shown in Fig. 9. Athigh temperature of 180 �C in the presence of excess terminalcarboxylic acids, small proportion of low-molecular weight speciesgenerated from the oxidative cleavage of unsaturated bond in theside chain would be dehydrated, generating non-hydroxylated C4:1and C6:1 fragments.

3.6. Thermal decomposition mechanism

FTIR, NMR and GC analyses were used as complementarytechniques to determine the molecular structure of the degraded

Fig. 12. A plausible chain cleavage mechanism in mcl-PHA during thermal degradationprocess. Stage I: Hydrolysis of the PHA ester bond. Stage II: Dehydration of hydroxylend group at 180 �C.

PHA, and with that to elucidate the mechanism of thermal degra-dation of the mcl-PHA.

The thermal degradation process appeared to involve a-chainscission of the mcl-PHA polymer chains via hydrolytic ester bondcleavage. The primary degradation products were a mixture of low-molecular weight oligomeric hydroxyalkanoic acids which resultedfrom the hydrolysis of the ester linkages. At 180 �C, a proportion ofthe degradation products have undergone dehydration of hydroxylend groups, giving rise to alkenoic acid as secondary products.Fig. 12 showed the plausible chain cleavagemechanism in mcl-PHApolymers during thermal degradation process.

A b-chain cleavage mechanism was proposed for the thermaldegradation of scl-PHA [25e31]. For scl-PHA such as PHB whichconsists of 4 carbon monomers with a methyl group as the sidechain, formation of a stable 6-membered ring ester intermediateand b hydrogen transfer at the ester linkage is possible during thethermal degradation. However, mcl-PHA is composed of six tofourteen carbon monomers, with the aliphatic R side chain having3, 5, 7, 9 or 11 carbons. These bulky R side chains could hinder theformation of a stable 6-membered ring ester intermediate. Thuscompared to the b hydrogen elimination reaction espoused for scl-PHA, hydrolysis of ester linkages might occur more readily for therandom chain scission reaction of mcl-PHA during thermaldegradation.

4. Conclusion

Thermodegradation of PHAwould have a dramatic effect on thethermal properties and physical properties of the polymer and canresult in a progressive reduction in molecular weight, at the sametime, generating useful hydroxyl acid oligomers. The latter could beused as building blocks or intermediates for the synthesis of otheruseful biomaterials. 180 �C-treated PHA contained the highest acidnumber, the highest concentration of terminal carboxylic acidsand the lowest number-average molecular weight. The acidity ofthe polymeric hydroxyl acids increased as the number-averagemolecular weight decreased. Analyses of FTIR, 1H NMR and GCsuggests thermal decomposition of mcl-PHA was due to thehydrolysis of the ester linkages, thus producing a mixture of lower-molecular weight hydroxyacids with proportional increase in thehydroxyl and carboxyl end groups. Heating at 180 �C would lead todehydration of the hydroxyl terminal groups, producing alkenoicacids as another end product.

Acknowledgement

This study was supported by funding, Vote F PS155/2009A fromUniversity of Malaya.

References

[1] Dawes EA, Senior PJ. The role and regulation of energy reserve polymers inmicroorganisms. Adv Microbiol Physiol 1973;10:135e266.

[2] Doi Y. Microbial polyesters. New York: VCH; 1990.[3] Du G, Yu J. Green technology for conversion of food scraps to biodegradable

thermoplastic polyhydroxyalkanoates. Environ Sci Technol 2002;36:5511e6.[4] Annuar MSM, Tan IKP, Ramachandran KB. Evaluation of nitrogen sources for

growth and production of polyhydroxyalkanoates from palm kernel oil byPseudomonas putida. Asia Pac J Mol Biol Biotechnol 2008;16(1):11e5.

[5] Wang JG, Bakken LR. Screening of soil bacteria for polyehydroxybutyric acidproduction and its role in the survival of starvation. Microbiol Ecol 1998;35:94e101.

[6] Madison LL, Huisman GW. Metabolic engineering of poly(3-hydrox-yalkanoates): from DNA to plastic. Microbiol Mol Biol Rev 1999;63:21e53.

[7] Lee SY, Lee Y, Wang FL. Chiral compounds from bacterial polyesters: sugars toplastics to fine chemicals. Biotechnol Bioeng 1999;65(3):363e8.

[8] Ballistreri A, Giuffrida M, Guglielmino SPP, Carnazza S, Ferreri A,Impallomeni G. Biosynthesis and structural characterization of medium-

Mei Chan Sin et al. / Polymer Degradation and Stability 95 (2010) 2334e23422342

chain-length poly(3-hydroxyalkanoates) produced by Pseudomonas aerugi-nosa from fatty acids. Bio Macromol 2001;29:107e14.

[9] Abe C, Taima Y, Nakamura Y, Doi Y. New bacterial copolyesters of 3-hydrox-yalkanoates and 3-hydroxy-ώ-fluoroalkanoates produced by Pseudomonasoleovorans. Polym Commun 1990;31:404e6.

[10] Fritzsche K, Lenz RW, Fuller RC. Production of unsaturated polyesters byPseudomonas oleovorans. Int J Biol Macromol 1990;12:85e91.

[11] Fritzsche K, Lenz RW, Fuller RC. Bacterial polyesters containing branched poly(-hydroxyalkanoate) units. Int J Biol Macromol 1990;12:92e101.

[12] Fritzsche K, Lenz RW, Fuller RC. An unusual bacterial polyester with a phenylpendant group. Macromol Chem 1990;191:1957e65.

[13] Kim YB, Lenz RW, Fuller RC. Poly(-hydroxyalkanoate) copolymers containingbrominated repeating units produced by Pseudomonas oleovorans. Macromol1992;25:1852e7.

[14] Choi MH, Yoon SC. Polyester biosynthesis characteristics of Pseudomonas cit-ronellolis grown on various carbon sources, including 3-methyl-branchedsubstrates. Appl Environ Microbiol 1994;60:3245e54.

[15] Preusting H, Nijenhuis A, Witholt B. Physical characteristics of poly(3-hydroxyalkanoates and poly(3-hydroxyalkenoates) produced by Pseudomonasoleovorans grown on aliphatic hydrocarbons. Macromol 1990;23:4220e4.

[16] Huisman GW, de Leeuw O, Eggink G, Witholt B. Synthesis of poly-3-hydrox-yalkanoates is a common feature of fluorescent pseudomonads. Appl EnvironMicrobiol 1989;55:1949e54.

[17] Timm A, Steinbuchel A. Formation of polyesters consisting of medium-chain-length 3-hydroxyalkanoic acids from gluconate by Pseudomonas aeruginosa andother fluorescent pseudomonads. Appl Environ Microbiol 1990;56:3360e7.

[18] Kiska DL, Gilligan PH. Pseudomonas. In: Murray PR, Baron EJ, Pfaller MA,Tenover FC, Yolken RH, editors. Manual of Clinical Microbiology. 7th ed. ASMP;1999. p. 517e25.

[19] Brandl H, Gross RA, Lenz RW, Fuller RC. Pseudomonas oleovorans as a source ofpoly(-hydroxyalkanoates) for potential applications as biodegradable poly-esters. Appl Environ Microbiol 1988;54:1977e82.

[20] Lageveen RG, Huisman GW, Preusting H, Ketelaar P, Eggink G, Witholt B.Formation of polyesters by Pseudomonas oleovorans: effect of substrates onformation and composition of poly(R)-3-hydroxyalkanoates and poly(R)-3-hydroxyalkenoates. Appl Environ Microbiol 1988;54:2924e32.

[21] Eggink G, de Waard P, Huijberts GNM. FEMS Microbiol Rev 1992;103:159.[22] Eggink G, van der Wal H, Huijberts GNM, de Waard P. Oleic acid as a substrate

for poly-3-hydroxybutyrate formation in Alcaligenes eutrophus and Pseudo-monas putida. Ind Crops Prod 1993;1:157e63.

[23] Tan IKP, Sudesh K, Theanmalar M, Gan SN, Gordon III B. Saponified palmkernel oil and its major free fatty acids as carbon substrates for the productionof polyhydroxyalkanoates in Pseudomonas putida PGA1. Appl Microbiol Bio-technol 1997;47:207e11.

[24] Platt DK. Biodegradable polymers market report. Smithers Rapra Limited;2006.

[25] Grassie N, Murry EJ, Holms PA. The thermal degradation of poly(-(D)-b-hydroxybutyric acid): part 3- the reaction mechanism. Polym Degrad Stab1984;6:127e34.

[26] Kunioka M, Doi Y. Thermal degradation of microbial copolyesters: poly(3-hydroxybutyrate-co-3-hydroxyvalerate) and (3-hydroxybutyrate-co-4-hydrox-ybut -yrate). Macromol 1990;23:1933e6.

[27] Nguyen S, Yu G, Marchessault R. Thermal degradation of poly(3-hydrox-yalkanoates): preparation of well-defined oligomers. Biomacromol 2002;3(1):219e24.

[28] Lehrle RS, Williams RJ. Thermal degradation of bacterial poly(hydroxybutyricacid): mechanism from the dependence of pyrolysis yields on sample thick-ness. Macromol 1994;27:3782e9.

[29] Lehrle RS, Williams RJ, French C, Hammond T. Thermolysis and methanolysisof poly(b-hydroxybutyrate): random scission assessed by statistical analysisof molecular weight distributions. Macromol 1995;28:4408e14.

[30] Ballistreri A, Montaudo G, Garozzo D, Giuffrida M, Montaudo MS. Micro-structure of barterial poly(b-hydroxybutyrate-co-b-hydroxyvalerate) by fastatom bombardment mass spectrometry analysis of the partial pyrolysisproducts. Macromol 1991;24:1231e6.

[31] Abate R, Ballistreri A, Montaudo G, Impallomeni G. Thermal degradation ofmicrobial poly(4-hydroxybutyrate). Macromol 1994;27:332e6.

[32] Erceg M, Kovacic T, Klaric I. Dynamic thermogravimetric degradation of poly(3-hydroxybutyrate)/aliphatic-aromatic copolyester blends. Polym DegradStab 2005;90:86e94.

[33] Lee JY, Choi HK, Shim MJ, Kim SW. Estimation of cure rate for DGEBA/MDA/PGE-AcAm system by Kissinger expression. Appl Chem 1997;1(2):714e7.

[34] Doyle CD. Kinetic analysis of thermogravimetric data. J Appl Polym Sci1961;5:285.

[35] Krishnan K, Viswanathan G, Kurian AJ, Ninan KN. Kinetics of decomposition ofnitramine propellant by differential scanning calorimetry. Def Sci J 1992;42(3):135e9.

[36] Radhakrishnan Nair MN, Thomas George V, Gopinathan Nair MR. Ther-mogravimetric analysis of PVC/ELNR blends. Polym Degrad Stab; 2007:189e96.

[37] Gonzalez A, Irusta L, Fernandez-Berridi MJ, Iriarte M, Iruin JJ. Application ofpyrolysis/ gas chromatography/ fourier transform infrared spectroscopy andTGA techniques in the study of thermal degradation of poly(3-hydrox-ybutyrate). Polym Degrad Stab 2005;87:347e54.

[38] Huijberts GNM, Eggink G, de Waard P, Huisman GW, Witholt B. Pseudomonasputida KT2442 cultivated on glucose accumulates poly(3-hydroxyalkanoates)consisting of saturated and unsaturated monomers. Appl Environ Microbiol1992;58(2):536e44.