NANO LETTERS
Enhanced Hydrogen Adsorptivity of Single-Wall Carbon Nanotube Bundles by One-Step C60-Pillaring Method
2009 Vol. 9, No. 11 3694-3698
Miki Arai,† Shigenori Utsumi,*,‡ Mamiko Kanamaru,† Koki Urita,† Toshihiko Fujimori,† Noriko Yoshizawa,§ Daisuke Noguchi,† Katsuhiro Nishiyama,‡ Yoshiyuki Hattori,| Fujio Okino,| Tomonori Ohba,† Hideki Tanaka,⊥ Hirofumi Kanoh,† and Katsumi Kaneko† Graduate School of Science, Chiba UniVersity, 1-33 Yayoi, Inage, Chiba 263-8522, Japan, Department of Mechanical Systems Engineering, Tokyo UniVersity of Science, Suwa, 5000-1 Toyohira, Chino 391-0292, Japan, DiVision of Energy Technology, Center for AdVanced Carbon Materials, National Institute of AdVanced Industrial Science and Technology (AIST), Tsukuba 305-8565, Japan, Faculty of Textile Science and Technology, Shinshu UniVersity, 3-15-1 Tokida, Ueda 386-8567, Japan, and Department of Chemical Engineering, Kyoto UniVersity, Nishikyo-ku, Kyoto 615-8510, Japan Received May 18, 2009; Revised Manuscript Received September 11, 2009
ABSTRACT Single-wall carbon nanotube (SWCNT) bundles were pillared by fullerene (C60) by the cosonication of C60 and SWCNT in toluene to utilize the interstitial pores for hydrogen storage. C60-pillared SWCNTs were confirmed by the shift in the X-ray diffraction peak and the expanded hexagonal and distorted tetragonal bundles revealed by high-resolution transmission electron microscopy. The H2 adsorptivity of the C60pillared SWCNT bundles was twice that of the original SWCNT bundles, indicating a design route for SWCNT hydrogen storage.
Single-wall carbon nanotubes (SWCNTs) are considered to be the most promising material for a new sustainable chemistry1-4 and particularly in hydrogen storage, because SWCNT bundles have both internal and interstitial nanospaces that strongly interact even with supercritical H2.5-7 Consequently, research on the purification, dispersion, and functionalization of SWCNTs has been intensively carried out in order to extend the inherent properties and develop superquality materials.8-10 One prominent point regarding SWCNTs must be stressed. A SWCNT is essentially an interfacial material, being remarkably different from other solid materials in that all component carbon atoms are exposed to both surfaces, each with different nanoscale curvatures. A SWCNT has a huge geometrical surface area of 2630 m2 g-1, the same as graphene. The effective surface area of SWCNTs for molecules varies with its tube diameter and the target molecular size. In addition to the large surface area, the differences between surfaces with positive and * To whom correspondence should be addressed. Tel: +81-266-73-9850. Fax: +81-266-73-1230. E-mail:
[email protected]. † Chiba University. ‡ Tokyo University of Science, Suwa. § AIST. | Shinshu University. ⊥ Kyoto University. 10.1021/nl9015733 CCC: $40.75 Published on Web 10/20/2009
2009 American Chemical Society
negative curvature can be exploited to establish unique material science and technology. However, when the interstitial pore width is just comparable to the size of a small molecule, the molecules preadsorbed in the interstitial nanospaces often block further adsorption, or the capacity of the interstitial pore spaces is too small compared with the internal nanospace capacity. Hata et al. developed highly pure SWCNTs without a bundle structure, stimulating interfacial research on SWCNTs.11 Very recently, the authors used the SWCNT samples developed by Hata et al. to uncover evidence that a monolayer of N2 molecules adsorbed on the internal wall (negative curvature) of a SWCNT is more ordered than a monolayer on the external wall (positive curvature).12 Thus, SWCNTs have an explicit bisurface nature for molecules, which would be applicable in the development of intriguing and novel materials with multiple interfacial functions. Ordinary SWCNTs associate to form an ordered bundle structure through dispersion interaction, providing interstitial pore spaces surrounded by carbon walls with positive curvature, which have the strongest molecular sites. Therefore, bundled SWCNTs have considerable potential for application to gas storage, the stabilization of unstable molecules, quantum molecular sieving, specific reaction fields, gas sensing, electrochemical energy storage,
Figure 1. Relation between the C60-doped amount on the SWCNTs and C60-toluene solution concentration. (a) C60-doped amount on 1 g -1 -1 of SWCNT (g gSWCNT ) is plotted against concentration of C60-dissolved toluene (concentration: 0-2.8 g LToluene ). (b) Magnified view of Figure 1a. The broken lines represent the calculated C60-doped amount corresponding to perfect filling of the interstitial nanospaces in the model structure of C60-pillared SWCNT bundles with hexagonal symmetry.
and so on.13 However, it is necessary to establish a means for tuning the bundle structure for providing the larger capacity of internal and interstitial nanospaces with an optimum size for the target function, as the volume of the interstitial nanospaces at the strongest sites is too small. Pillaring an SWCNT bundle is the best approach to control interstitial nanoporosity, realizing enhanced adsorptivity for supercritical gases such as H2, and strengthening the specificity of the molecular recognition function.14,15 Here we report the simple preparation of fullerene16 (C60)-pillared SWCNT bundles by sonication of SWCNTs in a C60 toluene solution and the consequent enhancement of the supercritical H2 adsorptivity of the SWCNTs. As C60 molecules have a conjugated π-electron structure similar to that of SWCNTs, the C60-pillared SWCNT system can be regarded as a new nanocarbon. SWCNTs that give an X-ray diffraction (XRD) peak due to their ordered hexagonal bundle structure17,18 were prepared by the laser ablation of a graphite rod in the presence of Ni and Co.19,20 Closed SWCNT samples were used to clearly show the effect of C60-pillaring; N2 adsorption measurements at 77 K revealed that the purified SWCNTs were closed. The tube diameter of the SWCNTs (dSWCNT) was 1.37 nm, determined by the radial breathing mode (RBM) band of the Raman spectrum21 (see Supporting Information, Figure S1 for the characterization of the purified SWCNT before C60pillaring). For C60-pillaring, we applied the methods used for the adsorption of organic substances on SWCNTs22 and the preparation of peapod SWCNTs;23 C60-pillared SWCNTs were prepared by a simple sonication of SWCNT in C60 toluene solution with different concentrations. The C60pillared SWCNTs are designated as SWCNT-C60(x), where x is the amount in gram of C60 doped to 1 g of SWCNTs. The C60-doped amount against the C60 concentration of the toluene solution (g L-1 Toluene) is shown in Figure 1 (see Supporting Information, Table S1 for details). The C60 uptake versus the C60 concentration curve has a step near 0.7 g -1 gSWCNT of uptake and 0.5 g L-1 Toluene of the C60 concentration; the step indicates the formation of a stable structure between C60 and the SWCNTs. The uptake at the step closely corresponds to amount required for perfect filling of the Nano Lett., Vol. 9, No. 11, 2009
interstitial spaces by C60 molecules, as estimated from the interstitial spaces in the model structure of an SWCNT bundle and the uptake of C60 for a trigonal arrangement (Supporting Information, Figure S2).24 This stable structure, that is, SWCNT-C60(0.646), provides the maximum nanopore volume in N2 adsorption measurements at 77 K, as shown in Figure 2a (and Supporting Information, Table S1. N2 adsorption isotherms can be seen in Supporting Information, Figure S3.). Hence, SWCNT-C60(0.646) should have the optimum C60-pillared structure for the acceptance of molecules in the interstitial nanospaces expanded by C60pillaring. A C60 concentration higher than 1.0 g L-1 Toluene should induce further C60-pillaring and coating of the external surface of the SWCNT bundle, thus reducing the nanopore volume. Figure 3 shows high-resolution transmission electron microscopy (HR-TEM) images of the bundle structure of SWCNT-C60(0.646), which should have the optimum structure to adsorb molecules. The wide-range observation (Figure 3a) shows a well-aligned bundle sheet even after ultrasonic C60-pillaring treatment in the toluene solution. The side-view observation (Figure 3b) indicates that C60 molecules are present on the SWCNT surfaces and there are no peapod SWCNTs (i.e., C60 in the SWCNTs). Figure 3c,d shows cross sections of the SWCNT-C60(0.646) bundle having expanded hexagonal and tetragonal arrays, respectively. The intertube distance of the expanded bundle is estimated to be ∼2.2 nm, leading to an interlayer distance d′ ) 1.9 nm for hexagonal symmetry and d′′ ) 1.8 nm for tetragonal symmetry, under the assumption of uniform bundle structure for each symmetry. The interlayer distances d′ and d′′ of the bundles of hexagonal and tetragonal superlattices of the C60-pillared SWCNT bundle are 2.03 and 1.92 nm, respectively, according to geometrical evaluation, which were close to the values determined from TEM, as shown in Figure 4. Thus, SWCNTC60(0.646) should have a mixed C60-pillared structure with both hexagonal and tetragonal symmetries. XRD patterns, as shown in Figure 5, support the above-mentioned C60pillared SWCNT structure. SWCNT-C60(0), which was ultrasonically treated in toluene without C60, gives an explicit peak at 2.81° (X-ray source: MoKR), corresponding to the 3695
Figure 2. Changes in nanoporosity and H2 adsorption amounts of SWCNT bundles with C60-pillaring. (a) Relation between the nanospace volume and the C60-doped amounts x. (b) Relation between the amount of H2 adsorbed and x (0.005 MPa (O), 0.01 MPa (b), and 0.1 MPa (0)) measured at 77 K.
Figure 3. TEM images of SWCNT-C60(0.646). (a) A well-aligned bundle sheet (low magnification). (b) Side view of a C60-pillared SWCNT. (c) Cross-section of an expanded hexagonal bundle. (d) Cross-section of a distorted-tetragonal array bundle.
interlayer distance (d ) 1.44 nm (experimental)) of a hexagonal lattice of SWCNT arrays (see Figure 4a), whereas individual SWCNTs have no diffraction peak in the concerned diffraction angle-range. This peak is weakened by the C60-pillaring treatment and a broad peak appears around 3696
2θ ) 2.0°. The new peak corresponds to an interlayer distance of ∼2.0 nm, which is the average of 2.03 and 1.92 nm, derived from the TEM-derived two-structure models. Thus, XRD clearly indicates the formation of C60-pillared SWCNT bundles. However, the pillared structure is not Nano Lett., Vol. 9, No. 11, 2009
Figure 4. Evaluation of interlayer distances for plausible pillared structures. Geometrical derivation of the interlayer distance d of an SWCNT bundle with a hexagonal arrangement (a), the interlayer distance d′ of a C60-pillared SWCNT bundle with a hexagonal arrangement (b), and the interlayer distance d′′ of a C60-pillared SWCNT bundle with a tetragonal array (c).
necessarily regular; hexagonal and tetragonal structures coexist, and therefore a broad superlattice peak is observed. Thermogravimetric analysis (TGA) (Supporting Information, Figure S4) showed pillaring of C60 in the SWCNT bundles. The sample weight decrease is much less than the C60-doped amounts, except in the region of low C60-doped amount. The considerable weight decrease in the low C60 region is caused by the elimination of amorphous carbon and oxygen functional groups. On the other hand, only a part of the C60doped amounts adsorbed on the external surface of the SWCNT bundles can be eliminated in the high C60 region. This indicates that major C60 molecules inserted in the strong potential sites of interstitial pores cannot be eliminated. Surface composition analysis with X-ray photoelectron spectroscopy (XPS) also supports pillaring of C60 molecules in SWCNT bundles. XPS analysis of the C1s spectrum25 of SWCNT-C60(0.646) by fitting with the C1s spectra of SWCNT and C60 shows the presence of 10% of C60 on the bundle surface, much less than the bulk content (39%) (Supporting Information, Figure S5). XPS detects electrons only from surface layers in the order of 1 nm; predominant C60 molecules are not on Nano Lett., Vol. 9, No. 11, 2009
the external surface of the SWCNT bundle, but inside of the SWCNT bundle as pillars. The absence of C60 on the bundle surface of SWCNT-C60(0.646) is revealed by Raman spectroscopy, which is also surface sensitive. The Raman peak for C60 at 1467 cm-1 appears only for the SWCNT-1 C60(x) samples whose x is larger than 1.68 g gSWCNT 26,27 (Supporting Information, Figure S6). Thus, all characterization results confirm that SWCNT-C60(0.646) has a promising C60-pillared structure with adequate nanoporosity. The change in the interaction strength of the adsorption sites can be sensitively detected by supercritical H2 adsorption. The adsorbed amounts of supercritical H2 on C60-pillared SWCNTs at 77 K are plotted against the C60-doped amount x, as shown in Figure 2b. As the critical temperature of H2 is 33 K, H2 at 77 K is supercritical, and thereby the adsorption of H2 even at 77 K needs intensive assistance with interaction potential from solid nanospaces.28,29 C60-pillaring markedly enhances the adsorption of H2, providing almost twice the adsorption amount of SWCNTs in the low-pressure region, and 1.3 times higher in the ambient pressure region. Upward concave H2 adsorption isotherms (see Supporting Information, Figure S7) 3697
utilization of interstitial nanopore spaces for various fields, such as electrochemical, adsorption, sensor, and separation technologies. Acknowledgment. The authors are grateful to Dr. Kunimitsu Takahashi at Institute of Research and Innovation (IRI) for providing SWCNT samples. This research was founded with Grant-in-Aids for Scientific Research (S) from Japanese Government. Supporting Information Available: Sample preparation procedure, characterization method and results, Raman and X-ray photoelectron spectra, TGA results, N2 and H2 adsorption isotherms at 77 K, and H2 adsorption heat derived from the DR plot. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) (2) (3) (4) (5)
Figure 5. XRD patterns of SWCNT-C60(x) and C60. The vertical solid line indicates the superlattice-peak position of an ordered SWCNT bundle with hexagonal symmetry. The broken line indicates the almost overlapped position of the superlattice peaks of hexagonal and tetragonal array models of the C60-pillared SWCNT bundle.
(6) (7) (8) (9) (10)
indicate the presence of strong adsorption sites, even for supercritical H2. In addition, the desorption branch is situated above the adsorption branch, and therefore there should be very narrow nanospaces having strong interaction potential where strongly bound H2 molecules mediate further adsorption and diffusion in the interstitial spaces. The isosteric heat of H2 adsorption evaluated from the supercritical Dubinin-Radushkevich (DR) analysis30 is 9.5-9.7 kJ mol-1 (see Supporting Information, Figure S8), and does not change remarkably with C60-pillaring. Therefore, enhancement of the H2 adsorptivity of C60-pillared SWCNTs results from the increase in the nanospace volume. However, this value is much greater than the condensation enthalpy of H2 molecules at the boiling point (∼20 K; 0.22 kJ mol-1) and the isosteric heat on the interstitial sites of (12, 12) SWCNT bundles (∼9 kJ mol-1) or on the slit pores of carbon (∼8 kJ mol-1) evaluated from Grand Canonical Monte Carlo simulation, indicating strong H2 molecule-interstitial pore interaction in the C60-pillared SWCNT bundles. In addition, C60, as the pillar, should deepen the interaction potential wall for H2. In conclusion, a one-step method for C60-pillaring in SWCNT bundles was proposed, that is, the cosonication of C60 and SWCNT in toluene. C60-pillared SWCNT bundle structure provides enhanced adsorptivity for supercritical H2. Indeed, the H2 adsorption amounts on the C60-pillared SWCNT bundles doubly increased, compared with nontreated SWCNT bundles. TEM observation revealed that the C60-pillared SWCNT bundle had expanded hexagonal and distorted tetragonal arrays. These expanded interstitial nanospaces were also substantiated by a new XRD peak corresponding to the interlayer distances of SWCNTs in which C60 molecules were surrounded by three or four SWCNTs. These results indicate a simple and promising tuning route for SWCNT bundle structures, allowing the 3698
(11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30)
Iijima, S. Nature 1991, 354, 56. Iijima, S.; Ichihashi, T. Nature 1993, 363, 603. Hirsch, A. Angew. Chem., Int. Ed. 2002, 41, 1853. Saito, R.; Takeya, T.; Kimura, T.; Dresselhaus, G.; Dresselhaus, M. S. Phys. ReV. B 1998, 57, 4145. Liu, C.; Fan, Y. Y.; Liu, M.; Cong, H. T.; Cheng, H. M.; Dresselhaus, M. S. Science 1999, 286, 1127. Wang, Q.; Johnson, K. J. J. Phys. Chem. B 1999, 103, 4809. Seung, M. L.; Young, H. L. Appl. Phys. Lett. 2000, 76, 2877. Dillon, A. C.; Gennett, T.; Jones, K. M.; Alleman, J. L.; Parilla, P. A.; Heben, M. J. AdV. Mater. 1999, 16, 1354. Li, F.; Cheng, H. M.; Xing, Y. T.; Tan, P. H.; Su, G. Carbon 2000, 38, 2041. Islam, M. F.; Rojas, E.; Bergey, D. M.; Johnson, A. T.; Yodh, A. G. Nano Lett. 2003, 3, 269. Hata, K.; Futaba, D. N.; Mizuno, K.; Namai, T.; Yumura, M.; Iijima, S. Science 2004, 306, 1362. Ohba, T.; Matsumura, T.; Hata, K.; Yumura, M.; Iijima, S.; Kanoh, H.; Kaneko, K. J. Phys. Chem. C 2007, 111, 15660. Banerjee, S.; Kahn, M. G. C.; Wong, S. S. Chem.sEur. J. 2003, 9, 1898. Abrams, Z. R.; Ioffe, Z.; Tsukernik, A.; Cheshnovsky, O.; Hanein, Y. Nano Lett. 2007, 7, 2666. Zhao, Y.; Sugai, T.; Shinohara, H.; Saito, Y. J. Phys. Chem. Solids 2007, 68, 284. Kroto, H. W.; Heath, J. R.; O’Brien, S. C.; Curl, R. F.; Smalley, R. E. Nature 1985, 318, 162. Thess, A.; Lee, R.; Nikolaev, P.; Dai, H.; Petit, P.; Robert, J.; Xu, C.; Lee, Y. H.; Kim, S. G.; Rinzler, A. G.; Colbert, D. T.; Scuseria, G. E.; Toma´nek, D.; Fischer, J. E.; Smalley, R. E. Science 1996, 273, 483. Abe, M.; Kataura, H.; Kira, H.; Kodama, T.; Suzuki, S.; Achiba, Y.; Kato, K.; Takata, M.; Fujiwara, A.; Matsuda, K.; Maniwa, Y. Phys. ReV. B 2003, 68, 041405. Yudasaka, M.; Kokai, F.; Takahashi, K.; Yamada, R.; Sensui, N.; Ichihashi, T.; Iijima, S. J. Phys. Chem. B 1999, 103, 3576. Kokai, F.; Takahashi, K.; Yudasaka, M.; Iijima, S. J. Phys. Chem. B 2000, 104, 6777. Kataura, H.; Kumazawa, Y.; Maniwa, Y.; Umezu, I.; Suzuki, S.; Ohtsuka, Y.; Achiba, Y. Synth. Met. 1999, 103, 2555. Gotovac, S.; Honda, H.; Hattori, Y.; Takahashi, K.; Kanoh, H.; Kaneko, K. Nano Lett. 2007, 7, 583. Yudasaka, M.; Ajima, K.; Suenaga, K.; Ichihashi, T.; Hashimoto, A.; Iijima, S. Chem. Phys. Lett. 2003, 380, 42. Williams, K. A.; Eklund, P. C. Chem. Phys. Lett. 2000, 320, 325. Utsumi, S.; Honda, H.; Hattori, Y.; Kanoh, H.; Takahashi, K.; Sakai, H.; Abe, M.; Yudasaka, M.; Iijima, S.; Kaneko, K. J. Phys. Chem. C 2007, 111, 5572. Dresselhaus, M. S.; Dresselhaus, G.; Eklund, P. C. J. Raman Spectrosc. 1998, 27, 351. Rao, A. M.; Eklund, P. C.; Hodeau, J.-L.; Marques, L.; NunezRegueiro, M. Phys. ReV. B 1997, 55, 4766. Kaneko, K.; Murata, K. Adsorption 1997, 3, 197. Xu, W.-C.; Takahashi, K.; Matsuo, Y.; Hattori, Y.; Kumagai, M.; Ishiyama, S.; Kaneko, K.; Iijima, S. Int. J. Hydrogen Energy 2007, 32, 2504. Kaneko, K.; Shimizu, K.; Suzuki, T. J. Chem. Phys. 1992, 97, 8705.
NL9015733 Nano Lett., Vol. 9, No. 11, 2009
