Fabrication of a polymer/aligned shish-kebab composite: microstructure and mechanical properties

Abstract

A shear-induced ultrahigh molecular weight polyethylene (UHMWPE) shish-kebab mat (USKM), in which shish-kebabs are well aligned, has been achieved from a dilute UHMWPE solution. Such a USKM is incorporated into linear low density polyethylene (LLDPE) as a reinforcement filler. Isothermal crystallization shows that the USKM acts as an effective nucleating agent for LLDPE crystallization, leading to enhanced crystallinity, long period, and lamellar thickness of the USKM/LLDPE composite. Instead of shish-kebabs, only aligned fibrils are observed in the composite. Such fibrils consist of a shish structure wrapped with LLDPE chains and good bonding at the interface between the fibril and matrix is achieved. Consequently, the interfacial adhesion is expected to improve and favorable stress transfer is realized. Together with good bonding, the superior strength of the shish structure and the above enhanced crystalline parameters contribute to a significant increment in the tensile strength and modulus of the composite, as compared to pure LLDPE.

---

1. Introduction

The shish-kebab structure has attracted tremendous attention since it was first observed in sheared dilute polyethylene solution.^1–3 Well documented, this structure typically consists of an extended-chain central fibrillar core, forming the shish, and disk-like folded chain crystals, forming the kebabs, which are always oriented perpendicularly to the shish.^4–8 Shish-kebabs are well known to be induced by flow, such as in sheared dilute polymer solutions^2,3 and sheared or stretched polymer melts.^9–11 As a highly oriented structure, a shish-kebab is endowed with superior strength and modulus along its longitudinal axis,^7,8 providing the potential to fabricate self-reinforced polymer products.^12–15 The content of shish-kebabs has become a decisive factor for the enhancement of polymer products.^9,16,17 However, products fabricated via common processing methods cannot get reinforced as expected. Take injection-moulded parts as an example: a very small amount of shish-kebabs can be in situ developed in the final injection-moulded parts, and moreover they are only located in an extremely thin layer close to the surface due to the intense shear and fast solidification therein.^18,19 Whereas, numerous spherulites instead of shish-kebabs are always developed in a very broad inner layer due to weak shear and sufficient relaxation of chains caused by the slow cooling of melts therein.²⁰

To date, studies on semi-crystalline polymer-based (e.g., polyethylene and polypropylene) composites have indicated that reinforced fillers are roughly classified into two categories: inorganic and organic ones. Due to the extraordinary mechanical properties, a wide range of inorganic fillers including clay,^21–24 calcium carbonate,^25,26 carbon nanotubes,^27,28 etc., have been utilized to prepare reinforced composites. As for the organic fillers, various types have attracted more and more attention because of their good compatibility with polymer matrices, such as polymer fiber^29,30 and electrospun polymer nanofiber.^31,32 In view of some of the intrinsic features of the shish-kebabs, e.g., superior strength and modulus along the longitudinal axis and the huge potential to fabricate self-reinforced polymer products, we are wondering whether shish-kebabs could be used as reinforcement fillers and directly incorporated into a heterogeneous matrix to realize the reinforcement of a polymer matrix. If so, this method can also overcome the aforementioned problem that products fabricated using common processing methods cannot get reinforced as expected due to the limited amount of shish-kebabs. Nevertheless, studies concerning this have been scarcely reported. The reasons may be as follows: (1) with respect to the shish-kebabs obtained from the sheared dilute polymer solutions, common blending with a heterogeneous matrix may have the risk of destroying the as-prepared oriented structure of the shish-kebabs; (2) the content of the shish-kebabs obtained from sheared or stretched polymer melts is extremely low, and moreover it is impossible for them to be detached completely from the total products.

In this study, our interest is to explore the feasibility of direct introduction of the shish-kebabs into a heterogeneous polymer matrix as ideal reinforced elements. An ultrahigh molecular weight polyethylene (UHMWPE) shish-kebab mat (USKM) was firstly prepared through the shear-induced crystallization of UHMWPE solution and then incorporated into a linear low density polyethylene (LLDPE) matrix. Subsequently, compression moulding was utilized to fabricate a USKM/LLDPE composite. Results indicate that even a small amount of the USKM can notably enhance the LLDPE matrix. This work opens a new gateway for directly incorporating the shish-kebabs into a polymer matrix with the purpose of preparing high-performance polymer products.

2. Experimental section

2.1 Materials

UHMWPE powder was provided by the Second Auxiliary Factory, Beijing, China, with a number-average molecular weight (M_n) in the range of 2.0–3.0 × 10⁶ g mol⁻¹. The nominal melting point of it is 142 °C (obtained by the differential scanning calorimetry (DSC) method, a sample of 5–8 mg was heated to 180 °C from room temperature at a heating rate of 10 °C min⁻¹). Commercially available LLDPE (Trade marked as 7042) with a melt flow index (MFI) of 2.0 g/10 min (190 °C, 2.16 kg) and weight-average molecular weight (M_w) of 1.4 × 10⁵ g mol⁻¹, was supplied by Maoming Petrochemical Corp., China. Its density is 0.92 g cm⁻³. Xylene (analytical reagent, 99%) was bought from Tianjin Chemical Reagents Plant.

2.2 Fabrication of the shear-induced USKM

A shear-induced USKM was fabricated through shear-induced crystallization of a UHMWPE solution. UHMWPE powder was firstly dissolved completely in xylene at 140 °C to obtain the UHMWPE solution with a concentration of 0.01 wt% and then rapidly transferred into a bath pot whose temperature was 105 °C. The shear-induced crystallization was carried out at this temperature using a mechanical stir bar at a speed of 500 rpm, which is shown in Fig. 1a. Ferrum frames were fixed on the stir bar so that the frames could rotate synchronously with the stir bar. Thus, the UHMWPE solution was subjected to shear arising from the rotating ferrum frames on which the USKM could be collected. After the shear-induced crystallization for 30 min, the collected shish-kebab mat was taken out from the solution and washed with xylene to remove the free polymers. Before characterization, the mat was dried at room temperature for 24 hours (see Fig. 1b).

Fig. 1 (a) A schematic of the shear-induced crystallization setup used in this study; (b) a digital photo of the shear-induced USKM; (c) an SEM image of the shear-induced USKM with low magnification. The numbers in (a) represent: 1-ferrum frame; 2-UHMWPE/xylene solution; 3-stir bar; 4-motor. The blue arrow in (b) refers to the shear direction.

2.3 Composite preparation

In view of interfacial adhesion, polyethylenes (PE), such as high density polyethylene (HDPE), LLDPE, etc., must be the ideal candidates for acting as a matrix because they have the same monomer as that of UHMWPE.^16,33 Moreover, the prerequisite for realizing the reinforcement of the composite is that the USKM must be intact as much as possible when the matrix is in the molten state during compression moulding. To this end, the melting behaviors of the candidates (e.g., HDPE, LLDPE) should be taken into account. As shown in Fig. 2, the melting zone of the shear-induced USKM generally ranges from 120 °C to 150 °C. That is to say, the end melting point of the candidates must be near or below 120 °C (viz., the onset melting point of the USKM) in order to preserve the as-prepared USKM. However, the melting of the as-received HDPE ends at about 137 °C, at which the USKM must be partially melted (see Fig. 2). As for LLDPE, the end melting point is about 125 °C, very close to the onset melting point of the USKM (120 °C). Therefore, LLDPE is the ideal candidate because the USKM would be maintained as almost intact when LLDPE is completely melted at about 125 °C.

Fig. 2 The DSC melting curves of the as-received LLDPE, HDPE granules and the shear-induced USKM (samples of 5–8 mg were respectively heated to 180 °C from room temperature at a heating rate of 10 °C min⁻¹).

The USKM/LLDPE composite was prepared according to the following procedure. After the shear-induced crystallization and drying process, the USKM supported by ferrum frames was obtained. However, the USKM was not stripped off from the ferrum frames in order to avoid shrinkage of the USKM. Firstly, the shear-induced USKM supported by the ferrum frame was preheated to 100 °C. Then, a LLDPE/xylene solution with a concentration of 10 wt% was only casted on a piece of the USKM supported by the ferrum frame. Thus, the USKM cannot float on the surface of the LLDPE solution due to the support of the ferrum frame. After the solvent had completely evaporated, the USKM was stripped off the ferrum frame. The USKM/LLDPE composite was finally obtained using a vacuum-assisted compression moulding machine (FM-11, Beijing Future Materials Science-Technology Co., Ltd). The benefit of such vacuum-assisted compression moulding machines is twofold: firstly, oxidative degradation will not occur under a vacuum condition; secondly, compared with the common compression moulding machine, such machines can eliminate bubbles in the compressed samples as much as possible. The detailed scenario of the compression moulding process was interpreted as follows: the casted USKM/LLDPE composite film was preheated at 130 °C for 10 min and pressed at a pressure of 0.3 MPa for 5 min, followed by isothermal crystallization at 115 °C for 1 hour. After that, it was cooled to room temperature and the USKM/LLDPE composite was finally obtained. The concentration of USKM in the USKM/LLDPE composite was 0.8 wt% and its calculation method can be found in the ESI.† For comparison purposes, pure LLDPE film in the absence of USKM was also prepared under the same conditions described above.

2.4 Characterization

A macroscopic USKM was first observed using a digital camera (Sony WX9).

To better characterize the surface morphology of the shear-induced USKM, scanning electron microscopy (SEM, JEOL JSM 7500F) was employed under an acceleration voltage of 5 kV. Moreover, an SEM measurement was also adopted for identifying the interfacial interactions between the UHMWPE shish-kebabs and the matrix. Before the SEM observation, the compression moulded USKM/LLDPE composite and the pure LLDPE were etched according to our previous work.³⁴ The samples were sputtered with gold before observation.

Atomic force microscopy (AFM, SPM-9500J3, Japan) was also employed to observe the surface morphology of the shear-induced USKM. It was conducted at ambient conditions (25 °C) using tapping mode probes with a constant amplitude. The AFM image was recorded at the resonance frequency of the cantilever with a scan rate of 1 Hz. Moreover, the AFM image was taken at 256 × 256 pixels.

In order to clarify the nucleating effect of USKM towards LLDPE, isothermal crystallization was conducted using differential scanning calorimetry (TA DSC-2920) under a pure nitrogen atmosphere. Both the USKM/LLDPE composite and the pure LLDPE of 5–8 mg were respectively heated to 130 °C at a heating rate of 10 °C min⁻¹ and held for 2 min to erase the possible thermo-mechanical history. Then, the samples were quickly cooled down to 115 °C at a cooling rate of 30 °C min⁻¹ for isothermal crystallization. The melting behaviors of the USKM/LLDPE composite and the pure LLDPE were also investigated. The detailed scenario of the DSC melting measurement was set as follows: samples of 5–8 mg were heated to 180 °C at a heating rate of 10 °C min⁻¹.

Two-dimensional wide angle X-ray diffraction (2D-WAXD) and two-dimensional small angle X-ray scattering (2D-SAXS) experiments were both conducted at the beam line BL16B1 of the Shanghai Synchrotron Radiation Facility (SSRF). An X-ray beam with a wavelength of 0.124 nm was used. To collect 2D-WAXD and 2D-SAXS patterns, a Mar165CCD detector (2048 × 2048 pixels with a pixel size of 80 μm) was employed. The samples were placed with their orientation (shear direction) perpendicular to the incident X-ray beam. The data acquisition time was 60 s per frame. For all the 2D-WAXD and 2D-SAXS profiles, the amorphous background was extracted. It should be noted that for X-ray measurements, to get the maximum diffraction/scattering intensity, several samples were orderly stacked (viz., all the stacked samples were parallel along the shear direction) to obtain a total thickness of about 1 mm.

A tensile test was carried out on a Linkam TST350 tensile stage with a 200 N load cell. Rectangular specimens with widths of 5 mm and lengths of 10 mm were cut from the centre of the compression moulded samples along the shear direction (see Fig. S1 in ESI†). The tensile direction was along the orientation direction of the shish-kebabs. The pulling speed was 100 μm s⁻¹ and the measured temperature was around room temperature (25 °C). The reported values were calculated as averages over at least five specimens.

3. Results and discussion

3.1 Morphology of shear-induced USKM

Fig. 1b shows the macroscopic morphology of the USKM obtained via shear-induced crystallization. The SEM image with low magnification in Fig. 1c demonstrates that the USKM is composed of numerous fibrils oriented along the shear direction. Furthermore, the detailed morphology observed using SEM with high magnification is shown in Fig. 3. It can be evidently seen that these fibrils are orderly aligned along the shear direction despite the locally disordered arrangement. The amplified image reveals that all these fibrils consist of a single fibrillar core and many crystalline lamellae decorated perpendicularly to the fibrillar core (see Fig. 3b). Such characteristics confirm that the fibrils are typical shish-kebabs.^4,5 Nevertheless, obvious breakage exists in the shish-kebabs, shown by the white arrows in Fig. 3b, which should result from the high voltage exerted on these feeble shish-kebabs. Furthermore, though the shish-kebabs are densely arranged as shown in Fig. 3, explicit voids are present between them, leading to the potential possibility for LLDPE/xylene solution penetration. In addition, the formation of the aforementioned shish-kebabs in a dilute UHMWPE solution under shear flow can be best explained using the concept of a coil-stretch transition proposed by de Gennes:³⁵ the UHMWPE chains are generally in an equilibrium with the coiled state in the dilute solution. Once shear flow is exerted, the long coiled chains could be stretched along the shear direction and then crystallized into shish, on which, the short coiled chains are subsequently adsorbed and form kebabs.³⁶

Fig. 3 The SEM images of the shear-induced USKM. The blue arrow refers to the shear direction.

AFM was also employed to observe the detailed surface morphology of the shear-induced USKM (1.25 × 1.25 μm), as displayed in Fig. 4. Individual lamellae can be easily identified. These UHMWPE lamellae are highly aligned along the shear direction. Combined with the detailed morphology observed using SEM shown in Fig. 3, it can be safely concluded that the aligned lamellae shown in Fig. 4 are actually the shish-kebab structure. However, the kebabs are so compact that the shish is invisible. On the whole, the AFM image is a further evidence that well aligned shish-kebabs have been prepared in this study.

Fig. 4 The AFM image of the shear-induced USKM. The shear direction is horizontal.

3.2 The effect of the USKM on the isothermal crystallization behavior of LLDPE

In order to detect the nucleating effect of the USKM towards LLDPE, isothermal crystallization experiments at 115 °C for both the USKM/LLDPE composite and pure LLDPE were carried out. The isothermal crystallization exotherms of the two samples are shown in Fig. 5a. The normalized relative crystallinity (X_t) is obtained from the ratio of the exothermic peak area at a given crystallization time (t) to the total area of this exothermic peak, i.e.,where dH/dt is the heat flow rate.

Fig. 5 (a) The DSC curves of isothermal crystallization for the USKM/LLDPE composite and pure LLDPE; (b) the normalized relative crystallinity (X_t) as a function of crystallization time.

The normalized relative crystallinity versus time for the two samples is shown in Fig. 5b. It is evident that the isotherms exhibit a sigmoid dependence with time. The USKM/LLDPE composite yields a shorter induction time (0.3 min) and completion time (3.9 min), while the pure LLDPE has a longer induction time (1.0 min) and completion time (7.1 min). That is to say, the crystallization process of the composite seems to proceed steadily faster than that of the pure sample. Furthermore, the half crystallization time (t_1/2) is also obtained and presented in Fig. 5b. Obviously, t_1/2 of the USKM/LLDPE composite and pure LLDPE is 1.4 min and 2.5 min, respectively. In other words, the crystallization rate of the composite increases about 1.8 times of that for the pure LLDPE, suggesting that the USKM could act as an effective nucleating agent for LLDPE crystallization.^37,38

3.3 Microstructure of the USKM/LLDPE composite

2D-WAXD and 2D-SAXS measurements were carried out to investigate the crystalline and lamellar structures of the pure LLDPE and the USKM/LLDPE composite, and the results are shown in Fig. 6. According to the 2D-WAXD patterns, only the isotropic diffraction circles designated to the (110)_o plane and (200)_o plane of orthorhombic lamellae can be observed for the pure LLDPE, indicating a random arrangement of molecular chains.³⁹ With respect to the USKM/LLDPE composite, apart from the typical (110)_o plane and (200)_o plane, additional wide diffraction arcs designated to the (200)_m plane of the monoclinic cell are also observed. Combined with the aforementioned results of the pure LLDPE, it is obvious that this (200)_m plane is ascribed to the incorporation of the shear-induced USKM.⁴⁰ Moreover, Fig. 7a shows the azimuthal distribution integrated from the (200)_m plane of the USKM/LLDPE composite. Clearly, two wide peaks, exhibiting broad azimuthal widths, are respectively located at 90° and 270° on the azimuthal curve. It seems that the USKM incorporated into the LLDPE matrix shows a low orientation level. However, as shown in Fig. 3 and 4, shish-kebabs in the as-prepared mat are well aligned. Therefore, one can conclude that the original orientation structure is maintained to some extent despite the inevitable relaxation of the oriented molecules or the disorder of the shish-kebabs that occurred during the compression moulding process. Furthermore, the isotropic diffraction circle referring to the (110)_o crystal plane of the USKM/LLDPE composite shown in Fig. 6a₂ demonstrates that the oriented structure in the LLDPE matrix is not formed during the compression process after the incorporation of the shear-induced USKM.

Fig. 6 The 2D-WAXD (a₁ and a₂) and 2D-SAXS (b₁ and b₂) patterns of the samples: pure LLDPE (a₁ and b₁) and USKM/LLDPE composite (a₂ and b₂). The shear direction of the USKM is horizontal.

Fig. 7 (a) Azimuthal distribution integrated from the (200)_m crystal plane of the USKM/LLDPE composite; (b) 1D-WAXD curves of the pure LLDPE and the USKM/LLDPE composite.

Similar to the 2D-WAXD pattern for the pure LLDPE, the 2D-SAXS pattern of the pure LLDPE is also isotropic, manifesting that spherulites prevail herein. Whereas, with respect to the USKM/LLDPE composite, the 2D-SAXS pattern exhibits a streak along the meridian with no evidence of equatorial maxima. Considering that the LLDPE chains are not oriented obviously (see the 2D-WAXD results), it can be deduced that the meridional streak originates from the shish structure in the USKM. This is a further evidence that the incorporated USKM is not completely destroyed during the compression moulding process and the original oriented structure shown in Fig. 3 and 4 can be preserved even when subjected to high temperature during the compression moulding process.

To quantify the crystalline content, one-dimensional (1D) WAXD profiles were derived from Fig. 6 through circularly integrating the intensity of the 2D-WAXD patterns (shown in Fig. 7b). Then a Gaussian fitting of multipeaks was utilized to obtain the areas of crystalline and amorphous peaks. The overall crystallinity (X_c) was calculated according to the following equation⁴¹

where A_cryst and A_amorp are the fitted areas of the crystalline and amorphous regions, respectively. The overall crystallinity is listed in Table 1. Obviously, compared with the pure LLDPE, an increase in X_c for the USKM/LLDPE composite is achieved, which is ascribed to the heterogeneous nucleating effect of the USKM towards LLDPE. The heterogeneous nucleation could prompt the normally noncrystallizing LLDPE chains to crystallize, leading to an increment in the final crystallinity.⁴²

Table 1 The crystalline parameters obtained from the 2D-WAXD and 2D-SAXS results

Samples	Xc (%)	Lc (nm)	L (nm)
LLDPE	36.44	5.68	12.78
USKM/LLDPE	39.13	5.80	13.02

---

From the 2D-SAXS intensity, the one-dimensional correlation function (Strobl method) K(z) is calculated by cosine Fourier transformation⁴³

where q is the scattering vector, q = 4π sin θ_B/λ, θ_B denotes the Bragg angle, ∑(q) is the differential cross section per unit volume and r_e is the classical electron radius. Fig. 8a presents an example of the one-dimensional electron density correlation function. As indicated by the black arrows, this curve can give various structural parameters, e.g., long period (L) and lamellar thickness (L_c), according to the methodology proposed by Strobl et al.^43,44 Fig. 8b shows the one-dimensional correlation function for the pure LLDPE and the USKM/LLDPE composite. For the purpose of quantitative analysis, the values of L and L_c, obtained from the correlation functions are listed in Table 1. Obviously, the USKM/LLDPE composite exhibits larger L and L_c. Moreover, the aforementioned results show that higher crystallinity of the USKM/LLDPE composite is also achieved. Consequently, it is reasonable to claim that the increment of these crystalline parameters (L, L_c, X_c) is attributed to the incorporation of the USKM. As an effective nucleating agent for LLDPE (see the DSC results shown in Fig. 5), the USKM can provide heterogeneous nucleation sites for the LLDPE matrix and enhance the crystallization kinetics, consequently leading to higher crystallinity and thicker lamellae.⁴⁵

Fig. 8 (a) The calculation method of the long period (L) and lamellar thickness (L_c) obtained from the one-dimensional correlation function K(z); (b) the curves of the one-dimensional electron density correlation function of the pure LLDPE and the USKM/LLDPE composite.

3.4 Interfacial features between the incorporated shish-kebabs and matrix

Interfacial features between the shear-induced shish-kebabs and the matrix in the composite were observed using SEM. As shown in Fig. 9a, some fibrils indicated by blue arrows can be clearly observed in the composite. Owing to incomplete etching, not all fibrils can be substantially exposed. However, more fibrils can be clearly found in Fig. S2 with lower magnification, shown in the ESI.† Considering the fact that the oriented structure of the LLDPE matrix does not develop during the compression moulding process after the incorporation of the shear-induced USKM, these fibrils must be shish, originated from the USKM in the absence of the kebabs. However, the diameter of the fibrils indicated by the blue arrows (see Fig. 9b) seems to be much larger than the original shish structure shown in Fig. 3 and 4. In terms of the effective nucleating effect of the USKM towards LLDPE, one can conclude that such a shish structure is substantially wrapped with LLDPE molecules, forming the fibrils shown in Fig. 9 during the compression moulding process. Once again, SEM observation is a further evidence that the shish structures in the USKM act as nucleating templates for LLDPE matrix crystallization.

Fig. 9 SEM images of the etched surface of the USKM/LLDPE composite. The white arrow refers to the flow direction of the USKM.

Fig. 9a also shows that these fibrils are generally parallel to each other, implying that the original oriented structure of shish-kebabs has been maintained to some extent. Of note, kebabs are not observed on these fibrils. This may be due to the weaker stability of kebabs and their relaxation during the compression moulding process.²⁸ In addition, the observed fibrils are not as densely distributed as those shish-kebabs shown in Fig. 3 and 4, which might be ascribed to the two following reasons: on the one hand, the fibrils are not exposed substantially due to the incomplete etching as mentioned above; on the other hand, relaxation of the shish-kebabs with small sizes occurs during the compression moulding process.

Moreover, good bonding between the fibrils and matrix is achieved because a gap or groove cannot be observed at the interface (see Fig. 9b). The nucleating role together with good bonding at the interface between the fibrils and matrix is expected to produce anchoring of the LLDPE molecules to the fibrils and thus improve the adhesion, generating better stress transfer and higher mechanical properties of the USKM/LLDPE composite compared to the pure LLDPE.⁴⁶

3.5 Thermal behaviors

The DSC melting curves of the pure LLDPE and the USKM/LLDPE composite are exhibited in Fig. 10a. As can be seen, there are two melting peaks for the pure LLDPE. One sharp peak (primary peak) located at around 121.5 °C corresponds to the fusion of the stable crystals with thicker lamellae, while the other shoulder peak appearing at 115 °C is attributed to the melting of the small or imperfect crystals containing thinner lamellae. It is well known that incorporation of α-olefins in the polyethylene backbone allows LLDPE to have many short chain branches (SCBs). The short chain length and non-uniform distribution in the polymeric chains of LLDPE results in crystals with a wide distribution of lamellar thickness.⁴⁷ As for the USKM/LLDPE composite, the primary peak seems to become sharper and shifts to a higher temperature (122.5 °C), as compared to that of the pure LLDPE. Furthermore, the shoulder peak at about 114.9 °C is almost absent. The reason for this may be that the SCBs are prone to participating in the formation of thicker lamellae due to the effective heterogeneous nucleating effect of the USKM towards the LLDPE matrix. However, the melting peak of the USKM at around 138.6 °C shown in Fig. 2 can not be observed in the melting curve of the USKM/LLDPE composite (see Fig. 10b). The reason leading to this could be that the content of the USKM incorporated into the LLDPE matrix is extremely low (only 0.8 wt%) and difficult to detect using the DSC technique.⁴⁸

Fig. 10 (a) The DSC melting curves of the pure LLDPE and the USKM/LLDPE composite; (b) amplified image of the area in the dashed square of (a).

It is also worth mentioning that a crystallization peak for the pure LLDPE emerges at 109.4 °C. As elucidated above, owing to the existence of the SCBs and nonuniform short chain length as well as its nonuniform distribution in the LLDPE chains, a lamellar structure with a wide distribution of thicknesses could be formed.⁴⁷ When a melting experiment is conducted, the tiny and imperfect crystals will firstly melt, whereas the thicker or perfect ones are stable and remain intact. Then the molten chains of the tiny and imperfect crystals will self-nucleate and recrystallize once the crystallization temperature of LLDPE (109.4 °C) has been reached. Thus, the crystallization peak located at about 109.4 °C could be formed during the heating scan. However, such a crystallization peak shifts to 112.2 °C for the USKM/LLDPE composite. This phenomenon could be ascribed to the heterogeneous nucleation of the USKM towards LLDPE, leading to enhanced crystallization kinetics and a higher crystallization temperature.^37,49 The crystallinity (X_c) of the two samples was calculated using the equation

where ΔH_i is the enthalpy of fusion and ϕ_i is the weight fraction of the matrix polymer (LLDPE). Herein, ΔH^m_i is the enthalpy of fusion for 100% crystallized PE, which is adopted as 293 J g⁻¹.⁵⁰ The calculated crystallinity of the pure LLDPE and USKM/LLDPE composite are 47.2% and 55.0%, respectively. That is to say, the crystallinity of the USKM/LLDPE composite is higher than that of the pure LLDPE, which is well in line with the calculated results from 2D-WAXD.

3.6 Tensile test

The representative stress–strain curves of the pure LLDPE and the USKM/LLDPE composite are shown in Fig. 11. In order to quantitatively estimate the variations of mechanical properties, the values of mechanical parameters, e.g., yield strength (σ_y), ultimate tensile strength (σ), Young’s modulus (E) and elongation at break (ε) are obtained from the stress–strain curves and summarized in Table 2. One can clearly see that the LLDPE matrix has been steadily reinforced due to the incorporation of the shear-induced USKM. For example, the Young’s modulus, a manifestation of stiffness, rises from 80.6 MPa for pure LLDPE to 112.5 MPa for the USKM/LLDPE composite. At the same time, compared with the pure LLDPE, the composite exhibits a dramatic improvement in yield strength, from 5.5 MPa to 10.5 MPa with an increment of 90.9%. The ultimate tensile strength is also elevated from 13.2 MPa for the pure LLDPE to 16.1 MPa for the USKM/LLDPE composite. The mechanism of superior mechanical reinforcement will be discussed based on the aforementioned microstructural investigations.

Fig. 11 Representative stress–strain curves of the pure LLDPE and the USKM/LLDPE composite.

Table 2 Mechanical parameters obtained from the stress–strain curves

Samples	E (MPa)	σy (MPa)	σ (MPa)	ε (%)
LLDPE	80.6 ± 5.3	5.5 ± 0.7	13.2 ± 0.6	1468.0 ± 32
USKM/LLDPE	112.5 ± 4.6	10.5 ± 1.2	16.1 ± 0.8	1084.2 ± 45

---

It is well established that the macroscopic mechanical performance of the composite is determined by many microstructural factors, such as the interfacial interaction, orientation level of the incorporated shish-kebabs and the matrix crystalline phase, crystalline characters (e.g., L, L_c, X_c) and so on.⁵⁰ However, according to the 2D-WAXD and 2D-SAXS results, no obvious orientation of the matrix is detected. It is reasonable to deduce that the effect of the orientation level in the matrix on the mechanical enhancement of the composite can be ruled out. Owing to the strong nucleating role of the shish structure towards the LLDPE matrix, the interfacial adhesion and stress transfer between the shish and the LLDPE matrix must be enhanced (see Fig. 9). Moreover, the aligned fibrils containing the shish and their superior strength will play an important role in the reinforcement of the matrix. In addition, the mechanical properties of the semi-crystalline polymers generally depend on the content of the crystalline domains. Thus, the enhanced X_c must be responsible for the mechanical strengthening of the composite. Meanwhile, as well established in other studies,^51,52 the increased L and L_c also contribute to the enhanced tensile properties. Therefore, the increased X_c, L and L_c of the USKM/LLDPE composite as well as the alignment and reinforcement of the shish synergistically contribute to the increase in the mechanical properties of the USKM/LLDPE composite.

4. Conclusions

In this study, an ultrahigh molecular weight polyethylene (UHMWPE) shish-kebab mat (USKM), in which shish-kebabs are well aligned with each other, has been achieved through shear-induced crystallization. Then the USKM is incorporated into a heterogeneous matrix to prepare a reinforced composite. Via isothermal crystallization experiments, it is concluded that the USKM is highly effective in nucleating the LLDPE matrix, leading to increased crystallinity, long period and lamellar thickness for the composite. Because the shish is wrapped with LLDPE molecules, better interfacial adhesion and stress transfer can be achieved. Together with the superior strength of the shish structure, the aforementioned crystalline characteristics and better interfacial adhesion synergistically contribute to the mechanical reinforcement of the USKM/LLDPE composite. As a new method of utilizing shish-kebabs as fillers to prepare reinforced products, we believe that the compression moulded USKM/LLDPE composite holds promising applications in preparing high-performance products. Further study will be carried out to realize the application of the USKM in the preparation of high-performance composites.

Acknowledgements

We express our great thanks to the National Natural Science Foundation of China (51173171, 11172271), HASTIT of Henan Province and the State Key Laboratory of Materials Processing and Die & Mould Technology for financial support.

References

1. A. Pennings and A. Kiel, Colloid Polym. Sci., 1965, 205, 160–162.
2. A. Peterlin, Pure Appl. Chem., 1966, 12, 563–586.
3. A. Peterlin, Adv. Macromol. Chem., 1968, 1, 225.
4. B. S. Hsiao, L. Yang, R. H. Somani, C. A. Avila-Orta and L. Zhu, Phys. Rev. Lett., 2005, 94, 117802.
5. W. Hu, D. Frenkel and V. B. Mathot, Macromolecules, 2002, 35, 7172–7174.
6. L. Balzano, N. Kukalyekar, S. Rastogi, G. W. Peters and J. C. Chadwick, Phys. Rev. Lett., 2008, 100, 048302.
7. J. Odell, D. Grubb and A. Keller, Polymer, 1978, 19, 617–626.
8. Z. Bashir, J. Odell and A. Keller, J. Mater. Sci., 1984, 19, 3713–3725.
9. H. R. Yang, J. Lei, L. Li, Q. Fu and Z. M. Li, Macromolecules, 2012, 45, 6600–6610.
10. O. O. Mykhaylyk, P. Chambon, C. Impradice, J. P. A. Fairclough, N. J. Terrill and A. J. Ryan, Macromolecules, 2010, 43, 2389–2405.
11. K. Cui, L. Meng, N. Tian, W. Zhou, Y. Liu, Z. Wang, J. He and L. Li, Macromolecules, 2012, 45, 5477–5486.
12. Y. F. Huang, J. Z. Xu, J. S. Li, B. X. He, L. Xu and Z. M. Li, Biomaterials, 2014, 35, 6687–6697.
13. Y. F. Huang, J. Z. Xu, J. Y. Xu, Z. C. Zhang, B. S. Hsiao, L. Xu and Z. M. Li, J. Mater. Chem. B, 2014, 2, 971–980.
14. L. Xu, Y. F. Huang, J. Z. Xu, J. Xu and Z. M. Li, RSC Adv., 2014, 4, 1512–1520.
15. Y. H. Chen, Z. Y. Huang, Z. M. Li, J. H. Tang and B. S. Hsiao, RSC Adv., 2014, 4, 14766–14776.
16. L. Xu, C. Chen, G. J. Zhong, J. Lei, J. Z. Xu, B. S. Hsiao and Z. M. Li, ACS Appl. Mater. Interfaces, 2012, 4, 1521–1529.
17. Y. Pan, S. Shi, W. Xu, G. Zheng, K. Dai, C. Liu, J. Chen and C. Shen, J. Mater. Sci., 2014, 49, 1041–1048.
18. J. Yang, K. Wang, H. Deng, F. Chen and Q. Fu, Polymer, 2010, 51, 774–782.
19. K. Wang, F. Chen, Q. Zhang and Q. Fu, Polymer, 2008, 49, 4745–4755.
20. S. Fellahi, B. Favis and B. Fisa, Polymer, 1996, 37, 2615–2626.
21. Z. Guo and B. Hagström, Polym. Eng. Sci., 2013, 53, 2035–2044.
22. R. R. Hegde, J. E. Spruiell and G. S. Bhat, Polym. Int., 2014, 63, 1112–1121.
23. K. Wang, S. Liang, J. Deng, H. Yang, Q. Zhang, Q. Fu, D. Xia, D. Wang and C. C. Han, Polymer, 2006, 47, 7131–7144.
24. C. Zhao, H. Qin, F. Gong, M. Feng, S. Zhang and M. Yang, Polym. Degrad. Stab., 2005, 87, 183–189.
25. Q. Yuan, Y. Yang, J. Chen, V. Ramuni, R. Misra and K. Bertrand, Mater. Sci. Eng., A, 2010, 527, 6699–6713.
26. M. Tanniru and R. Misra, Mater. Sci. Eng., A, 2005, 405, 178–193.
27. K. Wang, N. Li, K. Ren, Q. Zhang and Q. Fu, Composites, Part A, 2014, 61, 84–90.
28. J. Yang, C. Wang, K. Wang, Q. Zhang, F. Chen, R. Du and Q. Fu, Macromolecules, 2009, 42, 7016–7023.
29. S. Liang, K. Wang, H. Yang, Q. Zhang, R. Du and Q. Fu, Polymer, 2006, 47, 7115–7122.
30. H. H. Li, J. C. Liu, D. J. Wang and S. K. Yan, Colloid Polym. Sci., 2003, 281, 973–979.
31. U. Stachewicz, F. Modaresifar, R. J. Bailey, T. Peijs and A. H. Barber, ACS Appl. Mater. Interfaces, 2012, 4, 2577–2582.
32. S. Jiang, G. Duan, J. Schöbel, S. Agarwal and A. Greiner, Compos. Sci. Technol., 2013, 88, 57–61.
33. J. Chen, W. Yang, G. Yu, M. Wang, H. Ni and K. Shen, J. Mater. Process. Technol., 2008, 202, 165–169.
34. H. Zheng, B. Wang, G. Zheng, Z. Wang, K. Dai, C. Liu and C. Shen, Ind. Eng. Chem. Res., 2014, 53, 6211–6220.
35. P. de Gennes, J. Chem. Phys., 1974, 60, 5030–5042.
36. I. Dukovski and M. Muthukumar, J. Chem. Phys., 2003, 118, 6648–6655.
37. M. Sabino, G. Ronca and A. Müller, J. Mater. Sci., 2000, 35, 5071–5084.
38. H. Han, X. Wang and D. Wu, J. Chem. Technol. Biotechnol., 2013, 88, 1200–1211.
39. S. Liang, H. Yang, K. Wang, Q. Zhang, R. Du and Q. Fu, Acta Mater., 2008, 56, 50–59.
40. J. T. Yeh, S. C. Lin, C. W. Tu, K. H. Hsie and F. C. Chang, J. Mater. Sci., 2008, 43, 4892–4900.
41. A. Turner-Jones and A. Cobbold, J. Polym. Sci., Part B: Polym. Lett., 1968, 6, 539–546.
42. K. Wang, K. Mai, Z. Han and H. Zeng, J. Appl. Polym. Sci., 2001, 81, 78–84.
43. G. Strobl and M. Schneider, J. Polym. Sci., Polym. Phys. Ed., 1980, 18, 1343–1359.
44. H. Bai, F. Luo, T. Zhou, H. Deng, K. Wang and Q. Fu, Polymer, 2011, 52, 2351–2360.
45. R. R. Hegde, J. E. Spruiell and G. S. Bhat, Polym. Int., 2014, 63, 1112–1121.
46. F. Mai, D. Pan, X. Gao, M. Yao, H. Deng, K. Wang, F. Chen and Q. Fu, Polym. Int., 2011, 60, 1646–1654.
47. M. Camargo, F. M. M. Camargo and C. R. Wolf, Int. J. Polym. Anal. Charact., 2008, 13, 49–65.
48. S. Ratner, A. Weinberg and G. Marom, Polym. Compos., 2003, 24, 422–427.
49. H. Goossens, S. Jain, M. van Duin and P. Lemstra, Polymer, 2005, 46, 8805–8818.
50. F. Mai, K. Wang, M. Yao, H. Deng, F. Chen and Q. Fu, J. Phys. Chem. B, 2010, 114, 10693–10702.
51. B. Pukánszky, I. Mudra and P. Staniek, J. Vinyl Addit. Technol., 1997, 3, 53–57.
52. Y. Gao, K. Ren, N. Ning, Q. Fu, K. Wang and Q. Zhang, Polymer, 2012, 53, 2792–2801.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra07453c

---