Atomically sharp catalyst-free wurtzite GaAs/AlGaAs nanoneedles ...
APPLIED PHYSICS LETTERS 93, 023116 共2008兲
Atomically sharp catalyst-free wurtzite GaAs/ AlGaAs nanoneedles grown on silicon Michael Moewe, Linus C. Chuang, Shanna Crankshaw, Chris Chase, and Connie Chang-Hasnaina兲 Department of Electrical Engineering and Computer Sciences and Applied Science and Technology Group, University of California at Berkeley, Berkeley, California 94720, USA
共Received 16 February 2008; accepted 29 May 2008; published online 16 July 2008兲 We report a catalyst-free, self-assembled growth mode generating single-crystal wurtzite phase ultrasharp GaAs/ AlGaAs nanoneedles on both GaAs and Si substrates via low-temperature metal-organic chemical vapor deposition. The needles exhibit record-narrow tip diameters of 2 – 4 nm wide and sharp 6°–9° taper angles. The length is dependent on growth time and up to 3 – 4 m nanoneedles are attained. The structures do not exhibit twinning defects, contrary to typical GaAs nanowires grown by vapor-liquid-solid catalyzed growth. AlGaAs layered nanoneedle structures are also demonstrated. © 2008 American Institute of Physics. 关DOI: 10.1063/1.2949315兴 Nanoneedle structures with ultrasharp tips and narrow tapers1 are of great interest due to the strong2,3 electric field enhancement at the tips, which is inversely proportional to the tip diameter.4 This effect is commonly used for nonlinear optics such as tip-enhanced Raman spectroscopy.5 The enhancement is observed in metallic and semiconductor2 tips, and even single-wall carbon nanotubes.3,6,7 In this work we report the growth and characteristics of self-assembled single-crystal wurtzite GaAs nanoneedles with recordnarrow tip diameters, sharp 6°–9° taper angles and lengths up to 3 – 4 m. They grow vertically aligned on both 共111兲 Si and GaAs substrates, with bright room-temperature photoluminescence. The ultrasharp tip is a direct result of the growth mode being catalyst-free via surface deposition, contrary to the typical vapor-liquid-solid 共VLS兲 growth8 for onedimensional nanostructures. The nanoneedles are purely single-crystal wurtzite GaAs, free of twinning defects typically seen in nanowires. These nanoneedles could also prove useful for many other applications such as parallel scanningprobe microscopy devices, attoliter droplet delivery via etched core/shell nanoneedles for biological applications, or for direct integration of III-V material on Si for optoelectronic devices. The GaAs nanoneedles were grown without catalysts at low growth temperatures 共400– 420 ° C兲 on both GaAs and Si substrates, using a low-pressure metal-organic chemical vapor deposition reactor. The substrates were deoxidized and mechanically treated to initiate surface roughness to catalyze three-dimensional 共3D兲 GaAs island growth, which may be similar to 3D growth of GaN enabled by surfactants.9 The group III and V sources were triethylgallium and tertiarybutylarsine, which have relatively low decomposition temperatures 共300 and 380 ° C, respectively兲10,11 Figure 1共a兲 shows nanoneedles grown on a GaAs 共001兲 substrate. The nanoneedles are found to grow along the degenerate 具111典B orientations. We observe nanoneedles enveloping each other during growth. With increasing or decreasing the growth time, the nanoneedle length is correspondingly increased or decreased without changing its shape, angle or nanoneedle tip dimension. These observations elucidate the growth of a兲
Electronic mail: [email protected].
0003-6951/2008/93共2兲/023116/3/$23.00
nanoneedles being via continuous deposition on their initial 3D surface, favored along the 关0001兴 wurtzite crystal orientation. Typical scanning electron microscope 共SEM兲 images of nanoneedles grown on a GaAs 共111兲B substrate viewed both normal to the substrate and tilted 30° are shown in Fig. 1共b兲. The white hexagonal shapes in the first image indicate well-aligned vertical, sharp nanoneedles with a length of 2 – 3 m, which takes ⬃40 min of growth. The nanoneedle flat sidewalls align to the 具−1 − 12典 zinc blende substrate directions. Figure 1共c兲 shows a zoomed-in SEM image of a typical nanoneedle tip viewed nearly perpendicular to the growth axis. A linear array of nanoneedles is also attained on GaAs as shown in Fig. 1共d兲.
FIG. 1. 共Color online兲 SEM images of GaAs nanoneedles. 共a兲 Two needles which enveloped each other during growth, showing the sidewall and tip deposition growth mode. 共b兲 Nanoneedles grown on a GaAs 共111兲B substrate, viewed top down 共left兲 and tilted by 30° 共right兲 indicating the uniformity and alignment of the 关0001兴 nanoneedle growth axis to the 具111典B substrate directions. 共c兲 Nanoneedle viewed near to its side highlighting the facet smoothness, with the extremely sharp tip shown in the inset. 共d兲 A linear array of nanoneedles made by mechanically roughening only a very thin line. 共e兲 GaAs nanoneedle grown on 4° off-cut Si 共111兲 substrate, with views of 30° tilted and top down. The nanoneedle tilt indicates good epitaxial alignment to substrate despite the 4% lattice mismatch.
93, 023116-1
© 2008 American Institute of Physics
023116-2
Moewe et al.
FIG. 2. 共Color online兲 GaAs nanoneedle growth mode and the fabrication of hollow nanoneedles. 共a兲 The standard growth mode for the nanoneedles is via continual surface deposition, unlike VLS growth. This enables core-shell structures simply by adding additional gas precursors. 共b兲 Sonication of the AlGaAs-shell, GaAs-core nanoneedles resulted in broken tips. A selective etch was then performed to remove part of the core GaAs material to form hollow nanoneedles, demonstrating the surface-deposition growth mode. 共c兲 A planar-view SEM picture of the fabricated hollow nanoneedles. The AlGaAs cladding with thickness of ⬃80 nm is clearly seen after the removal of the GaAs core. 共d兲 A 20°-tilt view SEM picture of another fabricated hollow nanoneedle with the sonication break closer to the tip. The same thickness AlGaAs cladding can still be clearly seen.
The GaAs nanoneedles are grown on Si substrates using the same growth conditions and exhibit the same characteristics, despite the 4% lattice mismatch between GaAs and Si. Figure 1共e兲 shows 30° tilted and top-down views of a 4 m long nanoneedle grown on a 4° off-cut Si 共111兲 substrate. The lattice-mismatched GaAs nanoneedles grown on Si do not appear to be limited by the critical diameter effect seen in VLS-grown nanowires.12–14 This is probably due to the nanoneedle growth method which proceeds by sidewall and tip deposition and thus is not critically dependent on coherency to the substrate for larger nanoneedle base diameters. The typical nanoneedle density is ⬃107 / cm2 on GaAs substrates and ⬃5 ⫻ 105 / cm2 for Si substrates in the roughened areas. The growth mode of the nanoneedles is illustrated in Fig. 2共a兲. Unlike VLS growth, This growth mode facilitates the synthesis of core-shell heterostructures by simply adding additional precursors, such as nanoneedles with a GaAs core and an AlGaAs cladding layer. We grew such structures by adding trimethylaluminum for the last 80 nm of growth but keeping all other growth conditions constant. Sonication of these as-grown core-shell nanoneedles resulted in broken tips, exposing the 共0001兲 GaAs surface, as shown in Fig. 2共b兲. The sonicated sample was then selectively etched to remove the GaAs core while leaving the AlGaAs cladding intact, resulting in a hollow AlGaAs nanoneedle. Figure 2共c兲 shows some of the hollow structures fabricated by this method. The cladding is ⬃80 nm AlGaAs and the core at the broken tip is GaAs with a radius of ⬃300 nm. The etching was timed to remove ⬃150 nm of the GaAs core.15 Another etched nanoneedle is shown in Fig. 2共d兲, having the same AlGaAs wall thickness but with the sonication break closer to the tip. This further confirms the surface-deposition growth mechanism as well as the ability to grow smooth GaAs/ AlGaAs interfaces.
Appl. Phys. Lett. 93, 023116 共2008兲
FIG. 3. 共Color online兲 HRTEM images of GaAs nanoneedles. 共a兲 关1 − 100兴 zone axis HRTEM image of an as-grown GaAs nanoneedle, with a tip only 2 – 4 nm wide. The insets show the zoomed-out view, and also the image FFT. 共b兲 FFT from another nanoneedle on its 关1 − 210兴 zone axis with a distinct wurtzite pattern. The 共1 − 100兲 spacing is 3.45 Å. 共c兲 Top-down 关0001兴 TEM image of a nanoneedle. The image to the right shows a SAED pattern from the circled area, with distinct wurtzite 兵1 − 100其 spots matching the expected unique wurtzite 3.45 Å spacing. The chevron spot shape is due to electron scattering from the two sidewalls contained in the circled area.
Figure 3共a兲 shows a high resolution transmission electron microscopy 共HRTEM兲 image on the 关1 − 100兴 zone axis of a nanoneedle and its corresponding fast Fourier transform 共FFT兲. The tip in the image comes to an atomically sharp point of just 2 – 4 nm wide, which is one of the sharpest self-assembled semiconductor tips reported. The material remains single-crystal wurtzite all the way up until the tip, with no catalyst material observed. There is a surrounding 2 nm oxide layer which forms due to exposure to air. Figure 3共b兲 shows a FFT from another nanoneedle on the 关1 − 210兴 zone axis, showing the distinct wurtzite pattern, free of any zincblende phases. The c and a axes for these nanoneedles were determined to be 6.52 and 3.98 Å, respectively, within ⫾0.5%. This c / a ratio is 1.638, which is close to the ideal hexagonal c / a ratio of 1.633, and in close agreement with recent x-ray diffraction analysis of wurtzite GaAs in powder form created through high-pressure treatments.16 Figure 3共c兲 shows a TEM image from a different microscope of a nanoneedle oriented along its 关0001兴 growth axis with crystallographic directions labeled. A selected area electron diffraction 共SAED兲 pattern was recorded in the area indicated by the circle. The SAED chevron shape is due to scattering from the two nanoneedle side facets. The interplanar spacings uniquely match those of wurtzite GaAs. This 3.45 Å 兵1 − 100其 pattern spacing is distinct from the similarlooking 2.00 Å 兵2 − 20其 spacing in zinc blende 共111兲GaAs. Contrary to many reports on GaAs nanowire growth,17–19 these data show pure wurtzite phase material throughout the length of the nanoneedle. Very little is known experimentally about the wurtzite GaAs band structure.20 Hence further experiments on the large bulklike base structures of the nanoneedles could provide important experimental band structure data which could improve understanding of III-V nanowires with wurtzite
023116-3
Appl. Phys. Lett. 93, 023116 共2008兲
Moewe et al.
facilitate parallel, top-down device fabrication. TEM results indicate a single-crystal wurtzite structure free of zinc blende phases, which is rarely seen among other typical GaAs syntheses. Core-shell heterostructures and hollow structures were fabricated to exhibit the growth mechanism being surface deposition. The single nanoneedle photoluminescence was bright at both low and room temperatures. The nanoneedles are excellent candidates for many potential applications such as massively parallel scanned-probe microscopy and field emission applications, attoliter droplet injector circuits with the hollow needles, as well as to serve as a promising platform for Si-optoelectronic device integration. FIG. 4. 共Color online兲 -PL spectra of the pure GaAs nanoneedle and AlGaAs-coated GaAs nanoneedle. 共a兲 Low-temperature 共4 K兲 -PL spectrum. The pumping power is 100 W with a 2 m focused laser spot and a wavelength of 532 nm. The peak wavelength of the GaAs nanoneedle is at 1.510 eV. The AlGaAs-coated nanoneedle emits at a similar wavelength and is brighter than the pure GaAs nanoneedle by approximately a factor of 2. 共b兲 Room-temperature 共300 K兲 -PL spectrum. The pumping power is 300 W. The peak wavelength of the GaAs nanoneedle and AlGaAs-coated nanoneedle are both approximately 1.425 eV. The AlGaAs-coated nanoneedle is brighter than the GaAs nanoneedle by a factor of 4. The PL linewidths of the pure GaAs and AlGaAs-coated nanoneedles are 50 and 60 meV, respectively.
phase sections. The optical properties of single, as-grown nanoneedles grown on Si were characterized using microphotoluminescence 共-PL兲 with a 2 m focused laser spot and a wavelength of 532 nm. Figure 4共a兲 shows lowtemperature 共4 K兲 -PL spectra for a GaAs nanoneedle with a 300 nm base radius and a similarly sized GaAs nanoneedle with a 25 nm AlGaAs cladding layer. The peak wavelength of the GaAs nanoneedle is at 1.510 eV, which is redshifted by only 10 meV from zinc blende GaAs at 1.520 eV.21 No quantization effects are expected due to the large size of the nanoneedles. Hence, these results represent a reliable measurement of the bandgap of twin-free bulk wurtzite GaAs. The AlGaAs-coated nanoneedle emits at a similar wavelength and is brighter than the pure GaAs nanoneedle by approximately a factor of two at 4 K. This is not surprising since the AlGaAs layer can effectively passivate the GaAs surface. The room-temperature 共300 K兲 spectra of the same nanoneedles are shown in Fig. 4共b兲. The peak wavelength of the GaAs nanoneedle and AlGaAs-coated nanoneedle are both approximately 1.425 eV, very close to that of undoped, bulk zinc blende GaAs at 1.430 eV.21 The AlGaAs-coated nanoneedle is brighter than the GaAs nanoneedle by a factor of 4 at room temperature. To summarize, we report the growth, structural and optical properties of atomically sharp, single-crystal wurtzite GaAs/ AlGaAs nanoneedles. This and the growth mode both
This work was supported by DARPA HR0011-04-10040 共CONSRT兲 and HP-CITRIS grants. The authors acknowledge the fellowship support from the UC Berkeley EECS Fellowship Program, NSF-IGERT Program, and NSF Graduate Research Fellowship Program. 1
R. C. Smith, J. D. Carey, R. D. Forrest, and S. R. P. Silva, J. Vac. Sci. Technol. B 23, 632 共2005兲. 2 Y. B. Tang, H. T. Cong, Z. M. Wang, and H.-M. Cheng, Appl. Phys. Lett. 89, 253112 共2006兲. 3 J. Y. Huang, K. Kempa, S. H. Jo, S. Chen, and Z. F. Ren, Appl. Phys. Lett. 89, 053110 共2005兲. 4 G. T. Boyd, Th. Rasing, J. R. R. Leite, and Y. R. Shen, Phys. Rev. B 30, 519 共1984兲. 5 R. M. Stöckle, Y. D. Suh, V. Deckert, and R. Zenobi, Chem. Phys. Lett. 318, 131 共2000兲. 6 Y. Saito, Y. Tsujimoto, A. Koshio, and F. Kokai, Appl. Phys. Lett. 90, 213108 共2007兲. 7 K. A. Dean and B. R. Chalamala, J. Appl. Phys. 85, 3832 共1999兲. 8 R. S. Wagner and W. C. Ellis, Appl. Phys. Lett. 4, 89 共1964兲. 9 T. Satoru, S. Iwai, and Y. Aoyagi, Appl. Phys. Lett. 69, 4096 共1996兲. 10 P. W. Lee, T. R. Omstead, D. R. McKenna, and K. F. Jensen, J. Cryst. Growth 85, 165 共1987兲. 11 C. A. Larsen, N. I. Buchan, S. H. Li, and G. B. Stringfellow, J. Cryst. Growth 94, 663 共1989兲. 12 L. C. Chuang, M. Moewe, S. Crankshaw, and C. Chang-Hasnain, Appl. Phys. Lett. 92, 013121 共2008兲. 13 E. Ertekin, P. A. Greaney, D. C. Chrzan, and T. D. Sands, J. Appl. Phys. 97, 114325 共2005兲. 14 F. Glas, Phys. Rev. B 74, 121302共R兲 共2006兲. 15 M. C. Y. Huang, K. B. Cheng, Y. Zhou, B. Pesala, and C. J. ChangHasnain, IEEE Photonics Technol. Lett. 18, 1197 共2006兲. 16 M. I. McMahon and R. J. Nelmes, Phys. Rev. Lett. 95, 215505 共2005兲. 17 R. Banerjee, A. Bhattacharya, A. Genc, and B. M. Arora, Philos. Mag. Lett. 86, 807 共2006兲. 18 K. Hiruma, M. Yazawa, T. Katsuyama, K. Ogawa, K. Haraguchi, M. Koguchi, and H. Kakibayashi, J. Appl. Phys. 77, 447 共1995兲. 19 J. Johansson, L. S. Karlsson, C. P. T. Svensson, T. Mårtensson, B. A. Wacaser, K. Deppert, L. Samuelson, and Werner Seifert, Nat. Mater. 5, 574 共2006兲. 20 Z. Zanolli, F. Fuchs, J. Furthmüller, U. von Barth, and F. Bechstedt, Phys. Rev. B 75, 245121 共2006兲. 21 I. Vurgaftman, J. R. Meyer, and L. R. Ram-Mohan, J. Appl. Phys. 89, 5815 共2001兲.
Atomically sharp catalyst-free wurtzite GaAs/ AlGaAs nanoneedles grown on silicon Michael Moewe, Linus C. Chuang, Shanna Crankshaw, Chris Chase, and Connie Chang-Hasnaina兲 Department of Electrical Engineering and Computer Sciences and Applied Science and Technology Group, University of California at Berkeley, Berkeley, California 94720, USA
共Received 16 February 2008; accepted 29 May 2008; published online 16 July 2008兲 We report a catalyst-free, self-assembled growth mode generating single-crystal wurtzite phase ultrasharp GaAs/ AlGaAs nanoneedles on both GaAs and Si substrates via low-temperature metal-organic chemical vapor deposition. The needles exhibit record-narrow tip diameters of 2 – 4 nm wide and sharp 6°–9° taper angles. The length is dependent on growth time and up to 3 – 4 m nanoneedles are attained. The structures do not exhibit twinning defects, contrary to typical GaAs nanowires grown by vapor-liquid-solid catalyzed growth. AlGaAs layered nanoneedle structures are also demonstrated. © 2008 American Institute of Physics. 关DOI: 10.1063/1.2949315兴 Nanoneedle structures with ultrasharp tips and narrow tapers1 are of great interest due to the strong2,3 electric field enhancement at the tips, which is inversely proportional to the tip diameter.4 This effect is commonly used for nonlinear optics such as tip-enhanced Raman spectroscopy.5 The enhancement is observed in metallic and semiconductor2 tips, and even single-wall carbon nanotubes.3,6,7 In this work we report the growth and characteristics of self-assembled single-crystal wurtzite GaAs nanoneedles with recordnarrow tip diameters, sharp 6°–9° taper angles and lengths up to 3 – 4 m. They grow vertically aligned on both 共111兲 Si and GaAs substrates, with bright room-temperature photoluminescence. The ultrasharp tip is a direct result of the growth mode being catalyst-free via surface deposition, contrary to the typical vapor-liquid-solid 共VLS兲 growth8 for onedimensional nanostructures. The nanoneedles are purely single-crystal wurtzite GaAs, free of twinning defects typically seen in nanowires. These nanoneedles could also prove useful for many other applications such as parallel scanningprobe microscopy devices, attoliter droplet delivery via etched core/shell nanoneedles for biological applications, or for direct integration of III-V material on Si for optoelectronic devices. The GaAs nanoneedles were grown without catalysts at low growth temperatures 共400– 420 ° C兲 on both GaAs and Si substrates, using a low-pressure metal-organic chemical vapor deposition reactor. The substrates were deoxidized and mechanically treated to initiate surface roughness to catalyze three-dimensional 共3D兲 GaAs island growth, which may be similar to 3D growth of GaN enabled by surfactants.9 The group III and V sources were triethylgallium and tertiarybutylarsine, which have relatively low decomposition temperatures 共300 and 380 ° C, respectively兲10,11 Figure 1共a兲 shows nanoneedles grown on a GaAs 共001兲 substrate. The nanoneedles are found to grow along the degenerate 具111典B orientations. We observe nanoneedles enveloping each other during growth. With increasing or decreasing the growth time, the nanoneedle length is correspondingly increased or decreased without changing its shape, angle or nanoneedle tip dimension. These observations elucidate the growth of a兲
Electronic mail: [email protected].
0003-6951/2008/93共2兲/023116/3/$23.00
nanoneedles being via continuous deposition on their initial 3D surface, favored along the 关0001兴 wurtzite crystal orientation. Typical scanning electron microscope 共SEM兲 images of nanoneedles grown on a GaAs 共111兲B substrate viewed both normal to the substrate and tilted 30° are shown in Fig. 1共b兲. The white hexagonal shapes in the first image indicate well-aligned vertical, sharp nanoneedles with a length of 2 – 3 m, which takes ⬃40 min of growth. The nanoneedle flat sidewalls align to the 具−1 − 12典 zinc blende substrate directions. Figure 1共c兲 shows a zoomed-in SEM image of a typical nanoneedle tip viewed nearly perpendicular to the growth axis. A linear array of nanoneedles is also attained on GaAs as shown in Fig. 1共d兲.
FIG. 1. 共Color online兲 SEM images of GaAs nanoneedles. 共a兲 Two needles which enveloped each other during growth, showing the sidewall and tip deposition growth mode. 共b兲 Nanoneedles grown on a GaAs 共111兲B substrate, viewed top down 共left兲 and tilted by 30° 共right兲 indicating the uniformity and alignment of the 关0001兴 nanoneedle growth axis to the 具111典B substrate directions. 共c兲 Nanoneedle viewed near to its side highlighting the facet smoothness, with the extremely sharp tip shown in the inset. 共d兲 A linear array of nanoneedles made by mechanically roughening only a very thin line. 共e兲 GaAs nanoneedle grown on 4° off-cut Si 共111兲 substrate, with views of 30° tilted and top down. The nanoneedle tilt indicates good epitaxial alignment to substrate despite the 4% lattice mismatch.
93, 023116-1
© 2008 American Institute of Physics
023116-2
Moewe et al.
FIG. 2. 共Color online兲 GaAs nanoneedle growth mode and the fabrication of hollow nanoneedles. 共a兲 The standard growth mode for the nanoneedles is via continual surface deposition, unlike VLS growth. This enables core-shell structures simply by adding additional gas precursors. 共b兲 Sonication of the AlGaAs-shell, GaAs-core nanoneedles resulted in broken tips. A selective etch was then performed to remove part of the core GaAs material to form hollow nanoneedles, demonstrating the surface-deposition growth mode. 共c兲 A planar-view SEM picture of the fabricated hollow nanoneedles. The AlGaAs cladding with thickness of ⬃80 nm is clearly seen after the removal of the GaAs core. 共d兲 A 20°-tilt view SEM picture of another fabricated hollow nanoneedle with the sonication break closer to the tip. The same thickness AlGaAs cladding can still be clearly seen.
The GaAs nanoneedles are grown on Si substrates using the same growth conditions and exhibit the same characteristics, despite the 4% lattice mismatch between GaAs and Si. Figure 1共e兲 shows 30° tilted and top-down views of a 4 m long nanoneedle grown on a 4° off-cut Si 共111兲 substrate. The lattice-mismatched GaAs nanoneedles grown on Si do not appear to be limited by the critical diameter effect seen in VLS-grown nanowires.12–14 This is probably due to the nanoneedle growth method which proceeds by sidewall and tip deposition and thus is not critically dependent on coherency to the substrate for larger nanoneedle base diameters. The typical nanoneedle density is ⬃107 / cm2 on GaAs substrates and ⬃5 ⫻ 105 / cm2 for Si substrates in the roughened areas. The growth mode of the nanoneedles is illustrated in Fig. 2共a兲. Unlike VLS growth, This growth mode facilitates the synthesis of core-shell heterostructures by simply adding additional precursors, such as nanoneedles with a GaAs core and an AlGaAs cladding layer. We grew such structures by adding trimethylaluminum for the last 80 nm of growth but keeping all other growth conditions constant. Sonication of these as-grown core-shell nanoneedles resulted in broken tips, exposing the 共0001兲 GaAs surface, as shown in Fig. 2共b兲. The sonicated sample was then selectively etched to remove the GaAs core while leaving the AlGaAs cladding intact, resulting in a hollow AlGaAs nanoneedle. Figure 2共c兲 shows some of the hollow structures fabricated by this method. The cladding is ⬃80 nm AlGaAs and the core at the broken tip is GaAs with a radius of ⬃300 nm. The etching was timed to remove ⬃150 nm of the GaAs core.15 Another etched nanoneedle is shown in Fig. 2共d兲, having the same AlGaAs wall thickness but with the sonication break closer to the tip. This further confirms the surface-deposition growth mechanism as well as the ability to grow smooth GaAs/ AlGaAs interfaces.
Appl. Phys. Lett. 93, 023116 共2008兲
FIG. 3. 共Color online兲 HRTEM images of GaAs nanoneedles. 共a兲 关1 − 100兴 zone axis HRTEM image of an as-grown GaAs nanoneedle, with a tip only 2 – 4 nm wide. The insets show the zoomed-out view, and also the image FFT. 共b兲 FFT from another nanoneedle on its 关1 − 210兴 zone axis with a distinct wurtzite pattern. The 共1 − 100兲 spacing is 3.45 Å. 共c兲 Top-down 关0001兴 TEM image of a nanoneedle. The image to the right shows a SAED pattern from the circled area, with distinct wurtzite 兵1 − 100其 spots matching the expected unique wurtzite 3.45 Å spacing. The chevron spot shape is due to electron scattering from the two sidewalls contained in the circled area.
Figure 3共a兲 shows a high resolution transmission electron microscopy 共HRTEM兲 image on the 关1 − 100兴 zone axis of a nanoneedle and its corresponding fast Fourier transform 共FFT兲. The tip in the image comes to an atomically sharp point of just 2 – 4 nm wide, which is one of the sharpest self-assembled semiconductor tips reported. The material remains single-crystal wurtzite all the way up until the tip, with no catalyst material observed. There is a surrounding 2 nm oxide layer which forms due to exposure to air. Figure 3共b兲 shows a FFT from another nanoneedle on the 关1 − 210兴 zone axis, showing the distinct wurtzite pattern, free of any zincblende phases. The c and a axes for these nanoneedles were determined to be 6.52 and 3.98 Å, respectively, within ⫾0.5%. This c / a ratio is 1.638, which is close to the ideal hexagonal c / a ratio of 1.633, and in close agreement with recent x-ray diffraction analysis of wurtzite GaAs in powder form created through high-pressure treatments.16 Figure 3共c兲 shows a TEM image from a different microscope of a nanoneedle oriented along its 关0001兴 growth axis with crystallographic directions labeled. A selected area electron diffraction 共SAED兲 pattern was recorded in the area indicated by the circle. The SAED chevron shape is due to scattering from the two nanoneedle side facets. The interplanar spacings uniquely match those of wurtzite GaAs. This 3.45 Å 兵1 − 100其 pattern spacing is distinct from the similarlooking 2.00 Å 兵2 − 20其 spacing in zinc blende 共111兲GaAs. Contrary to many reports on GaAs nanowire growth,17–19 these data show pure wurtzite phase material throughout the length of the nanoneedle. Very little is known experimentally about the wurtzite GaAs band structure.20 Hence further experiments on the large bulklike base structures of the nanoneedles could provide important experimental band structure data which could improve understanding of III-V nanowires with wurtzite
023116-3
Appl. Phys. Lett. 93, 023116 共2008兲
Moewe et al.
facilitate parallel, top-down device fabrication. TEM results indicate a single-crystal wurtzite structure free of zinc blende phases, which is rarely seen among other typical GaAs syntheses. Core-shell heterostructures and hollow structures were fabricated to exhibit the growth mechanism being surface deposition. The single nanoneedle photoluminescence was bright at both low and room temperatures. The nanoneedles are excellent candidates for many potential applications such as massively parallel scanned-probe microscopy and field emission applications, attoliter droplet injector circuits with the hollow needles, as well as to serve as a promising platform for Si-optoelectronic device integration. FIG. 4. 共Color online兲 -PL spectra of the pure GaAs nanoneedle and AlGaAs-coated GaAs nanoneedle. 共a兲 Low-temperature 共4 K兲 -PL spectrum. The pumping power is 100 W with a 2 m focused laser spot and a wavelength of 532 nm. The peak wavelength of the GaAs nanoneedle is at 1.510 eV. The AlGaAs-coated nanoneedle emits at a similar wavelength and is brighter than the pure GaAs nanoneedle by approximately a factor of 2. 共b兲 Room-temperature 共300 K兲 -PL spectrum. The pumping power is 300 W. The peak wavelength of the GaAs nanoneedle and AlGaAs-coated nanoneedle are both approximately 1.425 eV. The AlGaAs-coated nanoneedle is brighter than the GaAs nanoneedle by a factor of 4. The PL linewidths of the pure GaAs and AlGaAs-coated nanoneedles are 50 and 60 meV, respectively.
phase sections. The optical properties of single, as-grown nanoneedles grown on Si were characterized using microphotoluminescence 共-PL兲 with a 2 m focused laser spot and a wavelength of 532 nm. Figure 4共a兲 shows lowtemperature 共4 K兲 -PL spectra for a GaAs nanoneedle with a 300 nm base radius and a similarly sized GaAs nanoneedle with a 25 nm AlGaAs cladding layer. The peak wavelength of the GaAs nanoneedle is at 1.510 eV, which is redshifted by only 10 meV from zinc blende GaAs at 1.520 eV.21 No quantization effects are expected due to the large size of the nanoneedles. Hence, these results represent a reliable measurement of the bandgap of twin-free bulk wurtzite GaAs. The AlGaAs-coated nanoneedle emits at a similar wavelength and is brighter than the pure GaAs nanoneedle by approximately a factor of two at 4 K. This is not surprising since the AlGaAs layer can effectively passivate the GaAs surface. The room-temperature 共300 K兲 spectra of the same nanoneedles are shown in Fig. 4共b兲. The peak wavelength of the GaAs nanoneedle and AlGaAs-coated nanoneedle are both approximately 1.425 eV, very close to that of undoped, bulk zinc blende GaAs at 1.430 eV.21 The AlGaAs-coated nanoneedle is brighter than the GaAs nanoneedle by a factor of 4 at room temperature. To summarize, we report the growth, structural and optical properties of atomically sharp, single-crystal wurtzite GaAs/ AlGaAs nanoneedles. This and the growth mode both
This work was supported by DARPA HR0011-04-10040 共CONSRT兲 and HP-CITRIS grants. The authors acknowledge the fellowship support from the UC Berkeley EECS Fellowship Program, NSF-IGERT Program, and NSF Graduate Research Fellowship Program. 1
R. C. Smith, J. D. Carey, R. D. Forrest, and S. R. P. Silva, J. Vac. Sci. Technol. B 23, 632 共2005兲. 2 Y. B. Tang, H. T. Cong, Z. M. Wang, and H.-M. Cheng, Appl. Phys. Lett. 89, 253112 共2006兲. 3 J. Y. Huang, K. Kempa, S. H. Jo, S. Chen, and Z. F. Ren, Appl. Phys. Lett. 89, 053110 共2005兲. 4 G. T. Boyd, Th. Rasing, J. R. R. Leite, and Y. R. Shen, Phys. Rev. B 30, 519 共1984兲. 5 R. M. Stöckle, Y. D. Suh, V. Deckert, and R. Zenobi, Chem. Phys. Lett. 318, 131 共2000兲. 6 Y. Saito, Y. Tsujimoto, A. Koshio, and F. Kokai, Appl. Phys. Lett. 90, 213108 共2007兲. 7 K. A. Dean and B. R. Chalamala, J. Appl. Phys. 85, 3832 共1999兲. 8 R. S. Wagner and W. C. Ellis, Appl. Phys. Lett. 4, 89 共1964兲. 9 T. Satoru, S. Iwai, and Y. Aoyagi, Appl. Phys. Lett. 69, 4096 共1996兲. 10 P. W. Lee, T. R. Omstead, D. R. McKenna, and K. F. Jensen, J. Cryst. Growth 85, 165 共1987兲. 11 C. A. Larsen, N. I. Buchan, S. H. Li, and G. B. Stringfellow, J. Cryst. Growth 94, 663 共1989兲. 12 L. C. Chuang, M. Moewe, S. Crankshaw, and C. Chang-Hasnain, Appl. Phys. Lett. 92, 013121 共2008兲. 13 E. Ertekin, P. A. Greaney, D. C. Chrzan, and T. D. Sands, J. Appl. Phys. 97, 114325 共2005兲. 14 F. Glas, Phys. Rev. B 74, 121302共R兲 共2006兲. 15 M. C. Y. Huang, K. B. Cheng, Y. Zhou, B. Pesala, and C. J. ChangHasnain, IEEE Photonics Technol. Lett. 18, 1197 共2006兲. 16 M. I. McMahon and R. J. Nelmes, Phys. Rev. Lett. 95, 215505 共2005兲. 17 R. Banerjee, A. Bhattacharya, A. Genc, and B. M. Arora, Philos. Mag. Lett. 86, 807 共2006兲. 18 K. Hiruma, M. Yazawa, T. Katsuyama, K. Ogawa, K. Haraguchi, M. Koguchi, and H. Kakibayashi, J. Appl. Phys. 77, 447 共1995兲. 19 J. Johansson, L. S. Karlsson, C. P. T. Svensson, T. Mårtensson, B. A. Wacaser, K. Deppert, L. Samuelson, and Werner Seifert, Nat. Mater. 5, 574 共2006兲. 20 Z. Zanolli, F. Fuchs, J. Furthmüller, U. von Barth, and F. Bechstedt, Phys. Rev. B 75, 245121 共2006兲. 21 I. Vurgaftman, J. R. Meyer, and L. R. Ram-Mohan, J. Appl. Phys. 89, 5815 共2001兲.
