Properties of Microinjection-Molded Polypropylene/Graphite Composites.
Recently, there has been growing interest in microparts of functional filler-loaded polymer composites due to their potential applications in electronics and microelectromechanical systems. Therefore, several microprocessing methods, such as microinjection molding ([micro]IM), hot embossing, injection-compression molding, and thermoforming  as well as ultrasonic molding , have been developed to meet the ever-increasing demands of the abovementioned applications. Among these methods, [micro]IM has been widely employed, thanks to its potential for mass production of microparts at a relatively low cost per part .
Like conventional injection molding (CIM), [micro]IM includes the same processing stages, that is, plasticization of the polymer pellets, metering and injecting polymer melt into mold cavity, packing of the injected melt, cooling the molded products, and finally demolding . However, [micro]IM is not simply a scaled-down version of the CIM process. According to Kukla et al. , microparts are identified in terms of part weight (on the order of a few milligrams), part dimensions, and/or the tolerance of part dimensions (within the micrometer range). Typical products from [micro]IM include microgears, microneedles, microfilters, and microfluidic devices [5, 6]. The common feature for these microparts is the very high surface-area-to-volume ratio (up to [10.sup.3]-[10.sup.6]/m) , which poses certain challenges for the mold filling process due to the very high cooling rate effects in [micro]IM [7, 8]. To avoid premature solidification of polymer melts, certain processing parameters in [micro]IM (such as melt and mold temperatures as well as injection velocities) are typically higher than those in CIM [7-9].
Many studies have been carried out regarding the micromolding of unfilled thermoplastics [3, 10, 11]. However, applications that require microparts having specified electrical, thermal, and mechanical properties can be accommodated by using polymeric composites with multifunctional fillers . To date, a few studies have been performed with the [micro]IM of multiwalled carbon nanotube (CNT)-loaded polymer composites [12-17]. For example, Abbasi et al.  reported that the percolation threshold ([p.sub.c]) for polycarbonate (PC)/CNT microparts shifted to higher filler concentrations when compared with their compression-molded counterparts, which was attributed to the preferential orientation of CNT along the melt flow direction (FD) arising from the predominant shearing effect in this specific direction. In a previous study, we  found that higher melt temperature and mold temperature are normally advantageous to the overall enhancement of electrical conductivity for polypropylene (PP)/CNT microparts. Moreover, an increase in injection velocity is conducive to the formation of conductive pathways along the FD, whereas it shows a detrimental impact on the construction of conductive pathways across the transverse direction . Therefore, the employed processing parameters are crucial to the development of microstructure in subsequent moldings. Moreover, we also  reported that the morphology and intrinsic properties of carbon fillers can significantly affect the electrical and morphological properties of subsequent moldings. Aside from the above-mentioned factors, the selection of host polymers also played a significant role in determining the microstructure of subsequent microparts, thereby affecting their electrical and morphological properties [17, 19]. However, to the best of our knowledge, little research has been carried out for plate-like fillers, for example, graphite-filled systems.
Graphite, which is naturally abundant or can be synthesized from the petroleum coke or other precursor materials (i.e., synthetic graphite [SG]) , is known as a type of cost-effective carbon filler to enhance the transport properties of polymer matrices thanks to its low density and high thermal and electrical conductivity as well as good dispersibility . Recently, it has been reported that the low-temperature expandable graphite (LTEG) is an efficient additive that can significantly enhance the overall thermal and electrical conductivity of thermoplastic polymers at a relatively lower filler concentration when compared with flake graphite (FG) [22-25]. For example, Luo et al.  found that the percolation concentration, [p.sub.c], for LTEG filled high-density polyethylene composites is about 20 wt%, which is half of that obtained from their FG-containing counterparts. Similar findings have been reported by Zhou et al. [21, 24] in terms of enhancing the thermal conductivity of graphite-loaded polyamide 6 (PA6) composites. The efficacy of utilizing LTEG in enhancing the electrical conductivity of thermoplastic polymers in [micro]IM is of interest, since the thermomechanical history experienced in both processing methods (i.e., compression molding and [micro]IM) is very different.
In the present study, plate-like SG and LTEG were introduced into PP as the functional additives via melt blending, followed by [micro]IM. To this end, a rectangular mold insert that has a three-step decrease in thickness  along the melt FD was employed to evaluate the effect of abrupt changes in mold geometry on the filler distribution within subsequent micromoldings.
Isotactic PP (average [M.sub.w]: -250,000; average [M.sub.n]: -67,000) was purchased from Sigma Aldrich (Oakville, Canada). The PP has a density of 0.9 g/[cm.sup.3] and a melt index of 12 g/10 min (230[degrees]C/2.16 kg). An SG (Grade: A99), obtained from Asbury Graphite Mills Inc. (Asbury, NJ), is adopted as the functional filler. According to Rew et al. , the SG has an average diameter of 20 [micro]m and a typical surface area of 8 [m.sup.2]/g. The LTEG was purchased from Shijiazhuang ADT Carbonic Material Factory (China). According to the supplier, the initial expanded temperature of LTEG is about 150[degrees]C (as corroborated by Wu et al. ) and the expansion ratio is about 230 mL/g. LTEG exhibits a particle size of nearly 180 [micro]m with a density of 2.20 g/[cm.sup.3]. The morphology of as-received carbon fillers is displayed in Fig. 1.
A series of PP/SG composites were prepared under conditions of 190[degrees]C and 50 rpm for 10 min, using a Brabender mixer. In addition, a blend of PP with 30 wt% LTEG was prepared under the same conditions. The formulations for preparing graphite-containing PP composites are listed in Table 1. Then, the obtained blends were mechanically crushed and used for [micro]IM in a Battenfeld Microsystem 50 (Wittmann Battenfeld GmbH, Austria). The all-electric molding machine features a plunger injection system, which consists of a screw plasticization unit, a metering unit, and an injection unit . The diameter of the adopted injection piston is 4.89 mm. Figure 2a displays the mold insert and a final molding. The microparts were divided into three sections based on thickness. All sections from the microparts have the same width of 2.40 mm and the thickness of thick, middle, and thin sections is 0.85, 0.50, and 0.20 mm, respectively. The thick and middle sections have a length of 5.00 mm, whereas the thin section has a length of 4.80 mm. The melt temperature and mold temperature were 260[degrees]C and 100[degrees]C, respectively. The injection velocity was 300 mm/s. According to a previous study , the electrical conductivity for microparts with lower SG loading fractions (<30 wt%) cannot be determined using a Keithley 6,514 electrometer, because the resistance of those samples exceeds the upper limit of the testing unit.
To identify each system clearly, the obtained PP/SG blends were named as follows: SG30 refers to PP/SG 30 wt% composite, whereas SG30 Thick is the thick section of microparts molded from SG30. Similarly, LTEG30 denotes the concentration of LTEG in PP is 30 wt% and LTEG30 Thick indicates the thick section of microparts molded from LTEG30.
DC Electrical Conductivity Measurement
To facilitate characterization, the microparts were cut at the transition areas that are marked by red arrows in Fig. 2b. Direct current (DC) electrical conductivity ([sigma], S/cm) for each section was measured across the thickness direction (TD) and along the melt FD. The [sigma] was determined using a two-probe method detailed in a previous study . Sections taken from the microparts were placed between two copper electrodes, and applied pressure was used to minimize the contact resistance between the copper electrodes and surfaces of the sample. The resistance (R, [OMEGA]) was thus measured using a Keithley 6,514 electrometer. Afterward, the obtained R was converted into [sigma] (S/cm) using Eq. 1 :
[sigma] = 1/[rho] = 1/AR (D
where [rho] is the volume electrical resistivity, L (cm) is the distance between the copper electrodes, that is, the thickness of samples, and A ([cm.sup.2]) is the contact area.
The morphology of LTEG was observed with a Keyence VHX 6000 optical microscope (Japan). The morphology of SG and the microstructure of SG-containing samples were characterized using a high-resolution scanning electron microscope (SEM, S-4500, Hitachi). The examined surfaces of respective samples were fractured in liquid nitrogen across the TD (as shown in Fig. 2c), followed by chemical treatment to remove the amorphous phase of PP . Afterwards, the etched surface was coated with a thin layer of platinum prior to observations.
Differential Scanning Calorimetry
The melting and crystallization behavior of PP and graphitefilled PP samples were evaluated using a differential scanning calorimeter (DSC, Q200, TA Instruments) under nitrogen. First, the sample was heated from 40[degrees]C to 220[degrees]C at a rate of 10[degrees]C/min (i.e., Cycle 1). Afterward, it was cooled from 220[degrees]C to -30[degrees]C at a rate of 5[degrees]C/min (Cycle 2). It was then scanned from -30[degrees]C to 220[degrees]C at a rate of 10[degrees]C/min (Cycle 3). The crystallinity ([chi]) for each sample was calculated using the following equation.
[chi] = [DELTA][H.sub.m](/ (1-w)x [DELTA] [H.sup.O.sub.m]
where [DELTA][H.sub.m] is the heat of fusion of PP, w is the weight fraction of SG or LTEG in PP, and [DELTA][H.sup.O.sub.m] is the heat of fusion of 100% crystalline PP, that is, 207 J/g .
RESULTS AND DISCUSSION
Electrical Conductivity Measurements
The DC electrical conductivity ([sigma]) for each section of corresponding microparts was determined with respect to measurement directions, which is displayed in Fig. 3. The FD [sigma] for the thin section is not measured, because it is currently infeasible to place this section properly between the copper electrodes for the two-probe method. The FD [sigma] for both the thick and middle sections of microparts is higher than the TD [sigma] of corresponding counterparts, which indicates a role for the preferential orientation of conductive fillers along the melt FD. A similar trend has been observed for their CNT-containing PP counterparts .
Overall, the values of [sigma] increase with an incremental loading fraction of SG in PP, suggesting that the [sigma] of subsequent moldings is related to the construction of conductive pathways within the host matrix. Interestingly, the trend of [sigma] for each section of SG30 microparts is different from that of higher SG-containing or LTEG30 counterparts, which can be related to the development of microstructure in each section of subsequent moldings.
For example, the average TD [sigma] for SG30 Middle (1.75 x [10.sup.-7] S/cm) is higher than the TD [sigma] of SG30 Thick (6.23 X [10.sup.-8] S/cm); the average FD [sigma] for SG30 Thick (3.74 x [10.sup.-6] S/cm) is, however, relatively higher than the average FD [sigma] for SG30 Middle (9.67 X [10.sup.-7] S/cm). According to Ref. 28, the percolation threshold ([p.sub.c]) for compression-molded PP/SG composites is about 30 wt% (i.e., 14.8 vol%). However, the [p.sub.c] for microinjection-molded samples shifted to higher filler concentrations when compared with the samples prepared with compression molding [14, 18, 19]. Abbasi et al.  reported that the pc for compression-molded PP/CNT and PC/CNT composites is 1 wt% and 3 wt%, respectively. However, corresponding values for PP/CNT and PC/CNT microparts increased to 4 and 6 wt%, respectively. Therefore, it can be inferred that the weight fraction of SG particles within PP is insufficient to form intact conductive pathways when filler concentration is 30 wt%, as displayed in Fig. 4a and b. Li and Shimizu  reported that there would be an improved dispersion of fillers in the polymer matrix when they were processed under elevated shearing conditions. In this scenario, although direct contact among the added fillers is limited, the improved dispersion of SG facilitates the transport of electrons by "hopping" or "tunneling" mechanism .
It is anticipated that enough conductive pathways can be constructed with further increasing the loading concentration of SG. Therefore, the preferential orientation of conductive particles is favorable for the enhancement of [sigma] along the FD. For example, the FD [sigma] for SG50 Thick (1.25 X [10.sup.-3] S/cm) is approximately 2.5 times higher than that obtained from its compression-molded counterpart (5.06 X [10.sup.-4] S/cm) . Moreover, the FD [sigma] for SG50 Middle is about 4.2 times higher than the FD a for SG50 Thick. The above observation can be attributed to the combined influence of higher shearing conditions in [micro]IM  and "convergence" of the adopted mold cavities along FD. For example, there is a sharp increase of maximum shear rates amid the step transition areas, which was corroborated by Moldflow simulation . Consequently, it is likely to obtain higher degree of filler orientation in the thinner section of the microparts arising from the increased shearing conditions .
Furthermore, there is a concurrent reduction of TD a along the melt FD of the stepped microparts, which can be related to the orientation of SG particles. The preferential alignment of conductive particles favors the construction of conductive pathways along the FD, whereas their ability to form an intact conductive network across the TD would, to a certain degree, be impaired , as shown in Fig. 3b and c. A similar trend was also detected for LTEG-containing microparts, as displayed in Fig. 3d. This indicates that sufficient conductive pathways are constructed within corresponding microparts when the LTEG concentration is 30 wt% (i.e., 14.9 vol%), suggesting that lower weight fractions of LTEG fillers are required to obtain enough conductive pathways within molded samples, which is consistent with the findings by Luo et al. . According to the literature [22-25], the in situ exfoliation of LTEG during the melt blending process is essential to the formation of conductive network in polymer composites. In addition, the average particle size of LTEG (about 180 [micro]m) is larger than that of SG (nearly 20 [micro]m). Therefore, less contact resistance amid adjacent conductive fillers is expected in order to generate enough conductive pathways via physical interconnection between graphite fillers. As a result, the addition of LTEG significantly enhances the [sigma] of subsequent moldings. Additionally, the achieved [sigma] for each section, with respect to the measurement directions, of LTEG30 microparts is higher than that obtained from SG50 counterparts, implying that LTEG shows a higher efficiency to create enough conductive pathways when compared with that of SG particles. The higher [sigma] for LTEG30 microparts could adequately fulfill the requirements in certain aspects of some applications, such as sensors, and heat and electrostatic dissipation .
Fig. 3 also shows that the TD a for the thin section is lower than that of their thick and middle section counterparts, signifying that the typical shearing conditions present in the thin section are unfavorable for the random construction of 3D conductive pathways. A similar phenomenon has been observed for CNT and carbon black-filled PP microparts . Pan et al.  reported that the cooling time drops dramatically with a reduction of mold cavity thickness. Therefore, it is anticipated that the generated structure in the thin section (0.2 mm) can instantly "freeze" and have little chance returning to a random orientation [7, 17]. In this scenario, the TD [sigma] for the thin section of corresponding microparts can be substantially reduced due to a lack of sufficient conductive pathways. In general, the [sigma] in the TD will be lower than the [sigma] in the FD. This is due, in part, to the observation that, during the flow of solid-liquid suspensions, the solid particles are kept away from the solid boundary wall. As a result, the thin skin serves as an insulator.
The microstructure of compression-molded PP/SG composites was observed from the transverse section. Figure 4a indicates that SG has a high tendency to form interconnected conductive pathways within PP, which is consistent with the rapid increase of [sigma] for PP/SG composites at 30 wt% . Furthermore, it is shown in Fig. 4b that the SG particles have already formed an intact conductive network within PP, which is essential to the enhancement of a for subsequent polymer composites.
The dominant shearing conditions in [micro]IM would determine the development of microstructure in subsequent moldings, which is crucial to the enhancement of [sigma] [15, 18]. Thus, the morphology for the thick and middle sections of SG-containing microparts is shown in Fig. 5. All images were taken across the TD, as shown in Fig. 1b. The presence of SG particles within PP is indicated by white arrows. Figure 5a shows that SG particles are somewhat randomly dispersed in the thick section, whereas they tend to align along the FD, in the middle section, as shown in Fig. 5b. This can be ascribed to the plate-like structure of SG particles (as shown in Fig. 1a) and the difference of maximum shear rate that prevails in different sections of microparts . As for SG50 microparts, the SG particles have a relatively uniform distribution in either the thick section (Fig. 5c) or the middle section (Fig. 5d). In this scenario, sufficient conductive pathways can be formed at such a high filler concentration, that is, 50 wt% SG. Besides, the shear-induced orientation of SG particles in [micro]IM is favorable for the enhancement of [sigma] along the FD. For example, FD [sigma] for SG50 Thick (1.25 X [10.sup.-3] S/cm) is about 2.5 times higher than corresponding compression-molded counterparts. Meanwhile, the preferred alignment of SG along the FD would, to some extent, limit or impair the network formation across the TD . Abbasi et al.  reported that the shearing effect in [micro]IM is a few orders of magnitude higher than compression molding. Consequently, the a for compression-molded SG50 (5.06 X [10.sup.-4] S/cm)  is higher than that of TD [sigma] for SG50 Thick (1.03 X [10.sup.-4] S/cm).
The characterization of LTEG is shown in Fig. 6. Figure 6b shows the expanded volume of 0.5 g LTEG after a thermal treatment at 190[degrees]C for 10 min in a conventional oven (Binder, ED 115, Germany). As given in Fig. 6b, the expanded volume of LTEG measured by a graduated cylinder was about 7 mL/g, which is significantly higher than that of 0.5 g pristine LTEG (see Fig. 6a). The result indicated that the intercalated agents can easily escape from the stacked graphite flakes of LTEG  during the melt blending process, thereby increasing the likelihood of forming intact conductive network within resultant composites.
The development of microstructure for both the core layer and shear layer of LTEG30 Thick is displayed in Fig. 7. It is clear that intact conductive pathways can be formed within the microparts due to the large particle size and, more importantly, the in situ exfoliation of LTEG during melt processing [23-25]. For example, Wu and Drzal  found that the polymer composites with larger graphite particle size showed higher thermal conductivity when compared with those with smaller graphite particles, which was attributed to the fact that samples with larger size graphite particles have less contact resistance. Moreover, Luo et al.  reported that the in situ exfoliation of LTEG can greatly reduce the distance amid neighboring conductive particles, thereby enhancing both the thermal and electrical conductivity of subsequent blends. The morphology for both the core and shear layers of LTEG30 Middle is exhibited in Fig. 8. Similarly, compact stacking of exfoliated graphite flakes in both the shear layer and core layer of LTEG30 Middle supported that the FD Caption: FIG. 5. SEM images of (a) SG30 thick, (b) SG30 middle, (c) SG50 thick, and (d) SG50 middle. for LTEG30 Middle is higher than that of the LTEG30 Thick counterpart.
Differential Scanning Calorimetry
The melting and crystallization behavior of SG- or LTEG-containing blends and corresponding microparts were evaluated using DSC. For clarity, all displayed DSC curves were shifted vertically. Table 2 tabulates the characteristic data obtained from DSC heating and cooling curves. Figure 9a shows that there is a slight deviation of the main melting peaks (i.e., [T.sub.m]) for PP/SG blends when compared with that of pure PP, indicating that the addition of SG particles affects the melting behavior of PP. The [T.sub.m] decreases with an increase in SG content in PP. It can thus be conjectured that the presence of SG particles might retard free motion of polymer chains, which is unfavorable for the formation of stable PP crystals, thereby leading to a downshift of [T.sub.m] with increasing SG concentration. However, such a suppression effect on [T.sub.m] is less obvious for the thick section of corresponding microparts. Although the crystallinity ([chi]) from the first heating process, that is, Cycle 1, is comparable between PP/SG blends and the thick section of corresponding microparts, flow-induced orientation of polymer chains might be advantageous to the formation of more stable PP crystals, thereby leading to a relatively higher [T.sub.m] in microparts when compared with their melt blended counterparts. In addition, the crystallization temperature ([T.sub.c]) increases significantly upon the addition of SG, as reflected in Fig. 9b and e, revealing that the added fillers act as the nucleating agent. In this scenario, there is a noticeable increase (>12[degrees]C) of [T.sub.c] for both SG10 and SG10 Thick, as given in Table 2. Moreover, the [T.sub.c] increases with an incremental loading fraction of SG in PP, revealing that the crystallization of PP chains was associated with the number of crystallization sites rendered by SG particles. Similar results were reported by Kazemi et al.  for CNT-filled PP composites. Figure 9(c) and (f) revealed that the difference in [T.sub.m] between the PP/SG blends and the thick section of subsequent microparts narrowed down, which was confirmed by similar characteristic data obtained from the second heating process, that is, Cycle 3. Table 2 also shows that the (Cycle 1) [chi] for PP/SG composites and the thick section of subsequent microparts decrease with an increase in filler concentrations, which can be ascribed to the confinement effect that imposed on polymer chains with increasing number of SG particles. However, a contradictory trend is detected for the (Cycle 3) [chi], since an increasing [chi] is coupled with an increase in SG content. Herein, the difference of [chi] between both Cycles 1 and 3 suggests that the typical thermomechanical history experienced by the polymer melts during the melt blending or [micro]IM has an influence on the crystallization of PP.
The effect of in situ expansion of LTEG on the melting and crystallization behavior of PP/LTEG composite and corresponding microparts as well as its comparison with pure PP and SG-containing counterparts is shown in Fig. 10. Figure lOd shows that a mild shoulder peak for the middle section of PP and SG30 microparts is detected around 160[degrees]C (Cycle 1), whereas such behavior is absent from Cycle 3. Moreover, the melting peak for the thin section of both PP and SG30 microparts becomes narrower when compared with their thick and middle section counterparts, indicating a more homogeneous distribution of lamellae thickness. The above observations can be ascribed to the influence of thermomechanical history, which differs greatly in different sections of the three-step microparts. However, the melting and crystallization behavior of LTEG30 composite and each section of subsequent microparts displayed differently in comparison with previous observations for pure PP and SG30 counterparts. First, the [T.sub.m] for either the LTEG30 blend or each section of corresponding microparts shifted to a lower temperature range, that is, around 160[degrees]C, which is much lower than that of PP or SG-containing counterparts, as given in Table 2. The observation is contradictory to that of LTEG-filled PA6 composites [24, 25], because they found that the presence of exfoliated LTEG has a trivial impact on the [T.sub.m] of PA6. Second, there are two [T.sub.m]s observed in Cycle 1 for each section of LTEG30 microparts, whereas such a behavior disappeared in Cycle 3, as shown in Fig. 10g and i, suggesting that the appearance of double [T.sub.m] can be related to both the presence of LTEG and prevailing high shearing effect presented in [micro]IM. For example, the large particle size of exfoliated LTEG may interfere with the formation of large size crystalline domains due to the limited space and the confinement effect that imposed by graphite flakes . However, the high shearing effect along the FD is advantageous to the formation of highly aligned shish-kabab structure , which leads to the appearance of higher [T.sub.m] for subsequent microparts, as displayed in Fig. 10g. It is noted that the higher [T.sub.m] strengthens with an increase of shearing effect along FD, which further suggested an influence of flow-induced crystallization.
The [T.sub.c] for LTEG30 composite or each section of subsequent microparts is higher than pure PP counterparts, suggesting that LTEG also acts as a nucleating agent that significantly accelerates the crystallization of PP chains. However, corresponding values are lower than those of SG30 counterparts (see Table 1), which can explain that the number of incorporated LTEG particles in PP is much lower than that of SG-containing counterparts because of the larger particle size of LTEG. Therefore, based on these observations, it is suggested that the crystallization of polymer chains is largely dependent on the number of crystallization sites other than the particle size.
Overall, the [chi] for different sections of pure PP, SG30, and LTEG30 microparts followed a similar trend along FD, irrespective of heating cycles. For example, the thick section of these microparts showed almost the highest [chi] value and the [chi] for the thin section is minimum, albeit the insignificant difference between SG30 Middle (44.9%) and SG30 Thin (45.0%) from Cycle 1, and between PP Thick (48.3%) and PP Middle (48.8%) from Cycle 3. Therefore, it could be attributed to the impact of the typical thermomechanical history (i.e., high shear and elongation force fields as well as large thermal gradients) that is experienced by polymer melts in different sections of microparts along the predominant melt FD .
In summary, two different types of graphite particles, that is, SG- and LTEG-loaded PP composites were prepared by melt blending, then followed by [micro]IM. A micropart that has a three-step decrease in thickness along the FD was used to prepare specimens for testing. To facilitate characterizations, the microparts were divided into three sections based on thickness, that is, thick, middle, and thin sections, respectively. The electrical conductivity for each section of the microparts was measured using a two-probe method, which was correlated with the evolution of microstructure. Results showed that LTEG is more effective than SG in terms of enhancing the electrical conductivity, which is mainly attributed to the in situ exfoliation of inclusion fillers during the melt blending process. In addition, the measured electrical conductivity along the FD is higher than that across the transverse direction, which reflects a role of orientation of the added graphite fillers. The melting and crystallization behavior for each sample was evaluated using DSC. Results suggested that both the incorporation of graphite particles and the typical thermomechanical history that is experienced by the polymer melts during processing had an influence on the melting and crystallization behavior of resultant samples.
This work was presented in part at the SPE ANTEC[R]-2018 conference held in Orlando, USA, May 2018. The authors thank the Natural Sciences and Engineering Research Council of Canada (NSERC) and the NSERC Network for Innovative Plastic Materials and Manufacturing Processes for their financial support. SZ acknowledges support from China Scholarship Council.
[1.] M. Heckele and W.K. Schomburg, J. Micromech. Microeng., 14, 1 (2004).
[2.] J. Grabalosa, I. Ferrer, O. Martinez-Romero, A. Elias-Zuniga, X. Planta, and F. Rivillas, J. Mater. Process. Technol., 229, 687 (2016).
[3.] C. Yang, X.H. Yin, and G.M. Cheng, J. Micromech. Microeng., 23, 093001 (2013).
[4.] C. Kukla, H. Loibl, H. Detter, and W. Hannenheim, Kunstst.Plast. Eur., 88, 6 (1988).
[5.] N. Zhang, S.Y. Choi, and M.D. Gilchrist, Macromol. Mater. Eng., 299, 1362 (2014).
[6.] M.R. Kamal, R. El Otmani, A. Derdouri, and J.S. Chu, Int. Polym. Process., 32, 590 (2017).
[7.] S. Zhou, A.N. Hrymak, and M.R. Kamal, Polym. Eng. Sci., 58, E226 (2018).
[8.] P. Mele and J. Giboz, J. Appl. Polym. Sci., 134, 45719 (2017).
[9.] J. Giboz, T. Copponnex, and P. Mele, J. Micromech. Microeng., 17, 96 (2007).
[10.] U.M. Attia, S. Marson, and J.R. Alock, Microfluid. Nanofluidics, 7, 1 (2009).
[11.] J. Chu, M.R. Kamal, S. Derdouri, and A. Hrymak, Polym. Eng. Sci., 50, 1214 (2010).
[12.] S. Zhou, A.N. Hrymak, and M.R. Kamal, Compos. A Appl. Sci., 103, 84 (2017).
[13.] T. Ferreira, M.C. Paiva, and A.J. Pontes, J. Polym. Res., 20, 301 (2013).
[14.] S. Abbasi, P.J. Carreau, and A. Derdouri, Polymer, 51, 922 (2010).
[15.] S. Zhou, A.N. Hrymak, and M.R. Kamal, Polym. Eng. Sci., 56, 1182 (2016).
[16.] C. Pagano, R. Surace, V. Bellantone, F. Baldi, and I. Fassi, J. Compos. Mater., 52, 645 (2018).
[17.] S. Zhou, A.N. Hrymak, and M.R. Kamal, Polym. Adv. Technol, 29, 1753 (2018).
[18.] S. Zhou, A. Hrymak, and M.R. Kamal, J. Appl. Polym. Sci., 134, 45462 (2017).
[19.] S. Abbasi, A. Derdouri, and P.J. Carreau, Polym. Eng. Sci., 51, 992(2011).
[20.] M. Wissler, J. Power Sources, 156, 142 (2006).
[21.] S. Zhou, Y. Chen, H. Zou, and M. Liang, Thermochim. Acta, 566, 84 (2013).
[22.] W. Luo, C. Cheng, S. Zhou, H. Zou, and M. Liang, Iran. Polym. J., 24, 573 (2015).
[23.] H. Wu, C. Lu, W. Zhang, and X. Zhang, Mater. Des., 52, 621 (2013).
[24.] S. Zhou, Y. Lei, H. Zou, and M. Liang, Polym. Compos., 34, 1816(2013).
[25.] S. Zhou, L. Yu, J. Chang, H. Zou, and M. Liang, J. Appl. Polym. Sci., 131, 39596 (2014).
[26.] Y. Rew, A. Baranikumar, A.V. Tamashausky, S. El-Tawil, and P. Park, Construct. Build Mater., 135, 394 (2017).
[27.] M.R. Kamal, J. Chu, S. Derdouri, and A. Hrymak, Plast. Rubber Compos., 39, 332 (2010).
[28.] S. Zhou, A.N. Hrymak, and M.R. Kamal, SPE ANTEC[R] 2018 proceeding, Orlando, FL, paper number: M14_238.
[29.] A. Motaghi, A. Hrymak, and G.H. Motlagh, J. Appl. Polym. Sci., 132,41744 (2015).
[30.] J. Park, K. Eom, O. Kwon, and S. Woo, Microsc. Microanal., 7, 276 (2001).
[31.] S.A. Arvidson, S.A. Khan, and R.E. Gorga, Macromolecules, 43, 2916 (2010).
[32.] Y. Li and H. Shimizu, Polymer, 48, 2203 (2007).
[33.] S. Zhou, A.N. Hrymak, and M.R. Kamal, Nanomaterials, 8, 779 (2018).
[34.] H. Pang, L. Xu, D.X. Yan, and Z.M. Li, Prog. Polym. Sci., 39, 1908 (2014).
[35.] Y. Pan, S. Shi, W. Xu, G. Zheng, K. Dai, C. Liu, J. Chen, and C. Shen, J. Mater. Sci., 49, 1041 (2014).
[36.] H. Wu and L.T. Drzal, Polym. Compos., 34, 2148 (2013).
[37.] Y. Kazemi, A.R. Kakroodi, S. Wang, A. Ameli, T. Filleter, P. Potschke, and C.B. Park, Polymer, 129, 179 (2017).
[38.] Z. Jiang, Y. Chen, and Z. Liu, J. Polym. Res., 21, 451 (2014).
Shengtai Zhou (iD), (1) Andrew N. Hrymak, (2) Musa R. Kamal (iD), (3) Renze Jiang (2)
(1) The State Key Laboratory of Polymer Materials Engineering, Polymer Research Institute of Sichuan University, Chengdu 610065, Sichuan, China
(2) Department of Chemical & Biochemical Engineering, The University of Western Ontario, London, Ontario N6A5B9, Canada
(3) Department of Chemical Engineering, McGill University, Montreal, Quebec H3A0C5, Canada
Correspondence to: A. N. Hrymak: e-mail: firstname.lastname@example.org
Contract grant sponsor: China Scholarship Council, contract grant sponsor: Natural Sciences and Engineering Research Council of Canada, contract grant sponsor: NSERC Network for Innovative Plastic Materials and Manufacturing Processes.
Published online in Wiley Online Library (wileyonlinelibrary.com).
Caption: FIG. 1. The morphology of (a) SG and (b) LTEG particles.
Caption: FIG. 2. (a) Images of the mold insert and a final molding; (b) three-step configurations of as-molded microparts (red arrows indicate the boundary of each section); (c) sampling positions of each section for morphology observation. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 3. The electrical conductivity ([sigma]) for different sections of microparts with respect to the measurement directions, i.e., across the TD and along the FD, respectively, (a) SG30, (b) SG40, (c) SG50, and (d) LTEG30.
Caption: FIG. 4. SEM images of compression-molded (a) SG30 and (b) SG50 composites.
Caption: FIG. 5. SEM images of (a) SG30 thick, (b) SG30 middle, (c) SG50 thick, and (d) SG50 middle.
Caption: FIG. 6. The images of 0.5 g of (a) pristine LTEG and (b) expanded LTEG after thermal shock at 190[degrees]C for 10 min. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 7. SEM images of the thick section of LTEG30 microparts, (a) and (b) Core layer; (c) and (d) shear layer.
Caption: FIG. 8. SEM images of the middle section of LTEG30 microparts. (a) Core layer; (b) shear layer.
Caption: FIG. 9. The heating and cooling curves obtained from PP/G blends and thick section of corresponding microparts, (a) and (d) For Cycle 1; (b) and (e) for Cycle 2; (c) and (f) for Cycle 3. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 10. The melting and crystallization behavior of pure PP, SG30, LTEG30, and each section of corresponding microparts, (a), (d), and (g) For Cycle 1; (b), (e), and (h) for Cycle 2; (c), (f), and (i) for Cycle 3. [Color figure can be viewed at wileyonlinelibrary.com]
TABLE 1. The formulations for preparing PP/graphite composites Sample Component Wt% Vol% SG10 PP/SG 10 4.31 SG20 PP/SG 20 9.20 SG30 PP/SG 30 14.8 SG40 PP/SG 40 21.3 SG50 PP/SG 50 28.8 LTEG30 PP/LTEG 30 14.9 TABLE 2. The characteristic data ([T.sub.m], [T.sub.c] and [chi]) obtained from DSC measurements [T.sub.m] [chi] (%) [T.sub.c] ([degrees]C) ([degrees]C) Designation Cycle 1 Cycle 3 Cycle 1 Cycle 3 Cycle 2 Pure PP 167.39 163.97 49.8 48.8 118.55 SG10 165.39 166.50 49.1 55.4 130.65 SG30 163.95 166.02 48.4 56.3 135.52 SG50 163.72 166.38 47.9 57.6 139.15 PP thick 166.56 163.83 45.9 48.3 118.49 PP middle 165.75 162.21 45.8 48.8 118.54 PP thin 166.25 163.13 42.6 46.8 117.96 SG 10 thick 166.61 166.68 49.2 54.9 130.58 SG30 thick 166.04 166.44 48.4 56.3 135.58 SG30 middle 165.19 165.81 44.9 53.6 135.54 SG30 thin 167.34 166.08 45.0 54.5 135.47 SG50 thick 164.37 166.76 47.0 56.7 139.11 LTEG30 160.24 159.40 48.3 49.5 132.09 LTEG30 thick 159.64 159.08 46.1 51.2 132.64 LTEG30 middle 159.49 158.61 43.5 50.2 132.69 LTEG30 thin 159.59 158.58 42.3 48.5 132.66
|Printer friendly Cite/link Email Feedback|
|Author:||Zhou, Shengtai; Hrymak, Andrew N.; Kamal, Musa R.; Jiang, Renze|
|Publication:||Polymer Engineering and Science|
|Article Type:||Case study|
|Date:||Aug 1, 2019|
|Previous Article:||Influence of Screw Configuration and Processing Temperature on the Properties of Short Glass Fiber Reinforced Polypropylene Composites.|
|Next Article:||Effect of Interfacial Chemistry on Crystallization of Polypropylene/Multiwall Carbon Nanotube Nanocomposites.|