Mechanical and thermal properties of functionalized multiwalled carbon nanotubes/cyanate ester composite.

INTRODUCTION

Cyanate ester resins (CEs) are among the most important engineering thermosetting polymers and have received attentions because of their outstanding physical properties such as low water absorptivity and outgassing, excellent mechanical properties, dimensional and thermal stability, and flame resistance (1-4). Unfortunately, highly crosslinked thermosets, such as CEs, tend to be brittle and have reduced impact resistance. To improve the toughness of a cured CE, a number of different modifiers have been used. These additives include reactive and nonreactive rubbers (5), (6), a variety of engineering thermoplastics (7-9) and some kind of thermosets such as epoxy and bismaleimide (10-15). All of these modifiers have proven useful to some extent. However, toughening usually occurs at the cost of other characteristics of CEs. For example, rubbers can improve the toughness of CEs largely, whereas the thermal property is generally reduced (16). Although a few examples of utilization of thermoplastic modifiers to improve mechanical properties without sacrificing thermal stability, most of them need special solvents in fabrication processes, which will unavoidably affect the final properties of the cured resin and add difficulty to the fabrication operation (17).

Since the first observation of the carbon nanotubes (CNTs) in 1991 by Iijima, the awareness of their unique mechanical properties, such as extremely high strength and stiffness and enormous aspect ratio, make them potentially excellent reinforcing fillers of polymer materials (18), (19). A small addition of CNTs can improve the mechanical property of polymer significantly with little side-effect on other properties. However, there are two problems which have to be overcome in order to achieve the goal. One is the lack of interfacial adhesion, which is critical for load transfer in composites. Indeed. CNTs' surfaces are atomically smooth, which may limit the transfer of load from the matrix to nanotubes reinforcement. Another is the poor dispersion of nanotubes in the polymer matrix, which is also significant for the fabrication of reinforced nanocomposite.

Functionalization of CNTs by covalent molecules is the most popular method to improve the compatibility and the interaction with the polymer matrixes. Up to now, various functional groups have been attached onto the convex surfaces of CNTs via covalent bonding (20). In general, major approaches include: (i) amidation or esteri-fication of carboxylated CNTs, (ii) side-wall covalent attachment of functional groups directly to the pristine CNTs. The chemical group that should be attached to the CNTs depends on the nature of the polymer to be reinforced, and the CNTs functionalized with epoxy groups are very important to cyanate ester modification, as epoxy groups can react with cyanate ester to form oxazlalidinone structure with high mechanical and thermal properties (21). In this article, we are mainly concerned with the functionalization of the multiwalled carbon nanotubes (MWNTs) with 3-glycidyloxypropyltrimethoxy si lane (coupling agent, KH-560) and the properties of the functionalized MWNTs/CE composites.

EXPERIMENTAL

Materials

The pristine MWNTs (p-MWNTs) were obtained from a commercial source (Shenzhen Nanotech Port, China). They were produced by chemical vapor deposition (CVD) and contained about 5% impurities, consisting primarily of amorphous carbon. The nanotubes were 5-15 [micro]m long and 10-20 nm in diameter. KH560, analytical grades, was purchased from Jingzhou Jianghan fine chemical. (Jingzhou, China) and used without further purification. Bisphenol A dicyanate (BADCy) was supplied by Shanghai Huifen Kemao. (Shanghai, China). Other reagents were all provided by Xi'an Chemical Reagent Company and used as received without any further treatment.

Functionalization of MWNTs

The p-MWNTs were functionalized in a two-step process. First, put p-MWNTs (1 g) to 50 ml acid-[K.sub.2][Cr.sub.2][O.sub.7] solution ([K.sub.2][Cr.sub.2][O.sub.7]: [H.sub.2]O: concentrated [H.sub.2][SO.sub.4] = 7:12:150, by weight), 1 h sonication at 40[degrees]C, and 2 h stirring at 100[degrees]C. The solution was extensively washed again with deionized water until pH value reached 5-6. The oxidation-treated nanotubes (o-MWNTs) were collected on a 0.45 [mu]m PTFE membrane by vacuum filtration and dried overnight in a vacuum oven at 60[degrees]C, subsequently, these o-MWNTs (0.5 g) were then dispersed into 25 ml anhydrous ethanol. KH-560 (10 g), diluted in 25 ml anhydrous ethanol, was added into the mentioned solution. The resulting mixture was stirred for 24 h at the temperature of 60[degrees]C, and the chemical modifications of MWNTs were formed. The functionalized MWNTs (denoted f-MWNTs) were extensively washed with anhydrous ethanol and collected on a 0.45 [micro]m PTFE membrane by vacuum filtration and dried overnight in a vacuum oven at 60[degrees]C.

Preparation of the MWNTs/BADCy Composites

The MWNTs/BADCy composites containing 2 wt% of p-MWNTs and f-MWNTs were prepared. MWNTs were dispersed in BADCy via an ultrasonicator for 30 min at 100[degrees]C, and then dibutyltin laurate (200 ppm) used as curing catalyst was added to the suspension, subsequently, the system was moved immediately to an oil-bath and reacted for 40 min at 140[degrees]C. The resultant mixture was poured into a preheated mold (150[degrees]C) with the release agent on the inner walls, and then, the mold was degassed under vacuum for 20 min at 140[degrees]C, followed by the curing cycle: 180[degrees]C/2 h + 220[degrees]C/2 h + 240[degrees]C/3 h.

Characterization

X-ray diffraction (XRD) measurements were conducted using D/Max-3C instrument (Anode Material: Cu; Generator Settings: 40 kV, 35 mA). Fourier transform infrared spectroscopy (FTIR) (WQF-300IR, Beijing Optical Instruments Factory) was used to assess the presence of the organic groups in f-MWNTs. The MWNT samples were measured by the method of KBr pellet. Flexural and impact tests were performed on a universal testing machine according to GB2570-81 standard and GB2571-81 standard of China, respectively. Scanning electron microscopy (SEM) images were obtained with Quanta 200-FEI to examine surfaces of the fractured specimens. Dynamic mechanical analysis (DMA 2980, TA Instruments) was carried out to determine the glass transition temperature ([T.sub.g]) of the composites, defined as the temperature where the loss tangent reaches a peak. Measurements were performed at a heating rate of 10[degrees]C/min and a load frequency of 3 Hz.

RESULTS AND DISCUSSION

Functionalization of MWNTs

The concentrated 3:1 [H.sub.2][SO.sub.4]/[HNO.sub.3] solution is the most common oxidation reagent for MWNTs treatment (22). However, there are some disadvantages for this kind oxidation reagent when compared with the acid-[K.sub.2][Cr.sub.2][O.sub.7] solution. One is that the concentrated 3:1 [H.sub.2][SO.sub.4]/[HNO.sub.3] solution is easy to cut the MWNTs short during the treatment, which is not expected in reinforcement modification, and another is that its treatment time is usually long (23-25). Figure 1 illustrates the reactions in the whole treatment processing.

[FIGURE 1 OMITTED]

Figure 2 shows the XRD patterns measured from powder samples of p-MWNTs, o-MWNTs, and f-MWNTs. The pattern for the p-MWNTs (Fig. 2a) exhibites three peaks at 25.84[degrees], 42.87[degrees], and 53.63[degrees]. The pattern of o-MWNTs (Fig. 2b) is the same as that of p-MWNTs, which indicates that the oxidation procedure did not cut or damage the MWNTs structure badly and the remained oxidation reagent had been washed clean. When compared with the peaks of o-MWNTs, the peaks of f-MWNTs (Fig. 2c) are lower and there are some new small peaks. These differences indicate that KH560 has been bonded to MWNTs as the bonded KH560 on the surface of MWNTs weakens the strength of X-ray diffraction and forms some new small diffractive peaks.

[FIGURE 2 OMITTED]

FTIR spectra also corroborate the attachment of KH560 to the surface of MWNTs, as shown in Fig. 3. For the p-MWNTs (Fig. 3a), the bands at 3430 [cm.sup.-1] and 1398 [cm.sup.-1] is attributed to the presence of hydroxyl groups (--OH) on the surface of the MWNTs, which are believed to result from oxidation during purification of the raw material (26). Another band at 1620 [cm.sup.-1] is assigned to the C=O stretching of quinone groups on the surface of MWNTs. For the oxidized MWNTs (Fig. 3b), a new band appears at 1720 [cm.sup.-1], which is attributed to the C = O stretching vibrations of the carboxylic groups (--COOH). The increase in the relative intensities of the bands at 3430 and 1398 [cm.sup.-1] suggests that there are more--OH groups on the MWNTs surface after the oxidation. In the f-MWNTs (Fig. 3c), the band at 3430 [cm.sup.-1] becames weak, and the band at 1398 [cm.sup.-1] disappears. On the other hand, the bands at 2940 and 2840 [cm.sup.-1], the bands at 1000-1200 [cm.sup.-1 and the band at 826 cm"1 are attributed to the methylene group, the Si--O group and the Si--C group, respectively. The weak band at 910 [cm.sup.-1] confirms the presence of the epoxy group on the f-MWNTs.

[FIGURE 3 OMITTED]

The Properties of MWNTs/BADCy Composites

Figure 4 compares the impact strength of MWNTs/BADCy composites with that of pure BADCy resin. The impact strength decreased a little (9%) for the samples containing 2 wt% p-MWNTs, but an significant (41%) improvement was observed for the 2 wt% f-MWNTs based BADCy composites. The same trend appears in the flexural strength (see Fig. 5). The flexural strength of the samples containing 2 wt% p-MWNTs decreased by 28%. However, the flexural strength of the samples containing 2 wt% f-MWNTs increased by 50%. Fracture surfaces morphology of the cured BADCy and of MWNTs/BADCy composites with 2 wt% MWNTs were examined by SEM (see Fig. 6). Composites prepared with p-MWNTs (Fig. 6b, X 1000) exhibited isolated regions of agglomerated MWNTs. The agglomerates cause cracks to initiate and propagate easily. The generated cracks reduce the strength of the composite. The dispersion of the f-MWNTs/BADCy composite is much better than that of the p-MWNTs/BADCy composite. It can be observed that the ridges, dimpled patterns, and crevices on the fracture surfaces are much more prominent for the f-MWNTs/BADCy composite (Fig. 6c, X 10,000). More energy dissipation features are displayed. The well dispersed f-MWNTs are more efficient than the aggregated p-MWNTs in transferring applied load. What is more, the interfacial adhesion between the f-MWNTs and the BADCy matrix is not only van der Waals forces but also chemical bonds which can bear more applied load. It is possible that these fracture features are the manifestation of the increased strength of the f-MWNTs/BADCy composite.

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

Figure 7 represents the storage modulus of the cured BADCy and MWNTs/BADCy composites. It shows that the p-MWNTs/BADCy composite and the f-MWNTs/BADCy composite have almost the same storage modulus when temperature is low than [T.sub.g] and both of them are higher than that of the cured BADCy resin. This result is easy to be understood because of the high storage modulus and aspect ratio of MWNTs. The f-MWNTs/BADCy composite presents higher storage modulus at higher temperature (250[degrees]C-300 [degrees] C). [T.sub.g] of the p-MWNTs/BADCy composite is the same as that of the cured BADCy resin and both of them are little lower than that of the f-MWNTs/BADCy composite (see Fig. 8). It appears that the functional groups of epoxy on f-MWNTs play a role in increasing the interfacial adhesion between the nano-tubes and the matrix. The chemical bonding between the matrix and the f-MWNTs restricts the movement of the polymer chains at high temperature.

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

CONCLUSION

A MWNTs/BADCy composite using epoxy group as a bridge was prepared. The acid-[K.sub.2][Cr.sub.2][O.sub.7] solution is an effective treatment agent for introducing--OH groups on the MWNTs without cutting the MWNTs short or damaging the structure of the MWNTs. The two-step functionalization of MWNTs is effective in achieving uniform dispersions of MWNTs in BADCy resin. The presence of epoxy group on the surface of MWNTs increases the interfacial adhesion between MWNTs and BADCy and improves the mechanical properties of the f-MWNTs/ BADCy composite greatly. The f-MWNTs also increase the [T.sub.g] and the storage modulus of the f-MWNTs/BADCy composite slightly.

REFERENCES

(1.) C.P.R Nair, D. Mathew, and K.N. Ninan, Adv. Polym. Sci., 155. I (2001).

(2.) S. Gangulia, D. Deana. K. Jordanb, G. Pricec, and R. Vaia. Polymer, 44. 1315 (2003).

(3.) C. Marieta. E. Schulz., and 1. Mondragon, Compos. Set Technol., 62, 299 (2002).

(4.) I. Hamerton, A.M. Emsleya, B.J. Howlina, P. Klewpatinonda, and S. Takeda, Polymer. 45. 2193 (2004).

(5.) J.L. Wang, G.Z. Liang, W. Zhao, S.H. Lv, and H.X. Yan, Polym. Eng. Sci. 46, 581 (2006).

(6.) J. Borrajo, C.C. Riccardi, and R.J.J. Williams. Polymer. 36, 3541 (1995).

(7.) J.W. Hwang, S.D. Park, K. Cho, J.K. Kim, and C.E. Park, Polymer, 38, 1835 (1997).

(8.) C. Marieta, P.M. Remiro, G. Garmendia, I. Harismendy. and 1. Mondragon, Eur. Polym. J., 39, 1965 (2003).

(9.) J.N. Suman, J. Kathi, and S. Tammishetti. Eur. Polym. J., 41, 2963 (2005).

(10.) A. Chaplina, I. Hamertonb, H. Hermanb, A.K. Mudharb, and S.J. Shawa, Polymer, 41, 3945 (2000).

(11.) I. Hamerton, H. Herman, A.K. Mudhar, A. Chaplin, and S.J. Shaw, Polymer, 43, 3381 (2002).

(12.) R.H. Lina, W.H. Lub, and C.W. Lina. Polymer. 45. 4423 (2004).

(13.) B.C. Guo, D.M. Jia, W.W. Fu, and Q.H. Qiu, Polym. Degrad. Stab., 79, 521 (2003).

(14.) M.S. Lakshmi and B.S.R. Reddy. Eur. Polym. J., 38. 795 (2002).

(15.) C.A. Fyfe, I. Niu, S.J. Rettig, D.W. Wang, and M.D. Poliks. J. Polym. Sci. Part A: Polym. Chem., 32, 2203 (1994).

(16.) B.S. Hayes and J.C. Seferis, Polym. Eng. Sci.. 40. 1344 (2000).

(17.) I.B. Recalde, I. Porcar, A. Campos, and CM. Gomez, J. Chromatogr. A., 942, 53 (2002).

(18.) X.L. Xie, Y.W. Mai. and X.P. Zhou, Mater. Sci. Eng. R. 49, 89 (2005).

(19.) R. Andrews and M.C. Weisenberger, Curr. Opin. Solid State Mater. Sci., 8, 31 (2004).

(20.) P. Liu. Eur. Polym. J., 41, 2693 (2005).

(21.) J.H. Wang, G.Z. Liang, H.X. Yan, and S.B. He, Express Polym. Lett., 2, 118 (2008).

(22.) S. Goyanes, G.R. Rubiolo, A. Salazar, A. Jimeno. M.A. Corcuera, and I. Mondragon, Diamond Relat. Mater., 16, 412 (2007).

(23.) J.Y. Li and Y.F. Zhang, Appl. Surf. Sci.. 252. 2944 (2006).

(24.) H.B. Jia, Y.F. Lian, M.O. Ishitsuka, T. Nakahodo, Y. Maeda, T. Tsuehiya, T. Wakahara, and T. Akasaka. Sci. Technol. Adv. Mater., 6, 571 (2005).

(25.) Z.L. Yang, H.T. Pu, and J.L. Yin, Mater. Leu., 59, 2838 (2005).

(26.) T. Ramanathan. F.T. Fisher, R.S. Ruo, and L.C. Brinson. Chem. Mater., 17, 1290 (2005).

Correspondence to: Guozheng Liang; e-mail: [e-mail]

Contract grant sponsor: Natural Science Foundation of Shaanxi Province

of China; contract grant number: 2007B10.

DOI 10.1002/pen.21277

Published online in Wiley InterScience (www.interscience.wiley.com).

[C]2009 Society of Plastics Engineers

Jinhe Wang, Guozheng Liang, Hongxia Yan, Shaobo He

Department of Applied Chemistry, School of Science, Northwestern Polytechnical University, Xi'an Shaanxi 710072, China