NANO LETTERS
Solution Phase Production of Graphene with Controlled Thickness via Density Differentiation
2009 Vol. 9, No. 12 4031-4036
Alexander A. Green and Mark C. Hersam* Department of Materials Science and Engineering and Department of Chemistry, Northwestern UniVersity, EVanston, Illinois 60208-3108 Received July 9, 2009; Revised Manuscript Received August 31, 2009
ABSTRACT Graphene flakes with controlled thicknesses are isolated in solution using density gradient ultracentrifugation. These stable graphene dispersions are produced using the bile salt sodium cholate, which promotes graphite exfoliation and results in graphene-surfactant complexes having buoyant densities that vary with graphene thickness. The sorted graphene flakes are characterized using atomic force microscopy and Raman spectroscopy. Graphene dispersions produced using density differentiation offer superior performance in transparent conductors than those produced using conventional sedimentation-based centrifugation techniques.
Graphene, a two-dimensional lattice of carbon only a single atom thick, has generated considerable attention as a result of its outstanding electronic,1 mechanical,2 and chemical properties.3 While the earliest graphene samples were produced using micromechanical cleavage,4 recent efforts have focused on developing larger scale methods of producing graphene. Such methods can be divided into two main approaches: those that employ substrates such as SiC5 or nickel6,7 for direct growth of graphene and those that use pristine or functionalized graphite flakes for direct exfoliation of graphene into various solvents.8-11 The latter solution phase techniques offer several significant advantages since they utilize inexpensive and readily available graphite flakes, do not require methods of transferring the graphene from the growth substrate, and can employ existing technologies for scale up to large volume processing. For example, functionalized graphite oxide has been used to produce graphene oxide flakes several micrometers in size in aqueous solution; however, these materials must be chemically reduced using a variety of chemical and thermal treatments11-13 to regain their electrical conductivity. Despite these treatments, the electronic properties of reduced graphene oxide remain different from those of pristine graphene.11 Small sheets of unfunctionalized graphene can also be exfoliated into solvents such as dimethylformamide,9 Nmethyl-pyrrolidone,14 and water-surfactant solutions.15 Since these solution phase techniques offer limited control over the number of layers in the dispersed graphene sheets, the resulting dispersions suffer from polydispersity in their properties. Bilayer graphene, for instance, has been shown * Corresponding author. E-mail:
[email protected]. 10.1021/nl902200b CCC: $40.75 Published on Web 09/25/2009
2009 American Chemical Society
to have a tunable bandgap in the infrared16,17 and exhibit unique quantum mechanical behavior18 compared to zero bandgap single layer graphene, whereas trilayer graphene is a semimetal whose band overlap can be controlled by an applied electric field.19,20 Consequently, the production of graphene with monodisperse thickness in solution is expected to enable the fabrication of high density graphene arrays with uniform properties via dielectrophoresis,21 LangmuirBlodgett,10 or related liquid phase deposition techniques. In this letter, we employ density gradient ultracentrifugation (DGU) to isolate graphene sheets with controlled thickness. This aqueous solution phase approach is enabled by the planar amphiphilic surfactant sodium cholate (SC),22 which forms a stable encapsulation layer on each side of the suspended graphene sheets. Atomic force microscopy (AFM) and Raman spectroscopy verify the ability of DGU to sort graphene flakes by thickness. Furthermore, a quantitative relationship is developed between the buoyant density and thickness of graphene in aqueous surfactant solutions. Finally, we incorporate these graphene dispersions into transparent conductive films whose performance match those of graphene oxide films without requiring high temperature pyrolytic chemical reduction. Graphene dispersions were prepared by horn ultrasonication of naturally occurring graphite flakes (Asbury Graphite Mills) in aqueous solution containing 2% w/v SC (full experimental details are presented in the Supporting Information). The highly hydrophobic graphene sheets exfoliated in this process were stabilized by the amphiphilic SC molecules, whose hydrophobic faces associated with the graphene leaving the opposing hydrophilic faces to interact with the
Figure 1. (A) Schematic illustration of the graphene exfoliation process. Graphite flakes are combined with sodium cholate (SC) in aqueous solution. Horn ultrasonication exfoliates few-layer graphene flakes that are encapsulated by SC micelles. (B) Photograph of a 90 µg mL-1 graphene dispersion in SC six weeks after it was prepared. (C) Schematic illustrating an ordered SC monolayer on graphene.
surrounding aqueous environment (Figure 1A). Horn ultrasonication not only produced thin graphene sheets but also thicker graphite flakes resulting in an unstable, gray-black graphene/graphite slurry. Similar to previous work, our initial experiments relied on simple sedimentation centrifugation. To remove the fast sedimenting thick graphite material from the dispersion, sedimentation centrifugation was performed at 15 krpm for 60 min, leaving a black supernatant consisting of predominantly few-layer graphene flakes. Such sedimented graphene dispersions were stable for several weeks at loadings in excess of 90 µg/mL (Figure 1B). The success of SC in dispersing graphene enables solution phase processing of graphene using DGU.23 In DGU, solution phase dispersions are centrifuged in a density gradient established by layering a centrifuge tube with solutions containing varying amounts of a dense solute. At high centripetal acceleration, the material suspended in the dispersion is driven by differences in its buoyant density and that of the medium to different points in the gradient. Once the material reaches its isopycnic point, where its buoyant density matches that of the medium, it ceases to be driven by density differences, and the material can then be collected as fractions taken layer by layer from the centrifuge tube. For DGUbased sorting to occur, the buoyant density of a material must vary as a function of its physical or electronic structure. For carbon nanotubes, this condition has been established by encapsulating the nanotubes in molecules such as DNA,24 SC,25 sodium deoxycholate,26 and sodium dodecyl sulfate.27,28 SC, in particular, has been shown to be particularly effective in establishing differences in buoyant density for carbon nanotubes based on wall number,29 diameter,25 and chiral handedness.30 This sensitive molecular recognition is indicative of the ability of SC to encapsulate carbon nanotubes uniformly and reproducibly. Given the planar structure of 4032
Figure 2. (A) Photograph of a centrifuge tube following the first iteration of density gradient ultracentrifugation (DGU). The concentrated graphene was diluted by a factor of 40 to ensure that all graphene bands could be clearly resolved in the photograph. Lines mark the positions of the sorted graphene fractions within the centrifuge tube. (B,C) Representative AFM images of graphene deposited using fractions f4 (B) and f16 (C) onto SiO2. (D) Height profile of regions marked in panels B (blue curve) and C (red curve) demonstrating the different thicknesses of graphene flakes obtained from different DGU fractions.
SC, graphene will also be encapsulated by well-ordered assemblies of SC in aqueous solution (Figure 1C). Similar to carbon nanotubes, these graphene-SC complexes possess buoyant densities that vary as a function of the thickness of the graphene sheet in the complex, thus enabling DGU-based sorting. Before performing DGU, freshly sonicated graphene dispersions were centrifuged in a step density gradient. In these step gradients, an underlayer of 1.32 g mL-1 density was added to the centrifuge tube followed on top by the 1 g mL-1 graphene dispersion. Subsequent ultracentrifugation caused the graphene sheets to sediment rapidly to the point where the density of the medium changed discontinuously. In this region, the few-layer graphene flakes with low buoyant densities halted their sedimentation as they reached their isopycnic points while the denser thick graphite flakes continued their motion until they eventually formed a pellet at the bottom of the centrifuge tube. This step gradient approach thereby eliminated a large portion of the thick components in the dispersion without removing valuable fewlayer material that could also be eliminated in a centrifugation experiment where a dense underlayer was not present. Moreover, since all the buoyant few-layer graphene sheets collected near the step in the gradient, the material collected from this region was a highly concentrated graphene solution. The concentrated graphene solution was collected and then injected at the bottom of a density gradient and centrifuged for 24 h at a maximum centripetal acceleration of 141 000 g. During this process, the graphene sheets moved upward in the centrifuge tube to their isopycnic points. Figure 2A is a photograph of the centrifuge tube following the separation Nano Lett., Vol. 9, No. 12, 2009
that clearly displays multiple gray bands at different locations inside the gradient. These bands were recovered in 1 mm steps using a piston gradient fractionator. Of the resulting 32 fractions, five were selected for a second iteration of DGU: fractions f4, f10, and f22, which corresponded to the thinnest bands observed; and fractions f16 and f28, which were part of broader bands. To achieve further density gradient refinement, the graphene solution was loaded at the top of a second density gradient forcing the graphene flakes to move downward in the centrifuge tube to their respective isopycnic points. This second DGU iteration removed slow moving, low buoyant density materials that did not reach their isopycnic points in the first iteration. Atomic force microscope (AFM) images of the graphene flakes deposited onto SiO2 reveal irregularly shaped sheets with dimensions ranging from 50 to several hundred nanometers. Representative images and line profiles of material from fractions f4 and f16 are shown in Figure 2B-D. The flakes from fraction f4 exhibit an average thickness of 1.1 nm. Single-layer graphene on SiO2 typically has an apparent thickness of ∼1 nm as a result of adsorbed water,31 and this thickness could be further increased by residual sodium cholate molecules on the graphene surface. In contrast, the graphene sheets from f16 were found to have an average thickness of 1.5 nm. Figure 3A,B present the thickness histograms of several sorted graphene fractions as well as the concentrated and sedimented graphene dispersions. These histograms were calculated from at least 100 individual flakes using multiple 2 µm × 2 µm AFM images with the average thickness measured over the area of each flake. Comparison of the thickness distributions of the sedimented, concentrated, and f4 graphene solutions in Figure 3A shows progressive sharpening of the distributions with increasing buoyant density refinement. 37% of the sedimented graphene solution was found to have thicknesses greater than 2 nm while these flakes constituted only 2.6% of the dispersion following concentration step gradient processing. Following DGU, 80% of the graphene flakes from fraction f4 were found to have thicknesses of 1.2 nm or less, most likely corresponding to single-layer graphene. In contrast, only 24% of the concentrated graphene consisted of single-layer material. The thickness distributions of the graphene sorted using DGU show a monotonic increase in the average flake thickness with increasing buoyant density (Figure 3B, Supporting Information for distributions of f10 and f22). To gain further insight into the ordering of SC on the graphene surface, we developed a geometrical model of the buoyant density of the graphene-SC complex similar to that used for carbon nanotubes32 (Figure 3C). In this model, the thickness of the graphene sheet is defined by N, which specifies the average number of graphene layers inside the sheet separated by the graphene interlayer distance tgr ) 0.34 nm. On both sides of the graphene sheet is an anhydrous layer of thickness tA containing the SC encapsulation layer. The SC molecules in this region coat the graphene surface with surface packing density σ. Surrounding this SC layer is an electrostatically bound hydration shell of thickness tH. This hydration layer has the lowest density of any of the Nano Lett., Vol. 9, No. 12, 2009
Figure 3. (A) Mean flake thickness histograms for sedimented (gray), concentrated (purple), and DGU fraction f4 (blue) graphene solutions calculated using AFM. (B) Mean flake thickness histograms plotted by relative frequency (mode thickness scaled to unity) for DGU fractions f4 (blue), f16 (red), and f28 (green). (C) Buoyant density model for SC-encapsulated graphene in which a flake of thickness N is coated by a surfactant with packing density σ and effective thickness tA and a hydration layer of thickness tH. (D) Fit of the geometrical density model to the experimental data. The uncertainty in graphene flake thickness was taken as the fwhm of the flake thickness distribution.
components in the complex and hence serves to decrease the buoyant density of the graphene-SC assembly. The resulting buoyant density F(N) is then F(N) )
FSN + 2mSCσ + 2FH2OtH (N + 1)tgr + 2tA + 2tH
where FS ) 7.66 × 10-8 g cm-2 is the sheet density of graphene, mSC ) 7.15 × 10-22 g is the mass of one SC molecule, and FH2O is the density of water. Following this model, the buoyant density of the graphene flakes is 4033
independent of lateral area, which suggests that the same density gradient can be used to isolate graphene with a large range of areas. The graphene buoyant density model was applied to the experimental data by assuming the anhydrous shell thickness tA was 0.355 nm, which corresponds to the value measured for SC-encapsulated carbon nanotubes.32 Furthermore, the apparent thickness for single-layer graphene was taken to be the average thickness of 1.1 nm measured for graphene sheets from fraction f4. N for subsequent fractions was then calculated from their average mean thicknesses and the graphite interlayer spacing. With these conditions in place, the model yields a SC surface packing density σ of 1.35 nm-2 and a hydration layer thickness tH of 3.3 nm. Since SC occupies approximately a ∼0.7 nm2 area32 on the graphene surface, this surface density corresponds to ∼94% surface coverage of SC (schematically illustrated in Figure 1C), and is similar to the 72 ( 16% coverage calculated for carbon nanotubes. The small area of graphene occupied by each SC molecule also implies that it is possible for the SC encapsulation layer to accommodate variations in graphene thickness on length scales of several nanometers. Consequently, a continuum of mean graphene flake thicknesses can be separated using DGU depending on how the flakes are exfoliated. The hydration layer thickness of the grapheneSC complex is approximately 80% larger than that measured for carbon nanotubes; however, both parameters result in similar total hydration layer volumes once typical graphene areas and nanotube lengths are taken into account. Graphene flakes deposited onto SiO2 were also characterized by Raman spectroscopy using an excitation wavelength of 514 nm. Spectra were obtained using a beam size of 1-2 µm on samples with a high surface coverage of graphene enabling multiple flakes to be probed in a single measurement. Typical Raman spectra from the sorted graphene samples display four main peaks: the G band at ∼1590 cm-1, the 2D (or G’) band at ∼2700 cm-1, and the disorder-related D and D′ peaks at ∼1350 and ∼1620 cm-1, respectively (Figure 4A-C). The Raman spectra show systematic changes in the G and 2D peaks as a function of the thickness distribution of the graphene flakes. To gather sufficient statistics for these variations, we collected spectra from at least 30 different locations and extracted mean and standard deviation information from these data. Figure 4D presents the I(2D)/I(G) ratio and full-width at half-maximum (fwhm) of the 2D band as a function of the average thickness of the sorted and concentrated graphene dispersions. As the thickness of the graphene flakes increases, I(2D)/ I(G) decreases monotonically from a high of 2.1 ( 0.2 for single-layer graphene to 0.8 ( 0.1 for quadruple-layer graphene. fwhm(2D), on the other hand, increases with graphene thickness nearly doubling between single- and quadruple-layer samples. Similar trends in both I(2D)/I(G) and fwhm(2D) as a function of graphene layer number have previously been observed for CVD grown graphene samples7 and should prove useful for more rapid screening of DGUsorted graphene. Similar to CVD-graphene samples and those grown on SiC,33 we find that the 2D band of our 4034
Figure 4. (A) Representative Raman spectra of sorted graphene flakes from fractions f4 (blue), f10 (orange), f16 (red), f22 (purple), and f28 (green) on SiO2 with the G band intensity normalized to unity. (B) Magnified Raman spectra of sorted graphene flakes in the D and G band region (fractions f10 and f22 omitted for clarity). (C) Magnified Raman spectra of sorted graphene flakes in the 2D band region. The 2D peak decreases in intensity compared to the G peak and broadens with increasing mean flake thickness. (D) The ratio of 2D and G band intensity I(2D)/I(G) and fwhm(2D) as a function of the mean graphene thickness. Triangles and squares represent I(2D)/I(G) and fwhm(2D), respectively. Gray open symbols mark values obtained from the concentrated graphene dispersion.
multilayer graphene is best described by a single Lorentzian line shape, as opposed to the four component line shape observed for samples produced by micromechanical cleavage.34 Since these four components arise from the close interaction between ABAB stacked graphene layers, their absence implies weak interlayer coupling and hence nonABAB stacking.33,35 This source of disorder could be caused during horn ultrasonication or from rebundling of previously exfoliated graphene sheets. The disorder-related D peak is relatively intense compared to the G band in these samples, Nano Lett., Vol. 9, No. 12, 2009
air at 250 °C for two hours decreased their sheet resistance (by factors of 2-4) while increasing their transmittance by ∼1%. This thermal treatment likely removed some of the iodixanol and sodium cholate remaining in the film and enabled the graphene sheets to reorder to improve flake-flake contacts. AFM images of the annealed films indicate the graphene flakes form a disordered network (Figure 5B). Folded flakes can be discerned in the images along with rough areas that are most likely caused by residual surfactants, iodixanol, or the filter membrane. The films possess high optical transmittance from ∼300 to 3300 nm, revealing a wide transmittance window that is well suited for infrared applications (Figure 5C).
Figure 5. (A) Photograph of two sorted graphene transparent conductive films on glass. (B) AFM image of an annealed transparent conductive film produced from fraction f4 displaying the disordered graphene film morphology. (C) Optical transmittance of a set of films produced from fraction f4. Inset: transmittance of films in the ultraviolet-visible range. (D,E) Transmittance of graphene transparent conductors produced from sedimented (gray), concentrated (purple), f4 (blue), f16 (red), and f28 (green) solutions as a function of their sheet resistance at wavelengths of 550 (D) and 1000 nm (E), respectively. Lines are drawn between points to aid the eye.
most likely as a result of defects within the graphene sheets or from the small size of the flakes, which should increase the number of disordered graphene edges probed in the measurement.36 The ratio of the intensities of the D and G peaks I(D)/I(G) remains fixed at ∼0.93 for both the sedimented and density-refined material, which indicates that ultracentrifugation did not contribute additional defects to the graphene. This I(D)/I(G) value is comparable to that observed in highly reduced graphene oxide13 and is less than that measured for lithographically patterned graphene nanoribbons.37 We assessed the electrical properties of the sorted graphene material by using it to form transparent conductive films as shown in Figure 5A. The graphene films were produced by vacuum filtration onto mixed cellulose ester membranes and transferred to glass by dissolving away the ester membrane in acetone.38 We found that annealing the graphene films in Nano Lett., Vol. 9, No. 12, 2009
Four-point probe measurements of the film sheet resistance indicated that DGU processing yields significant improvements in the transparent conductive properties of the graphene films (Figure 5D,E). First, all films produced using buoyant density sorting demonstrated improved conductivity compared to films produced from sedimented graphene solutions. For the films produced from the concentrated, f16, and f28 dispersions, this improvement is ∼45%. Second, the films produced using predominantly single layer graphene flakes offer the best transparent conductor performance. These highly refined materials exhibit sheet resistances that are approximately half that of the other density processed material. Analysis of the graphene flakes indicates that the f4 material has the largest mean area (16,000 nm2) of all the materials used for transparent conductors (see Supporting Information). A large lateral area should result in fewer graphene-graphene contacts required for charge transport across the film and thus implies increased film conductivity. However, the differences in transparent conductor performance between the concentrated and sedimented solutions suggest that the graphene thickness distribution can also play an important role in the connectivity of the graphene network. The larger proportion of thick graphene sheets in the sedimented solution may disrupt the ideal close packed structure of the network, which would reduce the overlap between neighboring sheets. In contrast, the highly flexible single-layer graphene sheets deposited from solution are expected to coat underlying layers with greater conformity, resulting in improved graphene-graphene contacts. While the performance of these graphene transparent conductors is inferior to that of indium tin oxide or optimized carbon nanotube films,38 they compete favorably with transparent conductors produced from other solution phase dispersions of graphene.9,10,12,15,39 Importantly for many applications, DGU processed transparent conductive films require relatively benign thermal processing. In contrast, for films produced using graphene oxide, high temperature pyrolysis at 1100 °C is required for transparent conductors of comparable performance,12,39 a processing step that is not compatible with most transparent substrates. Lastly, these measurements suggest that the strong disorder-related Raman spectroscopy peaks observed in DGU graphene dispersions do not have an adverse effect on their application in transparent conductive films. 4035
In conclusion, we have shown that SC can form stable graphene dispersions with graphene concentrations in excess of 90 µg mL-1. In quantitative agreement with a geometrical model, DGU separations of the SC-encapsulated graphene yield graphene sheets with mean thicknesses that increase as a function of their buoyant density. AFM and Raman spectroscopy confirm the improved monodispersity of the graphene samples following DGU processing, while electrical and optical characterization demonstrate enhanced performance in transparent conductive films. By demonstrating the utility of DGU for processing graphene, this work suggests that additional two-dimensional nanomaterials can also be sorted using density differentiation. Acknowledgment. This work was supported by the National Science Foundation (DMR-0520513, EEC-0647560, and DMR-0706067), the Office of Naval Research (N0001409-1-0180 and N00014-09-1-0795), and the Nanoelectronics Research Initiative. A Natural Sciences and Engineering Research Council of Canada Postgraduate Scholarship (A.A.G.) is also acknowledged. This research utilized instruments in the Keck-II facility of the NUANCE Center at Northwestern University, the Keck Biophysics Facility at Northwestern University, and the Center for Nanoscale Materials at Argonne National Laboratory. The Center for Nanoscale Materials is supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC02-06CH11357. We thank Prof. J. Widom and NanoIntegris for use of their ultracentrifuge rotors as well as Dr. D. J. Gosztola for assistance with Raman spectroscopy. Supporting Information Available: Graphene solution phase processing, AFM sample preparation, additional graphene and transparent conductor data, and absorbance and Raman spectroscopy measurement information. This material is available free of charge via the Internet at http:// pubs.acs.org. References (1) Bolotin, K. I.; Sikes, K. J.; Jiang, Z.; Klima, M.; Fudenberg, G.; Hone, J.; Kim, P.; Stormer, H. L. Solid State Commun. 2008, 146 (9-10), 351–355. (2) Lee, C.; Wei, X. D.; Kysar, J. W.; Hone, J. Science 2008, 321 (5887), 385–388. (3) Geim, A. K. Science 2009, 324 (5934), 1530–1534. (4) Novoselov, K. S.; Jiang, D.; Schedin, F.; Booth, T. J.; Khotkevich, V. V.; Morozov, S. V.; Geim, A. K. Proc. Natl. Acad. Sci. U.S.A. 2005, 102 (30), 10451–10453. (5) Berger, C.; Song, Z. M.; Li, X. B.; Wu, X. S.; Brown, N.; Naud, C.; Mayou, D.; Li, T. B.; Hass, J.; Marchenkov, A. N.; Conrad, E. H.; First, P. N.; de Heer, W. A. Science 2006, 312 (5777), 1191–1196. (6) Kim, K. S.; Zhao, Y.; Jang, H.; Lee, S. Y.; Kim, J. M.; Kim, K. S.; Ahn, J. H.; Kim, P.; Choi, J. Y.; Hong, B. H. Nature 2009, 457 (7230), 706–710. (7) Reina, A.; Jia, X. T.; Ho, J.; Nezich, D.; Son, H. B.; Bulovic, V.; Dresselhaus, M. S.; Kong, J. Nano Lett. 2009, 9 (1), 30–35. (8) Stankovich, S.; Dikin, D. A.; Dommett, G. H. B.; Kohlhaas, K. M.; Zimney, E. J.; Stach, E. A.; Piner, R. D.; Nguyen, S. T.; Ruoff, R. S. Nature 2006, 442 (7100), 282–286.
4036
(9) Blake, P.; Brimicombe, P. D.; Nair, R. R.; Booth, T. J.; Jiang, D.; Schedin, F.; Ponomarenko, L. A.; Morozov, S. V.; Gleeson, H. F.; Hill, E. W.; Geim, A. K.; Novoselov, K. S. Nano Lett. 2008, 8 (6), 1704–1708. (10) Li, X. L.; Zhang, G. Y.; Bai, X. D.; Sun, X. M.; Wang, X. R.; Wang, E.; Dai, H. J. Nat. Nanotechnol. 2008, 3 (9), 538–542. (11) Tung, V. C.; Allen, M. J.; Yang, Y.; Kaner, R. B. Nat. Nanotechnol. 2009, 4 (1), 25–29. (12) Becerril, H. A.; Mao, J.; Liu, Z.; Stoltenberg, R. M.; Bao, Z.; Chen, Y. ACS Nano 2008, 2 (3), 463–470. (13) Gao, W.; Alemany, L. B.; Ci, L.; Ajayan, P. M. Nature Chem. 2009, 1 (5), 403–408. (14) Hernandez, Y.; Nicolosi, V.; Lotya, M.; Blighe, F. M.; Sun, Z. Y.; De, S.; McGovern, I. T.; Holland, B.; Byrne, M.; Gun’ko, Y. K.; Boland, J. J.; Niraj, P.; Duesberg, G.; Krishnamurthy, S.; Goodhue, R.; Hutchison, J.; Scardaci, V.; Ferrari, A. C.; Coleman, J. N. Nat. Nanotechnol. 2008, 3 (9), 563–568. (15) Lotya, M.; Hernandez, Y.; King, P. J.; Smith, R. J.; Nicolosi, V.; Karlsson, L. S.; Blighe, F. M.; De, S.; Wang, Z. M.; McGovern, I. T.; Duesberg, G. S.; Coleman, J. N. J. Am. Chem. Soc. 2009, 131 (10), 3611–3620. (16) Ohta, T.; Bostwick, A.; Seyller, T.; Horn, K.; Rotenberg, E. Science 2006, 313 (5789), 951–954. (17) Zhang, Y. B.; Tang, T. T.; Girit, C.; Hao, Z.; Martin, M. C.; Zettl, A.; Crommie, M. F.; Shen, Y. R.; Wang, F. Nature 2009, 459 (7248), 820–823. (18) Novoselov, K. S.; McCann, E.; Morozov, S. V.; Fal’ko, V. I.; Katsnelson, M. I.; Zeitler, U.; Jiang, D.; Schedin, F.; Geim, A. K. Nat. Phys. 2006, 2 (3), 177–180. (19) Craciun, M. F.; Russo, S.; Yamamoto, M.; Oostinga, J. B.; Morpurgo, A. F.; Tarucha, S. Nat. Nanotechnol. 2009, 4 (6), 383–388. (20) Koshino, M.; McCann, E. Phys. ReV. B 2009, 79 (12), 125443–5. (21) Vijayaraghavan, A.; Sciascia, C.; Dehm, S.; Lombardo, A.; Bonetti, A.; Ferrari, A. C.; Krupke, R. ACS Nano 2009, 3 (7), 1729–1734. (22) Mukhopadhyay, S.; Maitra, U. Curr. Sci. 2004, 87 (12), 1666–1683. (23) Green, A. A.; Hersam, M. C. Mater. Today 2007, 10 (12), 59–60. (24) Arnold, M. S.; Stupp, S. I.; Hersam, M. C. Nano Lett. 2005, 5 (4), 713–718. (25) Arnold, M. S.; Green, A. A.; Hulvat, J. F.; Stupp, S. I.; Hersam, M. C. Nat. Nanotechnol. 2006, 1 (1), 60–65. (26) Yanagi, K.; Miyata, Y.; Kataura, H. Appl. Phys. Expr. 2008, 1, (3). (27) Niyogi, S.; Densmore, C. G.; Doorn, S. K. J. Am. Chem. Soc. 2009, 131 (3), 1144–1153. (28) Green, A. A.; Hersam, M. C. Nano Lett. 2008, 8 (5), 1417–1422. (29) Green, A. A.; Hersam, M. C. Nat. Nanotechnol. 2009, 4 (1), 64–70. (30) Green, A. A.; Duch, M. C.; Hersam, M. C. Nano Res. 2009, 2 (1), 69–77. (31) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang, Y.; Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Science 2004, 306 (5696), 666–669. (32) Arnold, M. S.; Suntivich, J.; Stupp, S. I.; Hersam, M. C. ACS Nano 2008, 2 (11), 2291–2300. (33) Faugeras, C.; Nerriere, A.; Potemski, M.; Mahmood, A.; Dujardin, E.; Berger, C.; de Heer, W. A. Appl. Phys. Lett. 2008, 92 (1), 0119143. (34) Ferrari, A. C.; Meyer, J. C.; Scardaci, V.; Casiraghi, C.; Lazzeri, M.; Mauri, F.; Piscanec, S.; Jiang, D.; Novoselov, K. S.; Roth, S.; Geim, A. K. Phys. ReV. Lett. 2006, 97, (18), 187401-4. (35) Hass, J.; Varchon, F.; Millan-Otoya, J. E.; Sprinkle, M.; Sharma, N.; de Heer, W. A.; Berger, C.; First, P. N.; Magaud, L.; Conrad, E. H. Phys. ReV. Lett. 2008, 100 (12), 125504–4. (36) Casiraghi, C.; Hartschuh, A.; Qian, H.; Piscanec, S.; Georgi, C.; Fasoli, A.; Novoselov, K. S.; Basko, D. M.; Ferrari, A. C. Nano Lett. 2009, 9 (4), 1433–1441. (37) Jiao, L. Y.; Zhang, L.; Wang, X. R.; Diankov, G.; Dai, H. J. Nature 2009, 458 (7240), 877–880. (38) Wu, Z. C.; Chen, Z. H.; Du, X.; Logan, J. M.; Sippel, J.; Nikolou, M.; Kamaras, K.; Reynolds, J. R.; Tanner, D. B.; Hebard, A. F.; Rinzler, A. G. Science 2004, 305 (5688), 1273–1276. (39) Wang, X.; Zhi, L. J.; Mullen, K. Nano Lett. 2008, 8 (1), 323–327.
NL902200B
Nano Lett., Vol. 9, No. 12, 2009