Introduction

As a kind of engineering plastic, the cost-effective poly(ethylene terephthalate) (PET) is widely used in many fields such as automobile industry, mechanical equipment, electrical and electronic appliance, textile and packaging [1,2,3,4,5,6]. However, PET has a relatively low crystallization rate compared to other polymers such as PE, which restricts its wider applications. Comparatively, the maximum growth rate of the spherulites for PE is 5000 µm min−1, but it is only 10 µm min−1 for PET [7]. So, how to effectively increase the crystallization rate and get the desired morphology is a key problem to improve the related properties of PET. Different nucleation agents, nanoparticles and fibers all play an important role in improving the crystallization rate of PET [8,9,10].

The organic/inorganic hybrids have been extensively investigated [11,12,13], which can effectively improve the properties of polymer such as crystallization, mechanical and thermal. Among the nanocomposites prepared, polymer/clay nanocomposites have been more widely investigated, which exhibits significant improvements in mechanical crystallization and gas barrier properties compared to conventionally filled polymer at very low concentration of inorganic filler. These nanocomposites can be prepared by various methods, including melt blending, in situ polymerization and solution intercalation using different polymer such as nylon [14,15,16], polypropylene [17, 18], polylactide [19], poly(ethylene oxide) [20] and polystyrene [21]. Melt compounding is a simple and efficient method to obtain such nanocomposite. The used clay is usually modified by the cationic exchange modifiers (e.g., quaternary ammonium salt) with a better compatibility with organic polymers [22]. But the thermal stability of alkyl ammonium surfactants is, generally, relative low and would cause their decomposition in melt blending preparation of engineering polymer nanocomposites due to the high processing temperatures, e.g., above 250 °C. The decomposition of the modifier can even initiate polymer degradation and cause other disadvantageous effects [23, 24]. To solve the above problem, one way is to improve the thermal stability of organic modifiers, in which there are some defects such as high cost and complex procedure limiting its wide application [25,26,27,28,29]. The other way is the direct utilization of the natural pristine clay without any organic treatment [30, 31], and this can avoid the decomposition of the conventional organic modifier. However, the intercalation of macromolecular chains into the interlayers of pristine sodium montmorillonite (Na+-MMT), generally, is difficult, usually leading to the formation of microcomposites instead of nanocomposites [32]. There have been a few studies on the polymer/pristine sodium clay nanocomposites. Korbee et al. [33] reported a novel compounding process using Na+-MMT water slurry for preparation of Nylon-6/Na+-MMT nanocomposites. Hasegawa et al. [34] further developed this method, in which the Na+-MMT slurry and PA-6 pellets are directly injected in the extruder melt. The main problem in the above method is the use of so much water to reduce the slurry viscosity, producing a large amount of waste water.

A novel solid-state shear milling (S3M) technology was developed in our laboratory. This technology is based on the pan-type milling equipment designed in our laboratory, which shows excellent pulverization, dispersion, mixing and mechanochemical effects on polymers and fillers [35,36,37]. In our previous work, we successfully prepared the PET/Na+-MMT nanocomposites through S3M method mentioned above [38], where the used clay (Na+-MMT) particles are pristine and have no organic modification treatment. The obtained PET/Na+-MMT nanocomposites show the obviously enhanced mechanical properties. Since the performance of a polymer nanocomposite is directly connected with its crystallization and morphology to a certain degree, in this study our attention was focused on investigating the isothermal crystallization and spherulite morphology of PET/Na+-MMT nanocomposite prepared through S3M method.

Experimental

Materials

Na+-montmorillonite (MMT) (Cloisite® Na+) was supplied by Southern Clay Products Inc with a cation exchange capacity (CEC) of 92.6 meq (100 g)−1. Commercial-grade PET as pellets (intrinsic viscosity = 0.875 and molecular mass (M w) = 26,800) was provided by Yizheng Chemical Fibre Co. Ltd (Yizheng, China). The PET and Na+-MMT were vacuum-dried at 120 and 80 °C prior to melt extrusion, respectively.

Samples preparation

The mixture of PET and pristine Na+-MMT with PET/MMT mass ratio of 50/50 was milled in a pan-mill equipment [38] to produce the PET/Na+-MMT co-powder (PETMP50). The PETMP50 master batch was diluted with pure PET to obtain PET/Na+-MMT nanocomposite samples with Na+-MMT loading of 1.5 wt% (PETSN1.5), 3.5 wt% (PETSN3.5) and 4.5 wt% (PETSN4.5) in a twin-screw extruder at 270 °C. The cooled extrudate was then cut into pellets. The dried pellets were finally injection-molded into test samples. As a comparison, the mixture of PET and pristine Na+-MMT with 3.5 wt% loading was directly melt-compounded in the same twin-screw extruder at 270 °C (this method is called the conventional one). The obtained dried pellets were also injection-molded into test samples (PETCN3.5) for comparison.

Characterization

The crystallization and melting curves of PET/Na+-MMT nanocomposite samples were recorded on a TA DSC-200 differential scanning calorimeter (DSC) (TA Instruments, USA). The samples with a mass of about 8 mg were first heated in an aluminum crucible to 290 °C (far lower than the initial decomposition temperature about 400 °C of PET) at a rate of 100 °C min−1 and kept there at 290 °C for 10 min to eliminate the thermal history. Then, the samples were cooled to the designated crystallization temperatures (T c) rapidly (100 °C min−1) to carry out the isothermal crystallization measurement. After the isothermal crystallization finished, the samples were heated again to 290 °C at a rate of 10 °C min−1. The polarized optical microscopy (PLM) observation was done on a X5Z-H polarized optical microscope made in Chongqing Optical Instrument Factory, China. The used samples were prepared by sandwiching a tiny pellet of PET/Na+-MMT samples between two glass plates, compressed at 290 °C for 15 min and then slowly cooled to the ambient temperature at a designated cooling rate.

Results and discussion

Morphology analysis

Figure 1 compares the SEM morphologies of the conventionally prepared PET/Na+-MMT (3.5 wt%) composite (unmilled) (PETCN3.5) and the pan-milled PET/Na+-MMT (3.5 wt%) nanocomposite (PETSN3.5). In order to observe the formation of nanostructures of the layered Na+-MMT (intercalation/exfoliation) in PET matrix under pan-milling conditions, the TEM result is also provided in Fig. 1. As can be seen, the fractured surface of the conventional PETCN3.5 sample is smooth and there are heavily aggregated natural sodium montmorillonite particles distributed in PET matrix. The above results show that the dispersion of Na+-MMT fillers is poor and the interfacial compatibility is not good. Comparatively, for the pan-milled PETSN3.5 sample, there are no agglomerated Na+-MMT particles found in PET matrix. Under the very strong shear force field of pan-milling, the size of the Na+-MMT particles is reduced so much that the filler particles cannot be seen in the current magnification (×1000) in PET matrix. The clay platelet particles are all almost encapsulated by the matrix, and the interfacial compatibility is greatly improved (the fractured surface is very rough). This means that the dispersion of the clay filler particles is very homogeneous in PETSN3.5 nanocomposite sample. The TEM image of PETSN3.5 sample (Fig. 1c) shows that the Na+-MMT clay particles are successfully intercalated and exfoliated and there are nano intercalation/exfoliation structures formed in PET matrix, indicating the formation of nanocomposite in PETSN3.5 sample.

Fig. 1
figure 1

Morphology of the fracture surface of different tensile samples along perpendicular to tensile direction: a SEM image of PETCN3.5, b SEM image of PETSN3.5 and c TEM image of PETSN3.5

Full size image

Isothermal crystallization and melting

During the PET melting process, there are three melting endotherms [39,40,41], including endotherms I, II and III, appearing. Generally, endotherm I is the microcrystallite formed in the boundary layer between the larger crystallites and is related to the crystals produced during the secondary crystallization. Endotherm II is related to the crystals produced during primary crystallization process, and endotherm III, however, is related to the crystals formed from the result of recrystallization when heating [39,40,41]. The pan-milled PET/Na+-MMT nanocomposite with different clay loading was firstly annealed at different isothermal crystallization temperature. Then, the differential scanning calorimeter was used to investigate the melting behavior of the annealed samples. Some interesting results are obtained.

The DSC heating curves of four samples with different isothermal crystallization temperature are shown in Fig. 2, and the related DSC parameters are included in Table 1, where Fig. 2a–d shows the DSC melting curves of pure PET, PETSN1.5, PETSN3.5 and PETCN3.5, respectively. Figure 2 and Table 1 show that pure PET has three melting peaks, namely endotherms I, II and III, which also similarly appeared in many other literatures [40,41,42]. With increasing the isothermal crystallization temperature (T c), both endotherms I and II shift to a higher temperature except for endotherm III. In addition, the area of peak I is much smaller than that of the other two melting peaks, and its displacement is bigger relative to that of peak II. With increasing the annealed temperature from 213 to 219 °C, peak II becomes increasingly stronger in intensity and peak III, however, becomes weaker and weaker in intensity till it disappears. The above analysis shows that if the annealed temperature is high enough, for pure PET, the endothermic peak III at the higher temperature would completely change to endotherm II. In fact, the changes in the peak position and size of these endotherms reflect the transformation of the formed crystals in the isothermal crystallization process with increasing the annealed temperature. (The size mentioned here and in the following parts refers to the peak area.) With the increase in the annealed temperature, the increase in the temperature of I and II shows that there are more perfect crystals formed in both primary and secondary crystallization processes. The final disappearance of endotherm III shows that the increase in the annealed temperature is more advantageous to the formation of the crystals in the primary crystallization process, and, at the same time, decreases the crystal formation in the recrystallization process. In addition, compared with peak I and III, peak II has always the larger area, indicating that in the PET crystallization process, the primary crystallization is always the predominant process. Table 1 shows that the crystallinity X c varies between different samples. Relative to pure PET, the Na+-MMT incorporated PET samples have the higher crystallinity. The reason for this could be ascribed to the heterogeneous nucleation effect of Na+-MMT particles. In addition, the milled PETSN3.5 sample also shows the higher crystallinity than the conventionally prepared PETCN3.5 sample. This again indicates that the pan-milling promotes the heterogeneous nucleation effect of Na+-MMT particles to a certain degree, leading to increase in crystallinity. However, the increase in clay loading in the milled PET/Na+-MMT nanocomposite does not influence the crystallinity much.

Fig. 2
figure 2

Melting endotherms of a pure PET, b PETSN1.5, c PETSN3.5 and d PETCN3.5 samples recorded at a heating rate of 10 °C min−1 after isothermal crystallization at different temperature

Full size image
Table 1 DSC parameters of the melting endotherms of different samples treated at different isothermal crystallization temperature
Full size table

As far as the Na+-MMT-filled PET system (Fig. 2b–d) is concerned, all samples have a DSC melting curve similar to that of pure PET, i.e., the three endotherms. With the increase in the isothermally treated temperature, the temperature of I and II increases and the one of III almost keep unchanged. However, there are still some features in DSC curves which are different from those of pure PET due to the addition of Na+-MMT particles and the mechanochemical milling effect. These differences can be analyzed as the following. The area of the three endotherms of Na+-MMT-filled system is larger than that of pure PET. In addition, the temperature of all endotherms I, II and III of Na+-MMT samples is higher than that of pure PET. Particularly, with increasing the treated temperature, the temperature of endotherm I and II of the former shifts to a higher temperature more obviously than that of the latter. The above analysis shows that the addition of Na+-MMT particles promotes the crystallization of PET macromolecular chains and also helps to form more perfect crystals. Also, under the condition of isothermally treated temperature 207 °C, for pure PET, the size of II is greater than that of III, and however, for Na+-MMT-filled system, the size of II is smaller than that of III. This indicates that at a lower treated temperature (207 °C) the addition of Na+-MMT particles is more advantageous to the formation of crystals in the recrystallization heating process. But, like pure PET, increase in the annealed temperature also obviously promotes the primary crystallization process of Na+-MMT-filled system due to the increase in size of II and decrease in size of III. However, the disappearance of III in pure PET system (213 °C) is earlier than in Na+-MMT-filled system (216 °C), indicating that the addition of Na+-MMT particles delays the disappearance of the crystals formed in recrystallization process with the increase in the treated temperature to a certain extent.

In addition, there are also the obvious differences in the multiple melting behaviors between PETSN3.5 and PETCN3.5 sample. Figure 2c, d shows that the temperature of I and II in PETSN3.5 sample is higher than that in PETCN3.5 sample at various treated temperature. For the samples treated at 207 °C, the difference in area between II and III in PETSN3.5 is smaller than that in PETCN3.5. In addition, with the increase in the annealed temperature, the increase degree of size of II in PETSN3.5 is greater than that in PETCN3.5, but the decrease degree of size of III in the former is obviously lower than that in the latter. The above analysis shows that under effect of the strong three-dimensional shear force field, the mechanochemical pan-milling with multi-functions such as strong pulverization, dispersion, mixing and stress-induced mechanochemical reaction has an important influence on the crystallization of PET macromolecular chains. The pan-milling has realized the intercalation and exfoliation of Na+-MMT layers in PET matrix (Fig. 1), making Na+-MMT particles well dispersed in PET matrix. The well-dispersed Na+-MMT particles, of course, have remarkably enhanced heterogeneous nucleation effect on the PET macromolecular chains, greatly accelerating the crystallization of PET molecular chains. As a result, more perfect crystals in the crystallization process could be formed. From the above analysis, it can be known that the effect of the pan-milling seems to be more advantageous to the formation of crystals in the primary crystallization process related to endotherm II, leading to reduction in the number of crystals formed in the recrystallization process. The reason is possibly due to the improved dispersion of Na+-MMT particles. Moreover, with increasing the annealed temperature, this tendency is more obvious.

In addition, it is found that the content of Na+-MMT has an important influence on the multiple melting behaviors of the prepared PET/Na+-MMT nanocomposite. The increase in the Na+-MMT content increases the temperature of both I and II and decreases the difference in size between II and I (from Fig. 2b, c). Undoubtedly, the increase in the Na+-MMT loading would promote the crystallization of PET molecular chains to a greater extent, forming more perfect crystals related to endotherms I and II, respectively. It is also noticed that the increase in the Na+-MMT loading is more advantageous to the primary crystallization than to the recrystallization at various isothermally treated temperature.

Polarized microscopic analysis

The crystal morphologies of pure PET and its nanocomposites are also observed using polarized microscope. The results are shown in Fig. 3. For pure PET, the spherulites appearing are big and the perfectly grown maltese cross-image can be observed, as shown in Fig. 3a, b. For Na+-MMT contained samples (Fig. 3c–j), the incorporation of Na+-MMT particles increases the number of spherulites (relative to pure PET), which can be regarded as the nucleation sites, proving the heterogeneous nucleation action of the incorporated Na+-MMT. In addition, compared with the conventionally prepared sample (PETCN3.5), the S3M sample (PETSN3.5) shows the more number of spherulites. This indicates that the pan-milled Na+-MMT particles show the much more obvious heterogeneous nucleation effect on PET than the unmilled ones. For PETSNx samples prepared through S3M method (Fig. 3c–h), the addition of Na+-MMT particles has a different influence on the spherulite morphology of PET nanocomposite with different Na+-MMT loading. At low Na+-MMT content, e.g., 1.5%, because the well-dispersed Na+-MMT particles have a remarkable heterogeneous nucleation effect on the PET crystallization, the small amount of Na+-MMT particles acting as nuclei can induce PET macromolecular molecular chains on their surface to grow into big spherulites. As shown in Fig. 3c, d, the spherulite size of PETSN1.5 sample is larger than that of pure PET. However, the increase in the Na+-MMT content would obviously decrease the spherulite size of PETSNx sample. As shown in Fig. 3e–h, the spherulite size of PETSN3.5 and PETSN4.5 is even smaller than that of pure PET. This is because with the increase in the Na+-MMT loading, the nuclei in system increase remarkably. The increased nuclei, of course, would induce more and more spherulites formation. Many growing spherulites impinge each other, finally leading to the smaller spherulite size [43, 44]. In addition, for the conventionally prepared PETCNx sample (Fig. 3i, j), the addition of 3.5% Na+-MMT also significantly decreases the spherulite size relative to the pure PET. However, compared with the milled PETSN3.5 sample, the spherulite size of PETCN3.5 sample is still larger. This indicates that pan-milling can lead to the decrease in the size of the formed spherulite.

Fig. 3
figure 3

PLM morphology of spherulite growth for pure PET (a, b) and its composites with different loading of Na+-MMT including PETSN1.5 (c, d), PETSN3.5 (e, f), PETSN4.5 (g, h) and PETCN3.5 (i, j)

Full size image

Isothermal crystallization kinetics

The exothermal curves of isothermal crystallization of pure PET and PET/Na+-MMT samples obtained by cooling from their molten state to various crystallization temperature (T c) are shown in Fig. 4a–d. It can be seen that the isothermal crystallization behaviors of pure PET and its nanocomposite are depended on their isothermal crystallization temperature to a greater extent. With increasing T c, the corresponding exothermal peak shifts toward the direction of increasing crystallization time, i.e., the time to reach the maximum crystallization degree increases, indicating that the increase in the crystallization temperature would lead to the decrease in the crystallization rate and T c is an important factor influencing the crystallization time. Comparing the isothermally crystallization curves (Fig. 4a–d), both PETSNx and PETCNx samples show the crystallization behavior similar to that of pure PET sample at different crystallization temperature. But, at the same crystallization temperature, the Na+-MMT-filled system has the shorter crystallization time than the pure PET sample. This indicates that the addition of Na+-MMT particles reduces the time to reach the maximum crystallization degree, i.e., increases the crystallization rate of PET macromolecular chains. In addition, at the same Na+-MMT loading, compared with PETCN3.5 sample, PETSN3.5 sample has the shorter crystallization time and stronger crystallization peak, showing that the mechanochemically milled natural clay particles can promote the PET crystallization more obviously and can hence further enhance the PET crystallization rate. This is attributed to the realization of the good dispersion of Na+-MMT particles in PET matrix and their strong interfacial interactions caused by the pan-milling. For the PETSNx system, the change in the crystallization peak position is not obvious with the increase in the crystallization temperature, indicating that the effect of crystallization temperature on the PET crystallization rate is small. However, the increase in Na+-MMT content has a great influence on the crystallization peak position, indicating that the clay content shows a relatively great effect on the crystallization rate. In summary, the addition of Na+-MMT increases the PET crystallization rate and the crystallization rate would increase with the increase in the clay content.

Fig. 4
figure 4

Isothermal crystallization DSC curves of sample a pure PET, b PETSN1.5, c PETSN3.5 and d PETCN3.5 at different temperature

Full size image

The relationship between the relative crystallization degree [X(t)] and the crystallization time (t) can be described as the following formula [45]:

$$X(t) = \frac{{X_{\text{c}} (t)}}{{X_{\text{c}} (t_{\infty } )}} = \frac{{\int_{0}^{t} {\frac{{{\text{d}}H(t)}}{{{\text{d}}t}}{\text{d}}t} }}{{\int_{0}^{{t_{\infty } }} {\frac{{{\text{d}}H(t)}}{{{\text{d}}t}}{\text{d}}t} }} = \frac{{\Delta H_{\text{t}} }}{{\Delta H_{\infty } }}$$
(1)

where dH(t)/dt is the rate of heat evolution, \(\Delta H_{\text{t}}\) is the heat generated at time t and \(\Delta H_{\infty }\) is the heat produced in the total crystallization process. Figure 5a–d shows the relative crystallinity of pure PET and Na+-MMT-filled system (PETSN1.5, PETSN3.5 and PETCN3.5) at different crystallization time in isothermal crystallization process. It can be seen that all samples show a similar tendency. The crystallinity increases with the increase in the crystallization time. To reach the same crystallinity, the crystallization time required for the sample isothermally treated at higher temperature is longer. In addition, at the same crystallization time the sample isothermally treated at lower temperature has higher crystallinity. The above analysis illustrates that the decrease in crystallization temperature is beneficial to the formation of the nuclei, and however, the increase in crystallization temperature is beneficial to the growth of the crystals. For the Na+-MMT-filled system, the addition of Na+-MMT particles effectively shortens the time to reach the same crystallinity. The comparison of S3M method (Fig. 5c) with conventional method (Fig. 5d) shows that the pan-milling mechanochemical treatment also shortens the crystallization time required to reach the same crystallinity. In addition, it can be also seen that the increase in clay loading has the effect similar to that of pan-milling. Through the above analysis, it can be concluded that the clay particle incorporation, clay content increase and pan-milling mechanochemical treatment all could effectively improve the PET crystallization and enhance the corresponding crystallization rate.

Fig. 5
figure 5

Dependence of relative crystallinity X(t) on crystallization time t for the isothermal crystallization of sample a pure PET, b PETSN1.5, c PETSN3.5 and d PETCN3.5

Full size image

The isothermal crystallization kinetics of polymer for different modes of nucleation and growth can be well approximated by the Avrami equation. Assuming that the relative crystallinity (X t) increases with the crystallization time (t), the Avrami equation can be used to describe the isothermal crystallization process of PET and Na+-MMT-filled PET system as follows [46, 47].

$$1 - X_{\text{c}} (t) = \exp ( - Kt^{n} )$$
(2)
$$\log [ - \ln (1 - X_{\text{c}} (t))] = n\log t + \log K$$
(3)

where X c(t) is the relative crystallinity at time of t, n is the Avrami exponent related to crystallization nucleation and growth manner of crystals and K is the crystallization growth rate constant related to both the nucleation and growth rate. The parameters K and n can be calculated from the intercept and slope of plot of log [−ln(1 − X c(t))] versus log t according to Eq. 3. The relation of log [−ln(1 − X c(t))] versus log t of various samples is shown in Fig. 6a–d. The obtained parameters are listed in Table 2. As we know, the crystallization process of organic polymer is, generally, composed of two stages, i.e., primary crystallization process and secondary crystallization process. Figure 6 shows that for both pure PET and PET/Na+-MMT samples, all fitted curves almost consist of two linear sections and show a good linear relation at the initial crystallization stage, but gradually deviate such a relationship in the later crystallization process. This indicates that both pure PET and PET/Na+-MMT samples prepared by whether conventional method or S3M method have the secondary crystallization process, which, generally speaking, was caused by the mutual impingement of the formed spherulites in the later stage of crystallization process. For example, at the same crystallization temperature (T c = 213 °C), the time of appearing secondary crystallization for sample PETSN1.5, PETSN3.5 and PETCN3.5 is 1.25 min (log t = 0.09), 0.56 min (log t = 0.23) and 0.95 min (log t = 0.01), respectively. The obtained time is much shorter than that of pure PET (t = 1.77 min, log t = 0.24). Compared with pure PET, the Na+-MMT incorporated PET samples show the more apparent linear relationship. The reason for this may be also explained by the secondary crystallization of PET [48]. Pure PET seems to show the stronger secondary crystallization capacity than the PET nanocomposites, leading to the more deviations from a linear relationship for the former. In addition, at the same clay loading, the time when the secondary crystallization of PETSN3.5 sample appears is also obviously shorter than that of the conventional sample. The above analysis shows that the mechanochemically milled Na+-MMT particles can make the PET primary crystallization process finish earlier. It is known that the spherulite growth of pure PET is slow, and hence, the longer time is needed to impinge each other between spherulites. However, the added Na+-MMT particles as the effective nucleation agents for PET can increase the crystal formation rate. Consequently, in the earlier stage, many crystallites or small spherulites formed would grow quickly to impinge each other, hence shortening the primary crystallization process of PET molecular chains.

Fig. 6
figure 6

Plots of log{−ln[1 − X(t)]} versus log t at the indicated temperature for the isothermal crystallization of sample a pure PET, b PETSN1.5, c PETSN3.5 and d PETCN3.5

Full size image
Table 2 Avrami kinetic parameters obtained from the Avrami equation for the isothermal crystallization of different samples
Full size table

The Avrami exponent n and the general crystallization rate constant K can be obtained from the Avrami equation. The corresponding data of various samples are listed in Table 2. It can be seen that the Avrami exponent n of pure PET and Na+-MMT-filled PET sample varies in the range of 2–5, e.g., for sample pure PET, PETSN1.5, PETSN3.5 and PETCN3.5, the corresponding n value ranges from 3.656 to 4.302, 3.260 to 4.654, 3.416 to 4.930 and 2.785 to 3.513, respectively. Generally, the Avrami exponent n reflects the different nucleation modes in the crystallization process and the dimensionality of spherulite growth. According to the literatures [48, 49], the Avrami exponent n of 2–3 corresponds to the two-dimensional growth mode [49] and the Avrami exponent n of 3–4 corresponds to the three-dimensional spherical growth mode [48, 49]. If the Avrami exponent n is more than 4, this means that the corresponding PET crystallization becomes much more complicated and may be related to the increased number of growth points (the branching points on the lamellae) in crystal nuclei [49]. Because the Avrami exponent n of pure PET is basically close to 4, the nucleation mode of pure PET is the three-dimensional spherulitic growth from homogeneous nuclei. The Avrami exponent n of the milled PETSNx samples is in the range of 3–4 at lower crystallization temperature (207–216 °C) and is greater than 4 at higher crystallization temperature (219 °C), respectively. This means that the nucleation mode of PETSNx sample at lower temperature exhibits the three-dimensional spherulitic growth from heterogeneous nuclei due to the heterogeneous nucleation effect of the Na+-MMT particles and, however, would become complicated at higher temperature possibly due to the increase in the number of growth points (the branching points on the lamellae) in crystal nuclei. The Avrami exponent n of the unmilled PETCN3.5 sample is basically in the range of 2–3 in the full crystallization temperature range of 207–219 °C investigated, indicating the nucleation mode of PET belongs to the two-dimensional growth from heterogeneous nuclei. Totally, PETSNx samples show the higher Avrami exponent than both pure PET and PETCNx. The increase in the value n means the increase in the dimensionality of spherulite growth. The above results show that under the strong three-dimensional shear force field the existence of the mechanochemically milled Na+-MMT particles is beneficial to increase in the dimensionality of PET spherulite growth to a greater extent.

Generally, the general crystallization rate constant K is related to the nucleation rate and spherulite growth process. As shown in Table 2, the K value of all samples decreases with the increase in the crystallization temperature, indicating that the increase in the crystallization temperature would lead to the decrease in the PET crystallization rate [50]. For PETSNx samples, the crystallization rate constant increases with the addition of Na+-MMT particles and would also increase with the increase in their loading. Meanwhile, at the same clay content, PETSN3.5 sample has a much higher K value than PETCN3.5 sample. The calculation shows that the K value of the former is 60 times bigger than that of pure PET and 7 times bigger than that of the latter. The above experimental results fully show that the well-dispersed, intercalated and exfoliated Na+-MMT nanoparticles obtained by the strong pan-milling effect, which can act as the effective nucleation agent for PET, are more beneficial to the enhancement of the corresponding crystallization rate than the conventionally prepared sample.

The crystallization half-time (t 1/2) is defined as the time to reach the half crystallinity. The t 1/2 can be calculated according to the formula t 1/2 = (ln 2/K)1/n. The obtained t 1/2 data are shown in Fig. 7 and Table 2. It can be seen that for all samples, with increasing the crystallization temperature, the t 1/2 also increases (Fig. 7). Generally, the longer the t 1/2, the slower the crystallization rate. The values of t 1/2 of PETCN3.5 and PETSNx samples are much lower than that of pure PET. In addition, for milled Na+-MMT system, the values of t 1/2 present a decreasing tendency with the increase in the clay loading. The above results show that either the addition of the milled Na+-MMT particles or the increase in the clay content would enhance the PET crystallization rate. In addition, it is also seen that PETSN3.5 sample has the lower t 1/2 value than PETCN3.5 sample, indicating that relative to the latter, the former has a higher crystallization rate. Pan-milling promotes the PET crystallization, which is accordance with the previous results. Also, according to Eq. 1, the crystallinity at the highest crystallization rate can be calculated assuming dQ(t)/dt = 0 (Q(t) is the heat flow rate dH(t)/dt). In this way, the time (t max) to reach the highest crystallization rate can be obtained by using the following formula t max = [(n − 1)/nK]1/n [51]. The calculated results are included in Table 2. It can be seen that t max shows a variation tendency very similar to that of t 1/2 and the value of the two parameters is also very close, indicating that the highest crystallization rate occurs near the t 1/2 time.

Fig. 7
figure 7

Dependence of t 1/2 on T c of different PET/pristine Na+-MMT nanocomposites during isothermal crystallization process

Full size image

Crystallization activation energy

Generally speaking, in the isothermal crystallization process, the crystallization rate constant K and the isothermal crystallization activation energy ΔE can be calculated from the following equations [52]:

$$K^{1/{\text n}} = K_{0} {\text{Exp}}( - \Delta E/RT_{\text{c}} )\quad {\text{or}}\quad \frac{1}{n}\ln K = \ln K_{0} - \Delta E/RT_{\text{c}}$$
(4)

where T c is the isothermal crystallization temperature, K 0 is a constant irrelevant to temperature, R is the gas constant and ∆E is the overall apparent activation energy [48]. Specifically, ∆E = ∆ϕ + ∆F, where ∆ϕ is the activation energy for forming critical sized nuclei (nucleation activation energy) at crystallization temperature and ∆F is an activation energy of a unit polymer crystallite crossing through the liquid–solid interface (transformation activation energy) [48]. Generally, when the polymer melt is cooled to crystallize at the crystallization temperature, ∆ϕ dominates and ∆F could be negligible [48]. Therefore, ∆E can be approximately seen as the ∆ϕ. Figure 8 shows the plotting of (1/n)ln K versus 1/T c. The isothermal crystallization activation energy ∆E can hence be obtained from the slope of the fitted line (Fig. 8). The calculated ∆E of pure PET, PETSN1.5, PETSN3.5 and PETCN3.5 samples is − 250.2, − 247.7, − 232.4 and − 285.5 kJ mol−1, respectively. It is noted that the obtained crystallization activation energy takes negative values, as also reported in many other literatures [53,54,55]. The reason for this may be that the crystallization activation energy corresponds to a temperature close to melting temperature, i.e., nucleation controlled temperature region [53]. (The melting temperature of endotherm I of PET/Na+-MMT sample we prepared is in the range of 215–230 °C and the crystallization temperature we used is in the range of 207–219 °C, so both are close.) For mechanochemically milled PETSNx system, the absolute value of crystallization activation energy presents a possible decrease tendency with increasing the Na+-MMT content. Also, compared with both pure PET and PETCN3.5 samples, PETSN3.5 sample has the lower absolute value of crystallization activation energy. This is possibly because the pan-milling enhances the heterogeneous nucleation effect of Na+-MMT particles on the PET macromolecular chains, thus decreasing the energy barrier for nucleation, i.e., reducing the crystallization activation energy. As a result, the enhanced interfacial interaction between PET macromolecular chains and the milled Na+-MMT particles would decrease the work needed to create a new nucleus, thus increasing the crystallization rate [48]. The above result again shows that the milled pristine Na+-MMT particles can make the PET macromolecular chains be more easily crystallized and could further enhance the PET crystallization rate in the isothermal crystallization process.

Fig. 8
figure 8

Plot of (1/n)ln k versus 1/T c for calculating the isothermal crystallization activation energy ∆E of different samples

Full size image

Conclusions

The multi-melting behaviors and isothermal crystallization kinetic of pure PET and PET/Na+-MMT samples prepared through S3M method and conventional method, respectively, are investigated. The results of multi-melting behavior analysis show that there are three endotherms, i.e., I, II and III, appearing in the DSC heating curves of the involved samples after the isothermal crystallization experiments. The increase in the annealed temperature would promote the primary crystallization process of the tested samples. The addition of Na+-MMT particles contributes to the crystallization of PET macromolecular chains and also the formation of more perfect crystals, and, however, delays the disappearance of the crystals formed in recrystallization process with the increase in the annealed temperature. The mechanochemical pan-milling proves to be beneficial to the formation of crystals in the primary crystallization process related to endotherm II and can decrease the crystals formed in the recrystallization process due to the improved dispersion of Na+-MMT particles. In addition, the increase in Na+-MMT loading is more advantageous to the primary crystallization process. It is also found that both pan-milling effect and increase in the used Na+-MMT loading can decrease the size of the formed spherulite. The results of the isothermal crystallization kinetics investigation show that the crystallization rate can be increased by the mechanochemical pan-milling treatment, the addition of Na+-MMT particles and the increase in the clay loading. The Avrami analysis indicates that the isothermal crystallization process is composed of the primary stage and the secondary stage. The secondary crystallization process exists equally in pure PET and PET/Na+-MMT samples. But the mechanochemically milled Na+-MMT particles can obviously shorten the PET primary crystallization process and are more beneficial to increase in the dimensionality of PET spherulite growth. The Avrami exponent n increases with increasing the milled Na+-MMT content, and the n value of PETSNx sample is higher than that of pure PET and PETCNx sample. In addition, the crystallization activation energy of PETSNx is lower than that of pure PET as well as PETCNx, indicating that the milled Na+-MMT particles make PET more easier and rapidly crystallize during the isothermal crystallization process.