Nanomechanical Properties of Silica-Coated Multiwall Carbon

3146

Langmuir 2005, 21, 3146-3152

Nanomechanical Properties of Silica-Coated Multiwall Carbon NanotubessPoly(methyl methacrylate) Composites M. Olek,† K. Kempa,‡ S. Jurga,§ and M. Giersig*,† Center of Advanced European Studies and Research (CAESAR), Bonn, Germany, Boston College, Boston, MA, and Adam Mickiewicz University, Poznan, Poland Received November 29, 2004. In Final Form: January 5, 2005 The mechanical properties of polymer composites, reinforced with silica-coated multiwall carbon nanotubes (MWNTs), have been studied using the nanoindentation technique. The hardness and the Young’s modulus have been found to increase strongly with the increasing content of these nanotubes in the polymer matrix. Similar experiments conducted on thin films containing MWNTs, but without a silica shell, revealed that the presence of these nanotubes does not affect the nanomechanical properties of the composites. While carbon nanotubes (CNTs) have a very high tensile strength due to the nanotube stiffness, composites fabricated with CNTs may exhibit inferior toughness. The silica shell on the surface of a nanotube enhances its stiffness and rigidity. Our composites, at 4 wt % of the silica-coated MWNTs, display a maximum hardness of 120 ( 20 MPa, and a Young’s modulus of 9 ( 1 GPa. These are respectively 2 and 3 times higher than those for the polymeric matrix. Here, we describe a method for the silica coating of MWNTs. This is a simple and efficient technique, adaptable to large-scale production, and might lead to new advanced polymer based materials, with very high axial and bending strength.

Introduction The unique mechanical, electrical, and optical properties1-4 of multiwall carbon nanotubes (MWNTs) make them very attractive for the fabrication of new advanced materials, particularly polymer composites with improved performance, or with new properties. Due to their exceptionally high strength, and axial Young’s modulus,1,5,6 MWNTs have been commonly considered as reinforcing fillers for high-strength materials. In this context, various polymers have been used as matrix materials and different preparation techniques employed.7-15 Numerous studies on MWNTs showed that an effective utilization of carbon * Corresponding author. E-mail: [email protected]. † CAESAR. ‡ Boston College. § Adam Mickiewicz University. (1) Yu, M.-F.; Lourie, O.; Dyer, M. J.; Moloni, K.; Kelly, T. F.; Ruoff, R. S. Science 2000, 287, 637-640. (2) Ajayan, P. M. Chem. Rev. 1999, 99, 1787-1799. (3) Rao, C. N. R.; Govindaraj, B. A.; Nath, M. ChemPhysChem 2001, 2, 78-105. (4) Demczyk, B. G.; Wang, Y. M.; Satishkumar, J. C.; Cumings, M. H.; Han, W.; Zettl, A.; Ritchie, R. O. Mater. Sci. Eng., A 2002, 334, 173-178. (5) Wong, E. W.; Sheehan, P. E. Science 1997, 277 (5334), 19711975. (6) Salvetat, J.-P.; Kulik, A. J.; Bonard, J.-M.; Briggs, A. D.; Sto¨ckli, T.; Metenier, K.; Bonnamy, S.; Beguin, F.; Burnham, N. A.; Forro´, L. Adv. Mater. 1999, 11, 161-165. (7) Li, D.; Zhang, X.; Sui, G.; Wu, D.; Liang, J. J. Mater. Sci. Lett. 2003, 22, 791-793. (8) Qian, D.; Dickey, E. C.; Andrews, R.; Rantell, T. Appl. Phys. Lett. 2000, 76 (20), 2868-2870. (9) Park, S. J.; Cho, M. S.; Lim, S. T.; Choi, H. J.; Jhon, M. S. Macromol. Rapid Commun. 2003, 24, 1070-1073. (10) Velsco-Santos, C.; Martı´nez-Herna˜ndez, A. L.; Fisher, F. T.; Ruoff, R.; Casta˜no, V. M. Chem. Mater. 2003, 15, 4470-4475. (11) Jin, Z.; Pramoda, K. P.; Xu, G.; Goh, S. H. Chem. Phys. Lett. 2001, 337, 43-47. (12) Jin, Z.; Huang, L.; Goh, S. H.; Xu, G.; Ji, W. Chem. Phys. Lett. 2000, 332, 461-466. (13) Safadi, B.; Andrews, R.; Grulke, E. A. J. Appl. Polym. Sci. 2002, 84, 2660-2669. (14) Qiao, Y. L.; Cui, L.; Liu, Y.; Cui, S.; Yang, J. S.; Yang, D. A. J. Mater. Sci. Lett. 2002, 21 (23), 1813-1815. (15) Olek, M.; Ostrander, J.; Jurga, S.; Mo¨hwald, H.; Kotov, N.; Kempa, K.; Giersig, M. Nano Lett. 2004, 4 (10), 1889-1895.

nanotubes (CNTs) in composite applications strongly depends on the ability to disperse the CNTs homogeneously throughout the matrix.16-18 Obviously, good interfacial bonding and interactions between nanotubes and polymers are the necessary conditions for improving the mechanical properties of the composites. In general, the tensile modulus and ultimate strengths of the CNT composites are reported to increase, although below the level of expectation.19-21 The nanoindentation technique has proved to be a useful tool for the determination of the mechanical properties of thin films, including polymers.22-27 Depth-sensing indentation allows the displacement of the indenter to be measured as a function of an applied, controlled load. The resulting load versus displacement curves, together with the indenter geometry, can be analyzed to obtain the elastic modulus and hardness of the tested samples. In this manner, the nanomechanical behavior of different systems, including composites, can be quantitatively characterized by a nanoindentation experiment. Recently, Li et al. reported the nanomechanical properties of singlewalled carbon nanotube (SWNT) reinforced epoxy composites measured by nanoindentaion and scratch tech(16) Ruan, S. L.; Gao, P.; Yang, X. G.; Yu, T. X. Polymer 2003, 44, 5643-5654. (17) Andrews, R.; Jacques, D.; Qian, D.; Rantell, T. Acc. Chem. Res. 2002, 35, 1008-1017. (18) Rouse, J. H.; Lillehei, P. T. Nano Lett. 2003, 3, 59-62. (19) Du, F.; Fischer, J. E.; Winey, K. I. J. Polym. Sci., Part B: Polym. Phys. 2003, 41 (24), 3333-3338. (20) Moore, E. M.; Ortiz, D. L.; Marla, V. T.; Shambaugh, R. L.; Grady, B. P. J. Appl. Polym. Sci. 2004, 93 (6), 2926-2933. (21) Sennett, M.; Chang, S.; Doremus, R. H.; Siegel, R. W.; Ajayan, P. M.; Schadler, L. S. Ceram. Trans. 2002, 134, 551-556. (22) Klapperich, C.; Komvopoulos, K.; Pruit, L. J. Tribol. 2001, 123, 624. (23) Pavoor, P. V.; Bellare, A.; Strom, A.; Yang, D.; Cohen, R. E. Macromolecules 2004, 37, 4865-4871. (24) Du, B.; Tsui, O. K. C.; Zhang, Q.; He, T. Langmuir 2001, 17, 3286-3291. (25) Briscoe, B. J.; Fiori, L.; Pelillo, E. J. Phys. D: Appl. Phys. 1998, 31, 2395-2405. (26) Lu, W.; Komvopoulos, K. J. Tribol. 2001, 123, 717. (27) Rau, K.; Singh, R.; Goldberg, E. Mater. Res. Innovat. 2002, 5, 151-161.

10.1021/la0470784 CCC: $30.25 © 2005 American Chemical Society Published on Web 02/09/2005

Nanomechanical Properties of Silica-Coated MWNTs

niques.28 In general, the hardness, the elastic modulus, and the scratch resistance of such materials increase with the increasing weight percentage of the CNTs. It was found that, due to the formation of aggregates of SWNTs within composites and a weak bonding between polymer and SWNTs, the measured mechanical properties were below their theoretically predicted values. Pavoor et al. studied the nanomechanical behavior of MWNT/polyelectrolyte composites produced by using the layer-by-layer (LBL) assembly deposition technique.29 Utilization of the LBL technique has proven to be an efficient method for incorporating CNTs into a polymer matrix, with reduced phase segregation, high homogeneity, good dispersion and interpenetration of the nanocolloids and polymers, and high CNT concentration.15,30 The nanoindentation studies of ref 29 showed that the hardness and Young’s modulus of thin LBL MWNT/poly(allylamine hydrochloride) (PAH) films are comparable to those of PAH. The individual MWNT can be displaced easily during indentation, leading to poor nanomechanical properties, close to those of the surrounding matrix. Also, the results of indentation tests carried out by Dutta et al. on SWNT reinforced epoxy composites revealed quantifiable but modest improvements in mechanical properties, with varying weight percentage of nanotubes in the epoxy resin. However, these studies were performed under high loads, resulting in microscale deformations.31,32 The small changes in the observed mechanical properties of SWNT composites were explained by the formation of bundles and the curved morphology of CNTs. All these reports demonstrate only a weak enhancement of the mechanized properties of the CNT filled polymers, in contrast to earlier expectations. The morphology, homogeneous dispersion of CNTs, strong interconnectivity between components, concentration of CNTs, and elastic properties of carbon nanotubes have been shown to affect the mechanical response of the composites.15,30 In this study, we perform a nanomechanical characterization of composites, reinforced with the silica-coated MWNT (MWNT@SiO2), within the poly(methyl methacrylate) (PMMA) matrix. Individual carbon nanotubes were coated with a uniform, thick layer of silica and incorporated into PMMA. The CNT-SiO2/PMMA composites were spin-coated on a silicon wafer substrate and subsequently investigated by the nanoindentation technique. The hardness and elastic modulus were measured and compared with values obtained for the films made of uncoated MWNTs in the PMMA matrix. Mechanical Characterization Nanomechanical tests were carried out using an atomic force microscope (AFM) (NanoScope IV Digital Instruments) with a conjugated TriboScope nanomechanical test instrument from Hysitron Inc. The Hysitron nanoindenter is a depth-sensing and load-control device, which is capable of providing measurements of elastic and plastic properties at the nanoscale level. The TriboScope instrument can be used to perform surface imaging indentation and friction testing at loads over a broad range, with a depth resolution equal to 0.0002 nm and a load resolution of ∼1 nN. (28) Li, X.; Gao, H.; Scrivens, W. A.; Fei, D.; Xu, X.; Sutton, M. A.; Reynolds, A. P.; Myrick, M. L. Nanotechnology 2004, 15, 1416-1423. (29) Pavoor, P. V.; Gearing, B. P.; Gorga, R. E.; Bellare, A.; Cohen, R. E. J. Appl. Polym. Sci. 2004, 92, 439-448. (30) Mamedov, A. A.; Kotov, N. A.; Prato, M.; Guldi, D. M.; Wicksted, J. P.; Hirsch, A. Nat. Mater. 2002, 1, 190-194. (31) Penumadu, D.; Dutta, A.; Pharr, G. M.; Files, B. J. Mater. Res. 2003, 18, 1849. (32) Dutta, A. K.; Penumadu, D.; Files, B. J. Mater Res. 2004, 19, 158.

Langmuir, Vol. 21, No. 7, 2005 3147

Diamond conical and Berkovich tips were employed in this study as indenters. The total included angle on the Berkovich tip is 142.3°, with a half angle of 65.35°. This makes this tip very flat and efficient for a wide range of materials, including polymers. A conical tip with a nominal radius of curvature equal to 1 µm is usually used for soft material indentation.35 The typical indentation test consists of three steps. The load step, with a well defined and controlled maximum force applied to the sample at the specified loading rate, the hold step at maximum force, used to minimize the effects of the specimen creep on the calculated nanomechanical properties,33 and finally the unload step, usually conducted at the same rate as that for the load step. The hardness and elastic moduli were calculated from the recorded unloading step of the depth-displacement curve, based on the method of Oliver and Pharr.34 A general relation between penetration depth, h, and load, P, is

P ) C(h - hf)m

(1)

where C contains all the geometric constants, sample elastic modulus, Poisson’s ratio, the indenter elastic modulus, and the indenter Poisson’s ratio. hf is the final unloading depth, and m is the power law exponent, which is related to the geometry of the indenter. We have also

Fmax Ac

(2)

x

(3)

H)

Er )

1 2

π S Ac

where H is the local hardness; Ac is the contact area; Er is the reduced Young’s modulus; S is the contact stiffness, defined as the slope of the unloading curve fitted to a power law equation (eq 1); and Er is an effective elastic modulus, combining the properties of the indenter and the sample under test. It is given by

(1 - νs2) (1 - νi2) 1 ) + Er Es Ei

(4)

where υi and υs are the Poisson’s ratios for the indenter and the sample, respectively, and Ei and Es are the respective elastic moduli. Since we used a diamond tip with a high elastic modulus (Ei ) 1170 GPa) and a low Poisson’s ratio (υi ) 0.07), the second term in eq 4 can be neglected. Since the contact area is a function of the contact depth, a tip calibration procedure has to be employed to determine the geometry of the indenter tip. For this purpose, a series of indentations at different contact depths was made in a sample with a known elastic modulus. The dependence of the contact area and contact depth on load was plotted, and the area function was found using the following polynomial expression

Ac ) C0hc2 + C1hc + C2hc1/2 + C3hc1/4 + C3hc1/8 + ...

(5)

where Cn (n ) 0, 1, 2...) are the fitting coefficients and hc (33) Chudoba, T.; Richter, F. Surf. Coat. Technol. 2001, 148, 191198. (34) Oliver, W. C.; Pharr, G. M. J. Mater. Res 1992, 7, 1564. (35) TriboScope Users Manual.

3148

Langmuir, Vol. 21, No. 7, 2005

is the contact depth. In this study, the tip calibration was carried out on poly(methyl methacrylate), instead of the commonly used hard standard fused quartz. Klapperich et al. called into question the validity of using hard standard materials for tip calibrations, when soft materials are investigated.22 Briscoe et al. remarked that calibration against a hard surface may produce an error and does not represent the contact situation with softer surfaces.25 Moreover, contact depths obtained in this study were on the order of several hundred nanometers. This range cannot be obtained from a calibration performed on hard materials. The calibration procedure that we performed on a polymer is described below. Experimental Section Materials. MWNTs (CVD method, purity >95%, diameter 10-20 nm, length 5-20 µm) were obtained from NanoLab (Newton, MA). In this study, we produced two different composites, based on multiwall carbon nanotubes in PMMA (Mw ) 320 000). The first one is based on functionalized MWNTs that were blended with PMMA. In the second composite, we employed silica-coated MWNTs that were incorporated into a polymer (PMMA) matrix. The first sample was prepared by an amide functionalization of MWNTs in order to obtain soluble carbon nanotubes in organic solvents. In the first step, carbon nanotubes were oxidized with a mixture of nitric acidsulfuric acid (1:3 v/v). A 200 mg portion of MWNTs was suspended in this mixture followed by sonication (150 W) for 4 h and left aside for 20 h. The excess concentrated acids were removed by centrifugation. The resulting black solid was intensively washed with deionized water. To solubilize the MWNTs, we used the method of ref 36. Briefly, oxidized MWNTs were stirred in 100 mL of thionyl chloride at 70 °C for 24 h to convert surface-bound carboxylic acid groups into acyl chloride groups. After centrifugation, the remaining solid was washed with anhydrous tetrahydrofuran (THF) and dried under vacuum at room temperature. A mixture of the resulting MWNTs and 3 g of octadecylamine (ODA) was stirred under N2 atmosphere at 80 °C for 96 h. After cooling to room temperature, the excess ODA was removed by intensive washing with ethanol (six times centrifuged). A dry black solid of MWNTs was dispersed in chloroform, and the appropriate amount of poly(methyl methacrylate) (Mw ) 320 000) was added to obtain the desired weight concentration of CNTs with respect to PMMA. The final mixture was then intensively mixed until a stable, blackcolored chloroform solution of MWNT/PMMA composite was achieved. For nanomechanical tests, we prepared five MWNT/PMMA samples with 1, 2, 3, 4, and 5 wt % MWNT. In another sample, silica-coated MWNTs were used as a filler in the poly(methyl methacrylate) polymer. The coating process steps are as follows. The MWNTs were functionalized with poly(allylamine hydrochloride) (PAH). CNTs (50 mg) were dispersed in a 0.5 wt % PAH (Mw ) 70 000) salt solution (0.5 M NaCl, 500 mL) and sonicated for 5 h. The excess of polymer was removed by centrifugation (five times) and washed with water. A residual black solid was redispersed in water, forming a stable, homogeneous CNT suspension. The CNT water dispersion was transferred to silica sol (mixture of TEOS, H2O, and ethanol; mass ratio 2:1:4) in a 5:1 volume ratio. To prevent phase separation of TEOS and MWNT/water, the mixture was sonicated. After 2 h, the solution became (36) Qin, Y.; Liu, L.; Shi, J.; Wu, W.; Zhang, J.; Guo, Z. X.; Li, Y.; Zhu, D. Chem. Mater. 2003, 15, 3256-3260.

Olek et al.

homogeneous and was set aside overnight at room temperature. After 12 h, the mixture was centrifuged (four times) to wash the carbon nanotubes with ethanol. The sediment was redispersed in a solution of ammonia in ethanol (4.2 vol % ammonia (28 wt % in water) in ethanol). Immediately after this, TES solution (10 vol % in ethanol) was added under stirring (5 mL of TES in 500 mL of ethanol solution of CNT). The reaction mixture was stirred for another 8 h and sonicated from time to time. Finally, the CNTs were washed with ethanol (four times centrifuged) and again redispersed. The process described above leads to the formation of a uniform and thick layer of silica on every individual MWNT (Figure 1). The pristine MWNTs have a diameter in the range 5-20 nm. The CNT thickness after silica coating increased to 70-80 nm for all carbon nanotubes. This indicates that the silica shell thickness is roughly equal to the thickness of the core nanotubes. The modified MWNTs were subsequently transferred to chloroform by functionalization with 3-aminopropyl trimethoxysilane (97%). The unreacted component was removed by washing in chloroform. An appropriate amount of PMMA was added to the silica-coated MWNT-chloroform solution to obtain the desired concentration of CNTs. This mixture was further homogenized in an ultrasonic bath for 1 h. For nanomechanical investigation, five different silica-coated MWNT/PMMA (MWNT-SiO2/PMMA) composites were prepared with 1, 2, 3, 4, and 5 wt % CNTs in a polymer matrix. The chloroform dispersions of both MWNT/PMMA and MWNT-SiO2/PMMA composites were spin-coated on a silicon wafer substrate. The thickness of films can be controlled by altering the concentration of composites in the solution or by the speed of spinning. In general, thin films with a thickness >3 µm were formed at speeds ranging from 1200 to 2500 rpm in 25 s, followed by drying in an oven (100 °C, 3 min). Before mechanical tests were performed, samples were left aside for a few days to achieve complete solvent evaporation. The thickness and an approximate roughness of spin-coated films were estimated utilizing a profilometer (Surface Profiler Kla-Tencor P-10). Tip Calibration. As was mentioned above, the calibration of indenters (conical and Berkovich) was carried out on soft material: poly(methyl methacrylate) was used as a standard material with an elastic modulus equal to 3.6 GPa.27 A chloroform solution of PMMA was spin-coated on a silicon wafer substrate. Thin polymer films, of a thickness >3 µm, exhibit high smoothness of the surface (roughness, R, of 30 µN/s, the elastic modulus does not show any

Olek et al.

Figure 4. Elastic modulus (a) and hardness (b) for PMMA, as a function of the load/unload rate of the indentation.

dependence on the indentation velocity. This investigation shows that, at low load/unload rates, the mechanical response may slightly differ from that obtained for higher rates. It must be also pointed out that this experiment was carried out under a relatively long hold time, as high as 20 s. This reduces the creep effect, because the material has enough time to minimize the mechanical disequilibrium before the unloading step begins.22,25 In this manner, the unload step is not much affected by the material creep. A significant influence of the hold period on the calculated values of H and Er was observed in another experiment conducted on PMMA. Indents with a peak load equal to 500 µN/s at a 40 µN/s load/unload rate were carried out for different hold time periods. Figure 5 illustrates decreasing values of the hardness and elastic modulus, while the hold time at a maximum load is increasing. These results show a great influence of the material creep on the computed values of the elastic modulus and hardness. This effect was confirmed for both Berkovich tips (Figure 5) and conical tips (not shown in this paper). The investigations described above were conducted on PMMA due to the high reproducibility of the results and the small standard deviation. To enable comparison between data sets, the experimental conditions for all specimens in this study, the hold time and the load/unload rate of indentations were set to 20 s and 40 µN/s, respectively. Figure 6 shows the nanomechanical characterization of MWNT/PMMA composites. The tests were performed under fixed experimental conditions (see above), using a Berkovich tip. The homogeneity of our samples is confirmed by the relatively small standard deviation of the data points. The hardness and Young’s modulus are shown

Nanomechanical Properties of Silica-Coated MWNTs

Figure 5. Elastic modulus (a) and hardness (b) for PMMA, as a function of the hold time of the indentation.

as a function of the contact depth for different MWNT contents (Figure 6). The nanoindentation studies reveal that the mechanical properties of the MNWT composites are comparable to those of thin films of PMMA. Moreover, H and Er as a function of contact depth exhibit exactly the same behavior as that shown for PMMA. The hardness shows a decrease with increasing contact depth. There are no significant changes in H values with increasing concentration of carbon nanotubes in polymer. The Young’s modulus of the thin films is independent of the indentation displacement, and the Er values are close to those obtained for PMMA. While our results show no significant changes in the nanomechanical properties of the composites with increasing weight percentage of the carbon nanotubes, Li et al.,28 working with composites with a different morphology, reported a modest but quantifiable increase of the hardness and elastic modulus with increasing content of nanotubes in an ester resin. The LBL technique has proven to be an efficient method for incorporating carbon nanotubes into a polymer matrix, allowing for high composite homogeneity, good dispersion, interpenetration, and high CNT concentration.15 However, nanoindentation investigations of LBL MWNT/PAH composites revealed that the hardness and elastic modulus are close to those of the surrounding polymer matrix.29 These results are consistent with our study and confirm that a high concentration and a homogeneous distribution of CNTs within a polymer matrix as well as strong adhesion between the structural components are insufficient to provide reinforcement of composites (in the mechanical sense). It was suggested that the flexibility of carbon nanotubes and their curved morphology may reduce the reinforcement action. Even strong inter-

Langmuir, Vol. 21, No. 7, 2005 3151

Figure 6. Elastic modulus (a) and hardness (b) for different CNT weight percentage contents in MWNT/PMMA composites, as a function of the contact depth.

connectivity between the CNTs and the host polymer (as is obtained for LBL structures) does not lead to a significant increase of the nanomechanical reinforcement under indentation load. The indenter can easily displace carbon nanotubes due to their flexibility and bending properties. As a result, the indenter essentially “feels” resistance only from the surrounding matrix, and the mechanical response of the composite is close to that of the polymer matrix. Wong and Sheehan determined the average bending strength for MWNTs to be 14.2 ( 8.0GPa, that is, several times smaller than that for SiC nanorods.37 Thus, nanomechanical improvement of a CNT/polymer composite examined by nanoindentation is limited by the relatively small bending strength of carbon nanotubes. A completely different situation occurs for the silicacoated CNT, as shown in Figure 7. The MWNT@SiO2 reinforced composites can exhibit a much higher hardness and elastic modulus than PMMA. Both those quantities increase with increasing concentration of the coated MWNTs in the polymer matrix. The results demonstrate the great influence of the silica reinforcing on the mechanical response of the CNT/polymer composite. The silica shell encasing the carbon nanotube surface changes its mechanical properties considerably. Such modified carbon nanotubes possess a higher stiffness. Relatively large data scatter for the MWNT@SiO2 films, as compared to the MWNT/PMMA composites, indicate the presence of some inhomogeneities and a nonuniform distribution of MWNTs throughout the samples (Figure 8). The average standard deviations of the mean for the (37) Wong, E. W.; Sheehan, P. E. Science 277, 1971.

3152

Langmuir, Vol. 21, No. 7, 2005

Figure 7. Elastic modulus (a) and hardness (b) for different CNT weight percentage contents in silica-coated MWNT/PMMA composites, as a function of contact depth.

Figure 8. SEM image of the 3% silica-coated MWNT composite.

Young’s modulus are (0.27, (1.06, (1.14, and (1.56 GPa for 1, 2, 3, and 4% of the silica-coated MWNT content, respectively. The average standard deviations of the hardness are (0.006, (0.02, (0.03, and (0.02GPa for 1, 2, 3, and 4% MWNT@SiO2 load, respectively. The error bars are not shown in Figure 7 for clarity. For the 5% CNT sample, the Young’s modulus varies from 5 to 20 GPa, which demonstrates increasing inhomogeneity with increasing concentrations of CNTs. The preparation method used in our study of MWNT@SiO2/PMMA composites does

Olek et al.

not provide a uniform distribution of CNTs within the film, due to the poor solubility of CNT@SiO2 in chloroform. This leads to large errors in Er and H, and therefore, we did not include those values in Figure 7. Nevertheless, the results shown demonstrate clearly a significant increase in the hardness and elastic modulus of MWNT@SiO2/PMMA thin films, emphasizing the importance of the silica reinforcement of the carbon nanotubes. For example, Figure 7a shows the Young’s modulus for the 4 wt % sample to be ∼3 times as high as that for PMMA. For this CNT concentration, the hardness increases ∼2 times in comparison to the polymer (Figure 7b). We point out that this is not due to the different preparation techniques used for the coated and uncoated CNTs. The results of ref 29 clearly demonstrate that, even for a strong interconnectivity between components and a high homogeneity at a very high weight percent of CNT, the hardness and Young’s modulus remain as low as those for a surrounding polymer. The presence of carbon nanotubes in polymeric materials does not improve nanomechanical properties due to the high elasticity and small bending strength of CNTs.37 Silica coating of MWNTs opens up possibilities for the production of new advanced reinforced materials for a variety of applications. While CNTs have a very high tensile strength due to the nanotube stiffness, composites fabricated with CNTs may exhibit inferior toughness.37 As demonstrated in this work, the high bending strength of silica reinforced MWNTs makes them a good candidate for high-toughness composite structures. Since silica is an insulator, a coated CNT can be used as a coated nanowire for some nanoelectrical applications. The electrical as well as dielectric properties of composites containing silica-coated CNTs are strongly affected by the insulator layer around each nanotube. It was recently shown38 that a super-dielectric can be made this way, which has a very large, low-frequency dielectric constant and low dielectric loss. This may lead to novel applications in biology, medicine, and nanolectronic devices. The preparation of silica-coated carbon nanotubes described here is simple and efficient and can be utilized for largescale production. Conclusions The nanomechanical properties of the CNT/polymer composites (e.g., hardness) are not affected by the presence of CNTs. We show that silica-coated MWCNTs improve the nanomechanical properties of polymeric composites. Since the bending strength of CNTs is improved by the silica shell, the hardness and elastic modulus of MWNT/ polymer composites increase with increasing content of MWNTs in the matrix. For example, a polymer composite at 4 wt % MWNT@SiO2 displays an ultimate hardness of 120 ( 20 MPa and a Young’s modulus of 9 ( 1 GPa. These are respectively 2 and 3 times higher than the values of the pure polymeric matrix. Note Added after ASAP Publication. This article was published ASAP on February 9, 2005. Figures 6 and 7 were published inadvertently as duplicates. The correct version was reposted on February 22, 2005. LA0470784 (38) Kempa, K.; Olek, M.; Correa, M.; Giersig, M.; Cross, M.; Benham, G.; Sennett, M.; Carnahan, D.; Kempa, T.; Ren, Z. Manuscript in preparation.