NANO LETTERS
From Germanium Nanowires to Germanium-Silicon Oxide Nanotubes: Influence of Germanium Tetraiodide Precursor
2009 Vol. 9, No. 2 583-589
Jinquan Huang,† Wai Kin Chim,*,† Shijie Wang,‡ Sing Yang Chiam,‡,§ and Lai Mun Wong‡ Department of Electrical and Computer Engineering, National UniVersity of Singapore, 4 Engineering DriVe 3, 117576 Singapore, Institute of Materials Research and Engineering, A*STAR (Agency for Science, Technology and Research), 3 Research Link, 117602 Singapore, and Department of Physics, Imperial College London, Blackett Laboratory, Prince Consort Road, London SW7 2BW, United Kingdom Received September 8, 2008; Revised Manuscript Received December 17, 2008
ABSTRACT Growth of semiconductor nanowires has attracted immense attention in the field of nanotechnology as nanowires are viewed as the potential basic building blocks of future electronics. The recent renewed interest in germanium as a material for nanostructures can be attributed to its higher carrier mobility and larger Bohr radius as compared to silicon. Self-assembly synthesis of germanium nanowires (GeNWs) is often obtained through a vapor-liquid-solid mechanism, which is essentially a catalytic tip-growth process. Here we demonstrate that by introducing an additional precursor, germanium tetraiodide (GeI4), in a conventional furnace system that produces GeNWs on silicon, tubular structures of germanium-silicon (GeSi) oxide can be obtained instead. Incorporation of GeI4 results in passivation of the metal catalyst, preventing the occurrence of supersaturation, a prerequisite for the catalytic tip growth. We infer that passivation of the metal catalyst impedes Ge incorporation into the catalyst, leaving the catalyst rim as the only active sites for nucleation of both Si and Ge and thus resulting in the growth of GeSi oxide nanotubes via a root-growth process.
Nanostructured one-dimensional (1D) materials represent realistic solutions for the continued advancement in information technology.1 This advancement has been aided by numerous efforts to integrate 1D wires/tubes into conventional devices, such as nanosized transistors, on a large scale.2-4 The immense potential of these nanosized transistors has been shown by demonstrations of superior electrical characteristics, such as higher gate-controlled source-drain current and transconductance, in several differently structured field-effect devices that employed semiconducting nanowires as channel materials.5-7 The huge interest in nanostructures stems from the interesting mechanical,8,9 electrical,5,10 and optical properties11 of the nanosized device. The confinement of charges and alteration of the density of states in nanostructures provide additional dimensions for manipulation in the search for suitable materials for faster and more energy efficient electronic devices. These unique properties are not * Corresponding author,
[email protected]. † Department of Electrical and Computer Engineering, National University of Singapore. ‡ Institute of Materials Research and Engineering, A*STAR. § Department of Physics, Imperial College London. 10.1021/nl8027137 CCC: $40.75 Published on Web 01/15/2009
2009 American Chemical Society
achievable with conventional bulk materials, making nanostructures an essential component in future electronics. As a result, growth of 1D nanostructures, such as nanowires (NWs) and nanotubes (NTs), is of great interest to both the research community and industry. In particular, it is important to explore and understand the different growth techniques that have potential for eventual widespread production. The most common semiconductor used in the fabrication of electronic devices is silicon (Si). This is because Si has a stable oxide in the form of silicon dioxide (SiO2). This, together with the excellent Si/SiO2 interfacial properties, makes possible the electrical operation of field-effect transistor devices. However, there has been increasing interest in using germanium (Ge) as a channel material for future transistors due to its significantly higher hole and electron mobilities.12,13 Moreover, the exciton Bohr radius of Ge is much larger than that of Si (Ge, 24.3 nm; Si, 4.9 nm). This means that quantum confinement effects in Ge can possibly be achieved with nanostructures of relatively larger size.14 Growth of Ge nanowires (GeNWs) using different methods, such as laser ablation,15 chemical vapor deposition (CVD),16-18 vaportransport,19,20 andsupercriticalfluid-liquid-solid
Table 1. Summary of Experimental Details on the Growth of GeNWs, GeSiOxNTs, and the Control Experiments
Figure 1. Experimental setup for growth of (a) germanium nanowires and (b) germanium-silicon oxide nanotubes.
synthesis,21 has been investigated. Generally, these methods rely on a catalytic tip-growth process based on the vapor-liquid-solid (VLS) mechanism, which was first proposed by Wagner and Ellis.22 While the growth techniques and mechanisms of GeNWs are relatively well understood, studies on Ge-based nanotubes are not as extensive. Most reported growths of Ge-based nanotubes are based on the vapor adsorption on templated nanostructures such as porous anodic alumina23 and carbon nanotubes (CNTs).24 A selfassembly type approach to the growth of Ge-based nanotubes remains difficult. Such semiconductor tubular structures can be potentially used to encapsulate materials to produce new nanowires and nanocrystals in a confined configuration.25-27 In this work, we report a simple setup to produce both GeNWs and germanium-silicon oxide nanotubes (GeSiOxNTs) under different growth conditions. We will show that this simple setup yields GeNWs with reasonable density and good crystalline quality. We will also demonstrate that the addition of a precursor, germanium tetraiodide (GeI4), is crucial in affecting the growth from which nanotubes can be obtained. The possible growth mechanisms of such tubular structures will also be discussed. The experimental setups for the growth of GeNWs and GeSiOxNTs are schematically depicted in Figure 1. Both growths were carried out in a three-zone furnace (Lindberg/ Blue STF55346C) and Si(111) substrates with gold (Au) dots of 5-40 nm diameter (Supporting Information) were used in both processes. Prior to the deposition of the metal catalyst, the Si substrates were ultrasonically degreased in acetone followed by isopropyl alcohol, each for a duration of 15 min. The degreased silicon wafers were then dipped in 2% hydrofluoric acid for 1 min to remove the native oxide formed at its surface. The oxide-free substrates were immediately transferred into a thermal evaporator (Edwards Auto 306) containing Au wire (99.99+% purity, Goodfellow). A blanket deposition of Au was then evaporated onto the silicon substrates which were maintained at room temperature. The thickness of the evaporated Au is 2 nm, as 584
monitored in situ by a quartz crystal microbalance and verified ex situ by an atomic force microscope (JEOL JSPM 5200). The gold film was broken up into dots by annealing the Au-coated Si substrates in vacuum (0.005 mbar) at 400 °C for 30 min. For the growth of GeNWs, a small amount of Ge powder (99.999% purity, Sigma-Aldrich) was loaded at the closed end of a small quartz tube while the Au-dotted Si substrates were placed near to the open end of the tube as shown in Figure 1a. The growth of GeSiOxNTs was obtained using the setup shown in Figure 1b. GeI4 powder (99.99% purity, Sigma-Aldrich) was placed outside the furnace zone to allow for separate controlled heating of the GeI4 precursor, which was achieved by using an external heating tape whose temperature was monitored using a thermocouple. The setup described ensures that the GeI4 vapor flows over the Ge powder before any reaction with the Si substrate occurs. For the growth of nanowires and nanotubes, a counterflow of argon (Ar) gas was used. The flow rate of Ar was set at 100 sccm, and the pressure during growth was constantly maintained at 2 mbar by an automatic metallic-throttle valve (MKS type 253B) with PID feedback control. The temperature settings for the experiments described in this article are summarized in Table 1. The GeNWs synthesized using the above-mentioned setup have smooth surfaces and show well-defined lattice fringes (Supporting Information). A hemispherical Au dot can also be observed at the tip of the wire demonstrating the tipbased growth from the VLS mechanism. The GeNWs obtained in this case show nonordered growth on the Si(111) substrate. This is unlike the ordered growth of GeNWs synthesized via CVD processes16-18 but consistent with wires grown using Ge vapor source.15,19,20 These observations are important indications that precursors may play an important role in affecting the growth of the wires. The low-magnification scanning electron microscopy (SEM) (Philips XL30 FEG-SEM) images in panels a and b of Figure 2 show the general density of the nanotubes synthesized. The diameters of the tubes are about 30-200 nm, while the lengths of the nanotubes can be as long as 10 µm. A close-up view of the nanotubes reveals a wavy surface on the walls of the tubular structure, and the open end of Nano Lett., Vol. 9, No. 2, 2009
Figure 2. (a) SEM image of the as-synthesized germanium-silicon oxide nanotubes (GeSiOxNTs). (b) Close examination of the nanotubes reveals that each nanotube is a long, tubular structure with uniform diameter. The bright dots in the background are Au dots that remain on the substrate surface. (c and d) SEM images showing the open-ended GeSiOxNTs and the wavy surface of the walls of the tubular structure. (e) TEM image of a single GeSiOxNT and (f) its image at a higher magnification.
the nanotube can be clearly observed. The open end indicates that the tubes are not synthesized through the catalytic tipgrowth mechanism, which is responsible for the GeNW growth. The absence of a metal catalyst at the tip of the nanotube also implies that the growth mechanisms of the metal-catalyzed silicon28 and silica29 tubular structures cannot be applied here. The wall of the nanotube has a thickness of ∼5 nm as estimated from the transmission electron microscopy (TEM) (JEOL JEM 2100) images. In addition, highresolution TEM images reveal that the walls of the tubular structures do not show any form of observable diffraction fringes. The thin nanotube walls can be completely oxidized by traces of oxygen in the furnace and/or during the exposure to the ambient, since oxidation can take place from both sides of the thin (∼5 nm) wall. The amorphous nature of the nanotubes can therefore be a result of oxidation after growth. Elimination of the sources of oxygen in a better vacuum and/ or in situ characterization will be required to determine the initial elemental content of the nanotube. The chemical compositions of the tubular structures were investigated by both X-ray photoelectron spectroscopy (XPS) (UHV-VG ESCALAB 220i-XL) (Figure 3), TEM energy dispersive X-ray (EDX) analysis (Figure 4), and STEM-EDX Nano Lett., Vol. 9, No. 2, 2009
Figure 3. (a) XPS spectra of Si 2p core level scans and (b) Ge 3d XPS spectra of GeSiOxNTs. In both cases, the sample-analyzer distance was adjusted such that the signal is initially collected from the surface. The focal point of collection was gradually moved away (as indicated by the arrow) until the signal is free from the contribution of electrons from the surface, as indicated by the disappearance of the Si-Si and Ge-Ge peaks. The shift in the binding energy of the Si-O and Ge-O peaks is due to the effect of differential charging resulting from the change of X-ray flux on the surface as the focal point is gradually shifted away.
Figure 4. TEM-EDX spectrum of a typical GeSiOxNT. The C and Cu peaks detected are due to the copper-carbon TEM grid used.
mapping (Figure 5). For the XPS measurement, the sample-analyzer distances were adjusted such that the 585
Figure 5. STEM-EDX mapping of (b) Ge, (c) O, and (d) Si of a typical GeSiOxNT in (a).
signals were first collected from the substrate surface. The presence of any signal from the Si substrate surface will be indicated by the detection of the elemental silicon bonding (Si-Si) of the Si(111) substrate. The sample-analyzer distance was then adjusted such that the collected electrons became increasingly void of the contribution from the Si substrate surface. This was achieved by moving the focal point of the collection away from the substrate, as indicated by the arrow in Figure 3a. With the disappearance of the Si-Si peak contributed by the substrate, one can be confident that the remaining signal belongs largely to the free-standing nanotubes. The same procedure was used to obtain the Ge 3d core level spectra as shown in Figure 3b. The XPS analysis shows that the nanotubes consist of both Si-O and Ge-O type bondings, indicating that the tubes are essentially made up of oxidized Si and Ge. This is further verified by TEM-EDX and STEM-EDX mapping, which detected the presence of germanium, silicon, and oxygen elements. An interesting phenomenon was observed during the TEM analysis after the as-grown nanotubes were exposed to a prolonged high-intensity flux from high-energy electron beam bombardment. When viewed under high magnifications in the TEM, the nanotube under observation shrank and finally collapsed into a solid nanowire as shown in Figure 6. This may be a result of localized heating that has been shown to be capable of deforming nanostructures30 and inducing phase changes.31 Furthermore, the thin walls of the tubes are likely to have a reduced melting point due to thermodynamic size effect.32,33 The lower melting point represents reduced bond strengths, and this may aid the observed deformation of the nanotubes. This observed ease of shape transformation may lead to some potential applications in nanofluidics.34 The 586
collapse of these tubes can be controlled by changing the electron flux per unit area (i.e., by changing the spot size and current of the electron beam). This means that one has the ability to possibly control and manipulate the length and, to a certain extent, the shape of a nanotube by controlling the electron irradiation at defined locations. In order to better understand the growth mechanism of the nanotubes, three control experiments were conducted (Table 1). The first experiment kept the reaction sources unchanged, and a clean Si sample was used instead of the Au-dotted substrates. In this experiment, no growth of nanotubes was observed. This highlights that Au is essential in obtaining the tubular structures. While not restricting our analysis to the conventional VLS process, it is reasonable to assume that the Au dots either function as a catalyst or provide catalytic nucleation sites for the growth materials of the nanotubes. The second and third control experiments were designed to investigate the influence of varying precursor sources. However, before describing the experimental details, it is worthwhile to briefly discuss the reactions involved in a germanium-iodine system. For the reaction of solid Ge and GeI4 vapor, an equilibrium exists for the following reversible reaction Ge(s) + GeI4(g) / 2GeI2(g)
(1)
It has been reported that at temperatures above 600 °C35 the forward reaction (left to right in eq 1) dominates and hence germanium diiodide (GeI2) will have a larger concentration in the reaction chamber. However, this does not necessarily exclude the existence of GeI4 or Ge vapor. In fact, at our reaction temperature of 900 °C, we will still expect ∼3.6% of GeI4 to be present at equilibrium even if all the GeI4 vapor Nano Lett., Vol. 9, No. 2, 2009
Figure 6. (a-c) TEM images of a single GeSiOxNT showing gradual shape transformation under electron beam bombardment in the TEM. The diameter gradually shrinks from (a) 100 nm to (b) 70 nm to (c) 55 nm. Increasing surface roughness/folding was observed. (d) TEM image of a GeSiOxNT of 80 nm diameter collapsing into (e) a solid nanowire of 50 nm diameter.
produced is assumed to have impinged on and reacted with the Ge powder.35 In addition, free energy calculation has shown that such GeI4 vapor does not decompose to form GeI2 and free iodine when heated to a temperature of less than 2000 K. Decomposition of GeI4 to give elemental germanium and iodine is also unlikely at temperatures less than the melting point of Ge (939 °C) at reduced pressure.36 Therefore, GeI4 would remain chemically stable in vapor form at our growth temperature (900 °C). In addition to GeI2 and GeI4, Ge vapor is also present in the furnace. This is because not only the germanium powder was loaded in large excess but also a high vapor pressure is expected of the Ge powder when it is maintained at close to its melting point.37 The presence of Ge vapor is supported by the fact that elemental Ge (i.e., Ge-Ge bonding) was detected on the substrate surface during the XPS analysis (see Figure 3b). The germanium detected on the substrate surface may be a result of direct condensation of Ge vapor. In the second and third control experiments, GeI4 and GeI2 (99.99% purity, Sigma-Aldrich) were used individually as the respective sole precursor. It is found that the two experiments yielded neither nanotubes nor nanowires. Together with the observed formation of GeNWs using only Ge powder as the vapor source as described earlier, the two experiments indicate that adopting either GeI4 or GeI2 as a sole precursor is unlikely to achieve the supersaturation condition of the Au catalyst that is crucial for the VLS process to occur. This may indicate that the Ge-I bond of GeI4 or GeI2 does not react and/or dissociate on the Au dots at our investigated temperature. It should be noted that Yang et al. has attributed their observed growth of GeNWs to the presence of GeI2 from a similar combination of Ge and GeI4.38 However, we believe that since Yang et al.’s growth was conducted in a sealed tube with both sources (Ge and GeI4) heated to 1000-1100 °C, the presence of Ge vapor cannot be ruled out. Furthermore, the higher growth temNano Lett., Vol. 9, No. 2, 2009
perature used in Yang et al.’s experiment may result in the dissociation of the Ge-I bond of GeI2/GeI4 at the Au dots. This in turn allows supersaturation of Ge, enabling the growth of GeNWs via the VLS mechanism. From the test of precursors in the second and third control experiments, it is clear that the presence of both Ge and GeI2/GeI4 is crucial in obtaining the GeSiOxNTs. With this information, a growth mechanism for the nanotubes is proposed. Since no GeNWs were observed when a combination of Ge and GeI4 was used as the precursors, we have reasons to believe that either GeI2 or GeI4 might have formed a passivating layer on the Au dots to prevent Ge incorporation. This passivation hypothesis is a necessary deduction for a couple of reasons. First, the lack of GeNW growth from a pure GeI2 or GeI4 source demonstrates that Ge incorporation into Au is severely reduced or limited as no supersaturation condition is reached despite a significant amount of Ge precursor used. Second, since (Ge + GeI4) does not represent a complete reaction at our reaction temperature,35 one may expect a mixture of GeNWs with GeSiOxNTs or even a hybrid heterostructure of nanowires and nanotubes, if the Au dots are unpassivated. Indeed, passivation of Au surfaces with iodide is commonly reported, and the passivation hypothesized in this work can share common similarities.39-41 The described passivation phenomenon also suggests that the tubes may not be grown via the conventional VLS mechanism, and this is evident in the lack of a Au tip at the top of the tubular structures as mentioned earlier. Since the presence of the Au dots remains an essential criterion for the growth of the tubes (as evidenced by control experiment 1), we believe that the remaining plausible explanation of the observed growth of the tubular structures is possibly a vapor-solid process. Gold is essential in the growth of the nanotubes by acting not as a catalyst but to provide 587
produces GeNWs via the VLS mechanism. The addition of GeI4 results in possible passivation of the Au dot surface, and the tubes are grown via a root-growth mechanism instead. Acknowledgment. This work is supported by the National University of Singapore under Research Grant R-263-000420-112. The provision of a research scholarship to Jinquan Huang by the National University of Singapore is gratefully acknowledged. The authors would also like to thank Professor Jing Zhang from Imperial College London for numerous useful discussions.
Figure 7. Schematic depiction of the growth mechanism of the GeSiOxNTs. Ge atoms at the rim are a result of surface diffusion from the surrounding areas of the passivated Au dot while Si atoms can be contributed from the Au-Si alloy as indicated by the arrows. Ge and Si atoms only react at the rim of the Au-Si interface where there is sufficient Ge to result in the formation of the GeSiOxNTs.
Supporting Information Available: Figures showing SEM and TEM images of the GeNWs synthesized and SEM image of the Au dots formed by annealing a 2 nm thick Au film, SEM image of Au catalyst after nanotube removal, and SEM-EDX analysis of the Au catalyst on a sample after nanotube removal. This material is available free of charge via the Internet at http://pubs.acs.org. References
nucleation sites for the growth materials. Si atoms, similar to the case of solid-liquid-solid nanowire growth,42,43 are most likely contributed from the Au-Si alloy since Si can diffuse into a Au overlayer easily,44 while Ge atoms most likely come from some form of surface diffusion, either on the Si substrate surface or on the passivated Au dots. As pointed out by Kolasinski,45 even with the substantial amount of published work on root-growth syntheses, there is little information on the basic mechanism of the root growth of nanowires/nanotubes. However, Wang et al. observed that the edge of the catalyst is the preferential site for the root growth of their silicon oxide nanowires.42 Similarly, we expect the edge of the Au dots in our work to be the most active nucleation site whereby Ge atoms can react with Si as illustrated in the schematic diagram in Figure 7. The incorporation of oxygen atoms into the nanotubes can take place during the growth stage or as a result of subsequent oxidation as discussed earlier. The proposed mechanism requires the Au catalyst to stay on the sample surface throughout the growth. The SEM image of a sample that has the nanotubes removed by ultrasonication proves that this is indeed the case (Supporting Information). Dense Au dots with diameters ranging from 20 to 200 nm are observed on the sample surface. The increase in the size of the Au dots means that significant aggregation has occurred through Ostwald ripening46 and possibly aided by iodine47 since the original Au dots are only 5-40 nm in diameter. The close match between the range of the Au dot sizes and the diameters of the GeSiOxNTs is an added verification of the growth mechanism proposed. In addition, SEM-EDX (Zeiss Ultraplus FESEMEDAX Apollo 40 SDD) analysis on these Au dots remaining on the surface did not reveal the presence of any Ge in the Au catalyst (Supporting Information). This is consistent with the deduction of the possible passivation as proposed. In conclusion, a high density of GeSiOxNTs has been synthesized by introducing an additional precursor (GeI4) to a conventional three-zone furnace system, which originally 588
(1) International Technology Roadmap for Semiconductors (ITRS) Emerging Research Devices (ERD) (2007) http://www.itrs.net/Links/2007ITRS/ 2007_Chapters/2007_ERD.pdf. Retrieved on July 12, 2008. (2) Goldberger, J.; Hochbaum, A. I.; Fan, R.; Yang, P. D. Nano Lett. 2006, 6, 973. (3) Ng, H. T.; Han, J.; Yamada, T.; Nguyen, P.; Chen, Y. P.; Meyyappan, M. Nano Lett. 2004, 4, 1247. (4) Nguyen, P.; Ng, H. T.; Yamada, T.; Smith, M. K.; Li, J.; Han, J.; Meyyappan, M. Nano Lett. 2004, 4, 651. (5) Xiang, J.; Lu, W.; Hu, Y.; Wu, Y.; Yan, H.; Lieber, C. M. Nature 2006, 441, 489. (6) Wang, D.; Wang, Q.; Javey, A.; Tu, R.; Dai, H.; Kim, H.; McIntyre, P. C.; Krishnamohan, T.; Saraswat, K. C. Appl. Phys. Lett. 2003, 83, 2432. (7) Schmidt, V.; Riel, H.; Senz, S.; Karg, S.; Riess, W.; Gosele, U. Small 2006, 2, 85. (8) Hsin, C. L.; Mai, W. J.; Gu, Y. D.; Gao, Y. F.; Huang, C. T.; Liu, Y. Z.; Chen, L. J.; Wang, Z. L. AdV. Mater. 2008, 20, 1. (9) Wong, E. W.; Sheehan, P. E.; Lieber, C. M. Science 1997, 277, 1971. (10) Wang, D.; Chang, Y. L.; Wang, Q.; Cao, J.; Farmer, D. B.; Gordon, R. G.; Dai, H. J. Am. Chem. Soc. 2004, 126, 11602. (11) Li, Y.; Qian, F.; Xiang, J.; Lieber, C. M. Mater. Today 2006, 9, 18. (12) CRC Handbook of Chemistry and Physics, 78th ed.; Lide, D. R., Ed.; CRC Press: Boca Raton, FL, 1998. (13) Sze, S. M. Physics of Semiconductor DeVices, 2nd ed.; Wiley: New York, 1981. (14) Maeda, Y.; Tsukamoto, N.; Yazawa, Y.; Kanemitsu, Y.; Masumoto, Y. Appl. Phys. Lett. 1991, 59, 3168. (15) Morales, A. M.; Lieber, C. M. Science 1998, 279, 208. (16) Dailey, J. W.; Taraci, J.; Clement, T.; Smith, D. J.; Drucker, J.; Picraux, S. T. J. Appl. Phys. 2004, 96, 7556. (17) Kamins, T. I.; Li, X.; Williams, R. S. Nano Lett. 2004, 4, 503. (18) Jagannathan, H.; Deal, M.; Nishi, Y.; Woodruff, J.; Chidsey, C.; Mclntyre, P. C. J. Appl. Phys. 2006, 100, 024318. (19) Nguyen, P.; Ng, H. T.; Meyyappan, M. AdV. Mater. 2005, 17, 549. (20) Sun, X. H.; Didychuk, C.; Sham, T. K.; Wong, N. B. Nanotechnology 2006, 17, 2925. (21) Hanrath, T.; Korgel, B. A. AdV. Mater. 2003, 15, 437. (22) Wagner, R. S.; Ellis, W. C. Appl. Phys. Lett. 1964, 4, 89. (23) Mei, Y. F.; Siu, G. G.; Li, Z. M.; Fu, R. K. Y.; Tang, Z. K.; Chu, P. K. J. Cryst. Growth 2005, 285, 59. (24) Han, W. Q.; Wu, L.; Zhu, Y.; Strongin, M. Nano Lett. 2005, 5, 1419. (25) Hu, J. Q.; Meng, X. M.; Jiang, Y.; Lee, C. S.; Lee, S. T. AdV. Mater. 2003, 15, 70. (26) Hippe, C.; Wark, M.; Lork, E.; Ekloff, G. S. Microporous Mesoporous Mater. 1999, 31, 235. (27) Svrcek, V. Mater. Lett. 2008, 62, 2578. (28) Li, C.; Liu, Z.; Gu, C.; Xu, X.; Yang, Y. AdV. Mater. 2006, 18, 228. (29) Tuan, H. Y.; Ghezelbash, A.; Korgel, B. A. Chem. Mater. 2008, 20, 2306. Nano Lett., Vol. 9, No. 2, 2009
(30) Liang, C.; Terabe, K.; Hasegawa, T.; Aono, M. Solid State Ionics 2006, 177, 2527. (31) Sun, X.; Yu, B.; Meyyappan, M. Appl. Phys. Lett. 2007, 90, 183116. (32) Buffat, P.; Borel, J. P. Phys. ReV. A 1976, 13, 2287. (33) Wu, Y. Y.; Yang, P. D. AdV. Mater. 2001, 13, 520. (34) Goldberger, J.; Fan, R.; Yang, P. D. Acc. Chem. Res. 2006, 39, 239. (35) Lever, R. F. J. Electrochem. Soc. 1963, 110, 775. (36) Rolsten, R. F. Iodide Metals and Metal Iodides; John Wiley & Sons: New York, 1961; pp 300-304. (37) Vapor pressure of Ge: 2 × 10-7 Torr at 900 °C and 6 × 10-7 Torr at 939 °C (melting point). Source: Veeco http://www.veeco.com/library/ Learning_Center/Growth_Information/Vapor_Pressure_Data_For_Selected_Elements/index.aspx. Retrieved on July 12, 2008. (38) Wu, Y. Y.; Yang, P. D. Chem. Mater. 2000, 12, 605. (39) Rodriguez, J. F.; Mebrahtu, T.; Soriaga, M. P. J. Electroanal. Chem. 1987, 233, 283. (40) Rodriguez, J. F.; Soriaga, M. P. J. Electrochem. Soc. 1988, 135, 616.
Nano Lett., Vol. 9, No. 2, 2009
(41) Tadayyoni, M. A.; Gao, P.; Weaver, M. J. J. Electroanal. Chem. 1986, 198, 125. (42) Wang, C. Y.; Chan, L. H.; Xiao, D. Q.; Lin, T. C.; Shih, H. C. J. Vac. Sci. Technol., B 2006, 24, 613. (43) Yan, H. F.; Xing, Y. J.; Hang, Q. L.; Yu, D. P.; Wang, Y. P.; Xu, J.; Xi, Z. H.; Feng, S. Q. Chem. Phys. Lett. 2000, 323, 224. (44) Hiraki, A.; Lugujjo, E.; Mayer, J. W. J. Appl. Phys. 1972, 43, 3643. (45) Kolasinski, K. W. Curr. Opin. Solid State Mater.Sci. 2006, 10, 182. (46) Hannon, J. B.; Kodambaka, S.; Ross, F. M.; Tromp, R. M. Nature 2006, 440, 69. (47) Cheng, W.; Dong, S.; Wang, E. Angew. Chem., Int. Ed. 2003, 42, 449.
NL8027137
589
