Thermal degradation of polymer–carbon nanotube composites
This chapter discusses the thermal degradation of polymer–carbon nanotube (CNT) composites which plays a crucial role in determining their processing and applications. The chapter first reviews the mechanisms of thermal degradation/stability improvement by carbon nanotubes, which is affected by the barrier effect, thermal conductivity of CNT, physical or chemical adsorption, radical scavenging action, and polymer–nanotube interaction. It then describes the thermal degradation/stability of polymer–CNT composites and their development trends. It is necessary and meaningful to further investigate the thermal degradation of polymer–CNTs.
Since the work of Iijima in 1991, many polymer–carbon nanotubes (CNT) composites (Chen et al., 2006; Yuen et al., 2007; Chen and Wu, 2007; Kanagaraj et al., 2007; Kim et al., 2009a) have been studied with promising applications due to the unique combination of electronic, thermal, optical, and mechanical properties of CNT (Schadler et al., 1998; Stephan et al., 2002; Dalton et al., 2003; Barrau et al., 2003; Thostenson and Chou, 2003; Coleman et al., 2006), as summarized in recent review articles (Lau and Hui, 2002; Breuer and Uttandaraman, 2004; Moniruzzaman and Winey, 2006; Chen et al., 2009a; Mohammed and Uttandaraman, 2009). However, there are only a few studies on the thermal properties, especially thermal degradation or thermal stability, which is very important in the application of polymer–CNT composites.
The thermal stability of polymer–CNT composites plays a crucial role in determining their processing and applications because it affects the final properties of polymer–CNT composites, such as the upper limit use temperature and dimensional stability. For the fabrication of polymer–CNT composites with better balance in processing and performance, it is necessary to characterize the thermal stability and decomposition kinetics of polymer composites. Thus, an understanding of the thermal decomposition behavior of polymer–CNT composites makes it possible to develop commercial applications in a broad range of industries. At the same time, the thermal stability, in turn, is closely connected to fire retardancy. The thermal degradation/stability improvement also facilitates enlarging the applications in the field of fire retardancy. So, in this chapter, the thermal degradation of polymer–CNT composites is reviewed.
In general, the thermal degradation of materials is assessed by thermogravimetry analysis (TGA). However, there is no uniform standard on the characteristic temperature of materials from TGA curves. The characteristic values in the TGA curve that are commonly used include the temperature of the initial weight loss (Ti), the temperature of maximum weight loss rate (Tmax), the onset temperature obtained from the intersection between the baseline and the tangent at maximum weight loss rate (Tonset); the temperature at which 10% degradation occurs (T0.1), the temperature at which 50% degradation occurs, the mid-point of the degradation process (T0.5); and the amount of residue, denoted as char. Sometimes, it is considered that T0.1 is consistent with Tonset. A typical TGA curve for a polyurea containing 0.2% MWNT is shown in Fig. 16.1.
It is now well accepted that the improved thermal stability of polymer–CNT composites is due to the following: barrier effect, thermal conductivity of CNTs, physical or chemical adsorption, radical scavenging action, and polymer–nanotube interaction. For each polymer–CNT composite, thermal stability may be due to one mechanism or the combined action of several processes, which depend on the different components, microstructures and exterior conditions.
Barrier effect: The nanotube barrier effect of polymer–CNT composite degradation processes is mainly a mass transport barrier, which slows the escape of the volatile products during the process of degradation and the permeation of O2 or air. It is related to the network structure and dispersion of nanotube in the composites and the compact char formation. The well-dispersed nanotubes could hinder the transport of polymer degradation products from the condensed phase to the gas phase compared to composites with poorly-dispersed nanotubes (Marosfoi et al., 2006). The accumulation of CNT with a network structure in the composites also tends to increase the mechanical integrity of a protective layer or barrier to prevent the evolution of degradation products to the gas phase. This protective layer or barrier will also limit air or oxygen diffusion into or within the polymer composite, and this is extremely efficient at low temperatures. However, on increasing the temperature up to a limiting value, the superficial protecting nanotube network will be destroyed and the large mass loss starts until degradation is complete. In addition, the formation of a compact char of CNT and polymer matrix during the thermal degradation is beneficial to the barrier labyrinth effect of CNT so that the diffusion of degradation products from the bulk of the polymer to the gas phase is slowed (Kashiwagi et al., 2002).
Thermal conductivity of CNTs: It is well known that CNTs, as compared to the polymers, show higher thermal conductivity in the axial direction. The addition of CNTs will assist in conducting heat from the surface and facilitate heat dissipation within the composites, resulting in slow degradation of the composites (Yu et al., 2000). CNTs in the composites form a network by connecting with each other in a random way. This special morphology can effectively promote the thermal conductivity of the CNT network in the entire composite to improve the thermal stability (Fujii et al., 2005). On the other hand, CNTs are sometimes functionalized or modified to obtain good dispersion, which may impair thermal conductivity due to separation of the CNTs (Stevens et al., 2003; Sakellariou et al., 2008; Bartholome et al., 2008; Gao et al., 2009).
Physical or chemical adsorption: CNTs have typical diameters in the range of ~ 1–50 nm, lengths of many microns and a large aspect ratio which can physically or chemically adsorb the decomposition products; in addition, polymers near the nanotubes might degrade more slowly due to decreasing activity, which would move degradation to higher temperatures. It has been reported that bundle exfoliation could remarkably increase the aspect ratio of the CNT in the composites, and result in a more effective absorption of the degradation products during the degradation processes (Zanetti et al., 2001; Yang et al., 2005).
Polymer–nanotube interaction: The interfacial interaction of the polymer–nanotube is significantly affected by the surface characteristics of the CNTs. The pure CNT is easy to agglomerate and the interface is poor. Surface functionalization could significantly modify the surface characteristics of CNTs and strengthen the interfacial bonding between CNTs and the polymer matrix (Markovic et al., 2006; Lee et al., 2007; Kim et al., 2008a). Good interfacial interaction will increase the activation energy of degradation. The higher the activation energy of degradation, the more stable the composite. It also reduces the interfacial thermal resistance (Yang et al., 2008). However, the functionalization of CNTs will also have a negative effect on the thermal stability of the CNT due to the degradation of polymers grafted onto the surface of the CNT. Therefore, the optimum degree of functionalization for the CNT must be considered (Wang, 2009). Figure 16.2 shows an example of the interfacial interaction between m-CNT and poly(ethylene 2,6-naphthalate) (PEN) matrix.
Radical scavenging action: Generally, it is known that the thermal degradation of most polymers occurs by random chain scission to form radicals that are susceptible to reagents capable of trapping the radicals. Fang et al. (2008) proposed the free radical-trapping mechanism of C60 to explain the enhanced thermal properties. At the initial degradation stage of polymers, C60 simultaneously traps macromolecular free radicals and other small free radicals, like H• and •OH, to form cross-link gel networks. One the other hand, the radical trapping effect of C60 also allows the macromolecular radicals more time to recombine, which also contributes to the formation of a gel network. Furthermore, at some elevated temperatures, the degradation of polymers will produce macromolecular radicals with several active sites. These active sites are easily attacked by the C60 particles and subsequently form a branching or 3D gel network structure. The CNTs have high electron affinities, similar to C60, and have been proposed to act as scavengers of free radicals (Watts et al., 2003). In fact, CNTs are usually purified, by using concentrated HNO3/H2SO4 mixtures before use. The surface of the oxidized CNT is covered to some extent with carboxylic (− COOH), carbonyl (C=O) and hydroxyl (− OH) groups. These groups can trap active free radicals to form more stable species during the degradation process of polymers such as PE, PP, and PMMA. Investigations also indicate that the iron oxides in CNTs could also act as a radical trapping component during the process of degradation (Zhu et al., 2001; Kashiwagi et al., 2002; Du et al., 2006; Sahooa et al., 2009). The entangled network structure of the CNT also helps the radical trapping through restricting the thermal motion of the polymer chains, and keeps the degrading polymer radicals together to present radical recombination reactions during the degradation process, which has been called nano-confinement (Gorrasi et al., 2007; Costache et al., 2007; Chen et al., 2007). When the radicals produced are relatively unstable, there is no time for them to participate in radical recombination reactions (Costache et al., 2006).
The intrinsic thermal stability of CNTs is important for the thermal properties of polymer–CNT composites. In general, pristine CNTs are thermally stable up to 3073 K in a vacuum, 1023 K in air and there is no significant mass loss in N2 below 1073 K (Wang, 2009). But CNTs always contain amorphous carbon, carbon nanoparticles, hemifullerenes and residues from the metal catalysts. These have an effect on the thermal stability of the CNT. The thermal stability of amorphous carbon, carbon nanoparticles, and hemifullerenes is comparatively lower than that of CNT itself. Therefore, the lower the purity of the CNT, the easier is the degradation. For residues from the metal catalysts, such as Fe3 +, Ni2 + or Co2 +, although the thermal degradation of the CNT is not affected, there are important influences on the thermal stability of polymers such as PMMA (Xi et al., 2005), EVA (Marosfoi et al., 2006) and PC (Schartel et al., 2008); these metal ions are Lewis acids which can bond with the carbonyl oxygen in PMMA and form an intermediate complex. Due to conjugation, free radicals are weakened, the activity of free radicals increases and the PMMA thermal degradation reaction occurs more easily. However, Kashiwagi et al. (2002) proposed that iron did not affect the thermal stability of PP in PP–iron-containing MWNTs. Since the iron particles are inside and at ends of the MWNTs, the chances of contact with PP chains during the TGA experiments would be low and would not occur until the walls of the nanotube were destroyed. Bocchini et al. (2007) also reported that impurities in MWNTs (e.g. metal oxides from the MWNT synthesis catalyst) do not influence the thermal stability of LLDPE in nitrogen.
The thermal stability of CNTs also depends on the structure of the CNT. Marosfoi et al. (2006) reported that DWNTs undergo a two-step degradation and start to lose mass at 723 K; the majority of the mass loss occurred in the second step with a degradation temperature of 798 K in air. In the case of MWNTs, the mass loss starts at 877 K and large mass loss occurs at 1003 K. One could even observe the mass loss at temperatures up to 1073 K. The differences in degradation behavior between DWNTs and MWNTs could be explained by the higher number of defects in DWNTs formed in the synthetic process, and these defects make DWNTs more subject to the degradative action of air compared to MWNTs. Liew et al. (2005) have studied the thermal stability of single-walled and multi-walled carbon nanotubes and found that S WNTs are thermally more stable than MWNTs, due to the presence of more than one layer in the MWNT. When a MWNT is subjected to thermal load, the atoms from different layers start to vibrate. At high temperatures, the vibrations become stronger and the atoms from one layer collide with the atoms from the neighboring layers, making it easier for the MWNT to degrade.
CNTs are highly polarizable carbon compounds with an attractive interaction of 0.5 eV per nanometer of inter-tube contact. The cohesive force makes it difficult to disperse CNTs into individual nanotubes, and limits the potential applications of CNTs (Wang, 2009). Chemical treatment of CNTs becomes an attractive strategy to overcome these barriers. One of the most common chemical techniques is the oxidative treatment in solution. While good dispersion of CNT could be achieved through modification, the thermal stability of the CNT is affected, depending on the modification methods. The enhanced thermal stability of functionalized MWNTs by different chemical treatments was reported by Yang et al. (2008). Figure 16.3 shows the synthesis and structure of functionalized MWNTs. Their TGA traces are shown in Fig. 16.4. It has been observed that the Tonset of pristine MWNT (0#), HNO3-treated MWNT(1#), H2O2-NH3-treated MWNT(2#), and EDA-treated MWNT(3#) appears at 890 K, 991 K, 916 K, and 954 K, respectively. The degradation temperature of MWNTs treated by nitric acid is higher than that oxidized by H2O2/NH3 or EDA. The onset degradation temperature difference between EDA-treated MWNT and nitric acid-treated MWNT is attributed to the amine-functionalization which may have weakened the interfacial bonding between graphite sheets of the MWNT. Impurities have been removed by chemical treatment in the preparation of peroxide-ammonia-treated-MWNTs, resulting in better intrinsic thermal stability than in the pristine material. Similar experimental results have been reported in other papers (Andrews et al., 2004; Ovejero et al., 2006).
16.4 The TGA traces of functionalized MWNTs by different chemical treatments. Reprinted with permission from Yang et al. (2008).
Defect functionalization is another effective chemical treatment. An obvious decrease in thermal stability of MWNT obtained by defect functionalization and ‘grafting to’ technique was observed by Xu et al. (2007). The structures are presented in Fig. 16.5. The acid-treated MWNT exhibited very good thermal stability which showed 8% mass loss at 873 K in N2. The mono-fluorene-functionalized MWNT displayed 15% mass loss at 873 K. Furthermore, the polyfluorene (PF)-functionalized MWNT started to decompose at about 423 K and lost about 50% of the mass at 873 K. The decreased decomposition temperature of the PF-functionalized MWNT is due to the bromo-end groups of the grafted polyfluorenes which dramatically promote the decomposition of conjugated polymers. Similar observations were presented for PLA-g-MWNT (poly(L-lactic acid) (Chen et al., 2005). Lower thermal stability was observed for PLA-g-MWNT systems compared to that of the acid-treated MWNT (MWNT-COOH). Moreover, the extent of the decrease depends on the different molecular weight of PLA. The molecule chain length of PLA grafted onto the MWNT used in this study is presented in Fig. 16.6. The thermal degradation of PLA-g-MWNT gradually increased with the increase of molecule weight. Jana and Cho (2008) also reported that Ti of poly(ε-caprolactone)diol-g-MWNT (PCL-g-MWNT) decreased regularly with the increase in wt% of PCL due to the increased proportion of the carbonyl groups coming from the f-MWNT. The carbon atoms of the f-MWNT attached to the carboxylic group are sp3 hybridized, which causes high inherent strain, making them more susceptible to thermal degradation. The Tonset of various modified carbon nanotubes are summarized in Table 16.1.
|Functionalized carbon nanotubes||Tonset (oC)||Ref.|
|MWNT-COOH||717||Yang et al. (2008)|
|NH2-MWNT-C=O||642||Yang et al. (2008)|
|MWNT-CONHCH=CHNH2||680||Yang et al. (2008)|
|MWNT-g-PLLA||250||Chen et al. (2005)|
|MWNT-g-PCL||280||Jana and Cho (2008)|
|MWNT-PA||200||Zhang et al. (2008b)|
|MWNT-g-PS||310||Wang et al. (2006b)|
|SWNT-g-PS||337||Wang et al. (2006b)|
In most cases, CNTs incorporated into the polyolefins can lead to a stabilization of the polymeric matrix. Gorrasi et al. (2007) revealed that the thermal degradation of LLDPE–MWNT composites in the air was significantly delayed with 1–10 wt% of MWNT. Tonset of LLDPE was enhanced by about 40 K for a 10 wt% loading of MWNT. In the Costache et al. (2007) study, two different CNTs were used in the preparation of PE–CNT composites. T0.1 and T0.5 of PE–MWNT were respectively increased by about 16 K and 5 K, respectively, while for PE–SWNT, the increases are 13 K and 3 K, compared to PE. No significant difference has been observed between MWNT and SWNT. For PP–MWNT a slight increase of Tmax was reported by Wiemann et al. (2005). These findings were related to the high thermal stability, good thermal conductivity, and network structure of CNTs. Meanwhile, impurities in the CNTs will be inevitably brought into the composites. These impurities can promote the degradation of polyolefins. Kanagaraj et al. (2007) observed that Tonset of HDPE–MWNT composites rapidly decreased with the addition of MWNTs (range of diameter 60–100 nm, length of the tubes 5–15 μm, and purity > 95%). Tonset are 668 K, 628 K, 600 K and 589 K for pure HDPE, 0.11%, 0.22% and 0.44% MWNT in PE, respectively. However, this could be avoided by purifying MWNT. McNally et al. (2005) reported that Tonset of PE–MWNT composites, where MWNTs were treated with acid and distilled water, was enhanced by about 20 K at a 10 wt% loading of MWNTs.
Experimental results indicated that chemically-modified CNTs could improve the mechanical properties, but this does not always enhance the thermal stability of polyolefin composites. The Kovalchuk et al. (2008) study showed that the addition of 0.1 wt% modified MWNTs obtained from methylaluminoxane reacting with – COOH groups on the MWNT increases Tmax of the iPP/modified MWNT by about 30 K in comparison with pristine iPP, and this parameter rises with a further increase in the filler content. By introducing just 3.5 wt% of modified MWNT, Tmax surpasses the characteristic value of the neat polymer by about 60 K. This thermal stabilization effect is mostly connected with the nanotube barrier effect. The thermal degradation temperature will be improved if the interfacial interaction between polyolefin and CNT is enhanced. Shofner et al. (2006) provided an improved thermal stable effect for PE–CNT composites through comparative experiments among PE, PE–pure SWNTs and PE–fluorinated SWNTs systems. Although the interfacial interaction of PE–fluorinated SWNTs was modified, PE–fluorinated SWNTs show smaller differences in the thermal behavior than PE–pure-SWNTs compared to the pure PE. The difference would be accounted for by a lower carbon loading in the fluorinated SWNTs because of fluorine mass included in the fluorinated SWNTs.
The formation of a protective char is another reason for the delayed thermal and oxidative degradation of polyolefin/CNT. Bocchini et al. (2007) did research on the thermal degradation of LLDPE–MWNT. SEM and ATR–FTIR experimental data indicated the presence of a thin protective film of MWNT–carbon char composite on the surface of the nanocomposites, and this has been used to explain the delayed thermal degradation temperature of LLDPE–MWNT in N2 or in air. This protective char was formed by the catalyzed oxidative dehydrogenation of LLDPE in the presence of MWNTs, which prevents oxygen diffusion towards the underlying polymer matrix and the loss of volatile products from the polymer. Kashiwagi et al. (2002) and Kashiwagi et al. (2004) also observed that a nanotube network layer is formed which insulates PP from the external radiant flux during burning of the PP–MWNT nanocomposites. Their research indicates that this nanotube layer could physically reduce the external heat flux up to one-half. The formation of nanotube layer also was confirmed by Zhao et al. (2006) and Seo and Park (2004).
Most research on the degradation of polyolefin/CNT indicated that free radical trapping plays an important role in the enhancement of thermal stability of the composites. The C-C bond in the main chains of polyolefin breaks at a high temperature to form carbon-based radicals, or the C-H bond is attacked by oxygen to generate HO• and alkoxy free radicals in the presence of O2, and subsequently leads to a series of polymer chain degradations through random scission. The addition of CNTs can delay the degradation of polyolefins by trapping free radicals. Watts et al. (2003) revealed that the thermal-oxidative stability of PE, PP and poly(vinylidene fluoride) (PVDF) is retarded by MWNT. The oxidation induction temperatures (OIT) for pure PP, CNT-PP (1:5) and CNT-PP(1:10) are 431 K, 443 K and 448 K, respectively. In the case of PVDF, it is also evident that the OIT shifts from 398 K for pure PVDF to > 433 K for CNT-PVDF. CNT contain localized trapping sites due to lattice defects (e.g. vacancies, dangling bonds, OH and C=O attachments). These localized states are acceptor-like, which can terminate radicals during polymer degradation. In Kodjie et al.’s (2006) work, the thermal stability of HDPE–CNT also showed dramatic enhancement based on the radical scavenging action, as high as 70 K/115 K improvement of Tmax in N2/air atmosphere, respectively.
In another study carried out by Wang et al. (2006a), it was demonstrated that physical or chemical adsorption of the MWNT (diameter 10–30 nm, length 0.5–40 μm) for the PP molecules could significantly increase the thermal degradation temperature. The improvements in the Tmax for PP at the 1% and 5% loading of unmodified MWNT are about 40 K and 70 K, respectively. TGA and DTG data suggest a two-step degradation in the PP–MWNT composites, which indicates that the incorporation of MWNT induces PP molecular to form two different aggregation structures; this has been explained by physical adsorption of MWNT. Nanotubes were covered with a layer of PP molecules through physical adsorption in the PP–MWNT composites prepared by the melt mixing method. The adsorbed PP molecules are much less active than those far from the nanotube surface, and their degradation is delayed. Similar physical or chemical adsorption of the MWNTs takes place in the investigation by Yang et al. (2005). In their work, MWNTs were added to the aPP, weight-averaged molecular weight and number averaged molecular weight are 41,402 and 1481, respectively, by melt blending at 353 K in a Barabender mixer. TGA results indicate that thermal stability of the aPP–MWNT composite in nitrogen is enhanced significantly by the addition of nanotubes. The peak temperature of the DTG curve for the nanocomposites with 5 wt% nanotube loading is about 70 K higher than that of pure aPP.
The defects and acid sites such as –COOH and –OH on the MWNTs caused by acid treatment can impair the thermal or thermo-oxidative stability of polyolefin. Bikiaris et al. (2008) pointed out that a decreasing trend is seen in the iPP-MWNT nanocomposites with an increase in the acid-treatment time of the MWNT. Thermal or thermo-oxidative decomposition data of the iPP–MWNT nanocomposites are presented in Table 16.2. These can be explained by more defects and acid sites on the MWNT which degrade at low temperature to induce the degradation of iPP.
Polyamides belong to the group of polar polymers. Li et al. (2006) investigated the thermal degradation difference between PA6–p-MWNT and PA6–f-MWNT, where f-MWNT was prepared by grafting 1, 6-hexamethylenediamine onto the MWNT. Ti in air of PA6 was improved by 5 K for PA6–p-MWNT and 14 K for PA6–f-MWNT at 0.5 wt% MWNT loading. The higher thermal stability of PA6–f-MWNT is due to the stronger interfacial interactions, which were evidenced by the degradation activation energy of composites in air obtained by the Kissinger method. The degradation activation energy of PA6–f-MWNT is 169 kJ/mol while that of PA6–p-MWNT composites is 165 kJ/mol. The effect of different functionalized MWNT on the thermal stability of PA6–MWNT composites was also investigated (Chen et al., 2006). TGA data suggested that the PA6–MWNT-NH2 composites exhibited greater thermal stability than PA6–MWNT-COOH composites at the same MWNT loading. This difference could be due to the strong interfacial interaction between MWNT-NH2 and PA6 and the PA6–MWNT-NH2 composite showed higher degradation activation energy compared to PA6–MWNT-COOH composite.
The thermal degradation of PA–CNT composites was also affected by the amount of CNT used in the composites. According to Huang et al. (2009), the thermal stability of polyamide 11 (PA 11)–MWNT composites is highly dependent on the concentration of MWNTs; in this study, CNT was prepared by catalytic chemical vapor deposition of methane on Co-Mo/MgO catalysts and then further treated in 2.6 M nitric acid. TGA data show that Tonset of the composite increases with an increase of the MWNT content, and the composites containing 1 wt% MWNT have the best thermal stability, and the decomposition temperature is improved about 20 K compared to PA11. The relationship between Tonset and MWNT concentration in the PA11–MWNT composites is shown in Fig. 16.7. This trend was confirmed in PA6 (Gao et al., 2005b; Shen et al., 2009).
A study by Sahoo et al. (2009) indicated that processing methods have an influence on the improvement of the thermal properties of the PA6–MWNT composites. The composite containing 10 wt% MWNT prepared in a Haake internal mixer (method I) showed delayed degradation compared to the same MWNT composition prepared by an extrusion process and subsequently injection molding (method II). The 10% and 50% decomposition temperatures of PA6–MWNT obtained from method I are higher by 10 K and 5 K compared to the composite from method II. Such an improvement can be associated with the better dispersion of MWNT in the Haake internal mixer. This investigation reveals that thermal stability of MWNT-filled polymer composites is highly dependent on the state of dispersion, mixing and processing conditions.
The thermal degradation of PA6–MWNT showed no obvious changes, but the flame retardancy of PA6–MWNT is significantly influenced by the incorporation of MWNTs (Schartel et al., 2005). The cone calorimeter results indicate that during the burning of PA6–MWNT, the MWNT interconnected network structure remained in the materials, since it was thermally stable and had strong mechanical properties. The rheological measurements suggested that the melt viscosity increased which not only prevented dripping and flowing, but also hindered the decomposition of volatiles feeding the flame.
The addition of CNTs does not change the thermal degradation mechanism of PMMA composites. However, the dispersion and concentration of SWNTs were found to significantly influence the thermal degradation of PMMA–SWNT in the work of Kashiwagi et al. (2005). Evidence from the cone calorimeter and optical microscopy experiments indicated that a nanotube-containing network layer without any cracks or openings was formed during the burning tests, and the entire surface of the sample with good dispersion was covered by this protective layer (Fig. 16.8), while the composites with poor dispersion or a low content of the nanotubes (0.2% by mass or less) formed numerous black discrete islands and vigorous bubbling was observed between the islands. The peak heat release rate of the composites forming the network layer is about one-half that of those which formed the islands. More experimental data from TGA, TGA-FTIR, GC-MS, and cone calorimetry (Costache et al., 2006) revealed that MWNTs served as a barrier in the degradation of PMMA–MWNT.
16.8 Effect of nanotubes’ dispersion on residues of PMMA–SWNT (0.5%): (a) nitrogen gasification residue, good dispersion sample; (b) burned residue, good dispersion sample; (c) nitrogen gasification residue, poor dispersion sample; (d) burned residue, poor dispersion sample. (Reprinted with permission from Kashiwagi et al., 2005.)
In the case of PMMA–CNT composites prepared by in-situ dispersion polymerization, many researchers (Dai et al, 2007; Kashiwagi et al., 2007; Kovalchuk et al., 2008; Zhang et al., 2008a; Choi et al., 2008) have found that TGA data of the thermal degradation temperature of PMMA–CNT composites could be improved by about 10–60 K. This improved thermal stability of PMMA–CNT composites was attributed to strong interfacial interaction and outstanding thermal conductivity, determined by the method of preparation. S WNTs or MWNTs can react with the initiator to open the π-bond, participate in PMMA polymerization and form strong interfacial interaction between CNT and PMMA (Jia et al., 1999; Park et al., 2003), and the heat dissipation rate of the interface between CNT and matrix can be improved. In addition, it is also related to the enhanced thermal conductivity of composites due to the addition of SWNTs or MWNTs compared to the pure PMMA (Bagchi and Nomura, 2006).
Yuen et al. (2007) reported that the presence of silane-modified MWNT increased the thermal stability of the PMMA–VTES polymer matrix both in N2 and in air. The thermal degradation data for the Si-MWNT–PMMA–VTES composites are summarized in Table 16.3. The authors stated that the increased thermal degradation characteristic temperature is caused by the interposition of Si. However, thermal conductivity data provided in the research suggested that this increased thermal degradation characteristic temperature should be related to the improved thermal conductivity between the MWNT and polymer matrix. The thermal conductivity of Si-MWNT–PMMA–VTES increased with the addition of Si-MWNT compared to neat PMMA–VTES, since Si-MWNT was bonded to PMMA–VTES via SiOx, improving the interfacial interaction between the MWNTs and the polymer matrix and facilitating heat dissipation.
Characteristic thermal degradation temperature of Si-MWNT–PMMA–VTES composites. Reprinted with permission from Yuen et al. (2007)
The addition of CNTs will change the degradation pathway of styrenic polymers. Kong and Zhang’s (2008) work revealed that the thermal stability of HIPS–CNT composites is better than that of HIPS in N2, in which the CNT was prepared by the method using the catalytic decomposition of ethanol with LaNiO3 as catalyst in an autoclave at 973 K.
Tonset and T0.5 of HIPS–CNT are 21 K and 10 K, respectively, higher than that of HIPS. Py-GC-MS was used to determine the reason for the high thermal stability of the HIPS–CNT composites. The routine degradation mechanism of pure HIPS starts by β-scission of a chain-end radical; the dimer and trimer are produced by an intermolecular transfer (backbiting) reaction. The abundant product is styrene, and a small amount of α-methylstyrene is also seen. However, GC-MS experiments suggested that in the process of degradation of HIPS–CNT, there is a high yield of α-methylstyrene. It is explained that the primary radical is willing to transform into the tertiary radical because of its increased stability after the formation of primary and secondary radicals by β-scission, which leads to the higher yield of α-methylstyrene (Jang and Wilkie, 2009). Schematics of β-cracking are shown in Fig. 16.9. The degradation follows this pathway when CNT is mixed with HIPS. This degradation model was also observed in HIPS–Fe-MMT nanocomposites (Kong et al., 2008).
Chemical modification of CNT is also an efficient way to affect the thermal stability of styrenic polymers–CNT composites. Chen et al. (2009c) reported a specific process of styryl-modified MWNT, as shown in Fig. 16.10, and proved that the degradation temperature of styryl-modified MWNT–PS has been improved compared to MWNT-COOH–PS. This improvement of PS-g-MWNT was also reported in other studies (Kong et al., 2004). All research indicated that this improved thermal stability was associated with the imposed restrictions due to the interaction between the styryl groups on the MWNT surface and the polymer chains which was enhanced by the chemical modification.
Polyaniline (PANI) is one of the most important hydrophilic-conducting polymers; the thermal stability of PANI–MWNT composites was studied in some research (Ma, et al., 2006; Markovic et al., 2006; Qi et al., 2008). Qi et al. (2008) reported that the incorporation of MWNTs (diameter: < 10 nm and length range: 5–15 μm, purity: 95%) achieved improved thermal stability for PANI–MWNT composites prepared by in-situ polymerization. Ti and the peak temperature of PANI–MWNTs were delayed about 40 K and 100 K, respectively. Another study on the thermal degradation of PANI–MWNT (8~20 nm in diameter and 2~20 pm in length) composites prepared by ultrasonically initiated, in-situ emulsion polymerization was reported by Markovic et al. (2006). Ti of PANI–MWNT composites increased by ~ 35 K. The phenomenon is related to the interactions between the aromatic ring π-bonds of PANI and graphitic structures of MWNT.
The incorporation of CNT into styrenic polymers does not always improve the thermal stability of composites. In the case of acrylonitrile-butadiene-styrene (ABS)–SWNT composite, one could see a two-step degradation process, however, the initial degradation temperature is decreased, and the second degradation temperature is increased, compared to those of pure ABS. A catalytic effect of SWNT was proposed although the full understanding of the role of SWNT in the radical degradation initiation is still needed (Yang et al., 2004). For the decreased initial degradation temperature of styrene-butadiene-styrene tri-block copolymer (SBS)–MWNT composites (Lu et al., 2009), a different mechanism was proposed. The presence of low molecular weight materials produced in the melt mixing process was taken as the source which led to the decrease of initial degradation temperature.
The thermal degradation of polyesters could be affected because of the barrier effect of CNTs. Research indicates that the thermal stability of CNT–polyesters composites including poly(ethylene 2,6-naphthalate)(PEN)–CNT composites (Kim and Kim, 2006; Hu et al., 2008; Kim et al., 2008b; Kim et al., 2009b), poly(butylene terephthalate)(PBT)–MWNT composites (Mathew et al., 2005; Wu et al., 2008a; Kim, 2009), poly(ethylene terephthalate)(PET)–MWNT composites (Lee et al., 2005; Anand et al., 2007; Vassiliou et al., 2009), poly(ε-caprolactone)diol (PCL)–MWNT composites (Sivalingam et al., 2003; Wu et al., 2007) and PC–MWNT composites (Schartel et al., 2008) was significantly affected compared to the pure polyesters. There is no doubt that the enhancement of thermal stability was due to the barrier effect of CNTs which was proved by the high char yield on TGA analysis. However, the improved thermal stability of polyester composites is sensitive to the purity and chemical modification of CNT. In the case of poly(L-lactic acid) (PLA)–CNT, one can see the difference in the thermal degradation of PLA–CNT caused by the purity of CNT (Moon et al., 2005; Kim et al., 2007; Kim et al., 2008a). The decomposition temperature of PLA–purified CNT composites is about 10 K higher than that of PLA–non-purified CNT composites (Chiu et al., 2008). The T0.1 of PC–MWNT composites shifted with increasing MWNT content up to 95 K towards lower temperatures which is also related to additional components or impurities incorporated during processing, such as remaining catalysts from nanotube synthesis (Schartel et al., 2008). To understand how the difference in the chemical modification affects the thermal stability of PLA–CNT composites, Wu et al. (2008b) prepared PLA–CNT composites using various functionalized CNT including carboxylic and hydroxy MWNT. The thermal stability order of composites is: PLA–hydroxy CNT < PLA<PLA–carboxylic CNT. After the initial stage of degradation, the PLA–hydroxy CNT sample has a higher degradation rate, even in contrast to that of neat PLA, since the hydroxy groups produce Lewis or Bronsted acid sites on the surface of the CNT at a high temperature and the relative poor dispersion state of the MWNT reduces thermal conductivity efficiency. Although the carboxylic groups can also produce Lewis acid sites on the surface of a MWNT, the good dispersion shown by SEM (Fig. 16.11) enhances the barrier effect of the carboxylic MWNT. The functionalized MWNT affected the thermal stability of PLA composites by changing the dispersion state of CNT in polymer matrix.
16.11 TEM images of (a) PLA–carboxylic (CNT); (b) PLA–hydroxy (CNT); and (c) PLA–pure (CNT) samples at a magnification of 596,000. Reprinted with permission from Wu et al. (2008b).
Lai et al. (2004) reported that the degradation temperature of poly(hydroxybutyrate-co-hydroxyvalerate)(PHBV)–MWNT composites was 16 K higher than that of pure PHBV due to the nanodispersion of CNT. In the case of polyacrylonitrile (PAN)–surface-oxidized MWNT composites, Tonset shifts from 541 K for pure PAN to 565 K for the PAN–MWNT (95/5) composites in N2. This change is attributed to the enhanced interfacial interaction due to the chemical binding of negatively charged –C≡N: groups on the MWNT (Ge et al., 2004). In Probst et al.’s (2004) work on poly(vinyl alcohol) (PVA)–SWNT, SWNT were found to promote the thermal degradation of PVA due to the impurities and defects on the SWNT. Similar decreased thermal stability was also reported in the cross-linked composites of PVA–MWNT prepared by the addition of glutaraldehyde to PVA–CNT dispersions to form cross-link PVA chains via an acetalization reaction (Basiuk et al., 2009) and PVDF–SWNT composites fabricated by dispersion of SWNT in an aqueous surfactant solution, followed by mixing with PVDF powder, filtration and hot pressing (Xu et al., 2006).
Recently, Chen et al. (2009b) studied the effect of the combined action of clay and carboxylated MWNT with 3 wt% carboxyl content and where 10–20 nm outer diameter on the thermal degradation of hydrogenated nitrile-butadiene rubber (HNBR) and the Kissinger, Flynn-Wall-Ozawa, and Friedman methods were used to evaluate the activation energy sequence of HNBR and its nanocomposites. Both the initial degradation temperature and activation energy sequence of HNBR and its nanocomposites are in the order: HNBR–clay–MWNT > HNBR–clay > HNBR. HNBR–clay–MWNT composites had higher char yield at 873 K than HNBR–clay, which was attributed to the interaction between clay and MWNT. The activation energies of HNBR and HNBR composites showed a sharp increase in the low conversion area and a slow increase in the high conversion region. A reduction in the diffusion speed of the degradation products gives the lower thermal degradation rate observed for HNBR–clay–CNT nanocomposites. The coexistence of clay and CNT could form compact char layers with better barrier properties than clay alone. Similar experimental results were found in the clay-supported CNT–poly (vinyl alcohol) (PVA) nanocomposites reported by Zhao et al. (2009).
Accelerated char formation due to presence of clay and MWNT was again indicated as a mechanism of thermal stability improvement by Gao et al. (2005a) in the ethylene vinyl acetate copolymer (EVA) composites with MWNT and clays. In the morphology of the chars in the composites (presented in Fig. 16.12) produced after cone calorimetry and the natural burning experiment, it was found that nanotubes act as the site of nucleation of graphitization leading to the formation of turbostratic and graphitic carbons. Perfect crystallites of graphite are shown by a broad peak at 0.3354 nm, exactly the interlayer distance of pure graphite, in the XRD pattern of the char of the composite containing both nanotube and clay. In addition, there is remarkable difference in the number of cracks in the char, shown in Fig. 16.12. The char is the barrier to the evolution of flammable volatiles to the vapor phase and oxygen ingress to the condensed phase while the cracks permit evolution and ingress. A similar thermal stability improvement of MWNT with magnesium hydroxide in EVA was also reported by Ye et al. (2009) and Beyer (2002).
16.12 The morphology of the chars produced from clay–EVA, clay–MWNT–EVA and MWNT–EVA nanocomposites following the cone calorimeter test (top row) and the natural burning (bottom row). (Reprinted with permission from Gao et al., 2005a.)
Considering the literature available on polymer–CNT composites, both degradability and stability for several composites have not yet been studied completely and the mechanisms of degradation have also not been clearly explained (Kashiwagi et al. 2002; Breuer and Uttandaraman, 2004; Bocchini et al. 2007; Kumara et al. 2009). For instance, how does one prove the thermal stability improvement or interactions at the molecular level between CNT and polymers? In particular, the combination of excellent properties and the gradually decreasing cost of CNT make them an ideal candidate as an advanced filler material for the next generation of high performance structural and multifunctional composites (Njuguna et al. 2008, Chen et al, 2009a), which will play an important role in a wide range of material applications and fundamental science in the future. It is therefore necessary and important to further investigate the thermal degradation of polymer–carbon nanotubes.
In conclusion, the thermal degradation of nanotube–polymer composites offers both great potential and great challenges, making it a vibrant area of work for years to come. The improvement and application of these composites at high temperature and their utility in fire retardancy will depend on how effectively we can handle the challenges. The significant progress in the understanding of these composite systems within the past few years points toward a bright future (Gelves et al. 2005; Moniruzzaman and Winey, 2006).
In this review, we provide an overview of research in the past decade on the thermal degradation of polymer–CNT composites and give insights into the factors that will ultimately control this. However, there are no universal conclusions that can be drawn because of the complexity of thermal degradation and the variety of composites. The following points are evident for the thermal degradation of polymer–CNT composites:
First, the particular structure and excellent thermal properties of CNT are particularly important in the thermal stability of polymer–CNT composites so that the incorporation of CNTs into polymers can significantly influence the thermal degradation of polymer–CNT composites, particularly the thermal stability improvement of composites.
Second, the thermal degradation behavior of polymer–CNT composites is affected by many factors, such as component, structure, interfacial interaction, exterior conditions and so on. The changes in the thermal stability of polymer–CNT composites may be ascribed to one dominant factor or a combination of factors.
Third, the mechanisms of thermal degradation of composites mainly include good matrix–nanotube interactions, good thermal conductivity of the nanotubes, barrier effect, the network character and dispersion of carbon nanotubes, and radical scavenger action and physical-chemical adsorption by carbon nanotubes.
Bikiaris, D., Vassiliou, A., Chrissafis, K., et al. Effect of acid treated multi-walled carbon nanotubes on the mechanical, permeability, thermal properties and thermo-oxidative stability of isotactic polypropylene. Polym. Degrad. Stab. 2008; 93:952–967.
Chen, G.X., Kim, H.S., Park, B.H., Yoon, J.S., et al. Controlled functionalization of multiwalled carbon nanotubes with various molecular-weight poly(l-lactic acid). J. Phys. Chem. B. 2005; 109:22237–22243.
Chen, S.G., Yu, H.Y., Ren, W.T. Thermal degradation behavior of hydrogenated nitrile-butadiene rubber (HNBR)/clay nanocomposite and HNBR/clay/carbon nanotubes nanocomposites. Thermochim. Acta. 2009. [doi 10, 1016/j.tca.03, 010].
Chen, X.H., Tao, F., Wang, J.F., et al. Concise route to styryl-modified multi-walled carbon nanotubes for polystyrene matrix and enhanced mechanical properties and thermal stability of composite. Mater. Sci. Eng. A. 2009; 499:469–475.
Choi, H.J., Lim, J.Y., Zhang, K., et al. Poly(methyl methacrylate)/multi-walled carbon nanotube microbead composites via dispersion polymerization under ultrasonication. Diamond & Related Materials. 2008; 17:1498–1501.
Costache, M.C., Heidecker, M.J., Wilkie, C.A., et al. The influence of carbon nanotubes, organically modified montmorillonites and layered double hydroxides on the thermal degradation and fire retardancy of polyethylene, ethylenevinyl acetate copolymer and polystyrene. Polymer. 2007; 48:6532–6545.
Costache, M.C., Wang, D.Y., Wilkie, C.A., et al. The thermal degradation of poly(methyl methacrylate) nanocomposites with montmorillonite, layered double hydroxides and carbon nanotubes. Polym. Adv. Technol. 2006; 17:272–280.
Ge, J.J., Hou, H.Q., Li, Q., et al. Assembly of well-aligned multiwalled carbon nanotubes in confined polyacrylonitrile environments: electrospun composite nanofiber sheets. J. Am. Chem. Soc. 2004; 126:15754–15761.
Gorrasi, G., Sarno, M., Bartolomeo, A.D., et al. Incorporation of carbon nanotubes into polyethylene by high energy ball milling: morphology and physical properties. J. Polym. Sci. Part B: Polym. Phys. 2007; 45:597–606.
Huang, S., Wang, M., Liu, T.X., et al. Morphology, thermal, and rheological behavior of nylon 11/multi-walled carbon nanotube nanocomposites prepared by melt compounding. Polym. Eng. Sci. 2009; 49:1063–1068.
Kashiwagi, T., Du, F., Winey, K.I., et al. Flammability properties of polymer nanocomposites with single-walled carbon nanotubes: effects of nanotube dispersion and concentration. Polymer. 2005; 46:471–481.
Kim, J.Y., Han, S., Kim, D.K., et al. Mechanical reinforcement and crystallization behavior of poly(ethylene 2,6-naphthalate) nanocomposites induced by modified carbon nanotube. Composites, Part A. 2009; 40:45–53.
Lee, H.J., Oh, S.J., Choi, J.Y., et al. In situ synthesis of poly(ethylene terephthalate) (PET) in ethylene glycol containing terephthalic acid and functionalized multiwalled carbon nanotubes (MWNTs) as an approach to MWNT/PET nanocomposites. Chem. Mater. 2005; 17:5057–5064.
Ma, X.F., Zhang, X.B., Li, Y., et al. Preparation of nano-structured polyaniline composite film via carbon nanotubes seeding approach and its gas-response studies. Macromol. Mater. Eng. 2006; 291:75–82.
Mathew, G., Hong, J.P., Rhee, J.M., et al. Preparation and characterization of properties of electrospun poly(butylene terephthalate) nanofibers filled with carbon nanotubes. Polymer Testing. 2005; 24:712–717.
Ovejero, G., Sotelo, J.L., Romero, M.D., et al. Multiwalled carbon nanotubes for liquid-phase oxidation. functionalization, characterization, and catalytic activity. Ind. Eng. Chem. Res. 2006; 45:2206–2212.
Park, S.J., Cho, M.S., Lim, S.T., et al. Synthesis and dispersion characteristics of multi-walled carbon nanotube composites with poly(methyl methacrylate) prepared by in situ bulk polymerization macromol. Rapid Commun. 2003; 24:1070–1073.
Sahoo, N.G., Cheng, H.K.F., Cai, J.W., et al. Improvement of mechanical and thermal properties of carbon nanotube composites through nanotube functionalization and processing methods. Comp. Sci. Technol. 2009; 69:313–320.
Shen, Z.Q., Bateman, S., Wu, D.Y., et al. The effects of carbon nanotubes on mechanical and thermal properties of woven glass fiber reinforced polyamide-6 nanocomposites. Comp. Sci. Tech. 2009; 69:239–244.
Stevens, J.L., Huang, A.Y., Peng, H.Q., et al. Sidewall amino-functionalization of single-walled carbon nanotubes through fluorination and subsequent reactions with terminal diamines. Nano Let. 2003; 3:331–336.
Vassiliou, A.A., Chrissafis, K., Bikiaris, D.N., et al, Thermal degradation kinetics of in situ prepared PET nanocomposites with acid-treated multi-walled carbon nanotubes. J. Therm. Anal. Calorim. Published on line 2009; doi: 10.1007/s10973-009-0426-4.
Wiemann, K., Kaminsky, W., Gojny, F.H., et al. Synthesis and properties of syndiotactic poly(propylene)/carbon nanofiber and nanotube composites prepared by in situ polymerization with metallocene/MAO catalysts. Macromol. Chem. Phys. 2005; 206:1472–1478.
Wu, D.F., Wu, L., Sun, Y.R. Rheological properties and crystallization behavior of multi-walled carbon nanotube/poly(ε-caprolactone) composites. J. Polym. Sci. Part B: Polym. Phys. 2007; 45:3137–3147.
Yang, K., Gu, M.Y., Han, H.B., et al. Influence of chemical processing on the morphology, crystalline content and thermal stability of multi-walled carbon nanotubes. Mater Chem. Phys. 2008; 112:387–392.
Ye, L., Wu, Q.H., Qu, B.J. Synergistic effects and mechanism of multiwalled carbon nanotubes with magnesium hydroxide in halogen-free flame retardant EVA/MH/MWNT nanocomposites. Polym. Degrad. Stab. 2009; 94:751–756.
Yuen, S.M., Ma, C.C.M., Chiang, C.L., et al. Silane-modified MWCNT/PMMA composites–preparation, electrical resistivity, thermal conductivity and thermal stability. Composites, Part A: Applied Science and Manufacturing. 2007; 38:2527–2535.
Zhang, K., Lim, J.Y., Choi, H.J., et al. Core-shell structured carbon nanotube/poly (methyl methacrylate) composite and its electrorheological characteristics. Diamond & Related Materials. 2008; 17:1604–1607.