Extrinsic toughening of recycled carbon fibers in polypropylene composites in the absence of plasticity penalty

Materials

A commercially available grade of PP homopolymer (Adstif HA801U from LyondellBasell, Houston, USA) with flexural modulus of ∼2100 MPa was used in this study. The melt flow rate of Adstif HA801U was measured as 64 dg min^-1, suitable for high-speed injection molding of thin-walled parts. A maleic anhydride–grafted PP, PP-g-MA (Polybond 3200, Crompton Corporation, Connecticut, USA), with a maleic anhydride grafting level of 2.0 wt% and the weight-average molecular weight of 84,000 g mol^-1 was used as coupling agent at a concentration of 3 wt%. 40-mm chopped recycled carbon fibers with fiber diameter of ca. 5 µm (Recafil® from SGL Carbon Company, Germany), tensile strength of 4.4 GPa, and 1% sizing amount on the surface of the fibers (epoxy resin compatible) were introduced into the matrix at a loading of 20 wt%. 2 wt% of an antioxidant (Cyanox 2777 supplied by Univar) and 1 wt% of a carbon black masterbatch (PP-940 from Modern Dispersions, Leominster, USA) were also introduced to the compounds. PP-20CF and PP-20CF-3MA are nomenclatures used to denote PP composites containing 20 wt% recycled carbon fibers without and with 3 wt% PP-g-MA, respectively.

Sample preparation

The carbon fiber–reinforced composites were produced by melt compounding. PP, antioxidant, colorant, and compatibilizer were introduced through the main feeder, and carbon fibers were fed through a side feeder after the only mixing zone. The samples were compounded using a co-rotating twin-screw extruder (D = 27 mm and L/D = 36) at 100 rpm and a temperature profile of 180°C–230°C from the hopper to the die. Screw geometry with one mixing zone, comprising elliptical kneading elements with staggering angles of 30° and 60°, was used to minimize fiber attrition. The extrudates were dried for 24 h in a vacuum oven prior to molding test bars using a Battenfeld injection molding machine at 230°C.

Characterization

A Tinius Olsen machine equipped with a tensometer was used to evaluate tensile and flexural characteristics of the samples at a rate of 50 mm min^-1 and 2 mm min^-1, according to ISO 527 and ISO 178, respectively. The notched Charpy impact strength at 23°C was determined according to ISO 179 using a pendulum impact tester (Testing Machines Inc.). Between five to ten samples were tested for evaluating the mechanical properties. Melt flow rate of the samples were measured using MP600 (Tinius Olsen) according to ISO 1133 at 230°C using a weight of 2.16 kg. Heat deflection temperature (HDT) characterization was conducted under 1.82 MPa bending load according to ISO 75-1 by using 603 HDTM instrument (Tinius Olsen). The density of the composites was calculated based on ASTM D792 by incorporating the Archimedes method by means of a Mettler A1201 density measurement device.

The state of fiber dispersion in various samples was examined by scanning electron microscopy (SEM) using a Hitachi S4700 microscope operated at a voltage of 10 kV. The samples containing carbon fibers were cryogenically fractured using liquid nitrogen followed by sputter, depositing a thin layer (∼4 nm) of gold on their surfaces.

Differential scanning calorimetry (DSC) of the samples was performed on a DSCQ1000 (TA Instruments) under a helium atmosphere. The samples were heated from room temperature to 200°C and held at that temperature for 3 min to eliminate thermal history. They were then cooled to 20°C and heated again to 200°C at a constant rate of 10°C min^-1. An empty pan run and a calibration run with sapphire under identical conditions were performed to obtain the heat capacity, C _p , curves for the neat sample and the composites containing 20 wt% carbon fibers.

Dynamic mechanical analysis (DMA) was performed from ambient temperature to 140°C at 3°C min^-1 using Q800 TA Instruments. The storage modulus was examined at a fixed frequency of 1 Hz in the linear viscoelastic regime by using a dual cantilever bend fixture of 35 mm length.

Results and Discussion

The morphology of the composites at the surface, cross section, and fiber/matrix interface was characterized by SEM. Figure 1 presents the surface morphology of the composite samples. Irrespective of the high concentration of the recycled carbon fibers used for reinforcing the composites, they are homogenously distributed at the surface of the samples. The surface morphology reveals the tendency of the fibers to orient along the flow direction as a result of the strong shear stresses during the injection molding. It could also be observed that the fibers are to some degree wetted with the matrix in the compatibilized composites (Figure 1(d)), whereas the fibers in the non-compatibilized samples seem to be remained uncovered (Figure 1(b)).OPEN IN VIEWER

The fractured cross-section of the composites, observed from different perspectives, illustrates that the recycled carbon fibers are generally aligned along the flow direction within the composites, analogous to their uniaxial orientation at the surface (Figure 2). The matrix of both composites is fractured in a brittle failure mode, and a remarkable population of broken carbon fibers is evident. No indication of stretched strips of the matrix at the smooth fracture zone of the composites is observed, which is consistent with the tensile strain characteristics of the composites. The SEM micrographs exhibit that the fibers are uniformly distributed and individually dispersed within the matrix and no trace of fiber agglomeration (e.g., fiber bundles) is observed in the reinforced composites. Fu et al. ³⁰ studied PP composites reinforced with short carbon fibers and short glass fibers prepared by extrusion compounding followed by injection molding techniques. The preferential orientation of both glass fibers and carbon fibers in the fellow direction were confirmed by SEM micrographs for different samples.OPEN IN VIEWER

Figure 3 shows the level of interfacial bonding between the fibers and matrices of the composites. While the uncompatibilized composites feature a clean fiber surface, the strong interfacial bonding led to carbon fibers coated with some residual polymer for those containing the coupling agent. A portion of the interfacial shear stress transfer between the matrices and fibers is due to the mechanical friction force on the interface exerted by the thermal shrinkage of PP. ²⁶ This shrinkage-induced friction force at the interface of both types of composites is large enough to reinforce PP and transfer part of the applied load from matrix to the fibers. The SEM micrographs demonstrate that cohesive failure of the matrix close to the interface caused interfacial microfailure in the compatibilized samples, rather than fiber/matrix debonding. In other words, polymeric layers link the recycled carbon fibers to the matrix in the compatibilized composites, enhancing the interphase strength and assuring that the fibers will support the matrix. The verified adhesion at the interface of the matrix and fibers illustrates that potential interactions may exist between the maleic anhydride groups of the compatibilizer and some functional groups on the surface of the recycled fibers.^13,31 The potential functional groups may have formed during the manufacturing process or remained from the original epoxy matrix reinforced by the virgin fibers. The outcome of enhanced affinity between the components of the compatibilized composites is expected to reflect in their thermal and mechanical characteristics.OPEN IN VIEWER

Mechanical properties of the recycled carbon fiber–reinforced composites are presented in Figure 4. The tensile strength was improved by 35% and 141% for the reinforced composites without and with the compatibilizer, respectively (Figure 4(a)). The results unfold that the reinforcement potential of the fibers is greatly utilized in the compatibilized composites featuring efficient stress transfer across the fiber–matrix interface. Rezaei et al. ²⁷ developed PP composites containing 10 wt% carbon fibers and compared their specific mechanical properties with those of commercial steel used in passenger vehicles’ hood. Specific strengths of 35,780 and 53,060 N m kg^-1 were reported for the commercial steel bonnet (with a density of 7.426 g cm^-3) and the composites, respectively. In this study, the developed composites without and with the compatibilizer exhibited specific strengths of 46,519 and 82,748 N m kg^-1, respectively, at a measured density of ∼0.997 g cm^-3 using the Archimedes method. While the reported value for the tensile strength of the commercial steel bonnet (265.7 MPa) is higher than that obtained for the developed composites in this study (i.e., 46.4 and 82.5 MPa), the specific strength of the developed composites is higher than that of the steel hood. The elongation at break of the stiff PP homopolymer used as the matrix is relatively low (∼3%); therefore, the introduction of the nearly non-deformable carbon fibers has no pronounced effect on deteriorating failure strain (Figure 4(b)).OPEN IN VIEWER

The ability of composites to tolerate bending deformations is critical in structural parts supporting an applied load. A three-point bending configuration was used to measure stiffness of the fiber-reinforced composites (Figure 4(c)). The introduction of recycled carbon fibers led to ca. 430% and 470% increase in the bending modulus in uncompatibilized and compatibilized composites, respectively. It appears that the interface modification plays a more crucial role in tensile strength than stiffness (flexural modulus), which is evaluated at small bending deformations. That is, the interfacial frictional sliding between the fibers and the matrix dissipates a considerable amount of energy; hence, the quality of the interface in stress transfer between the matrix and fibers plays a significant role in stress at break. On the other hand, the flexural modulus is evaluated at very low deformations; hence, its sensitivity to the fiber–matrix interface is marginal.

The notched Charpy impact strength of the brittle matrix increased from 1.1 to 4.7 and 5.2 kJ m^-2 for the composites without and with the coupling agent, respectively (Figure 4(d)). The introduction of carbon fibers led to significant enhancement of impact strength, particularly in the compatibilized samples, through crack shielding mechanisms highlighted in Figure 5.OPEN IN VIEWER

Figure 5 exhibits the extrinsic energy dissipating mechanisms that operate during the fracture of the recycled carbon fiber–reinforced composites. The extrinsic toughening mechanisms, such as fiber bridging, fiber pull-out, and fiber–matrix debonding, are denoted as A, B, and C, respectively, in the SEM micrograph of Figure 5(a) and schematically depicted in Figure 5(b). In this study, a stiff and strong matrix with low impact strength was used; hence, the fiber-reinforced composites demonstrate improved performance in fracture toughness. In contrast, if a tough matrix (e.g., a ductile TPO compound) is formulated, its toughness may reduce upon the incorporation of the discontinuous fibers.^6,32 In the case of a tough TPO matrix, part of the fracture energy is dissipated by elongation and cavitation of the elastomer particles, formation of plastic zones, shear yielding, and matrix deformation.^33,34 Therefore, it is reasonable to expect a reduction in fracture energy by incorporating fibers into a TPO matrix. In other words, replacing a portion of the plastic zones by a certain volume fraction of brittle short fibers, which do not significantly contribute to the energy dissipated by deformation in the composites, has an adverse effect on impact strength of the matrix. Besides, the presence of the fibers may inhibit deformation mechanisms and confine plastic zones to the interfibrillar spacing ahead of the crack tip. Consequently, the reduction in impact energy corresponds not only to the reduction of plastic zone size but also to the changes in the strain to failure of the highly restricted plastic zones between the fibers, associated with interfacial interactions.^32,35–39

In fiber-reinforced polymer composites, the reduction of matrix plasticity (plasticity penalty), induced by the presence of stiff and brittle fibers, needs to be compensated by extrinsic toughness-improving mechanisms of the fibers. In relatively tough matrices reinforced with fibers, the plasticity penalty due to toughness reduction of the matrix might be more than offset by the fiber-shielding toughening; thus, the ultimate impact strength reduces. In contrast, the plasticity penalty associated with fiber reinforcement of a brittle matrix is marginal; hence, extrinsic toughness plays the dominant role and results in impact strength enhancement. Accordingly, the influence of short fibers on impact strength of polymers depends on the trade-off between plasticity penalty and extrinsic toughening, which is associated with the initial toughness of the matrix. In our previous study, a tough TPO matrix was formulated with a notched impact strength of 14 kJ m^-2.³² The notched impact strength of the TPO reduced to ca. 5 kJ m^-2 upon incorporation of 20 wt% carbon fibers, confirming dramatic reduction of matrix plasticity. In contrast, the matrix used in this study is a highly crystalline PP homopolymer with low impact strength; hence, the plasticity penalty associated with fiber reinforcement is trivial. Therefore, incorporation of the fibers led to ∼373% enhancement in impact strength. Interestingly, the developed composites in this study, based on a brittle matrix, exhibit the same value of notched impact strength as tough TPO samples reinforced with the same loading of the recycled carbon fibers. In this study, no sign of intrinsic toughening mechanism (plasticity) induced by the matrix was observed for the composites, which is in good agreement with their marginal strain at the break.

The dynamic storage modulus (E^′) of the neat and reinforced composites is shown in Figure 6 as a function of temperature. The reinforced composites exhibit a significantly higher storage modulus compared to the neat samples, with a marginal difference between those with and without the compatibilizer, in agreement with the flexural modulus trend observed in Figure 4. The load-bearing capability of the carbon fiber–reinforced composites at high temperatures, even up to ca. 130°C, is still higher than that of the neat homopolymer at ambient temperature. The dynamic storage modulus of the composites is ca. 3 times and 7 times higher than that of the neat samples at ambient temperature and maximum testing temperature, respectively. In other words, the role of the recycled carbon fibers in load-bearing becomes more significant at higher temperatures, where the matrix loses its strength.OPEN IN VIEWER

Thermal characteristics of the neat samples and composites are illustrated in Figure 7 and summarized in Table 1. The melting temperature tends to remain constant regardless of the presence of carbon fibers and compatibilizers. The hot crystallization temperature, T _hc , of the homopolymer reduces with the introduction of 20 wt% carbon fibers. It appears that the presence of carbon fibers at a high concentration restricts the migration and diffusion of polymer macromolecules to the surface of the nuclei, which is more pronounced in the compatibilized composites. Several factors affect nucleation and growth of crystals in polymers and composites. For instance, at high fiber concentrations, the heterogenous nucleating effect of fibers on T _hc is suppressed by macromolecular mobility hindrance.^40–46 The degree of crystallinity, χ c , of the samples was calculated from the melting enthalpy, ΔH _m ; matrix weight fraction, w_m; and the enthalpy of fusion of a pure crystalline PP, ΔH ^ο ; equal to 207 J g^-1, ⁴⁷ according to equation (1). As no trace of cold crystallization was detected, the melting enthalpy of the samples is equal to their corresponding hot crystallization enthalpy, ΔH _hc

χ c = Δ H m w m × Δ H ° .

(1)

OPEN IN VIEWER

Figure 7.

Thermal properties of the neat samples and reinforced composites containing 20 wt% recycled carbon fibers.

OPEN IN VIEWER

Table 1.

DSC analysis results of the neat samples and reinforced composites containing 20 wt% recycled carbon fibers.

Sample name	T hc (°C)	Δ H hc (J g-1)	T m (°C)	Δ H m (J g-1)	χ c (%)
Neat PP	132.1	113.1	166.9	113.8	55.0
PP-20CF	128.5	91.0	166.7	91.0	55.0
PP-20CF-3MA	122.7	85.7	167.1	85.8	51.8

The heat capacity of the neat and reinforced composites (Figure 7(b)) increases linearly in the measured temperature range as C _p is higher when the macromolecules are free to move. The introduction of carbon fibers, with a lower heat capacity of 0.92 J g^-1 K^-1, ⁴⁸ reduces the specific heat capacity of the composites, which is more substantial in the compatibilized sample. Kada et al. ¹⁸ also reported lower values of specific heat capacity for the PP composites containing carbon fibers. A monotonic reduction of specific heat capacity was observed with carbon fiber concentration.

The MFI and HDT of the neat matrix and reinforced composites are shown in Figures 8(a) and (b), respectively. The MFI of the matrix reduced from 64 to ∼8 dg min^-1 by incorporating carbon fibers. Large parts with complex geometries made of viscous compounds are prone to processing issues, such as warpage, if processing parameters are not tuned properly. To maximize the ease of injection molding (processability), the composites were formulated using a PP homopolymer with a relatively high MFI. The higher viscosity of the composites is attributed to the tendency of the fibers to entangle and form networks and perturbing the flow of the polymer chains. Additionally, the substitution of the matrix macromolecules with rigid carbon fibers reduces the matrix portion of the composites, which is the only source of flowability in the system. Furthermore, the strong interactions between the carbon fibers and the matrix restrict the molecular mobility of the polymer chains in the interfaces, thus increasing viscosity of the composites.OPEN IN VIEWER

The HDT measurements were performed to assess the maximum temperature at which the developed composites can be used as a rigid material. The HDT of the neat samples increased from an average value of 72°C–147°C and 151°C for the composites without and with the compatibilizer, respectively. These remarkable enhancements in HDT is ascribed to the improvement of modulus, reduction of high-temperature creep, and restricted relaxation of macromolecules in the presence of fibers at high temperatures.^49–53 In an earlier study, we observed that the introduction of 20 wt% short carbon fibers to a thermoplastic polyolefin matrix increased the HDT by ∼100°C, verifying the remarkable effect of the fibers in improving the HDT of the matrix. ³² The introduction of 20 wt% glass fibers and carbon fibers to a polyamide-6 matrix improved the HDT of the matrix from ∼65°C to 125°C and 135°C, respectively, confirming the higher efficiency of carbon fibers in enhancing the HDT properties of the matrix. ⁵⁴

Improved interfacial adhesion between the fibers and matrix induced by the compatibilizer led to a higher flexural modulus and a more efficient stress transfer across the fiber–matrix interface. The improved bonding between the fibers and matrix also led to a higher restriction of molecular mobility in the interphase, which was reflected in thermal characterizations as a lower hot crystallization temperature and a higher HDT in the compatibilized composites.