Preparation and characterization of high molecular weight poly(L-lactic acid) by chain extending

Abstract Poly(L-lactic acid) (PLLA) of high molecular weight was prepared by chain extending reaction in a micro-compounder. Phosphites were used as chain extenders to increase the molecular weight of the PLLA prepolymer that was prepared by bulk polycondensation of L-lactic acid. Chemistry Magazine Online shows the effects of the amount of the phosphite, temperature and the screw speed were studied on the torque of PLLA melt in mixing. There’s also another thing that this magazine provides, something you can do in your free time. We truly recommend you to visit our friends on BestFreeSlots, if you would like to try some new and entertaining thing to do when relaxing. It’s definitely worth your visit. Under the optimal conditions, the molecular weight of the PLLA could increase from 62,100 g/mol to 126,000 g/mol by the triphenyl phosphite. The chemical structure were characterized by FTIR and 1H NMR and crytallinity of PLLA were characterized by DSC and XRD.
Keywords poly(L-lactic acid); polycondensation; chain extending; phosphite

1 INTRODUCTION
Poly(L-lactic acid) (PLLA) is a bio-based and biodegradable polyester which is increasingly used in the medical and industrial field. In most application, PLLA of high molecular weight (>100,000 g/mol) is demanded. The high molecular weight PLLA is generally prepared by ring opening polymerization of the lactide, while the lactide is produced by oligomer of the lactic acid [1]. Another way to yield PLLA of high molecular weight was solution polycondensation of L-lactic acid [2]. It is difficult to achieve a high molecular weight PLLA by direct bulk polycondensation of L-lactic acid due to the unfavorable reaction equilibrium [3].
Moon et al [4] examined the melt polycondensation of L-lactic acid catalyzed by tin chloride dihydrate/p-toluene sulfonic acid binary system and obtained PLLA of the high molecular weight in 20 h with 67% yield. Chen et al [5] prepared the high molecular weight PLLA by using titanium (IV) butoxide as catalyst. The chain extenders such as ethylene carbonate, heterocyclic compounds, and diisocyanates were used to increase the molecular weight of the PLLA [6]. However, these chain extenders can change chain structure of the PLLA. The triphenyl phosphite as chain extender has been used to increase the viscosity of PET and PBT, while the chain structures of PET and PBT can be maintained [7-8]. Reactive extrusion process is a very attractive approach to prepare the high molecular weight PLLA. Jacobsen studied the polymerization of the lactide in the twin-screw extruder and obtained PLLA of the molecular weight of 100,000 in 7 min [9].
In this paper, the phosphites were used in the chain extending of pre-PLLA through reactive extrusion in a twin-screw micro-compounder, and PLLA of the high molecular weight was obtained.

2 EXPERIMENTAL
2.1 Materials 

L-lactic acid (LLA), as an 88 wt % aqueous solution, was purchased from Fengyuan Corp; tin (II) chloride dihydrate (SnCl2·2H2O) and tin (II) octoate (Sn(Ot)2) were purchased from Beijing Chemical Corp; p-toluenesulfonic acid (TSA), triphenyl phosphite (TPPi), tri-nonylatedphenyl phosphite (TNPi) and tis-(2,4-di-tert bulyl phenyl) phosphite (TBPi) were purchased from Shanghai Chemical Corp. All these materials were used without further purification. The other chemicals were reagent grade.
2.2 Preparation of PLLA
PLLA-prepolymer (PLLA0) was prepared through bulk-polycondensation of L-lactic acid, which properties are listed in Table 1. The prepolymer dried at 60 °C for 2h was mixed with certain amounts of the phosphite in a mortar, and then fed into the twin-screw micro-compounder (Haake MiniLab) at the temperature of 160 °C. The feeding time was in 5 min and the residence time was controlled in a range of 0-30 min. The mixing process in the micro-compounder was protected by the nitrogen atmosphere. Finally, the products were extruded from the die and moulded by a mini-injection moulding machine connected to the micro-compounder.

Table 1 Characterizations of PLLA0

Mη(g/mol) Mw(g/mol) Mw/Mn color acid value
86,700 62,100 1.54 white power 1.5

2.3 Characterizations
The average molecular weight and polydispersity of PLLA were determined by GPC (Waters 1515 HPLC pump with Waters 2424 RI Detector), calibrated with polystyrene standards, and THF was used as eluent.
1H NMR spectra were obtained with a Bruker DMX-400 NMR spectrometer with CDCl3 as a solvent and TMS as an internal standard.
FTIR analysis was carried out with a Nicolet 5DX Spectrometer. The potassium bromide (KBr) pellet method was used for FTIR study.
Differential scanning calorimetry (DSC) analysis was carried out by TA Instruments Q100 from -20 oC to 200 oC at a heating rate 10 oC/min under the nitrogen atmosphere.
X-ray diffraction (XRD) experiments were carried out with a Rigaku D/MAX-RA X-ray diffractometer using Cu Ka radiation. The scattering angle (2q) is varied from 5o to 30o.

3 RESULTS AND DISCUSSIONS
The twin screw micro-compounder was taken as a reactor to continue the polycondensation of PLLA0. The residence time of the PLLA melt in the compounder can be controlled through its circumfluence.
As shown in Figure 1, for the pure PLLAthe melt torque (M) in the circumfluence decreases with the residue time. It means that the PLLA is prone to degrade in the compounding process at melting state. The similar relationship between the melt torque and molecular weight was observed by Jacques [10].

Figure 1 Torque of PLLA melts in compounding process with addition of phosphites at 160 ℃

3.1 Effect of phosphite
Three kinds of the phosphites (1.0 wt %) were added to PLLA0 to inhibit the degradation of PLLA during the compounding process. In Figure 1, TPPi made the melt torque increase quickly and reach a maximum at the residue time of 20 min; TNPi made the torque rise slightly with the time; TBPi kept the torque unchanged. These means that the addition of TPPi obviously increases the viscosity of the PLLA melt, that is the molecular weight of PLLA, while TNPi does it to certain extent. Because the degradation of PLLA is inevitable, the increase of the melt torque means that the chain extending reactions occurs between the PLLA due to addition of the phosphites. In the phosphites, TPPi is the most prominent in the chain extending. For the TPPi, when the residue time is over 20 min in Figure 1, the melt torque decreases, which is attributed to the complete consumption of the TPPi and the dominant degradation of PLLA.
3.2 Effect of temperature 
Figure 2 gives a comparison for the melt torques with the residue time at 160oC and 180oC. It is noted that the torques are lower at 180 oC than at 160 oC. It seems that high temperature is not helpful to accelerate the chain extending of PLLA by TPPi, but result in more serious degradation.

Figure 2 Torques of PLLA melts in compounding process at 160oC and 180oC

3.3 Screw speed 
Three kinds of the screw speeds of 30, 60 and 90 rpm were brought to bear on the screw in the micro-compounder. The curves of the melt torques to residence time are shown in Figure 3. It is seen that at the speed of 90 rpm, the melt torque was high in the beginning, but it reduced quickly after 15 min. It implies that the high screw speed lead PLLA to degrade due to strong shearing.

Figure 3 Torque of PLLA melts with 1.0 wt % TPPi added at 160oC at the different screw speeds

3.4 Amount of TPPi
The relationship between the amount of the TPPi and the melt torque are shown in Figure 4. Addition of the TPPi of 1.0, 2.0 and 5.0 wt % induced a large increase of the melt torque in mixing compared with the pure PLLA. Nevertheless, with the TPPi of 5.0 wt %, the torque was lower than with the 2.0 wt %. It means that the large amount of the TPPi is not advantageous to the chain extending reaction of PLLA.

Figure 4 Torque of PLLA melts with different amounts of TPPi added at 160oC

The molecular weights of the PLLA at the maximum torque are shown in Table 2. It is noted that the Mw (weight average molecular weight) of the PLLA increased, while the polydispersity broadened with addition of the TPPi. When the TPPi is at 2.0 wt %, the largest Mw was achieved, which is consistent with the highest torque in Figure 4.
The chain extending reaction of the PLLA by the TPPi may take place on the end hydroxyl groups, the calculated mole ratios of TPPi/OH groups in the reactive system are listed in Table 2. When the molar ratio of TPPi/OH is very high, the TPPi may react with the end carboxyl groups, while the reaction inhibits the PLLA chain to extend. That is why the PLLA with 5.0 wt % TPPi has a lower Mw than with 2.0 wt % TPPi.

Table 2 Molecular weights of PLLA by GPC

PLLA Amount of TPPi (wt %) TPPi /OH mole ratioa GPC
Mw (g/mol) Mw/Mn
PLLA0 1.0 0.96 93,000 1.60
PLLA0 2.0 1.92 126,000 1.69
PLLA0 5.0 4.80 102,000 1.72

a: Number of end OH groups in the PLLA0 were determined by titrimetric methods with a supposition that number of the end OH groups is equal to that of the end COOH groups.

3.5 Structure characterization of PLLA
Structure characterizations were achieved for the PLLA0 and the PLLA-TPPi (2.0 wt %). Figure 5 shows the FTIR spectra of the two PLLA samples. They have completely similar infra-red absorbance peaks except the appearance of a small peak at 1600 cm-1 (stretching vibration of the phenyl group) and small differences in the relative intensities for the several peaks.
Figure 6 is the 1H NMR spectra of the PLLA0 and PLLA-TPPi. For the PLLA0, only two peaks are observed at the chemical shifts 1.5 and 5.2 ppm (the peak at 7.3 ppm for the solvent). In the spectra of the PLLA-TPPi, the small new peaks appear, the peaks at the 1.5 and 5.2 ppm broaden. For the PLLA-TPPi, the peak at the 4.3 ppm for the methine in the end groups and the broadened peaks at 1.5 and 5.2 ppm indicate that there exist PLLA of small molecular weight in the product. The peaks at the 6.8 and 6.9 ppm are corresponding to the phenyl proton induced by TPPi.

Figure 5 FTIR spectra of PLLAand PLLA-TPPi

Figure 6 1H NMR spectra of PLL A0 and PLLA-TPPi

3.6 DSC and XRD analysis
The thermal behaviors of the PLLA0 and PLLA-TPPi were studied by DSC. Figure 7 shows the heating DSC curves for the two samples. In the curves, there are the three thermal processes: the glass transition with Tg, the crystallization with Tc and the crystal melting with Tm. The Tof the PLLA-TPPi (2.0 wt % TPPi) is at 49.4 oC which is lower than that (57.1 oC) of PLLA0.

Figure 7 DSC curves of the PLLA-TPPi and PLLA0.

Figure 8 WAXD curves of the PLLA-TPPi and PLLA0

For the PLLA0, a cold crystallization process caused a mild exothermic peak at about 125 oC; for the PLLA-TPPi, the exothermic peak sharply appears at 100 oC. These indicate that the crystallization for the PLLA-TPPi is faster and earlier than that for the PLLA0. The PLLA-TPPi and PLLA0 show a similar melting process at 149oC which is corresponding to the a-crystal, but the PLLA-TPPi has another small melting peak at 135 oC which may be corresponding to the b-crystal..
Figure 8 shows the X-ray diffraction patterns of the PLLA-TPPi and PLLA0. For the PLLA0 the sharpest peak is at about 16.7° (2q, 020 reflection), the other peaks are at 14.8° (101 reflection) and 19.1° (023 reflection), respectively; for the PLLA-TPPi the similar diffraction peaks appear at the same 2q angles as for the PLLA0.

4 CONCLUSIONS
PLLA of high molecular weight (Mw=126,000 g/mol) can be prepared through the chain extending reactions by the phosphites in the twin screw micro-compounder. In the three kinds of the phosphites (TPPi, TNPi and TBPi), the TPPi is the best chain extender for the PLLA.
Existence of the TPPi can make polycondensation of the PLLA continue and inhibit degradation of the PLLA to certain extent. The chain extending mechanism of the phosphite for the PLLA is considered to be esterification promotion.

Acknowlegements
 This work was supported by the Science and Technology Developmemt Project of Baoding (No.09ZG016).

REFERENCES
[1] Jacobsen S.; Fritza H.G.; Dege’e Ph.; Dubowas, Ph. Jerome R. Polymer 2000, 41: 3395-3401.
[2] Ajioka M.; Enomoto K.; Suruki K.; Yamaguchi A. Bull. Chem. Soc. Jpn. 1995, 68: 2125-2232.
[3] Sandor K.; Ildiko B.; Jeno B.; Gyo1rgy D.; Miklo’s Z. J. Phys. Chem. 2001, B105: 2833-2838.
[4] Moon S. I.; Lee C. W.; Masatoshi M. J. Polym. Sci., Polym. Chem. Ed. 2000, 38: 1673-1679.
[5] Chen G. X.; Kim H. S.; Kim E. S.; Yoon J. S. European Polymer Journal 2006, 42: 468-471.
[6] Woo S. I.; Kim B. Polymer Bulletin 1995, 35: 415-419.
[7] Tuominen J. and Seppala J. V. Macromolecules 2000, 33: 3530-3537.
[8] Jacques B.; Devauxt J.; Legras R. Polymer 1996, 37: 4085-4090.
[9] Jacobsen S.; Fritza H.G.; Dege’eb Ph. et al. Polymer 2000,41: 3395-3401.
[10] Jacques B.; Devauxt J.; Legras R. Polymer 1996, 37: 4085-4091.