A Rapid Synthesis of Iron Phosphate Nanoparticles via Surface

Hyun Jin Yang,† Hyun Jae Song,† Hyun Joon Shin,§ and Hee Cheul Choi*,†. Department of Chemistry and Pohang Accelerator Laboratory and Departmen...
0 downloads 0 Views 483KB Size
Download PDF
9098

Langmuir 2005, 21, 9098-9102

A Rapid Synthesis of Iron Phosphate Nanoparticles via Surface-Mediated Spontaneous Reaction for the Growth of High-Yield, Single-Walled Carbon Nanotubes Hyun Jin Yang,† Hyun Jae Song,† Hyun Joon Shin,§ and Hee Cheul Choi*,† Department of Chemistry and Pohang Accelerator Laboratory and Department of Physics, Pohang University of Science and Technology, San 31, Hyoja-Dong, Nam-Gu, Pohang, 790-784 South Korea Received June 5, 2005. In Final Form: July 23, 2005 The direct formation of iron phosphate nanoparticles on hydroxyl-terminated SiO2/Si substrates with a narrow size distribution (average diameter ) 2.2 nm) is achieved by a simple room temperature spontaneous reaction of ferric chloride and phosphoric acid. Single-walled carbon nanotubes (SWNTs) are grown in high yield from the synthesized iron phosphate nanoparticles by the thermal chemical vapor deposition (CVD) method, as confirmed by atomic force microscopy (AFM) and Raman spectroscopy. Furthermore, threeterminal, p-type, nanotube network field effect transistor (FET) devices are successfully fabricated using the synthesized SWNTs via the photolithography technique. The reduced solubility of Fe(III) ions when they form iron phosphate salts in aqueous media is the main driving force for the nanoparticle formation. Systematic control experiments reveal that the surface property, concentration, and pH of the reaction solution play equally important roles in the formation of nanoparticles.

Introduction Along with semiconductor nanoparticles, metal and metal oxide particles in the quantum size regime have drawn much interest because of their unique optical, electrical, and catalytic properties for various potential applications such as nanoscale optoelectronic devices,1 chemical sensors,2 display,3 photovoltaics,4 and nanobiosensors.5 One important application for metal nanoparticles is to exploit their catalytic activities for the growth of diameter-controlled, one-dimensional (1D) nanostructures, such as carbon nanotubes, metal, or semiconductor nanowires.6 During the growth of 1D nanostructures using chemical vapor deposition (CVD) or vapor-liquid-solid (VLS) processes, discrete metal nanoparticles provided a confined space in which precursor moieties were dissolved and saturated. Hence, the resulting 1D nanostructures dictate their diameters, similar to those of catalyst nanoparticles.7 To manufacture size-controlled metal or metal oxide nanoparticles, various synthetic techniques have been * E-mail: [email protected]. † Department of Chemistry. § Pohang Accelerator Laboratory and Department of Physics. (1) (a) Hickman, J. J.; Ofer, D.; Laibinis, P. E.; Whitesides, G. M.; Wrighton, M. S. Science 1991, 252, 688. (b) Chen, S.; Ingram, R. S.; Hostetler, M. J.; Pietron, J. J.; Murray, R. W.; Schaaff, T. G.; Khoury, J. T.; Alvarez, M. M.; Whetten, R. L. Science 1998, 280, 2098. (c) Markovich, G.; Collier, C. P.; Henrichs, S. E.; Remacle, F.; Levine, R. D.; Heath, J. R. Acc. Chem. Res. 1999, 32, 415. (d) Willner, I.; Willner, B. Pure Appl. Chem. 2001, 73, 535. (2) Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Letsinger, R. L.; Mirkin, C. A. Science 1997, 277, 1078. (3) Colvin, V. L.; Schlamp, M. C.; Alivisatos, A. P. Nature 1994, 370, 354. (4) Huynh, W. U.; Peng, X.; Alivisatos, A. P. Adv. Mater. 1999, 11, 923. (5) (a) Bruchez, M., Jr.; Moronne, M.; Gin, P.; Weiss, S.; Alivisatos, A. P. Science 1998, 281, 2013. (b) Chan W. C. W.; Nie, S. Science 1998, 281, 2016. (c) Mattoussi, H.; Mauro, J. M.; Goldman, E. R.; Anderson, G. P.; Sundar, V. C.; Mikulec, F. V.; Bawendi, M. G.; J. Am. Chem. Soc. 2000, 122, 12142. (d) Gomez-Romero, P. Adv. Mater. 2001, 13, 163. (6) (a) Dai, H. Acc. Chem. Res. 2002, 35, 1035. (b) Gudiksen, M. S.; Wang, J.; Lieber, C. M. J. Phys. Chem. B 2001, 105, 4062. (7) Xia, Y.; Yang, P.; Sun, Y.; Wu, Y.; Mayers, B.; Gates, B.; Yin, Y.; Kim, F.; Yan, H. Adv. Mater. 2003, 15, 353.

developed. They include conventional solution-based syntheses, which generally employ soluble precursors, reducing agents, and structure directing agents.8 Electrochemical, sonochemical, and photocatalytic approaches9 as well as template-assisted syntheses using organic or biological host molecules, such as dendrimers, block copolymers, ferritin, etc., have also been introduced.10 These methods generally entail external reducing sources and proper organic capping molecules, in addition to complex experimental equipment, to handle air-sensitive precursor molecules in inert reaction environments.11,12 Our recent efforts focused on the development of spontaneous chemical reactions for the formation of organic-free, discrete metal nanoparticles. Spontaneous reactions enable the direct formation of target nanoparticles on specific substrates by means of simple chemical reactions in solution phase under ambient conditions. An example of the recently developed chemical pathways for the spontaneous nanoparticle formation is a galvanic displacement reaction in aqueous media. Ionic species of Au, Pt, and Pd were spontaneously reduced to form respective discrete nanoparticles on germanium and various metal surfaces by the spontaneous exchange of electrons and holes between metal ions and solid substrates at room temperature.13 A similar concept has been (8) Rotello, V. M. Nanoparticles: Building Blocks for Nanotechnology; Kluwer Academic/Plenum Publishers: New York, 2004, pp 1-27. (9) (a) Penner, R. M. J. Phys. Chem. B 2002, 106, 3339. (b) Suslick, K. S.; Price, G. J. Annu. Rev. Mater. Sci. 1999, 29, 295. (c) Itakura, T.; Torigoe, K.; Esumi, K. Langmuir 1995, 11, 4129. (10) (a) Choi, H. C.; Kim, W.; Wang, D.; Dai, H. J. Phys. Chem. B 2002, 106, 12361. (b) (i) Fu, Q.; Huang, S.; Liu, J. J. Phys. Chem. B 2004, 108, 6124. (ii) Han, S.; Yu, T.; Park, J.; Koo, B.; Joo, J.; Hyeon, T.; Hong; S.; Im, J. J. Phys. Chem. B 2004, 108, 8091. (c) Li, Y.; Kim, W.; Zhang, Y.; Rolandi, M.; Wang, D.; Dai, H. J. Phys. Chem. 2001, 105, 11424. (11) (a) Bawendi, M.; Murray, C.; Norris, D. J. Am. Chem. Soc. 1993, 115, 8706. (b) Talapin, D.; Rogach, A.; Kornowski, A.; Haase, M.; Weller, H. Nano Lett. 2001, 1, 207. (12) (a) Hyeon, T.; Lee, S. S.; Park, J.; Chung, Y.; Na, H. B. J. Am. Chem. Soc. 2001, 123, 12798. (b) Cheon, J.; Kang, N.-J.; Lee, S.-M.; Lee, J.-H.; Yoon, J.-H.; Oh, S. J. J. Am. Chem. Soc. 2004, 126, 1950. (13) Porter, L. A., Jr.; Choi, H. C.; Ribbe, A. E.; Buriak, J. M. Nano Lett. 2002, 2, 1067.

10.1021/la051484l CCC: $30.25 © 2005 American Chemical Society Published on Web 08/30/2005

Rapid Synthesis of Iron Phosphate Nanoparticles

further applied to generate noble metal nanoparticles that are selectively populated on the sidewalls of single-walled carbon nanotubes (SWNTs).14 More recently, we demonstrated a simple chemical approach for iron nanoparticle formation by the reaction of Fe(III) ions and hydroxylamine in the presence of SiO2/Si substrates.15 In this case, hydroxylamine serves as both a clustering agent for Fe(III) ions and an anchoring molecule between iron clusters and surface hydroxyl groups. Herein, we report the spontaneous formation of iron phosphate nanoparticles on SiO2/Si surfaces by utilizing the solubility drop that is induced when Fe(III) ions that are adsorbed on hydroxyl-terminated SiO2/Si substrates react with phosphate ions to form water-insoluble iron phosphate salts at room temperature. Detailed studies about the morphological characteristics, formation mechanism, and chemical identities of iron phosphate nanoparticles were performed using atomic force microscopy (AFM) and synchrotron X-ray photoelectron spectroscopy (XPS). The synthesized iron phosphate nanoparticles were successfully applied for the growth of high-yield SWNTs using the CVD method. Furthermore, the fabrication of carbon nanotube network field effect transistor (FET) devices using the prepared SWNTs was also demonstrated. Experimental Section All of the nanoparticle formation reactions were carried out on highly doped, p-type silicon wafers on which a 500-nm-thick SiO2 layer was thermally grown (SiO2/Si). Formation of Iron Phosphate Nanoparticles. SiO2/Si substrates were cleaned with Piranha solution (H2SO4:H2O2 ) 3:1), followed by a thorough rinsing with 2-propanol (IPA) and deionized (DI) water (18 MΩ), and then dried with a N2 stream. (caution: Piranha is a highly oxidizing agent, so careful handling is required) Ferric chloride (FeCl3‚6H2O, 98%, Aldrich) and phosphoric acid (H3PO4, 85 wt %, Yakuri Pure Chemicals) were used as received. Nanoparticle formation was accomplished by an immersion of precleaned SiO2/Si substrates into watercontaining scintillation vials followed by the sequential addition of appropriate volumes of 10 mM ferric chloride and 10 mM phosphoric acid. The total volume of the solution was maintained at 10 mL. Note that the ferric chloride solutions were freshly prepared before use to avoid the unwanted aggregation of hexaaquo complexes. After three min of reaction, the substrates were removed and rinsed with DI water and then dried with a N2 stream. The population and morphological information of the iron phosphate nanoparticles were investigated using AFM (tapping mode, Nanoscope III, Digital Instruments). For the qualitative elemental analysis, an XPS study was performed in an ultrahigh vacuum chamber equipped with a monochromatic linear-polarized soft X-ray at 1200 eV of photon energy out of an undulator radiation beamline (8A1 U7-beamline, Pohang Accelerator Laboratory). Growth of SWNTs and Fabrication of Nanotube FET Devices. SiO2/Si substrates containing iron phosphate nanoparticles were brought into the CVD system equipped with gas flow controllers and a tube furnace (Atomate, Inc.). The samples were heated under a 500 sccm of Ar gas flow environment until the furnace temperature reached 850 °C, and then Ar gas was replaced with H2, CH4, and C2H4 at the rates of 500, 1000, and 20 sccm, respectively. After 10 min of reaction at 900 °C, the furnace was cooled to room temperature under an Ar atmosphere. The successful growth of SWNTs was confirmed by AFM and further characterized with Raman spectroscopy with a Renishaw RM 1000-Invia set to 20 mW of excitation power and 632.8 nm (1.96 eV) of laser wavelength. Nanotube network FET devices were fabricated using a conventional photolithographic technique, (14) Choi, H. C.; Shim, M.; Bangsaruntip, S.; Dai, H. J. Am. Chem. Soc. 2002, 124, 9058. (15) Choi, H. C.; Kundaria, S.; Wang, D.; Javey, A.; Wang, Q.; Rolandi, M.; Dai, H. Nano Lett. 2003, 3, 157.

Langmuir, Vol. 21, No. 20, 2005 9099

Figure 1. AFM images of iron phosphate nanoparticles in low magnification (a) (the scale bar is 300 nm) and in high magnification (b) (the scale bar is 50 nm). (c) A histogram of the diameter distribution of spontaneously formed iron phosphate nanoparticles on a SiO2/Si substrate. The average diameter is 2.2(3) nm. as described previously.16 Their p-type transistor characteristics (I-Vg) were measured using a semiconductor analyzer (Keithley, 4200-SCS).

Results and Discussion The spontaneous formation of iron phosphate nanoparticles on SiO2/Si substrates was achieved by an immersion of precleaned SiO2/Si substrates into a solution containing ferric chloride and phosphoric acid. Figure 1a,b shows typical AFM images of spontaneously formed iron phosphate nanoparticles when a precleaned SiO2/Si substrate is immersed into a solution containing 200 µL of 10 mM phosphoric acid and 200 µL of 10 mM ferric chloride for 3 min (the pH of the final solution is 3.2). To obtain detailed information about the size distribution of the prepared nanoparticles, 115 nanoparticles were selected from various regions, and their diameters were individually measured from the height profiles using AFM (Figure 1c). The histogram of the diameter distribution indicates that the average diameter is 2.2(8) nm. Although the total range of diameters spans from 0.6 to 4.5 nm, more than 80% of the nanoparticles were found in the range of 1-3 nm. When one considers the fact that there is no organic capping molecule involved, it is quite astonishing to obtain nanoparticles with such a narrow size distribution without severe aggregation. The major driving force for the nanoparticle formation is believed to be the reduced solubility of the Fe(III) ions in aqueous media when they form iron phosphate salts. The solubility product constant (Ksp) of iron phosphate dihydrate is 9.91 × 10-16 at 25 °C in an aqueous environment, which indicates that iron phosphate is insoluble in water.17,18 The deposition of a water-insoluble iron phosphate nanoparticle is supposed to be dominant on hydroxyl-terminated SiO2 surfaces because the electrostatic interactions between the Fe(III) ions and the surface hydroxyl groups increase the local concentration (16) Ozel, T.; Gaur, A.; Rogers, J. A.; Shim, M. Nano Lett. 2005, 5, 905. (17) Solubility Product Constants. In CRC Handbook of Chemistry and Physics, Internet Version 2005; Lide, D. R., Ed.; http://www.hbcpnetbase.com; CRC Press: Boca Raton, FL, 2005. (18) (a) Scaccia, S.; Carewska, M.; Bartolomeo, A.; Prosini, P. Thermochim. Acta 2002, 383, 145. (b) Gadgil, M. M.; Kulshreshtha, S. K. J. Solid State Chem. 1994, 111, 357.

9100

Langmuir, Vol. 21, No. 20, 2005

Yang et al.

Figure 2. A schematic view of iron phosphate nanoparticle formation on hydroxyl-terminated SiO2/Si substrate. (a) The increased local concentration of Fe(III) ions caused by the electrostatic interaction between Fe(III) ions and the hydroxyl group on the surface. (b) The formation of water-insoluble iron phosphate nanoparticles after the addition of phosphoric acid.

Figure 4. AFM images showing the effects of the concentration of the reactants on the formation of iron phosphate nanoparticles. The scale bars are 300 nm.

Figure 3. Effects of the surface hydroxyl group population on the formation of iron phosphate nanoparticles. AFM images of nanoparticles formed on SiO2/Si substrates that were (a) cleaned with solvents only, (b) cleaned with Piranha, and (c) calcined at 800 °C in air. Note that the calcination process diminishes the population of hydroxyl groups by converting them into siloxane forms. The scale bars are 300 nm.

of Fe(III) ions around the surface (Figure 2). To prove these hypotheses, we systematically investigated the effects of the surface property, concentration, and pH of the reaction solutions on the formation of iron phosphate nanoparticles. Effects of Surface Property on Iron Phosphate Nanoparticle Formation. During nanoparticle formation, hydroxyl (-OH) groups on the surface seem to play an important role. As shown in Figure 3a, nanoparticles were formed in poor quality on the softly cleaned surfaces (only rinsed by copious amounts of DI water, acetone, and IPA), where most of the surface hydroxyl groups may be hindered by various contaminants. However, Piranhatreated surfaces show high-quality nanoparticle formation in terms of population, size distribution, and degree of aggregation (Figure 3b). In this case, Piranha treatment certainly helps to remove naturally stained organic residues almost completely and consequently maximizes the population of hydroxyl groups on the surfaces. We further examined substrates on which the population of hydroxyl groups is intentionally decreased by calcining them at 800 °C in air for 10 min (Figure 3c).15 These results clearly show that the population of surface hydroxyl groups is critical for the determination of the local concentration of Fe(III) ions near the surface, which directly affects the formation of water-insoluble iron phosphate nanoparticles. Effects of Concentration and pH on Iron Phosphate Nanoparticle Formation: Dramatic changes in the populations and sizes of the nanoparticles were observed when the concentrations of phosphoric acid and ferric chloride were systematically changed (Figure 4). Both the sizes and the populations of the nanoparticles were continuously increased when the phosphoric acid concentration was increased from 10 µM to 4 mM while the ferric chloride concentration was fixed at 200 µM

(Figure 4c-e). However, the further increase of phosphoric acid concentration to 10 mM substantially decreased the population of nanoparticles (Figure 4f). Next, when ferric chloride concentration was changed from 10 µM to 200 µM at a fixed concentration of phosphoric acid (200 µM), the populations and sizes of the nanoparticles were increased as the concentration of ferric chloride increased (Figure 4g,d). Again, when the ferric chloride concentration was further increased to 4 mM, the number of nanoparticles was drastically decreased (Figure 4a). A similar tendency was also observed when the ferric chloride concentration was varied against a different fixed phosphoric acid concentration (4 mM, Figure 4b,e,h). Because the nonlinear dependence of nanoparticle formation on reactant concentration seems to be related to the pH alteration of the solutions, we further tested pH effects on the formation of nanoparticles. Large variations in the populations and size distributions of the nanoparticles were observed when the pHs of the solutions were varied from 2.1 to 10.7 (Figure 5). Interestingly, we found that little or no particle was formed at both extreme levels of pH. When 4 µmol of HCl was added into the original reaction solution (200 µM FeCl3 + 200 µM H3PO4; natural pH ) 3.2), the pH drop was negligible (pH ) 3.0); thus, the population as well as the size of nanoparticles were almost identical (Figure 5b,c). When an additional 80 µmol of HCl was added, very few nanoparticles were formed with a significant pH drop (pH ) 2.1; Figure 5a). The addition of 4 µmol of NaOH into the original solution increased the pH to 3.38, resulting in the severe aggregation of the nanoparticles (Figure 5d), whereas the further increase of the pH to 10.7 yielded almost no particles (Figure 5e). These phenomena can be explained as follows. At low pH, the dissociation constants of phosphoric acid (pKa1 ) 2.16, pKa2 ) 7.21, and pKa3 ) 12.32 at 25 °C)19 are considerably decreased; hence, the nanoparticle formation is suppressed because of the lack of available phosphate ions (HnPO4(3-n)- for n ) 0, 1, and 2). At high pH, the local concentration of Fe(III) ions on a SiO2/Si surface significantly diminishes because of the increased population of free hydroxide ions as well as the available phosphate ions in the solution. XPS Investigation. Figure 6 shows the XPS spectra of iron phosphate nanoparticles on SiO2/Si substrate. All (19) Dissociation Constants of Inorganic Acids and Bases. In CRC Handbook of Chemistry and Physics, Internet Version 2005, Lide, D. R., Ed.; http://www.hbcpnetbase.com; CRC Press: Boca Raton, FL, 2005.

Rapid Synthesis of Iron Phosphate Nanoparticles

Langmuir, Vol. 21, No. 20, 2005 9101

Figure 5. Effects of pH on the formation of iron phosphate nanoparticles. AFM images of nanoparticles when the pHs of the final solutions are (a) 2.14 (by the addition of 80 µmol of HCl), (b) 3.00 (by the addition of 4 µmol of HCl), (c) 3.21 (natural pH), (d) 3.38 (by the addition of 4 µmol of NaOH), and (e) 10.73 (by the addition of 20 µmol of NaOH). The scale bars are 300 nm.

of the spectra were internally calibrated with a Au 4f7/2 peak from Au film that was mounted with the samples. In the survey spectrum, the C 1s peak (284.6 eV) comes from residual impurities adsorbed on the surface when the sample was exposed to air, whereas the Si peak at 103.3 eV originates from the substrate. Because the nanoparticles were formed at a submonolayer level, the Fe and P peaks are relatively small compared to the C, O, or Si peaks. The detailed spectra of the Fe and P regions show that the binding energies of Fe 2p3/2 and P 2p3/2 are 712.5 and 133.4 eV, respectively. These results indicate that the binding energies of each element of the nanoparticles are almost identical to those of bulk iron phosphate (712.8 and 133.75 eV for Fe 2p3/2 and P 2p3/2, respectively).20 We also tried to convert iron phosphate nanoparticles into iron phosphide (FeP) nanoparticles by reducing them in a H2 atmosphere at various elevated temperatures.21 The conversion, however, was not successful, as determined by XPS, which failed to show any peak shift of P

3p3/2 from iron phosphate. This might be because the removal of phosphorus as a form of phosphoric acid from iron phosphate nanoparticles is thermodynamically more favorable than the formation of iron phosphides in the reducing environment. Synthesis of High-Yield SWNTs and Fabrication of Nanotube Network FETs. The highly efficient catalytic activities of the iron phosphate nanoparticles for the growth of SWNTs were demonstrated. AFM studies show that SWNTs that are 0.6-3 nm in diameter are successfully grown in high yield from iron phosphate nanoparticles on SiO2/Si substrates by the thermal CVD method (Figure 7a). Resonant micro-Raman spectroscopy was also used to characterize the average diameters of the SWNTs at 3 µm of the spot radius. The diameters of the SWNTs (d, in nm) were calculated from the Raman shifts (ωRBM, in cm-1) of their radial breathing modes by the equation ωRBM ) 248/dt.22 A typical Raman spectrum of SWNTs is shown in Figure 7b. The six peaks in the region between 90 and 250 cm-1 were identified as radial breathing modes, and the peak at 1580 cm-1 was identified as a G band, which comes from an in-plane graphene vibrational mode. For the diameter distribution statistics, 102 peaks taken from 50 different spots were investigated. As shown in Figure 7d, the average diameter of the carbon nanotubes was 1.5(4) nm, and this value agrees well with the average diameter of 1.6(6) nm that was obtained by AFM measurements (Figure 7c). Using high-yield SWNTs grown from iron phosphate nanoparticles, we also fabricated nanotube network FET devices by a conventional photolithographic technique. Figure 8a,b shows a schematic and an optical image of a three-terminal FET device having multiple nanotube connections between the source and the drain electrodes. A characteristic current versus gate voltage (I-Vg) curve of a p-type nanotube network FET device was measured by sweeping gate voltages from -10 to 10 V through an Si back gate (Figure 8c). Although both metallic and semiconducting SWNTs coexist between the electrodes, the fabricated devices generally show a more than 50% conductance change via electrostatic gating at a 10 mV bias (Vds) because of the higher percentage of the semiconducting SWNT. This result is comparable to the previously reported data obtained from devices made of carbon nanotubes grown using a Fe/Mo cocatalyst on alumina-supporting catalysts.23 One of the advantages of using a spontaneously formed iron phosphate nanoparticle catalyst over the alumina-supported Fe/Mo catalyst is the absence of a micron-sized supporting catalyst, which does

(20) Wagner, C. D.; Naumkin, A. V.; Kraut-Vass, A.; Allison, J. W.; Powell, C. J.; Rumble, J. R., Jr. NIST X-ray Photoelectron Spectroscopy Database, Version 3.4 (Web Version); National Institute of Standards and Technology: Gaitherburg, 2003. (21) . Stamm, K. L.; Garno, J. C.; Liu, G.-y.; Brock, S. L. J. Am. Chem. Soc. 2003, 125, 4038.

(22) Jorio, A.; Saito, R.; Hafner, J. H.; Lieber, C. M.; Hunter, M.; McClure, T.; Dresselhaus, G.; Dresselhaus, M. S. Phys. Rev. Lett. 2001, 86, 1118. (23) (a) Qi, P.; Vermesh, O.; Grecu, M.; Javey, A.; Wang, O.; Dai, H.; Peng, S.; Cho, K. J. Nano Lett. 2003, 3, 347. (b) Kong, J.; Soh, H. T.; Cassell, A. M.; Quate, C. F.; Dai, H. Nature 1998, 395, 878.

Figure 6. XPS spectra of iron phosphate nanoparticles. A survey spectrum (a) as well as high-resolution spectra of Fe (b) and P (c) are shown.

9102

Langmuir, Vol. 21, No. 20, 2005

Yang et al.

Figure 7. (a) AFM image of high-yield SWNTs grown from iron phosphate nanoparticles using CVD. The scale bar is 1 µm. (b) A typical micro-Raman spectrum of the carbon nanotubes taken from the sample shown in a. Also shown are histograms of size distribution obtained by (c) AFM height profiles (average diameter ) 1.6(6) from 92 nanotubes) and (d) micro-Raman spectra (average diameter ) 1.5(4) from 102 RBM peaks).

fabricated nanotube network FET devices, we are currently investigating the characteristics of the internal Schottky junctions and their effects on sensing performance.

Figure 8. (a) A schematic view and (b) an optical microscope image of a p-type nanotube network FET device containing multiple SWNT connections between the source and the drain electrodes. The scale bar is 200 µm. (c) A characteristic I-Vg curve obtained from a nanotube network FET device (Vds ) 10 mV).

not have the ability to control the diameters of nanotubes and causes problems for the atomic scale investigation of carbon nanotubes, for example, with AFM. Another characteristic feature of the nanotube network FET device is that the SWNTs between two electrodes have a weblike geometry, resulting in numerous quantum cross-junctions, such as metallic-metallic, metallicsemiconducting, and semiconducting-semiconducting nanotube-nanotube junctions. Although most of the chemical sensing experiments using nanotube FETs currently focus only on nanotube-metal contact regions or the surface of the carbon nanotubes, systematic studies about these quantum cross-junctions in nanotube network FET devices will provide great opportunities to develop more selective and sensitive sensor systems. Using the

Conclusion A simple route was developed for the formation of iron phosphate nanoparticles through surface-involved spontaneous chemical reactions between ferric chloride and phosphoric acid in an ambient condition. Iron phosphate nanoparticles showed excellent catalytic activity for the growth of SWNTs in high yield with relatively narrow size distribution. Using a conventional photolithography technique, p-type nanotube network FET devices were successfully fabricated with high-yield SWNTs. Because of the easiness to manipulate the catalyst, reasonably efficient gate dependence, and multiple internal crossjunctions, the nanotube network FET devices will provide great opportunities for the development of high-performance nanotube-based sensor devices with better understanding. We are currently trying to generalize this method to various metal species. The success in this will open a new window for the synthesis of unprecedented metal nanostructures directly on solid substrates in a straightforward and efficient way. Acknowledgment. This work was supported by grant No. R01-2004-000-10210-0 from the Basic Research Program of the Korea Science & Engineering Foundation. We also thank the Center for Integrated Molecular Systems (CIMS). Prof. Young Hee Lee and Mr. Hee Jin Jeong in SKKU are thanked for taking the Raman spectra. We also thank Dr. Junwoo Lim at NECST, KNU for his help in the FET device fabrication. LA051484L