Graphene induced microstructural changes of PLA/MWCNT biodegradable nanocomposites: rheological, morphological, thermal and electrical properties

Abstract

Two types of carbon nanofillers, i.e. multiwalled carbon nanotubes (MWCNTs) and graphene nanoplatelets (GnPs), were chemically functionalized and incorporated into poly(lactic acid) (PLA) through solution mixing to prepare single filler and hybrid PLA nanocomposites. The functionalization of the nanofillers was characterized in detail using FTIR and TGA analysis. X-ray diffraction (XRD) showed that functionalized GnPs (fGnPs) were fully exfoliated in the nanocomposites. The dispersion state of the single and hybrid nanofillers in the polymer matrix was studied in detail by means of melt viscoelastic experiments together with electron microscopy results, which revealed full exfoliation of fGnPs and high dispersion of fCNTs. It was shown that the incorporation of fCNTs–fGnP hybrids into PLA created a large surface area which established strong interfacial adhesion between the efficient hybrid nanofiller networks and the matrix. Positive deviation was observed from the mixture law in the electrical, thermal and mechanical properties. Thorough analysis of the results showed that formation of efficient hybrid networks enhanced the mechanical properties of hybrid filler nanocomposites through increasing effective stress transfer, retardation of flaw formation, emerging microcrack growth and dissipation of additional mechanical energy.

---

Introduction

The past few decades have seen a growing demand for biodegradable polymeric materials as a viable solution for creating a proper balance between the increasing consumption of plastics from non-renewable resources and the limited availability of landfills and raw materials. Being one of the few biodegradable synthetic polyesters, poly(lactic acid) (PLA) is often perceived as one the most appealing candidates to meet this urgent need. The promise of this polymer comes partly from the fact that it is derived from renewable resources. Currently, it is produced either by ring opening polymerization of lactide or through polycondensation of lactic acid monomers, which are obtained from the fermentation of corn or other renewable raw agricultural resources.^1–4

Among other remarkable features that have created wide applications for this polymer are outstanding processability, high strength and modulus, and low carbon footprint. As with any other polymer, however, there are a number of drawbacks which place a limit on its use in industrial applications such as packaging. These include poor mechanical and thermal resistance and inadequate gas barrier properties. In order to enhance the thermomechanical resistance of PLA, a number of techniques such as copolymerization, blending, and reinforcing have been employed.^5–7 The last one has recently focused on improving the properties of PLA by incorporation of nanoparticles such as clay, carbon nanofillers, etc. to achieve high performance biodegradable polymer nanocomposites.^8–14

Polymer nanocomposites containing carbon nanofillers such as one dimensional carbon nanotubes (1D CNTs) and two dimensional graphene nanoplatelets (2D GnPs) have recently drawn considerable interest, due to their superior mechanical and gas barrier properties, dimensional stability, and electrical conductivity. In general, these properties depend on the intrinsic characteristics of the carbon nanofillers including their morphology (dimension, aspect ratio, alignment, etc.) as well as their dispersion state within the polymeric matrix.^15–18

The interfacial interaction between the nanofillers and the polymer matrix is the most important contributing factor in determining the state of nanofiller dispersion. CNTs easily tend to entangle and agglomerate due to their size and high aspect ratio, while GnPs tend to restack due to their strong van der Waals and strong π–π interactions. This leads to the relatively poor interaction between the nanofillers and the polymeric matrix. Different dispersion methods as well as nanofiller surface functionalization through oxygen plasma treatment, oxidation with acid, and reduction with base and the use of a variety of dispersing agents have been developed to improve their dispersion and compatibility in the polymer matrix. Good dispersion of the CNTs and GnPs leading to the formation of efficient three dimensional networks is the key to exploiting their extraordinary potential in advanced materials.^19–21

Over the last decade, increasing activities have been directed towards using hybrid filler systems as one of the successful methods of developing advanced polymeric nanocomposites with improved and, in many cases, synergistic properties. Previous works have shown that hybrid filler nanocomposites containing a mixture of carbon nanofillers with different geometries show a remarkable improvement effect in mechanical, electrical and thermal properties.^22–31 In other words, preparation of the CNTs–GnPs/polymer hybrid system is an alternative means to obtain cost-effective nanocomposites with balanced properties. Yue et al.,²⁵ for instance, have observed that a combination of MWCNTs and GnPs in 8 : 2 ratio synergistically increase flexural properties and reduce the electrical percolation threshold for the epoxy composites. They demonstrated that small amounts of graphene platelets can act as a dispersing agent for MWCNTs and improve their dispersion. Al-Saleh²⁶ has reported that at constant overall nanofiller concentration, GnPs–MWCNTs/PP hybrids show more effective dispersion of nanotubes and stronger adhesion at the MWCNTs/PP interface. Poosala et al.²⁷ have prepared GnPs–MWCNTs/PC nanocomposites to investigate their electrostatic dissipative application. They proposed that surface modification of GnPs, for instance by oxidation with acid, is required to improve the dispersion and intercalation of GnPs in the PC matrix. Li et al.²⁸ have reported that incorporation of MWCNTs–GnPs hybrids into epoxy resulted in optimum dispersion of MWCNTs and GnPs as well as better interfacial adhesion, which in turn enhances load transfer effectiveness. They also showed that remarkably enhanced mechanical properties were achieved at ultra-low hybrid loadings (0.5 wt%). Chatterjee et al.²⁹ have used two types of GnPs with different dimension. They reported that the high aspect ratio of CNTs combined with the larger surface area of GnPs is responsible for the synergistic effect of the hybrid samples. Im et al.³⁰ investigated thermal conductivity of a GO–MWCNTs/epoxy hybrid composite and showed that the increased thermal conductivity is due to the formation of three dimensional heat conduction paths created as a result of MWCNTs addition. Yang et al.³¹ have shown that stacking of individual multi-graphene platelets (MGPs) is effectively inhibited by introducing MWCNTs, resulting in a remarkable synergetic effect in the mechanical properties and thermal conductivity. They demonstrated that long and tortuous MWCNTs can bridge adjacent MGPs and inhibit their aggregation.

Therefore, CNTs–GnPs hybrids with uniform dispersion of both nanofillers (CNTs and GnPs) hold great promise as high performance composite systems to achieve satisfactory enhancement efficiency. Two plausible explanations have been proposed for the improved dispersion state in CNTs–GnPs/polymer hybrid nanocomposites. The first one suggests that with multilayer GnPs, long and tortuous CNTs can bridge the distances between multilayer GnPs and inhibit their restacking/reaggregation. The second one assumes that exfoliated GnPs inhibit reaggregation of CNTs, which improves their dispersion. Therefore, more work is required to gain a deeper insight into the mechanisms through which dispersion of nanofillers in hybrid filler nanocomposites is improved.

The objective of the present study was to investigate the enhancement of electrical, thermal and mechanical properties of PLA nanocomposites upon addition 1D MWCNTs and 2D GnPs as single fillers and in the form of hybrid fillers. First, surface functionalization was performed through oxidation using a mixture of acids (H₂SO₄/HNO₃) to facilitate the dispersion of the carbon nanofillers and increase their compatibility with the PLA matrix. Then, exfoliated fGnPs were used to inhibit reaggregation of fCNTs and improve their dispersion. Their microstructural development was also assessed to cast further light on the improvement of properties. The tree dimensional networks of fCNTs–fGnPs hybrid nanofillers exhibit the greatest efficiency in the enhancement of the properties and showed a certain degree of positive deviation from the properties predicted by the mixture law.

Experimental

Materials

The PLA matrix used in this study was NatureWorks 6350D with a melt flow index of 65 g/10 min supplied by NatureWorks LLC, USA. The carbon nanotubes were commercially available as multiwalled carbon nanotubes (MWNTs), NC-7000, from Nanocyl Inc., Belgium, which had a carbon purity of 90%, average outer diameter of 9.5 nm, lengths up to 1.5 μm (average aspect ratio of about 158), and surface area of 250–300 m² g⁻¹. The graphene nanoplatelets (GnPs), N002-PDR, from Angstron Materials, USA, have a thickness < 1 nm, average diameter < 10 μm (average aspect ratio of about 4000) and surface area of 400–800 m² g⁻¹ are comprised of stacks of 1–3 monolayer graphene sheets. SEM images of the pristine MWCNTs and GNPs powders used in this work are shown in Fig. 1. Chloroform (CHCl₃, 99.8%), sulfuric acid (H₂SO₄, 99.9% v/v) and nitric acid (HNO₃, 65% v/v) were purchased from Merck chem. Co., Germany, and used without further purification.

Fig. 1 SEM micrographs of the pristine (a) GnPs and (b) MWCNTs.

Preparation of chemically functionalized carbon nanofillers

Both carbon nanofillers (MWCNTs and GnPs) were chemically functionalized with carboxyl groups through oxidation by an acid mixture to enhance their adhesion to the PLA matrix. The nanofillers were first dispersed in a 3 : 1 (v/v) mixture of concentrated H₂SO₄ and HNO₃ at 80 °C for 4 h. After 2 h of sonication at the same temperature, the nanofillers suspension was stirred for 12 h at 60 °C. The obtained suspension was then diluted with plenty of deionized water until there was no trace of acid (pH ∼ 7) in the bulk of the nanofillers. Finally, the carboxylated carbon nanofillers were dried at 80 °C for 24 h in a vacuum oven.^32,33 Functionalized carbon nanofillers will be referred to as fCNTs and fGnPs in the following.

Preparation of nanocomposites

PLA film and its nanocomposites were prepared via solution casting using chloroform as the solvent. PLA, pristine and functionalized carbon nanofillers were dried in a vacuum oven at 80 °C for 12 h before use. PLA solution was prepared by step-wise addition of 20 g of PLA in 200 mL of chloroform while agitating vigorously in laboratory capped containers and at ambient temperature (20 °C) for 2 days. The solution was cast onto glass molds. For the preparation of single and hybrid filler filled PLA nanocomposites, predetermined amounts of carbon nanofillers were added to chloroform (200 mL) and stirred in laboratory capped containers for 1 day at ambient temperature using a magnetic stirrer. These suspensions were then homogenized at 8000 rpm for 15 min in a high shear homogenizer (IKA T25-digital ultra turrax, Germany), followed by sonication in a high intensity ultrasonic processor (Haver & Boecker, Germany, 35 kHz, 400 W) for 15 min at ambient temperature. The suspensions were mixed with the previously prepared PLA solution and then stirred for 6 h with a magnetic stirrer. The solutions were homogenized at 8000 rpm for 15 min followed by another round of sonication for 15 min before they were cast onto glass molds. After drying at ambient temperature for 1 day, PLA and its nanocomposite cast films were further dried in a vacuum oven at 80 °C for 12 h to remove the entrapped solvent.^34,35 The thickness of the obtained films was about 0.5 mm. For each single filler nanocomposite, PLA was mixed with 1 wt% of the pristine and functionalized carbon nanofiller. For hybrid filler nanocomposites, different ratios of fGnPs to fCNTs, i.e. 3 : 1, 2 : 2, and 1 : 3, were chosen, with the total concentration of the nanofillers fixed at 1 wt%. In the following, they will be referred to as 0.75% fGnPs–0.25% fCNTs (3 : 1), 0.5% fGnPs–0.5% fCNTs (2 : 2) and 0.25% fGnPs–0.75% fCNTs (1 : 3).

Characterization methods

Fourier transform infrared (FTIR) spectra were recorded on a spectrum GX-PerkinElmer (USA) in the wave number range of 4000–500 cm⁻¹ with a resolution of 4 cm⁻¹. 5 scans were performed for each sample. The FTIR spectra of pristine and acid modified nanofillers were obtained in transmittance mode by mixing a small amount of the materials with KBR powders.

X-ray diffraction (XRD) was performed at room temperature using an X-ray diffractometer (Philips model X'Pert, Netherlands) over an angular range (2θ) of 5–40° at a rate of 1° min⁻¹ and a step size of 0.02. The X-ray beam was a Co Kα radiation (λ = 1.78897 Å) using a 40 kV voltage generator and a 30 mA current.

The oscillatory shear rheological measurements were carried out at 190 °C with an MCR 301 rheometer (Physica Anton Paar, Austria) using parallel plate geometry with a diameter of 25 mm and a gap of 1 mm. The dynamic strain amplitude sweep experiments were performed in order to determine the linear viscoelastic region by monitoring storage modulus.

The morphology of the samples was examined by using scanning electron microscopy (SEM), with an AIS-2100 from Seron Co. (South Korea) at 15 kV. The cryogenically fractured surface of the samples was coated with gold using a sputter coater under argon gas atmosphere.

The electromagnetic interference shielding efficiency (EMI SE) of the PLA and its nanocomposite samples was measured using HP 8410C Network Analyzer (Japan) in the frequency range of 5–12 GHz; the sample thickness was ∼0.5 mm.

Tensile experiments were performed according to ASTM D882 using a universal testing machine (Zwick/Roell, Germany) at room temperature. For each sample, results from three measurements were averaged.

Dynamic mechanical thermal analyses were performed at a frequency of 1 Hz and a heating rate of 2 °C min⁻¹ using a PerkinElmer Pyris Diamond DMA dynamic mechanical analyzer (USA). Thermogravimetric analysis curves were recorded with a PerkinElmer Pyris Diamond TG/DTA Thermogravimetric/Differential Thermal Analyzer (USA). The samples were heated from 50 to 500 °C at a heating rate of 10 °C min⁻¹ under nitrogen atmosphere.

Results and discussion

Characterization of functionalized carbon nanofillers

Fig. 2 shows the FTIR spectra, SEM images, TGA thermograms of the carbon nanofillers and their possible interactions after functionalization. As it was expected, the FTIR spectra of the carbon nanofillers show similar peaks due to their structural resemblance. One of these peaks is the one at 1550 cm⁻¹, corresponding to the stretching of C=C bond (see Fig. 2a and b). The peaks between 2800 and 3000 cm⁻¹, corresponding to the stretching vibration of C–H bonds, are also common between the two spectra. It can also be seen that the FTIR spectrum of the pristine MWCNTs showed two set of peaks at ∼1200, and ∼3400 cm⁻¹ corresponding to the stretching bands of C–C–O, and O–H, respectively. In functionalized sample, similar characteristic bands appeared but with significantly higher intensities. This could possibly be an indication of increase in the number of attached OH groups (to the surface of the carbon nanofillers) caused by the functionalization process. The spectrum of the pristine GnPs showed peaks at ∼1100, ∼1600, and ∼3450 cm⁻¹, corresponding to C–O, C=O, and O–H stretching bands, respectively. The functionalized sample had these characteristic bands, although with different intensities (in proportion to the degree of modification). Moreover, functionalized carbon nanofillers show new peaks between 1650 and 1750 cm⁻¹ (C=O stretching band), which originate from the carboxylic acid groups attached covalently to the surface of nanofillers.

Fig. 2 FTIR spectra of pristine and functionalized (a) GnPs and (b) MWCNTs, SEM images of (c) fGnPs and (d) fCNTs, (e) TGA curves of pristine and functionalized carbon nanofillers, and (f) possible interactions between the functionalized nanofillers and PLA.

In order to evaluate the aggregation/stacking state of the carbon nanofillers during the functionalization process, their surface morphology is shown in Fig. 2c and d. In comparison with Fig. 1, neither the bundles of fCNTs nor fGnPs layers formed extensively aggregated structures after the functionalization process. Moreover, there is no perceptible damage, including shortening in the length of the MWCNTs bundles or crushing the graphene layers.

According to TGA results shown in Fig. 2e, the residue weight at 450 °C was 97.53 wt% for pristine MWCNTs and 93.46 wt% for fCNTs, while this figure was 90.29 wt% and 85.31 wt% for pristine GnPs and fGnPs, respectively. As it can be observed, loss of weight in the functionalized samples is more significant and occurs at much lower temperatures compared to the pristine samples, suggesting that the oxygen containing functional groups had mostly been attached to the surface of the carbon nanofillers. Similar results were reported by Yoon et al.³⁶ The incorporation of functionalized carbon nanofillers into PLA can enhance interfacial adhesion between the nanofillers and the PLA matrix through polar and hydrogen bonding formed between carboxylic acid groups and the lactide group of PLA (Fig. 2f).

XRD results

X-ray diffraction analysis was used to evaluate the exfoliation state of the graphene nanoplatelets and also to determine the crystallographic properties of the PLA nanocomposites. Fig. 3 shows the XRD patterns of the pristine and functionalized GnPs, neat PLA, and single and hybrid nanofillers filled PLA. As it can be seen, the pristine and functionalized GnPs showed no diffraction peak around 2θ ∼ 26°, indicating that they were already in a completely exfoliated state (individual layers). This suggests that the functionalization process did not affect the interlayer spacing of GnPs and their structure remained in relatively the same exfoliated state. The crystalline region of both unfilled and nanofiller filled PLA, displayed a typical sharp diffraction peak pattern, and the amorphous region showed a broad scattered area. This figure also shows that while neat PLA has X-ray diffraction peaks at 2θ = 16.6° and 19°, the nanocomposites containing 1 wt% of pristine GnPs show a small diffraction peak at 2θ = 26.4°, which suggests some restacking of pristine GnPs due to their van der Waals and strong π–π interactions. However, the single filler and hybrid PLA nanocomposites containing fGnPs show only the diffraction peaks of neat PLA with no diffraction peak around 2θ = 26°, which indicate that the exfoliated layers of fGnPs did not restack during the mixing process. This can be attributed to strong interfacial interactions between PLA chains and exfoliated fGnPs.³⁵

Fig. 3 XRD patterns of pristine and functionalized GnPs, neat PLA, and its nanocomposite samples.

Rheological measurements

Determination of the linear viscoelastic region. In order to differentiate the linear and nonlinear viscoelastic regions and investigate dispersion stability of carbon nanofillers, strain amplitude sweep experiments were performed at the controlled frequency of 1 rad s⁻¹ in the overall range of 0.05–1000%. The normalized storage modulus versus strain amplitude for PLA matrix and its single filler nanocomposites is shown in Fig. 4. It can be seen that the transition from linear to nonlinear viscoelastic behavior in the nanocomposites containing functionalized nanofillers occurs at smaller strains compared to those with pristine nanofillers. This can be related to changes in microstructure induced by carbon nanofillers and the breakup of some of the crucial elastic links in the nanofillers three dimensional networks, whose existence is more predominant in the case of functionalized nanofillers. Moreover, normalized storage modulus in the PLA nanocomposite containing 1 wt% of fCNTs decreases rapidly, suggesting that the three dimensional network structures were stronger with fCNTs than with fGnPs.^37,38 Considering these results, further linear rheological measurements were all performed in the linear viscoelastic region (with an amplitude of 1%).

Fig. 4 Normalized storage modulus versus strain amplitude for neat PLA and its single filler and hybrid nanocomposites at 190 °C.

Viscoelastic behavior of nanocomposites containing pristine and functionalized carbon nanofillers. In order to further investigate the superiority of functionalized carbon nanofillers, key viscoelastic properties were measured in small amplitude oscillations as a function of frequency and time. Fig. 5a plots the values of storage modulus versus angular frequency (ω) for neat PLA and its nanocomposite samples containing 1% of pristine and functionalized carbon nanofillers. At low frequencies, the neat PLA sample exhibits the classic terminal viscoelastic behavior, with scaling of G′ ∼ ω². The nanocomposites containing pristine carbon nanofillers show a small deviation from the classic behavior in storage modulus at lower frequencies. However, the oscillatory behavior of the functionalized carbon nanofillers based nanocomposites is distinctly different. For these samples, storage modulus shows a nonterminal plateau or solid like behavior at low frequencies and has higher values compared with those of neat PLA and pristine carbon nanofillers filled PLA. These results suggest a relatively fine dispersion of nanofillers as a consequence of the strong interfacial interaction between the PLA matrix and the functionalized nanofillers compared to the unfunctionalized ones. At high frequencies, the values of storage modulus for all samples are similar to neat PLA, indicating that segmental motion of the PLA matrix determines the material response at short time scales.

Fig. 5 (a) Storage modulus, and (b) relaxation modulus of neat PLA and its single filler nanocomposites at 190 °C.

Stress relaxation test was also performed at 190 °C with strain magnitudes of γ₀ = 0.05. Fig. 5b shows the relaxation modulus, G(t), of neat PLA and its nanocomposite samples containing 1% of pristine and functionalized carbon nanofillers. It is observed that the addition of nanofillers reduces the stress relaxation rate which can be related to nanofiller induced slowing down of the PLA chain dynamic, as in the case of dynamic oscillatory shear measurements. The values of relaxation modulus of the samples with functionalized carbon nanofillers are much larger than those of neat PLA or pristine carbon nanofillers based nanocomposites. In other words, functionalized carbon nanofillers based nanocomposites relaxation resembles a solid material while the others relax quickly.

The results of both experiments indicate that the addition of functionalized carbon nanofillers to the PLA matrix significantly affects the long time relaxation behavior of the nanocomposites by increasing the relaxation time due to the formation of strong three dimensional network structures. In the case of fCNTs based nanocomposites, longer relaxation time and higher relaxation modulus indicate that the strong structure of this nanocomposite creates a significant energetic barrier against the molecular motion during the shear flow. It is probably due to the presence of fCNTs domains dispersed on a molecular level in the PLA matrix as well as the interfacial interaction between PLA chains and fCNTs.

Having considered the distinct effect of functionalization discussed above (XRD and rheological results), the rest of the experiments were carried out on functionalized single and hybrid nanofillers filled PLA.

Dispersion of carbon nanofillers in PLA matrix. Fig. 6 shows angular frequency dependence of storage modulus, complex viscosity and damping factor for PLA and its functionalized single filler and hybrid nanocomposites measured at 190 °C. The single filler and hybrid nanocomposite samples exhibit a pronounced viscosity upturn and a strong nonterminal (or the so-called solid like) behavior at low frequency ranges, which are indications of the formation of three dimensional networks between individual particles (particle–particle) and/or the particles and the matrix (particle–matrix). The strength of these networks or the degree to which they are formed is enhanced by increasing degree of dispersion and concentration of nanofillers.^38–40

Fig. 6 (a) Storage modulus, (b) complex viscosity, and (c) damping factor of neat PLA and its single filler and hybrid nanocomposites at 190 °C.

A comparison between the two single nanofiller filled PLA revealed a stronger solid like response in the case of the fCNTs based nanocomposite, implying that their ability to form three dimensional networks is greater than that of fGnPs. This can be attributed to the flexible geometry of fCNTs, which can boost the formation of network structures. Furthermore, simultaneous incorporation of fGnPs and fCNTs increases the viscosity and elasticity of the samples, which are characteristic of nanostructured materials. With increasing fCNTs content, the nonterminal behavior in storage modulus and viscosity upturn at low frequencies becomes more pronounced.

The results of the damping factor (tan δ) are shown in Fig. 6d. The nanocomposite samples exhibited low damping factors with a peak at a higher frequency in comparison with the neat PLA, which is due to the existence of interfacial bonding between the carbon nanofillers and the polymer chains. This means that the movement of the chains was restricted and that their relaxation times increased. This is more noticeable with fCNTs based nanocomposites and for hybrid filler samples (particularly those with higher fCNTs contents).³⁸

Morphology characterizations of samples

SEM micrographs of the cryofractured surface of PLA and the nanocomposite samples each containing 1 wt% of the single nanofillers or fGnPs/fCNTs hybrid nanocomposites are presented in Fig. 7. The fractured surface of the neat PLA matrix displays a smooth surface typical of brittle fracture (Fig. 7a). As it can be observed in Fig. 7b and c, both single nanofiller filled PLA samples show a good state of dispersion of CNTs for PLA/1% fCNTs and exfoliated GnPs for PLA/1% fGnPs, respectively. Although, a few domains of aggregated fCNTs can still be observed. Fig. 7d–f show that the hybrid filled nanocomposites display a good dispersion/distribution of both nanofillers with no rich fCNTs domains, or aggregates, which demonstrates the adequacy of the mixing process (the arrows indicate the location of the nanofillers). Similar morphologies have been reported by other researchers, who proposed two plausible explanations for the improved dispersion state in CNTs–GnPs/polymer hybrid nanocomposites:^28,37 (1) in the case of multilayer GnPs, the long and tortuous CNTs can bridge the distances between multilayer GnPs and inhibit their restacking, (2) for the exfoliated GnPs, the single layer of GnPs inhibits reaggregation of CNTs, which improves their dispersion. The results obtained in this work revealed that the second is the dominant mechanism for the hybrid nanofiller filled PLA.

Fig. 7 SEM micrographs of (a) neat PLA, (b) PLA/1% fGnPs, (c) PLA/1% fCNTs, (d) PLA/0.75% fGnPs–0.25% fCNTs, (e) PLA/0.5% fGnPs–0.5% fCNTs, and (f) PLA/0.25% fGnPs–0.75% fCNTs.

Thus, thanks to exfoliated fGnPs platelets and considering the large lengths of the flexible fCNTs, which can act as chelating arms between fGnPs platelets, a significant increase in the contact area of hybrid nanofillers and the PLA matrix seems to be perfectly plausible. These results suggest that PLA has a greater dispersing capability when both nanofillers are present and therefore, stronger three dimensional networks are formed in hybrid filler systems.

EMI SE measurements

The results of electromagnetic interference shielding effectiveness (EMI SE) versus frequency for fGnPs/fCNTs hybrid nanocomposites are compared with EMI shielding of the single filler nanocomposites over the frequency range of 5–12 GHz (Fig. 8a). In case of the single filler nanocomposite samples each containing 1 wt% of fCNTs and fGnPs, the average EMI SE values over the 5–12 GHz range was 4.82 and 2.54 dB, respectively. This revealed a higher shielding effectiveness in the case of the fCNTs filled PLA, which can be attributed to the significant improvement in the electrical conductivity of the fCNTs. EMI SE is strongly dependent on the dispersion degree of carbon nanofillers in the polymer matrix. The incorporation of the electroconductive fCNTs and fGnPs with high aspect ratios and low percolation thresholds and thus lower loading levels required to achieve a given EMI SE value, can enhance absorption and reflection through the primary and secondary shielding mechanisms.^41,42 It is clearly observed that the value of the EMI SE in the hybrid filler systems depends strongly on the fGnPs/fCNTs ratio, increasing with the loading of fCNTs.

Fig. 8 (a) EMI SE of neat PLA and its single filler and hybrid nanocomposites, (b) average EMI SE of hybrid filler nanocomposites with different fGnPs : fCNTs ratios at overall concentrations of 1 wt%.

Fig. 8b shows average EMI SE values of hybrid filler nanocomposites plotted as a function of the studied mix ratios (at overall filler concentrations of 1 wt%.). Compared with the mixture law prediction (the dashed line) on EMI SE, the hybrid filler nanocomposites reveal a certain degree of positive deviation in practice, especially when the ratio of fGnPs : fCNTs is 1 : 3. This enhancement may be potentially attributed to the formation of efficient conductive networks in hybrid filler systems. For commercial applications, the target value of the EMI SE is around 20 dB, which is equal to or 1% lower than the transmission of the electromagnetic wave. Hence, hybrid filler nanocomposites show greater promise for commercial applications compared to their single filler counterparts.

Tensile properties

It has been shown that in carbon nanofiller reinforced composites, the mechanical reinforcement inclines at a faster clip above a critical nanofiller concentration where percolating filler networks are formed, known as percolation threshold. The percolation threshold (P_c) for MWCNTs and GnPs can be estimated from 0.5/η and 21.2/α respectively, where η = l/d and α are the average aspect ratios.²⁵ We have assumed that destructive factors such as ultrasonication have caused minimal damage to carbon nanofillers during the functionalization and sample preparation procedures and thus their aspect ratios are preserved. According to the technical data sheet for the MWCNTs and GnPs used in this study, their average aspect ratios are about η = 158 and α = 4000, respectively. Considering the density of fCNTs (1.8 g cm⁻³), fGnPs (2.2 g cm⁻³), and PLA (1.24 g cm⁻³), the calculated percolation threshold for fCNTs and fGnPs is 0.31 and 0.53 vol%, which corresponds to weight concentration of 0.39 and 0.67 wt%, respectively. In our study, the overall nanofillers concentration was 1 wt%, well above the estimated percolation thresholds.

The tensile properties of the neat PLA and its single filler and hybrid nanocomposites are summarized in Table 1. It can be seen that the incorporation of nanofillers into the PLA leads to enhancement in the modulus and strength.

Table 1 Tensile test data of the neat PLA and its single filler and hybrid nanocomposites

Samples	Tensile modulus (MPa)	Tensile strength (MPa)	Elongation at break (%)
PLA	1317.58 ± 30.53	35.41 ± 1.54	4.4 ± 0.2
PLA/1% fGnPs	1744.94 ± 51.26	40.95 ± 2.14	3.1 ± 0.1
PLA/0.75% fGnPs–0.25% fCNTs	1870.68 ± 54.85	43.87 ± 2.65	3 ± 0.2
PLA/0.5% fGnPs–0.5% fCNTs	1976.56 ± 48.13	46.75 ± 2.47	2.8 ± 0.1
PLA/0.25% fGnPs–0.75% fCNTs	2024.16 ± 60.47	48.97 ± 2.81	2.6 ± 0.2
PLA/1% CNTs	2032.82 ± 58.47	49.63 ± 2.73	2.4 ± 0.2

---

The nanocomposites containing 1 wt% of fCNTs exhibited a significant increase of 40.2 and 54.3% in tensile strength and modulus, respectively. Meanwhile, the nanocomposites containing 1 wt% of fGnPs improved tensile strength and modulus by 15.7 and 32.4%, respectively.

It is well known that the reinforcement mechanism of polymer nanocomposites is governed by the load transfer, stress concentration and defect distribution. According to the literature, the modulus of nanoparticle reinforced polymers depends mainly on the composite constituents (modulus, rigidity, volume fraction and geometrical dimensions), rather than the interfacial interaction between the nanoparticles and the polymer matrix. It is no surprise that PLA/1 wt% fCNTs nanocomposite had a higher modulus that PLA/1 wt% fGnPs nanocomposite, because fCNTs have a higher modulus and volume fraction than fGnPs inside the PLA. On the other hand, the strength of polymer nanocomposites is largely determined by the interfacial interaction between the nanoparticles and polymer matrix. In effect, sufficient interfacial adhesion leads to a high degree of load transfer from polymer matrix to the nanoparticles, resulting in enhanced mechanical properties. In this study, functionalization was carried out on the carbon nanofillers to achieve this goal. The elongation at break of single filler and hybrid nanocomposites is lower that of neat PLA. The fracture strain of the nanocomposite samples is probably due to voids and defects caused by the agglomerations and slippage of the overlapped nanofillers. In other words, poor dispersion of the reinforcements caused formation of nano or micro flaws, which result in local stress concentration in the matrix. Some of the unappealing aspects of these carbon nanofillers, such as waviness of pristine MWCNTs and their reaggregation, as well as the restacking of GnPs (arising from van der Waals forces) and their tendency to bend and roll up, contribute to the decline in their effective aspect ratios and surface area. This deterioration leads to a weak interfacial adhesion between matrix and nanoparticles which may hamper stress transfer and limit reinforcement efficiency.²⁸

Fig. 9 plots the tensile modulus and strength, and the fracture strain of the hybrid filler nanocomposites as a function of the mix ratios at the overall filler concentrations of 1 wt%. The hybrid filler nanocomposites show a positive deviation in the tensile modulus and strength when compared with the mixture law (the dashed line). The clarification of these observations needs a closer look into the microstructure of these systems. The exfoliated fGnPs inhibit the reaggregation of fCNTs, which leads to stronger three dimensional hybrid nanofiller networks. Consequently, a large surface area and strong interfacial adhesion between the nanofillers networks and the PLA matrix is established, which contributes to the effectiveness of stress transfer. Moreover, these networks are not conducive to the formation of flaws, which is very important for stress transfer from the nanofillers to the PLA matrix. Under small strains, it is much more difficult to change the orientation of the fCNTs constituting the hybrid networks compared to the single/free fCNTs, since they now have stable junctions with fGnPs and a strong interfacial adhesion with the PLA chains. This can explain the higher tensile modulus observed in the hybrid filler nanocomposites. At higher strains, however, the rotation of fGnPs and the orientation of fCNTs turn into factors which can give rise to the rearrangement of the nanofillers networks. It can be suggested that during this rearrangement process, some additional mechanical energy was dissipated, which inhibited the rapid growth of the emerging micro cracks by weakening the stress concentrations. All of the above explanations seem plausible enough to account for the improvement in the mechanical properties of hybrid filler nanocomposites, manifested in a positive deviation from the mixture law.

Fig. 9 (a) Tensile strength, (b) tensile modulus and (c) elongation at break of hybrid filler nanocomposites with different fGnPs : fCNTs ratios at overall concentrations of 1 wt%.

Thermal analyses

Dynamic mechanical behavior of samples. Fig. 10 shows the results of storage modulus (E′) and damping factor (tan δ) versus temperature for the neat PLA and its single filler and hybrid nanocomposites. Of all the nanocomposite samples, the one with 1 wt% of fCNTs showed the highest storage modulus. In the case of the hybrid filler nanocomposites, the storage modulus generally increased with increasing fCNTs content under the glass transition region, as shown in Fig. 10a. This result is due to the stronger reinforcing effect of the fCNTs, which restrict the motion of PLA chain segments and enhance the storage modulus. On the other hand, we suggest that fCNTs dispersion was improved by simultaneous incorporation of exfoliated fGnPs, which results in formation of more cross-linking physical bonds between the PLA chains and fCNTs and thus the movement of chains is further restricted and the storage modulus increased significantly. As it can be observed from Fig. 10b, PLA shows a peak corresponding to its glass transition temperature (T_g) at ∼59 °C. The peak position shifted to slightly higher temperatures with increasing fCNTs content. The peak height also reduced because of the sufficient incorporation of fGnPs and well dispersed fCNTs, which restricted the mobility of the PLA chains, which casts further light on the improvements in mechanical properties. The values of tan δ peaks for hybrid filler nanocomposites were plotted against the mix ratios in the inset of Fig. 10b, which shows a negative deviation from the mixture law (the dashed line) suggesting that the polymer chains' motion was more restricted by the better dispersed fCNTs and consequently strong hybrid networks in hybrid filler systems.

Fig. 10 DMA plots of (a) storage modulus and (b) damping factor of neat PLA and its single filler and hybrid nanocomposites.

For all the samples, an increase in the storage modulus is observed with rising temperature in the cold crystallization range (about 85 °C for neat PLA and about 78–85 °C for the nanocomposite samples) owing to reinforcing by the crystallites being formed. The decrease in cold crystallization temperature can be attributed to heterogeneous nucleation of carbon nanofillers, which is considerable for nanocomposites containing fCNTs.⁴³

Thermal decomposition study. The thermal stability of single filler and hybrid nanocomposites was investigated by thermogravimetric and differential thermogravimetric analysis (TG-DTA) (Fig. 11). From Fig. 11a, the relative thermal stability of the samples was compared by means of the temperature of 5% weight loss (T_d,5%), 70% weight loss (T_d,70%), and the char yields at 450 °C. In order to track the changes in weight loss more closely, certain temperature ranges have been magnified in the inset. Moreover, the temperature of the maximum rate of mass loss (T_max) is illustrated in Fig. 11b, while T_max values of hybrid filler nanocomposites were plotted against the mix ratios at the overall filler concentrations of 1 wt% in the inset. The thermogravimetric data are summarized in Table 2.

Fig. 11 (a) TGA, (b) DTG, and (c) DTA curves of neat PLA and its single filler and hybrid nanocomposites.

Table 2 TG-DTA data for the neat PLA and its single filler and hybrid nanocomposites

Samples	Td,5% (°C)	Td,70% (°C)	Tmax (°C)	Char residue (%) at 500 °C
PLA	312.78	361.35	359.94	0.79
PLA/1% fCNTs	323.34	368.03	370.11	1.71
PLA/0.25% fGnPs–0.75% fCNTs	320.47	370.95	373.02	1.92
PLA/0.5% fGnPs–0.5% fCNTs	318.4	374.41	376.21	2.01
PLA/0.75% fGnPs–0.25% fCNTs	328.24	377.16	378.18	2.38
PLA/1% fGnPs	332.56	378.55	379.07	2.59

---

In the case of nanocomposites, the T_d,5%, T_d,70% and T_max were enhanced upon addition of the carbon nanofillers, especially for the nanocomposites containing fGnPs, indicating that the incorporation of fGnPs effectively improved the thermal stability. This can be assigned to the well-known protective effect of the 2D platelets of fGnPs (acting as a mass transfer barrier) against thermal decomposition of the nanocomposites.^44,45 Much more residual char was also obtained for the nanocomposite samples compared with the neat PLA sample. From the inset of Fig. 11b, enhancement of T_max was more noticeable for hybrid filler nanocomposites, which showed an extent of positive deviation compared with the mixture law (the dashed line). This improvement can be attributed to fGnPs-assisted dispersion of fCNTs, which resulted in the formation of a more effective barrier network in terms of protection against mass and heat transfer. Meanwhile, the chars hinder the transfer of volatile products and heat during the decomposition process.

The heat flow curves of the neat PLA and its nanocomposite samples during decomposition process in the temperature ranges of 250 to 450 °C are shown in Fig. 11c. Obviously, four main endothermic peaks appear in the DTA curve of the neat PLA, while only two or three are found for all nanocomposite samples, indicating that different decomposition processes (i.e. chain scission processes) occur between two and four periods. However, two endothermic peaks in case of the nanocomposite samples mean that the decomposition processes mentioned above may have overlapped.

Putting all TGA/DTA results together, it does not seem that the improvement of thermal stability can be easily rationalized in terms of the barrier model, which suggests that the decomposition rates of the nanocomposites must have been limited by hindering the diffusion of gaseous decomposition products. In addition, the presence of surface barrier should not affect the total heat of decomposition. Although the barrier model makes sense from both an experimental and a theoretical perspective, there may be other factors behind the enhanced thermostability. Consequently, it is reasonable to speculate that the thermal decomposition mechanism of the nanocomposite samples is different from that of the neat PLA.

Conclusions

In this work, single filler and hybrid PLA nanocomposites containing previously functionalized MWCNTs and GnPs were prepared by a solution mixing method. XRD revealed that the fGnPs were fully exfoliated throughout the PLA matrix. A homogeneous dispersion of single filler and hybrid of fCNTs and fGnPs was achieved in the PLA matrix as evidenced by dynamic rheological tests and scanning electron microscopy. The strong interaction between the functionalized carbon nanofillers and the PLA matrix greatly enhanced the dispersion as well as the interfacial adhesion, which resulted in strong hybrid fCNTs–fGnPs networks. The hybrid filler nanocomposites showed a certain degree of positive deviation in electrical, mechanical and thermal properties when compared to the mixture law. The electromagnetic interference shielding efficiency (EMI SE) of the nanocomposite samples was enhanced with increasing the loading level of fCNTs. On the contrary, TG-DTA results indicate that the presence of fGnPs proved beneficial for further increase of thermal stability properties due to their favorable protective effect. Finally, detailed investigation of the tensile test results showed that effective stress transfer, retarded formation of flaws, slowed growth of microcracks and dissipation of additional mechanical energy resulting from efficient hybrid networks were responsible for the enhanced mechanical properties of hybrid filler nanocomposites.

Notes and references

1. M. Singhvi and D. Gokhale, RSC Adv., 2013, 3, 13558–13568.
2. R. A. Auras, L.-T. Lim, S. E. Selke and H. Tsuji, Poly(lactic acid): synthesis, structures, properties, processing, and applications, John Wiley & Sons, 2011.
3. M. G. A. Vieira, M. A. da Silva, L. O. dos Santos and M. M. Beppu, Eur. Polym. J., 2011, 47, 254–263.
4. S. Gu, M. Yang, T. Yu, T. Ren and J. Ren, Polym. Int., 2008, 57, 982–986.
5. Z. Qiu and W. Guan, RSC Adv., 2014, 4, 9463–9470.
6. H. Wang and Z. Qiu, Thermochim. Acta, 2011, 526, 229–236.
7. Z. Qiu and P. Zhou, RSC Adv., 2014, 4, 51411–51417.
8. V. H. Sangeetha, H. Deka, T. O. Varghese and S. K. Nayak, Polym. Compos., 2016 DOI:10.1002/pc.23906.
9. J.-B. Zeng, K.-A. Li and A.-K. Du, RSC Adv., 2015, 5, 32546–32565.
10. R. M. Rasal, A. V. Janorkar and D. E. Hirt, Prog. Polym. Sci., 2010, 35, 338–356.
11. H. Norazlina and Y. Kamal, Polym. Bull., 2015, 72, 931–961.
12. R. B. Valapa, G. Pugazhenthi and V. Katiyar, RSC Adv., 2015, 5, 28410–28423.
13. I. Kelnar, J. Kratochvil, I. Fortelny, L. Kapralkova, A. Zhigunov, M. Nevoralova, M. Kotrisova and V. Khunova, RSC Adv., 2016, 6, 30755–30762.
14. W. Yang, F. Dominici, E. Fortunati, J. M. Kenny and D. Puglia, RSC Adv., 2015, 5, 32350–32357.
15. W. K. Chee, H. N. Lim, N. M. Huang and I. Harrison, RSC Adv., 2015, 5, 68014–68051.
16. M. Zhang, Y. Li, Z. Su and G. Wei, Polym. Chem., 2015, 6, 6107–6124.
17. G. Mittal, V. Dhand, K. Y. Rhee, S.-J. Park and W. R. Lee, J. Ind. Eng. Chem., 2015, 21, 11–25.
18. Z. Spitalsky, D. Tasis, K. Papagelis and C. Galiotis, Prog. Polym. Sci., 2010, 35, 357–401.
19. M. Rahmat and P. Hubert, Compos. Sci. Technol., 2011, 72, 72–84.
20. N. G. Sahoo, S. Rana, J. W. Cho, L. Li and S. H. Chan, Prog. Polym. Sci., 2010, 35, 837–867.
21. T. Ramanathan, A. Abdala, S. Stankovich, D. Dikin, M. Herrera-Alonso, R. Piner, D. Adamson, H. Schniepp, X. Chen and R. Ruoff, Nat. Nanotechnol., 2008, 3, 327–331.
22. S. Maiti, N. K. Shrivastava, S. Suin and B. Khatua, ACS Appl. Mater. Interfaces, 2013, 5, 4712–4724.
23. A. Yu, P. Ramesh, X. Sun, E. Bekyarova, M. E. Itkis and R. C. Haddon, Adv. Mater., 2008, 20, 4740–4744.
24. M. Safdari and M. S. Al-Haik, Carbon, 2013, 64, 111–121.
25. L. Yue, G. Pircheraghi, S. A. Monemian and I. Manas-Zloczower, Carbon, 2014, 78, 268–278.
26. M. H. Al-Saleh, Synth. Met., 2015, 209, 41–46.
27. A. Poosala, W. Kurdsuk, D. Aussawasathien and D. Pentrakoon, Chiang Mai J. Sci., 2014, 41, 1274–1286.
28. W. Li, A. Dichiara and J. Bai, Compos. Sci. Technol., 2013, 74, 221–227.
29. S. Chatterjee, F. Nafezarefi, N. Tai, L. Schlagenhauf, F. Nüesch and B. Chu, Carbon, 2012, 50, 5380–5386.
30. H. Im and J. Kim, Carbon, 2012, 50, 5429–5440.
31. S.-Y. Yang, W.-N. Lin, Y.-L. Huang, H.-W. Tien, J.-Y. Wang, C.-C. M. Ma, S.-M. Li and Y.-S. Wang, Carbon, 2011, 49, 793–803.
32. P.-C. Ma, N. A. Siddiqui, G. Marom and J.-K. Kim, Composites, Part A, 2010, 41, 1345–1367.
33. V. Georgakilas, M. Otyepka, A. B. Bourlinos, V. Chandra, N. Kim, K. C. Kemp, P. Hobza, R. Zboril and K. S. Kim, Chem. Rev., 2012, 112, 6156–6214.
34. S. Dadashi, S. M. Mousavi, Z. Emam-Djomeh and A. Oromiehie, International Journal of Nanoscience and Nanotechnology, 2014, 10, 245–256.
35. M. Sabzi, L. Jiang, F. Liu, I. Ghasemi and M. Atai, J. Mater. Chem. A, 2013, 1, 8253–8261.
36. J. T. Yoon, Y. G. Jeong, S. C. Lee and B. G. Min, Polym. Adv. Technol., 2009, 20, 631–638.
37. A. Rostami, H. Nazockdast, M. Karimi and Z. Javidi, Adv. Polym. Technol., 2016, 35, 80–89.
38. A. Rostami, M. Masoomi, M. J. Fayazi and M. Vahdati, RSC Adv., 2015, 5, 32880–32890.
39. A. Saadat, H. Nazockdast, F. Sepehr and M. Mehranpour, Polym. Eng. Sci., 2010, 50, 2340–2349.
40. E. Nazockdast and H. Nazockdast, Appl. Rheol., 2011, 21, 25434.
41. M. H. Al-Saleh, W. H. Saadeh and U. Sundararaj, Carbon, 2013, 60, 146–156.
42. S. R. Dhakate, K. M. Subhedar and B. P. Singh, RSC Adv., 2015, 5, 43036–43057.
43. A. Nuzzo, E. Bilotti, T. Peijs, D. Acierno and G. Filippone, Polymer, 2014, 55, 4908–4919.
44. P. Visakh and O. B. Nazarenko, Thermal Degradation of Polymer Blends, Composites and Nanocomposites, Springer, 2015.
45. S. Bourbigot and G. Fontaine, Polym. Chem., 2010, 1, 1413–1422.

---