Crystallization behavior of functional polypropylene grafted graphene oxide nanocomposite

Abstract

With an aim to understand the role of polymer grafted graphene oxide (GO) in crystallization processes, a functional polypropylene (PP) grafted GO nanocomposite, prepared by isocyanate group-contained polypropylene (PP-g-TMI) reacting with hydroxyl and carboxyl groups on GO, was investigated in terms of isothermal and non-isothermal crystallization by differential scanning calorimetry. Comparing with the PP-g-TMI/natural graphite (NG) nanocomposite, a fully exfoliated and uniformly dispersed GO in the PP-g-TMI matrix for the PP-g-TMI/GO nanocomposite was affirmed by transmission electron microscopy and a large enhancement of its storage modulus. For isothermal crystallization at 140 °C, the addition of GO into PP-g-TMI accelerates the crystallization rate dramatically more than that of NG, indicating the grafted GO can act as a very efficient heterogeneous nucleation agent to increase nucleation dramatically. However, for isothermal crystallization at 126 °C, the crystallization rate of PP-g-TMI accelerated by grafted GO is inferior to that by NG, which can deduce that the grafted GO can restrict migration and diffusion of polymer molecular chains to the surface of the nucleus for spherulitic growth due to strong covalent binding with PP-g-TMI, the formed highly viscous and dense GO layers. These two converse effects of the grafted GO on crystallization can also explain the interesting non-isothermal crystallization behavior that the grafted GO increases the crystallization rate of PP-g-TMI at a low cooling rate, while it decreases the crystallization rate at a high cooling rate.

---

1. Introduction

Polymer nanocomposites have been an area of active research due to their potential to achieve great property improvement by dispersing a small amount of nano-fillers in polymer matrices. It is particularly important for fabricating high-performance semi-crystalline polymer materials because the introduction of nano-fillers can not only act as conventional reinforcing agents but also change their crystallinities and crystalline morphologies which play a critical role in determining the ultimate properties of these nanocomposites. For example, a novel β-nucleating agent supported onto the surface of octadecylamine functionalized graphene oxide (GO) with only 0.1 wt% loading can result in a maximum increase in the impact strength of polypropylene (PP) upto almost 100% and a simultaneous improvement of the tensile strength upto about 30%.¹

Through examining various inorganic nano-fillers such as montmorillonite,² carbon carbonate,³ silica,⁴ and carbon nanotube^5,6 as nucleating agents, it is well accepted that the excellent nucleating ability of nano-fillers generally results from their extremely high specific surface areas and large aspect ratios. In particular, graphene, a two-dimensional sheet consisting of sp² carbon atoms arranged in a honeycomb structure, has attracted considerable attention due to its high specific surface area (2600 m² g⁻¹) and fascinating properties such as outstanding mechanical strength, extreme thermal conductivity, and thermal stability.^7,8 Despite a great deal of efforts have been devoted to prepare graphene-based polymer nanocomposites, it is difficult to prevent the aggregation and stacking of graphene due to poor interfacial adhesion between nanosheets and surrounding matrices. Thus, GO, an important derivative of graphene with hydroxyl and carboxyl functional groups located at its basal planes and edges, is suitable as an alternative candidate to allow for achieving its uniform dispersion in polar polymers such as poly(methyl methacrylate), and polyamide. More importantly, the presence of these functional groups allowed GO being functionalized by physical adsorption/grafting protocols to achieve good dispersion in non-polar polymer. For instance, modified GO can be homogeneously dispersed in PP.⁹

It has been verified that GO can exhibit a high nucleation activity resulting in an obvious improvement in crystallinity and crystalline temperature of semi-crystalline polymer such as PP and polylactic acid (PLA).^10–14 Xu et al.¹⁴ suggested that the introduction of GO into PP leading to the acceleration of crystallization rate was attributed to the GO-induced intrachain conformation ordering because they found that long ordered structures of PP were formed under the presence of GO especially in the early stage of crystallization. While, Fan et al.¹⁵ noted that the number of nucleation sites and the rate of crystallization of PP/GO composites decreased with the increase of annealing time due to in situ thermal reduction of GO. The surface modification of GO also acts as an important factor to affect the crystallization of polymer matrix as well as the dispensability of the particles. For example, Ryu et al.¹⁶ revealed the crystallinity and crystalline temperature would increase with the length of alkylamine chain on the GO surface because GO modified by a longer alkylamine chain had stronger interaction with PP to promote the dispersion of GO in PP. Yuan et al.¹⁷ showed that a high content of GO modified with p-phenylenediamine and cyanuric chloride could act as a β-nucleating agents for PP crystallization. Although such interactions can promote the dispersion of GO in polymer matrix and facilitate polymer chain arrangement in lattice, they also restrict the movement of polymer chains, and consequently lead to a decrease in crystallization kinetics of the polymer matrix. As reported by Huang et al.,¹⁸ the nucleation effect of GO was dominant to achieve accelerated overall crystallization kinetics of PLA matrix at low GO concentration of 0.25 and 0.5 wt%; while, as the GO concentration rose up to 1.0 wt%, the mobility and diffusion of PLA chains were constrained significantly resulting in crystallization rate from promotion to restriction. Therefore, for the development of semi-crystalline polymer/GO nanocomposite, it is very necessary to understand the effect of the physical and/or chemical interaction between GO and polymer matrix on its crystallization behavior. Nevertheless, a comprehensive understanding for the role of GO grafted by polymer chain, which has the identical chemical structure to that of the matrix polymer, on the crystallization behavior of polymer nanocomposite has not received any attention.

In this work, PP is adopted as a model polymer to probe the effect of GO grafted by polymer chain on its crystallization process. Herein, an isocyanate group-contained polypropylene (PP-g-TMI) is firstly synthesized via free-radical grafting of 3-isopropenyl-α,α-dimethylbenzyl isocyanate (TMI) onto PP. Then, the modified PP reacts with hydroxyl and carboxyl groups on GO to form a functional polypropylene grafted GO nanocomposite. The crystallization behavior of obtained nanocomposite is investigated by differential scanning calorimeter to understand the role of grafted GO in different crystallization process of polymer.

2. Experimental

2.1 Materials

PP pellets used in this study are a commercial grade (F401) from Sinopec Yangzi Petrochemical Co., Ltd., China. Its melt-flow index is 2.6 g per 10 min at 230 °C under loading of 2.16 kg. TMI with a molecular weight of 201.27 g mol⁻¹ and a boiling point of 270 °C per 760 mmHg is purchased from Aldrich Co. LLC., USA. Dicumyl peroxide (DCP) is used as the free radical initiator for TMI grafting PP. Natural flake graphite (NG) with 99% purity and an average lateral size of 40 nm size is purchased from Aladdin Industrial Co., China. Xylene and N-methyl-2-pyrrolidone (NMP) are used as the solvents for preparation of PP/carbon particle nanocomposite.

2.2 Synthesis of PP-g-TMI

PP-g-TMI was synthesized by a free-radical grafting reaction in a Brabender torque rheometer which had a volume capacity of 50 ml and was equipped with two screws. In the free radical grafting process, DCP (0.34 g) was first dissolved in liquid TMI (2.5 ml) and styrene (1.25 ml) followed by mixing with PP pellets (50 g). Those components were charged simultaneously into the mixing chamber, and were mixed with a rate of 65 revolutions per minute (rpm) at 190 °C. After 8 min of reaction, samples were quickly taken from the chamber. The resulting samples were dissolved in xylene with mechanical stirring at 120 °C, and then were charged into acetone at room temperature. PP-g-TMI was precipitated out while the unreacted TMI monomer and copolymerized TMI remains in acetone. Finally, PP-g-TMI was separated by filtration, and then dried in a vacuum oven at 80 °C for 24 h. According to the method described elsewhere,¹⁹ TMI's content in PP-g-TMI is 1.1 wt%.

2.3 Preparation of graphene oxide

Graphene oxide (GO) was prepared from NG powder using a modified Hummers method.²⁰ In a typical preparation process, NG powder (2 g) was first introduced into a solution of H₂SO₄ (98%, 52 ml) and NaNO₃ (1 g) at 0 °C and stirred for 10 minutes. Then, 6 g of KMnO₄ was divided into five portions and put into the solution in every ten minutes. After this reaction lasted for 4 h at below 5 °C, the solution was heated to 35 °C and stirred again for half an hour. Over this time, the solution thickened and turned a brownish-gray in color. Subsequently, 90 ml of water was slowly and carefully introduced to the reaction vessel, at this time, temperature of the solution rose to about 95 °C. After the solution was kept at this elevated temperature for 15 min, 300 ml of water with 3% hydrogen peroxide (H₂O₂) was added to the reaction vessel. The obtained products were washed with 5% hydrochloric acid (HCl) until sulfate ions were no longer detectable with barium chloride (BaCl₂), and then products were washed with water five times to remove impurities. Lastly, the GO powder was obtained through drying the slurry in a vacuum oven at 60 °C.

2.4 Preparation of polypropylene composite

Various PP composites were prepared by solution blending. The detailed processes are as follows: a desired amount of GO or NG was first dispersed in NMP by ultrasonication at room temperature for 1 h followed by mechanical stirring for another 2 h; at the same time, a desired amount of PP-g-TMI was dissolved in xylene by mechanical stirring at 120 °C; then, the dispersion solution of GO or NG in NMP was dropped into the PP-g-TMI solution, and then stirred for 2 h; finally, the resulting solution was introduced into acetone to coagulate the composite. The composite was obtained by filtration using a membrane (0.22 μm) and drying in a vacuum oven at 70 °C for 24 h.

2.5 Characterization of the graphene and nanocomposite

X-ray photoelectron spectroscopy (XPS). X-ray photoelectron spectroscopy (XPS) is used to evaluate the atomic compositions of NG and GO. Those measurements were carried out on an ESCALAB electron spectroscopy from VG ESCALAB MARK II with 1253.6 eV Mg Ka radiation. The binding energies are referenced to the C1s line at 284.8 eV from adventitious carbon.

X-ray diffraction (XRD). X-ray diffractometries of NG, GO and polymer nanocomposite were performed using an X'Pert PRO (PANalytical) X-ray diffractometer (40 kV and 45 mA) in the 2θ range of 5–40° at a rate of 0.04° s⁻¹.

Differential scanning calorimetry (DSC). The crystallization behaviors of PP-g-TMI and its composites were studied using a differential scanning calorimetry (DSC) of type DSC7 Perkin-Elmer under nitrogen atmosphere.

For the non-isothermal crystallization process, samples were subjected to the following thermal cycle: they were first heated from 30 °C to 200 °C at a rate of 10 °C min⁻¹, and maintained at 200 °C for 3 min to erase their previous thermal history; and then they were cooled down to 30 °C at different cooling rate (5, 10, and 15 °C min⁻¹). From DSC curves of non-isothermal crystallization, the relative crystallinity as a function of temperature can be defined as:

In which, dH denotes the measured enthalpy of crystallization, T_i is the instantaneous crystallization temperature, and T_o and T_e denote the onset and endset crystallization temperature, respectively. With a constant cooling rate (Ø), T_i can be converted to the crystallization time, t, using the following relationship:

Thus, the relative crystallinity as function of time, X_t, is defined as:

For the isothermal crystallization process, samples were first annealed at 200 °C for 5 min to erase their previous thermal histories, then were cooled to the desired crystallization temperature at a rate of 20 °C min⁻¹, finally were maintained at set crystallization temperature until the crystallization was completed. All thermal traces were recorded for later data analysis.

Rheological characterization. A HAAKE rheometer (RS 6000) was used to characterize the rheological behavior of PP-g-TMI and its composites. Before measurement, samples were thermal-pressed into disks of 20 mm in diameter and 1 mm in thick at 180 °C. A dynamic frequency sweep mode was performed to measure the elastic modulus (G′), loss modulus (G′′), and complex viscosity (η*) as a function of angular frequency (ω). The strain amplitude was set at 1%, which was in the range of the linear viscoelastic shear oscillation. All tests were carried out at 180 °C within the frequency range from 0.01 to 100 rad s⁻¹.

Structural and morphological characterization. The morphologies of polymer nanocomposites were characterized by transmission electron microscope (TEM) of type JEM-1200EX (Martinez-Salazar and Cannon, 1984). Prior to TEM analysis, specimens were cut to films of less than 100 nm thick with a cryomicrotome equipped with a glass knife at −90 °C. TEM measurements are generated with an accelerating voltage of 90 kV.

3. Results and discussion

3.1 Characterization of GO

Fig. 1 shows the diffraction patterns of NG and GO. As expected, NG has a strong diffraction peak at 26.5° while it does not appear in the XRD pattern of GO, which indicates a complete oxidation of NG by the modified Hummers method. Moreover, a new peak in the diffraction pattern of GO is observed at about 10.2°, corresponding to 0.87 nm of layer-to-layer distance. This distance is larger than that of NG due to the intercalating oxide functional groups, which further indicates a complete oxidation of NG into GO.

Fig. 1 XRD patterns of NG and GO.

The atomic composition of GO was analyzed by XPS, as shown in Fig. 2. The sharp peaks at 285 eV and 530 eV are attributed to carbon atom and oxygen atom, respectively. Based on these two peak areas, it can be known that the mole ratio of carbon atom and oxygen atom in GO is 2.65 : 1. In order to keep the same carbon content in polymer composites, it should keep in mind that in this work, the loadings of GO in the composites are 0.5 wt%, 1.0 wt% and 2.0 wt% corresponding to 0.34 wt%, 0.67 wt% and 1.34 wt% loadings of NG, respectively.

Fig. 2 XPS spectra (a) and deconvolution of high resolution spectra corresponding to Cls (b) of as-prepared GO.

Moreover, in order to understand the functional groups on the GO, a deconvolution of high resolution spectrum corresponding to C1s at about 285 eV is also shown in Fig. 2b. It is obvious that there are some oxygen-containing functional groups such as –C–O–, –C=O, and –COOH on the GO surface corresponding to the characteristic peaks at 286.6 eV, 287.8 eV, and 289.0 eV, respectively.²¹

3.2 Characterization of PP-g-TMI composite

Fig. 3 shows the FTIR spectra of PP-g-TMI and its composite with 0.5 wt% GO. PP-g-TMI can be characterized by a peak at 2255 cm⁻¹ corresponding to the isocyanate group of TMI and another one at 2722 cm⁻¹ corresponding to PP backbone. Moreover, it can be noted that the FTIR absorption ratio between 2255 and 2722 cm⁻¹, A(2255)/A(2722), will decrease obviously once the introduction of GO via a solution mixing technique, which indicates that the reaction between the isocyanate groups in PP-g-TMI and hydroxyl and carboxyl groups on GO can occur to form the PP-g-TMI grafted GO. So this grafted GO may produce a strong interaction with unreacted PP-g-TMI, leading to a higher degree of exfoliated and dispersed GO in PP-g-TMI matrix. In order to confirm this point, TEM is used to evaluate the morphologies and the dispersion/distribution states of GO in the PP-g-TMI composites, as shown in Fig. 4. TEM images of PP-g-TMI composites with 0.67 wt% and 1.34 wt% NG are also presented in Fig. 4 for comparison. Clearly, for PP-g-TMI/NG composites as shown in Fig. 4A and C, it is a thick tactoid-like structure with agglomerates of NG due to the intrinsic immiscibility between NG and PP-g-TMI. However, GO can be well exfoliated and homogeneously dispersed in the PP-g-TMI matrix, as show in Fig. 4B and D, which can further confirm that there is a strong interaction between GO and polymer matrix.

Fig. 3 FTIR spectra of PP-g-TMI (a) and PP-g-TMI/GO 0.5 wt% (b).

Fig. 4 TEM micrographs of PP-g-TMI composites with (A) 0.67 wt% NG, (B) 1 wt% GO, (C) 1.34 wt% NG, and (D) 2 wt% GO.

Rheology behavior can be another effective tool for quantifying the dispersion of polymer composite.^22,23 Fig. 5 shows the shear storage moduli (G′) of PP-g-TMI and its composites from a dynamic frequency scan measurement at 180 °C. It can be seen that the G′ values of composites are higher than that of neat PP-g-TMI in the studied frequency range. This may ascribe to the presence of high modulus graphite sheets. It can also be seen that that PP-g-TMI/GO composite exhibits a higher G′ value than that of PP-g-TMI/NG composite, which may mean the presence of better dispersed GO sheets in the former composite. Moreover, it should be noted that the former composite shows an obvious “tail” in G′ at low frequencies while it is not case for the latter, which further confirm that GO is better dispersed throughout the PP-g-TMI matrix as it is generally reported in the literature that a “tail” in the storage modulus can be observed when nanoparticle sheets are fully exfoliated.^24–27

Fig. 5 Storage modulus versus frequency for neat PP-g-TMI and its composites with 2 wt% NG and 2 wt% GO at 180 °C.

3.3 Non-isothermal crystallization behaviors of PP-g-TMI and its composites

The non-isothermal crystallization behaviors of the neat PP-g-TMI and its composites were investigated by DSC. Fig. 6 presents representative exothermal curves of heat flow as a function of temperature during the non-isothermal crystallization for neat PP-g-TMI and its composites at a cooling rate of 10 °C min⁻¹. As shown in Fig. 6, it is obvious that when the loading of 0.67 wt% NG is introduced, a significant elevation in crystallization peak temperature (T_c) of PP-g-TMI is observed from 113.8 °C to 118.2 °C. However, it is unexpected that a better dispersed PP-g-TMI composite with 1 wt% GO exhibits only 0.2 °C increase in T_c compared with neat PP-g-TMI. Similar results are also further affirmed at different loadings of NG and GO, as summarized in Fig. 7. Clearly, as the NG loading rises, T_c first increases dramatically, and then gradually reaches a platform beyond 2 wt%. However, T_c is only slightly elevated with the increase of GO loading. More importantly, the curve of T_c as a function of the particle loading for PP-g-TMI/NG composites is still above that for PP-g-TMI/GO composites. It seems that the effect of such grafted GO on PP-g-TMI's crystallization is inferior to that of NG as an additive irrespective of the particle content in composite. However, interestingly, it is not the case that when cooling rate drops to 5 °C min⁻¹, the addition of NG and GO exhibits a similar significant elevation on T_c of PP-g-TMI. More specifically, comparing with the neat PP-g-TMI, T_c of PP-g-TMI composites with 0.67 wt% NG and 1 wt% GO increase from 121.8 °C to 125.2 °C, and 125.1 °C, respectively.

Fig. 6 Non-isothermal crystallization curves of PP-g-TMI, PP-g-TMI/NG 0.67 wt%, and PP-g-TMI/GO 1 wt% at a cooling rate of 10 °C min⁻¹.

Fig. 7 Crystallization peak temperatures of PP-g-TMI composites with different contents of NG and GO at a cooling rate of 10 °C min⁻¹.

In order to further investigate this interesting non-isothermal crystallization behavior induced by grafted GO, the crystallinities (X_c) of PP-g-TMI and its composites at different cooling rates are compared in Table 1. From Table 1, it can be seen that irrespective of cooling rate, the introduction of NG and GO will decrease the crystallinites of PP-g-TMI, especially for GO. The development of relative crystallinities (X_t) of PP-g-TMI and its composites as a function of crystallization time at different cooling rates are further compared in Fig. 8. From these curves in Fig. 8, the time to reach 50% of relative crystallinity (t_1/2), indicator of the crystallization rate, can be obtained, as shown in Fig. 9. From Fig. 8 and 9, it can be seen that irrespective of PP-g-TMI or its composites, the curves of relative crystallinity as a function of time shift toward the left, and t_1/2 values decrease obviously with an increase of cooling rate, which indicates that the crystallization rate is faster at a higher cooling rate. More interestingly, at a cooling rate of 3 or 5 °C min⁻¹, t_1/2 values of the PP-g-TMI composites with 0.67 wt% NG and 1 wt% GO are almost same, and are much lower that of the neat PP-g-TMI, which indicates that a strong heterogeneous nucleation influence on PP-g-TMI crystallization occurs in the present of NG or GO. As the cooling rate rises to 10 °C min⁻¹, there is only 0.08 min reduction in t_1/2 from 0.96 to 0.88 min as the introduction of 1 wt% GO into PP-g-TMI. Its acceleration on crystallization rate is much inferior to that of 0.67 wt% NG that reduces t_1/2 from 0.96 to 0.64 min. Moreover, further increasing the cooling rate to 20 or 25 °C min⁻¹, the curve of relative crystallinity as a function of time for PP-g-TMI composite with 1 wt% GO locates a longer time domain, and t_1/2 increases instead of the decrease as compared with neat PP-g-TMI. This may imply that such grafted GO may inhibit, rather than promote, the overall crystallization rate of PP-g-TMI at the cooling rate above 20 °C min⁻¹. While for PP-g-TMI composite with 0.67 wt% NG, t_1/2 is still shorter than that of neat PP-g-TMI, which indicates that NG always exhibits an enhancement on crystallization of PP-g-TMI.

Table 1 Crystallinities of PP-g-TMI and its composites with 0.67 wt% NG and 1 wt% GO at different cooling rates

Samples	Xc (%)
	3 °C min^−1	5 °C min^−1	10 °C min^−1	20 °C min^−1	25 °C min^−1
PP-g-TMI	48.0	47.1	46.1	45.4	45.2
PP-g-TMI/NG 0.67%	47.6	47.0	45.1	44.3	43.9
PP-g-TMI/GO 1%	45.6	44.4	43.5	43.4	43.2

---

Fig. 8 Development of relative crystallinities as a function of crystallization time of PP-g-TMI and its composites with 0.67 wt% NG and 1 wt% GO at different cooling rates. (a) 3 °C min⁻¹; (b) 5 °C min⁻¹; (c) 10 °C min⁻¹; (d) 20 °C min⁻¹; and (e) 25 °C min⁻¹.

Fig. 9 Effect of the cooling rate on t_1/2 of PP-g-TMI and its composites with 0.67 wt% NG and 1 wt% GO in the non-isothermal crystallization process.

The foregoing results show GO can react with PP-g-TMI to achieve a perfect exfoliation and homogeneous dispersion in the PP-g-TMI matrix, while NG is dispersed in PP-g-TMI as agglomerate states. Usually, a better dispersion of nanoparticle means a larger number of nuclei to enhance the crystallization of polymer. However, in this work, only at a low cooling rate, GO exhibits an almost same influence on the crystallization rate of PP-g-TMI with that of NG, while a poorer enhancement on overall crystallization rate of PP-g-TMI than that of NG at a high cooling rate. The reason for such an interesting crystallization behavior may result from the antipodal effect of grafted GO on crystallization grow at different cooling rates but still unclear. Therefore, it is necessary to explore the isothermal crystallization to gain a deeper insight into the role of grafted GO in the crystallization behavior of PP-g-TMI.

3.4 Isothermal crystallization behaviors of PP-g-TMI and its composites

As reported in literature,¹⁵ GO can be reduced during the isothermal crystallization at a high temperature, and then may influence the crystallization behavior of PP/GO composite. The isothermal crystallization for PP-g-TMI composite with 1 wt% GO at 140 °C was repeated for three times, as shown in Fig. 10. From Fig. 10, it can be seen that the DSC thermograms of repeated measurements are almost superposition completely, which indicates that the process of isothermal crystallization at temperature below 140 °C has hardly effect on the crystallization behavior of PP/GO composite.

Fig. 10 DSC thermograms of isothermal crystallization for PP-g-TMI composite with 1 wt% GO at 140 °C for three times repeated measurements.

Fig. 11 compares exothermic curves of PP-g-TMI and its composites with 0.67 wt% NG and 1 wt% GO during the isothermal crystallization at 140 °C, 136 °C, 132 °C, and 126 °C. It can be seen that with the decrease of isothermal crystallization temperature, all exothermic curves of PP-g-TMI and its composites shift toward a short time domain and the peaks become narrow, indicating a fast crystallization rate at a high supercooling degree. Moreover, it is worth noting that at the same isothermal crystallization temperature, the completion of crystallization process for PP-g-TMI composite is obviously faster than that of neat PP-g-TMI, which can be concluded that the addition of NG or GO can significantly accelerate the crystallization rate of PP-g-TMI in the investigated crystallization temperature range. This can be more clearly affirmed by their t_1/2 values, as shown in Table 2. For example, at 136 °C, the introduction of 0.67 wt% NG or 1 wt% GO is seen to reduce t_1/2 value of PP-g-TMI from 8.73 to 6.15 min, or 5.23 min, respectively.

Fig. 11 DSC thermograms of isothermal crystallization for PP-g-TMI and its composites with 0.67 wt% NG and 1 wt% GO at 140 °C (a), 136 °C (b), 132 °C (c), and 126 °C (d).

Table 2 Crystallization kinetic parameters for neat PP-g-TMI and its composites with 0.67 wt% NG and 1 wt% GO at different isothermal crystallization temperatures

Samples	Tc (°C)	n	k × 10^−4 (min^−n)	t1/2 (min)
PP-g-TMI	140	3.9	0.1	17.63
	136	3.5	4.0	8.73
	132	3.4	56	4.07
	126	3.4	3467	1.25
PP-g-TMI/NG 0.67 wt%	140	3.3	0.9	14.43
	136	3.4	13.8	6.15
	132	3.4	182	2.87
	126	3.4	12 023	0.85
PP-g-TMI/GO 1 wt%	140	3.0	5.0	11.18
	136	3.3	31.6	5.23
	132	3.2	328	2.73
	126	3.3	10 233	0.88

---

More interestingly, at 140 °C, the beginning of isothermal crystallization process for PP-g-TMI composite with 1 wt% GO is obvious earlier than that with 0.67 wt% NG, and t_1/2 value of the former is 23% shorter than that of the latter, which implies that the induction ability of GO on the crystallization of PP-g-TMI is stronger than that of NG. As the isothermal crystallization temperature drops to 132 °C, t_1/2 values of PP-g-TMI composite with 1 wt% GO and 0.67 wt% NG decrease to 2.87 min and 2.73 min, respectively. There is less than 5% difference in t_1/2 values between these two PP-g-TMI composites. Further decreasing the isothermal crystallization temperature to 126 °C, the t_1/2 value of PP-g-TMI composite with 1 wt% GO is 3.5% longer, rather than shorter, than that with 0.67 wt% NG, which may infer that the enhancement of GO on the crystallization kinetic of PP-g-TMI is inferior to that of NG. Therefore, there is an interesting crystallization behavior that at a high isothermal crystallization temperature, the efficiency of GO to accelerate the crystallization of PP-g-TMI is superior to that of NG, while it is reversed at a low isothermal crystallization temperature.

It has been well documented that the incorporation of nanoparticle into semi-crystalline polymer can affect the crystalline morphology and crystallization kinetic. In order to explore whether the difference of crystallization rate between PP-g-TMI and its composites is caused by their crystal forms, XRD is employed to analyze their crystal morphologies after isothermal crystallization at 144 °C and 130 °C, as shown in Fig. 12. All six samples appear the same characteristic diffracting peaks at 14.1°, 16.8°, 18.6°, 21.0° and 21.9° correspond to the (110), (040), (130), (111), and (131/041) planes of α-form crystals, respectively. This indicates that the presence of NG and GO has no effect on the crystal form of PP-g-TMI in the studied range of isothermal crystallization temperature.

Fig. 12 XRD curves of PP-g-TMI and its composites with 0.67 wt% NG and 1 wt% GO after isothermal crystallization at 130 °C and 144 °C, respectively.

Based on the traditional crystallization viewpoint, the overall crystallization rate depends on two stages: one is nucleation, and the other is growth. In order to further understand the above interesting crystallization behaviors of PP-g-TMI and its composites, the nucleation and growth during the isothermal crystallization are analyzed using the derived Avrami equation as following:

lg(−ln(1 − X_t)) = n lg t + lg k (4)

where n is the Avrami exponent related to the type of nucleation and growth geometry of crystal, and k is the crystallization rate constant involving nucleation and growth rate. These two kinetic parameters can be obtained from the slop and intercept of linear fitting from the plot of lg(−ln(1 − X_t)) as a function of lg t, respectively. Table 2 summarizes n and k values of PP-g-TMI and its composites at different isothermal crystallization temperatures.

As seen in Table 2, n values for PP-g-TMI and its composites are around 3–4 in the investigated range of isothermal crystallization temperature, which implies that their crystallizations mostly follow by a three-dimensional spherulitic growth in a sporadic and instantaneous nucleation. However, it should be also noted that comparing to PP-g-TMI and its composite with 0.67 wt% NG, an obvious smaller n value as the incorporation of 1 wt% GO into PP-g-TMI can be found, especially at a high isothermal crystallization temperature. For example, at 140 °C, the introduction of GO into PP-g-TMI can drop the n value from 3.9 to 3.0. This decline of n value can usually ascribe to the following reasons: (1) nanoparticles, such as NG and GO, can act as heterogeneous nucleation agents to increase the instantaneous nucleation; (2) the crystal growth is confined from the impingement on adjacent single crystal due to a high nucleation density, and the restriction from a crowed and dense nanoparticle network. Therefore, a smaller n value of PP-g-TMI/GO composite may result from the strong covalent binding between GO and PP-g-TMI that not only can provide a template for the conformational ordering of PP-g-TMI chains to strongly promote the heterogeneous crystallization but also can build in a well dispersed and dense GO network, as indicated by TEM image (Fig. 4) and rheological behavior (Fig. 5), to produce a severely confined space for the growth of PP-g-TMI nucleus.

From Table 2, it can also be seen that k values of PP-g-TMI and its composites display a same trend with t_1/2 values as described above: (1) those k values increase with decreasing crystallization temperature irrespective of PP-g-TMI or its composites, indicating a high supercooling is beneficial to improve the crystallization rate; (2) at the same crystallization temperature, k values of PP-g-TMI composites are higher than that of neat PP-g-TMI, which imply the incorporation of NG and GO into PP-g-TMI can accelerate the crystallization rate; (3) with decreasing the isothermal crystallization temperature, k value of PP-g-TMI composite with 1 wt% GO is first higher and then lower than that with 0.67 wt% NG, which may infer that the grafted GO has a dual influence on the crystallization rate of PP-g-TMI. One is a positive effect on the crystallization, as it can act as an effective heterogeneous nucleation agent to increase k value; the other is a negative effect on the crystallization, as the strong force between GO and PP-g-TMI and the increased system viscosity lead to a decrease in the migration and diffusion of polymer molecular chains to the surface of the nucleus for spherulitic growth, giving rise to the decrease of k value.

These kinetic parameters, n and k, depend on the crystallization temperature, as shown in Table 2. As a matter of fact, according to Lauritzen–Hoffman nucleation theory,^28,29 the relationship of the crystallization rate (G) and crystallization temperature can be expressed as:

in which, G_o is the pre-exponential factor, U* is the activation energy for the transport of polymer segments to the crystallization site, R is the gas constant, T_∞ is a hypothetical temperature at which all chain motions cease, K_g is the nucleation constant, ΔT is the supercooling defined as the difference between T_c and the equilibrium melting point (T^o_m), f is a corrective factor given by . The first exponential term of eqn (5) is considered as the contribution of the chain migration to nucleus growth. The second exponential term is ascribed as the contribution of nucleation process. Based on eqn (5), it can be known that with crystallization temperature decrease, the first exponential term is declined while the second exponential term is elevated. Therefore, the crystallization temperature has an antipodal influence on the crystallization rate.

When the isothermal crystallization is preceded at a high temperature, e.g. 140 °C, the crystal nucleus for neat PP-g-TMI is difficult to form. However, once NG or GO is added into PP-g-TMI, the crystallization rate is dramatically improved, as illustrated in Table 2 and Fig. 11a. At that temperature, it is obvious that nucleation should be the dominant factor in determining the overall crystallization rate. From Table 2 and Fig. 11a, it can be clearly known that GO exhibits a better heterogeneous nucleating efficiency to reduce the nucleating barrier and accelerate the overall crystallization than that NG does. Thus, the covalent binding of GO and PP-g-TMI not only can promote the dispersion of GO to provide more nucleation sites but also adsorb the PP-g-TMI chains on the surface of GO to promote the conformational ordering and reduce the nucleating barrier. As the crystallization temperature drops, the nucleation rate significantly increases, while the enhancement on viscosity decreases the mobility of polymer chain, resulting in the crystal growth retardation. The determined factor for the crystallization rate turns gradually from nucleation to growth. Besides a well dispersed and dense GO network provides a severely confined space for PP-g-TMI crystal growth, the existence of strong covalent binding of oxygen-containing functional groups on GO layers and isocyanate functional groups in PP-g-TMI dramatically suppresses the mobility and diffusion of matrix chain, giving rise to a decrease in the positive effect on the overall crystallization rate from its heterogeneous nucleation so the difference of the crystallization rate between PP-g-TMI/GO composite and PP-g-TMI, and PP-g-TMI/NG composite reduces gradually. Further decreasing the crystallization temperature, the migration of PP-g-TMI chain to the crystal growth fronts becomes the dominant factor in determining the overall crystallization rate. The heavily constrained crystal growth for PP-g-TMI/GO composite shows a slower crystallization rate comparing to PP-g-TMI/NG composite at 126 °C. It is rational to speculate that at a lower crystallization temperature, the crystallization rate of the PP-g-TMI/GO composite will even become slower than that of neat PP-g-TMI.

Based on the above discussions, it isn't difficult to understand that GO and NG as additives exhibit different interesting influences on crystallization behaviors of PP-g-TMI during the non-isothermal crystallization process. At a low cooling rate, there is enough time for crystal growth so that nucleation is the dominant factor to determine the overall crystallization rate. Thus, the introduction of GO or NG into PP-g-TMI exhibits a high heterogeneous nucleation efficiency, resulting in a much faster crystallization rate comparing to neat PP-g-TMI. However, at a high cooling rate, crystal growth becomes the dominant factor for the overall crystallization rate. The grafted GO will dramatically constrain the mobility of PP-g-TMI chain for crystal growth, leading to a slower crystallization rate than that of PP-g-TMI/NG composite, even neat PP-g-TMI.

4. Conclusions

An isocyanate group-contained polypropylene (PP-g-TMI) was first synthesized via free-radical grafting of 3-isopropenyl-α,α-dimethylbenzyl isocyanate (TMI) onto polypropylene (PP), and subsequently reacted with hydroxyl and carboxyl groups on graphene oxide (GO) to form a functional polypropylene grafted GO composite. TEM images and rheology behaviors show that GO can be fully exfoliated and uniformly dispersed in the PP-g-TMI matrix. The crystallization behavior of this obtained composite was studied by differential scanning calorimetry to understand the role of the grafted GO in different crystallization processes of PP-g-TMI. The crystallization behavior of PP-g-TMI/NG composite with a poor dispersion was also investigated for comparison. For the isothermal crystallization at a high temperature in which nucleation is the dominant factor in determining the crystallization rate, the introduction of GO into PP-g-TMI more dramatically accelerates the crystallization rate of PP-g-TMI than that NG does; while at a low temperature in which crystal growth becomes the dominant factor in determining the crystallization rate, the grafted GO for the acceleration of the crystallization rate is inferior to that of NG. This indicates that the grafted GO has a two-fold influence on crystallization kinetics of semi-crystalline polymer. On the one hand, the grafted GO can act as a very efficient heterogeneous nucleation agent to increase dramatically nucleation, as a positive effect; on the other hand, the grafted GO can restrict the migration and diffusion of polymer molecular chain to the surface of the nucleus for spherulitic growth due to the strong covalent binding with PP-g-TMI, the formed high viscosity, and the dense GO layers, as a negative effect. The two antipodal influences can explain the interesting non-isothermal crystallization behavior that at a low cooling rate, the grafted GO can increase the crystallization rate of PP-g-TMI, while at a high cooling rate, it doesn't increase but decreases the crystallization rate.

Acknowledgements

The authors thank the National Natural Science Foundation of China (51203133), the Fundamental Research Funds for the Central Universities (2015FZA4026), and the Research Foundation of State Key Laboratory of Chemical Engineering for their financial support.

References

1. R. Y. Bao, J. Cao, Z. Y. Liu, W. Yang, B. H. Xie and M. B. Yang, J. Mater. Chem. A, 2014, 2, 3190–3199.
2. Z. Z. Yu, G. H. Hu, J. Varlet, A. Dasari and Y. W. Mai, J. Polym. Sci., Part B: Polym. Phys., 2005, 43, 1100–1112.
3. J. Ding, J. N. Shangguan, W. H. Ma and Q. Zhong, J. Appl. Polym. Sci., 2013, 128, 3639–3651.
4. S. S. Hwang and P. P. Hsu, J. Ind. Eng. Chem., 2013, 19, 1377–1383.
5. R. D. K. Misra, D. Depan and J. Shah, Phys. Chem. Chem. Phys., 2013, 15, 12988–12997.
6. J. Z. Xu, T. Chen, C. L. Yang, Z. M. Li, Y. M. Mao, B. Q. Zeng and S. H. Benjamin, Macromolecules, 2010, 43, 5000–5008.
7. D. F. Wu, Y. X. Cheng, S. H. Feng, Z. Yao and M. Zhang, Ind. Eng. Chem. Res., 2013, 52, 6731–6739.
8. V. Singh, D. Joung, L. Zhai, S. Das, S. I. Khondaker and S. Seal, Prog. Mater. Sci., 2011, 56, 1178–1271.
9. F. You, D. Wang, X. Li, M. Liu, G. H. Hu and Z. M. Dang, RSC Adv., 2014, 4, 8799–8807.
10. J. Z. Xu, Y. Y. Liang, H. D. Huang, G. J. Zhong, J. Lei, C. Chen and Z. M. Li, J. Polym. Res., 2012, 19, 9975.
11. Y. S. Yun, Y. H. Bae, D. H. Kim, J. Y. Lee, I. J. Chin and H. J. Jin, Carbon, 2011, 49, 3553–3559.
12. H. Wang and Z. B. Qiu, Thermochim. Acta, 2011, 526, 229–236.
13. S. H. Ryu and A. M. Shanmugharaj, Mater. Chem. Phys., 2014, 146, 478–486.
14. J. Z. Xu, Y. Y. Liang, G. J. Zhong, H. L. Li, C. Chen, L. B. Li and Z. M. Li, J. Phys. Chem. Lett., 2012, 3, 530–535.
15. J. S. Fan, R. J. Huang, S. B. Ye, T. J. Li and J. C. Feng, Polymer, 2014, 55, 4341–4347.
16. S. H. Ryu and A. M. Shanmugharaj, Chem. Eng. J., 2014, 244, 552–560.
17. B. H. Yuan, C. L. Bao, L. Song, N. N. Hong, K. M. Liew and Y. Hu, Chem. Eng. J., 2014, 237, 411–420.
18. H. D. Huang, J. Z. Xu, Y. Fan, L. Xu and Z. M. Li, J. Phys. Chem. B, 2013, 117, 10641–10651.
19. G. H. Hu, H. X. Li and L. F. Feng, J. Appl. Polym. Sci., 2003, 88, 1799–1807.
20. W. S. Hummers and R. E. Offeman, J. Am. Chem. Soc., 1958, 80, 1339.
21. S. Stankovich, D. A. Dikin, R. D. Piner, K. A. Kohlhaas, A. Kleinhammes, Y. Y. Jia, Y. Wu, S. T. Nguyen and R. S. Ruoff, Carbon, 2007, 45, 1558–1565.
22. M. Moniruzzaman and K. I. Winey, Macromolecules, 2006, 39, 5194–5202.
23. C. Penu, G. H. Hu, A. Fernandez, P. Marchal and L. Choplin, Polym. Eng. Sci., 2012, 52, 2173–2181.
24. C. M. Koo, M. J. Kim, M. H. Choi, S. O. Kim and I. J. Chung, J. Appl. Polym. Sci., 2003, 88, 1526–1535.
25. Z. H. Xu, Y. H. Niu, L. Yang, W. Y. Xie, H. Li, Z. H. Gan and Z. G. Wang, Polymer, 2010, 51, 730–737.
26. R. Rezanavaz, M. K. R. Aghjeh and A. A. Babaluo, Polym. Compos., 2010, 6, 1028–1036.
27. C. A. Avila-Orta, F. J. Medellín-Rodríguez, M. V. Dávila-Rodríguez, Y. A. Guirre-Figueroa, K. Yoon and B. S. Hsiao, J. Appl. Polym. Sci., 2007, 106, 2640–2647.
28. J. D. Hoffman, R. L. Miller, H. Marand and D. B. Roitman, Macromolecules, 1992, 25, 2221–2229.
29. H. Marand, J. Xu and S. Srinivas, Macromolecules, 1998, 31, 8219–8229.

---