Journal of Thermal Analysis and Calorimetry, Vol.
95 (2009) 3, 957–964
THERMAL PROPERTIES OF POLYLACTIDES
Effect of molecular mass and nature of lactide isomer
J. Ahmed, J.-X. Zhang, Z. Song and S. K. Varshney*
Polymer Source, Inc. 124 Avro street, Dorval, Montreal, Quebec, H9P 2X8, Canada
A thermal analysis of a series of polylactides (PLA) was carried out based on the number of average molecular mass (Mn), and the
nature of isomer (D, L and DL). It is confirmed that the glass transition temperature (Tg) of PLA increased as a function of molecular
mass irrespective of isomer type except sample with a high polydispersity index. The melting temperature (Tm) and enthalpy of
crystal fusion (DHf) of L-isomer increased as the Mn was increased from 1100 to 27500. The degree of crystallinity (cc%) increased
as a function of molecular mass. However no crystallization peak was detected in the lower molecular mass range (550–1400). The
non-isothermal crystallization behavior of the PLA melt was significantly influenced by the cooling rate. Both D and L isomers ex-
hibited insignificant difference in thermal properties and DL lactides exhibited amorphous behavior at identical molecular masses.
Change in microstructure showed significant difference between two isomers. Analysis of the FTIR spectra of these PLA samples in
the range of 1200–1230 cm–1 supported DSC observation on crystallinity.
Keywords: crystallinity, glass transition temperature, melting temperature, polylactides
Introduction D-lactide and meso-lactide. An equimolar mixture of D
and L isomer results in DL-lactide. Among the various
Lactic acid based polymers are known as polylactides polylactides, optically pure polylactides, poly(L-lactide)
(PLA). The interests in polylactides have increased (PLLA) and poly(D-lactide) (PDLA), are crystalline
significantly due to their biodegradability and polymers with a melting point around 180°C [5, 6]. The
biocompatibility [1, 2]. Polylactide is used in various atactic polymer, poly-(DL-lactide) is an amorphous ma-
medical applications including as carriers for con- terial with a glass transition temperature of 50–57°C
trolled release drug delivery systems, and in biode- [7, 8]. It is well established that the properties of
gradable pins, screws, and cell scaffolds in tissue en- polylactides vary to a large extent depending on the ra-
gineering. Recently, polyanhydrides based on PLA tio and the distribution of the two isomers.
have been accepted for implantations. These are in- In recent years there has been renewed interest in
cluded in architecturally fabricated fully bio-absorb- the crystallization of lactides. Nature derived lactides
able coronary stents for treating vascular disease, thus (agricultural source) are mostly in L-form and exhibit
replacing conventional metallic stents [3]. These bio- crystalline behavior. The crystallization of behavior
absorbable stents are deployed as present stent. Fur- of polylactides depends on the thermal history (i.e.
thermore, they utilize an anti-inflammatory, therapeu- annealing) [9], amount and type of additives [10] and
tic drug-delivery polymer permitting delivery of stereo sequence distribution [11]. In addition, crystal-
multiple drugs with tunable release kinetics. The stent lization depends upon optical purity. It is reported
acts like a scaffold for tissue growth during the initial that at least 72% threshold optical purity in composi-
six months. Once the artery has healed, the stent is ab- tion is required to obtain crystallinity of polylactides
sorbed by the body, leaving no permanent implant. [11, 12]. Semi-crystalline nature of PLA has been ob-
The advantage of polylactide based biopolymers over served during cooling of melt [13]. The heat of crys-
conventional metallic or non-biodegradable polymers tallization measured by differential scanning calorim-
is to ease of removal by the body system itself and re- etry may depend on the sterosequence distribution in
tention of shape during time. Attention is also paid to the polymer. It is also important to know the percent
PLA due to solid waste accumulation [4] and thus, of crystallinity since it strongly influences the physi-
there are significant efforts to produce newer biode- cal properties of the polymer and composite [14, 15].
gradable packaging materials based on PLA. Understanding of thermodynamic properties of
Because of the chirality of the lactyl unit, lactide lactides requires quantitative thermal analysis and
exists in three diasteroisomeric forms: L-lactide, reliable interpretation of the underlying molecular
1388–6150/$20.00 Akadémiai Kiadó, Budapest, Hungary
© 2008 Akadémiai Kiadó, Budapest Springer, Dordrecht, The Netherlands
AHMED et al.
mobility [16]. Interestingly, polylactides exhibit both Experimental
amorphous and semi-crystallinity and therefore a wide
variation in thermal properties are expected for poly- Materials
lactides. Numerous publications are available on ther- Lactide monomer was obtained from Purac Co. USA
mal properties and crystallinity of polylactides, but most and used as received. Both poly L(–), D(+) and racemic
of those studies have carried out with limited number of (DL±)-lactides were synthesized by tin-catalyzed
samples. An attempt is made to present a broad spec- ring-opening bulk polymerization. Diethylene glycol
trum of thermal characteristics using a series of poly- monomethyl ether, ethylene glycol and diethylene gly-
lactides (L, D and LD) synthesized in our laboratory. col were used as initiators. The obtained polymers were
The main objectives of the present study is to purified by dissolving in acetone and further precipi-
investigate thermal properties (melting, crystallization tated in ethanol/hexane 2:8 v/v mixture. The precipita-
and glass transition) of polylactides using calorimetric tion repeated thrice to remove un-reacted lactide mono-
technique as function of molecular mass, isomer, and mer. FTIR was used to ensure the complete removal of
microstructure. In addition, the effect of cooling rate unreacted lactide monomer from the polymer by mea-
on melt crystallization behavior during non-isothermal suring a characteristic absorbance peak at 1250 cm–1 as-
cooling of L-lactides has also been investigated. signed for the cyclic monomer structure. Finally sam-
ples were oven-dried at 38°C under vacuum for 48 h.
Table 1a Characteristics of PLA (L-form) samples: heating rate (10°C min–1)
Mn Mw/ Tg / Tm / DH / Tc / DH / cc/
Product ID
·103 Mn °C °C J g–1 °C J g–1 %
P7163-(PLLA1)
0.55 1.30 –3.81 – – – – –
P7158-(PLLA2)
1.1 1.08 21.96 88.02 23.44 – – –
P7150-(PLLA3)
1.4 1.12 30.75 112.27 41.02 – – –
P3937-(PLLA4)
4.7 1.09 45.60 157.82 55.50 98.25 47.82 8.26
P3929-(PLLA5)
7.2 1.10 51.79 148.99 58.85 100.85 49.88 11.27
P3938-(PLLA6)
17.5 1.15 62.54 173.98 60.54 105.71 44.82 16.90
P6465-(PLLA7)
19 1.95 64.20 172.12 51.76 106.44 42.75 9.69
P3934-(PLLA8)
27.5 1.12 66.83 179.09 70.59 109.34 42.35 30.37
P2290-(PLLA9)
40.4 1.35 62.71 180.57 50.06 101.63 10.01 43.06
P6467-(PLLA10 )
150 1.70 62.69 179.78 42.08 98.38 03.38 56.67
958 J. Therm. Anal. Cal., 95, 2009
THERMAL PROPERTIES OF POLYLACTIDES
Table 1b Characteristics of PLA (DL-form) samples
Product ID Mn·103 Mw/Mn Tg/°C
P8860-(PDLLA1)
0.95 1.15 –6.67
P8861-(PDLLA2)
3.0 1.12 26.24
P8862-(PDLLA3)
4.9 1.08 42.11
P8863-(PDLLA4)
8.5 1.05 46.12
P6464-(PDLLA5)
250 1.80 50.86
Characterization of materials PDLLA1) are presented in Fig. 1. The number-aver-
Both number average (Mn) and the mass average (Mw) age molecular mass and polydispersity index (Mw/Mn)
molecular masses were determined by size exclusion of PLA samples (L and DL) used in this study are
chromatography (SEC) on a Varian instrument given in Tables 1a and b, respectively.
equipped with refractive index, UV and dual detec-
tors from Viscotek Co. (light scattering, and viscosity Measurement of properties
detectors). All three columns were procured from
Supelco (G6000-4000, 2000-HXL) and used for the Thermal analysis was carried out with a TA Q100
SEC at 35°C where THF was used as the eluent con- Differential Scanning Calorimeter (DSC) (TA Instru-
taining 4.0 triethyl amine (V/V%). The light scattering ments) which was calibrated against an indium stan-
detectors were calibrated with polylactide standards. dard. An empty aluminum pan was the reference.
The molecular masses of the obtained polymers were The samples (6–8 mg) were run at a 10°C min–1 heat-
further verified by their 1H NMR spectra by compar- ing/cooling ramp in two heating-cooling cycles in a
ing the integration values from the initiator moiety nitrogen atmosphere (flow rate 50 mL min–1). In the
and from lactide units. The values were found close in first cycle, PLA samples were heated from –20 to
agreement with the SEC results. Two representatives 200°C and isothermed for 5 min; cooled back to
1
H NMR spectra for lactide samples (PLLA2 and –20°C and isothermed for 2 min. Same steps were fol-
Fig. 1a 1H NMR spectra of polylactide samples PLLA2 Fig. 1b 1H NMR spectra of polylactide samples PDLLA1
J. Therm. Anal. Cal., 95, 2009 959
AHMED et al.
lowed in the second cycle. The melting temperature tion for PLA [17]. PLLA samples exhibit crystallinity,
(Tm) was taken at the end of the melting peak. The however, no exothermic peak of crystallization was
crystallization temperature (Tc) was considered as the found for low molecular mass (550 to 1400) samples
minimum of the exothermic peak. The Tm was deter- (Fig. 2). Low molecular mass polymers (short chains)
mined from the first heating cycle while crystalliza- are generally weaker in strength. Although they are
tion and glass transition temperature were calculated crystalline, only weak Van der Waals forces hold the
by the second heating-cooling cycle. The area under lattice together. This allows the crystalline layers to slip
the curve was calculated as the enthalpy from the in- past one another causing a break in the material and
strument software. The non-isothermal crystallization causes the absence of distinct crystalline peak for low
studies were carried out in the cooling rate ranged molecular mass PLA samples.
from 2.5 to 20°C min–1. For PLA samples containing PLA samples with high Mn (>40000) signifi-
racemic mixture (DL) were annealed at 200°C for cantly affected the crystallization behavior and
5 min and cooling–heating cycle was followed to lo- showed lower crystallization temperature (Tc) and
cate the glass transition temperature. Thermal scans enthalpy associated with the crystallization (DHc).
for each sample were carried out in duplicate and the The crystallization peaks attributed by both PLLA9
average values are reported in this paper. and PLLA10 were relatively lower (10°C min–1). The
Fourier transform infrared spectra were obtained crystallization peak was untraceable in the second
with a FTIR spectrophotometer (Nicolet-Avatar 330 scan even at excessive high cooling rate (40°C min–1).
model) using He–Ne laser by the transmission The latter indicates that the molecular mobility of
method. Polylactide samples were cast onto KBr crys- high molecular mass polymers became restricted dur-
tal from solutions in chloroform, and the solvent was
evaporated prior to measurement. No evidence of re-
sidual solvent was observed spectroscopically. Spec-
tra were recorded in the range 400–4000 cm–1. The in-
strument settings were kept constant (16 scans; reso-
lution, 4 cm–1; velocity, 0.6329; aperture, 100).
Results and discussion
Thermal behaviors of polylactides L and DL isomers
are presented in Table 1. The glass transition tempera-
ture of semi-crystalline L-isomer increased from –3.8
to 66.8°C as number-average molecular mass was in-
creased from 550 to 27500 and leveled thereafter (Ta-
ble 1a). A lowering of glass transition temperature
was pronounced for high molecular mass samples
(PLLA9 and 10) which is probably affected by the
high polydispersity index and the plasticizing effect
contributed by low molecular mass polymers. The Tg
of DL form increased systematically from –6.7 to
about 50.9°C with an increase of Mn from about 1000
to 250000. The glass transition temperature of L form
showed slightly higher values than DL form. This
trend is very common for polymeric materials that
bear crystallinity.
It is observed that each PLLA sample exhibits an
endothermic melting peak during the thermal scan ex-
cept the low molecular mass sample (PLLA1). The
melting temperature (Tm) and enthalpy of crystal fusion
(DHm) of PLLA samples increased with Mn from 88 to
181°C and from 23.4 to 70.6 J g–1, respectively. Samples
with high polydispersity index exhibited lowering of
DHm though the melting temperature leveled (Table 1a). Fig. 2 Effect of number average molecular mass (Mn) on heat-
The melting temperature range and enthalpy obtained in ing-cooling curves of PLLA2 and PLLA6; A – Mn of
this study are close in agreement with earlier observa- 1100 and B – Mn of 25000
960 J. Therm. Anal. Cal., 95, 2009
THERMAL PROPERTIES OF POLYLACTIDES
ing high cooling rate and could not achieve the de- Table 2 Effect of heating rate on non-isothermal crystalliza-
sired crystallinity during cooling of the melt. tion of PLA Sample ID # P3938 (PLLA7 and
Mn=17500)
The degree of crystallinity (cc) of PLA samples
can be calculated based on the enthalpy value of a Heating rate/°C min–1 Tc/°C DHc/J g–1 cc/%
100% crystalline PLA sample [18] from the following 2.5 123.2 55.8 6.7
equation:
5 115.3 50.4 9.4
( DH m - DH c )
%c c = ×100 10 109.0 46.6 11.7
93 20 101.8 23.9 32.9
where DHm is the melting enthalpy, DHc is the crystal-
lization enthalpy and 93 J g–1 the melting enthalpy of
totally crystallized PLA sample. This calculation
leads to degree of crystallinity ranging from 8.3 to
56.7%. The degree of crystallinity values is close in
agreement with the earlier reported values for PLA
samples [17].
The thermal behavior of the polylactides de-
pends on the thermal history of the samples and rate
of heating. Figure 3 shows the typical heating-cooling
pattern for PLLA sample at two selected heating rates
(10 and 20°C min–1). The melting peak is very sharp
at a higher heating rate as compared to lower one. At a
slower heating rate (10°C min–1), the melting of
PLLA was noticed at about 160°C with relatively low
enthalpy of crystal fusion (DHm =58.5 J g–1) and the
corresponding values at higher heating rate were
163°C and 59.4 J g–1, respectively. While the sample Fig. 4 Effect of cooling rate on crystallization of PLLA6
was cooled from the melt at two different cooling
rates, a higher crystallization exothermic peak at found to decrease significantly. It has been observed
108.3°C was observed for a cooling rate of 10 over that the crystallization peak temperature increased
20°C min–1 (102.4°C). The corresponding crystalliza- and shifted towards a lower value as cooling rate was
tion enthalpies equaled about 48.2 and 45.6 J g–1, re- increased. This observation is very common for the
spectively. crystallization of polymeric materials and the phe-
Table 2 and Fig. 4 demonstrate that the non-iso- nomenon can be attributed to the kinetic effect in a
thermal melt-crystallization of polylactides is a func- nucleation-controlled region [19]. At a lower cooling
tion of cooling rate. As the cooling rate gradually in- rate, there is sufficient time for the nuclei with a
creased from 2.5 to 20°C min–1, the crystallization larger size that could stabilize at a higher temperature.
temperature (Tc) and associated enthalpy (DHc) were On the other hand, at a fast cooling rate the polymer
molecules experienced rapid decrease in temperature
and produce fine nuclei that could stabilize at consid-
erably lower temperature. A 2nd order polynomial
equation can be used to predict the crystallization
temperature-cooling rate behavior of PLA samples.
2
æ dT ö æ dT ö
Tc =007
. ç ÷ -2.75ç ÷+128.75 (R =0.99)
2
è dt ø è dt ø
The degree of crystallinity (%cc) was increased
linearly from 6.7 to 33% as a function of cooling rate
and the observation is supported by earlier reported
values [17].
A comparison of thermal properties among D
isomer, L isomer and racemic mixture (DL) of PLA is
presented in Table 3. Both D and L isomers exhibited
semi-crystalline behavior with possession of both
Fig. 3 Effect of heating rate on melting and crystallization of melting and glass transition temperature. It is evident
PLLA11
J. Therm. Anal. Cal., 95, 2009 961
AHMED et al.
Table 3 Effect of isomer on thermal properties of PLA
Mn Mw / Tg/ Tm/ DH/ Tc/ DH/
Product ID Isomer
·103 Mn °C °C J g–1 °C J g–1
P3937-(PLLA4)
4.7 1.09 L 45.60 157.82 55.50 98.25 47.82
P7052-(PDLLA6)
4.3 1.90 DL 44.70 – – – –
P3936-(PLLA11)
7.0 1.09 L 67.94 159.87 58.78 108.28 48.31
P7379-(PDLLA7)
7.3 1.16 DL 44.14 – – – –
P3922-(PDLA1)
13.8 1.19 D 65.69 170.26 67.01 107.57 52.38
P3936-(PLLA12)
14.0 1.12 L 66.83 173.26 61.07 110.27 48.12
P3923-(PDLA2)
16.5 1.20 D 69.10 173.49 64.55 109 51.60
P2293-(PLLA13)
16.8 1.32 L 58.64 173.42 61.38 105 38.14
Table 4 Effect of end group on thermal properties of PLA
Mn Mw / R Tg/ Tm/ DH/ Tc/ DH/
Product ID
·103 Mn group* °C °C J g–1 °C J g–1
P7150- (PLLA3)
1.4 1.12 2 30.75 112.27 41.02 – –
P8030-(PLLA14)
1.4 1.13 1 37.66 136.67 44.20 100.15 41.93
P2283- (PLLA15)
24.6 1.33 1 68.31 178.48 56.61 136.68 55.80
P3938-(PLLA6)
25.0 1.15 2 62.54 179.09 70.59 109.34 42.35
*1: R=CH3–CH–(CH3)–O– 2: R=CH3–O–CH2–CH2–O–
962 J. Therm. Anal. Cal., 95, 2009
THERMAL PROPERTIES OF POLYLACTIDES
from the table that there is no significant difference in tions are duly supported by earlier FTIR work on PLA
thermal properties of PLA samples with almost equal samples [21].
molecular mass although their polydispersity indexes
are different. It infers that thermal properties of PLA
are dictated by lactyl unit not by the isomer. DL sam- Conclusions
ples exhibited amorphous behavior with distinct glass
transition temperature. Thermal analysis of a series of polylactides indicated
Effect of microstructure on the thermal behavior that the microstructure, number average molecular
of L-polylactide is presented in Table 4. A relatively mass and isomer type have considerable effects on
lower Tg and Tm values were observed for PLLA3 glass transition temperature, melting and crystalliza-
compared to PPLA14 although both had similar mo- tion behavior. Semi-crystalline polylactides showed
lecular masses. Additionally, PLLA3 failed to dem- higher Tg over amorphous sample at similar Mn range
onstrate the crystallinity and no distinct crystalliza- and the Tg leveled at higher Mn. Lower molecular
tion exotherm was found for the sample. The crystal- mass semi-crystalline samples did not exhibit melting
lization behavior of high molecular mass PLLA sam- and crystalline peak during thermal scan whereas, sig-
ples (PLLA15 and PLLA6) differed significantly nificant drop in crystallization temperature and crys-
where as both showed similar Tg and Tm. This finding tallization enthalpy were observed for high molecular
confirmed that the microstructure played a significant mass polylactides. Cooling rate significantly affected
role in crystallization process and crystallinity of the degree of crystallinity from the melt and the crys-
polylactides. talline peak was untraceable at higher cooling rate. A
Figure 5 shows the FTIR spectra (from 2nd order polynomial equation has been proposed to
1300–1150 cm–1 in wavelength) of selected PLA sam- predict the crystallization temperature-cooling rate
ples. The spectrum of both D and L isomer of PLA behavior of PLA samples. DSC observation of
(PLLA13 and PDLA2) are almost identical. Bands at crystallinity was supported by FTIR spectra between
1269 and 1215–1186 cm–1 are known as crystalline 1186 and 1269 cm–1.
bands in PLLA structures. These consist of vibra-
tional components i.e., C–COO stretching, O–CH
stretching, CH3 rocking, and CH bending [20]. Both Acknowledgements
D and L samples exhibited a shoulder between 1240
and 1200 cm–1 that clearly distinguished from the DL The authors wish to acknowledge Ms. Neelima Agarwal for
sample (PDLLA8) where the shoulder was not pres- locating the financial support from Polymer Source under the
on-going research and development program for the synthesis
ent. A small shift of the crystalline band might be af-
of new materials. Terry Varshney is appreciated for editing
fected by preparation of sample, molecular mass and the manuscript.
other factors. The absence of crystalline band in low
molecular mass sample between 1215 and 1186 cm–1
justifies the absence of crystallinity. These observa- References
1 L. J. Fetters, D. J. Lohse, D. Richter, T.A. Witten and
A. Zirkel, Macromol., 27 (1994) 4639.
2 A. Tullo, Chem. Eng. News, 78 (2000) 13.
3 S. K. Varshney, O. Hnojewyj, J. Zhang and P. Rivelli,
US Patent Application. 0225472, 2007.
4 R. G. Sinclair, J. Macromol. Sci., Part A: Pure Appl.
Chem., 33 (1996) 585.
5 J. Kline and H. H. Kline, Die Makromol. Chem.,
30 (1959) 23.
6 V. C. Schulz and J. J. Schwaab, Die Makromol. Chem.,
87 (1965) 90.
7 K, Jamshidi, S. H. Hyon and Y. Ikada, Polymer,
29 (1998) 2229.
8 M. Varma-Nair, R. Pan and B. J. Wunderlich, J. Polymer
Sci. Part B: Polym. Phys., 29 (1991) 1107.
9 H. Tsuji and Y. Ikada, Polymer, 36 (1995) 2709.
10 S. Brochu, R. E. Prud’homme, I. Barakat and R. JérÛme,
Macromol., 28 (1995) 5230.
Fig. 5 FTIR spectra for selected PLA samples
J. Therm. Anal. Cal., 95, 2009 963
AHMED et al.
11 K. A. M. Thakur, R. T. Kean, E. S. Hall, J. J. Kolstad, 18 E. W. Fischer and H. J. Sterzel Wegner, Z. Kolloid.
T. A. Lindgren, M. A. Doscotch, J. I. Siepmann and Polym., 251 (1973) 980.
E. J. Munson. Macromol., 29 (1996) 8844. 19 P. Supaphol, P. Thanomkiat, J. Junkasem and
12 H. Tsuji, F. Horii, M. Nakagawa, Y. Ikada, H. Odani and R. Dangtangee, Polymer Test., 26 (2007) 20.
R. Kitamaru. Macromol., 25 (1992) 4114. 20 S. H. Kang, S. L. Hsu, H. D Stidham, P. B. Smith,
13 B. Kalb and A. J. Penings, Polymer, 21 (1980) 607. M. A. Leugers and X. Yang, Macromol, 34 (2001) 4542.
14 G. Perego, G. D. Cella and C. Bastioli, J. Appl. Polym. 21 H. Urayama, S. I. Moon and Y. Kimura., Macromol.
Sci., 59 (1996) 37. Mater. Eng., 288 (2003) 137.
15 M. Day, A. V. Nawaby and X. Liao, J. Therm. Anal. Cal.,
86 (2006) 623.
16 M. Pyda, The Nature of Biological Systems as Revealed Received: January 25, 2008
by Thermal Methods, D. Lõrinczy, Ed., Kluwer Academic Accepted: July 10, 2008
Publisher, Amsterdam 2004. OnlineFirst: November 12, 2008
17 J. R. Sarasua, R. E. Prud’homme, M. Wisniewski,
A. Le Borgne and N. Spassky, Macromol.,
DOI: 10.1007/s10973-008-9035-x
31 (1998) 3895.
964 J. Therm. Anal. Cal., 95, 2009
