Choking Effect of Single-Wall Carbon Nanotubes on Solvent

The radial breathing mode (RBM) frequency in Raman spectra of single-wall carbon nanotubes (SWNTs) is upshifted (3−9 cm-1) by immersion of the SWNTs...
1 downloads 0 Views 65KB Size
3220

2007, 111, 3220-3223 Published on Web 02/06/2007

Choking Effect of Single-Wall Carbon Nanotubes on Solvent Adsorption in Radial Breathing Mode Hiroaki Honda,† Cheol-Min Yang,‡ Hirofumi Kanoh,†,‡ Hideki Tanaka,§ Tomonori Ohba,†,‡ Yoshiyuki Hattori,| Shigenori Utsumi,‡ and Katsumi Kaneko*,†,‡ Graduate School of Science and Technology, Chiba UniVersity, 1-33 Yayoi, Inage, Chiba 263-8522, Japan, Department of Chemistry, Faculty of Science, Chiba UniVersity, 1-33 Yayoi, Inage, Chiba 263-8522, Japan, Department of Chemical Engineering, Kyoto UniVersity, Nishikyo, Kyoto 615-8510, Japan, and Faculty of Textile Science and Technology, Shinshu UniVersity, 3-15-1 Tokida, Ueda 386-8567, Japan ReceiVed: NoVember 27, 2006; In Final Form: January 9, 2007

The radial breathing mode (RBM) frequency in Raman spectra of single-wall carbon nanotubes (SWNTs) is upshifted (3-9 cm-1) by immersion of the SWNTs in various alcohols and water. However, the immersion in the solvents does not cause any visible variation in the tangential mode frequency. The degree of this RBM frequency upshift induced by the solvent adsorption corresponds to that caused by applying a high pressure of 1 GPa (6-10 cm-1/GPa) in the literature. The degree of the RBM frequency upshift increases with increases in the molecular weight of the solvents. RBM frequency upshift caused by immersion of SWNTs in water is much larger than that by the molecular weight of alcohols, indicating the cluster formation of water molecules.

Introduction Single-wall carbon nanotubes (SWNTs) are unique solids for interface chemistry because all carbon atoms of SWNTs are faced to interfaces of positive and negative curvatures. Furthermore, SWNTs have a tendency to form a bundle structure, offering interstitial and intratube nanopores surrounded by positively and negatively curved single-wall carbons, respectively. Both kinds of bundled SWNT nanopores have a deep interaction potential well for various molecules and ions. The pore structure of the SWNT bundle has been evaluated using molecular adsorption techniques.1-4 At the same time, Raman spectroscopy can determine the diameter of SWNTs through the radial breathing mode (RBM) and X-ray diffraction can provide meaningful information on the lattice structure of the SWNT bundle after removal of the interference effect.5,6 Therefore, the pore structure of SWNT bundles can be evaluated using a combination of the adsorption technique, Raman spectroscopy, and X-ray diffraction. In particular, the effective diameter of the intratube nanopore of SWNTs can be determined by analysis of the RBM Raman spectra. Hence, further study of the governing factors of RBM will be useful for expanding the chemical applications of SWNTs. We also need to understand the essential nature of the single-wall interface system when interacting with the surrounding molecules. The dynamic nature of the single-wall interface should be remarkably sensitive to the surrounding environments in the nanometer scale. * To whom correspondence should be addressed. E-mail: kaneko@ pchem2.s.chiba-u.ac.jp. † Graduate School of Science and Technology, Chiba University. ‡ Department of Chemistry, Chiba University. § Kyoto University. | Shinshu University.

10.1021/jp067856w CCC: $37.00

Active studies on the adsorption properties of SWNTs for supercritical H2 and CH4 have been carried out.7 The quantum molecular sieving effect of SWNTs has been theoretically predicted.8 Tanaka et al. clearly revealed experimentally and theoretically that single-wall carbon nanohorns, being one of the single-wall nanocarbons that can be prepared without metallic catalysts, exhibit an explicit adsorption difference between H2 and D2 even at 77 K.9 These adsorption studies need precise information on the effective pore size distribution of the SWNT bundle. Therefore, environmental factors influencing the RBM must be clearly understood to exactly determine SWNT diameters. The RBM frequency of SWNTs is sensitive to influence by chemical and physical factors as well as tube diameters.5 The chemical factor is chemical doping by electronacceptor or electron-donor dopants (upshift or downshift due to charge-transfers between the dopants and SWNTs).10,11 The most representative physical factors are bundle size (upshift for larger size),12 pressure (upshift at higher pressure),13-15 and temperature (downshift at higher temperature)16 effects, which are related mainly to van der Walls forces and strength of carbon-carbon force constants of SWNTs. Here, we report for the first time on the “choking effect” of SWNTs through RBM by solvation in organic solvents. Experimental Section The SWNTs prepared using a HiPco-process were obtained from Carbon Nanotechnologies, Inc. The SWNTs were not purified, and the Fe content, determined by thermogravimetric analysis, was about 10 wt %. The SWNTs have bundled assembly structures that are about 20-30 nm in diameter (Supporting Information, Figure S1). The SWNT samples were immersed in the solvents at 298 K just after drying the samples © 2007 American Chemical Society

Letters

Figure 1. RBM region in Raman spectra (excitation laser with wavelength of 514.5 nm) of SWNTs immersed in various alcohols; peak 1 at 247 cm-1, peak 2 at 262 cm-1, and peak 3 at 271 cm-1.

at 383 K for 8 h in air. We chose alcohols with a straight-chain structure (methanol, ethanol, propanol, butanol, hexanol, octanol, and decanol) and deionized water with a polar nature. Evaporation of the solvents from the SWNTs was performed at room temperature for 3 days in air. The adsorption of ethanol vapor on the SWNTs was performed at 298 K, after pre-evacuation under a vacuum pressure of less than 10-4 Pa at 433 K for 2 h. The saturated vapor pressure of the ethanol at 298 K is 58 Torr. Raman spectra were taken in the back-scattering configuration and collected with a JASCO NRS-2100 laser Raman spectrometer using an Ar ion laser (514.5 nm) at a laser power of 0.5 mW and 4 µm beam diameter. The excitation laser power was selected carefully in order to avoid the heating of the sample. A silicon wafer and indene were used to calibrate the Raman shift, which guarantees the Raman shift error of 0.5 cm-1 at the peak position. And we obtained a spectral resolution of 1 cm-1 from full width at half-maximum of Rayleigh scattering. The diameters of SWNTs were determined by the relation of ω (cm-1) ) 234/d (nm) + 10 for SWNT bundles, where ω and d are RBM frequency and SWNT diameter, respectively.17-19 Results and Discussion Figure 1 shows Raman spectral changes of the RBM region for SWNTs immersed in various alcohols with different straightchain lengths. We identified six characteristic RBM peaks fitted by using a Lorentzian function (Figure 1). After immersion of SWNTs in various alcohols, all RBM peaks were considerably shifted (3-9 cm-1) to a higher frequency side. Venkateswaran et al. reported hydrostatic pressure-induced shifts of RBM and tangential mode (TM) frequencies in the Raman spectra of SWNT bundles, which is closely related to stiffening of the carbon-carbon bonds in an intratube, intertube interaction within a bundle, and distortion in the tube cross section under compression.13,14 It is interesting that the degree of RBM peak upshift of SWNTs by our liquid-phase solvent adsorption effect

J. Phys. Chem. C, Vol. 111, No. 8, 2007 3221 corresponds to that by high pressure of about 1 GPa; previous results have shown that the rates of pressure-induced upshift in RBM frequency are in the range of 6-10 cm-1/GPa.13-15 Therefore, we propose that liquid-phase adsorption of alcohol molecules on the SWNTs should strengthen the carbon-carbon force constant in the circumferential direction of the SWNT, resulting in an upshift in RBM frequency. Alternatively, the G-band peak frequency at 1590 cm-1 did not change after the immersion in alcohols, which is in good contrast to the pressureinduced G-band upshift (Supporting Information, Figure S2).13 With increases in the pressure, intertube distances within a bundle decrease and ultimately the nanotube geometry of the circular cross section is distorted, resulting in an upshift in RBM and G-band.15 Components influencing the RBM and TM vibrations of SWNTs are intratubular carbon-carbon bond strength and intertubular van der Waals interaction. Here, we assume that the liquid-phase alcohol adsorption on SWNTs does not have an influence on the van der Waals interaction between SWNTs because alcohol molecules are adsorbed only into interstitial channels and the external surfaces of the bundles.20 Therefore, we can consider only the changes in the intratubular vibrations. Liquid-phase alcohol adsorption gives rise to the considerable influence on the vibration in the radial direction, whereas it might not exert any visible influence on the vibration in the tangential direction. When hydrostatic pressure is applied to SWNTs, compressive force should be isotropically applied to the SWNTs, resulting in upshift of both RBM and TM frequencies, which was reported by Eklund and co-workers.13 Alternatively, SWNT surface coverage of the alcohol molecules might be effective for suppressing the vibration motion of the radial direction in preference to that of the tangential direction, resulting in upshift of only RBM frequency. This should be attributed to highly anisotropic geometry due to the high aspect ratio of the SWNTs. Cronin et al. reported that the G-band frequency of an individual SWNT is downshifted by uniaxial strain, whereas the RBM frequency remains unchanged, which probably originates from an elongation of the carbon-carbon bonds caused by strain in the tangential direction.21 Therefore, their results provide a good comparison to ours. This unchanged G-band frequency also provides conclusive evidence that the upshift in RBM frequency after liquid-phase alcohol adsorption is not related to charge transfer between alcohol molecules and SWNTs because charge transfers induced by chemical doping shift both RBM and TM to higher or lower frequency, depending on the type of dopants.10,11 Here, we call our solvent effect the “choking effect” that means suppression effect in breathing motion of RBM, which is different from the compression effect by hydrostatic pressure; the physical adsorption on SWNT surfaces does not induce the choking effect in “geometry or structure of SWNTs” but choking effect in “vibration motion of RBM”. We compared the change of the RBM peak position against the molecular weight of the alcohols for SWNTs with different diameters in order to explain dependence of the alcohol molecular weight on the RBM frequency (Figure 2). We considered only three SWNTs with distinct peak shapes (peak 1 at 247 cm-1, peak 2 at 262 cm-1, and peak 3 at 271 cm-1) to improve the accuracy in demonstrating the dependence of alcohol molecular weight for individual RBM peaks. The degrees of the RBM peak shift were linearly proportional to the molecular weight of the alcohols, indicating a strong dependence on the alcohol molecular weight of the RBM frequency. Dispersion force enables alcohol molecules to approach a single-wall interface easily, and the alcohol mol-

3222 J. Phys. Chem. C, Vol. 111, No. 8, 2007

Figure 2. Relation between RBM peak position and molecular weight of alcohols; peak 1, peak 2, and peak 3 with the corresponding SWNT diameters of 0.99, 0.93, and 0.90 nm, respectively.

Letters ably upshifted after immersion of SWNTs in water; the primary process of wetting is adsorption. This effect of liquid-phase water adsorption on the RBM frequency is also reversible (Supporting Information, Figure S3). The immersion of SWNTs in water does not affect the G-band frequency (Supporting Information, Figure S6), showing a tendency similar to SWNTs immersed in alcohols. The degree of the RBM peak shift caused by immersion in water was significantly higher than the RBM peak position caused by the molecular weight of alcohols. Water molecules can be expected to behave on the SWNTs as if they had a larger molecular weight. The deviation degree is consistent with the cluster size of water molecules observed in graphitic slit nanopores, which was evidenced by the grand canonical Monte Carlo (GCMC) simulation and in situ small-angle X-ray scattering studies (Supporting Information, Figure S5).22,23 Therefore, this large upshift in RBM peaks after water adsorption probably stems from the cluster formation of water molecules on SWNTs. Conclusions In conclusion, we showed that the RBM frequency of SWNTs is reversibly upshifted by alcohols and water adsorption, whereas the G-band frequency is insensitive to solvent adsorption. We assume that the solvent adsorption leads to the choking effect on the breathing motion in the radial direction due to covering of the SWNT surfaces by adsorbed layers of the solvent molecules, resulting in RBM frequency upshift. We again emphasize that degree of this solvent-induced RBM shift corresponds to that by application of 1 GPa pressure. We also showed that the solvent-induced RBM frequency shift depends heavily on the solvent molecular weight.

Figure 3. RBM region in Raman spectra (excitation laser with wavelength of 514.5 nm) of SWNTs immersed in water: (a) raw SWNT, (b) water-SWNT.

ecules then surround the walls of SWNTs. Because the dispersion force increases with molecular weight of adsorbed solvents, alcohol molecules with larger molecular weight interact more strongly with SWNT walls. We can therefore assume that the molecular weight dependence of the RBM frequency upshift can be attributed mainly to the interaction between adsorbed alcohols and SWNT walls. The RBM peak positions were completely recovered after evaporation of alcohols from SWNTs, indicating a reversible choking effect on RBM (Supporting Information, Figure S3). We also observed the RBM frequency variation after adsorption of ethanol vapor on SWNTs at various relative pressures (P/P0) (Supporting Information, Figure S4). The RBM frequency of SWNTs shifted to a higher frequency side with increasing P/P0. A steep increase in the rate of RBM frequency shift was observed at higher P/P0. This can probably be attributed to perfect surface coverage of the ethanol molecules on SWNTs due to completion in monolayer adsorption. The amount of RBM frequency upshift by ethanol vapor adsorption near P/P0 ) 1 is almost equal to that caused by the ethanol immersion treatment. To confirm the effect of solvent polarity in RBM frequency variation, we also used nonpolar solvents such as heptane and decane. After the immersion of SWNTs in nonpolar solvents, the RBM peaks were also shifted to a higher frequency side (not shown here). Hence, this indirectly evidences that the solvent polarity is not a predominant factor in the observed RBM shift. Figure 3 presents an RBM peak change in Raman spectra for SWNTs immersed in water. All RBM peaks were consider-

Acknowledgment. This work was supported by the Grantin-Aid for Scientific Research (S) of the Japan Science Promotion Society (JSPS). We thank Dr. N. Ichikuni and Ms. Y. Komai of Chiba University for their cooperation. Supporting Information Available: Figure S1, TEM image of raw SWNTs; Figure S2, TM region in Raman spectra of SWNTs immersed in various alcohols; Figure S3, RBM and TM regions in Raman spectra after desorption of the alcohols and water from SWNTs; Figure S4, RBM region in Raman spectra of SWNTs adsorbed with ethanol vapor under various relative pressures at 298 K; Figure S5, snapshot of cluster structure of water molecules adsorbed within graphitic slit nanopore; Figure S6, TM region in Raman spectra of SWNTs immersed in water. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Yang, C.-M.; Kaneko, K.; Yudasaka, M.; Iijima, S. Nano Lett. 2002, 2, 385. (2) Yang, C.-M.; Kim, D. Y.; Lee, Y. H. Chem. Mater. 2005, 17, 6422. (3) Williams, K. A.; Eklund, P. C. Chem. Phys. Lett. 2000, 320, 352. (4) Ohba, T.; Kaneko, K. J. Phys. Chem. B 2002, 106, 7171. (5) Rao, A. M.; Richter, E.; Bandow, S.; Chase, B.; Eklund, P. C.; Williams, K. A.; Fang, S.; Subbaswamy, K. R.; Menon, M.; Thess, A.; Smalley, R, E.; Dresselhaus, G.; Dresselhaus, M. S. Science 1997, 275, 187. (6) 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.; Tomanek, D.; Fischer, J. E.; Smalley, R. E. Science 1996, 273, 483. (7) Dillon, A. C.; Jones, K. M.; Bekkedahl, T. A.; Kiang, C. H.; Bethune, D. S.; Heben, M. J. Nature 1997, 386, 377. (8) Challa, S. R.; Sholl, D. S.; Johnson, J. K. Phys. ReV. B 2001, 63, 245419. (9) Tanaka, H.; Kanoh, H.; Yudasaka, M.; Iijima, S.; Kaneko, K. J. Am. Chem. Soc. 2005, 127, 7514.

Letters (10) Rao, A. M.; Eklund, P. C.; Bandow, S.; Thess, A.; Smalley, R. E. Nature 1997, 388, 257. (11) Skakalova, V.; Kaiser, A. B.; Dettlaff-Weglikowska, U.; Hrncarikova, K.; Roth, S. J. Phys. Chem. B 2005, 109, 7174. (12) Milnera, M.; Kurti, J.; Hulman, M.; Kuzmany, H. Phys. ReV. Lett. 2000, 84, 1324. (13) Venkateswaran, U. D.; Rao, A. M.; Richter, E.; Menon, M.; Rinzler, A.; Smalley, R. E.; Eklund, P. C. Phys. ReV. B 1999, 59, 10928. (14) Venkateswaran, U. D.; Masica, D. L.; Sumanasekera, G. U.; Furtado, C. A.; Kim, U. J.; Eklund, P. C. Phys. ReV. B 2003, 68, 241406. (15) Peters, M. J.; McNeil, L. E.; Lu, J. P.; Kahn, D. Phys. ReV. B 2000, 61, 5939. (16) Raravikar, N. R.; Keblinski, P.; Rao, A. M.; Dresselhaus, M. S.; Schadler, L. S.; Ajayan, P. M. Phys. ReV. B 2002, 66, 235424.

J. Phys. Chem. C, Vol. 111, No. 8, 2007 3223 (17) Ku¨rti, J.; Kresse, G.; Kuzmany, H. Phys. ReV. B 1998, 58, R8869. (18) Henrard, L.; Herna´ndez, E.; Bernier, P.; Rubio, A. Phys. ReV. B 1999, 60, R8521. (19) Pfeiffer, R.; Kuzmany, H.; Kramberger, C.; Schaman, C.; Pichler, T.; Kataura, H.; Achiba, Y.; Ku¨rti, J.; Zolyomi, V. Phys. ReV. Lett. 2003, 90, 225501. (20) Yang, C.-M.; Kanoh, H.; Kaneko, K.; Yudasaka, M.; Iijima, S. J. Phys. Chem. B 2002, 106, 8994. (21) Cronin, S. B.; Swan, A. K.; Unlu, M. S.; Goldberg, B. B.; Dresselhaus, M. S.; Tinkham, M. Phys. ReV. Lett. 2004, 93, 167401. (22) Ohba, T.; Kanoh, H.; Kaneko, K. J. Am. Chem. Soc. 2004, 126, 1560. (23) Ohba, T.; Kanoh, H.; Kaneko, K. Nano Lett. 2005, 5, 227.