- Email: [email protected]
epoxy composites by using a pre-curing treatment
Composites Science and Technology 71 (2011) 765–771
Contents lists available at ScienceDirect
Composites Science and Technology journal homepage: www.elsevier.com/locate/compscitech
Improving the flexural and thermomechanical properties of amino-functionalized carbon nanotube/epoxy composites by using a pre-curing treatment S.G. Prolongo ⇑, M.R. Gude, A. Ureña Dpt. Materials Science and Engineering, ESCET, University Rey Juan Carlos, c/Tulipán s/n 28933 Móstoles, Madrid, Spain
a r t i c l e
i n f o
Article history: Received 29 September 2010 Received in revised form 24 January 2011 Accepted 30 January 2011 Available online 20 February 2011 Keywords: A. Nanocomposites B. Mechanical properties B. Electrical properties B. Thermomechanical properties
a b s t r a c t A prior thermal (pre-curing) treatment of mixtures of epoxy monomer and amino-functionalized carbon nanotubes (CNTs) was used to promote a chemical reaction between the matrix and the reinforcement, favouring the formation of a strong interface. Samples of epoxy resin and different weight percentages of amino-functionalized multi-walled CNTs were prepared with and without the pre-curing treatment (150 °C, 1 h). The degree of dispersion of the nanofiller was better when this pre-curing treatment was used. This allowed a higher CNT content while keeping a high sample homogeneity. Without the pre-curing step, the addition of CNTs increases both the flexural strength and strain to failure by 45%. Moreover, with the pre-curing step, the nanocomposite with 0.25 wt.% CNTs presents an increase of flexural strength by 58% and strain to failure by 68% regard to neat epoxy resin. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Composites reinforced with CNTs are being widely researched for over a decade. Dramatic improvements are expected in their thermal, mechanical and electrical behaviour. Epoxy resins are well established as thermosetting matrices of advanced composites displaying a series of promising characteristics for a wide range of applications. So the incorporation of CNTs into epoxy resins presents numerous potential applications, including aerospace and automobile materials, due to the expected increase of their use temperature, their low density and their improved mechanical properties. Besides, the development of CNT/epoxy composites opens new perspectives for multi-functional materials, e.g., conductive polymers with improved mechanical performance and with a perspective of damage sensing and ‘‘life’’-monitoring [1,2]. Despite the promising developments for CNT/epoxy composites, the inherent processing difficulties, such as dispersion and interphase control, have caused mechanical performance to fall short of theoretical predictions. The most common procedure used to overcome this problem is to try to get an effective CNT surface grafting, addressing both issues, dispersion and interface stress transfer [3–5]. Numerous grafting methods have been employed to introduce a variety of molecules onto CNTs [6–8]. One of them is based on amino-functionalization of CNTs. Kim et al. [9] confirmed that the amino-functionalized CNTs exhibit higher surface energy and much better wettability with epoxy resin than the pristine CNTs, inhibiting the nanotube re-agglomeration during the ⇑ Corresponding author. Fax: +34 91 4888150. E-mail address: [email protected] (S.G. Prolongo). 0266-3538/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.compscitech.2011.01.028
curing of the resin. In fact, several works confirm the better thermal and mechanical performance of epoxy composites reinforced with amino-functionalized CNTs when compared to the untreated ones [9–11]. However, direct evidences of covalent interactions between the modified CNT and epoxies have not been reported. This is presumably due in part to the low concentration of functional groups on the CNT surface when dispersed into epoxy resin/curing agent mixture which makes it difficult to pinpoint the reaction. Abdalla et al. [12] confirmed that the differences in the way the epoxy agent reacts with the modified nanotubes can result in steric and electronic differences that can in turn affect the cure and this invariably translate to differences in the crosslinking topology and hence the macroscale properties. When the curing reaction is carried out on a mixture of epoxy monomer and amino curing agent with dispersed amino-functionalized CNTs, both kind of amine groups (from curing agent and from CNTs) compete to react with epoxide groups. Taking into account the low concentration of CNT added and the low percentage of amine groups linked over their surface, together with steric hindrances and diffusion problems of rigid nanotubes, it is obvious that the chemical reaction with the curing agent is more favoured. This implies that the amount of covalent bonds between epoxy matrix and amino-functionalized CNTs must be low. Therefore, we have investigated the effect of inducing the chemical reaction between amino-functionalized CNTs with the epoxy monomer before curing process, in order to assure the formation of covalent linkages at the interface, which facilities stress transfer on the composite. For this, a previous thermal treatment, named pre-curing, is applied to the mixture of functionalized CNTs and epoxy monomer, before the addition of the curing agent.
766
S.G. Prolongo et al. / Composites Science and Technology 71 (2011) 765–771
In previous works, authors have tried to find experimental evidences, by indirect methods, of the chemical reaction between amino groups anchored on CNTs and the oxirane rings of the epoxy monomer when the pre-curing treatment was applied. One of these methods was the measurement of the viscosity of mixtures of the epoxy monomer with different contents of amino-functionalized carbon nanotubes before and after applying the pre-curing treatment [13]. For low CNT contents the viscosity was almost the same irrespective of whether pre-curing treatment was carried out or not. The main reason is the very low concentration of amine groups, close to 2000:1 (epoxy:hydrogen amine). However for 0.5 wt.% CNTs, the viscosity was approximately 70% higher for the pre-cured mixtures regard to non-treated ones, suggesting stronger interfacial epoxy–nanotube interactions because of the chemical reaction epoxy-amine. Differential scanning calorimetry (DSC) was used with the aim of determining the enthalpy of the pre-curing reaction (epoxy monomer/CNT mixtures), but due again to the low amount of functional groups on the CNTs, any exothermic peak was not detected. However, a small change was observed in the glass transition temperature (Tg) of the epoxy monomer from 18 to 16 °C for precured CNT/epoxy mixtures, indicating that chemical reaction between amino-functionalized CNT and epoxy rings of the monomer had occurred. Also, this was confirmed indirectly determining the curing enthalpy of the epoxy monomer/hardener/CNT mixtures pre-cured and without any thermal treatment. In this case, the heat of reaction was 2–3 kJ/epoxy equivalent lower for pre-cured composites. This could mean that some of the oxirane rings of the epoxy monomer have previously reacted, during the pre-curing treatment, and thus they are not available for the curing with the hardener [14]. Once it was proven the effectiveness of the pre-curing treatment in promoting the chemical reaction between the epoxy monomer and the amino-functionalized multi-walled carbon nanotubes, in this work we study its effect on the properties of the CNT/epoxy composites. A stronger interface between the nanofiller and the matrix should improve the load transfer, increasing the mechanical properties.
the same temperature. The resulting mixture was degassed at 90 °C for more than 12 h in order to eliminate the remaining chloroform and the trapped air. Half of the samples were subjected to pre-curing treatment, involving of heating at 150 °C for 1 h. This treatment was carried out with mechanical stirring at approximately 150 rpm. Then the hardener was added in stoichiometric ratio. The composite resin was molded in a steel mould and cured at 210 °C for 3 h. Composites, pre-cured and non-pre-cured, containing different weight fraction (i.e. 0, 0.25 and 0.40 wt.%) of amino-functionalized CNTs were prepared. 2.3. Characterisation In order to qualitatively determine the dispersion of CNTs, the composites were observed by Field Emission Gun Scanning (FEG– SEM) Electron Microscopy (Nova NanoSEM FEI 230). The surface of the samples was sputter coated by a thin layer (5–10 nm) of Au (Pd). The experimental conditions of the sputtering were 30 mA and 120 s (Bal-tec, SCD-005 sputter). A Mettler Toledo balance, with ±0.001 mg, equipped with a density determination kit by means the buoyancy technique, was used to evaluate the change of density in composites. Ten measurements of density for each sample were carried out. Thermal dynamic mechanical properties were determined with a TA Instrument DMTA Q800 operating in single cantilever mode with an oscillation frequency of 1.0 Hz. Data were collected from 100 °C to 300 °C at a scanning rate of 2 °C/min. The dimensions of DMTA specimens were 1.5 15 35 mm3. The mechanical characterisation of composites was carried out by flexural test (Instron 4465), following the ASTM D-790 at a crosshead speed of 0.8 mm/min. The fracture surfaces were also covered with Au (Pd) and observed by Scanning Electron Microscopy (ESEM, Philips XL30) in order to study the fracture mechanisms. Finally, the electrical measurements were made using Megohmmeter Chauvin Arnoux (CA 6533), according to ASTM D4496. The specimens were silver-pasted to minimize the contact resistance between the composite and the electrodes. 3. Results and discussion
2. Experimental 2.1. Materials The experiments were performed with multi-walled carbon nanotubes (MWCNTs) produced through catalytic carbon vapour deposition by Nanocyl (Belgian company). This CNT material was partially functionalized with amino groups (0.5 wt.% NH2 measured by the manufacturer). Their specific surface was around 300 m2/g. The diameter of CNTs was measured close to 10 nm, formed by eight walls, measured by Transmission Electron Microscopy (TEM). The length ranges were estimated from one hundred nanometers to one micrometer. Composites were made from epoxy, diglycicyl ether of bisphenol A (DGEBA) whose equivalent weight is 178 g/epoxy equivalent, and a curing agent, 4,40 -diaminediphenylsulfone (DDS) whose equivalent weight is 62.1 g/amine hydrogen. Both compounds were purchased from Sigma–Aldrich Chemical. 2.2. Fabrication of CNT/epoxy composites According to the procedure optimized in a previous work [15], CNTs were first dispersed in chloroform, by magnetic stirring at 45 °C for 30 min, before adding the monomer epoxy. This DGEBA–CNT–chloroform mixture was stirred with high shear mechanical mixer at 45 °C for 30 min and then was sonicated for 45 min at
3.1. Morphology of composites In order to evaluate the dispersion state of the amino-functionalized CNTs in the epoxy resins, the fracture surfaces obtained by flexural test of the composites were observed by FEG–SEM microscopy. Figs. 1 and 2 show the composites with 0.25 and 0.4 wt.% CNTs, non-treated (left) and whose CNT/DGEBA mixtures were previously thermal treated at 150 °C for 1 h (right). At first glance, it is confirmed that the dispersion degree reached on the pre-cured samples is better than the non-treated ones. Large CNT agglomerates were clearly seen from the non-treated composites. These agglomerates were almost absent in the pre-cured samples, confirming that: (i) the dispersion of CNTs was enhanced significantly due to the thermal pre-curing treatment and (ii) the improved dispersion was stabilised and maintained after the epoxy resin was fully cured. Also, long nanotubes were pulled out from the matrix when the samples were not pre-cured (Fig. 1b and c), suggesting weak CNT–matrix adhesion. In contrast, the pull-out CNTs were very short in a small number (Fig 1e and f) on pre-cured composites, because the pre-curing treatment induces a strong adherence to the matrix due to the covalent bonds [9]. There are several possible fracture mechanisms of CNTs, such as pull out caused by CNT/matrix debonding in the case of weak interfacial adhesion, rupture of CNT–strong interfacial adhesion and bridging and partial debonding of interface-local bonding to the matrix enables crack bridging and interfacial failure in the non-
S.G. Prolongo et al. / Composites Science and Technology 71 (2011) 765–771
Fig. 1. FEG–SEM micrographs of epoxy non-treated (a–c) and thermal pre-cured (d–f) composites with 0.25 wt.% CNTs.
Fig. 2. FEG–SEM micrographs of epoxy non-treated (a) and thermal pre-cured (b) composites with 0.4 wt.% CNTs.
767
S.G. Prolongo et al. / Composites Science and Technology 71 (2011) 765–771
Loss modulus E´´ (MPa)
250 prec- 0.25%
200
0.4%
100
50
0
a
100
150
200
250
T (ºC) 4000
0,8
prec- 0.40%
epoxy
0.25%
prec- 0.25% prec- 0.25%
3000
0,6
0.40% 0.25% prec- 0.40%
epoxy
0.40%
2000
0
b
0,0 50
100
150
200
250
T (ºC)
The addition of carbon nanotubes should improve the thermal and mechanical performance of epoxy resins. Fig. 4 shows the variation of storage modulus, loss modulus and loss tangent with the temperature obtained from DMTA analysis. Table 1 collects the values of a-relaxation and b-relaxation temperature of the composites, measured as the maximum of the tan d curves, which were calculated by dividing the loss modulus by the corresponding storage modulus. The values of storage modulus in the glass region (at 30 °C) and in the rubbery plateau (at 250 °C) are also shown.
1,26 non-pre-cured pre-cured 1,25 3
Fig. 4. Loss modulus (a), storage modulus and loss tangent (b) of neat epoxy resin, non-treated and thermal pre-cured CNT/epoxy composites.
A strong effect of CNTs on the storage modulus of the composites in the glass region was clearly observed. The storage modulus proportionally increases with the amount of CNTs added. Further, the samples which were subjected to a previous pre-curing treatment present more pronounced increases. So, the increase on the modulus of pre-cured composite with 0.40% CNTs was 45% at room temperature regard to the one of neat epoxy resin. This effect can be attributed to the better CNT dispersion and improved interfacial interaction between amino-functionalized CNTs and epoxy matrix due to the formation of covalent bonds for the pre-curing thermal treatment. Non-clear differences were observed on the b-relaxation temperature. However, the a-relaxation temperature, which can be associated to the glass transition temperature, lightly increases
1,24 Table 1 Themo-mechanical properties of epoxy composites.
1,23
1,22 epoxy
0.25%
0,4
0,2
1000
3.2. Thermal and mechanical properties of composites
Density (g/cm )
0.25%
prec- 0.4%
150
epoxy
Loss tangent tan δ
bonded regions [10]. As observed in Fig. 1, in the pre-cured composites, the nanotubes are embedded and tightly held to the matrix and they are broken instead of bending pull out during bending test, which indicates the existence of interfacial bonding between amino-functionalized CNTs and the epoxy matrix [11]. Fig. 3 shows the changes of density of the composites as function of the amount of CNT added and the application of the precuring treatment. The changes observed must be related to the presence of nanoreinforcements because it was confirmed that all the studied nanocomposites are totally cured, as it will be demonstrated with the glass transition temperature values measured by DMTA. The influence of the CNT content on the density of composites is different on the pre-cured and non-pre-cured samples. Theoretically [3], taking into account the rule of mixtures commonly applied to traditional composite materials, the density should increase lightly with the content of CNTs, but this increase would be negligible within the experimental scatter. In fact, this is what happens with 0.25% CNT (the lower nanoreinforcement content studied). The density of the pre-cured composite with 0.4% CNT, in turn, increases more than expected. Yaping et al. observed similar results, founding that the density of composites prepared with amino-functionalized CNTs was greater than that of unmodified CNT/epoxy resin [3]. This is explained because the pre-curing process improves the dispersion of the nanoreinforcements and their chemical interaction with the matrix, occupying higher fraction of the free volume of the epoxy network and raising the compactness. On the other hand, the density of non-treated composites with high nanoreinforcement percentage drops. This is associated with the presence of defects, such as the formation of large agglomerations.
Storage modulus E´ (MPa)
768
0.4%
wt % CNT Fig. 3. Density of neat epoxy resin, non-treated and thermal pre-cured CNT/epoxy composites.
Epoxy 0.25% Non-precured 0.40% Non-precured 0.25% Pre-cured 0.40% Pre-cured
Ta (°C)
Tb (°C)
E0 (MPa) at 30 °C
E0 (MPa) at 250 °C
229 229
59 59
2.60 2.70
31 39
234
53
3.13
32
232 233
50 50
3.30 3.91
38 43
769
S.G. Prolongo et al. / Composites Science and Technology 71 (2011) 765–771
a
Flexural strength (MPa)
160 140
non-pre-cured pre-cured
120 100 80 60 40 20 0 epoxy
0.25%
0.40%
wt % CNT 7 6
Flexural strain (%)
with the CNT addition due to the reduction of the mobility of the local resin around CNTs [16]. The pre-curing treatment did not seem induce any effect in the relaxation temperatures, which can be again explained by the low number of amino-groups on the nanotubes. The peak height of loss modulus presents a continuous decrease with the percentage of CNTs added together with a widening of peak. However, for the same nanotube content, the pre-cured composites have narrower and higher loss modulus peaks. This behaviour can be associated to the improvement on the CNTs dispersion when the pre-curing treatment is applied. The good dispersed nanotubes must dissipate energy due to resistance against viscoelastic deformation of the surrounding epoxy matrix [17]. Also, the covalent bonds between amino-functionalized CNTs and epoxy improved the efficiency of load transfer from matrix to fillers, resulting in an increase in loss modulus due to more energy loss and dissipation in composites [9]. The decrease of E00 for higher nanotube contents can be interpreted by an increasing susceptibility of agglomeration, leading to less energy dissipating in the system under visco-elastic deformation. This effect has been already observed by other authors [17]. On the other hand, the tan d peaks associated to glass transition were broader and shoulders with both, CNT content and pre-curing treatment. These phenomena can be attributed to the covalent bonds between epoxy and amino-functionalized CNTs, inducing different crosslinking regions into the epoxy matrix. It is known that the mechanical properties of fibre-reinforced composites strongly depend on extend of load transfer between the matrix and filler [18]. As shown in Fig. 5, the general tendency of the flexural strength and strain is an increase by the addition of CNTs. The epoxy resin reinforced with 0.40% CNTs presents increases of 45% in both, flexural strength and strain. The flexural modulus slightly increases, from 3.4 GPa for neat epoxy resin to 3.6 GPa for the nanocomposite with 0.4 wt.% CNTs. However, at lower CNT contents, the composite with 0.25% CNT, which was subjected to pre-curing thermal treatment, shows improved mechanical properties, increasing 58% and 68% in strength and strain, respectively, regard to the pristine epoxy resin. Comparing the mechanical properties of epoxy nanocomposites with 0.25 wt.%, the thermal pre-curing treatment induces an increase of 24% and 39% in strength and strain to failure, respectively. This is indicating that the pre-curing treatment induces interfacial bonding, which enables an effective stress transfer between epoxy matrix and the amino-functionalized CNTs. This also means that the addition of amino-functionalized CNTs into thermoset polymer can increase the fracture energy because the crack propagation can be blunted by bringing up the crack faces of nanotubes [18]. It is interesting to stand out the important increase of the strain of composites regard to the brittle epoxy resin. Most current toughening methods, e.g. rubber and thermoplastic blends, can effectively increase the toughness, but with sacrifice of the mechanical strength [19]. Also, with strong covalent bonding, the nanotubes offer extra benefits to increase the strain to failure. In conventional fibre-reinforced composites, the strain usually drastically drops. However, carbon nanotubes present a particular reinforcement with high aspect ratio and highly flexible elastic behaviour during load, which are very different from micrometer-size fibres [19]. Additionally, the curved CNTs are typically twisted and entangled and can therefore continuously stretched. By means of strong interfacial bonding with crosslinked matrix, promoted by the pre-curing treatment, such behaviour contributes to continuous absorption of energy and results in increased strain in the epoxy thermoset. Taking into account experimental scatter, the pre-cured treatment seems not so effective at the maximum CNT content studied (0.4 wt.%). In fact, the strength and strain to failure for these nano-
b
non-pre-cured pre-cured
5 4 3 2 1 0 epoxy
0.25%
0.40%
wt % CNT Fig. 5. Ultimate flexural strength and strain of neat epoxy resin, non-treated and thermal pre-cured CNT/epoxy composites.
composites are lower than the ones of pre-cured sample with 0.25 wt.%. This is due to the presence of defects associated to the high nanoreinforcement content. This allows us to state that in the manufacture of this kind of nanocomposites, it is more important to reach a good dispersion degree and a right load transfer between nanoreinforcement and matrix than the increase of the added content of nanotubes. The fracture surfaces of epoxy composites are shown in Fig. 6, which were captured by SEM. The epoxy resin presents the typical characteristics of brittle fracture. The surface is smooth, with very directionally deformation lines, which means that the crack propagated is not interrupted. However, the fracture surfaces of composites seem be rougher. The ridge patterns and rivers marks indicate that the composites presented a more ductile fracture. The pre-cured composites, specially that reinforced with 0.25 wt.% CNT, also show changes of fracture planes, the crack deviated from its original plane, indicating crack path deflection. This means that the energy required for the propagation of the crack increased. This is in accordance to the obtained results on mechanical characterisation (Fig. 5), where it was concluded that the sample with best mechanical properties (with highest strength and strain to failure) was the pre-cured one with 0.25 wt.% CNT. 3.3. Electrical properties of composites The epoxy resins are electrical insulator materials, whose electrical resistivity reaches values up to 1017–1018 X cm. When its application requires it, in order to decrease its very high electrical
770
S.G. Prolongo et al. / Composites Science and Technology 71 (2011) 765–771
Fig. 6. SEM micrographs of neat epoxy resin (a), non-treated composites with 0.25 (b) and 0.40% (c) and thermal pre-cured composites with 0.25 (d) and 0.4 wt.% CNTs (e).
4. Conclusions This study has demonstrated that the application of previous thermal treatment on amino-functionalized CNT/epoxy mixtures,
4
9,0x10
Electrical resistivity (ohm.cm)
resistivity, the epoxy resins are commonly filled with electrical particles, such as carbon black or silver pellets. These fillers do not increase generally the mechanical properties. For it, one of the alternative approaches more studied in the last years, it is the addition of carbon nanoreinforcement due to the well known high electrical conductivity of the carbon nanotubes. The electrical resistivity of the composites was measured as function of CNT content and the pre-curing treatment applied. Fig. 7 shows the obtained results. As it can be seen, the electrical resistivity drastically drops with the addition of CNTs, in more than 13 magnitude orders. But this fall does not depend on the nanotube contents. This means that the studied nanoreinforcement contents are higher than the percolation threshold, which is the minimum filler concentration from which a three dimensional conductive network is formed and the electrical resistivity falls drastically. The thermal pre-curing treatment decreases even more the electrical resistivity of the composites. This could be associated to the better dispersion of the nanometric reinforcements which avoids the appearance of agglomerates, which could act as impediments to the flow of electrical current.
non-pre-cured pre-cured
4
6,0x10
4
3,0x10
0,0
0.25%
0.4%
wt % CNT Fig. 7. Electrical resistivity of non-treated and thermal pre-cured CNT/epoxy composites.
before adding the curing agent, improves the formation of strong interface. The pre-cured composites present better nanotube dispersion states. Both effects result in much improved mechanical properties and a light increase of the electrical conductivity of
S.G. Prolongo et al. / Composites Science and Technology 71 (2011) 765–771
CNT/epoxy composites compared to non-treated ones containing the same amount of nanotubes. Acknowledgements The authors want to thank the participation of Marta Boada Collado and the financial support of Ministerio de Ciencia e Innovación of Spain (Project MAT2007-61178) and Consejería de Educación of Comunidad de Madrid (programme S2009MAT/1585). M.R. Gude thanks Consejería de Educación de la Comunidad de Madrid (Spain) and Fondo Social Europeo for awarding a research contract. References [1] Li C, Thostenson ET, Chou TW. Sensors and actuators based on carbon nanotubes and their composites: a review. Compos Sci Technol 2008;68:1227–49. [2] Alexopoulos ND, Bartholome C, Poulin P, Marioli-Riga Z. Structural health monitoring of glass fiber reinforced composites using embedded carbon nanotube (CNT) fibers. Compos Sci Technol 2010;70:260–71. [3] Yaping Z, Aibo Z, Quinghua C, Jiaoxia Z, Rongchang N. Functionalized effect on carbon nanotube/epoxy nano-composites. Mater Sci Eng A 2006;435– 436:145–9. [4] Xie X, Mai YW, Zhou XP. Dispersion and alignment of carbon nanotubes in polymer matrix: a review. Mater Sci Eng 2005;49:89–112. [5] Sahoo NG, Rana S, Cho JW, Li L, Chan SH. Polymer nanocomposites based on functionalized carbon nanotubes. Prog Polym Sci 2010;35:837–67. [6] Chen W, Lu H, Nutt SR. The influence of functionalized MWCNT reinforcement on the thermomechanical properties and morphology of epoxy nanocomposites. Compos Sci Technol 2008;68:2535–42. [7] Fiedler B, Gojny FH, Wichmann MHG, Nolte MCM, Schulte K. Fundamental aspects of nano-reinforced composites. Compos Sci Technol 2006;16:3115–25.
771
[8] Nadler M, Werner J, Mahrholz T, Riedel U, Hufenbach W. Effect of CNT surface funcionalisation on the mechanical properties of multi-walled carbon nanotube epoxy-composites. Composites A 2009;40:932–7. [9] Ma PC, Mo SY, Tang BZ, Kim JK. Dispersion, interfacial interaction and reagglomeration of functionalized carbon nanotubes in epoxy composites. Carbon 2010;48:1824–34. [10] Gojny FH, Wichmann MHG, Fiedler B, Schulte K. Influence of different carbon nanotubes on the mechanical properties of epoxy matrix composites – a comparative study. Compos Sci Technol 2005;65:2300–13. [11] Shen J, Huang W, Wu L, Hu Y, Ye M. The reinforcement role of different aminofunctionalized multi-walled carbon nanotubes in epoxy nanocomposites. Compos Sci Technol 2007;67:3041–50. [12] Abdalla M, Dean D, Adibempe D, Nyairo E, Robinson P, Thompson G. The effect of interfacial chemistry on molecular mobility and morphology of multiwalled carbon nanotubes epoxy nanocomposite. Polymer 2007;48:5662–70. [13] Prolongo SG, Gude MR, Ureña A. Rheological behaviour of nanoreinforced epoxy adhesives of low electrical resistivity for joining carbon fiber/epoxy laminates. J Adhes Sci Technol 2010;24:1097–112. [14] Prolongo SG, Gude MR, Ureña A. The curing process of epoxy/aminofunctionalized MWCNTs: calorimetry, molecular modelling, and electron microscopy. J Nanotechnol 2010 [article ID 420432]. [15] Prolongo SG, Burón M, Gude MR, Chaos-Morán R, Campo M, Ureña A. Effects of dispersion techniques of carbon nanofibers on the thermo-physical properties of epoxy nanocomposites. Compos Sci Technol 2008;68:2722–30. [16] Ramanathan T, Liu H, Brinson LC. Functionalized SWNT/polymer nanocomposites for dramatic property improvement. J Polym Sci: Part B: Polym Phys 2005;43:2269–79. [17] Gojny FH, Schulte K. Functionalisation effect on the thermo-mechanical behaviour of multi-wall carbon nanotube/epoxy composites. Compos Sci Technol 2004;64:2303–8. [18] Shen J, Huang W, Wu L, Hu Y, Ye M. Thermo-physical properties of epoxy nanocomposites reinforced with amino-functionalized multi-walled carbon nanotubes. Composites A 2007;38:1331–6. [19] Zhu J, Peng H, Macias FR, Margrave JL, Khabashesku VN, Imam AM, et al. Reinforcing epoxy polymer composites through covalent integration of functionalized nanotubes. Adv Funct Mater 2004;17:643–8.
Contents lists available at ScienceDirect
Composites Science and Technology journal homepage: www.elsevier.com/locate/compscitech
Improving the flexural and thermomechanical properties of amino-functionalized carbon nanotube/epoxy composites by using a pre-curing treatment S.G. Prolongo ⇑, M.R. Gude, A. Ureña Dpt. Materials Science and Engineering, ESCET, University Rey Juan Carlos, c/Tulipán s/n 28933 Móstoles, Madrid, Spain
a r t i c l e
i n f o
Article history: Received 29 September 2010 Received in revised form 24 January 2011 Accepted 30 January 2011 Available online 20 February 2011 Keywords: A. Nanocomposites B. Mechanical properties B. Electrical properties B. Thermomechanical properties
a b s t r a c t A prior thermal (pre-curing) treatment of mixtures of epoxy monomer and amino-functionalized carbon nanotubes (CNTs) was used to promote a chemical reaction between the matrix and the reinforcement, favouring the formation of a strong interface. Samples of epoxy resin and different weight percentages of amino-functionalized multi-walled CNTs were prepared with and without the pre-curing treatment (150 °C, 1 h). The degree of dispersion of the nanofiller was better when this pre-curing treatment was used. This allowed a higher CNT content while keeping a high sample homogeneity. Without the pre-curing step, the addition of CNTs increases both the flexural strength and strain to failure by 45%. Moreover, with the pre-curing step, the nanocomposite with 0.25 wt.% CNTs presents an increase of flexural strength by 58% and strain to failure by 68% regard to neat epoxy resin. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Composites reinforced with CNTs are being widely researched for over a decade. Dramatic improvements are expected in their thermal, mechanical and electrical behaviour. Epoxy resins are well established as thermosetting matrices of advanced composites displaying a series of promising characteristics for a wide range of applications. So the incorporation of CNTs into epoxy resins presents numerous potential applications, including aerospace and automobile materials, due to the expected increase of their use temperature, their low density and their improved mechanical properties. Besides, the development of CNT/epoxy composites opens new perspectives for multi-functional materials, e.g., conductive polymers with improved mechanical performance and with a perspective of damage sensing and ‘‘life’’-monitoring [1,2]. Despite the promising developments for CNT/epoxy composites, the inherent processing difficulties, such as dispersion and interphase control, have caused mechanical performance to fall short of theoretical predictions. The most common procedure used to overcome this problem is to try to get an effective CNT surface grafting, addressing both issues, dispersion and interface stress transfer [3–5]. Numerous grafting methods have been employed to introduce a variety of molecules onto CNTs [6–8]. One of them is based on amino-functionalization of CNTs. Kim et al. [9] confirmed that the amino-functionalized CNTs exhibit higher surface energy and much better wettability with epoxy resin than the pristine CNTs, inhibiting the nanotube re-agglomeration during the ⇑ Corresponding author. Fax: +34 91 4888150. E-mail address: [email protected] (S.G. Prolongo). 0266-3538/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.compscitech.2011.01.028
curing of the resin. In fact, several works confirm the better thermal and mechanical performance of epoxy composites reinforced with amino-functionalized CNTs when compared to the untreated ones [9–11]. However, direct evidences of covalent interactions between the modified CNT and epoxies have not been reported. This is presumably due in part to the low concentration of functional groups on the CNT surface when dispersed into epoxy resin/curing agent mixture which makes it difficult to pinpoint the reaction. Abdalla et al. [12] confirmed that the differences in the way the epoxy agent reacts with the modified nanotubes can result in steric and electronic differences that can in turn affect the cure and this invariably translate to differences in the crosslinking topology and hence the macroscale properties. When the curing reaction is carried out on a mixture of epoxy monomer and amino curing agent with dispersed amino-functionalized CNTs, both kind of amine groups (from curing agent and from CNTs) compete to react with epoxide groups. Taking into account the low concentration of CNT added and the low percentage of amine groups linked over their surface, together with steric hindrances and diffusion problems of rigid nanotubes, it is obvious that the chemical reaction with the curing agent is more favoured. This implies that the amount of covalent bonds between epoxy matrix and amino-functionalized CNTs must be low. Therefore, we have investigated the effect of inducing the chemical reaction between amino-functionalized CNTs with the epoxy monomer before curing process, in order to assure the formation of covalent linkages at the interface, which facilities stress transfer on the composite. For this, a previous thermal treatment, named pre-curing, is applied to the mixture of functionalized CNTs and epoxy monomer, before the addition of the curing agent.
766
S.G. Prolongo et al. / Composites Science and Technology 71 (2011) 765–771
In previous works, authors have tried to find experimental evidences, by indirect methods, of the chemical reaction between amino groups anchored on CNTs and the oxirane rings of the epoxy monomer when the pre-curing treatment was applied. One of these methods was the measurement of the viscosity of mixtures of the epoxy monomer with different contents of amino-functionalized carbon nanotubes before and after applying the pre-curing treatment [13]. For low CNT contents the viscosity was almost the same irrespective of whether pre-curing treatment was carried out or not. The main reason is the very low concentration of amine groups, close to 2000:1 (epoxy:hydrogen amine). However for 0.5 wt.% CNTs, the viscosity was approximately 70% higher for the pre-cured mixtures regard to non-treated ones, suggesting stronger interfacial epoxy–nanotube interactions because of the chemical reaction epoxy-amine. Differential scanning calorimetry (DSC) was used with the aim of determining the enthalpy of the pre-curing reaction (epoxy monomer/CNT mixtures), but due again to the low amount of functional groups on the CNTs, any exothermic peak was not detected. However, a small change was observed in the glass transition temperature (Tg) of the epoxy monomer from 18 to 16 °C for precured CNT/epoxy mixtures, indicating that chemical reaction between amino-functionalized CNT and epoxy rings of the monomer had occurred. Also, this was confirmed indirectly determining the curing enthalpy of the epoxy monomer/hardener/CNT mixtures pre-cured and without any thermal treatment. In this case, the heat of reaction was 2–3 kJ/epoxy equivalent lower for pre-cured composites. This could mean that some of the oxirane rings of the epoxy monomer have previously reacted, during the pre-curing treatment, and thus they are not available for the curing with the hardener [14]. Once it was proven the effectiveness of the pre-curing treatment in promoting the chemical reaction between the epoxy monomer and the amino-functionalized multi-walled carbon nanotubes, in this work we study its effect on the properties of the CNT/epoxy composites. A stronger interface between the nanofiller and the matrix should improve the load transfer, increasing the mechanical properties.
the same temperature. The resulting mixture was degassed at 90 °C for more than 12 h in order to eliminate the remaining chloroform and the trapped air. Half of the samples were subjected to pre-curing treatment, involving of heating at 150 °C for 1 h. This treatment was carried out with mechanical stirring at approximately 150 rpm. Then the hardener was added in stoichiometric ratio. The composite resin was molded in a steel mould and cured at 210 °C for 3 h. Composites, pre-cured and non-pre-cured, containing different weight fraction (i.e. 0, 0.25 and 0.40 wt.%) of amino-functionalized CNTs were prepared. 2.3. Characterisation In order to qualitatively determine the dispersion of CNTs, the composites were observed by Field Emission Gun Scanning (FEG– SEM) Electron Microscopy (Nova NanoSEM FEI 230). The surface of the samples was sputter coated by a thin layer (5–10 nm) of Au (Pd). The experimental conditions of the sputtering were 30 mA and 120 s (Bal-tec, SCD-005 sputter). A Mettler Toledo balance, with ±0.001 mg, equipped with a density determination kit by means the buoyancy technique, was used to evaluate the change of density in composites. Ten measurements of density for each sample were carried out. Thermal dynamic mechanical properties were determined with a TA Instrument DMTA Q800 operating in single cantilever mode with an oscillation frequency of 1.0 Hz. Data were collected from 100 °C to 300 °C at a scanning rate of 2 °C/min. The dimensions of DMTA specimens were 1.5 15 35 mm3. The mechanical characterisation of composites was carried out by flexural test (Instron 4465), following the ASTM D-790 at a crosshead speed of 0.8 mm/min. The fracture surfaces were also covered with Au (Pd) and observed by Scanning Electron Microscopy (ESEM, Philips XL30) in order to study the fracture mechanisms. Finally, the electrical measurements were made using Megohmmeter Chauvin Arnoux (CA 6533), according to ASTM D4496. The specimens were silver-pasted to minimize the contact resistance between the composite and the electrodes. 3. Results and discussion
2. Experimental 2.1. Materials The experiments were performed with multi-walled carbon nanotubes (MWCNTs) produced through catalytic carbon vapour deposition by Nanocyl (Belgian company). This CNT material was partially functionalized with amino groups (0.5 wt.% NH2 measured by the manufacturer). Their specific surface was around 300 m2/g. The diameter of CNTs was measured close to 10 nm, formed by eight walls, measured by Transmission Electron Microscopy (TEM). The length ranges were estimated from one hundred nanometers to one micrometer. Composites were made from epoxy, diglycicyl ether of bisphenol A (DGEBA) whose equivalent weight is 178 g/epoxy equivalent, and a curing agent, 4,40 -diaminediphenylsulfone (DDS) whose equivalent weight is 62.1 g/amine hydrogen. Both compounds were purchased from Sigma–Aldrich Chemical. 2.2. Fabrication of CNT/epoxy composites According to the procedure optimized in a previous work [15], CNTs were first dispersed in chloroform, by magnetic stirring at 45 °C for 30 min, before adding the monomer epoxy. This DGEBA–CNT–chloroform mixture was stirred with high shear mechanical mixer at 45 °C for 30 min and then was sonicated for 45 min at
3.1. Morphology of composites In order to evaluate the dispersion state of the amino-functionalized CNTs in the epoxy resins, the fracture surfaces obtained by flexural test of the composites were observed by FEG–SEM microscopy. Figs. 1 and 2 show the composites with 0.25 and 0.4 wt.% CNTs, non-treated (left) and whose CNT/DGEBA mixtures were previously thermal treated at 150 °C for 1 h (right). At first glance, it is confirmed that the dispersion degree reached on the pre-cured samples is better than the non-treated ones. Large CNT agglomerates were clearly seen from the non-treated composites. These agglomerates were almost absent in the pre-cured samples, confirming that: (i) the dispersion of CNTs was enhanced significantly due to the thermal pre-curing treatment and (ii) the improved dispersion was stabilised and maintained after the epoxy resin was fully cured. Also, long nanotubes were pulled out from the matrix when the samples were not pre-cured (Fig. 1b and c), suggesting weak CNT–matrix adhesion. In contrast, the pull-out CNTs were very short in a small number (Fig 1e and f) on pre-cured composites, because the pre-curing treatment induces a strong adherence to the matrix due to the covalent bonds [9]. There are several possible fracture mechanisms of CNTs, such as pull out caused by CNT/matrix debonding in the case of weak interfacial adhesion, rupture of CNT–strong interfacial adhesion and bridging and partial debonding of interface-local bonding to the matrix enables crack bridging and interfacial failure in the non-
S.G. Prolongo et al. / Composites Science and Technology 71 (2011) 765–771
Fig. 1. FEG–SEM micrographs of epoxy non-treated (a–c) and thermal pre-cured (d–f) composites with 0.25 wt.% CNTs.
Fig. 2. FEG–SEM micrographs of epoxy non-treated (a) and thermal pre-cured (b) composites with 0.4 wt.% CNTs.
767
S.G. Prolongo et al. / Composites Science and Technology 71 (2011) 765–771
Loss modulus E´´ (MPa)
250 prec- 0.25%
200
0.4%
100
50
0
a
100
150
200
250
T (ºC) 4000
0,8
prec- 0.40%
epoxy
0.25%
prec- 0.25% prec- 0.25%
3000
0,6
0.40% 0.25% prec- 0.40%
epoxy
0.40%
2000
0
b
0,0 50
100
150
200
250
T (ºC)
The addition of carbon nanotubes should improve the thermal and mechanical performance of epoxy resins. Fig. 4 shows the variation of storage modulus, loss modulus and loss tangent with the temperature obtained from DMTA analysis. Table 1 collects the values of a-relaxation and b-relaxation temperature of the composites, measured as the maximum of the tan d curves, which were calculated by dividing the loss modulus by the corresponding storage modulus. The values of storage modulus in the glass region (at 30 °C) and in the rubbery plateau (at 250 °C) are also shown.
1,26 non-pre-cured pre-cured 1,25 3
Fig. 4. Loss modulus (a), storage modulus and loss tangent (b) of neat epoxy resin, non-treated and thermal pre-cured CNT/epoxy composites.
A strong effect of CNTs on the storage modulus of the composites in the glass region was clearly observed. The storage modulus proportionally increases with the amount of CNTs added. Further, the samples which were subjected to a previous pre-curing treatment present more pronounced increases. So, the increase on the modulus of pre-cured composite with 0.40% CNTs was 45% at room temperature regard to the one of neat epoxy resin. This effect can be attributed to the better CNT dispersion and improved interfacial interaction between amino-functionalized CNTs and epoxy matrix due to the formation of covalent bonds for the pre-curing thermal treatment. Non-clear differences were observed on the b-relaxation temperature. However, the a-relaxation temperature, which can be associated to the glass transition temperature, lightly increases
1,24 Table 1 Themo-mechanical properties of epoxy composites.
1,23
1,22 epoxy
0.25%
0,4
0,2
1000
3.2. Thermal and mechanical properties of composites
Density (g/cm )
0.25%
prec- 0.4%
150
epoxy
Loss tangent tan δ
bonded regions [10]. As observed in Fig. 1, in the pre-cured composites, the nanotubes are embedded and tightly held to the matrix and they are broken instead of bending pull out during bending test, which indicates the existence of interfacial bonding between amino-functionalized CNTs and the epoxy matrix [11]. Fig. 3 shows the changes of density of the composites as function of the amount of CNT added and the application of the precuring treatment. The changes observed must be related to the presence of nanoreinforcements because it was confirmed that all the studied nanocomposites are totally cured, as it will be demonstrated with the glass transition temperature values measured by DMTA. The influence of the CNT content on the density of composites is different on the pre-cured and non-pre-cured samples. Theoretically [3], taking into account the rule of mixtures commonly applied to traditional composite materials, the density should increase lightly with the content of CNTs, but this increase would be negligible within the experimental scatter. In fact, this is what happens with 0.25% CNT (the lower nanoreinforcement content studied). The density of the pre-cured composite with 0.4% CNT, in turn, increases more than expected. Yaping et al. observed similar results, founding that the density of composites prepared with amino-functionalized CNTs was greater than that of unmodified CNT/epoxy resin [3]. This is explained because the pre-curing process improves the dispersion of the nanoreinforcements and their chemical interaction with the matrix, occupying higher fraction of the free volume of the epoxy network and raising the compactness. On the other hand, the density of non-treated composites with high nanoreinforcement percentage drops. This is associated with the presence of defects, such as the formation of large agglomerations.
Storage modulus E´ (MPa)
768
0.4%
wt % CNT Fig. 3. Density of neat epoxy resin, non-treated and thermal pre-cured CNT/epoxy composites.
Epoxy 0.25% Non-precured 0.40% Non-precured 0.25% Pre-cured 0.40% Pre-cured
Ta (°C)
Tb (°C)
E0 (MPa) at 30 °C
E0 (MPa) at 250 °C
229 229
59 59
2.60 2.70
31 39
234
53
3.13
32
232 233
50 50
3.30 3.91
38 43
769
S.G. Prolongo et al. / Composites Science and Technology 71 (2011) 765–771
a
Flexural strength (MPa)
160 140
non-pre-cured pre-cured
120 100 80 60 40 20 0 epoxy
0.25%
0.40%
wt % CNT 7 6
Flexural strain (%)
with the CNT addition due to the reduction of the mobility of the local resin around CNTs [16]. The pre-curing treatment did not seem induce any effect in the relaxation temperatures, which can be again explained by the low number of amino-groups on the nanotubes. The peak height of loss modulus presents a continuous decrease with the percentage of CNTs added together with a widening of peak. However, for the same nanotube content, the pre-cured composites have narrower and higher loss modulus peaks. This behaviour can be associated to the improvement on the CNTs dispersion when the pre-curing treatment is applied. The good dispersed nanotubes must dissipate energy due to resistance against viscoelastic deformation of the surrounding epoxy matrix [17]. Also, the covalent bonds between amino-functionalized CNTs and epoxy improved the efficiency of load transfer from matrix to fillers, resulting in an increase in loss modulus due to more energy loss and dissipation in composites [9]. The decrease of E00 for higher nanotube contents can be interpreted by an increasing susceptibility of agglomeration, leading to less energy dissipating in the system under visco-elastic deformation. This effect has been already observed by other authors [17]. On the other hand, the tan d peaks associated to glass transition were broader and shoulders with both, CNT content and pre-curing treatment. These phenomena can be attributed to the covalent bonds between epoxy and amino-functionalized CNTs, inducing different crosslinking regions into the epoxy matrix. It is known that the mechanical properties of fibre-reinforced composites strongly depend on extend of load transfer between the matrix and filler [18]. As shown in Fig. 5, the general tendency of the flexural strength and strain is an increase by the addition of CNTs. The epoxy resin reinforced with 0.40% CNTs presents increases of 45% in both, flexural strength and strain. The flexural modulus slightly increases, from 3.4 GPa for neat epoxy resin to 3.6 GPa for the nanocomposite with 0.4 wt.% CNTs. However, at lower CNT contents, the composite with 0.25% CNT, which was subjected to pre-curing thermal treatment, shows improved mechanical properties, increasing 58% and 68% in strength and strain, respectively, regard to the pristine epoxy resin. Comparing the mechanical properties of epoxy nanocomposites with 0.25 wt.%, the thermal pre-curing treatment induces an increase of 24% and 39% in strength and strain to failure, respectively. This is indicating that the pre-curing treatment induces interfacial bonding, which enables an effective stress transfer between epoxy matrix and the amino-functionalized CNTs. This also means that the addition of amino-functionalized CNTs into thermoset polymer can increase the fracture energy because the crack propagation can be blunted by bringing up the crack faces of nanotubes [18]. It is interesting to stand out the important increase of the strain of composites regard to the brittle epoxy resin. Most current toughening methods, e.g. rubber and thermoplastic blends, can effectively increase the toughness, but with sacrifice of the mechanical strength [19]. Also, with strong covalent bonding, the nanotubes offer extra benefits to increase the strain to failure. In conventional fibre-reinforced composites, the strain usually drastically drops. However, carbon nanotubes present a particular reinforcement with high aspect ratio and highly flexible elastic behaviour during load, which are very different from micrometer-size fibres [19]. Additionally, the curved CNTs are typically twisted and entangled and can therefore continuously stretched. By means of strong interfacial bonding with crosslinked matrix, promoted by the pre-curing treatment, such behaviour contributes to continuous absorption of energy and results in increased strain in the epoxy thermoset. Taking into account experimental scatter, the pre-cured treatment seems not so effective at the maximum CNT content studied (0.4 wt.%). In fact, the strength and strain to failure for these nano-
b
non-pre-cured pre-cured
5 4 3 2 1 0 epoxy
0.25%
0.40%
wt % CNT Fig. 5. Ultimate flexural strength and strain of neat epoxy resin, non-treated and thermal pre-cured CNT/epoxy composites.
composites are lower than the ones of pre-cured sample with 0.25 wt.%. This is due to the presence of defects associated to the high nanoreinforcement content. This allows us to state that in the manufacture of this kind of nanocomposites, it is more important to reach a good dispersion degree and a right load transfer between nanoreinforcement and matrix than the increase of the added content of nanotubes. The fracture surfaces of epoxy composites are shown in Fig. 6, which were captured by SEM. The epoxy resin presents the typical characteristics of brittle fracture. The surface is smooth, with very directionally deformation lines, which means that the crack propagated is not interrupted. However, the fracture surfaces of composites seem be rougher. The ridge patterns and rivers marks indicate that the composites presented a more ductile fracture. The pre-cured composites, specially that reinforced with 0.25 wt.% CNT, also show changes of fracture planes, the crack deviated from its original plane, indicating crack path deflection. This means that the energy required for the propagation of the crack increased. This is in accordance to the obtained results on mechanical characterisation (Fig. 5), where it was concluded that the sample with best mechanical properties (with highest strength and strain to failure) was the pre-cured one with 0.25 wt.% CNT. 3.3. Electrical properties of composites The epoxy resins are electrical insulator materials, whose electrical resistivity reaches values up to 1017–1018 X cm. When its application requires it, in order to decrease its very high electrical
770
S.G. Prolongo et al. / Composites Science and Technology 71 (2011) 765–771
Fig. 6. SEM micrographs of neat epoxy resin (a), non-treated composites with 0.25 (b) and 0.40% (c) and thermal pre-cured composites with 0.25 (d) and 0.4 wt.% CNTs (e).
4. Conclusions This study has demonstrated that the application of previous thermal treatment on amino-functionalized CNT/epoxy mixtures,
4
9,0x10
Electrical resistivity (ohm.cm)
resistivity, the epoxy resins are commonly filled with electrical particles, such as carbon black or silver pellets. These fillers do not increase generally the mechanical properties. For it, one of the alternative approaches more studied in the last years, it is the addition of carbon nanoreinforcement due to the well known high electrical conductivity of the carbon nanotubes. The electrical resistivity of the composites was measured as function of CNT content and the pre-curing treatment applied. Fig. 7 shows the obtained results. As it can be seen, the electrical resistivity drastically drops with the addition of CNTs, in more than 13 magnitude orders. But this fall does not depend on the nanotube contents. This means that the studied nanoreinforcement contents are higher than the percolation threshold, which is the minimum filler concentration from which a three dimensional conductive network is formed and the electrical resistivity falls drastically. The thermal pre-curing treatment decreases even more the electrical resistivity of the composites. This could be associated to the better dispersion of the nanometric reinforcements which avoids the appearance of agglomerates, which could act as impediments to the flow of electrical current.
non-pre-cured pre-cured
4
6,0x10
4
3,0x10
0,0
0.25%
0.4%
wt % CNT Fig. 7. Electrical resistivity of non-treated and thermal pre-cured CNT/epoxy composites.
before adding the curing agent, improves the formation of strong interface. The pre-cured composites present better nanotube dispersion states. Both effects result in much improved mechanical properties and a light increase of the electrical conductivity of
S.G. Prolongo et al. / Composites Science and Technology 71 (2011) 765–771
CNT/epoxy composites compared to non-treated ones containing the same amount of nanotubes. Acknowledgements The authors want to thank the participation of Marta Boada Collado and the financial support of Ministerio de Ciencia e Innovación of Spain (Project MAT2007-61178) and Consejería de Educación of Comunidad de Madrid (programme S2009MAT/1585). M.R. Gude thanks Consejería de Educación de la Comunidad de Madrid (Spain) and Fondo Social Europeo for awarding a research contract. References [1] Li C, Thostenson ET, Chou TW. Sensors and actuators based on carbon nanotubes and their composites: a review. Compos Sci Technol 2008;68:1227–49. [2] Alexopoulos ND, Bartholome C, Poulin P, Marioli-Riga Z. Structural health monitoring of glass fiber reinforced composites using embedded carbon nanotube (CNT) fibers. Compos Sci Technol 2010;70:260–71. [3] Yaping Z, Aibo Z, Quinghua C, Jiaoxia Z, Rongchang N. Functionalized effect on carbon nanotube/epoxy nano-composites. Mater Sci Eng A 2006;435– 436:145–9. [4] Xie X, Mai YW, Zhou XP. Dispersion and alignment of carbon nanotubes in polymer matrix: a review. Mater Sci Eng 2005;49:89–112. [5] Sahoo NG, Rana S, Cho JW, Li L, Chan SH. Polymer nanocomposites based on functionalized carbon nanotubes. Prog Polym Sci 2010;35:837–67. [6] Chen W, Lu H, Nutt SR. The influence of functionalized MWCNT reinforcement on the thermomechanical properties and morphology of epoxy nanocomposites. Compos Sci Technol 2008;68:2535–42. [7] Fiedler B, Gojny FH, Wichmann MHG, Nolte MCM, Schulte K. Fundamental aspects of nano-reinforced composites. Compos Sci Technol 2006;16:3115–25.
771
[8] Nadler M, Werner J, Mahrholz T, Riedel U, Hufenbach W. Effect of CNT surface funcionalisation on the mechanical properties of multi-walled carbon nanotube epoxy-composites. Composites A 2009;40:932–7. [9] Ma PC, Mo SY, Tang BZ, Kim JK. Dispersion, interfacial interaction and reagglomeration of functionalized carbon nanotubes in epoxy composites. Carbon 2010;48:1824–34. [10] Gojny FH, Wichmann MHG, Fiedler B, Schulte K. Influence of different carbon nanotubes on the mechanical properties of epoxy matrix composites – a comparative study. Compos Sci Technol 2005;65:2300–13. [11] Shen J, Huang W, Wu L, Hu Y, Ye M. The reinforcement role of different aminofunctionalized multi-walled carbon nanotubes in epoxy nanocomposites. Compos Sci Technol 2007;67:3041–50. [12] Abdalla M, Dean D, Adibempe D, Nyairo E, Robinson P, Thompson G. The effect of interfacial chemistry on molecular mobility and morphology of multiwalled carbon nanotubes epoxy nanocomposite. Polymer 2007;48:5662–70. [13] Prolongo SG, Gude MR, Ureña A. Rheological behaviour of nanoreinforced epoxy adhesives of low electrical resistivity for joining carbon fiber/epoxy laminates. J Adhes Sci Technol 2010;24:1097–112. [14] Prolongo SG, Gude MR, Ureña A. The curing process of epoxy/aminofunctionalized MWCNTs: calorimetry, molecular modelling, and electron microscopy. J Nanotechnol 2010 [article ID 420432]. [15] Prolongo SG, Burón M, Gude MR, Chaos-Morán R, Campo M, Ureña A. Effects of dispersion techniques of carbon nanofibers on the thermo-physical properties of epoxy nanocomposites. Compos Sci Technol 2008;68:2722–30. [16] Ramanathan T, Liu H, Brinson LC. Functionalized SWNT/polymer nanocomposites for dramatic property improvement. J Polym Sci: Part B: Polym Phys 2005;43:2269–79. [17] Gojny FH, Schulte K. Functionalisation effect on the thermo-mechanical behaviour of multi-wall carbon nanotube/epoxy composites. Compos Sci Technol 2004;64:2303–8. [18] Shen J, Huang W, Wu L, Hu Y, Ye M. Thermo-physical properties of epoxy nanocomposites reinforced with amino-functionalized multi-walled carbon nanotubes. Composites A 2007;38:1331–6. [19] Zhu J, Peng H, Macias FR, Margrave JL, Khabashesku VN, Imam AM, et al. Reinforcing epoxy polymer composites through covalent integration of functionalized nanotubes. Adv Funct Mater 2004;17:643–8.