NANO LETTERS
Thermal Conversion of Bundled Carbon Nanotubes into Graphitic Ribbons
2005 Vol. 5, No. 11 2195-2201
H. R. Gutie´rrez,†,‡ U. J. Kim,†,‡ J. P. Kim,| and P. C. Eklund*,‡,§ Department of Physics and Department of Materials Science, The PennsylVania State UniVersity, UniVersity Park, PennsylVania 16802, and Korea Basic Science Institute, Busan, South Korea Received July 4, 2005; Revised Manuscript Received September 12, 2005
ABSTRACT High temperature heat treatment (HTT) of bundled single-walled carbon nanotubes (SWNTs) in vacuum (∼10-5 Torr) has been found to lead to the formation of two types of graphitic nanoribbons (GNRs), as observed by high-resolution transmission electron microscopy. Purified SWNT bundles were first found to follow two evolutionary steps, as reported previously, that is, tube coalescence (HTT ≈ 1400 °C) and then massive bond rearrangement (HTT ≈ 1600 °C), leading to the formation of bundled multiwall nanotubes (MWNTs) with 3−12 shells. At HTT > 1800 °C, we find that these MWNTs collapse into multishell GNRs. The first type of GNR we observed is driven by the collapse of diameterdoubled single-wall nanotubes, and their production is terminated at HTT ≈ 1600 °C when the MWNTs also start to form. We propose that the collapse is driven by van der Waals forces between adjacent tubes in the same bundle. For HTT > 2000 °C, the heat-treated material is found to be almost completely in the multishell GNR form.
Carbon nanotubes, discovered by Iijima1,2 and Ajayan3 and their research groups, have captured the attention of a large international community interested in the fundamental nanoscience4,5 of these molecular filaments as well as the possibilities for new technology.6,7 Likewise, the thermal evolution of different carbon nanostructures such as nanocones8 and bundles of double-walled (DWNT)9,10 and singlewalled carbon nanotubes (SWNTs)11-16 at very hightemperature remains an interesting fundamental issue that has attracted intense research efforts in the last few years. Two surprising morpholological changes have already been observed when annealing bundles of SWNTs at high temperatures in an inert gas or vacuum.11-16 The first change observed was at ∼1300-1400 °C in bundled SWNTs prepared by pulsed laser vaporization (PLV). Smalley and co-workers at Rice University found that they coalesce into larger diameter tubes.11 The coalescence was found to be significant (i.e., 70-90%) and rapid (∼1 h).11 Thermal coalescence has since been observed in tubes produced by the high-pressure disproportion of CO (HiPCO)12 and also by the electric arc discharge (ARC)13,14 process. The second interesting morphological transformation in bundled SWNTs occurs at a higher heat treatment temperature (HTT) ∼16002000 °C. In this temperature range, original and coalesced tubes undergo massive bond rearrangement, transforming into * Corresponding author. E-mail:
[email protected]. † These authors contributed equally to this work. ‡ Department of Physics, The Pennsylvania State University. § Department of Materials Science, The Pennsylvania State University. | Korea Basic Science Institute. 10.1021/nl051276d CCC: $30.25 Published on Web 10/06/2005
© 2005 American Chemical Society
three- to six-shell multiwalled carbon nanotubes (MWNTs).15,16 These past transformation studies on bundled SWNTs were carried out on unpurified material containing significant amorphous carbon and a growth catalyst. The effect of these impurity phases on the thermal transformation is not known. In this letter, we report the observation of new multishell carbon filaments that we call graphitic nanoribbons (GNR). These GNRs were observed to form from purified bundled SWNTs heat treated at high temperature (HTT) above ∼1800 °C. For HTT > 2000 °C, the GNR filaments are the dominant form of carbon in the sample, as observed by highresolution transmission electron microscopy (HRTEM). To the best of our knowledge, these multishell GNRs are reported here for the first time. Interestingly, a single graphene sheet can be found at the center of some of these GNRs. SWNTs made from the electric arc discharge tubes of Ni-Y loaded carbon electrodes were obtained from CarboLex, Inc. They were purified as follows: Amorphous carbon was removed by dry oxidation at ∼350 °C for ∼20 min, and then the metal catalyst residue was removed by a 24 h reflux in 6 N HCl. The purification is described in more detail elsewhere.17 The purified material was then placed in a high-temperature vacuum furnace (“Red Devil”, R. D.Webb, Inc.) and degassed in 10-5 Torr at 200 °C overnight. This was done to remove strongly bound O2 at temperatures below which it might react with the sample (i.e., removing carbon as CO2). After this long-term, low-T degassing step, the furnace was then ramped at 10 °C /min and then held at a
Figure 1. (a) High-resolution TEM image of the cross section of a typical as-grown SWNT bundle, produced by Arc discharge process. (b) Schematic representation of the triangular spatial arrangement in the bundle. (c-e) diameter distributions of singlewall nanotubes after heat-treatment temperatures (HTT) of 1800, 1600, and 200 °C, respectively.
specific HTT for ∼4 h. The material was then cooled to room temperature over ∼8 h. Thermal evolution of the Arc material annealed at HTT ) 1100, 1400, 1600, 1800, 2000, and 2200 °C was studied mainly by TEM and Raman spectroscopy. A detailed discussion of the Raman study will be presented elsewhere.18 Samples for transmission electron microscopy were prepared by dispersing the material in EtOH (Aldrich); drops of this dispersion were deposited onto lacey carbon grids (from Ted Pella, Inc.). For TEM images of the bundle crosssection, we applied nanotube powder (without dispersion in ethanol) directly to the grids. This procedure seemed to increase the number of bundles aligned perpendicular to the plane of the grid and therefore parallel to the e beam. Highresolution transmission electron microscope (HRTEM) images were obtained with a JEOL 2010F (200 kV) (point-topoint resolution 1.9 Å). Raman spectra were collected under ambient conditions from thin films of the material deposited onto glass slides. These thin films were prepared from the same dispersions used for TEM analysis. A JY Horiba T64000 spectrometer with an Olympus BX40 confocal microscope (∼1 µ diameter focal spot size) and cooled CCD detector was used to collect spectra. Excitation was provided by an Ar-Kr laser at 1-3 mW incident power, as measured at the sample position. Both H- and V-polarized light were accepted in the scattered radiation. In Figure 1a, we show a representative HRTEM image of SWNT bundle cross section of ARC material (CarboLex, Inc.) that had been vacuum-degassed at 200 °C for ∼8 h. This temperature is too low to promote any C-C bond rearrangement. The diameter distributions, obtained from 2196
TEM images, are also shown in Figure 1c-e for HTT ) 200, 1600, and 1800 °C. The as-grown Arc SWNTs were found to exhibit a most probable diameter dh ) 1.4 nm and a relatively narrow diameter distribution largely contained between 1.2 and 1.8 nm (Figure 1e). This narrow distribution promotes a triangular lattice packing of the SWNT in the rope (Figure 1b). A triangular lattice provides tubes surrounded by six nearest-neighbors. Each pair of tubes in this arrangement has almost the same probability of coalescence. Coalescence can be observed from the nanotube diameter distribution after HTT ) 1600 °C and 1800 °C. The effects of “diameter doubling” on the distributions (Figure 1d and e) are clearly visible. A few tubes with diameter d ) 3 × dh are observed for HTT ) 1800 °C. It is important to notice that after annealing at 1600 and 1800 °C, tubes with diameters d < 1.4 nm represent only a small fraction of the sampe, that is, 4-6% of the sample. This suggests that a d ≈ 1.4 nm is the smallest diameter nanotube that can withstand coalescence in the 1600-1800 °C range. Our data suggest that once the nanotubes double (or triple) their diameter, they become significantly more stable against coalescence for this range of HTT. The width of the diameter distribution increases significantly from HTT ) 1600-1800 °C (Figure 1); and the distribution shifts to larger diameters. This suggests that the coalescence after HTT ) 1600 °C involves the smaller diameter primary nanotubes (d e 1.6 nm); and at 1800 °C, even the larger diameter primary nanotubes (e.g., d g 1.6 nm) appear to coalesce. By “primary” we mean tubes produced in the original (ARC) growth process (not by coalescence). However, the formation of tubes with d g 3.4 nm at HTT ) 1800 °C could also be originated by the simultaneous coalescence of three SWNTs, which may be important at higher temperatures. This view assumes that two successive coalescence transformations involving the same tube (i.e., coalescence to ∼2dh, followed by coalescence of 2dh and dh tubes to yield a 3dh tube) have a low probability because of the significantly higher stability of diameterdoubled (2dh) tubes. Almost no SWNTs with diameters 4 times larger than the mean starting diameter were observed. This observation is in agreement with recent molecular dynamic (MD) simulations19 carried out at effective high temperature in which the simultaneous coalescence of three SWNTs in an initial triangular arrangement were found to lead to the largest coalesced tube. Their simulations also suggest that strong bond rearrangement activity involving more than four closepacked tubes leads to the formation of a MWNT.19 Other simulations suggest that if a very large diameter SWNT is generated, the tube will collapse into a two-layer graphitic ribbon20,21 and will therefore not exhibit a cylindrical crosssection in the product. HRTEM images of these collapsed tubes could easily be misinterpreted as a two-layer graphitic sheet with “edge” atoms, depending on the orientation of the electron beam.21 At higher temperature, T g 1600 °C, we also observe the second interesting transformation to MWNTs reported previously15,16 (Figure 2). In this transformation, bundles of Nano Lett., Vol. 5, No. 11, 2005
Figure 2. HRTEM images of purified ARC material after HTT ) 1800 °C. (a and b) massive bond rearrangement leads to bundles of MWNTs. (c) Map of number of walls vs inner diameter for MWNTs at this HTT. (d) Type-I graphitic nanoribbon (GNR); the closed ends suggest the collapse of a bundle of diameter-doubled SWNTs.
coalesced tubes are observed to transform into bundles of MWNTs. We also observed a few graphitic nanoribbons (GNRs) after HTT ) 1600 °C. We found that most of the SWNTs no longer exhibit straight walls; corrugated walls are observed instead. GNRs were detected in the HRTEM images as very long filaments of parallel sp2 layers (Figure 2d). After HTT ) 1800 °C, most of the MWNTs were found in HRTEM images to exhibit well-formed tubules with a small number of defects. Also, the material was found to contain a significantly higher fraction of GNRs. Three different structures were observed: (1) bundled MWNTs (Figure 2a and b), (2) bundles of diameter-doubled SWNTs (Supporting Information) and (3) a “type-I” GNR whose interesting end is shown (Figure 2d). The image of the type-I GNR appears to indicate that a stack of coalesced SWNTs has collapsed (i.e., each diameter-doubled tube in the stack has become flattened in cross section). Each tube in the stack has a bulb-like closed end (located by the arrows). The bulbs have lateral dimension of ∼1.6 nm, slightly larger than the mean diameter for as-grown material (dh ) 1.4 nm). We assume that the cross section of this bulb is approximately elliptical, and we are observing the minor diameter. In Figure 2c, we plot the number (n) of walls we observe in the HTT-produced MWNTs versus their inner diameter. The data show that the number of walls per MWNT is contained in the range 3 < n Rmax the collapsed shape is energetically favored, and for Rmin < R < Rmax, the collapsed shape is metastable. However, this analysis was done for individual tubes; we believe that the latter case (metastable collapsed shape) can also be induced by external forces, or by van der Waals interaction between tubes in a bundle. This could explain why GNRs (type-II) appear in bundles. This view is also consistent with our observation that isolated MWNTs do not collapse into a GNR under HTT as high as 2200 °C. We also observed in a simultaneous study that HiPCO tubes thermally transform into isolated, rather than bundled, MWNTs. These isolated MWNTs remain stable (without collapsing) up to 2200 °C and type-II GNRs were not formed in this case; these results will be discussed in detail elsewhere.26 In a recent paper, Kim et al.27 also reported that isolated MWNTs are stable against collapse up to HTT ) 2800 °C. However, under high TEM acceleration voltage (800 keV), previous studies by Crespi and co-workers28 found that isolated MWNTs can collapse into a GNR. This collapse may be driven by defects created by direct knock-on collisions of the electrons with the atomic nuclei in the sample.37 In our experiments, the acceleration voltage is much lower (200 keV) than that reported in ref 28 and we took special care to minimize the sample exposure-time to the e-beam. The morphological changes via HTT are interesting as seen in low-magnification TEM; also, images at different positions in the samples were taken routinely in order to verify that we were observing general features of the transformation (Supporting Information). Although a detailed Raman study of the whole thermal transformation will be published elsewere,18 we show in Figure 4 the Raman spectra collected using 514.53-nm excitation for three important stages of thermal transformation process that correspond to HTT ) 200, 1800, and 2200 °C. The Raman spectra each exhibit three structures. (1) the radial breathing modes (“R band”) in the ∼100-300 cm-1 region; (2) the “D band” between 1230 and 1370 cm-1; and (3) a “G band” between 1500 and 1600 cm-1 associated with modes with c-atom vibrations tangential to the tube surface. The substructure in the Raman G-band of the SWNTs is derived from the 1582 cm-1 interlayer modes of graphite and is another consequence of the cylindrical symmetry of a small-diameter carbon nanotube. The frequency of the radial breathing modes in the R band is approximately related to the tube diameter by the relation ωRBM ≈ 223.7/dt + 12 cm-1, where dt is the tube diameter in nanometers and ωRBM is the wavenumber.29 The additive constant (12 cm-1) approximates the effect of tube-tube interactions in a bundle. As can be seen in Figure 4, after HTT ) 1800 °C the nanotubes exhibit a decrease in the intensity of the high-frequency R-band components associated with small-diameter tubes. This decrease is a consequence of the coalescence process; diameter-doubled SWNTs are forming at the expense of small-diameter tubes. Also, a new broad R-band is observed at ∼93 cm-1 that is identified Nano Lett., Vol. 5, No. 11, 2005
Figure 4. Room-temperature Raman spectra of samples heat-treated at HTT ) 200, 1800, and 2200 °C. The spectra were obtained using 514-nm radiation.
with new diameter-doubled tubes (d ≈ 2.4 nm). By HTT ) 2200 °C, the radial breathing mode scattering has completely disappeared, consistent with the disappearance of all SWNT structures, as observed by TEM. For the HTT ) 200 °C sample, the G-band region in Figure 4 was fitted with only three Lorentzians. The composite fit is shown superimposed on the experimental spectrum, and the individual components used in the fit are shown below the G band. After HTT ) 1800 °C, an additional weak Lorentzian component at ∼1582 cm-1 must be added to the SWNT spectral components to fit the data. This new peak signals the evolution of the sample to graphitic or MWNT carbons, in agreement with our HRTEM. After HTT ) 2200 °C, HRTEM indicates that GNRs are the dominant structure in the sample. The G band can now be well fitted by a single Lorentzian located at 1582 cm-1. The fine structure of the G band associated with the SWNTs has disappeared completely, and the G-band maximum has downshifted from 1592 to ∼1582 cm-1. The G band in a well-ordered graphite crystal, or in an MWNT with a large inner diameter (i.e., greater than 3-5 nm), are nearly identical; both are well-described by a single Lorentzian component located at ∼1582 cm-1 with a fwhm of ∼6 cm-1.35 After HTT ) 2200 °C, we also observe a single Lorentzian G-band with a fwhm of ∼28 cm-1, about 5 times broader than that in MWNTs or graphite. Presumably, this band is dominated by a GNR contribution. This 5-times enhancement in the width may be due to disorder in the sp2 network or possibly to the finite lateral width of the GNR. At this moment, our HRTEM studies cannot provide much information about the GNR width or assess the in-plane order of the GNR. Further studies will be needed to pin down which of these two possibilities dominates the G-band line width. It is also possible that there is a less-important contribution to the G band from small graphitic flakes also present in the sample. Nano Lett., Vol. 5, No. 11, 2005
Scattering in the ∼1300-1370 cm-1 range (D band) is present in all known forms of disordered sp2 carbons.35 The D band is associated with symmetry-breaking phenomena such as finite basal plane dimensions in polycrystalline material or particle size effects.30-34 It can also be due to missing C atoms and associated bond disorder. Interestingly, in this study an increase in the D-band scattering is observed with increasing HTT up to 2200 °C. Normally, sp2 carbon systems exhibit the reverse behavior, that is, D-band scattering decreases with the increasing HTT. Our HRTEM images show side views of the ribbon, which indicate that the plane spacing in the GNR is 3.4 ( 0.2 Å (see the Supporting Information),36 close to the value 3.35 Å reported for graphite. Usually, as this small value of the plane spacing is reached, the (hk0) diffraction peaks also sharpen, consistent with improved in-plane order in the network. We therefore tentatively assign the intense D-band scattering from samples exposed to HTT > 2000 °C to the finite width of the GNR. We summarize our observations about the thermal transformation of bundled ARC SWNTs in Figure 5. We suspect that two distinct pathways for tube coalescence can occur, as shown in Figure 5a: chainlike (i) and close-packed (ii) coalescence. These particular pathways are also motivated by the MD simulations of Lopez et al.19,37 For example, they have found that in the close-packed configuration three is the maximum number of SWNTs that can coalesce to produce one triple-diameter SWNT. Their MD simulations also indicate that simultaneous reaction and bond rearrangement involving more than four SWNTs can produce a MWNT. Furthermore, they could also produce a chainlike coalescence process by “artificially” introducing wall defects on the outer SWNTs in small seven-tube bundles (the defects were not on the central SWNT). With this simulation, they did not reproduce the usual experimental value (0.34 nm) for interwall spacing in a MWNT. In fact, their MWNTs show a very open architecture with much larger intershell 2199
Figure 5. (a) Schematic representation of chainlike (i) and closepacked (ii) coalescence processes. (b) Diagram of the thermal evolution of large and small SWNT bundles. Large SWNT bundles evolve forming MWNT bundles that collapse at higher temperatures to form type-II GNRs. Small SWNT bundles can produce isolated MWNT or bundles of diameter-doubled SWNT that can eventually collapse to form a type-I GNR.
spacing than that observed here or in previous work.38,39 However, they were able to simulate the basic thermal transformation of a SWNT bundle to MWNTs, and they argued that this was probably due to chainlike coalescence. In summary, using HRTEM we have observed that purified bundled SWNTs heat-treated in a vacuum to temperatures exceeding ∼1600-1800 °C will evolve into a new form of an all-carbon filament that we call a type-II graphitic nanoribbon (GNR). Type-I GNRs were also observed and appear to involve the flattening of diameter-doubled SWNTs. Type-I GNRs are far less common than type-II GNRs and appear after HTT ≈ 1600 °C, that is, just after the coalescence of single-wall nanotubes. Their production is terminated by the formation of MWNTs. After the formation of the MWNTs at HTT ≈ 1800 °C, type-II GNRs are observed to form. We propose that the type-II GNR is the result of the collapse of bundled MWNTs driven by van der Waals forces between adjacent MWNTs. After ∼2200 °C, all of the MWNT bundles have evolved into type-II GNRs. Acknowledgment. This work was supported by NSFDMR-0103585 and NSF-DMR-0304178. We gratefully acknowledge stimulating discussions with Professor Vin Crespi. Supporting Information Available: Low-magnification TEM pictures were taken at different positions of the samples in order to verify their homogeneity. Typical images of the samples after heat treatment at 200, 1800, 2000, and 2200 °C are shown. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Iijima, S. Nature 1991, 354, 56. (2) Iijima, S.; Ichihashi, T. Nature 1993, 363, 603. 2200
(3) Ebbesen, T. W.; Ajayan, P. M. Nature 1992, 358, 220. (4) Dresselhaus, M. S.; Dresselhaus, G.; Jorio, A. Annu. ReV. Mater. Res. 2004, 34, 247. (5) Dresselhaus, M. S.; Dresselhaus, G.; Charlier, J. C.; Herna´ndez, E. Philos. Trans. R. Soc. London, Ser. A 2004, 362, 2065. (6) Endo, M.; Hayashi, T.; Kim, Y. A.; Terrones, M.; Dresselhaus, M. S. Philos. Trans. R. Soc. London, Ser. A 2004, 362, 2223. (7) Rao, C. N. R.; Satishkumar, B. C.; Govindaraj, A.; Nath, M. ChemPhysChem. 2001, 2, 78. (8) Mun˜oz-Navia, M.; Dorantes-Da´vila, J.; Terrones, M.; Hayashi, T.; Kim, Y. A.; Endo, M.; Dresselhaus, M. S.; Terrones, H. Chem. Phys. Lett. 2005, 407, 327. (9) Kim, Y. A.; Muramatsu, H.; Hayashi, T.; Endo, M.; Terrones, M.; Dresselhaus, M. S. Chem. Phys. Lett. 2004, 398, 87. (10) Endo, M.; Hayashi, T.; Muramatsu, H.; Kim, Y.-A.; Terrones, H.; Terrones, M.; Dresselhaus, M. S. Nano Lett. 2004, 4, 1451. (11) Nikolaev, P.; Thess, A.; Rinzler, A. G.; Colbert, D. T.; Smalley, R. E. Chem. Phys. Lett. 1997, 266, 422. (12) Bronikowski, M. J.; Willis, P. A.; Colbert, D. T.; Smith, K. A.; Smalley, R. E. J. Vac. Sci. Technol., A 2001, 19, 1800. (13) Li, H. J.; Guan, L. H.; Shi, Z. J.; Gu, Z. N. J. Phys. Chem. B 2004, 108, 4573. (14) Yudasaka, M.; Sensui, N.; Takizawa, M.; Bandow, S.; Ichihashi, T.; Iijima, S. Chem. Phys. Lett. 1999, 312, 155. (15) Metenier, K.; Bonnamy, S.; Beguin, F.; Journet, C.; Bernier, P.; de La Chapelle, M. L.; Chauvet, O.; Lefrant, S. Carbon 2002, 40, 1765. (16) Yudasaka, M.; Ichihashi, T.; Kasuya, D.; Kataura, H.; Iijima, S. Carbon 2003, 41, 1273. (17) Furtado, C. A.; Kim, U. J.; Gutierrez, H. R.; Pan, L.; Dickey, E. C.; Eklund, P. C. J. Am. Chem. Soc. 2004, 126, 6095. (18) Kim, U. J.; Gutie´rrez, H. R.; Kim, J. P.; Eklund, P. C. Submitted for publication. (19) Lopez, M. J.; Rubio, A.; Alonso, J. A. IEEE Trans. Nanotechnol. 2004, 3, 230. (20) Gao, G. H.; Cagin, T.; Goddard, W. A. Nanotechnology 1998, 9, 184. (21) Benedict, L. X.; Chopra, N. G.; Cohen, M. L.; Zettl, A.; Louie, S. G.; Crespi, V. H. Chem. Phys. Lett. 1998, 286, 490. (22) Koshio, A.; Yudasaka, M.; Iijima, S. Chem. Phys. Lett. 2002, 356, 595. (23) (a) Harigaya, K.; Enoki, T. Chem. Phys. Lett. 2002, 351, 128. (b) Harigaya, K.; Kobayashi, Y.; Kawatsu, N.; Takai, K.; Sato, H.; Ravier, J.; Enoki, T.; Endo, M. Physica E 2004, 22, 708. (24) Nakada, K.; Fujita, M.; Dresselhaus, G.; Dresselhaus, M. S. Phys. ReV. B 1996, 54, 17954. (25) Tang, T.; Jagota, A.; Hui, C. Y.; Glassmaker, N. J. J. Appl. Phys. 2005, 97, 074310. (26) Kim, U. J.; Gutie´rrez, H. R.; Kim, J. P.; Eklund, P. C. J. Phys. Chem. B, in press. (27) Kim, Y. A.; Hayashi, T.; Osawa, K.; Dresselhaus, M. S.; Endo, M. Chem. Phys. Lett. 2003, 380, 319. (28) Crespi, V. H.; Chopra, N. G.; Cohen, M. L.; Zettl, A.; Louie, S. G. Phys. ReV. B 1996, 54, 5927. (29) 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. (30) Kim, U. J.; Furtado, C. A.; Liu, X.-M.; Chen, G.; Eklund, P. C. J. Am. Chem. Soc., submitted for publication. (31) Matthews, M. J.; Pimenta, M. A.; Dresselhaus, G.; Dresselhaus, M. S.; Endo, M. Phys. ReV. B 1999, 59, R6585. (32) Tunistra, F.; Koenig, J. L. J. Chem. Phys. 1970, 53, 1126. (33) Knight, D. S.; White, W. B. J. Mater. Res. 1989, 4, 385. (34) McGuirea, K.; Gothard, N.; Gai, P. L.; Dresselhaus, M. S.; Sumanasekera, G.; Rao, A. M. Carbon 2005, 43, 219. (35) Dresselhaus, M. S.; Dresselhaus, G.; Pimenta, M. A.; Eklund, P. C. In Analytical Applications of Raman Spectroscopy; Pelletier, M. J., Ed.; Blackwell Science: Cambridge, MA, 1999; Chapter 9. (36) The spacing between graphene layers was obtained from the Fourier transform of high-resolution images of the GNRs. The uncertainty was estimated from the width of the FT ring as shown in the Supporting Information. This uncertainty should not be confused with the point-to-point resolution (Rayleigh criterion) of the instrument (1.9 Å) that represents the upper limit to which intuitive image
Nano Lett., Vol. 5, No. 11, 2005
interpretation can be made. For more information about resolution definitions and information limits on HRTEM read the following: (a) Smith, D. J. Rep. Prog. Phys. 1997, 60, 1513. (b) Van Dyck, D.; Van Aert, S.; den Dekker, A. J. Microsc. Microanal. 2004, 10, 153; and references therein.
Nano Lett., Vol. 5, No. 11, 2005
(37) Banhart, F. Rep. Prog. Phys. 1999, 62, 1181. (38) Liu, M.; Cowley, J. M. Carbon 1994, 32, 393. (39) Cowley, J. M. Microsc. Res. Technol. 1999, 46, 75.
NL051276D
2201
