Letter pubs.acs.org/NanoLett
A General Approach toward Shape-Controlled Synthesis of Silicon Nanowires W. Molnar,† A. Lugstein,*,† P. Pongratz,‡ M. Seyring,§ M. Rettenmayr,§ C. Borschel,∥ C. Ronning,∥ N. Auner,⊥ C. Bauch,⊥ and E. Bertagnolli† †
Institute of Solid State Electronics, TU-Wien, Floragasse 7, A-1040 Vienna, Austria Institute of Solid State Physics, TU-Wien, Wiedner Hauptstrasse 8/052, A-1040 Vienna, Austria § Institute of Materials Science and Technology, University of Jena, Löbdergraben 32, D-07743 Jena, Germany ∥ Institute of Solid State Physics, University of Jena, Max-Wien-Platz 1, D-07743 Jena, Germany ⊥ Spawnt Research GmbH, Entwicklungszentrum Wolfen, Kunstseidenstrasse 6, D-06766 Bitterfeld-Wolfen/Goethe-University Frankfurt a.M., Max-von-Laue-Strasse 7, 60438 Frankfurt am Main, Germany ‡
S Supporting Information *
ABSTRACT: Controlling the morphology, electronic properties, and growth direction of nanowires (NWs) is an important aspect regarding their integration into devices on technologically relevant scales. Using the vapor−solid−solid (VSS) approach, with Ni as a catalyst and octachlorotrisilane (Si3Cl8, OCTS) as a precursor, we achieved epitaxial growth of rectangular-shaped Si-NWs, which may have important implications for electronic mobility and light scattering in NW devices. The process parameters were adjusted to form cubic α-NiSi2 particles which further act as the shaping element leading to prismatic Si-NWs. Along with the uncommon shape, also different crystallographic growth directions, namely, [100] and [110], were observed on the very same sample. The growth orientations were determined by analysis of the NWs’ azimuths on the Si (111) substrates as well as by detailed transmission electron microscopy (TEM) and selected area electron diffraction (SAED) investigations. KEYWORDS: Silicon, octachlorotrisilane, nanowire synthesis, vapor−solid−solid, shape control
Q
may deteriorate the electronic device’s performance, in particular, for example, shorten the nonradiative lifetime of charge carriers, alternative catalysts are highly desired. So far alternative metals like Al,21 Ti,21 Ga,22 Pt,23 and Ag24 have successfully been implemented as Au-substitutes. Further it was shown that the catalyst plays an important role in defining the NW’s morphology and growth orientation.25 achieved growth of [111] oriented Si-NWs with Al via VSS growth, however requiring UHV conditions. An approach with Pt was shown by Baron et al.23 leading also to [111] oriented Si-NWs catalyzed by PtSi islands. Sunkara and Sharma22 have shown that Ga catalyzes the growth of preferably [100] oriented Si-NWs by utilizing the VLS mechanism in a microwave plasma reactor. Using Ag particles of different sizes, Wittemann et al.24 achieved Si-NWs of different orientation, and finally Arbiol et al.26 have utilized Cu leading to vapor−solid−solid (VSS)grown [110] and [1210] oriented Si-NWs with diamond and also wurtzite structures within the same NW. In this Letter we present an approach to control the shape of Si-NWs grown via the VSS mechanism. This mechanism offers great potential for NW synthesis regarding reduced growth temperatures, more uniform diameter distributions, better
uasi-one-dimensional nanowires (NWs) have attracted tremendous attention as a playground to study fundamental mesoscopic effects such as quantum confinement1,2 or single-electron transistor phenomena,3 and as potential technological applications, enabling extraordinary progress for nanoscale electronics,4,5 sensors,6,7 photonic devices,8,9 and solar cells,10 as well as catalysis11 and life science.12 However, for developing these applications at technologically relevant scales, it is essential to control the electrical, optical, and mechanical properties of NWs. These properties, aside of the composition, critically depend on the morphology, crystallographic orientation, and defect structure of the NWs. Synthesis techniques using chemical vapor deposition (CVD),13 metal organic CVD,14 molecular-beam epitaxy,15,16 and laser ablation techniques17 have therefore been directed at producing single crystal NWs with uniform geometry and uniform growth direction. However, despite great advances in synthesis, the ability to control the NW’s shape has been significantly limited. The particular type we focus on here is a catalyst-assisted method superior for the fabrication of nanometer-sized wires. For the CVD synthesis of Si-NWs, today mostly SiH4,18 SiCl4,19 and Si2H620 serve as precursor gases. Au is usually the catalyst of choice for the growth at moderate temperatures due to the low eutectic temperature with Si and its chemical inertness. However, as Au causes deep level traps in Si which © 2012 American Chemical Society
Received: August 24, 2012 Revised: November 29, 2012 Published: December 5, 2012 21
dx.doi.org/10.1021/nl303152b | Nano Lett. 2013, 13, 21−25
Nano Letters
Letter
control of growth orientation, increased purity, and more abrupt heterostructure interfaces.27 The synthesis of prismatic NWs was performed in a hot wall atmospheric pressure CVD system with a horizontal tube furnace where octachlorotrisilane (OCTS, Si3Cl8) was supplied via a saturator. For the synthesis of prismatic Si-NWs we used pieces of Si(111) as substrates. After a careful removal of the native oxide with buffered hydrofluoric acid (BHF; 7:1) we sputter deposited a 2 nm thin layer of Ni as the NW growth promoting catalyst. After a further BHF dip the samples were placed on the quartz sample holder of a magnetic specimen− transport system and instantly introduced into the CVD system. After 10 min of evacuation, the tube was purged with He, and the furnace was heated up. When the final growth temperature was reached the sample holder was transferred into the growth region by the magnetic specimen−transport system. The precursor was supplied by routing feed gas consisting of He/H2 through a saturator at room temperature leading to a partial pressure of 0.03 mbar OCTS in the feed gas. In a previous study we have shown that OCTS, in contrast to the synthesis approach with SiCl4 as a precursor, provides Si without the addition of H2 to the feed gas.28 However, under H2 atmosphere the chemical byproduct SiCl4 originating from the OCTS decomposition breaks up into Si and HCl, thus making additional Si available for NW growth. After typically 60 min growth the sample holder was pulled out of the heating zone, the precursor flow was stopped, and the quartz tube was purged with 200 sccm He for further 5 min. In contrast to our previous growth approach mentioned above, we optimized the growth conditions with respect to feed gas composition and received remarkably long Si-NWs at a growth temperature of 900 °C with diameters ranging from 100 to 200 nm by adding H2 to the feed gas (see Figure 1a). The addition of 20 sccm H2
Figure 2. (a) SEM image of epitaxially grown Si-NWs; (b) schematic of epitaxially grown [110] Si-NWs featuring an azimuth of 120° between the NWs and an angle of 54.7° with respect to the Si(111) substrate; (c) schematic of epitaxially grown [100]-NWs featuring an azimuth of 120° between them and an angle of 35.3° with respect to the Si substrate.
with respect to the (111) oriented Si substrate (see Figure 3a) forms a triangular network in the top view SEM image. According to the schematic shown in Figure 2b, these NWs are of [110] growth orientation. The second group that can be identified with the same azimuth but at an angle of 35.3° to the substrate (see Figure 3a) is supposed to be [100]-oriented NWs. As shown in the schematic of Figure 2c, such [100] oriented NWs form also a triangular network, shifted by 60° with respect to the [110] oriented NWs (see Figure 2a). Figure 3a exemplifies the angles of the epitaxial grown NWs to the substrate in excellent agreement with the tabulated
Figure 1. (a) Scanning electron microscopy (SEM) image of Si-NWs synthesized at 900 °C with Ni as a catalyst, OCTS as Si precursor, and 50 sccm He and 20 sccm H2 as feed gas. (b) Square-shaped NWs achieved for an increased gas flow of 100 sccm He and 40 sccm H2.
Figure 3. (a) Side view SEM image of [100] and [110] oriented SiNWs with angles to the substrate of 35.3° and 54.7° respectively. (b) Highly resolved SEM image of an individual, prismatic Si-NW.
to the 50 sccm He feed gas leads to straight NWs with clearly detectable catalyst particles atop of every NW. Figure 1b shows a sample grown at the same temperature and gas feed ratio but with doubled feed gas flow (100 sccm He, 40 sccm H2). The NWs grown this way appeared to be prismatic and epitaxial with diameters of 80−200 nm, lengths up to 5 μm at a maximum growth rate of 166 nm/min and a very low yield of ∼0.3 NWs/μm3. Remarkably, two types of growth directions were observed on the very same sample and therefore under identical growth conditions (see Figure 2a). The first group of epitaxial Si-NWs with an azimuth of 120° between them and an angle of 54.7°
values of 35.3° for [100] NWs and 54.7° for [110] NWs. The prismatic habitus of such grown NWs is unambiguously shown in the SEM image of Figure 3b. Their distinct shape and growth directions which were already deduced from their orientational relationship due to the epitaxial growth on the Si (111) substrate were further confirmed by transmission electron microscopy (TEM) imaging and diffraction analysis.29 The respective micrographs and selected area electron diffraction (SAED) patterns are shown in Figure 4. 22
dx.doi.org/10.1021/nl303152b | Nano Lett. 2013, 13, 21−25
Nano Letters
Letter
[110], [111], and [112] as the most probable directions of Hterminated Si-NWs. However, it is widely accepted that larger diameter Si-NWs most likely grow along the [111] directions.32 Si atoms precipitating upon the (111) surface during growth produce the largest decrease in Gibbs free energy, as these planes of Si have the largest density of surface atoms.33 For thinner NWs, the free energy of the side faces must be taken into account, and the [110] growth direction becomes more favorable. One can assume that there may exist a possibility to define and to stabilize the growth orientation by controlling the surface conditions already at the onset of the nucleation. This means that, during the NW nucleation event, the surface energies influence the nucleus structure and thus again the growth direction. However, many aspects of NW growth are beyond the scope of simple arguments based on near equilibrium growth controlled by surface energies. Kawashima et al.34 showed that there are strong correlations between growth rate, crystallinity, and catalyst particle formation, all of which are controlled by the amount of Si supplied to the catalyst. However, to the best of our knowledge, VLS synthesized Si-NWs with [100] growth direction have been reported only by Sunkara et al.22 using molten Ga as catalyst, and Lew et al.35 using alumina membranes. In general, Zhang et al.36 have shown theoretically that different shapes, hexagonal, square-shaped, octagonal, and triangular to name a few, are possible for Si-NWs of various orientations, and Ciobanu et al.37 have calculated that it is energetically possible to grow square shaped [110]-NWs, although they have to be extremely thin and H-passivated. With Ni as catalyst now we achieved prismatic Si-NWs of different orientations with a square-shaped α-NiSi2 nanoparticle atop. The growth temperature of 900 °C is 66 °C lower than the lowest liquid eutectic of Ni/Si,38 and the size of our catalytic particles is still too big to produce a massive drop in the melting temperature.39 Thus, it should be considered more of a VSS (vapor−solid−solid)-growth, where the usual thermodynamic preferences of a classical VLS growth do not apply anymore. We therefore suggest that the catalyst particle is in the solid state during growth of Si-NWs and that a solidphase diffusion process, either in the bulk or on the surface, or both, must be responsible for NW growth. Nominally
Figure 4. TEM images and the respective SAED patterns of the Nicatalyzed square-shaped Si-NWs proves the (a,b) [110] and (d,e) [100] growth orientations. The dark speckles within the wires below the catalytic particles arise from diffraction contrast produced by lattice defects. (c) The SEM image displays a clear impression of the square habitus of a catalytic particle in the nucleation phase. The SAED patterns were recorded in a perpendicular orientation to the wire axis so that the growth directions (arrows) can be determined from the indexed diffraction vectors. The zone axis of both diffraction patterns is the [001] direction.
Both [110] as well as [100]-oriented NWs feature a square habitus independently of their growth orientation with a fair amount of stacking faults and dislocation loops subsequent to the catalytic particle (Figure 4a and e). The diffraction patterns in Figure 4b and d show the NWs’ respective orientation. EDX measurements and SAED investigations of a catalytic particle as shown in Figure 4c revealed a composition of about 33% Ni and 66% Si and the space group to be Fm3m ̅ /225 with a lattice constant to be 5.406 Å, in good agreement to cubic α-NiSi2.30 Detailed calculations according the growth directions of VLS grown Si-NWs performed by Zhang et al.31 revealed [100],
Figure 5. Schemata of prismatic (a) [110] and (b) [100] oriented Si-NWs with the corresponding diffraction patterns of the Si-NW and α-NiSi2 tips recorded perpendicular to the wire axis. The zone axis of all diffraction patterns is the [001] direction. 23
dx.doi.org/10.1021/nl303152b | Nano Lett. 2013, 13, 21−25
Nano Letters
■
subeutectic NW synthesis using Au as a catalyst has been achieved for several semiconductor materials, including Ge,40 InP,41 GaAs,42 and ZnSe.43 For Si-NWs synthesis our findings are conceptually similar to a previous report by Kamins et al.44 of Ti-promoted Si-NW growth for temperatures below the Si− Ti eutectic point. Thus, it can be assumed that surface tension and surface energy are altered considerably from usual growth conditions of a VLS mechanism. Considering the small difference in the lattice constants of Si (5.429 Å) and α-NiSi2 (5.406 Å), it seems feasible that the cubic α-NiSi2 particle worked as a shaping element, which forces its shape on the NWs during growth. An explanation for the two different growth orientations on the very same sample and thus identical growth conditions can by deduced on the basis of a work of Hsu et al.45 They observed that annealing of nanogrooves in Si filled with Ni leads to the nucleation of [100] as well as [110] oriented NiSi2−NWs with good crystallinity. Also Tung et al.46 observed the formation of NiSi2-nanolayers, by annealing thin Ni films on Si(111), with two different crystallographic orientations which they identified as [100] and [110]. In addition they have shown that for Ni films with a thickness of a few nanometers, the distribution of the two orientations depended critically on the thickness of the initial Ni film. For our NW growth approach a layer of 2 nm Ni was sputter-deposited as the growth promoting catalyst on the Si (111) substrates. As for such thin sputtered layers we cannot expect a homogeneous Ni layer; variations in the film thickness will promote the NiSi2particle nucleation oriented in [100], [110], or even both directions which will be transferred to the Si-NWs. This assumption is further supported by TEM analysis, which showed the same orientation of the lattice axes of the Si-NWs and the catalytic particles atop for both growth directions (Figure 5a and b). This is proven by the coincidence of the {220} reflexes of the Si-NWs and the α-NiSi2 tip as well as the coincidence of the {400} reflexes, both shown in the SAED patterns of NW and tip (insets of Figure 5). From the indexing of the diffraction patterns the crystallographic planes of the lateral surface facets were determined. The [110] oriented NW exhibits both (001) and (−110) planes at its surface (Figure 5a), while [100] oriented NW shows {100} planes parallel to the lateral surface (Figure 5b). Based on the growth conditions and TEM results, we assume that the catalytic particle adopts a cubic structure during the initial growth stage as Ni reacts with Si to form α-NiSi2. With this cubic catalytic particle as the shaping element, prismatic SiNWs were successfully synthesized by a VSS growth mediated by the silicide. The epitaxial growth of the Si-NWs in [100] and [110] directions is a result of the preferred nucleation of αNiSi2 on Si(111) in these directions which will be transferred to the Si-NWs. To our knowledge, this is the first report of shape transfer from a catalytic particle to the NW which has to be seen as a proof-of-concept. Given the immense variety of potential binary alloy seeds, we suppose that this approach should not be limited solely to the materials discussed hereother precursors or catalysts should extend this method to other materials.
■
Letter
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors acknowledge the support by the Austrian FWF projects nos. 20937-N14 and P19414. Technical support by USTEM TU Wien is gratefully acknowledged.
■
REFERENCES
(1) Jalilian, R.; Sumanasekera, G. U.; Chandrasekharan, H.; Sunkara, M. K. Phys. Rev. B 2006, 74, 155421. (2) Demichel, O.; Calvo, V.; Pauc, N.; Besson, A.; No, P.; Oehler, F.; Gentile, P.; Magnea, N. Nano Lett. 2009, 9, 2575. (3) Thelander, C.; Martensson, T.; Bjork, M. T.; Ohlsson, B. J.; Larsson, M. W.; Wallenberg, L. R.; Samuelson, L. Appl. Phys. Lett. 2003, 83, 2052. (4) Huang, Y.; Duan, X.; Cui, Y.; Lauhon, L. J.; Kim, K. H.; Lieber, C. M. Science 2001, 294, 1313. (5) Thelander, C.; Nilsson, H. A.; Jensen, L. E.; Samuelson, L. Nano Lett. 2005, 5, 635. (6) Cui, Y.; Lieber, C. M. Science 2001, 291, 851. (7) Wang, W. U.; Chen, C.; Lin, K. H.; Fang, Y.; Lieber, C. M. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 3208. (8) Huang, M. H.; Mao, S.; Feick, H.; Yan, H.; Wu, Y.; Kind, H.; Weber, E.; Russo, R.; Yang, P. Science 2001, 292, 1897. (9) Barrelet, C. J.; Greytak, A. B.; Lieber, C. M. Nano Lett. 2004, 4, 1981. (10) Kayes, B. M.; Atwater, H. A.; Lewis, N. S. J. Appl. Phys. 2005, 97, 114302. (11) Valden, M.; Lai, X.; Goodman, D. W. Science 1998, 281, 1647. (12) Patolsky, F.; Timko, B. P.; Zheng, G.; Lieber, C. M. MRS Bull. 2007, 32, 142−9. (13) Wagner, R. S.; Ellis, W. C. Appl. Phys. Lett. 1964, 4, 89. (14) Arakawa, Y. Solid-State Electron. 1994, 37, 523. (15) Tchernycheva, M.; Harmand, J. C.; Patriarche, G.; Travers, L.; Cirlin, G. E. Nanotechnology 2006, 17, 4025. (16) Martelli, F.; Piccin, M.; Bais, G.; Jabeen, F.; Ambrosini, S.; Rubini, S.; Franciosi, A. Nanotechnology 2007, 18, 125603. (17) Wang, N.; Tang, Y. H.; Zhang, Y. F.; Lee, C. S.; Lee, S. T. Phys. Rev. B 1998, 58, R16024. (18) Cui, Yi.; Lauhon, L. J.; Gudiksen, M. S.; Wang, J.; Lieber, C. M. Appl. Phys. Lett. 2001, 78, 2214. (19) Hochbaum, A. I.; Rong, F.; Rongrui, H.; Peidong, Y. Nano Lett. 2005, 5, 457−460. (20) Park, W. I.; Zheng, G.; Jiang, X.; Tian, B.; Lieber, C. M. Nano Lett. 2008, 8, 3004−3009. (21) Sharma, S.; Kamins, T. I.; Williams, R. S. J. Cryst. Growth 2004, 267, 613−618. (22) Sunkara, M. K.; Sharma, S. Nanotechnology 2004, 15, 130−134. (23) Baron, T.; Gordon, M.; Dhalluin, F.; Ternon, C.; Ferret, P.; Gentile, P. Appl. Phys. Lett. 2006, 89, 233111. (24) Wittemann, J. V.; Münchgesang, W.; Senz, S.; Schmidt, V. J. Appl. Phys. 2010, 107, 096105. (25) Wang, Y.; Schmidt, V.; Senz, S.; Gö sele, U. Nature Nanotechnology 2006, 1, 186. (26) Arbiol, J.; Kalache, B.; Cabarrocas, P. R.; Morante, J. R.; Morral, A. F. Nanotechnology 2007, 18, 305606. (27) Lensch-Falk, J. L.; Hemesath, E. R.; Perea, D. E.; Lauhon, L. J. J. Mater. Chem. 2009, 19, 849−857. (28) Molnar, W.; Lugstein, A.; Pongratz, P.; Auner, N.; Bauch, C.; Bertagnolli, B. Nano Lett. 2010, 10, 3957−3961. (29) Seyring, M.; Song, X. Y.; Rettenmayr, M. ACS Nano 2011, 5, 2580−2586. (30) Fujitani, H.; Asano, S. Phys. Rev. B 1995, 51, 18019−18021.
ASSOCIATED CONTENT
S Supporting Information *
EDS analysis of the catalytic particle and Si-NW and SEM analysis of the catalytic particle during nucleation. This material is available free of charge via the Internet at http://pubs.acs.org. 24
dx.doi.org/10.1021/nl303152b | Nano Lett. 2013, 13, 21−25
Nano Letters
Letter
(31) Zhang, R. Q.; Lifshitz, Y.; Ma, D. D. D.; Frauenheim, T.; Lee, S. T.; Yong, S. Y. J. Chem. Phys. 2005, 123, 144703. (32) Schmidt, V.; Senz, S.; Gösele, U. Nano Lett. 2005, 5, 931−935. (33) Ghandhi, S. K. VLSI Fabrication Principles: Silicon and Gallium Arsenide, 1st ed.; Wiley: New York, 1983; Chapter 1. (34) Kawashima, T.; Mizutani, T.; Nakagawa, T.; Torii, H.; Saitoh, T.; Komori, K.; Fujii, M. Nano Lett. 2008, 8, 362. (35) Lew, K.-K.; Reuther, C.; Carim, A. H.; Redwing, J. M.; Martin, B. R. J. Vac. Sci. Technol. B 2002, 20, 389. (36) Wang, N.; Cai, Y.; Zhang, R. Q. Mater. Sci. Eng. 2008, R 60, 1− 51. (37) Ciobanu, C. V.; Chan, T.-L.; Chuang, F.-C.; Lu, N.; Wang, C.Z.; Ho, K.-M. Nano Lett. 2006, 6, 277−281. (38) Massalski, T. B.; Okamoto, H.; Subramanian, P. R.; Kacprzak, L. Binary alloy phase diagrams, 2nd ed.; National Institute of Standards and Technology (U.S.), ASM International: Materials Park, OH, 1990. (39) Buffat, P.; Morel, J. P. Phys. Rev. 1976, A13 (2287), s1976d. (40) Greytak, A. B.; Lauhon, L. J.; Gudiksen, M. S.; Lieber, C. M. Appl. Phys. Lett. 2004, 84, 4176. (41) Dick, K. A.; Deppert, K.; Martensson, T.; Mandl, B.; Samuelson, L.; Seifert, W. Nano Lett. 2005, 5, 761. (42) Persson, A. I.; Larsson, M. W.; Stenström, S.; Ohlsson, B. J.; Samuelson, L.; Wallenberg, L. R. Nat. Mater. 2004, 3, 677. (43) Colli, A.; Hofmann, S.; Ferrari, A. C.; Ducati, C.; Martelli, F.; Rubini, S.; Cabrini, S.; Franciosi, A.; Robertson. J. Appl. Phys. Lett. 2005, 86 (153), 103. (44) Kamins, T. I.; Williams, R. S.; Basile, D. P.; Hesjedal, T.; Harris, J. S. J. Appl. Phys. 2001, 89, 1008. (45) Hsu, H. F.; Tseng, C. H.; Chen, T. H. J. Nanosci. Nanotechnol. 2010, 10, 4533. (46) Tung, R. T.; Gibson, J. M.; Poate, J. M. Phys. Rev. Lett. 1983, 50 (6), 429.
25
dx.doi.org/10.1021/nl303152b | Nano Lett. 2013, 13, 21−25
