Toward Local Growth of Individual Nanowires on Three-Dimensional

Sep 7, 2011 - 39, CH-3602 Thun, Switzerland. ‡. Institute of Photonic Technology, Albert-Einstein-Strasse 9, D-07745 Jena, Germany. §. Max Planck I...
0 downloads 0 Views 2MB Size
Download PDF
LETTER pubs.acs.org/NanoLett

Toward Local Growth of Individual Nanowires on Three-Dimensional Microstructures by Using a Minimally Invasive Catalyst Templating Method Martin G€unter Jenke,*,†,|| Damiana Lerose,‡,|| Christoph Niederberger,† Johann Michler,† Silke Christiansen,‡,§ and Ivo Utke*,† †

EMPA, Swiss Federal Laboratories for Materials Science and Technology, Feuerwerkerstr. 39, CH-3602 Thun, Switzerland Institute of Photonic Technology, Albert-Einstein-Strasse 9, D-07745 Jena, Germany § Max Planck Institute for the Science of Light, G€unther Scharowsky-Strasse 1, 91058 Erlangen, Germany ‡

ABSTRACT: We present a novel minimally invasive postprocessing method for catalyst templating based on focused charged particle beam structuring, which enables a localized vapor liquid solid (VLS) growth of individual nanowires on prefabricated three-dimensional micro- and nanostructures. Gas-assisted focused electron beam induced deposition (FEBID) was used to deposit a SiOx surface layer of about 10  10 μm2 on top of a silicon atomic force microscopy cantilever. Gallium focused ion beam (FIB) milling was used to make a hole through the SiOx layer into the underlying silicon. The hole was locally filled with a gold catalyst via FEBID using either Me2Au(tfac) or Me2Au(acac) as precursor. Subsequent chemical vapor deposition (CVD)-induced VLS growth using a mixture of SiH4 and Ar resulted in individual high quality crystalline nanowires. The process, its yield, and the resulting angular distribution/crystal orientation of the silicon nanowires are discussed. The presented combined FIB/FEBID/CVD-VLS process is currently the only proven method that enables the growth of individual monocrystalline Si nanowires on prestructured substrates and devices. KEYWORDS: Focused ion beam (FIB), focused electron beam induced deposition (FEBID), vapor liquid solid growth (VLS), silicon nanowire, nanowire, chemical vapor deposition (CVD), postprocessing

M

ono- or polycrystalline nanowires (NWs) are used for a wealth of applications in quantum electronics,1 electro2,3 nics, photonics,4 biosensorics, 5 7 energy harvesting,8 and electricity generation.9 Most of the applications either need highly regular monocrystalline NW arrays of high quality on large areas10 or require single NWs at specific locations with nanoscale precision.11 Various methods have been developed for the fabrication of individual NWs at planar surfaces,12 18 but none for nonplanar surfaces, which often contain mechanically delicate features. The few existing methods that show individual NW growth on planar substrates and that could also be used on nonplanar surfaces comprise “Dip-Pen” nanolithography,19 atomic force microscopy (AFM) nanolithography,20 and electron beam induced mask deposition for nanolithography.21 However, none of these methods were reported to work on nonplanar surfaces, probably due to the small scan ranges19,20 or plasma etching processes21 involved in the methods. The method presented here uses a combination of the two minimally invasive direct writing methods, focused ion beam (FIB) milling and focused electron beam-induced deposition (FEBID) with vapor liquid solid (VLS) growth. FIB milling uses ions to physically sputter material in a well-defined manner.22 Gas-assisted FEBID is an emerging lithography concept, which enables maskless local deposition of various materials.23 It relies on the local r 2011 American Chemical Society

dissociation of functional, surface physisorbed molecules by a focused electron beam. The volatile dissociation reaction products desorb into the vacuum chamber of the scanning electron microscope (SEM) and are removed by the vacuum system, while the nonvolatile dissociation products will form a deposit. Since the molecules are continuously supplied via a gas injection system to the substrate surface, the deposition can be extended to (almost) any desired three-dimensional shape by moving the focused electron beam or keeping it stationary.22 Once the desired structure is written, the gas injection system is closed and the nondissociated adsorbates desorb, leaving the local lithographic deposit pattern on the surface. VLS growth is a special chemical vapor deposition (CVD) method, which was first described in 1964 by Wagner and Ellis24 and has been studied in detail now for several decades.25,26 Recently real time in situ transmission electron microscopy (TEM) studies have given a new insight into this important SiNW growth method.26,27 For VLS growth, a CVD chamber is evacuated and a silicon substrate with a metal dot on top is heated until the metal dot (often made of gold) liquefies. A silane precursor gas is added into the chamber. The gold dot Received: June 24, 2011 Revised: August 24, 2011 Published: September 07, 2011 4213

dx.doi.org/10.1021/nl2021448 | Nano Lett. 2011, 11, 4213–4217

Nano Letters

Figure 1. Schematic representation of selective in-place growth of an individual nanowire (NW) on a 3D nano/microstructure (AFM cantilever) by using a combination of FEBID/FIB/VLS-CVD. (a) The AFM cantilever and the corresponding cross section before the process. (b) A part of the AFM cantilever was covered with an SiO2 layer by FEBID. (c) FIB milling of a hole where the nanowire will be grown. (d) Gold-containing material is deposited at the bottom of the hole by FEBID. (e) VLS growth of the SiNW.

catalyzes the decomposition of silane precursor gas permitting the silicon atoms to dissolve in the liquefied metal. As soon as the gold dot becomes supersaturated in silicon, a silicon nanowire begins to grow at the solid liquid interface of the liquid gold dot and the solid silicon substrate. The newly developed combined FIB/FEBID/VLS-CVD process is shown schematically in Figure 1. It comprises the following steps: (a) placing of a wafer with prestructured objects such as nano/microelectro-mechanical systems (NEMS/MEMS) or nano/ micro-opto-electro-mechanical systems (NOEMS/MOEMS) into a dual beam FIB/SEM chamber (we used a wafer with

LETTER

atomic force microscope cantilevers); (b) covering of the surface at the very end of one cantilever with a thin 10  10 μm2 SiOx film using FEBID; (c) drilling of a hole into the SiOx layer and the underlying silicon by using the Ga+-ion beam of a FIB; (d) depositing gold containing material (in the following called gold dots) into the FIB holes with a second FEBID process. The templating procedures (b d) were repeated for all cantilevers inside the FIB/SEM chamber. After this minimally invasive templating procedure, performed at room temperature, the wafer was transferred into a CVD machine; there the VLS growth of the SiNWs was performed using silane as a precursor gas (e). The processing steps are described in more detail in the following text. The procedure (Figure 1) started by introducing a wafer with cantilevers into a dual beam Lyra FIB-SEM instrument from TESCAN. A roughly 10  10 μm2 and 50 nm thin SiOx film was deposited onto the cantilever surface using FEBID (30 kV, 3500 pA, 508 nm full width at half-maximum (fwhm), 0.15 beam overlap, 5 μs dwell time, 18 min, 56 s total deposition time, serpentine raster scan) and an Alemnis GmbH gas injection system (GIS) with tetraethyl-orthosilicate (TEOS) (CAS 78-10-4) as precursor molecule, see Figure 1b. The SiOx film is used as a protection layer against unwanted chemisorption of catalyst and parasitic NW growth. It could in principle be produced by any method compatible with the fabrication process of the final device, for example, plasma-enhanced CVD, wet or dry oxidation. We used the FEBID process due to its local and minimally invasive nature, which is advantageous for most fully fabricated micro/nanostructures. However, we have also successfully used dry oxidation on flat Si-wafers. The SiOx film was subsequently FIB structured using a Ga+ ion beam (30 kV, 10 pA, about 8 nm fwhm, circular raster scan, milling durations: milliseconds to a few seconds) (Figure 1c). Holes with nominal diameters between 40 and 800 nm and depths between 50 nm and 1.5 μm were milled into the SiOx film and the silicon substrate (Figure 2a) to ensure epitaxial SiNW growth according to the crystallographic direction preferred under the VLS growth conditions used. The experimentally determined milling rate was 0.4979 μm3/(nA 3 s) for the SiOx layer alone and 0.6145 μm3/(nA 3 s) for the combination of the SiOx layer and the bulk Si underneath. The sputter yield of the SiOx layer alone was 1.76 SiO2 atoms/Ga+ ion, which is only slightly smaller than the value of 2 SiO2 atoms/Ga+ ion reported in literature.28,29 Consecutively, gold-containing dots were created in the bottom of the FIB-milled holes using FEBID (Figure 1d and 2b). Two precursors were used: either dimethyl(trifluoroacetylacetonate)gold(III) (Me2Au(tfac), CAS 63470-53-1) or dimethyl(acetylacetonate)gold(III) (Me2Au(acac), CAS 14951-50-9). These commercially available precursors are known to deposit structures with reproducible composition.22,30 The Me2Au(tfac) precursor resulted in FEB deposits with a composition of gold 39.0 at. %, carbon 37.9 at. %, and oxygen 23.1 at. %. A four times lower gold content in the FEB deposit was obtained with the Me2Au(acac) precursor: gold 8.4 at. %, carbon 63.0 at. %, and oxygen 28.6 at. %. Higher gold contents (even pure gold) could be obtained using other precursors, such as the carbon-free molecule AuCl(PF3).31 However, these precursors are not commercially available, degrade relatively fast and are very sensitive to light and humidity. Of important note is that also other metal catalysts could be deposited using FEBID, especially if gold needs to be avoided due to its fast diffusion rate in silicon; for a list of molecules see a recent review of Utke et al.22 Typical process parameters for the deposition of the gold carbon containing dots with the Me2Au(tfac) or Me2Au(acac) precursor were an electron acceleration 4214

dx.doi.org/10.1021/nl2021448 |Nano Lett. 2011, 11, 4213–4217

Nano Letters

LETTER

Figure 2. SEM images of FIB cross sections performed at different fabrication steps. (a) After FIB milling of the hole. (b) After FEBID gold deposition. (c) After O2 plasma. (d) After SiNW growth. The cross sections shown are from different holes. Note, that the granular structure of the SiNW in (d) results from the cross-section preparation that was necessary to improve the image contrast; the SiNWs are actually monocrystalline, see Figure 6.

Figure 3. SEM images of an Si(100) AFM cantilever on which a single SiNW has been selectively grown at a predefined position. (a) The AFM cantilever and chip. (b) The overview of the SiNW on the very end of the AFM cantilever. (c) Zoom of the VLS-CVD grown nanowire with the gold catalyst cap. (d) Top view on the SiNW.

voltage of 30 kV and electron beam currents between 9 and 30 pA, corresponding to a fwhm of 17 and 30 nm, respectively. The exposure was performed with a stationary electron beam with deposition times between 10 and 30 s. Our experiments showed that the carbon content (up to 70 at. %) in the deposited gold dots needed to be reduced in order to obtain individual silicon nanowires in the final VLS-CVD growth step. We found two successful strategies to do this. One was to perform an O2 plasma (200W, 40 kHz, 320 cm3/min, 99.9% O2, process time: 30 to 45 min) step directly after the deposition (Figure 2c). The second was to use a He/SiH4 mixture instead of Ar/SiH4 mixture during the VLS CVD growth.30 Both purifying steps also reduced the size of the resulting pure gold dot compared to the unpurified deposit; this is favorable for the growth of small diameter SiNWs. Prior to the VLS CVD process, the SiOx protection layer could be removed, using for example an HF dip. This optional process step suppresses any parasitic SiO2 NW formation, which was, however, not very pronounced in our experiments. The final SiNW synthesis processing step was performed in a homemade CVD reactor by using the VLS method (Figure 1e and Figure 2d). Ar (5 sccm) and SiH4 (4 sccm) were introduced into the chamber that had a background pressure of 10 7 mbar. Under these conditions, the chamber pressure rose to a process pressure of 0.5 mbar. The substrate holder was heated to about 520 C. The treatment time was about 15 min, to reach the desired SiNW length of >1.5 μm. The SiNW growth was performed on two different substrates; a silicon (100) cantilever and a silicon (111) wafer. The crystallographic orientations of all cantilevers and wafers were subsequently investigated by electron back scattered detection (EBSD). Figure 3a c shows a typical example of a SiNW grown selectively on a silicon (100) AFM cantilever by using the Me2Au(tfac) precursor and an O2 plasma step prior to the VLS-CVD step. The SiNW has a diameter of 90 nm, a length of 2.1 μm, an angle of about 40 with respect to the cantilever

surface, and an angle of 23 with respect to the cantilever direction (Figure 3a d). The top view angular distribution of the SiNW on the cantilevers is shown in Figure 4a. About 43% of the SiNWs grow along the [111] direction, which is parallel to the cantilever Æ011æ axis, see the SEM images in Figure 3a,d. The other Æ111æ directions (along 180, 270, and 0) are less frequent or not pronounced. In addition, two other directions at 330 and 67 were observed, which could be attributed either to the [111] direction obtained by kinking, or to directed growth of the SiNW in the [110] or [112] direction.12,32 Figure 5 shows SiNWs grown selectively on a silicon (111) wafer by using a Me2Au(tfac) precursor or a Me2Au(acac) precursor for templating. The gold-containing catalyst dots were purified either by using an O2 plasma step prior to the VLS CVD growth or by adding He instead of Ar as carrier gas during the VLS CVD growth process. Neither the different FEBID precursors nor the different purification methods were found to have any noticeable effect on the crystallographic orientations or size of the SiNWs. Most of the SiNWs had an angle between 20 and 90 to the (111) substrate surface. The top view angular NW distribution in Figure 4b shows several growth orientations that can be attributed to Æ110æ or Æ111æ growth axes that are differing only in their angle with the substrate by 53 and 20, respectively, and to kinking, e.g., a 60 rotation on the (111) plane.12 The other orientations cannot be confidently attributed; they may result either from kinking or NWs growing in the Æ112æ direction. Growth into the Æ111æ direction seems to be less pronounced, because wires with a 90 or 20 angle to the (111) substrate surface are less often seen. It is known from the literature that the growth direction and crystalline orientation of SiNWs is not uniquely determined by the epitaxial relation with the substrate. It depends in addition on the diameter of the SiNW,33 the CVD chamber temperature, and the partial pressure of the SiH4 gas in the CVD chamber.12,32 We obtained “straight” SiNW growth (Figure 5a) on the (111) Si substrates with diameters of 80 nm and above at 520 C and 4215

dx.doi.org/10.1021/nl2021448 |Nano Lett. 2011, 11, 4213–4217

Nano Letters

Figure 4. Angular distribution of the SiNWs. (a) Grown on Si(100) cantilevers; the blue arrows indicate the projected Æ111æ growth directions. (b) Grown on Si(111) wafers; the red arrows indicate the projected Æ110æ and Æ111æ growth directions, which are differing only in the angle (53 and 20) with the substrate. The green dotted arrows indicate the projected directions obtained by kinking, e.g., 60 rotation on the (111) plane.12

167 mTorr partial SiH4 pressure. Whereas the SiNW diameters agree with those reported in the literature, the CVD parameters are higher than the optimal process parameters (480 C and 50 100 mTorr) reported in ref 12. One reason for this might be that the VLS growth is occurring on amorphous and Gaimplanted material due to the FIB milling step (see Figure 1c).22 This might also be the reason why the SiNWs often show kinking (Figure 5b) or faceting (Figure 5c) and sometimes faint multiwire growth (Figure 5d). The best conditions to produce single, straight (as close as possible orthogonal to the (111) substrate surface) SiNWs without these effects were to (a) create a deep (at least 1 μm) hole, (b) deposit a gold dot with the same diameter as the FIB hole (c) position the gold dot in the bottom center of the hole (d) perform the VLS growth with the above parameters. Another feature of this procedure was that the resulting SiNW diameter was usually about the same as the hole diameter, a fact that allows the control of the diameter of the SiNW. However, kinking and faceting could not be fully suppressed by this way. We believe that this could be significantly enhanced by improving the CVD procedure, for example, by controlling the oxygen content in the CVD chamber as suggested by Kodambaka et al.34 This is a part of ongoing work and outside the scope of this letter. With regard to crystallinity, SEM images with secondary, backscattered and transmitted electrons (SE, BSE, and STEM) of an individual SiNW grown on a Si(111) substrate were

LETTER

Figure 5. SEM images of silicon nanowires (SiNWs) grown on Si(111) wafers. The SiNWs show different growth behaviors: (a) straight growth, (b) kinking, (c) faceting, and (d) multiwire growth. The best conditions to produce single and straight SiNW growth are discussed in the text.

recorded with a Hitachi S4800 microscope (Figure 6). In the SE image the surface and some residual contaminants from the transfer and imaging process are clearly visible. In the BSE image, the gold cap and a layer around the SiNW are visible. The outer layer is clearly recognizable in the STEM image. It covers the whole SiNW and is probably either an SiO2 layer that results from the air-exposure after the CVD process35 or a carbon layer that results from the contamination during SEM observation. The SiNW is very homogeneous, only a few darker areas are visible. The darker areas at the bottom of the wire are most probably due to thickness variations. The homogeneity of the SiNW in the STEM image indicates that the grown SiNWs are single crystalline. This is in agreement with the literature, which shows that VLS grown SiNWs are always single crystalline.12,32,33,36 In conclusion, we presented a minimally invasive templating method based on a combination of gas-assisted FEBID and FIB milling. The method allows the growth of individual single crystalline silicon nanowires on nonplanar prestructured surfaces at specific, chosen positions by CVD VLS. The SiNWs produced on Si(111) wafers have several growth orientations. The SiNWs on the Si(100) cantilevers grow preferentially along the Æ111æ direction, which is parallel to the cantilever. Such SiNWs with gold caps grown on AFM cantilevers could be employed as tip-enhanced Raman spectroscopy (TERS) tips and replace the complicated nanomanipulation procedures involved when SiNWs are “glued” on AFM tips.11 Moreover, due to the parallel nature of the VLS CVD process the procedure can be employed 4216

dx.doi.org/10.1021/nl2021448 |Nano Lett. 2011, 11, 4213–4217

Nano Letters

Figure 6. Secondary electron (SE), backscattered electron (BSE) and transmission electron (TE) image of an SiNW proving its single crystalline state.

at wafer scale. The integration of the method into micro- and nanofabrication processes for applications (such as TERS, highresolution AFM tips, electron harvesting, 3D circuitry) was discussed. Given the sub-10 nm resolution of FIB milling and FEBID37,38 and the dependence of the SiNW diameter on the diameter of the hole, the method should allow the growth of SiNWs with comparable small diameters. In addition, the method could be extended to other suitable catalyst and substrate material combinations compatible with FEBID and catalystbased growth mechanisms.

’ AUTHOR INFORMATION Corresponding Author

*(I.U.) E-mail: [email protected]. Phone: +41 58 765 62 57. Fax: +41 33 228 44 90. )

Author Contributions

These two authors contributed equally to this work.

’ ACKNOWLEDGMENT This research was financially supported by both the Swiss Commission for Technology and Innovation CTI (Project No. 10710.1 PFNM-NM) and by the European Commission within the frame of the FP7 project ROD-SOL (Project Reference 227497). We thank Matthias Schamel for help with the nanomanipulation and the STEM measurements. We also thank Dr. Thomas Stelzner for helpful discussions. ’ REFERENCES (1) Nadj-Perge, S.; Frolov, S. M.; Bakkers, E. P. A. M.; Kouwenhoven, L. P. Nature 2010, 468 (7327), 1084–1087. (2) Liao, L.; Lin, Y. C.; Bao, M. Q.; Cheng, R.; Bai, J. W.; Liu, Y. A.; Qu, Y. Q.; Wang, K. L.; Huang, Y.; Duan, X. F. Nature 2010, 467 (7313), 305–308. (3) Xiang, J.; Lu, W.; Hu, Y. J.; Wu, Y.; Yan, H.; Lieber, C. M. Nature 2006, 441 (7092), 489–493. (4) Pauzauskie, P. J.; Yang, P. Mater. Today 2006, 9 (10), 36–45. (5) Tian, B. Z.; Cohen-Karni, T.; Qing, Q. A.; Duan, X. J.; Xie, P.; Lieber, C. M. Science 2010, 329 (5993), 830–834.

LETTER

(6) Zheng, G. F.; Gao, X. P. A.; Lieber, C. M. Nano Lett. 2010, 10 (8), 3179–3183. (7) Cui, Y.; Wei, Q. Q.; Park, H. K.; Lieber, C. M. Science 2001, 293 (5533), 1289–1292. (8) Hochbaum, A. I.; Chen, R. K.; Delgado, R. D.; Liang, W. J.; Garnett, E. C.; Najarian, M.; Majumdar, A.; Yang, P. D. Nature 2008, 451 (7175), 163–176. (9) Stelzner, T.; Pietsch, M.; Andra, G.; Falk, F.; Ose, E.; Christiansen, S. Nanotechnology 2008, 19 (29), 295203. (10) Lerose, D.; Bechelany, M.; Philippe, L.; Michler, J.; Christiansen, S. J. Cryst. Growth 2010, 312 (20), 2887–2891. (11) Becker, M.; Sivakov, V.; Andra, G.; Geiger, R.; Schreiber, J.; Hoffmann, S.; Michler, J.; Milenin, A. P.; Werner, P.; Christiansen, S. H. Nano Lett. 2007, 7 (1), 75–80. (12) Schmid, H.; Bjork, M. T.; Knoch, J.; Riel, H.; Riess, W.; Rice, P.; Topuria, T. J. Appl. Phys. 2008, 103 (2), 024304. (13) Cheng, C.; Lei, M.; Feng, L.; Wong, T. L.; Ho, K. M.; Fung, K. K.; Loy, M. M. T.; Yu, D. P.; Wang, N. ACS Nano 2009, 3 (1), 53–58. (14) Fan, H. J.; Lee, W.; Scholz, R.; Dadgar, A.; Krost, A.; Nielsch, K.; Zacharias, M. Nanotechnology 2005, 16 (6), 913–917. (15) Lee, S. H.; Lee, H. J.; Oh, D.; Lee, S. W.; Goto, H.; Buckmaster, R.; Yasukawa, T.; Matsue, T.; Hong, S. K.; Ko, H.; Cho, M. W.; Yao, T. F. J. Phys. Chem. B 2006, 110 (9), 3856–3859. (16) Martensson, T.; Borgstrom, M.; Seifert, W.; Ohlsson, B. J.; Samuelson, L. Nanotechnology 2003, 14 (12), 1255–1258. (17) Martensson, T.; Carlberg, P.; Borgstrom, M.; Montelius, L.; Seifert, W.; Samuelson, L. Nano Lett. 2004, 4 (4), 699–702. (18) Peng, K. Q.; Zhang, M. L.; Lu, A. J.; Wong, N. B.; Zhang, R. Q.; Lee, S. T. Appl. Phys. Lett. 2007, 90 (16), 163123. (19) Li, J. Y.; Lu, C. G.; Maynor, B.; Huang, S. M.; Liu, J. Chem. Mater. 2004, 16 (9), 1633–1636. (20) He, J. H.; Hsu, J. H.; Wang, C. W.; Lin, H. N.; Chen, L. J.; Wang, Z. L. J. Phys. Chem. B 2006, 110 (1), 50–53. (21) Guan, Y. F.; Fowlkes, J. D.; Retterer, S. T.; Simpson, M. L.; Rack, P. D. Nanotechnology 2008, 19 (50), 505302. (22) Utke, I.; Hoffmann, P.; Melngailis, J. J. Vac. Sci. Technol., B 2008, 26 (4), 1197–1276. (23) Utke, I.; Golzhauser, A. Angew. Chem., Int. Ed. 2010, 49 (49), 9328–9330. (24) Wagner, R. S.; Ellis, W. C. Appl. Phys. Lett. 1964, 4 (5), 89–90. (25) Givargizov, E. I. J. Cryst. Growth 1975, 31 (Dec), 20–30. (26) Ross, F. M. Rep. Prog. Phys. 2010, 73 (11), 114501. (27) Kallesoe, C.; Wen, C. Y.; Molhave, K.; Boggild, P.; Ross, F. M. Small 2010, 6 (18), 2058–2064. (28) Steckl, A. J.; Chyr, I. J. Vac. Sci. Technol., B 1999, 17 (2), 362–365. (29) Melngailis, J.; Musil, C. R.; Stevens, E. H.; Utlaut, M.; Kellogg, E. M.; Post, R. T.; Geis, M. W.; Mountain, R. W. J. Vac. Sci. Technol., B 1986, 4 (1), 176–180. (30) Hochleitner, G.; Steinmair, M.; Lugstein, A.; Roediger, P.; Wanzenboeck, H. D.; Bertagnolli, E. Nanotechnology 2011, 22 (1), 015302. (31) Utke, I.; Hoffmann, P.; Dwir, B.; Leifer, K.; Kapon, E.; Doppelt, P. J. Vac. Sci. Technol., B 2000, 18 (6), 3168–3171. (32) Westwater, J.; Gosain, D. P.; Tomiya, S.; Usui, S.; Ruda, H. J. Vac. Sci. Technol., B 1997, 15 (3), 554–557. (33) Schmidt, V.; Senz, S.; Gosele, U. Nano Lett. 2005, 5 (5), 931–935. (34) Kodambaka, S.; Hannon, J. B.; Tromp, R. M.; Ross, F. M. Nano Lett. 2006, 6 (6), 1292–1296. (35) Stelzner, T.; Andra, G.; Wendler, E.; Wesch, W.; Scholz, R.; Gosele, U.; Christiansen, S. Nanotechnology 2006, 17 (12), 2895–2898. (36) Cui, Y.; Lauhon, L. J.; Gudiksen, M. S.; Wang, J. F.; Lieber, C. M. Appl. Phys. Lett. 2001, 78 (15), 2214–2216. (37) Gierak, J. Semicond. Sci. Technol. 2009, 24 (4), 043001. (38) van Dorp, W. F.; van Someren, B.; Hagen, C. W.; Kruit, P. Nano Lett. 2005, 5 (7), 1303–1307.

4217

dx.doi.org/10.1021/nl2021448 |Nano Lett. 2011, 11, 4213–4217