Crystallization Behavior and Morphology of Polylactide and PLA/Clay Nanocomposites in the Presence of Chain Extenders
N. Najafi, M.C. Heuzey, P.J. Carreau Department of Chemical Engineering, Center for Applied Research on Polymers and Composites (CREPEC), Ecole Polytechnique, Montreal, Quebec, Canada
The effect of clay and chain extender on the nonisothermal, isothermal crystallization kinetics, and morphology of polylactide (PLA) was investigated in this study. PLA and PLA-based nanocomposites containing 2 wt% organoclay were prepared via melt compounding. Three commercially available chain extenders were used: polycarbodiimide (PCDI), tris(nonylphenyl) phosphite (TNPP), and Joncryl ADR4368F. The nanoclay particles were found to act as nucleating agents. Chain extender incorporation, however, had diverse effects on both crystallization rate and degree of crystallinity. Nonisothermal DSC results revealed that the addition of PCDI increased the cold-crystallization temperature (Tc) from 106 to 1148C, reduced the degree of crystallinity from 6.3 to 5.3%, and resulted in the formation of bimodal melting peaks in PLA. On the other hand, the reduction of chain ends in the presence of TNPP resulted in a significant increase of the crystallization rate and degree of crystallinity from 6.3 to 15.2%. In the case of Joncryl, its incorporation led to the formation of a long-chain branching structure, which disrupted the chain packing. Therefore, the degree of crystallinity (from 6.3 to 1.6%) and the rate of crystallization decreased, while Tc was increased from 106 to 1228C in the presence of Joncryl. POLYM. ENG. SCI., 53:1053–1064, 2013. ª 2012 Society of Plastics Engineers
INTRODUCTION The environmental impact of petroleum-based polymers and their waste management has motivated their substitution by more environmentally benign products. The most common biodegradable polymers are polylactic acid (PLA), polycaprolactone (PCL), polybutylene adipate terephthalate (PBAT), and polyhydroxy butyrate (PHB) [1]. PLA has attracted the most attention among them, due to its high strength, high modulus, good processability, transparCorrespondence to: M.C. Heuzey; e-mail:
[email protected] Contract grant sponsor: NSERC (Natural Science and Engineering Research Council of Canada). DOI 10.1002/pen.23355 Published online in Wiley Online Library (wileyonlinelibrary.com). C 2012 Society of Plastics Engineers V
POLYMER ENGINEERING AND SCIENCE—-2013
ency after processing, and commercial availability [2, 3]. PLA is a linear aliphatic thermoplastic polyester derived from renewable plant sources such as starch and sugar beet [4]. PLA can be synthesized either by direct condensation polymerization of lactic acid (hence called polylactic acid) [5] or by ring-opening polymerization of cyclic lactide (polylactide) [6], leading to the production of low and high-molecular weights, respectively. Despite all the advantages of PLA, there are nevertheless some drawbacks such as poor gas-barrier properties, low-mechanical resistance, and melt strength [2] that restrict its practical applications. Many attempts have been made to overcome such shortcomings and improve PLA properties. Some of these efforts have been directed toward the incorporation of nanoscale particles including organomodified layered silicate, carbon nanotubes, and TiO2 [7–10]. Although the mechanical and barrier properties of PLA can be improved by the incorporation of organically modified clay, thermal degradation of PLA seems to be intensified with clay loading, leading to a loss of molecular weight [11–13]. However, it has been shown that the use of a chain extender during compounding can compensate for the molecular weight decrease, with an overall positive impact on mechanical and physical properties [12–15]. It is well known that crystalline properties play an essential role in physical, mechanical, and gas-barrier properties. Meanwhile, the crystallization rate and ultimate crystalline morphology are impacted significantly by the thermal history [16]. As a consequence, a considerable attention has been devoted to the fundamental understanding of PLA-crystallization kinetics [3, 7, 10, 16–20]. Crystallization is a process associated with partial alignment of polymer chains, starting from nucleation and followed by subsequent growth. In addition to thermal history, other factors strongly affect the nucleation process. Crystal nucleation is considerably influenced by impurities, dyes, fillers, plasticizers, etc. Once the nuclei are formed, segments of a chain pull out of the amorphous phase, fold together, and sequentially attach to the growth front, forming an ordered structure called lamellae. The resulting lamellar crystals then organize themselves into
FIG. 1. Chemical structure of (a) polycarbodiimide (PCDI) [24], (b) tris(nonylphenyl) phosphate (TNPP) [23], c) Joncryl ADR, where x, y, and z are all between 1 and 20 [25], and (d) the organomodifier of Cloisite 30B, where T is tallow [26].
larger spheroidal entities named spherulites [21]. Detailed investigation in both melt crystallization and cold crystallization of PLA and its stereocomplex have been extensively performed [17, 19, 22]. It has been reported that the degree of crystallization, crystallization growth rate, and the resulting crystalline morphology are highly influenced by chain structure, concentration of the residual monomer, nucleating agent, thermal history, and environmental factors like shear/stretching flow [17, 19, 22]. Besides these factors, the molecular weight is also a key parameter governing the crystallization kinetics. As previously mentioned, the incorporation of organoclay particles into PLA leads to further thermal degradation, decreasing the molecular weight and mechanical properties. To control such degradation in PLA nanocomposites, chain extenders can be used, resulting in a reattachment of cleaved polymer chains [12]. The impact of different chain extenders, polycarbodiimide (PCDI), tris(nonylphenyl) phosphite (TNPP), hexamethylene diisocyanate (HDI), pyromellitic dianhydride (PMDA), and Joncryl 1ADR 4368F on the thermal degradation of PLA and PLA/clay nanocomposites was investigated in our previous works [12, 13]. The effect of chain extender and processing conditions on clay delamination and the resulting rheological, mechanical, and gas-barrier properties were also investigated in our previous studies [9, 12]. Although the impact of PCDI and TNPP on the nonisothermal crystallization of PLA was briefly reported in Refs. [23, 24], to our knowledge, no comprehensive report has been published on the influence of these chain extenders on the nonisothermal and isothermal crystallization behavior of PLA-based nanocomposites. The objective of this work 1054 POLYMER ENGINEERING AND SCIENCE—-2013
is to show how the presence of a chain extender affects the crystallization behavior of PLA and PLA-based nanocomposites using differential scanning calorimetry (DSC) and polarized optical microscopy (POM).
EXPERIMENTAL Materials The PLA investigated in this study, PLA 4032D, is a semicrystalline grade from NatureWorks, USA, having a DLA content of 2 mol%. The organomodified clay is Cloisite1 30B, supplied by Southern Clay Products. In addition, three different chain extenders were used in this work: PCDI, a carboxyl-reactive chain extender, TNPP, and Joncryl1 ADR 4368F. The two former ones were purchased from Sigma Aldrich, Canada, and the latter was supplied by BASF, Germany. All these products were used as received. The molecular formulae of the chain extenders and the organomodifier of Cloisite 30B are presented in Fig. 1. PCDI (Fig. 1a) is a carboxyl reactive chain extender where the carbodiimide (N¼ ¼C¼ ¼N) groups react with the COOH groups of PLA to decrease the number of active sites for degradation [12, 24]. TNPP (Fig. 1b) has phosphate groups that react with the hydroxyl end groups PLA and produce longer PLA chains through transesterification [12, 23]. Joncryl (Fig. 1c) is a modified acrylic copolymer with epoxy functions. The epoxy groups of Joncryl can theoretically react with both hydroxyl and carboxyl groups of the polyester, leading to the formation of branched polymer chains [12, 25]. DOI 10.1002/pen
Plasti-Corder1 internal mixer. Before mixing, PLA and clay were dried at 708C in a vacuum oven for 48 h. The dried PLA was compounded in the molten state with 2 wt% dried clay and chain extender in the internal mixer. The nomenclature used for the nanocomposites is as follows: PLA-2C–PCDI, PLA-2C–TNPP, and PLA-2C–J for the systems containing 2 wt% PCDI, 1 wt% TNPP, or 1 wt% Joncryl as a chain extender and PLA-2C for PLA and clay only. For comparison purposes, PLA without and with chain extender but without clay was also compounded under the same conditions and are, respectively, named neat PLA, PLA–PCDI, PLA–TNPP, and PLA–J. The mixing was conducted under nitrogen atmosphere at a rotation speed of 100 rpm for 11 min, while the temperature was set at 1908C. After mixing, the various systems were immediately immersed in liquid nitrogen to avoid thermo-oxidative degradation during cooling. Thereafter, the processed materials were placed in a vacuum desiccator at ambient temperature for further use. Characterization
FIG. 2. DSC thermograms of the second heating for PLA and PLAbased nanocomposites with and without chain extenders. The heat flow axis has been shifted for clarity. The dotted vertical indicates neat PLA cold-crystallization temperature.
Finally, Cloisite1 30B (Fig. 1d) is an organically modified clay with two hydroxyl groups. The interaction between the C¼ ¼O groups of PLA and the hydroxyl groups of the organomodifier makes it highly compatible with the PLA matrix [26].
Material Processing Melt compounding of PLA with clay and chain extender was carried out in a counter-rotating Brabender DOI 10.1002/pen
The nonisothermal melt-crystallization behavior of the various specimens was investigated using a TA Instruments differential scanning calorimeter (DSC-Q 1000) under nitrogen atmosphere. Samples (10 mg) were encapsulated in aluminum standard pans. To eliminate any initial thermal history, the samples were heated at a scanning rate of 108C/min from 30 to 2508C, held for 2 min and then cooled to 308C at the same rate. The crystallization enthalpy (DHc), melting enthalpy (DHm), and degree of crystallinity were determined from the second heating cycle performed at the same heating rate. To study the kinetics of isothermal cold-crystallization behavior (from glassy state) in the temperature range of 80– 1208C, all the specimens were first heated at 608C/min to 2508C and held there for 5 min. Then the molten samples were cooled at 608C/min to 308C and subsequently reheated at the same rate to the desired crystallization temperature (Tc). The protocol used to study the isothermal crystallization behavior was similar to Refs. [27, 28]. The samples were kept at Tc until the crystallization was complete. To compare the kinetics of isothermal melt with that of cold crystallization, the specimens were heated to 2508C at 608C/min and held there for 5 min. Then the molten samples were cooled at the same rate to the desired crystallization temperature (Tc). The crystallization was monitored using a Nikon Optiphot-2 polarizing microscope to follow the formation and growth of the spherulites. Thin films with a thickness of 100 lm were prepared using compression molding. A film was placed between two glass slides and heated on a programmable hot stage at a heating rate of 508C/min to 1908C, where the sample was kept for 5 min to eliminate the initial thermomechanical history. Subsequently, the sample was cooled at the same rate to a set temperature of 1308C to observe crystallization. POLYMER ENGINEERING AND SCIENCE—-2013 1055
TABLE 1. Cold crystallization and melting temperatures, enthalpies, and degree of crystallinity of neat PLA and PLA-based nanocomposites with and without chain extender.
Neat PLA PLA-PCDI PLA-TNPP PLA-J PLA-2C PLA-2C-PCDI PLA-2C-TNPP PLA-2C-J
Tc (8C)
Tm (8C)
DHc (J/g)
DHm (J/g)
Xc (%)
106 114 106 122 103 107 103 105
167 159 and 167 166 165 166 167 168 165
30.2 31.1 22.3 26.8 29.0 26.5 24.1 21.8
36.1 36.0 36.5 28.3 36.9 34.0 38.0 25.7
6.30 5.30 15.2 1.60 8.70 8.30 15.4 4.30
Data were obtained from the second heating cycle in DSC.
Fourier transform infrared absorption spectra were obtained using a Perkin Elmer FTIR spectrometer. The attenuated total reflectance mode was used to measure the IR absorption in the solution state. Samples (4 g) were dissolved in 10 mL of chloroform. The spectral resolution and scanning speed were adjusted to 4 cm21 and 32 kHz, respectively. The spectra were acquired after subtraction of the chloroform absorption obtained under the same conditions. RESULTS AND DISCUSSION Nonisothermal Analysis PLA is known to crystallize slowly in comparison with other polyesters such as PCL and polyethylene terephthalate (PET) due to the rigid segments in its main chain [19]. The nonisothermal melt-crystallization behavior of PLA and PLA-based nanocomposites with and without chain extender is shown in the DSC thermograms in Fig. 2. The second heating is shown. From these thermograms, the degree of crystallinity (Xc) is determined according to Eq. 1: 1 DH DH m c A 100 Xc ð % Þ ¼ @ 0 fPLA DHm 100
increased the cold-crystallization temperature from 106 to 1148C, while it slightly decreased the extent of crystallinity (from 6.3 to 5.3%). This behavior has also been observed by Ding et al. [30] in a PCDI-poly(p-dioxanone) system. Based on their explanation, the molecular mobility is confined due to the introduction of rigid segments (including phenyl groups) into the PLA, and hence the crystallization ability is declined. To verify this explanation, the FTIR spectra of PLA, PCDI, and PLA treated by PCDI were obtained and presented in Fig. 3. The reactive functional groups of PCDI, carbodiimide (N¼ ¼C¼ ¼N) exhibit a characteristic infrared band at 2130 cm21 in the FTIR spectrum of PCDI alone. However, this peak is absent in the FTIR spectrum of PLA treated by PCDI, indicating that the carbodiimide groups have been consumed by the reaction with the chain extender. The phenyl rings exhibit a characteristic infrared band at 1530 cm21 [31]. The FTIR results indicate the presence of this group in the spectrum of PCDI alone and the PLA–PCDI system, hence confirming the incorporation of PCDI into PLA. For this system, a bimodal melting peak is observed in Fig. 2 at 159 and 1678C. Such a multiple melting behavior has also been observed by other authors for PLA systems [18, 22, 32] and reflects the melting process of crystals having different degree of perfection. Considering that the incorporation of PCDI into PLA further reduces the segmental mobility of polymer chains, a more imperfect crystallites are expected to form in the PLA–PCDI system when compared with neat PLA. Such small and imperfect crystallites tend to melt at a lower temperature in comparison with the more perfect crystals. Based on this explanation, the low and high-melting peaks in the second heating run of the PLA–PCDI system may correspond to the melting points of less and more perfect crystallites, respectively.
0
(1)
where DHm and DHc are the measured melting and crystallization enthalpies, respectively, and fPLA is the PLA weight percent in the sample. For the enthalpy of fusion (DHom ) of a perfectly crystalline PLA, a value of 93.6 J/g has been used [20, 29]. The main features of the DSC thermograms are summarized in Table 1. Crystallization of the neat PLA during the cooling process was insignificant (not shown), whereas it easily crystallized during the heating process (cold-crystallization). A cold-crystallization exotherm followed by a melting peak is observed for the neat PLA at 106 and 1678C, respectively, quite close to the values reported in previous references [3, 17]. The incorporation of PCDI into PLA 1056 POLYMER ENGINEERING AND SCIENCE—-2013
FIG. 3. FTIR spectra of PLA, PCDI, and PLA treated by PCDI (adapted from [12]).
DOI 10.1002/pen
FIG. 4. Isothermal crystallization thermograms of (a) PLA and PLA nanocomposites with and without different chain extenders at 1108C and (b) Joncryl-based nanocomposite at different crystallization temperatures.
It has been shown that the addition of TNPP to PLA increases the thermal stability and, in some cases, its molecular weight [12, 23, 33]. Theoretically, the crystallization rate and degree of crystallization should decrease as the molecular weight increases due to a reduction in chain mobility. In contrast to expectations, a significant increase in the degree of crystallinity (from 6.3 to 15.2%) is reported in Table 1 for the PLA treated by TNPP. The reason for this increase is discussed later. The DSC thermograms of Fig. 2 and the data collected in Table 1 indicate that the addition of Joncryl to PLA decreases the extent of crystallinity from 6.3 to 1.6%, while the crystallization temperature is shifted from 106 to 1238C. In our previous work, it was found that Joncryl led to the formation of a long-chain branched (LCB) structure in Joncryl-treated PLA and that it had a profound effect on molecular weight [12]. The presence of branches disrupts the packing of polymer chains, thus preventing crystallization. The decreased chain mobility caused by the increased molecular weight and LCB, on the other hand, is responsible for the increased cold-crystallization temperature. DOI 10.1002/pen
The presence of clay nanoparticles is found to affect the nucleation and crystal growth rate of PLA nanocomposites [7, 8, 10]. Figure 2 and Table 1 indicate that the crystallization temperature (Tc) of PLA without chain extender was slightly reduced (from 106 to 1038C), and the degree of crystallinity increased from 6.3 to 8.7% after clay loading (PLA-2C). These findings suggest that the nanosized dispersed clay particles may act as nucleating agents in the PLA nanocomposites, leading to enhanced nucleation and facilitating the crystallization process [7, 8, 10]. The nucleating effect of clay particles is more pronounced in the PLA nanocomposite treated with PCDI (PLA-2C–PCDI). The resulting nanocomposite has a higher degree of crystallinity (8.3%) and a lower Tc (1078C) in comparison with the PLA–PCDI system (Xc ¼ 5.3% and Tc ¼ 1148C). However, the lower mobility of polymer chains caused by the presence of PCDI rigid segments still leads to an increase in cold-crystallization temperature from 103 to 1078C when compared with the nanocomposite with no chain extender (PLA-2C). A comparison of the nanocomposite containing TNPP (PLA-2C–TNPP) with the PLA–TNPP system reveals that their degree of crystallization is comparable, although the Tc of the resultant nanocomposite is lower (1038C) than that of TNPP-treated PLA (1068C). In spite of that, its degree of crystallinity (15.4%) is higher than that of PLA (6.3%) and the PLA nanocomposite without chain extender (8.7%). The DSC results of the PLA nanocomposite with Joncryl (PLA-2C–J) are also shown and reported in Fig. 2 and Table 1. On the basis of our last results [9, 12], the effect of Joncryl in the PLA nanocomposite is not as spectacular as for neat PLA due to the degradation of the matrix, highly favored by the presence of the clay. Therefore, polymer chains of a lower molecular weight and less LCB are expected to form in the resulting nanocomposite when compared with the PLA–Joncryl system. Such a decrease in molecular weight accompanied by less LCB, as well as the nucleating effect of the clay particles, is responsible for the decreased cold-crystallization temperature from 122 to 1048C and the increased degree of crystallization of the Joncryl-based nanocomposite from 1.6 to 4.3% compared to the PLA–Joncryl system. However, the LCB structure of the resultant nanocomposite obstructs chain packing, leading to a reduction of its crystallinity (1.6%) in comparison with that of the neat PLA (6.3%) and PLA nanocomposite without chain extender (8.7%). Isothermal Analysis The isothermal crystallization behavior of PLA and PLA-based nanocomposites with and without chain extender was investigated by DSC in a temperature range between 80 and 1208C. The results in the form of thermograms for all systems at Tc ¼ 1108C and for Joncryl-based nanocomposite at different crystallization temperatures are presented in Fig. 4a and b, respectively. The thermograms for the other crystallization conditions showed very similar trends and, hence, are not presented, but all the correspondPOLYMER ENGINEERING AND SCIENCE—-2013 1057
TABLE 2. Summary of isothermal crystallization data of PLA and PLA-based nanocomposites in the presence of different chain extenders, along with parameters for Avrami equation. Sample PLA
PLA-PCDI
PLA-TNPP
PLA-J
PLA-2C
PLA-2C-PCDI
PLA-2C-TNPP
PLA-2C-J
T (8C) 80 90 100 110 120 80 90 100 110 120 80 90 100 110 120 80 90 100 110 120 80 90 100 110 120 80 90 100 110 120 80 90 100 110 120 80 90 100 110 120
t1/2
exp
(min)
DH (J/g)
k (min2n)
n
m
R2
18.9 21.5 27.2 26.1 31.9 17.2 21.3 23.6 27.1 31.8 20.5 24.6 29.0 33.3 30.1 10.1 20.0 22.1 17.3 25.7 20.0 23.1 27.5 31.5 33.3 19.2 21.6 25.5 26.1 32.1 21.1 24.0 29.6 34.6 30.4 14.5 14.6 23.4 27.6 23.8
2.86 3 1027 5.24 3 1024 1.24 3 1022 9.85 3 1022 5.44 3 1022 6.32 3 1029 5.64 3 1026 2.24 3 1023 7.54 3 1022 2.57 3 1022 4.65 3 1027 1.19 3 1022 0.23 1.02 0.72 7.54 3 10210 2.83 3 1024 2.77 3 1023 1.03 3 1022 5.33 3 1023 1.86 3 1027 6.04 3 1023 0.11 4.43 1.14 1.68 3 1027 1.99 3 1023 2.65 3 1022 0.96 0.31 4.99 3 1026 2.43 3 1022 0.22 4.53 0.91 1.34 3 1026 5.44 3 1023 7.19 3 1022 1.30 0.59
4.41 4.25 4.61 3.86 5.27 5.47 6.14 5.75 4.57 5.63 5.10 4.29 5.48 4.45 6.53 5.39 3.26 3.55 4.11 5.15 5.23 3.83 4.71 4.01 5.36 5.21 4.74 5.96 4.37 6.06 4.66 4.38 5.12 4.27 6.04 4.60 4.39 4.60 3.52 5.79
0.41 0.25 0.61 20.14 1.27 1.47 2.14 1.75 0.57 1.63 1.1 0.29 1.48 0.45 2.53 1.39 20.74 20.45 0.11 1.15 1.23 20.17 0.71 0.01 1.36 1.21 0.74 1.96 0.37 2.06 0.66 0.38 1.12 0.27 2.04 0.6 0.39 0.6 20.48 1.79
0.998 0.996 0.992 0.996 0.990 0.997 0.990 0.992 0.994 0.987 0.996 0.998 0.994 0.998 0.961 0.999 0.995 0.996 0.990 0.988 0.998 0.992 0.993 0.989 0.974 0.999 0.995 0.995 0.998 0.978 0.999 0.998 0.997 0.989 0.967 0.998 0.991 0.981 0.993 0.974
28.4 5.40 2.36 1.67 1.61 29.5 6.69 2.70 1.61 1.74 16.1 2.56 1.21 0.93 0.96 46.2 10.8 4.72 2.75 2.55 18.1 3.38 1.47 0.62 0.90 18.5 3.40 1.72 0.92 1.11 12.6 2.13 1.24 0.61 0.93 17.4 2.98 1.60 0.83 1.01
ing enthalpy values are listed in Table 2. Similarly to the nonisothermal crystallization case, the addition of a chain extender and clay particles strongly influenced the crystallization behavior. The addition of organoclay as a filler and TNPP as a chain extender slightly increased the measured values of DH, while the incorporation of PCDI and/or Joncryl into PLA and the PLA-based nanocomposite decreased the enthalpy values. To quantify how the chain extenders affect the rate of crystallization, the relative degree of crystallinity, Xt, versus the crystallization time, t, is considered and plotted in Fig. 5. The relative degree of crystallinity is defined as Eq. 2: Xt ¼
DHðtÞ DHð1Þ
(2)
where DH(t) is the enthalpy of isothermal crystallization at time t and DH(!) is the enthalpy of complete crys1058 POLYMER ENGINEERING AND SCIENCE—-2013
t½
cal
(min)
28.1 5.42 2.37 1.66 1.62 29.5 6.71 2.70 1.62 1.77 16.2 2.58 1.22 0.92 0.99 46.1 10.9 4.73 2.78 2.56 18.1 3.40 1.48 0.63 0.91 18.5 3.43 1.73 0.93 1.14 12.7 2.15 1.24 0.64 0.96 17.5 3.01 1.63 0.83 1.02
tallization, both calculated from Eq. 3 and the data from Fig. 4:
DHðtÞ ¼
Z1
dH :dt dt
0
DHð1Þ
Z1 dH ¼ :dt dt
(3)
0
To further quantitatively describe the evolution of the crystallinity during isothermal crystallization, we used the well-known Avrami model, which is the most common and easiest approach to obtain relevant parameters characterizing the crystallization kinetics [34, 35]. The relative crystallinity can be calculated from Eq. 4: XðtÞ ¼ 1 expðktn Þ
(4) DOI 10.1002/pen
tors such as nucleation type, nucleation density, crystal growth dimension, and restriction of crystalline formation due to surrounding fillers [17]. The index value of 1, 2, and 3 indicates one, two, and three-dimensional spherulites growth, respectively. The fractional Avrami index value or value greater than 4 can theoretically be explained by the introduction of the nucleation index, m (Eq. 6), describing the nucleation mechanism throughout the crystallization process [37]: m ¼ n n0 1
(6)
where n0 is the dimensionality index (e.g., n0 ¼ 1 for rods, n0 ¼ 2 for disks, and n0 ¼ 3 for spheres). m values between 1 and 0 represent instantaneous and sporadic nucleation gradually decreasing with time and approaching a constant value at a certain time, while a value of 0 indicates sporadic nucleation steadily increasing with time. An m value from 0 to 1 and greater than 1 indicates a sporadic nucleation, which increases and markedly
FIG. 5. Relative crystallinity (Eq. 4) of (a) PLA and PLA nanocomposites with and without different chain extenders at 1108C and (b) Joncryl-based nanocomposite at different crystallization temperatures.
where X(t) is the relative degree of crystallinity (from Eq. 2), n is the Avrami index, and k is the overall crystallization rate constant including nucleation and crystal growth contributions. The Avrami constants k and n can be calculated by fitting the experimental data to Eq. 5, obtained after taking the double logarithm of Eq. 4: lnð lnð1 XðtÞÞÞ ¼ ln k þ n lnðtÞ
(5)
Because the Avrami model rarely describes the whole conversion range [36], the relative crystallinity data between 5 and 80% were used to calculate k and n. The plots of ln [2ln (1 2 X (t)] as a function of ln (t) for the data obtained at Tc ¼ 1108C for the PLA and PLA-based nanocomposites with and without chain extenders are presented in Fig. 6. The behavior at the four other crystallization temperatures was similar, and, for conciseness, the results are not presented here. However, the calculated k, n, and the correlation coefficient of the fit (R2) are all summarized in Table 2. Most correlation coefficient values are larger than 99%, indicating good fits of the experimental data in Fig. 5. The Avrami index, n, is a constant with a typically integer value between 1 and 4, depending on different facDOI 10.1002/pen
FIG. 6. Avrami fits of the isothermal crystallization data of (a) PLA and PLA nanocomposites with and without different chain extenders at 1108C and (b) Joncryl-based nanocomposite at different crystallization temperatures.
POLYMER ENGINEERING AND SCIENCE—-2013 1059
increases with time, respectively [37]. A large variety of n values have been reported for PLA in the literature [17, 38]. The data summarized in Table 2 reveal that the obtained Avrami exponent ranges from 3 to 6 in these samples, suggesting a three-dimensional spherulite growth (n0 ¼ 3). The calculated m values, reported in Table 2, are mainly greater than zero but smaller than 1 indicating that the nucleation process is most likely governed by sporadic nucleation, which increases with time. The k values of Table 2 as well as the data of Figs. 4 and 5 show that the addition of PCDI to the neat PLA leads to a reduction in the crystallization rate. For example, at the crystallization temperature of 1008C, the k value varies from 1.3 3 1022 min2n for the neat PLA to 2.2 3 1023 min2n for the PLA treated with PCDI. The temperature of 1008C is chosen in the comparison of the k values; however, the trend is the same at all crystallization temperatures. The incorporation of rigid PCDI fragments into the PLA confines the molecular mobility, as shown by the FTIR results, and subsequently reduces the rate of crystallization. It is interesting to note that the crystallization of the PLA and PLA–PCDI system almost terminates at the same time in Fig. 4, even though the PLA crystallization begins sooner in comparison with that in the PLA–PCDI system. This behavior indicates that the crystallization rate of the PLA is gradually decreased at later stage of crystallization when compared with that of the PLA–PCDI system (see Fig. 5). The faster growth of PLA crystallites at early stage of the crystallization may lead to crystal impingements. Such impingements, therefore, retard the crystallization kinetics at later stage of the crystallization. This explanation is in good agreement with what is observed by optical microscopy as discussed later. In contrast with PCDI, the incorporation of TNPP to PLA results in a significant increase of the PLA crystallization kinetics, even though the molecular weight is also increased [12]. Based on the explanation of Carvalho et al. [39], decreasing the number of chain ends in the system as a consequence of short chains removal is responsible for the increased crystallization rate. Also, Richter and coworkers [40] explained that chain ends are much more mobile than mid-chain segments. The high mobility of the chain ends decreases the likelihood of their attaching to the growth front in comparison with a stem from a mid-chain region. Consequently, the chain ends are more resistant to being folded into a compact crystalline form, acting as an entropic defect at the crystal growth front, and subsequently decrease the crystal growth rate. As reported in our previous work [12], the addition of TNPP decreases the number of chain ends per mass in PLA, resulting in a decrease of the defects at the growth front and promoted crystallization kinetics when compared with neat PLA (the k value changes from 1.3 3 1022 to 0.23 min2n). The decreased number of chain ends per mass resulting from the TNPP incorporation is also responsible for the 1060 POLYMER ENGINEERING AND SCIENCE—-2013
increased degree of crystallinity in PLA and PLA-based nanocomposites containing TNPP during nonisothermal crystallization, as shown in Table 1. On the contrary, the addition of Joncryl to PLA significantly reduces the crystallization kinetics (k value changes from 1.3 3 1022 to 2.77 3 1023 min2n) as evidenced in Table 2 and Figs. 4 and 5. The epoxy groups, present in Joncryl (Fig. 1), react with hydroxyl and carboxyl groups of the polyester and form a LCB structure [12]. The LCB structure and branch points hinder the chain-folding phenomenon, which is required for the incorporation of the chains into growing crystalline lamellae. As reported by other authors [7, 8, 10], the presence of clay particles increases the crystallization kinetics of PLA. Generally, two major factors, nucleation and mobility of chain segments, control the crystallization behavior. As shown in Fig. 4 and Table 2, the inclusion of clay particles in neat PLA and PLA with chain extender decreases the crystallization halftime from 2.36 to 1.47 min at Tc ¼ 1008C. This is due to heterogeneous nucleation that results in narrower crystallization peaks, indicating that the rate of crystallization is increased (k ¼ 0.11 min2n) when compared with that of neat PLA (k ¼ 1.24 3 1022 min2n). In the case of the PLA-based nanocomposite-containing PCDI and Joncryl, the nucleating effect of clay particles promotes the crystallization kinetics on one hand. On the other hand, less chain mobility in PCDIenriched system and the LCB structure in the Joncryl nanocomposite results in the retardation of the crystallization process; the k value changes from 0.11 to 2.6 3 1022 and 7.2 3 1022 min2n in nanocomposite-containing PCDI and Joncryl, respectively. The comparison of the crystallization kinetics between the nanocomposite-containing PCDI and those treated by Joncryl is also interesting. Joncryl-based nanocomposite shows a larger crystallization rate than the nanocomposite-containing PCDI at the early stage of the crystallization, while a reverse trend is observed for the nanocomposite-containing PCDI as the relative crystallinity increases, since the crystallization is finished at a shorter time (see Fig. 5). Based on what is reported in Refs. [41, 42], the crystallization rate is first controlled by nucleation and then crystal growth and packing. We assume that the formation of a branched structure in the nanocomposite-containing Joncryl leads to an enhancement of the density of nuclei, increasing the crystallization rate at the early stage of crystallization. However, suppression or disruption of chain packing within the crystallites, caused by a LCB structure, then retards the crystal growth at later stages. Comparing the k parameter of the various samples shows that its value first increases with increasing crystallization temperature, followed by a decrease above 1108C. Such a behavior is expected in polymer due to a balance between two opposing effects. At low-crystallization temperatures, close to the glass transition temperature (Tg 608C), the decreased chain mobility significantly retards the crystallization rate, whereas at high-crystallizaDOI 10.1002/pen
TABLE 3. Summary of melt-isothermal crystallization data of PLA and PLA-based nanocomposites in the presence of TNPP as a chain extender, along with parameters for Avrami equation. Sample PLA PLA–TNPP PLA-2C–TNPP
T (8C)
DH (J/g)
90 110 90 110 90 110
22.3 23.2 23.1 26.2 23.3 27.6
t1/2
exp
k (min2n)
(min)
6.18 4.89 5.54 4.48 5.24 4.05
6.12 2.30 1.90 3.05 2.60 4.58
tion temperatures, close to equilibrium melting temperature (Tm0), a considerable decrease in nucleation density hinders the crystallization growth, although the chain mobility is high. Moreover, the comparison of different systems at a given crystallization temperature shows that the incorporation of PCDI and Joncryl into PLA decreases the k value, especially at low-crystallization temperatures, close to Tg, where the segmental mobility is the dominant factor controlling the crystallization rate. However, the TNPP and clay addition result in an increased k value, in complete agreement with what is explained earlier. To further verify whether these systems closely follow the Avrami model, the obtained k and n are used to calculate the crystallization halftime (t1/2) using Eq. 7. ln 2 1 t1 =n = =2cal k
(7)
The calculated t1/2 is also reported in Table 2. The comparison between the experimental and calculated t1/2 shows that the Avrami model can precisely predict t1/2, confirming that the crystallization kinetics of all the systems is correctly described by this model for the relative degree of crystallinity between 5 and 80%. To compare the overall crystallization rates from the melt with glassy states, the isothermal crystallization of the neat molten PLA and some nanocomposite samples was also studied at 90 and 1108C. All the corresponding data are listed in Table 3, although the thermograms are not presented for the sake of brevity. The results indicate that the overall crystallization rate in the molten state is much slower than that from the glassy state (i.e., k value of TNPP-enriched PLA decreased from 1.19 3 1022 and 1.02 to 1.9 3 1023 and 3.05 3 1023 min2n at 90 and 1108C, respectively). Similar results have been observed in polypropylene by Supaphol [43]. Based on the Lauritzen and Hoffman’s theory [44], the crystal growth rate is expected to be only a function of the crystallization temperature, Tc. Supaphol [43] suggested that faster crystallization from the glassy state resulted from a higher contribution of nucleation mechanisms (either as an enhancement of nucleation rate or density). Indeed, the total number of activated nuclei that can serve as predetermined homogeneous nuclei during isothermal crystallization at Tc increased upon the DOI 10.1002/pen
3 3 3 3 3 3
1024 1023 1023 1023 1023 1023
n
m
R2
3.85 3.58 3.45 3.62 3.36 3.58
20.15 20.42 20.55 20.38 20.64 20.42
0.998 0.998 0.999 0.999 0.999 0.998
t1/2
cal
(min)
6.20 4.92 5.53 4.48 5.27 4.06
quenching process and led to an increase in the overall crystallization rate. This argument can also be confirmed by the calculated nucleation index, m, presented in Tables 2 and 3. Contrary to crystallization from the glassy state (Table 2), m has a value less than zero in melt isothermal crystallization (Table 3). This indicates that the instantaneous and sporadic nucleation gradually decrease with time and approach a constant value in melt isothermal crystallization, while such nucleation rates markedly increase with time [37] in cold isothermal crystallization. The crystal growth rate is usually defined as the inverse of the crystallization halftime (G ¼ 1/t1/2), and the variation of Gexp is plotted as a function of crystallization temperature in Fig. 7. As expected, Gexp is strongly dependent on the crystallization temperature. Gexp increases first with increasing crystallization temperature up to 1108C, passes through a maximum, and then is reduced. The trend is consistent with the crystallization kinetics (k values). A comparison of the neat PLA with the PLA–PCDI and PLA–Joncryl systems reveals that the restriction of chain movement in the PCDI and Joncrylenriched systems reduces Gexp when compared with that of the neat PLA. However, the reduction of defects in polymer chains promotes Gexp as TNPP is added to the
FIG. 7. Crystallization rate (G) of the PLA and PLA-based nanocomposites in the presence of different chain extenders as a function of crystallization temperature.
POLYMER ENGINEERING AND SCIENCE—-2013 1061
Polarized Microscopy
FIG. 8. Optical polarizing micrographs of (a) neat PLA, (b) PLAPCDI, (c) PLA-TNPP, (d) PLA-J, (e) PLA-2C, (f) PLA-2C-PCDI, (g) PLA-2C-TNPP, and (h) PLA-2C-J isothermally crystallized at 1308C for column 1: 10 min and column 2:15 min. Column 3 shows spherulite impingement. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
neat PLA. The effect of clay loading on the crystallization kinetics is evident in Fig. 7, especially at high-crystallization temperature where nucleation is the dominant factor controlling the crystallization kinetics. The formation of additional nucleation sites by the clay particles is responsible for such enhancement of Gexp. 1062 POLYMER ENGINEERING AND SCIENCE—-2013
POM was used to directly observe the effect of clay and chain extenders on the crystal morphology of PLA. Figure 8 displays the crystal morphology of PLA and PLA nanocomposites with and without the presence of a chain extender. The micrographs are arranged in three columns. Columns 1 and 2 represent, respectively, the polarized optical micrographs of samples isothermally crystallized at 1308C for 10 and 15 min, while column 3 presents an approximate spherulite density and the occurrence of spherulite impingement. The temperature of 1308C was chosen since the nucleation process is slower at that temperature, and it allows the observation of the spherulite grow. Figure 8a shows that the spherulites are formed for the neat PLA after 10 min, and their size systematically increases with time until the occurrence of spherulite impingement. The impingement starts as adjacent growing spherulites meet each other at a point, preventing spherulite growth at the contact region. Figure 8b reveals that the spherulite growth rate is clearly decreased as PCDI is added to PLA, leading to the postponement of the spherulite impingement. This is an expected observation, because the introduced rigid segments as observed from the FTIR results diminish the chain mobility and subsequently retard the chain-folding process. On the contrary, the incorporation of TNPP into PLA promotes the spherulite growth (Fig. 8c) based on the reasons explained earlier in relation to the results shown in Figs. 5 and 6. The crystal morphology of Joncryl-treated PLA, presented in Fig. 8d, on the other hand, shows the formation of dense finer crystals in comparison with the neat PLA. This finding is similar to what was observed in [13] and indicates that the branching units obstruct packing of the polymer chains and thus cause the cessation of the spherulite growth. The compounding of organoclay with PLA with and without chain extender slightly enlarges the spherulite size, as illustrated in Figs. 8e–h, and leads to the further occurrence of spherulite impingement. Rangasamy et al. [45] also observed a similar trend in PP–nanoclay system. They believed that the increase of nucleation speed in the nanocomposite promotes the crystallization rate, which may change the crystal size. The comparison of the crystal size of Joncryl-based nanocomposite (Fig. 8h) with Joncryl-treated PLA (Fig. 8d) reveals that the spherulite sizes in the nanocomposite appear to be larger than that observed in the Joncryl–PLA system. As mentioned earlier, a lower molecular weight and less long-chain branching (LCB), which favor the folding of polymer chains, are responsible for the increased crystal size.
CONCLUSION In this study, the nonisothermal and isothermal crystallization behaviors of PLA and PLA–organoclay nanocomDOI 10.1002/pen
posites in the presence of three different chain extenders (polycarbodiimide, PCDI, tris (nonylphenyl) phosphite, TNPP, and Joncryl) were investigated using differential scanning calorimetry (DSC) and polarized optical microscopy (POM). Both isothermal and nonisothermal crystallization studies suggested that organoclay particles act as nucleating agents and yield a faster overall rate of crystallization. The addition of PCDI increased the cold-crystallization temperature and decreased the degree of crystallinity. It was assumed that the lower mobility of polymer chains, reacted with this chain extender, reduced spherulite size and the crystallization growth rate, resulting in an increased crystallization halftime for the isothermal case. In contrast, TNPP was found to significantly intensify the crystallization process by reducing the number of chain ends, acting as defects to incorporate into the crystals. The degree of crystallinity, crystal growth rate, and subsequently spherulite size increased as TNPP was added. The Joncryl incorporation, however, confined the crystallization process due to the formation of a LCB structure, resulting in the disruption of chain packing. Therefore, the degree of crystallinity and the rate of crystallization decreased, while the cold-crystallization temperature was dramatically increased in nonisothermal crystallization. The isothermal crystallization growth rate of all the compounds was determined in the temperature range of 80–1208C. It first increased to attain a maximum at 1108C and was then followed by a reduction. The Avrami analysis was used to quantify the effects of clay and chain extenders on the isothermal crystallization behavior. It was found that all systems satisfactorily follow the Avrami equation and more likely exhibit a 3D spherulitic growth. A comparison of the overall crystallization rate parameters calculated from both cold and melt isothermal crystallization processes revealed that the rate of crystallization from the glassy state proceeded faster than that from the melt state. This indeed suggests that the quenching process increased the total number of activated nuclei that can act as predetermined homogeneous nuclei upon subsequent crystallization at Tc. REFERENCES 1. A. Demirbas, Energy Sour.: Part A, 29, 419 (2007). 2. S. Gumus, G. Ozkoc, and A. Aytac, J. Appl. Polym. Sci., 123, 2837 (2012). 3. G.Z. Papageorgiou, D.S. Achilias, S. Nanaki, T. Beslikas, and D. Bikiaris, Thermochim. Acta, 511, 129 (2010). 4. J. Lunt, Polym. Degrad. Stab., 59, 145 (1998). 5. F. Achmad, K. Yamane, S. Quan, and T. Kokugan, Chem. Eng. J., 151, 342 (2009). 6. R. Mehta, V. Kumar, H. Bhunia, and S.N. Upadhyay, J. Macromol. Sci.: Polym. Rev., 45, 325 (2005).
DOI 10.1002/pen
7. Y. Di, S. Iannace, E. Dimsio, and N. Luigi, J. Polym. Sci.: Polym. Phys., 43, 689 (2005). 8. J.H. Lee, T.G. Park, H.S. Park, D.S. Lee, Y.K. Lee, S.C. Yoon, and J.D. Nam, Biomaterials, 24, 2773 (2003). 9. N. Najafi, M.C. Heuzey, and P.J. Carreau, Comp. Sci. Tech., 72, 608 (2012). 10. M. Pluta, J.K. Jeszka, and G. Boiteux, Eur. Polym. J., 43, 2819 (2007). 11. K. Fukushima, D. Tabuani, and G. Camino, Mater. Sci. Eng. C, 29, 1433 (2009). 12. N. Najafi, M.C. Heuzey, P.J. Carreau, and P.M. WoodAdams, Polym. Degrad. Stab., 97, 554 (2012). 13. Q. Meng, M.C. Heuzey, and P.J. Carreau, Polym. Degrad. Stab., 97, 2010 (2012). 14. H. Li and M.A. Huneault, J. Appl. Polym. Sci., 122, 134 (2011). 15. J. Liu, L. Lou, W. Yua, R. Liao, and R. Li, Polymer, 51, 5186 (2010). 16. H. Xiao, W. Lu, and J.T. Yeh, J. Appl. Polym. Sci., 113, 112 (2009). 17. M. Day, A.V. Nawaby, and X. Liao, J. Therm. Anal. Calorim., 86, 623 (2006). 18. X. Ling and J.E. Spruiell, J. Polym. Sci.: Polym. Phys., 44, 3200 (2006). 19. T. Miyata and T. Masuko, Polymer, 39, 5515 (1998). 20. L. Wang and C.M. Dong, J. Polym. Sci.: Polym. Chem., 44, 2226 (2006). 21. L. Mandelkern, Crystallization of Polymers: Kinetics and Mechanisms, Cambridge University Press, Cambridge, United Kingdom (2004). 22. H. Abe, M. Harigaya, Y. Kikkawa, T. Tsuge, and Y. Doi, Biomacromolecules, 6, 457 (2005). 23. J.A. Cicero, J.R. Dorgan, S.F. Dec, and D.M. Knauss, Polym. Degrad. Stab., 78, 95 (2002). 24. L.X. Yang, X.S. Chen, and X.B. Jing, Polym. Degrad. Stab., 93, 1923 (2008). 25. M. Villalobos, A. Awojulu, T. Greeley, G. Turco, and G. Deeter, Energy, 31, 3227 (2006). 26. V. Krikorian and D.J. Pochan, Chem. Mater., 15, 4317 (2003). 27. J. Yu and Z. Qiu., Ind. Eng. Chem. Res., 50, 12579 (2011). 28. V. Ojijo, T. Malwela, S.S. Ray, and R. Sadiku, Polymer, 53, 505 (2012). 29. J.W. Huang, Y.C. Hung, Y.L. Wen, C.C. Kang, and M.Y. Yeh, J. Appl. Polym. Sci., 12, 3149 (2009). 30. S.D. Ding, Z.P. Liu, and T. Yang, J. Polym. Res., 17, 63 (2010). 31. M. Rybachuk, A. Hertwig, M. Weise, M. Sahre, M. Mann, and J. Bell, Appl. Phys. Lett., 96, 909 (2010). 32. M. Pluta, J.K. Jeszka, and G. Boiteux, Eur. Polym. J., 43, 2819 (2007). 33. J.R. Dorgan, J. Janzen, and M.P. Clayton, J. Rheol., 49, 607 (2005). 34. M. Avrami, J. Chem. Phys., 9, 177 (1941). 35. M.C. Tobin, J. Polym. Sci.: Polym. Phys., 14, 2253 (1976) 36. A.T. Lorenzo, M.L. Arnal, J. Albuerne, and A.J. Muller, Polym. Test., 26, 222 (2007).
POLYMER ENGINEERING AND SCIENCE—-2013 1063
37. P. Supaphol, Thermochim. Acta, 370, 37 (2001). 38. H. Tsuji, H. Takai, and S.K. Saha, Polymer, 47, 3826 (2006). 39. J.L. Carvalho, S.L. Cormier, N. Lin, and K.D. Veress, Macromolecules, 45, 1688 (2012). 40. M. Zamponi, M. Monkenbusch, L. Willner, A. Wischnewski, B. Farago, and D. Richter, Europhys. Lett., 72, 1039 (2005).
1064 POLYMER ENGINEERING AND SCIENCE—-2013
41. S.H. Tabatabaei, P.J. Carreau, and A. Ajji, Chem. Eng. Sci., 64, 4719 (2009). 42. J. Tian, W. Yu, and C. Zhou, J. Appl. Polym. Sci., 104, 3592 (2007). 43. P. Supaphol and J.E. Spruiell, Polymer, 42, 699 (2001) 44. P. Supaphol, J. Appl. Polym. Sci., 79, 1603 (2001). 45. L. Rangasamy, E. Shim, and B. Pourdeyhimi, J. Appl. Polym. Sci., 121, 410 (2011).
DOI 10.1002/pen