Synthesis of Group III Antimonide Nanowires

Transcription

1 J. Phys. Chem. C 2007, 111, Synthesis of Group III Antimonide Nanowires S. Vaddiraju,, M. K. Sunkara,*, A. H. Chin, C. Z. Ning,*,, G. R. Dholakia, and M. Meyyappan Department of Chemical Engineering, UniVersity of LouisVille, LouisVille, Kentucky 40292, NASA Ames Research Center, Moffett Field, California 94035, and Center for Nanophotonics and Department of Electrical Engineering, Arizona State UniVersity, Tempe, Arizona ReceiVed: December 26, 2006; In Final Form: March 17, 2007 Introduction Synthesis of nanowires of Group III antimonides (GaSb and InSb) is studied in detail using two approaches: (i) direct antimonidization of Group III metal droplets and (ii) reactive vapor transport of Group III metals in the presence of antimony in the vapor phase. The diameter of the GaSb nanowires ranged from 30 to 700 nm and length from a few to hundreds of micrometers. GaSb nanowires as long as 1 millimeter have been synthesized using direct antimonidization of large (several millimeters) sized gallium droplets. Reactive vapor transport of Group III metals in the presence of antimony in the vapor phase led to the formation of homoepitaxially oriented GaSb nanowire arrays on top of GaSb crystals. In the case of InSb, 100-nm-thick nanowires were obtained by direct antimonidization of indium droplets. Optical and electrical measurements of the GaSb nanowires, performed using photoluminescence and scanning tunneling spectroscopy, reveal a band gap of 0.72 ev, similar to that of bulk GaSb. Synthesis of inorganic semiconductor nanowires has attracted much interest over the last 10 years 1-4 with potential applications in nanoscale lasers, detectors, sensors, and other information technology components. 5-8 So far, most of the synthesis work focused on wide band gap III-V orii-vi materials where optical transitions are in the UV-visible-blue wavelength ranges. 5-7 There is, however, a vast wavelength range in the infrared regime, where detectors and lasers are highly desired. Nanowire technology might be able to provide unique solutions for these wavelength ranges. GaSb and InSb, with respective band gaps of 0.72 ev and 0.17 ev, have long been recognized as important materials for midwave and long-wave infrared (IR) detection and lasing applications In addition, the low effective masses of the carriers and the large exciton Bohr radii of these antimonides make them ideal candidates for studying the quantum confinement effects in nanowires. 9 Despite much interest in antimonide semiconductors, the synthesis of one-dimensional structures of GaSb and InSb has not been reported extensively in the literature, although the synthesis of the same in bulk 11 and quantum dot 12 forms has been reported. Very few reports, including the synthesis of GaSb nanowires using a single source precursor 13 and InSb nanowires using electrodeposition in porous templates, 14 do exist in the literature. Recently, we have synthesized GaSb nanowires and successfully demonstrated lasing from a single GaSb submicrometer-sized wire in the IR wavelength regime. 15 Here, we present a detailed study on the synthesis of both GaSb and InSb nanowires. The goal of the work is to determine whether antimonidization reactions behave similar to nitridation reactions in terms of nanowire growth from the Group III metal * To whom correspondence should be addressed. mahendra@louisville.edu (M.K.S.); cning@asu.edu (C.Z.N.). University of Louisville. NASA Ames Research Center. Arizona State University. Current Address: Massachusetts Institute of Technology, 25 Ames Street, , MA droplets. The original concept for bulk nucleation and growth of nanowires from low melting molten metal pools was demonstrated and developed for Si, Ge, and later for Group III nitrides, oxides, and sulfides. 3,16-18 Another goal is to determine whether Group III metal droplet-led growth via self-catalysis can be employed for antimonide nanowire synthesis, similar to previous demonstration of InN growth. 19 In addition to the synthesis, results from band gap and surface electronic structure characterization are presented. In the direct antimonidization of gallium (indium) approach, droplets of gallium (indium) are exposed to antimony vapor. The dissolution and supersaturation of gallium (indium) with antimony leads to the nucleation of GaSb (InSb) crystals on top of droplets. Further, growth of these droplets into nanowires occurs by basal attachment. In contrast, the reactive vapor transport approach involves the vapor transport and subsequent reaction of gallium and antimony, both supplied through the vapor phase. In this case, the formation of GaSb crystal nuclei occurs on the substrate in the initial stages. This is followed by selective wetting and gallium droplet formation on top of the GaSb crystals. Hence, further growth of these crystal nuclei occurs in a tip-led growth fashion leading to the formation of GaSb nanowires. Experimental Methods Synthesis of GaSb nanowires was performed in a hot-wall chemical vapor deposition (HWCVD) setup (Figure 1) which consists of a 1-in. quartz tube housed inside a programmable, single temperature zone Lindberg oven. The oven is capable of heating up to 1100 C with an accuracy of (1 C. Growth of GaSb nanowires was accomplished by both spontaneous nucleation 3,16-18 and reactive vapor transport techniques For the direct antimonidization approach, amorphous quartz substrates with molten gallium droplets were exposed to antimony. Antimony was supplied using either pure antimony or antimony chloride (SbCl 3 ). Experiments using pure solid antimony as the source were performed by placing antimony powder in a quartz crucible on the upstream side of the reactor /jp068943r CCC: $ American Chemical Society Published on Web 05/01/2007

2 7340 J. Phys. Chem. C, Vol. 111, No. 20, 2007 Vaddiraju et al. Figure 1. (a) Schematic representation of the reactive vapor transport setup employed for Group III antimonide nanowire synthesis. The inset shows the schematic representation of the setup of the substrates used for both direct antimonidization and reactive vapor transport approaches. TABLE 1: Experimental Conditions Employed for the Synthesis of GaSb and InSb Nanowires no. group III metal source antimony source process conditions result 1 direct antimonidization of molten gallium droplets 2 reactive vapor transport method 3 direct antimonidization of molten indium droplets pure gallium droplets supported on amorphous quartz substrates pure gallium vapor transported onto amorphous quartz substrates pure indium droplets supported on amorphous quartz substrates indium droplets on amorphous quartz substrates in the presence of chlorine pure antimony vapor transported from a crucible onto the gallium droplets SbCl 3 vapor transported from a sample cylinder onto the gallium droplets pure antimony vapor transported from a crucible onto gallium droplets InSb powder supported in a crucible InSb powder supported in a crucible 100 Torr, C, 50 sccm of 10% Ar in H m Torr, C, 50 sccm of 10% Ar in H m Torr, 1000 C, 50 sccm of 10% Ar in H nm-thick GaSb nanowires nm-thick GaSb nanowires nm-thick GaSb nanowires 150 m Torr, 450 C, 50 sccm of H nm-thick InSb nanowires 150 m Torr, 450 C, 50 sccm of H nm-thick InSb nanowires In that case, the antimony powder source and the molten gallium covered substrates were at the same temperature because of the use of a single temperature-zone furnace. In the case of the SbCl 3 source, a sample cylinder, maintained at 80 C, was attached to the HWCVD setup. A gas mixture of 50 sccm hydrogen and 20 sccm argon was used as the carrier gas in all the above experiments, for the vapor-phase transport of antimony/sbcl 3 onto the molten gallium at pressures ranging from 30 mtorr to 100 Torr and substrate temperatures of C. Reactive vapor transport experiments for the synthesis of GaSb nanowires were performed under similar conditions wherein gallium (from a pure gallium source) was supplied in the vapor phase along with antimony. Here, the reaction of antimony and gallium vapors occurred on a blank quartz substrate placed directly above the gallium source. The placement of the substrates, for both the direct antimonidization and reactive vapor transport approach, is shown in the inset to Figure 1. In all the cases, the substrates were supported on quartz boats (inset to Figure 1) inside the quartz tube reactor setup. The substrates for direct antimonidization approach were placed inside the quartz boat (Figure 1) and the temperature of the substrates is expected to follow that of the oven. However, the substrates for reactive vapor transport experiments were placed on top of the boat, mechanically supported by the walls of the boat. In this configuration, the temperature of the substrates, measured using a pyrometer, was observed to be about 100 C lower than the oven temperature. For the case of InSb, only direct antimonidization of indium droplets was used. Substrates pre-coated with droplets of indium using simple evaporation were used in these experiments. The substrates were placed on a boron nitride crucible filled with indium. The crucible was then placed on a ceramic heater (GE

3 J. Phys. Chem. C, Vol. 111, No. 20, Synthesis of Group III Antimonide Nanowires Figure 2. (a) Micrograph of GaSb nanowires synthesized using direct antimonidization of gallium droplets. (b) Optical micrograph of a 1-mmlong nanowire synthesized using this approach. (c) Micrograph of a molten gallium droplet with GaSb nanowires coming out radially in all directions. (d) SEM image showing rectangular faceting of a GaSb nanowire. (e) SEM image showing hexagonal faceting of a GaSb nanowire. (f) XRD of the bulk synthesized nanowires showing that they have a diamond cubic crystal structure, with a lattice parameter of Å. (g) Raman spectrum of the synthesized nanowires, showing a primary mode at 159 cm-1, corresponding to GaSb. Advanced Ceramics) and heated in the presence of hydrogen atmosphere at a temperature of 800 C to obtain droplets of indium on the substrates. The size of the indium droplets on the substrate under these conditions varied from 20 nm to 2 µm. Antimony was then supplied onto these droplets in the quartz tube setup mentioned above using pure InSb powder as the source. The antimonidization of indium was performed by placing substrates on top of the quartz boat containing the InSb powder, with the indium droplet coated side of the substrate facing the InSb source. In this case, the temperature of the substrate is lower than that of the InSb source by 100 C (450 vs 550 C). Experiments were performed both with and without the presence of chlorine. The addition of chlorine was accomplished by simply dipping the indium droplet-coated quartz

4 7342 J. Phys. Chem. C, Vol. 111, No. 20, 2007 Vaddiraju et al. Figure 3. (a) and (b) Micrographs of GaSb nanowires synthesized using antimony chloride (SbCl 3) as the source of antimony. (c) Low magnification transmission electron micrograph of a GaSb nanowire. (d) High magnification transmission electron micrograph and the corresponding diffraction pattern indicating the growth direction of the nanowires to be [110]. substrates in 30 vol % hydrochloric acid solution before the experiments. The experimental conditions used for the synthesis of both GaSb and InSb nanowires are summarized in detail in Table 1. The STS measurements were performed in vacuum (10-6 Torr), using an Oxford mini cryo scanning tunneling microscope (STM) equipped with mechanically cut Pt-Ir tips. Typical set point bias voltage and tunnel currents employed were 1 V and na, respectively. Results and Discussion (a) GaSb Nanowire Synthesis. First set of experiments were performed by supplying antimony through the vapor phase onto large (several millimeters) sized gallium droplets at 100 Torr pressure. The scanning electron microscopy (SEM) image in Figure 2a shows a high density of nucleation and growth of well-faceted GaSb nanowires from a millimeter-sized gallium droplet. Figure 2a shows only a small part of a droplet with nanowires coming out of it. Nanowires as long as 1 mm were synthesized under these conditions. An optical micrograph of a 1-mm-long nanowire is presented in Figure 2b. In this approach, the direct antimonidization of gallium droplets lead to bulk supersaturation of molten gallium with antimony, which leads to spontaneous nucleation of GaSb crystal nuclei on top of the droplet. Further growth of the nuclei occurs via basal attachment. This leads to the formation of GaSb nanowires growing outward from the droplet in all directions as shown in Figure 2c similar to our earlier observations with Si, Ge, and III nitrides. 3,16-18 All the nanowires are well faceted, and they predominantly exhibit a rectangular cross-section consistent with the [110] growth direction (see Figure 2d). In some cases, hexagonal cross-sectional faceting is also observed (Figure 2e), which is due to the appearance of (111) facets along with (110) facets. X-ray diffraction (XRD) spectrum shown in Figure 2f for the as-synthesized nanowires contains primary reflections corresponding to the diamond cubic structure of GaSb with a lattice parameter of Å. 10 Raman spectrum of the nanowires in Figure 2g shows a longitudinal phonon mode peak (159 cm -1 ) corresponding to GaSb. In addition to the primary phonon mode, higher order phonon modes are also seen. The SEM images in parts a and b of Figure 3 show the GaSb nanowires synthesized using SbCl 3 as the source for antimony. Nanowires with diameters in the range of nm (Figure 3c) and 5 µm in length were obtained. The growth of nanowires in this case followed the same model as above, i.e., the spontaneous nucleation and basal growth from large gallium droplets. Here, the presence of chlorine aided in suppressing the lateral growth while promoting growth of GaSb crystal nuclei in one dimension. High-resolution TEM image in Figure 3d shows that these nanowires grow in [110] direction. Selected area diffraction pattern of a nanowire obtained along the [-111] zone axis is also shown as inset in Figure 3d. GaSb nanowires were also synthesized using the reactive vapor transport approach wherein both gallium and antimony were supplied onto blank quartz substrates. The growth of nanowires was observed only on the side of the quartz substrates facing the gallium source, similar to the growth of InN nanowires reported previously. 19 The diameter of the nanowires was approximately nm. In this case, the nanowires were observed to grow from the underlying crystals as shown in Figure 4a with gallium droplets at their tips. In this selfcatalytic mode, the nucleation of the crystal initially occurs and then the Group III metal droplets on top of these crystals lead

5 Synthesis of Group III Antimonide Nanowires J. Phys. Chem. C, Vol. 111, No. 20, Figure 4. (a) Low magnification micrograph of GaSb nanowires synthesized using reactive vapor transport approach. (b) High magnification micrograph showing homoepitaxial growth of GaSb nanowires on GaSb crystals. (c) Image of a single nanowire growing at an angle of 54.7 on a (111) surface. (d) Image of a single nanowire growing at an angle of 90 on a (100) GaSb surface. (e) High-resolution TEM image of a GaSb nanowire synthesized under these conditions, showing the growth direction of the nanowires to be [110]. TABLE 2: Gibbs Free Energy Change for the Reactions Involved in the Formation of GaSb, at a Pressure of 1 atm reactive vapor transport temperature (K) Ga(g) + 1/2Sb2(g) f GaSb(s) ( G, J/mol) spontaneous growth Ga(g) + Sb(g) f GaSb(s) Ga(g) + SbCl3(g) + 3/2H2(g) f GaSb(s) + 3HCl(g) Ga(s+l) + Sb(g) f GaSb(s) Ga(s+l) + 1/2Sb2(g) f GaSb(s) Ga(s+l) + SbCl3(g) + 3/2H2(g) f GaSb(s) + 3HCl(g) to one-dimensional growth via liquid-phase epitaxy with the underlying crystal. These results are similar to InN19 and GaN20 nanowire growth, where the Group III metal is vapor transported onto the substrates in a dissociated ammonia environment. In all these cases, the size of the resulting nanowires is primarily controlled by the size of metal droplet formed. The droplet size could increase with time, thus leading to extensive tapering upward. The diameters of the nanowires could be as small as 10 nm at the base and could reach a few hundreds of nanometers at the tip. Here, the increase in the size of the droplet could be due to the change in the wetting behavior and hence the change in the contact angle between the molten metal droplet and the underlying crystal as the growth proceeds. In some cases, the formation of large crystals of GaSb was observed on the substrate, and the nanowires were found to be in epitaxy with the underlying crystal, as shown in parts c and d of Figure 4. The nanowires were found to be oriented relative to both the (111) and (100) planes of the underlying crystals. In particular, the nanowires were either oriented at an angle of 54.7 to the (111) plane (Figure 4c) or were parallel to the (100) plane (Figure 4d) of the underlying crystal. On the (100) plane, the nanowires were observed to grow perpendicular to the surface in the initial stages. This is followed by a sharp bend and the growth of the nanowires parallel to the (100) plane. In addition, the formation of etch pits was also observed under the nanowires grown on the (100) plane, which might be due to the dissolution of antimony from the crystal into gallium during the wire formation by liquid-phase epitaxy. On the basis of the orientation of the nanowires relative to the crystal planes, the growth direction of the nanowires is expected to be [110]. TEM analysis (Figure 4e) of the nanowires shows that the growth direction is [110]. The high magnification micrograph in Figure 4e shows the plane spacing to be nm, which corresponds to the [110] planes of GaSb. The results indicate the possibility of obtaining epitaxially oriented nanowire arrays over single-crystal substrates. Also, the results show that the

6 7344 J. Phys. Chem. C, Vol. 111, No. 20, 2007 Vaddiraju et al. forms a dimer in the vapor. Gallium also evaporates at high enough temperatures. The reactions of gallium with Sb 2 for the formation of GaSb are also spontaneous and represented below in eqs 3 and 4 Ga (s+l) Sb 2(g) f GaSb (s) (3) Ga (g) Sb 2(g) f GaSb (s) (4) Hence, it is thermodynamically possible to react gallium with antimony and obtain GaSb. The Gibbs free energy changes associated with each of the above reactions at various temperatures are presented in Table 2. Further, the reaction of gallium and antimony chloride for the formation of GaSb is also spontaneous in both spontaneous nucleation and reactive vapor transport approaches. The reactions are represented below Ga (s+l) + SbCl 3(g) H 2 f GaSb (s) + 3HCl (g) (5) Figure 5. Photoluminescence of GaSb nanowires. (a) GaSb nanowires (red) compared to GaSb wafer (black) at room temperature. (b) temperature-dependent photoluminescence of a single GaSb nanowire. nanowires tend to attain a steady-state growth direction based on the process conditions even though the growth direction could be determined by epitaxial conditions. Thermodynamic analysis of the gallium-antimony system shows that the reaction between gallium and antimony for the formation of solid GaSb is spontaneous in both the spontaneous nucleation and basal growth and reactive vapor transport approaches, 21 as shown in eqs 1 and 2, respectively Ga (l) + Sb (s) f GaSb (s) (1) Ga (g) + Sb (g) f GaSb (s) (2) Even though the sublimation of solid antimony into vapor is not spontaneous, it is well-known that antimony evaporates at high temperatures. In addition, it is also known that antimony Ga (g) + SbCl 3(g) H 2 f GaSb (s) + 3HCl (g) (6) The gallium-antimony phase diagram 22 has a eutectic point at a temperature of 29.7 C in the excess gallium region. The melting point of GaSb is C. From the phase diagram, it can also be observed that the solubility of antimony in gallium is negligible at temperatures below 400 C. Hence, in principle, it should be possible to spontaneously nucleate GaSb crystal nuclei by supersaturating molten gallium droplets with antimony. Similar to our earlier results with silicon and germanium nanowire growth from molten gallium droplets, the nucleation of GaSb nanowires from molten gallium droplets also occurred in a spontaneous manner. The spontaneity of nucleation can be clearly inferred from the uniformity of the lengths and also the monosize distribution of the obtained nanowires. 23 The spontaneity of nucleation could be explained based on the thermodynamic arguments proposed by Chandrasekeran et al., i.e., there exists a supersaturation limit at which the nucleation of GaSb from gallium melt tends to be spontaneous. Because of the lack of data to account for the variation of the above parameters with varying gas-phase composition, the theoretical prediction of the diameters of the nanowires is difficult. This is because of the dependence of surface tension of gallium, interfacial energy, and the interaction parameters on the gasphase composition along with temperature. In addition, the dynamics of nuclei in the initial stages, depending upon the Figure 6. (a) Current-voltage (I-V) characteristics of GaSb nanowires obtained using STS measurements. (b) The di/dv curve clearly shows a band gap of 0.7 ev, consistent with that obtained using PL measurements.

7 J. Phys. Chem. C, Vol. 111, No. 20, Synthesis of Group III Antimonide Nanowires Figure 7. InSb nanowires synthesized using indium (a) without the presence of any chlorine and (b) with the presence of chlorine. Pure InSb was used as the source of antimony in both the cases. (c) XRD of the bulk synthesized nanowires showing that they have a diamond cubic crystal structure, with a lattice parameter of A. (d) Raman spectrum of the synthesized nanowires, showing modes corresponding to InSb. (e) Low-magnification transmission electron micrograph of an InSb nanowire, and (f) high-magnification transmission electron micrograph and the corresponding diffraction pattern indicating the growth direction of the nanowires to be [110]. temperature and the gas phase, could alter the final nanowire size. For example, the nuclei size estimated using the supersaturation with the concentrations at the spinodal limit (the concentrations at which spontaneous nucleation occurs) and the equilibrium solubility may not be preserved as the final nanowire size. The observed variation in the diameter of the obtained

8 7346 J. Phys. Chem. C, Vol. 111, No. 20, 2007 Vaddiraju et al. Figure 8. (a) I-V characteristics of the synthesized InSb nanowires, and (b) tunneling conductance of the nanowires showing negligible DOS. GaSb nanowires with varying gas-phase composition (presence or absence of chlorine) is suspected to be as follows. Addition of chlorine to the gas phase (through the SbCl 3 source) leads to an increase in the surface tension of the gallium, leading to the suppression of lateral growth, thereby leading to the formation of thin nanowires. However, in the absence of chlorine in the gas phase, the coalescence of the smaller GaSb nuclei may result in bigger crystals, leading to the formation of thick nanowires. A second possibility is the lateral growth of thin nanowires formed into thicker nanowires due to the better wetting of GaSb by gallium in the absence of chlorine. Photoluminescence (PL) studies show that the GaSb nanowires exhibit a band gap of 0.72 ev ( 1720 nm). The roomtemperature PL spectrum of the nanowires along with that obtained from a commercial single-crystal GaSb wafer is presented in Figure 5a. The spectrum obtained from the nanowires match well with that for the bulk wafer. Furthermore, our lasing experiment 15 showed that the GaSb nanowires are of high optical quality. Figure 5b shows the temperaturedependent PL of a single GaSb nanowire, showing the expected blue-shifting of the PL peak with decreasing temperature. In addition to the PL measurements, scanning tunneling spectroscopy (STS) of the nanowires also confirmed the band gap of GaSb to be approximately 0.7 ev (Figure 6). Current-voltage characteristics of the GaSb nanowires obtained using STS are presented in Figure 6a. The I-V curve is asymmetric, with a steeper slope in the negative bias region, corresponding to electrons tunneling from the valence band of the semiconducting nanowires. The normalized differential conductivity [(di/dv)- (V/I)], proportional to the density of states at (E F ) obtained numerically from the measured I-V characteristics, is presented in Figure 6b. The curve clearly reveals the conduction and the valence band edges and shows a broad band gap of 0.7 ev. (b) InSb Nanowire Synthesis. The growth conditions corresponding to the direct antimonidization of indium droplets for the spontaneous nucleation and growth of InSb nanowires are presented in Table 1. These nanowires are approximately 100 nm in diameter and about 2 µm long as seen in Figure 7a. In some cases, chlorine chemistry was used for enhancing the surface tension of the indium melt surface for reducing the lateral growth. These experiments were performed by dipping the indium droplet coated substrates in 30% (by volume) hydrochloric acid for 2 min before the experiment. The SEM image in Figure 7b indicates high densities of straight InSb nanowire arrays on top of indium droplets. The use of chlorinated indium metal source allows for the formation of spherical shape of indium metal droplets (Figure 7a) compared to a more flat film like appearance observed in the absence of any chlorine (Figure 7b). Nevertheless, both cases exhibited spontaneous nucleation and growth of high densities of InSb nanowires from pure indium metal. Chemical composition and phase of the obtained nanowires were verified using Raman spectroscopy and XRD, respectively. XRD of the nanowires, in Figure 7c, shows that they have a diamond cubic phase with a lattice parameter of Å. Raman spectrum (Figure 7d) of the nanowires shows two primary modes of InSb at 179 and 190 cm The growth direction of the InSb nanowires is [110] as confirmed by the TEM image in Figure 7f. In the case of GaN, the growth direction obtained in the spontaneous nucleation and growth approach is always C-plane oriented, and different from the A-plane orientation observed in reactive vapor transport approach. In contrast, GaSb has a diamond cubic crystal structure and the growth direction expected is [111] and not the observed [110] direction. However, the growth direction can be varied in the case of InSb by varying the growth kinetics using temperature as a parameter. This was previously observed for the case of Ge nanowires synthesized from molten pools of gallium in a spontaneous nucleation and growth approach. 22 Further, high vapor pressures of indium at the experimental temperatures always led to indium-rich conditions by forming indium droplets in the initial stages. The antimonidization of these droplets always led to spontaneous nucleation and growth of a high density of nanowires as shown in parts a and b of Figure 7. Nevertheless, experiments using metal organic compounds of indium and/or other compounds could be used for developing self-catalysis schemes of InSb nanowire arrays onto single-crystal substrates similar to that observed for GaSb nanowires. STS analysis performed on the synthesized InSb nanowires is presented in Figure 8. The I-V curve obtained from InSb nanowires is more asymmetric, compared to that observed in the case of GaSb. Further, InSb nanowires show a U-shaped di/dv with negligible density of states, making it difficult to estimate the exact value of the gap accurately, consistent with InSb being a very narrow band gap semiconductor having a bulk gap of 0.17 ev. Conclusions In summary, we have presented a detailed study on the direct synthesis of GaSb and InSb nanowires using gallium/indium pure metal sources and antimony or SbCl 3 as the antimony source. The results clearly illustrate that the direct antimonidization reaction of Group III metal droplets lead to spontaneous nucleation and growth of a high density of nanowires from larger sized metal droplets. In the second approach, the vapor transport of gallium in the presence of antimony in the gas phase, led to gallium droplet at tips, and thereby the tip-led growth of nanowires with liquid-phase epitaxy. The synthesis concept

9 Synthesis of Group III Antimonide Nanowires J. Phys. Chem. C, Vol. 111, No. 20, shown here could be extended to other materials, such as aluminum antimonide (AlSb). In addition, our results show that homoepitaxially oriented antimonide nanowires can be synthesized over large areas by employing single-crystal substrates and by controlling the supply of Group III metal. Acknowledgment. This work was supported by a NASA contract to U. of Louisville (NASA-JRI-NNA05CS49A) during S.V. s stay at NASA Ames. Research at U. of L. was supported by Kentucky Science and Engineering Foundation (KSEF ). C.Z.N. during this work was employed by the University Affiliated Research Center (UARC) at NASA Ames, operated by the University of California, Santa Cruz. A.H.C. and G.R.D. are employed by ELORET Corporation and their work at NASA Ames is supported by a subcontract from the UARC. References and Notes (1) Trentler, T. J.; Hickman, K. M.; Goel, S. C.; Viano, A. M.; Gibbons, P. C.; Buhro, W. E. Science 1995, 270 (5243), (2) Morales, A. M.; Lieber, C. M. Science 1998, 279 (5348), (3) Sunkara, M. K.; Sharma, S.; Miranda, R.; Lian, G.; Dickey, E. C. Appl. Phys. Lett. 2001, 79 (10), (4) Zhang, Y. F.; Tang, Y. H.; Lam, C.; Wang, N.; Lee, C. S.; Bello, I.; Lee, S. T. J. Cryst. Growth 2000, 212 (1-2), (5) Huang, M. H.; Mao, S.; Feick, H.; Yan, H.; Wu, Y.; Kind, H.; Weber, E.; Russo, R.; Yang, P. Science 2001, 292 (5523), (6) Duan, X.; Huang, Y.; Agarwal, R.; Lieber, C. M. Nature 2003, 421 (6920), (7) Johnson, J. C.; Yan, H. Q.; Yang, P. D.; Saykally, R. J. J. Phys. Chem. B 2003, 107 (34), (8) Favier, F.; Walter, E. C.; Zach, M. P.; Benter, T.; Penner, R. M. Science 2001, 293 (5538), (9) Vurgaftman, I.; Meyer, J. R.; Ram, Mohan, L. R. J. Appl. Phys. 2001, 89 (11), (10) Dutta, S.; Bhat, H. L.; Kumar, V. J. Appl. Phys. 1997, 81 (9), , and references therein. (11) Moravec, F. J. Cryst. Growth 1993, 128 (1-4), (12) Alphandery, E.; Nicholas, R. J.; Mason, N. J.; Zhang, B. Appl. Phys. Lett. 1999, 74 (14), (13) Kuczkowski, A.; Schultz, S.; Assenmacher, W. J. Mat. Chem. 2001, 11 (12), (14) Zhang, X. R.; Hao, Y. F.; Meng, G. W.; Zhang, L. D. J. Electrochem. Soc. 2005, 152 (10), C664-C668. (15) Chin, A. H.; Vaddiraju, S.; Maslov, A. V.; Ning, C. Z.; Sunkara, M. K.; Meyyappan, M. App. Phys. Lett. 2006, 88, (16) Chandrasekeran, H.; Sunkara, M. K. Proceedings of the Materials Research Society, 693, , (17) Sharma, S.; Sunkara, M. K. J. Amer. Chem. Soc. 2002, 124 (41), (18) Sunkara, M. K.; Sharma, S.; Chandrasekeran, H.; Krogman, K.; Bhimarasetti, G. J. Mater. Chem. 2004, 14 (4), (19) Vaddiraju, S.; Mohite, A.; Chin, A.; Meyyappan, M.; Sumanasekera, G.; Alphenaar, B. W.; Sunkara, M. K. Nano Lett. 2005, 5 (8), (20) Li, H. W.; Chin, A. H.; Sunkara, M. K. AdV. Mat. 2006, 18 (2), 216. (21) Kancke, O.; Kubaschewski, O.; Hesselmann, K. Thermochemical Properties of Inorganic Substances; Springer-Verlag Publishers: New York, (22) Hansen, M. Constitution of Binary Alloys; McGraw-Hill Publishers: New York, (23) Chandrasekeran, H.; Sumanasekera, G. U.; Sunkara, M. K. J. Phys. Chem. B 2006, 110 (37), (24) Chandrasekeran, H. Rationalizing nucleation and growth in the vapor-liquid-solid (VLS) methods. Ph.D. Dissertation, University of Louisville, Louisville, (25) Aoki, K.; Anastassakis, E.; Cardona, M. Phys. ReV. B1984, 30 (2),