NANO LETTERS
Simple and Scalable Gel-Based Separation of Metallic and Semiconducting Carbon Nanotubes
2009 Vol. 9, No. 4 1497-1500
Takeshi Tanaka,*,† Hehua Jin,† Yasumitsu Miyata,† Shunjiro Fujii,† Hiroshi Suga,† Yasuhisa Naitoh,† Takeo Minari,† Tetsuhiko Miyadera,† Kazuhito Tsukagoshi,† and Hiromichi Kataura†,‡ Nanotechnology Research Institute, National Institute of AdVanced Industrial Science and Technology (AIST), Ibaraki 305-8562, Japan, and JST, CREST, Kawaguchi 330-0012, Japan Received November 18, 2008; Revised Manuscript Received January 20, 2009
ABSTRACT We report a rapid and scalable method for the separation of metallic and semiconducting single-wall carbon nanotubes (SWCNTs); the separation is performed by the selective adsorption of semiconducting SWCNTs on agarose gel. The most effective separation was realized by a simple procedure in which a piece of gel containing SWCNTs and sodium dodecyl sulfate was frozen, thawed, and squeezed. This process affords a solution containing 70% pure metallic SWCNTs and leaves a gel containing 95% pure semiconducting SWCNTs. Field-effect transistors constructed from the separated semiconducting SWCNTs have been demonstrated to function without any electrical breakdown.
Single-wall carbon nanotubes (SWCNTs) have a wide range of promising applications, particularly in electronic devices such as field-effect transistors (FETs),1,2 and as such have received considerable attention in recent years.3 A key obstacle preventing the development of SWCNT-based electronics is the coexistence of metallic and semiconducting nanotubes in as-synthesized materials. Current SWCNTs syntheses afford a mixture of metallic and semiconducting nanotubes, and the two types must be separated in order to be utilized. The separation of metallic and semiconducting SWCNTs (MS separation) has been attempted by many different approaches, including dielectrophoresis,4-6 selective oxidation,7 amine extraction,8,9 aromatics extraction,10,11 polymer wrapping,12-15 and density-gradient ultracentrifugation.16,17 However, all of the methods proposed to date remain unsuitable for industrial applications,18 achieving only low yields, purity, or throughput, or involving substantial cost. Recently, the present authors discovered that MS separation occurs during electrophoresis of an SWCNTcontaining agarose gel.19 Agarose gel electrophoresis is widely employed in biology for the separation of DNA.20,21 The SWCNT-containing gel can be prepared simply by dispersing as-prepared SWCNTs in sodium dodecyl sulfate (SDS) solution, and then gelling the mixture with liquid * To whom correspondence should be addressed. E-mail: tanaka-t@ aist.go.jp. † National Institute of Advanced Industrial Science and Technology (AIST). ‡ JST, CREST. 10.1021/nl8034866 CCC: $40.75 Published on Web 02/25/2009
2009 American Chemical Society
agarose. In electrophoresis, the application of a constant direct current to the gel causes most of the metallic SWCNTs to migrate in solution out of the starting gel, leaving the semiconducting SWCNTs in the gel itself. In the present communication, the further development of this process is reported, and the principle of gel-based MS separation is described. The gel-squeezing method appears to be effective due to the selective adsorption of semiconducting SWCNTs to the agarose gel matrix, whereas metallic nanotubes remain in the free state as SDS micelles in the interstitial solution of the gel (Figure 1a). In the electrophoresis process in which segregation was originally observed, an external electric field induces the separation of the solution and gel phases.19 To demonstrate the mechanism of the proposed method, four gel-based separation techniques are presented here: manual squeezing of the frozen and thawed gel, centrifugation, diffusion, and permeation (Figure 1b-d). Manual squeezing using finger pressure after freezing and thawing the gel is a method commonly employed to recover DNA from agarose gel (Figure 1b).22 The general procedure is as follows (see detailed methods in Supporting Information). SWCNTs synthesized by laser vaporization (LV1, 1.2 ( 0.1 nm in diameter) are dispersed by ultrasonication in a 2% SDS solution, and then the mixture is purified by centrifugation. The resulting SWCNT dispersion is prepared as a gel, which is then frozen (Figure 2a) and thawed. Squeezing with the fingers (Figure 2b) then causes the
Figure 1. (a) Model of MS separation using agarose gel. Red, semiconducting SWCNTs; beige, agarose gel matrix; green, metallic SWCNTs; yellow, SDS. (b-e) Schematic diagrams showing steps of MS separation using agarose gel: (b) freeze and squeeze, (c) centrifugation, (d) diffusion, and (e) permeation. M, metallic SWCNT; S, semiconducting SWCNT.
solution phase to separate from the gel. This process does not require the use of special equipment other than a domestic freezer. Without the freezing and thawing steps, the gel is simply crushed into small fragments when squeezed, and the solution cannot be recovered. The solution obtained by this method is gray in color (Figure 2b), while the remaining gel is greenish (Figure 2c), reflecting the partitioning between metallic and semiconducting nanotubes.19 The SWCNTs in the squeezed gel could be recovered as a solution by heating, centrifugation, and redispersion (Figure 2d, middle, S). The ratio of metallic to semiconducting SWCNTs in each of the phases was estimated according to the absorbance in the M11 (500-700 nm) and S22 (700-1000 nm) bands (Figure 2e, middle), where the subscripts denote the order 1498
Figure 2. MS separation by freeze and squeeze method. Frozen agarose gel containing LV1-SWCNTs dispersed with SDS (a) is thawed and squeezed by hand (b). The gel is greenish after squeezing, indicating segregation of semiconducting nanotubes (c). (d) Solutions of separated metallic and semiconducting SWCNTs. (e) Absorption spectra of separated SWCNTs: (left) HiPco SWCNTs (1.0 nm diameter), (middle) LV1 SWCNTs (1.2 mm), and (right) LV2 SWCNTs (1.4 mm). Black spectra are the results for the SWCNT dispersion before separation.
of the optical transitions.23 These results clearly show that the metallic and semiconducting components were segregated with a relatively high degree of purity in the two respective phases by this squeezing process. Calculation of the areas of the M11 and S22 bands in reference to the bands for the unsqueezed sample containing 33% metallic and 67% semiconducting SWCNTs indicates that the semiconducting SWCNT content in fraction S is ca. 95%, and the metallic content in fraction M is ca. 70%.24 MS separation of the HiPco nanotubes (1.0 ( 0.3 nm) and thicker LV nanotubes (LV2, 1.4 ( 0.1 nm) was also successful, as shown in Figure 2d, and the separations were confirmed from the absorption spectra (Figure 2e). This method is thus suitable for separating SWCNTs of various diameters or diameter distributions. Centrifugation was also examined as a means of recovering the solution from the agarose gel (Figure 1c). A range of surfactants and gels were tested for this method of separation Nano Lett., Vol. 9, No. 4, 2009
Figure 3. Spectral results for MS separation by centrifugation method using HiPco SWCNTs at various concentrations of agarose (0.05-1.0%). (a) Gel fraction. (b) Solution fraction.
(Supporting Information, Figure S1). Using agarose gel, good separation of the metallic and semiconducting nanotubes was achieved with the SDS surfactant. When sodium cholate was used in place of SDS, we could detect MS separation to some extent, but not in the case of sodium dodecylbenzene sulfonate or sodium deoxycholate at all. The use of agar gel, a mixture of agarose and agaropectin, resulted in partial MS separation, whereas no separation was achieved with either Gelrite or gellan gum (polysaccharide gels). The specific combination of gel and surfactant is therefore very important for MS separation by this method. It was also found that the purity of the metallic and semiconducting SWCNTs obtained by centrifugation could be improved by optimization of the agarose gel concentration (Figure 3). The purity of semiconducting SWCNTs in the compressed gel was found to increase with decreasing agarose concentration in the starting gel (Figure 3a), while the purity of the metallic SWCNTs in the solution fraction improved as the agarose concentration increased up to approximately 1.0% (Figure 3b), beyond which the metallic nanotube purity increased only slightly. When the separation was repeated for the solution fraction concentrated metallic SWCNTs, further enrichment of metallic SWCNTs was detected. We also performed Raman spectroscopic analysis and confirmed the MS separation (see detail in Supporting Information, Figure S2). The feasibility of two nonmechanical separation methods, diffusion and permeation, was also examined. In the diffusion method, the SWCNT-containing gel was soaked in an elution buffer containing SDS (Figure 1d). To facilitate effective diffusion of SWCNTs, a gel with small particle size (ca. 2.5 mm in diameter) was employed. MS separation was sucNano Lett., Vol. 9, No. 4, 2009
Figure 4. FET fabrication using separated semiconducting SWCNTs. (a) Optical image of FET devices with pairs of source/drain electrodes (scale bar, 1 mm). (b) Schematic diagram of a backgated transistor (S, source; D, drain; G, gate). FET geometry was top contact with channel length of 20 µm and channel width of 200 µm. A 200 nm SiO2 layer was used as the gate dielectric. (c) Atomic force microscopy image of SWCNT networks on SiO2/Si substrate (scale bar, 250 nm). (d) Transfer characteristics (Vds ) -1 V), forward and reverse sweeps of drain current (Ids) vs gate voltage (Vg). (e) Output characteristics, Ids vs drain voltage (Vds) characteristics measured at gate voltages of -40 (top) to -10 V (bottom) with 5 V steps. (f) Histogram of distribution of Ion/Ioff ratios. The 17 and 22 devices were characterized for the semiconducting SWCNTs (red) and unseparated SWCNTs (blue), respectively.
cessfully observed in this process, although the purities of the respective phases were relatively poor (Supporting Information, Figure S3). In the opposite permeation process, agarose gel (without SWCNTs) was added to the SWCNT dispersion (Figure 1e, Supporting Information, Figure S4). The selective accumulation of semiconducting SWCNTs on the gel was observed. Cross-sectional observations of the individual gel particles indicate that the semiconducting SWCNTs were adsorbed throughout the bulk of the gel particles in addition to the surface. The SWCNTs in the gel 1499
could be further purified by centrifugation to remove contaminating metallic SWCNTs (Supporting Information, Figure S4). These results also provide evidence of the strong and selective interaction between semiconducting SWCNTs and agarose gel, while the metallic nanotubes remain in the free state as SDS micelles (Figure 1a). The electronic properties of the separated semiconducting SWCNTs were confirmed by examining the performance of FETs prepared from nanotube networks (Figure 4a-c, see Supporting Information, Methods). Figure 4d,e shows the typical transfer and output characteristics of the FET. The FET exhibited p-type behavior with on/off ratio of 104. Mobility was estimated to be 0.7 cm2/Vs. Figure 4f shows histogram of on/off ratios for devices constructed from the semiconducting SWCNTs and unseparated SWCNTs. It was confirmed that SWCNTs densities on the both devices were equivalent (see Supporting Information, Figure S5). Almost all devices from the semiconducting SWCNTs (94%) displayed a high on/off ratio of more than 104 without electrical breakdown,2 while no devices using unseparated SWCNTs displayed the high on/off ratio. These results indicate that the electronic performance of FET device was improved by using semiconductor-enriched SWCNTs. The present results demonstrate that the electric field applied in the original electrophoresis experiment is not critical in order to achieve MS separation using agarose gel. However, the specific combination of surfactant and gel, SDS and agarose, was found to be very important for achieving effective separation, consistent with previous reports that have indicated that certain mixtures of surfactants are required for MS separation.16,17 The concentration of agarose that defines the pore size of the gel affects the purities of metallic and semiconducting SWCNTs (Figure 3), suggesting the correlation between the lengths of SWCNTs and the purities in this separation method. The method of MS separation presented in this study is superior to previously reported methods, providing rapid and inexpensive recovery of both metallic and semiconducting SWCNTs with yields approaching 100%. The method is readily scalable and is not restricted by equipment limitations (see Supporting Information, Figure S6). FET devices prepared directly from the separated nanotubes were also demonstrated to be feasible even without optimization of the separation process. With further improvement, this gel-based separation method has profound potential for the industrial production of metallic and semiconducting SWCNTs and is expected to stimulate and facilitate basic and applied research on SWCNT electronics. Acknowledgment. This work was supported in part by Grant-in-Aid for Young Scientists (18710108) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan to T.T. and by the Industrial Technology
1500
Research Grant Program (2008) of the New Energy and Industrial Technology Development Organization (NEDO) of Japan to T.T. We thank M. Shiraishi, T. Takenobu, H. Okimoto, H. Miyazaki, S. Wang, K. Yanagi, and D. Nishide for useful discussions, M. Ogura for illustrational assistance, and Y. Urabe for technical assistance. Supporting Information Available: Description of detailed experimental procedures and supplementary figures. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Tans, S. J.; Verschueren, A. R. M.; Dekker, C. Nature 1998, 393, 49. (2) Collins, P. C.; Arnold, M. S.; Avouris, P. Science 2001, 292, 706. (3) Baughman, R. H.; Zakhidov, A. A.; de Heer, W. A. Science 2002, 297, 787. (4) Krupke, R.; Linden, S.; Rapp, M.; Hennrich, F. AdV. Mater. 2006, 18, 1468. (5) Krupke, R.; Hennrich, F.; Weber, H. B.; Kappes, M. M.; von Lohneysen, H. Nano Lett. 2003, 3, 1019. (6) Krupke, R.; Hennrich, F.; von Lohneysen, H.; Kappes, M. M. Science 2003, 301, 344. (7) Miyata, Y.; Maniwa, Y.; Kataura, H. J. Phys. Chem. B 2006, 110, 25. (8) Maeda, Y.; Kimura, S.; Kanda, M.; Hirashima, Y.; Hasegawa, T.; Wakahara, T.; Lian, Y. F.; Nakahodo, T.; Tsuchiya, T.; Akasaka, T.; Lu, J.; Zhang, X. W.; Gao, Z. X.; Yu, Y. P.; Nagase, S.; Kazaoui, S.; Minami, N.; Shimizu, T.; Tokumoto, H.; Saito, R. J. Am. Chem. Soc. 2005, 127, 10287. (9) Maeda, Y.; Kanda, M.; Hashimoto, M.; Hasegawa, T.; Kimura, S.; Lian, Y. F.; Wakahara, T.; Akasaka, T.; Kazaoui, S.; Minami, N.; Okazaki, T.; Hayamizu, Y.; Hata, K.; Lu, J.; Nagase, S. J. Am. Chem. Soc. 2006, 128, 12239. (10) Li, H. P.; Zhou, B.; Lin, Y.; Gu, L. R.; Wang, W.; Fernando, K. A. S.; Kumar, S.; Allard, L. F.; Sun, Y. P. J. Am. Chem. Soc. 2004, 126, 1014. (11) Wang, W.; Fernando, K. A. S.; Lin, Y.; Meziani, M. J.; Veca, L. M.; Cao, L.; Zhang, P.; Kimani, M. M.; Sun, Y. P. J. Am. Chem. Soc. 2008, 130, 1415. (12) Zheng, M.; Diner, B. A. J. Am. Chem. Soc. 2004, 126, 15490. (13) Zheng, M.; Jagota, A.; Semke, E. D.; Diner, B. A.; Mclean, R. S.; Lustig, S. R.; Richardson, R. E.; Tassi, N. G. Nat. Mater. 2003, 2, 338. (14) Nish, a.; Hwang, J. Y.; Doig, J.; Nicholas, R. J. Nat. Nanotechnol. 2007, 2, 640. (15) Chen, F. M.; Wang, B.; Chen, Y.; Li, L. J. Nano Lett. 2007, 7, 3013. (16) Arnold, M. S.; Green, A. A.; Hulvat, J. F.; Stupp, S. I.; Hersam, M. C. Nat. Nanotechnol. 2006, 1, 60. (17) Yanagi, K.; Miyata, M.; Kataura, H. Appl. Phys. Express 2008, 1, 034003. (18) Hersam, M. C. Nat. Nanotechnol. 2008, 3, 387. (19) Tanaka, T.; Jin, H.; Miyata, Y.; Kataura, H. Appl. Phys. Express 2008, 1, 114001. (20) Sambrook, J.; Russell, D. W. Molecular Cloning: a laboratory manual, 3rd ed.; Cold Spring Harbor Laboratory: Cold Spring Harbor, NY, 2001; p 5.1. (21) Tanaka, T.; Takahashi, F.; Fukui, T.; Fujiwara, S.; Atomi, H.; Imanaka, T. J. Bacteriol. 2005, 187, 7038. (22) Tautz, D.; Renz, M. Anal. Biochem. 1983, 132, 14. (23) Kataura, H.; Kumazawa, Y.; Maniwa, Y.; Umezu, I.; Suzuki, S.; Ohtsuka, Y.; Achiba, Y. Synth. Met. 1999, 103, 2555. (24) Miyata, Y.; Yanagi, K.; Maniwa, Y.; Kataura, H. J. Phys. Chem. C 2008, 112, 13187.
NL8034866
Nano Lett., Vol. 9, No. 4, 2009
