SEC-MALS Characterization of Microbial Polyhydroxyalkanoates

Biomacromolecules 2004, 5, 628-636

628

SEC-MALS Characterization of Microbial Polyhydroxyalkanoates Ema Zˇ agar* and Andrej Krzˇ an National Institute of Chemistry, Hajdrihova 19, 1000 Ljubljana, Slovenia Received November 5, 2003; Revised Manuscript Received December 8, 2003

Characterization of poly-3-hydroxybutyric acid (PHB) and poly-3-hydroxybutyric-co-valeric acid (PHBV, 13% valerate) in chloroform was performed using size exclusion chromatography coupled to a multi-angle light scattering detector (SEC-MALS). Absolute molar mass averages, molar mass distribution, and the radius of gyration were determined. Three sample preparation methods were examined: dissolution in chloroform (1) at room temperature, (2) at 60 °C, and (3) after thermal pretreatment of samples (annealing at 180 °C with subsequent quenching in liquid nitrogen). Dissolution at 60 °C and dissolution of thermally pretreated samples gave molecularly dissolved PHB and PHBV. At 60 °C using acid free chloroform, there was no indication of degradation for up to 120 min dissolution time, whereas thermal degradation of polymers did take place during annealing at 180 °C. The degradation rate constants for number and weight average degree of polymerization at 180 °C were slightly higher for PHB (5.19 × 10-5 min-1, 4.95 × 10-5 min-1) than for PHBV (4.99 × 10-5 min-1, 4.54 × 10-5 min-1). The dependence of the radii of gyration on molar mass showed that both polymers form random coils in chloroform. The relationship between the absolute molar masses and relative SEC results was determined. DSC and NMR characterization also gave evidence of the progress of degradation. Introduction Biodegradable polymers synthesized through biosynthetic routes have received much attention as environmentally friendly materials able to replace biostable plastics for packaging and other uses. A leading group among these materials are polyhydroxyalkanoates (PHAs), aliphatic polyesters based on simple hydroxyacids with differing alkyl side chains as their monomer structures (hydroxy-butyric, valeric, hexanoic acids, etc), as well as their copolymers. Since PHAs are biocompatible and bioabsorbable materials, they have many potential applications in the medical field, for example, in tissue engineering and controlled drug release applications.1-6 Determination of molar mass averages (MMA) and molar mass distribution (MMD) of PHA is important in evaluating the biosyntheses carried out by various strains of microorganisms on different feedstocks as well as in monitoring PHA degradation. Detailed investigations focused on PHA MMA and MMD determination methods are scarce.1 PHA numberaverage (M h n) and weight-average (M h w) molar masses and MMD are usually obtained by size exclusion chromatography (SEC) with chloroform as the preferred solvent.7-12 For low molar mass PHAs, tetrahydrofuran (THF) and dioxane were also used as solvents.7,9 The SEC columns were calibrated using various polymer standards (polystyrene, polyethylenoxyde)7-11 or a universal calibration method.12-15 Absolute weight average molar masses were determined in 2,2,2-trifluoroethanol (TFE) using static light scattering measurements11,14,16-20 or sedimentation measurements in an * To whom correspondence should be addressed. E-mail: [email protected].

ultracentrifuge.16 The Mark-Houwink-Sakurada parameters relating the intrinsic viscosity, [η], to the molar mass have been determined for poly-3-hydroxybutyric acid (PHB) in TFE and chloroform.16-18 The major problem with PHA molar mass determinations is their poor solubility in common organic solvents due to their high molar masses and high degree of crystallinity.17,21 It is difficult for the solvent to penetrate into crystalline domains which can thus also delay the solvation of the amorphous part of the polymer.21 Furthermore, the polar groups of PHA can interact with the column packing material, leading to the adsorption of the polymer on the column and therefore to erroneous molar mass determinations. Recently, Marchessault et al.21 studied PHA solubility and classified 39 solvents for PHB according to the threedimensional solubility parameters defined by Hansen.22 This semiempirical approach is limited in its applicability to amorphous PHA only and cannot account for the considerable effects of crystallinity and molar mass on solubility. Based on the solubility parameter concept, Marchessault et al. have suggested two approaches for preparing PHA solutions. One is by dissolving PHA in a solvent at elevated temperature, and the other is by transforming the semicrystalline PHB into an amorphous material by melting and subsequently quenching the sample in liquid nitrogen prior to dissolving. As both proposed methods involve thermal treatment of PHA by itself or in solution, there is a question as to whether the sample may undergo degradation during this process. PHA is known to degrade at elevated temperatures (thermal

10.1021/bm030073l CCC: $27.50 © 2004 American Chemical Society Published on Web 01/30/2004

SEC-MALS Characterization of Microbial PHAs

degradation)8,13,23-25 or under acidic or basic conditions (hydrolytic degradation).14 Thermal decomposition of PHA proceeds through a random scission process involving a sixmembered ring ester intermediate (β-elimination reaction), which results in the formation of unsaturated olefinic and carboxyl end groups.8,23-24 Some authors have suggested that the early stages of this process are delayed due to condensation between terminal hydroxyl and carboxyl groups23 or due to intramolecular exchange reactions.25 In this work, we present a study on the determination of absolute MMA and MMD of two biodegradable aliphatic polyesters (homopolyester poly-3-hydroxybutyric acid, PHB, and polyhydroxybutyric-co-valeric acid copolyester, PHBV) by size exclusion chromatography coupled to a multi-angle laser light scattering photometer (SEC-MALS) in chloroform. The aim of our work was to define an appropriate procedure for sample solution preparation in which the polymers will be dissolved on a molecular level with no degradation effects on MMA and MMD. We considered three procedures for preparing the sample solutions, as well as the possible effect of solution concentration on MMA and MMD. Polymer conformation was inferred from the dependence of the radius of gyration on molar mass. Degradation rate constants at 180 °C were determined. The absolute PHA M h w values were compared to their relative M h w values determined by SEC using polystyrene calibration curve, to establish their relationship under the experimental conditions used in SEC. Experimental Section Materials. The microbially synthesized, optically active homopolyester poly-3-hydroxybutyric acid (PHB) and polyhydroxybutyric-co-valeric acid copolyester (PHBV, 3-hydroxyvaleric acid content 13 mol %, determined by 1H NMR) used in this study were supplied by Biocycle, PHB Industrial S/A, Brazil. The powders of both polyester samples were dried to a constant weight at room temperature under vacuum and stored in a desiccator prior to use. Acid free chloroform (J. T. Baker) was used for sample preparation. Sample Preparation for SEC-MALS Measurements. Three procedures were employed for the preparation of SECMALS samples: (1) dissolution in chloroform at room temperature, (2) dissolution in chloroform at 60 °C for 60, 90, and 120 min, and (3) thermal pretreatment prior to dissolution in chloroform. In the later procedure, the samples were heated to 180 °C in a closed DSC pan and kept at that temperature for 1, 2, 4, 6, 8, and 10 min. Then the samples were quenched by submerging the pans with the encapsulated PHA into liquid nitrogen. Chloroform was added to the thus treated samples, and after a few minutes, clear solutions were obtained. Characterization. SEC-MALS measurements were performed at 25 °C using a Hewlett-Packard 1100 series pump coupled to a Dawn-DSP laser photometer (LS detector) equipped with a He-Ne laser (λ0 ) 633 nm) and an OptilabDSP interferometric refractometer (DRI detector, both instruments are from Wyatt Technology Corp., U.S.A.). Separations were carried out using a 5 µm PSS SDV linear XL column (300 mm length and 8 mm ID) with a precolumn

Biomacromolecules, Vol. 5, No. 2, 2004 629

(Polymer Standards Service GmbH, Mainz). The PSS SDV linear XL column contains a mixture of individual particle pore size materials, accurately blended to cover a range of molar masses from 500 g mol-1 to 3 × 106 g mol-1. Acid free chloroform was used as the eluent with a nominal flow rate of 1.0 mL min-1. The mass of the samples injected onto the column was typically (1.5-2.5) × 10-4 g, whereas the solution concentration was in the range of (0.3-0.5) %. The calculation of M h w from MALS requires a sample specific refractive index increment (dn/dc), which was determined from the differential refractometer response assuming a 100% sample mass recovery from the column. Data acquisition and evaluation was done using Astra 4.73.04 software (Wyatt Technology Corp.). Since the MALS detector is not particularly sensitive to low molar mass species, the sample molar mass averages were recalculated using Corona 1.40 software (Wyatt Technology Corp.), where the scattered data points at high elution volumes were fitted using linear regression.26 The molecular conformation was determined from the log-log plot of the root-mean-square (RMS) radius versus molar mass (M), since the RMS radius of linear macromolecules is proportional to its characteristic size parameter, which is specific for the particular particle shape.27 SEC measurements were performed on a Perkin-Elmer liquid chromatograph equipped with an LC-30 differential refractometer (DRI). The column was calibrated with polystyrene (PS) standards of low polydispersities. All other experimental conditions were as in SEC-MALS measurements. Differential scanning calorimetry (DSC) experiments were performed on a Pyris 1 Perkin-Elmer DSC apparatus for determining sample melting behavior and nonisothermal crystallization. The melting point (Tm) and heat of fusion (∆Hm) of Indium were used for apparatus calibration. Melting endotherms after annealing were determined by heating the samples from 0 to 180 °C at a heating rate of 200 °C min-1 followed by annealing for 2, 6, 10, and 300 min. They were then cooled to 0 °C at a cooling rate of 200 °C min-1 and after 5 min subsequently heated (heating rate 10 °C min-1) to 180 °C to obtain the DSC second heating curves. Nonisothermal crystallization exotherms of the polymers obtained in the melt or cold crystallization process are characterized by the peak temperature (Tc), where the value of heat flow is at a maximum and the sample crystallizes the fastest, and by the values of crystallization enthalpies (∆Hc). The 1H (300 MHz) and 13C (75 MHz) NMR spectra were recorded on a Varian VXR 300 NMR spectrometer using CDCl3 as the solvent and tetramethylsilane (TMS) as the internal reference. The solution concentration was (1-5) wt %, depending on whether 1H or 13C NMR spectra were recorded. 13C spectra were recorded with a relaxation delay of 5 s, an acquisition time of 3 s and up to 20 000 repetitions, while 1H spectra were recorded with a relaxation delay of 5 s, an acquisition time of 3 s and up to 500 repetitions. Results and Discussion Dissolution in Chloroform at Room Temperature. At room temperature, PHB and PHBV are poorly soluble in

630

Zˇ agar and Krzˇ an

Biomacromolecules, Vol. 5, No. 2, 2004

Figure 1. Molar mass vs elution volume curve (‚‚‚) and SEC-MALS chromatogram of PHB dissolved in CHCl3 at room temperature: (s) LS response at 90° angle, (- - -) RI response M h n ) 2.462 × 105 g mol-1, M h w ) 6.065 × 105 g mol-1, M h z ) 4.340 × 107, In ) 2.46, Iw ) 72, Rz ) 77 nm, 91% mass recovery.

Figure 2. Light scattering elution curves of PHB at 16 different observation angles.

chloroform. Figure 1 shows a SEC-MALS chromatogram of PHB dissolved in chloroform at room temperature followed by filtration through a 0.45 µm filter prior to injection. The differential refractometer (DRI) shows a monomodal MMD, whereas light scattering at 90° angle (LS) shows a bimodal MMD. The intense signal at low elution volume in the 90° LS chromatogram represents only about 1% of the injected sample mass and was therefore barely detected by the DRI detector. The calculated M h n and M h w of the low elution volume peak in the 90° LS chromatogram are about 2 orders of magnitude higher (107) than that of the second peak at larger elution volume (105). The fact that the first peak in the LS chromatogram represents very large particles of high molar masses is confirmed by a strong decrease in the intensity of the LS signal with increasing angle of observation as a consequence of intramolecular interference of scattered light27 (Figure 2). The high molar mass species can be attributed to incompletely dissolved PHB. This would suggest that molar mass averages for a PHB prepared in this manner sample are overestimated and that the molar mass distribution is broader than it should be for a molecularly dissolved sample. Due to incomplete sample dissolution, the recovered mass of the sample from

Figure 3. Molar mass vs elution volume curves and SEC-MALS chromatograms of PHB and PHBV dissolved in CHCl3 at 60 °C, 60 min: (s) LS response at 90° angle, (- - -) RI response.

the column was only 91%. The fact that dissolution of PHB in chloroform at room temperature does not give a molecularly dissolved polymer is missed when using relative SEC with a DRI detector, since it gives a normal monomodal sample molar mass distribution. Dissolution in Chloroform at 60 °C. Clear solutions of both PHB and PHBV in chloroform at 60 °C were obtained after a dissolution time of 60 min. Since aliphatic polyesters are prone to undergo hydrolytic degradation in acidic or basic conditions,14 acid free chloroform was used for solution preparation. The SEC-MALS chromatograms of both polymers gave a monomodal molar mass distribution in both DRI and LS traces (Figure 3) indicating complete dissolution at the molecular level. Comparison between the two polymers of the same molar mass shows that PHBV elutes at a lower elution volume than PHB (Figure 3). This can be explained by the fact that at the same molar mass PHBV has a larger hydrodynamic volume than PHB due to the larger ethyl side chains of hydroxyvalerate (HV) units compared to the smaller methyl side group of hydroxybutyrate (HB) units. The determined MMA and Rz of PHB are higher, whereas its MMD is somewhat narrower, compared to the corresponding values for PHBV (Tables 1 and 2). The relation between h w value absolute M h w and polystyrene (PS)-equivalent M obtained under the used chromatographic conditions is M h w(absolute) ) 0.91 × M h w(PS-equivalent) for PHB and M h w(absolute) ) 0.82 × M h w(PS-equivalent) for PHBV. The relation between absolute and PS-equivalent M h w values for PHB was previously determined by Doi et al.,11 using PS-equivalent M h w values determined at 40 °C. The quotient of seven PHB samples with different molar mass averages varied between 0.65 and 0.85. These values are lower than in our case, 0.91, most probably due to the higher temperature used in SEC. The MMA and MMD values and the z-average root-meansquare (RMS) radius (Rz) of PHB and PHBV were not influenced by the dissolution time, 60, 90, and 120 min


SEC-MALS Characterization of Microbial PHAs Table 1. SEC-MALS Results of PHB in CHCl3a

dissolution in CHCl3 at 60 °C

M hn × averageb

10-5,

g

mol-1

M hw ×

1.783

10-5,

g mol-1

M h z × 10-5, g mol-1

In

Iw

Rz, nm

5.754

1.92

1.68

43.5

3.424

dissolution of thermally pretreated and quenched sample in CHCl3 180 °C annealing time, min

M h n × 10-5, g mol-1

M h w × 10-5, g mol-1

M h z × 10-5, g mol-1

In

Iw

Rz, nm

1 2 4 6 8 10

1.495 1.322 1.162 1.092 1.018 0.852

2.862 2.592 2.233 1.994 1.836 1.694

4.748 4.180 3.588 3.214 2.831 2.623

1.91 1.96 1.92 1.83 1.80 1.99

1.66 1.61 1.61 1.61 1.54 1.55

37.9 33.3 29.0 26.2 20.5 18.7

dependence on solution concentration 180 °C, 1 min injected mass, g

M hn ×

1.52 × 10-4 2.03 × 10-4 2.52 × 10-4 average

10-5,

g

mol-1

M h w × 10-5, g mol-1

M h z10-5, g mol-1

In

Iw

Rz, nm

2.895 2.843 2.848 2.862

4.928 4.695 4.621 4.748

2.05 1.88 1.83 1.91

1.70 1.65 1.62 1.66

37.6 37.9 38.3 37.9

1.413 1.513 1.559 1.495

a Molar mass averages were calculated using Corona software. b The quoted average molar masses are average values determined at dissolution times 60, 90, and 120 min. The injected mass of sample on the column was approximately 2.5 × 10-4 g.

Table 2. SEC-MALS Results of PHBV in CHCl3a dissolution in CHCl3 at 60 °C

M hn × averageb

10-5,

g

mol-1

M hw ×

0.738

10-5,

g mol-1

M h z × 10-5, g mol-1

In

Iw

Rz, nm

2.978

2.12

1.91

30.4

1.562

dissolution of thermally pretreated and quenched sample in CHCl3 180 °C annealing time, min

M h n × 10-5, g mol-1

M h w × 10-5, g mol-1

M h z × 10-5, g mol-1

In

Iw

Rz, nm

1 2 4 6 8 10

0.672 0.65 0.625 0.567 0.55 0.531

1.491 1.393 1.278 1.199 1.119 1.076

2.745 2.446 2.147 2.095 1.835 1.757

2.22 2.14 2.04 2.11 2.03 2.03

1.84 1.76 1.68 1.75 1.64 1.63

28.6 25.4 23.8 22.5 21.0 20.5

dependence on solution concentration 180 °C, 1 min injected mass, g 10-4

1.54 × 2.05 × 10-4 2.44 × 10-4 average

M hn ×

10-5,

g

mol-1

M h w × 10-5, g mol-1

M h z × 10-5, g mol-1

In

Iw

Rz, nm

1.521 1.463 1.490 1.491

2.773 2.666 2.780 2.740

2.44 2.13 2.11 2.22

1.82 1.82 1.86 1.84

29.6 28.8 27.4 28.6

0.623 0.687 0.706 0.672

a Molar mass averages were calculated using Corona software. b The quoted average molar masses are average values determined at dissolution times 60, 90, and 120 min. The injected mass of sample on the column was approximately 2.5 × 10-4 g.

(Tables 1 and 2), indicating that the polymers did not undergo degradation during dissolution in acid free chloroform at 60 °C. The conformation of PHB and PHBV macromolecules was inferred from the slopes of the log-log plots representing the radius of gyration as a function of molar mass27,28 (Figure 4, parts A and B). For both samples, the slope was between 0.5 and 0.6, indicating a random coil conformation in chloroform. These results are consistent with the results of Miyaki et al.,17 who found that indices for molar mass dependence of intrinsic viscosity and radius of gyration of PHB approached the limits for random coil molecular conformation in chloroform, and with the NMR results of Doi et al.,29 who showed that PHA molecules are flexible chains in chloroform.

Thermal Pretreatment of Samples Prior to Dissolution in Chloroform. In this procedure, the semicrystalline PHB and PHBV samples were transformed into an amorphous material by melting them at 180 °C, followed by quenching to the temperature of liquid nitrogen. Chloroform was then immediately added to the quenched samples. After several minutes, clear solutions were obtained. Completely dissolved samples were obtained after annealing times of 1 min or more. To minimize thermal degradation, we chose the lowest temperature at which the melting of both samples was completed (180 °C) and the shortest annealing time (1 min), which allowed complete dissolution of the samples in chloroform. SEC-MALS chromatograms of thermally pretreated samples at 180 °C for 1 min gave monomodal MMD

632



Figure 4. Plot of RMS radius vs molar mass in CHCl3. (a) PHB, slope 0.55; (b) PHBV; slope 0.58.

Figure 5. DRI (A) and 90° LS (B) chromatograms of thermally pretreated (from the left to the right 1, 6, and 10 min at 180 °C) and then quenched PHB in CHCl3.

Figure 6. DRI (A) and 90° LS (B) chromatograms of thermally pretreated (from the left to the right 1, 6, and 10 min at 180 °C) and then quenched PHBV in CHCl3.

(Figures 5 and 6). The calculated MMA were lower by approximately 16% and 7% for PHB and PHBV, respectively, relative to the values obtained for the samples dissolved in chloroform at 60 °C (Tables 1 and 2). The molar mass averages of thermally pretreated samples were independent of solution concentration (Tables 1 and 2) thus indicating dissolution at the molecular level.

With increasing annealing time, the peaks in DRI and LS chromatograms of both polymers moved to higher elution volumes. In addition, the intensity of the LS signal decreased considerably with increasing annealing time (Figures 5 and 6). These results suggest a decrease in polymer hydrodynamic volume, which is caused by a decrease in sample molar mass as a result of increasing degradation. At 10 min annealing



time, the obtained molar mass averages of PHB and PHBV were lower by approximately 51% and 29%, respectively, compared to the corresponding molar mass values determined for samples dissolved in chloroform at 60 °C. The MMA, polydispersity index and Rz as a function of annealing time are presented in Tables 1 and 2. Thermal Degradation of PHB and PHBV at 180 °C. SEC-MALS Analysis. The mechanism of thermal degradation of PHA in the temperature range 170-200 °C involves random chain scission at the ester group accompanied by a decrease in molar mass.8,23,24 During the random degradation process, each polymeric repeat unit has the same probability of undergoing scission. The nonuniformity indices of numh w/P hn ber- (In) and weight- (Iw) distribution functions (In ) P and Iw ) P h z/P h w) of randomly degrading polymers theoretically approach the values of 2.0 and 1.5, respectively,30-33 which are characteristic of the ideal Schulz-Flory distribution. In our experiments, the In and Iw of PHB during annealing approached the values of 1.91 and 1.55, respectively, whereas In and Iw for PHBV approached the values of 2.03 and 1.63, respectively (Tables 1 and 2). Another more detailed criterion for random chain scission is that the reciprocal value of the average degree of polymerization (1/P h x,t; x ) n, w, z) should be a linear function of time (t). The following mathematical equations describe h w,t), and zthe time dependence of number- (P h n,t), weight- (P (P h z,t) average degree of polymerization of a randomly degrading polymer:32-34

Figure 7. Relationship between 1/P h x,t - 1/P h x,0 and annealing time (t) at 180 °C for PHB; x ) n ([), w (9), and z (2).

1 1 ) kd,nt P h n,t P h n,0

(1)

1 1 ) (kd,w/3)Iwt P h w,t P h w,0

(2)

Figure 8. Relationship between 1/P h x,t - 1/P h x,0 and annealing time (t) at 180 °C for PHBV; x ) n ([), w (9), and z (2).

1 1 ) (kd,z/2)(Iz - 2/3)t P h z,t P h z,0

(3)

Table 3. Rate Constants, kd,n, kd,w, of Thermal Degradation at 180 °C for PHB and PHBV

where P h n,0, P h w,0, P h z,0 are the initial number-, weight-, and z-average degrees of polymerization, respectively. In our h w,0, and P h z,0 represent the values obtained for case, P h n,0, P h z/P h w and Iz samples dissolved in chloroform at 60 °C. Iw ) P ) P h z+1/P h z are the indices of molecular nonuniformity of weight and z-distribution functions of the degree of polymerization. kd,n, kd,w, and kd,z are the rate constants of thermal degradation for sample number-, weight-, and z-average degree of polymerization, respectively. The linear relationship between 1/P h x,t - 1/P h x,0 (x ) n, w or z) and time t for PHB and PHBV shown in Figures 7 and 8 confirms that PHB and PHBV molar mass averages during thermal degradation at 180 °C follow a kinetic model of random chain scission. These results are in agreement with the results of Kunioka and Doi,8 who studied thermal degradation of PHB and PHBV copolymers between 100 and 200 °C, with the total pyrolysis study of PHB at 350 °C by Lehrle et al.,25 and with a degradation study of PHB in a nonaqueous chloroform-methanol mixture in the presence of 4-toluensulfonic or imidazole catalyst14 by Majid et al.14 Degradation rate constants (kd,n, kd,w) for both polymers were determined from the slopes of the plots (eqs 1 and 2, Table

PHB PHBV

kd,n × 105, min-1

kd,w × 105, min-1

5.19 4.99

4.95 4.54

3) and are slightly higher for PHB than for PHBV. The degradation rate constants for the number average degree of polymerization (kd,n) at 180 °C were determined for PHB and PHBV with varying contents of HV units by Doi et al.;8 their values are comparable to the kd,n values obtained in our study. DSC Analysis. The degradation in dependence of annealing time at 180 °C was also studied by DSC. The DSC thermogram of the original PHB before thermal treatment (first heating) showed two melting peaks (Tm1 ) 150.1 °C and Tm2 ) 172.4 °C, Figure 9A, Table 4) due to segregation of crystalline material of lower crystal thickness from the regions with larger crystalline phases.13 After annealing at 180 °C for different annealing times (2, 6, and 10 min), PHB crystallized from the melt during cooling. The Tc values of the samples annealed for up to 10 min increased slightly with longer annealing times (from 95.8 to 97.3 °C, Figure 10, Table 4). This is most probably a

634



Table 4. DSC Results of PHB for Various Annealing Times at Annealing Temperature 180 °C first heating; 10 °C min-1 PHB

Tm1/°C

Tm2/°C

150.1

172.4

∆Hm/Jg-1 36.92

79.43

First cooling; 200 °C min-1 Tc/°C -∆Hc/Jg-1

annealing time at 180 °C/min 2 6 10 300

95.8 96.5 97.3 93.5

57.17 59.08 59.52 63.58

second heating; 10 °C min-1 2 6 10 300

Tc/°C

-∆Hc/Jg-1

92.2 92.7 92.8 92.8

10.24 10.47 10.58 7.70

second heating; 10 °C min-1 2 6 10 300

Tm/°C

∆Hm/Jg-1

170.2 170.1 169.7 151.6; 161.6

90.45 90.24 90.07 51.73; 41.20

Figure 9. A. DSC melting traces of original PHB and annealed PHB for 2, 6, 10, and 300 min at 180 °C and then crystallized nonisothermally from the melt at the cooling rate 200 °C min-1 (heating rate 10 °C min-1). B. DSC melting trace of annealed PHB for 6 min at 180 °C and then crystallized nonisothermally from the melt at the cooling rate 200 °C min-1 (heating rate 10 °C min-1).

consequence of decreasing molar mass causing higher mobility of macromolecular chains and improved packing of chain segments.9 Therefore, the lower molar mass PHB needed less supercooling and the beginning of crystallization started at higher temperatures. In contrast, the Tc of the PHB sample annealed for 300 min shifted to a lower temperature, 93.5 °C (Figure 10, Table 4). This is most probably due to steric hindrances created by the larger amount of free end groups produced by degradation, which have a greater effect on the beginning of crystallization than the decreasing molar mass.9 Heating from the glassy state (second heating) revealed some cold crystallization of PHB with a Tc located around 93 °C (Figure 9). Cold crystallization enthalpies, ∆Hc, are significantly lower than those observed in crystallization from the melt (Table 4). With further heating, a melting endotherm was obtained. Samples annealed at 180 °C for up to 10 min showed only one melting peak, which is contrary to the first heating trace. The Tm were slightly lower than the Tm2 of the original PHB. The PHB sample annealed for 300 min showed two melting peaks, similar to the original PHB; the lower Tm1 values were comparable, whereas the higher Tm2

Figure 10. Crystallization exotherms of PHB during nonisothermal crystallization after annealing at 180 °C for 2, 6, 10, and 300 min (cooling rate 200 °C min-1).

dropped by 11 °C. The ∆Hm1 was higher than ∆Hm2, which is reverse to the values of the original sample in first heating scan (Figure 9A, Table 4). The significant differences in the DSC trace of the 300 min annealed PHB are a consequence of substantial degradation that results in a higher amount of macromolecules with molar masses below approximately 20.000 g mol-1. Below this threshold molar mass the lamellar thickness and Tm rapidly drop with decreasing molar mass.7 DSC thermogram of original PHBV (first heating scan) showed a broad melting endotherm (Figure 11, Table 5), due to the stereo-irregular ethyl groups of HV comonomer units, which are included in HB lattice and cause a high degree of crystal imperfection.35 At the cooling rate of 200 °C min-1, PHBV with 13 mol % HV almost did not crystallize from the melt, since HV units inhibit the crystallization process35



Table 5. DSC Results of PHBV for Various Annealing Times at Annealing Temperature 180 °C first heating; 10 °C min-1 ∆Hma/Jg-1

Tma/°C PHBV

168.4

126.2

95.04

first cooling; 200 °C min-1 Tc/°C -∆Hc/Jg-1

annealing time at 180 °C/min 2 6 10 300

83.3 78.9 58.6 -

11.28 6.60 0.57 -

second heating; 10 °C min-1

Figure 11. DSC melting traces of original PHBV and annealed PHBV for 2, 6, 10, and 300 min at 180 °C and then crystallized nonisothermally from the melt at the cooling rate 200 °C min-1 (heating rate 10 °C min-1).

2 6 10 300

Tc/°C

-∆Hc/Jg-1

56.9 58.9 60.8 68.8

48.99 51.72 52.89 58.39 second heating; 10 °C min-1

2 6 10 300

Tma/°C

∆Hma/Jg-1

165.9 167.0 166.9 153.6; 127.4

61.68 59.59 58.79 49.05

a Poorly separated peaks. ∆H is given for the whole melting endom therm.

Figure 12. Crystallization exotherms of PHBV during nonisothermal crystallization after annealing at 180 °C for 2, 6, 10, and 300 min (cooling rate 200 °C min-1).

(Figure 12, Table 5). Tc and ∆Hc decreased with increasing annealing time most probably due to additional hindrances to crystallization caused by the increasing amount of free end groups formed during degradation. These results indicate that in the case of PHBV sterical hindrances caused by HV ethyl groups and chain ends of degradation products has a stronger effect on the crystallization rate than the decreasing molar mass. During the second heating scan, the almost amorphous PHBV cold crystallized from the glassy state. Tc and ∆Hc increased with longer annealing time, indicating greater superheating of lower molar mass PHBV. After cold crystallization, PHBV showed a broad melting endotherm, which is shifted toward lower temperatures and had lower ∆Hm compared to the melting endotherm of the original PHBV (Figure 11). With longer annealing time, ∆Hm decreased slightly (Table 5). As in the case of PHB, the decrease in PHBV Tm and ∆Hm was most pronounced at the longest annealing time (300 min). NMR Analysis. Thermally degraded PHB and PHBV after 300 min annealing showed signals of degradation products

in 1H NMR spectra at δ 5.80 ppm, and δ 6.95 ppm for the olefinic protons, and at δ 1.80 ppm for the methyl protons in the chain end units (CH3-CHdCH-), and in 13C NMR spectra at δ 172 ppm for carboxyl groups.7 These results offer additional support for the degradation mechanism involving the β-elimination reaction and chain scission at the ester group. With annealing times up to 10 min, we could not observed differences in the intensities of the mentioned signals due to the limited sensitivity of NMR spectrometry under the experimental conditions used. Conclusions Dissolution of PHB and PHBV in chloroform on a molecular level was achieved by two procedures of sample preparation. In the first procedure, the samples were dissolved in chloroform at 60 °C for at least 60 min. Lengthening the dissolution time up to 120 min did not affect sample MMA and MMD indicating that degradation was negligible during dissolution. The second procedure involved thermal pretreatment (annealing at 180 °C, subsequent quenching) of samples before dissolution. Clear solutions were obtained after annealing time for at least 1 min. MMA and MMD of samples prepared by either procedure were not affected by solution concentration thus indicating complete dissolution. The MMA of thermally pretreated samples were lower than these obtained for samples dissolved at 60 °C in chloroform indicating some thermal degradation. Prolonging the annealing time further reduced MMA. The determined rate constants of thermal degradation at 180 °C were slightly higher for PHB than for PHBV.

636



The samples degraded at 180 °C for up to 10 min showed minor changes in their DSC thermograms, whereas NMR spectroscopy proved to have insuficient sensitivity to monitor intensity changes of the signals originating from degradation products. With a longer annealing time such as 300 min, both PHB and PHBV samples showed substantial changes in their DSC thermograms as well as in their NMR spectra. Acknowledgment. The authors gratefully acknowledge the financial support from the EC 5th RTD (project WHEYPOL) and from the Ministry of Education, Science and Sport of the Republic Slovenia (program P0-514-104). The authors also thank Dr. Sylvio Ortega of PHB Industrial SASerrana (Brasil) for having provided through Prof. Emo Chiellini of the University of Pisa the PHB and PHBV samples. References and Notes (1) Sudesh, K.; Abe, H.; Doi, Y. Prog. Polym. Sci. 2000, 25, 1503. (2) Braunegg, G.; Genser, K.; Bona, R. Macromol. Symp. 1999, 144, 375. (3) Braunegg, G.; Lefebvre, G.; Genser, K. F. J. Biotechnol. 1998, 65, 127. (4) Doi, Y.; Fukuda, K. Biodegradable plastics and polymers; Elsevier: Amsterdam, The Netherlands, 1994. (5) Doi, Y. Microbial polyesters; VCH: New York, 1990. (6) Holmes, P. A. Phys. Technol. 1985, 16, 32. (7) Yu, G.; Marchessault, R. H. Polymer 2000, 41, 1087. (8) Kunioka, M.; Doi, Y. Macromolecules 1990, 23, 1933. (9) Janigova, I.; Lacı´k, I.; Choda´k, I. Polym. Degrad. Stab. 2002, 77, 35. (10) Bloembergen, S.; Holden, D. A.; Bluhm, T. L.; Hamer, G. K.; Marchessault, R. H. Macromolecules 1989, 22, 1663. (11) Kusaka, S.; Iwata, T.; Doi, Y. J. Macromol. Sci., Pure Appl. Chem. 1998, A35(2), 319. (12) Rouxhet, L.; Legras, R. Nucl. Instrum. Methods Phys. Res. 2000, B 171; 487.

(13) Gogolewski, S.; Jovanovic, M.; Perren, S. M.; Dillon, J. G.; Hughes, M. K. Polymer Degrad. Stab. 1993, 40, 313. (14) Majid, M. I. A.; Ismail, J.; Few, L. L.; Tan, C. F. Eur. Polym. J. 2002, 38, 837. (15) Bradel, R.; Kleinke, A.; Reichert, K. H. Makromol. Chem., Rapid Commun. 1991, 12, 583. (16) Herrera de Mola, A.; Marx-Figini, M.; Figini, R. V. Macromol. Chem. 1975, 176, 2655. (17) Akita, S.; Einaga, Y.; Miyaki, Y.; Fujita, H. Macromolecules 1976, 9, 774. (18) Miyaki, Y.; Einaga, Y., Hirosye, T.; Fujita, H. Macromolecules 1977, 10, 1356. (19) Cornibert, J.; Marchessault, R. H.; Benoit, H.; Weill, G. Macromolecules 1970, 3, 741. (20) Huglin, M. B.; Radwan, M. A. Polymer, 1991, 32, 1293. (21) Terada, M.; Marchessault, R. H. Int. J. Biol. Macromol. 1999, 25, 207. (22) Hansen, C. M. J. Paint. Technol. 1967, 39(104), 511. (23) Grassie, N.; Murray, E. J.; Holmes, P. A. Polym. Degrad. Stab. 1984, 6, 95. (24) Morikawa, H.; Marchessault, R. H. Can. J. Chem. 1981, 59, 2306. (25) Lehrle, R.; Williams, R.; French, C.; Hammond, T. Macromolecules 1995, 28, 4408. (26) Corona Software 1.40 (User’s guide); Wyatt Technology Deutchland GmbH, 1996. (27) Kratochvil, P. Classical light scattering from polymer solutions; Elsevier: Amsterdam, 1987; Chapter 6.4. (28) Wyatt, P. J. Anal. Chim. Acta 1993, 272, 1. (29) Doi, Y.; Kunioka, M.; Tamaki, A.; Nakamura, Y.; Soga, K. Makromol. Chem. 1988, 189, 1077. (30) Flory, J. P. J. Am. Chem. Soc. 1936, 58, 1877. (31) Montroll, E. W.; Simha, R. J. Chem. Phys. 1940, 8, 721. (32) Simha, R. J. Appl. Phys. 1941, 12, 569. (33) Basedow, M.; Ebert, K. H.; Ederer, H. E. Macromolecules 1978, 11, 774. (34) Ballauf, M.; Wolf, B. A. Macromolecules 1981, 14, 654. (35) Scandola, M.; Ceccorulli, G.; Pizzoli, M.; Gazzano, M. Macromolecules 1992, 25, 1405.

BM030073L

SEC-MALS Characterization of Microbial Polyhydroxyalkanoates

Recommend Documents