Langmuir 2008, 24, 9763-9766
9763
Synthesis and Characterization of Monolithic Carbon Aerogel Nanocomposites Containing Double-Walled Carbon Nanotubes Marcus A. Worsley, Joe H. Satcher, Jr., and Theodore F. Baumann* Chemistry, Materials, Earth and Life Sciences Directorate, Lawrence LiVermore National Laboratory, 7000 East AVenue, LiVermore, California 94551 ReceiVed April 14, 2008. ReVised Manuscript ReceiVed July 1, 2008 We report the synthesis and characterization for the first examples of monolithic low-density carbon aerogel (CA) nanocomposites containing double-walled carbon nanotubes. The CA nancomposites were prepared by the sol-gel polymerization of resorcinol and formaldehyde in an aqueous surfactant-stabilized suspension of double-walled carbon nanotubes (DWNTs). The composite hydrogels were then dried with supercritical CO2 and subsequently carbonized under an inert atmosphere to yield monolithic CA structures containing uniform dispersions of DWNTs. The microstructures and electrical conductivities of these CA nanocomposites were evaluated for different DWNT loadings. These materials exhibited high BET surface areas (>500 m2/g) and enhanced electrical conductivities relative to pristine CAs. The details of these results are discussed in comparison with theory and literature.
Introduction Carbon aerogels (CAs) are novel mesoporous materials with many interesting properties, such as low mass densities, continuous porosities, high surface areas and high electrical conductivity.1-4 These properties are derived from the aerogel microstructure, which is a network of interconnected primary particles with characteristic diameters between 3 and 25 nanometers. Because of their unusual chemical and textural characteristics, carbon aerogels are promising materials for use as electrode materials for super capacitors and rechargeable batteries, adsorbents and advanced catalyst supports.1,3 To expand the potential application for these unique materials, recent efforts have focused on the design of CA composites with the goal of modifying the structure, conductivity or catalytic activity of the aerogel.6-10 Carbon nanotubes (CNTs) possess a number of intrinsic properties that make them promising materials in the design of composite materials. For example, CNTs can have electrical conductivities11 as high as 106 Sm-1, thermal conductivities12 as high as 3000 Wm-1 K-1 and elastic moduli on the order of 1 TPa.13 In addition, the large aspect ratios (100-1000) of CNTs means that small additions (less than 1 vol * To whom correspondence should be addressed. E-mail: baumann2@ llnl.gov. (1) Hrubesh, L. W. J. Non-Cryst. Solids 1998, 225, 335. (2) Pekala, R. W.; Kong, F. M. Abstracts Papers Am. Chem. Soc. 1989, 197, 113. (3) Lu, X. P.; Nilsson, O.; Fricke, J.; Pekala, R. W. J. Appl. Phys. 1993, 73, 581. (4) Pekala, R. W. J. Mater. Sci. 1989, 24, 3221. (5) Pekala, R. W.; Farmer, J. C.; Alviso, C. T.; Tran, T. D.; Mayer, S. T.; Miller, J. M.; Dunn, B. J. Non-Cryst. Solids 1998, 225, 74. (6) Wang, J.; Glora, M.; Petricevic, R.; Saliger, R.; Proebstle, H.; Fricke, J. J. Porous Mater. 2001, 8, 159. (7) Fu, R. W.; Zheng, B.; Liu, J.; Weiss, S.; Ying, J. Y.; Dresselhaus, M. S.; Dresselhaus, G.; Satcher, J. H.; Baumann, T. F. J. Mater. Res. 2003, 18, 2765. (8) Baumann, T. F.; Satcher, J. H. Chem. Mater. 2003, 15, 3745. (9) Tao, Y.; Noguchi, D.; Yang, C. M.; Kanoh, H.; Tanaka, H.; Yudasaka, M.; Iijima, S.; Kaneko, K. Langmuir 2007, 23, 9155. (10) Bordjiba, T.; Mohamedi, M.; Dao, L. H. J. Power Sources 2007, 172, 991. (11) Thess, A.; Lee, R.; Nikolev, P.; Dai, H.; Petit, P.; Forro, L.; Robert, J.; Xu, C.; Lee, Y. H.; Kim, S. G.; Rinzler, A. G.; Colbert, D. T.; Scuseria, G.; Tomanek, D.; Fischer, J. E.; Smalley, R. E. Science 1996, 273, 483. (12) Kim, P.; Shi, L.; Majumdar, A.; McEuen, P. L. Phys. ReV. Lett. 2001, 87, 215502. (13) Qi, H. J.; Teo, K. B. K.; Lau, K. K. S.; Boyce, M. C.; Milne, W. I.; Robertson, J.; Gleason, K. K. J. Mech. Phys. Solids 2003, 51, 2213.
%) of CNTs can produce a composite with novel properties.14 Therefore, the homogeneous incorporation of CNTs into a CA matrix provides a viable route to new carbon-based composites with enhanced thermal, electrical and mechanical properties.9,10,15,16 One of the main challenges in preparing CNT composites is achieving a good uniform dispersion of nanotubes throughout the matrix. CAs are typically prepared through the sol-gel polymerization of resorcinol with formaldehyde in aqueous solution to produce organic gels that are supercritically dried and subsequently pyrolyzed in an inert atmosphere. Therefore, a significant issue in fabricating CNT-CA composites is dispersing the CNTs in the aqueous reaction media. Previous work in the design of CNT-CA composites have addressed this issue by using organic solvents in the sol-gel reaction to facilitate dispersion of the CNTs.9,10 To our knowledge, no data has been published involving the preparation of CA composites containing CNTs dispersed in aqueous media. In this report, we describe a new method for the synthesis of monolithic CNT-CA composites that involves the sol-gel polymerization of resorcinol and formaldehyde in an aqueous solution containing a surfactantstabilized dispersion of double-walled carbon nanotubes (DWNT). One of the advantages of this approach is that it allows one to uniformly distribute CNTs in the CA matrix without compromising the synthetic control that is afforded by traditional organic sol-gel chemistry over the CA structure. Using this approach, we have prepared a series of DWNT-CA nanocomposites that exhibit high surface areas and enhanced electrical conductivities relative to pristine CAs. We will describe the physical characterization of these novel materials as well as the influence of DWNT loading on the electrical conductivity of the CA composite.
Experimental Section Materials. All reagents were used without further purification. Resorcinol (99%) and formaldehyde (37% in water) were purchased from Aldrich Chemical Co. Sodium carbonate (anhydrous) was (14) Foygel, M.; Morris, R. D.; Anez, D.; French, S.; Sobolev, V. L. Phys. ReV. B 2005, 71. (15) Bryning, M. B.; Milkie, D. E.; Islam, M. F.; Hough, L. A.; Kikkawa, J. M.; Yodh, A. G. AdV. Mater. 2007, 19, 661. (16) Bordjiba, T.; Mohamedi, M.; Dao, L. H. J. Electrochem. Soc. 2007, 155, A115.
10.1021/la8011684 CCC: $40.75 2008 American Chemical Society Published on Web 08/09/2008
9764 Langmuir, Vol. 24, No. 17, 2008 purchased from J.T. Baker Chemical Co. Sodium dodecylbenzene sulfonate (SDBS) was purchased from Fluka Chemical Corp., Inc. Purified DWNTs (99%) were purchased from Carbon Nanotechnologies, Inc. Sample Preparation. The DWNT-CA composites were prepared using traditional organic sol-gel chemistry.4 In a typical reaction, purified DWNTs were suspended in an aqueous surfactant solution containing SDBS and thoroughly dispersed using a Bronwill Biosonik IV tip sonicator operating at 25% of maximum power at high frequency. To determine the optimal conditions for DWNT dispersion, a range of sonication times (1-4 h) and SDBS-to-DWNT ratios (10:1, 5:1 and 2.5:1) were evaluated. Once the DWNT were dispersed, resorcinol (1.235 g, 11.2 mmol), formaldehyde (1.791 g, 22.1 mmol) and sodium carbonate catalyst (5.95 mg, 0.056 mmol) were added to the reaction solution. The resorcinol to catalyst ratio (R/C) employed for the synthesis of the composites was ∼200. The sol-gel mixture was then transferred to glass molds, sealed and cured in an oven at 85 °C for 72 h. The resulting gels were then removed from the molds and soaked in acetone for 3 days (changing the acetone every 24 h) to remove all the water from the pores of the gel network. The wet gels were subsequently dried with supercritical CO2 (31.1 °C and 7.38 MPa) and pyrolyzed at 1050 °C (ramp rate: 2 °C/min) under a N2 atmosphere for 3 h. The composite materials were isolated as black cylindrical monoliths. Carbon aerogel composites with DWNT loadings ranging from 0 to 8 wt % (0 to 1.3 vol %) were prepared by this method. For comparison purposes, pristine CAs as well as SDBS-loaded CAs were also prepared using the method described above, except without the addition of the DWNT. Characterization. Bulk densities of the DWNT-CA composites were determined from the physical dimensions and mass of each sample. The volume percent of DWNT in each sample was calculated from the initial mass of DWNTs added, assuming a CNT density of 1.3 g/cm3, and the final volume of the aerogel. Scanning electron microscopy (SEM) characterization was performed on a JEOL 7401F. SEM sample preparation included sputtering a few nanometer layer of Au on the aerogel sample. Imaging was done at 5-10 keV (20 µA) in SEI mode with a working distance of 2-8 mm. Surface area determination and pore volume and size analysis were performed by Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halenda (BJH) methods using an ASAP 2000 Surface Area Analyzer (Micromeritics Instrument Corporation).17 Samples of approximately 0.1 g were heated to 300 °C under vacuum (10-5 Torr) for at least 24 h to remove all adsorbed species. Electrical conductivity was measured using the four-probe method similar to previous studies.3 Metal electrodes were attached to the ends of the cylindrical samples. The amount of current transmitted through the sample during measurement was 100 mA and the voltage drop along the sample was measured over distances of 3-6 mm. Seven or more measurements were taken on each sample. The error bars represent one standard deviation.
Worsley et al.
Figure 1. SEM images of CA nanocomposites containing (a) 1 and (b) 8 w t% DWNTs.
Results and Discussion
work were very hydrophobic, an effective dispersion process was required. Previous work19-27 had identified various methods for the dispersal of carbon nanotubes in water, including the suspension of carbon nanotubes in aqueous solutions of surfactants via sonication. For the synthesis of the DWNT-CA composites, we investigated various surfactants as well as sonication methods to disperse the DWNTs in the sol-gel reaction. Based on our initial results, we found that tip sonication of DWNTs in aqueous solution of sodium dodecylbenzene sulfonate (SDBS) provided the most uniform dispersion of DWNTs in the CA matrix and, therefore, this approach was used to prepare the nanocomposites presented in this report. In a typical synthesis, the DWNT were added to a solution of SDBS in water and dispersed using a tip sonicator. Resorcinol, formaldehyde and the reaction catalyst were then added to the solution and the reaction mixture was cured at elevated temperatures, during which time, black monolithic gels formed. These wet gels were then supercritically dried and carbonized to afford the DWNT-CA composites. Interestingly, during the solvent exchange step prior to supercritical drying, the fluid washed from the pores of the wet gel was clear, indicating that the majority of DWNTs had been incorporated into the aerogel structure. Using this approach, a
Our objective in this work was the design of monolithic CA composites that exhibit enhanced electrical properties through the incorporation of CNTs into the CA matrix. Previous work has shown that the addition of a critical volume fraction of carbon nanotubes to various matrices, if uniformly dispersed, yields significant increases in the electrical conductivity.10,14,18 To achieve uniform distribution of CNTs in CA materials, our approach was to disperse the CNTs in the sol-gel reaction prior to polymerization so that the CNTs can be readily incorporated into the network structure as the polymer framework forms. The resulting RF polymer network containing the CNTs could then be dried and carbonized to afford the CA composite. The main challenge associated with this approach was dispersing the CNTs in water, the reaction medium for the sol-gel reaction, and maintaining the dispersion during polymerization to avoid settling or agglomeration of the CNTs. Since the DWNTs used for this
(17) Gregg, S. J.; Sing, K. S. W. Adsorption, Surface Area and Porosity, 2nd ed.; Academic: London, 1982. (18) Haggenmueller, R.; Guthy, C.; Lukes, J. R.; Fischer, J. E.; Winey, K. I. Macromolecules 2007, 40, 2417. (19) Bandyopadhyaya, R.; Nativ-Roth, E.; Regev, O.; Yerushalmi-Rozen, R. Nano Lett. 2002, 2, 25. (20) Huang, W. J.; Lin, Y.; Taylor, S.; Gaillard, J.; Rao, A. M.; Sun, Y. P. Nano Lett. 2002, 2, 231. (21) Matarredona, O.; Rhoads, H.; Li, Z. R.; Harwell, J. H.; Balzano, L.; Resasco, D. E. J. Phys. Chem. B 2003, 107, 13357. (22) Zheng, M.; Jagota, A.; Semke, E. D.; Diner, B. A.; Mclean, R. S.; Lustig, S. R.; Richardson, R. E.; Tassi, N. G. Nat. Mater. 2003, 2, 338. (23) Islam, M. F.; Rojas, E.; Bergey, D. M.; Johnson, A. T.; Yodh, A. G. Nano Lett. 2003, 3, 269. (24) Dyke, C. a.; Tour, J. M. J. Phys. Chem. A 2004, 108, 11151. (25) Grossiord, N.; Regev, O.; Loos, J.; Meuldijk, J.; Koning, C. E. Anal. Chem. 2005, 77, 5135. (26) Arnold, M. S.; Green, A. A.; Hulvat, J. F.; Stupp, S. I.; Hersam, M. C. Nat. Nanotechnol. 2006, 1, 60. (27) Vaisman, L.; Wagner, H. D.; Marom, G. AdV. Colloid Interface Sci. 2006, 128, 37.
Monolithic Carbon Aerogel Nanocomposites
Langmuir, Vol. 24, No. 17, 2008 9765
Table 1. Physical Properties for a Pristine CA and CA Nanocomposites Containing Dispersions of DWNTs sample
DWNT (wt %)
SDBS-to-DWNT ratio
density (g/cm3)
BET surface area (m2/g)
electrical conductivity (S/cm)
CA DWNT-CA1 DWNT-CA2 DWNT-CA3
0 8 8 8
2.5 5 10
0.170 0.231 0.143 0.139
650 557 580 585
2.5 8.1 5.3 5.0
series of CA composites with DWNT loading ranging from 0 to 8 wt % (0 to 1.3 vol %) were prepared. The microstructures of the DWNT-CA composites were evaluated using scanning electron microscopy. As shown in Figure 1, the network structures of the CAs consist of interconnected networks of primary carbon particles, as would be expected based on the sol-gel reaction formulation. This observation is important as it shows that the formation of the aerogel network is not negatively impacted by the presence of either the surfactant or the DWNTs. These images also show the distribution of DWNTs throughout the CA framework. Clearly, the combination of SDBS surfactant and sonication was effective in maintaining the dispersion of DWNT during the sol-gel polymerization reaction. Based on the SEM images, the DWNTs are dispersed as bundles with diameters of less than 10 nanometers, while the lengths of these bundles are on the order of ∼1 µm. Not surprisingly, the composites prepared with higher loading levels of DWNTs clearly show a higher population of nanotubes in the SEM images. As shown in Table 1, the addition of DWNTs and SDBS to the CA matrix results in a small decrease in BET surface area relative to the pristine CA. The observation is consistent with previous work.10 Nevertheless, these results show that this approach can be used to uniformly disperse CNTs within the CA framework while preserving the desirable textural properties of the pristine CA structure. To determine the effect that incorporation of DWNTs into the CA matrix has on the electrical properties of these materials, the electrical conductivity of the DWNT-CA composites were determined using the four-point probe method. As shown in Figure 2, the electrical conductivity of each composite material is enhanced relative to their respective pristine CA reference. The electrical conductivity enhancement, σenhanced, is given by
σenhanced )
σDWNT-CA - σCA σCA
(1)
where σDWNT-CA and σCA are the measured electrical conductivities of the DWNT-CA composite and the pristine CA, respectively.
Figure 2. Enhancement in electrical conductivity relative to a pristine CA as a function of DWNT concentration for a range of SDBS/DWNT ratios.
This relative enhancement in electrical conductivity was chosen over absolute electrical conductivity so that samples of different densities could be directly compared. In general, the electrical conductivity of the nanocomposites increases as a function of DWNT concentration (Figure 2). The most consistent increases across the range of DWNT concentrations are for the DWNTCAs prepared using 10:1 and 5:1 SDBS:DWNT ratios. In general, samples prepared with SDBS:DWNT ratio of 10:1 showed slightly larger improvements in conductivity than the 5:1 ratio in this case, but the 2.5:1 ratio was decidedly inferior to both. This observation is likely due to improved dispersion of the DWNTs in the reaction mixture at higher surfactant concentrations, as has been seen in previous work with aqueous suspensions of CNTs.21,23 The largest improvements in electrical conductivity were observed in DWNT-CAs with 8 wt% (1.3 vol%) DWNTs, showing a 2-fold increase in conductivity. To verify that these enhancements were attributable to the incorporated DWNTs and not the SDBS surfactant, we also measured the electrical conductivity of reference CA materials that were prepared with SDBS and without the DWNTs. While the data for these materials show modest improvements in conductivity relative to the pristine CA, the effect is small relative to the overall enhancements seen in the DWNT-CA composites. Therefore, these improvements can be attributed to the incorporation of DWNTs into the CA framework. Unlike other CNT composites,9,18,28-31 a clear percolation effect was not detected in the conductivity data for these materials. The observation is likely due to the relatively small contrast in the electrical conductivities between CNTs and CAs as compared to the differences in conductivities between CNTs and other composite matrices, such as organic polymers.32 The relative ratio in electrical conductivity between CNTs and many organic polymers (σnanotube/σpolymer) can be as large as 1014 to 1019. In these materials, below the percolation limit, electrons are forced to travel through the insulating polymer, but, at higher CNT-loading levels, the electrons are able to conduct along the nanotube network and a discontinuity in conductivity (percolation effect) is observed. For the materials investigated here, the relative ratio of electrical conductivities (σnanotube/σCA) is significantly smaller (∼105) as the electrical conductivity of the pristine CA materials can range from 1 to 10 S cm-1. Therefore, in these DWNT-CA composites, a larger fraction of electrons can conduct through the CA framework, relative to organic polymer systems, thus eliminating the percolation effect. If a significant fraction of electrons are indeed traveling through the CA matrix (i.e., percolation effects can be neglected), an effective medium theory can be used to estimate the improvement in electrical conductivity. As in previous studies,33,34 the effective electrical conductivity of a composite consisting of a system of randomly oriented rods (28) Zhan, G. D.; Kuntz, J. D.; Garay, J. E.; Mukherjee, A. K. Appl. Phys. Lett. 2003, 83, 1228. (29) Bryning, M. B.; Islam, M. F.; Kikkawa, J. M.; Yodh, A. G. AdV. Mater. 2005, 17, 1186. (30) Regev, O.; ElKati, P. N. B.; Loos, J.; Koning, C. E. AdV. Mater. 2004, 16, 248. (31) Park, C.; Ounaies, Z.; Watson, K. A.; Crooks, R. E.; Smith, J.; Lowther, S. E.; Connell, J. W.; Siochi, E. J.; Harrison, J. S.; Clair, T. L. S. Chem. Phys. Lett. 2002, 364, 303. (32) Winey, K. I.; Kashiwagi, T.; Mu, M. F. MRS Bull. 2007, 32, 348.
9766 Langmuir, Vol. 24, No. 17, 2008
Worsley et al.
Figure 3. Semilog plot of the enhancement in electrical conductivity relative to pristine CA as a function of DWNT concentration.
in a host medium is
σeff ) σm
(3 + φ(β⊥ + β||)) (3 - φβ⊥)
(2)
where
β⊥ )
2(d(σCNT - σCA) - 2RσσCNTσCA) d(σCNT + σCA) + 2RσσCNTσCA
(2.1)
L(σCNT - σCA) - 2RσσCNTσCA LσCA + 2RσσCNTσCA
(2.2)
β|| )
and σCA and σCNT are the electrical conductivities of the CA and CNT, respectively. For σCA, the electrical conductivity data for the pristine CA materials prepared in this study were used, while a value of 106 S m-1 was used for σCNT.11 Based on the SEM images, the DWNTs are dispersed as bundles with diameters (d) of ∼10 nm and lengths (L) of 1000 nm. The interfacial electrical resistance, Rσ, was used as the fitting parameter. The enhancement in electrical conductivity was calculated as in eq 1, substituting σeff for σDWNT-CA. The results are plotted as a function of volume fraction, φ, in Figure 3. The effective medium theory appears to be in good agreement with the experimental data. The value of Rσ obtained from this data is on the order 10-12 Ω m2. Further support for this theory can be found in the scaling law dependence of the DWNT-CA composites. As shown in Figure 4, the electrical conductivities of the nanocomposites and pristine CAs scale with aerogel density in a similar fashion.3 If the majority of electrons were simply traveling through the DWNT network rather than the CA framework, different scaling law dependence would be expected.
Figure 4. Log-log plot of the electrical conductivity as a function of density for DWNT-CAs (1 vol% DWNTs, 10:1 SDBS/DWNT) and pristine CAs.
of DWNTs. These DWNT-CA composites exhibit high surface areas and enhanced electrical conductivities relative to their respective pristine CAs. The electrical conductivity enhancement as a function of DWNT loading did not show a percolation effect, but followed an effective medium theory. In the course of designing these composites, we felt that further improvements in electrical conductivity might be realized through utilization of dispersion methods that did not require surfactants. Previous work33 has shown that the thermal conductivity enhancement of CNT composites can be degraded by increased interfacial resistance due to the presence of surfactant. In addition, several studies35-38 have shown that the electrical conductivity of carbon nanotubes can be compromised when foreign moieties interact with their surfaces. In the present study, the surfactant was introduced for the specific purpose of interacting with the DWNT surface to aid in dispersion. Therefore, while this approach did yield new CA materials with enhanced electrical conductivity, it would be desirable to know if the presence of SDBS surfactant does indeed limit these improvements. Therefore, we are currently investigating new methods for the dispersion of CNTs in water that would yield CA nanocomposites free of surfactant. Acknowledgement This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract No. DE-AC5207NA27344 and funded by the DOE Office of Energy Efficiency and Renewable Energy. LA8011684
Conclusions This paper describes the synthesis and characterization of monolithic CA nanocomposites containing uniform dispersions (33) Bryning, M. B.; Milkie, D. E.; Islam, M. F.; Kikkawa, J. M.; Yodh, A. G. Appl. Phys. Lett. 2005, 87. (34) Nan, C. W.; Liu, G.; Lin, Y. H.; Li, M. Appl. Phys. Lett. 2004, 85, 3549.
(35) Goldsmith, B. R.; Coroneus, J. G.; Khalap, V. R.; Kane, A. A.; Weiss, G. A.; Collins, P. G. Science 2007, 315, 77. (36) Kruger, M.; Buitelaar, M. R.; Nussbaumer, T.; Schonenberger, C.; Forro, L. Appl. Phys. Lett. 2001, 78, 1291. (37) Lee, Y. S.; Marzari, N. Phys. ReV. Lett. 2006, 97. (38) Mannik, J.; Goldsmith, B. R.; Kane, A.; Collins, P. G. Phys. ReV. Lett. 2006, 97.
