AlGaAs Core Multishell Nanowire-Based Light-Emitting Diodes

Apr 8, 2010 - Hyunseok Kim , Wook-Jae Lee , Alan C. Farrell , Juan S. D. Morales , Pradeep Senanayake , Sergey V. Prikhodko , Tomasz J. Ochalski , and...
14 downloads 13 Views 2MB Size
Download PDF
pubs.acs.org/NanoLett

GaAs/AlGaAs Core Multishell Nanowire-Based Light-Emitting Diodes on Si Katsuhiro Tomioka,*,†,‡,§ Junichi Motohisa,† Shinjiroh Hara,‡ Kenji Hiruma,‡ and Takashi Fukui*,†,‡ †

Graduate School of Information Science and Technology, Hokkaido University, North 14 West 9, Sapporo 060-0814, Japan, ‡ Research Center for Integrated Quantum Electronics (RCIQE), Hokkaido University, North 13 West 8, Sapporo 060-8628, Japan, and § PRESTO, Japan Science and Technology Agency (JST), 4-1-8 Honcho Kawaguchi, Saitama 332-0012, Japan ABSTRACT We report on integration of GaAs nanowire-based light-emitting-diodes (NW-LEDs) on Si substrate by selective-area metalorganic vapor phase epitaxy. The vertically aligned GaAs/AlGaAs core-multishell nanowires with radial p-n junction and NWLED array were directly fabricated on Si. The threshold current for electroluminescence (EL) was 0.5 mA (current density was approximately 0.4 A/cm2), and the EL intensity superlinearly increased with increasing current injections indicating superluminescence behavior. The technology described in this letter could help open new possibilities for monolithic- and on-chip integration of III-V NWs on Si. KEYWORDS Nanowire, core-shell, doping, III-V on Si, LED

M

onolithic integration of electronic and optical devices on Si has been attracting much interest for use in next-generation technology. Rapid progress has been made in the area of Si photonics,1 which are photonics made with Si-based materials.2-4 The other major approach is the integration of III-V compound semiconductor photonic devices on Si.5-7 This approach has been a challenge for a long time because of the difficulties of mismatches in lattice constant and thermal coefficient for heterepitaxy of III-Vs on Si. Such difficulties, however, can be overcome by III-V compound semiconductor nanowires (NWs) because of their nanometer-scaled footprint, and these NWs have attracted much attention as building blocks for fusing electronic and photonic functions of NWs8-12 on Si chips. Challenges have been met in integrating the III-V NW-based electron13-15 and optical devices16,17 on a Si wafer in a monolithic complementary metal oxide semiconductor (CMOS); compatible processes have been made using advanced epitaxial techniques with the III-V NWs on Si.18-27 Apart from the lattice mismatching, growth of polar materials on nonpolar substrates was an issue in the heteroepitaxy of III-V NWs on Si. It has also been a challenge to control the growth directions of III-V NWs on Si, which is important for device applications. Since typical III-V NWs grow preferentially in the [111]B direction (or [111]A direction), III-V NWs normal to the substrate can be formed on (111)B [or (111)A]-oriented surfaces. On the other hand, the Si(111) surface has a nonpolar nature and four equivalent

{111} planes occurs on its surface. Thus, III-V NWs might grow in those four equivalent 〈111〉 directions on the Si(111) surface; one is normal to the (111) surface and the others are 19.5°-tilted 〈111〉 directions.9,14,17 These tilted directions make the processing for device fabrication difficult and the geometric advantage of NWs is lost in terms of high density integrations. We have overcome this problem by using a specific Si(111) surface pretreatment in selective-area MOVPE (SA-MOVPE),17,21 and we succeeded in making the vertically aligned and position-controlled III-V NWs on Si substrates. One application of the III-V NWs on Si is nanoscale light sources and detectors for on-chip integration, which can replace Cu-based intrachip connections by optical interconnections with high performance in a small area. In particular, the large surface-to-volume ratio of the radial p-n junctions in core-multishell (CMS) NWs can simply enlarge the junction area as compared to that of planar substrate with the same surface area.28,29 The vertical CMS NWs with radial p-n junctions are, therefore, desirable because of their areaeffectiveness and because they can enhance performance in light-emitting-diodes (LEDs), photodetectors, and solar cells. LEDs based on CMS NWs have been fabricated in wide bandgap semiconductors.30-33 The NW-based light sources, which operated at wavelength from 800 to 900 nm, are feasible for on-chip integrations with Si-photodiodes (PDs) and avalanche PDs system. However, there have been little investigation of III-V NWs-based LEDs on Si for the nearinfrared region.8,34,35 In this letter, we report on integration of GaAs/AlGaAs CMS NWs-based LEDs on Si substrate. To enhance efficiencies in the NW-based LED, we have designed and grown double-heterostructure (DH) CMS NWs on Si as shown in Figure 1. The structure consists of an

* To whom correspondence should be addressed. E-mail: (K.T.) tomioka@ rciqe.hokudai.ac.jp; (T.F.) [email protected]. Received for review: 12/17/2009 Published on Web: 04/08/2010 © 2010 American Chemical Society

1639

DOI: 10.1021/nl9041774 | Nano Lett. 2010, 10, 1639–1644

FIGURE 1. (a) 30°-tilted view SEM image of CMS-NWs. (b) Illustration of CMS NWs (c) SEM image of a CMS NW cross-section cut following dashed line (1-1′) in (b).

n-type GaAs NW as a core and n-AlGaAs, p-GaAs, p-AlGaAs, and p-GaAs shells for the innermost to outermost shells, which are sequentially grown on the sidewall of the GaAs core NW. The n-type and p-type AlGaAs layers are cladding layers for confinements of electrons and holes in the inner p-GaAs layer and also for photons generated in the p-GaAs layers. The p-GaAs wedged between the n- and p-AlGaAs layers is quantum well (QW) tube, while, the outer p-GaAs is a capping layer for Ohmic contacts. In the experiment, n-Si(111) (F < 0.02 Ωcm) substrates were used as a starting materials. The process of selectivearea growth has been previously reported.21,26 First, 20 nm thick SiO2 was formed using thermal oxidation process to make a mask of SA-MOVPE growth. Next, the periodic openings were formed in the mask using electron beam lithography and wet chemical etching. The opening diameter, d0, was 100 nm. Before the partially masked substrate was loaded into the MOVPE reactor, it was degreased with organic solvents in an ultrasonic bath and slightly etched with buffered HF (BHF) solution to remove native oxides formed on the opening. Prior to the growth, thermal cleaning in H2 ambient was carried out at 900 °C in the reactor to further ensure the removal of the native oxide. The NWs were grown using a low-pressure (0.1 atm) MOVPE system. The carrier gas was pure hydrogen (H2). Trimethylgallium (TMGa), trimethylalminum (TMAl), and arsine (AsH3) were used for the III-V growths. Silane (SiH4) and diethylzinc (DEZn) were used as n-type and p-type dopants. The vertical GaAs NWs were grown at 750 °C for 60 min. The partial pressures of TMGa and AsH3 for the NW growth were 8.2 × 10-7 and 2.5 × 10-4 atm. After the growth of GaAs NWs, the n- and p-AlGaAs shell layers were formed at 700 °C for 5 min in each. The partial pressures of TMAl, TMGa, and AsH3 were 1.2 × 10-6, 8.2 × 10-7, and 1.3 × 10-4 atm. The partial pressures of SiH4 and DEZn were 2.5 × 10-8 and 2.8 × 10-6 atm for the n- and p-type layers, respectively. Also, the growths of p-GaAs well layer and capping layer were performed at 700 °C for 3 min in each. Nominal carrier concentrations of planar GaAs and p-GaAs were 3.5 × 1017 cm-3 and 4.0 × 1018 cm-3. The donor and acceptor concentrations, ND and NA, of planar n-AlGaAs and p-AlGaAs were 9.0 × 1017 and 1.0 × 1018 cm-3. The characterization of the doping level for each layers of the CMS-NWs is slightly difficult. There is a possibility that these carrier concentrations for the CMS-NWs were lower than © 2010 American Chemical Society

those for the planar epitaxial layers as mentioned in previous report regarding core-shell InP NW solar cell.36 Photoluminescence (PL) measurement was carried out at 4.2 K. The excitation light was He-Ne (632.8 nm) focused to a 2 µm spot size using a 100× objective lens. Electroluminescence (EL) measurement was performed with the same set up at room temperature using DC bias. PL and EL spectra were taken with a CCD detector. Figure 1a shows the growth results of the DH structure CMS NW array on Si. Vertically aligned regular hexagonal NWs with diameter, d, of about 220 nm were fabricated on Si substrate. Unintentional kink and taper formations of the NWs resulted from high doping are not observed in these images. Thus, each NW had uniform shell layer thickness. The average heights were 3 µm. Figure 1b is an illustration of the CMS NW structure and Figure 1c shows an SEM image taken from a cleaved and selectively etched CMS NW. The selective etching was carried out for 2 s with 1 NH4OH/1 H2O2/10 H2O solution. In this image, the lateral thicknesses of both n- and p-AlGaAs layers are 25 nm, and the lateral growth rate was estimated to be 2.5 nm/min. Figure 2a shows the PL spectra of the CMS-NWs at 4.2K under various excitation power densities. Curve fitting with the Gaussian functions for Figure 2b shows two main peaks at 1.654 eV (P1) with a full width at half-maximum (fwhm) of 74 meV and 1.598 eV (P2) with fwhm of 72 meV. Also, additional sharp PL peak resulting from GaAs QW layer is observed at 1.574 eV (P3) with fwhm of 8 meV. In this deconvolution, we assumed two numbers of Gaussian line shape for PL band of AlGaAs layer, a Gaussian line shape for the GaAs QW layers, and four numbers of Gaussian line shape for the GaAs NWs. The deconvolution was made by means of a least-squares peak. Also, we fixed only the number of the line shape with varying their peak positions and fwhm. Individual contributions of these peaks are represented by the dashed lines in Figure 2b. It is evident from Figure 2b that the calculated spectrum showed relatively good agreement with the experimental data. The excitation power dependence of each luminescent peak in Figure 2c indicates that the peaks P1 and P2 were blue shifted with increased excitation power density, and P3 was still constant with inclement of the excitation powers. In this case, we used the same numbers of Gaussian line shape for the deconvolution of each PL spectra. The blue shifts of the peak position estimated from the deconvoluted P1 and P2 1640

DOI: 10.1021/nl9041774 | Nano Lett. 2010, 10, 1639-–1644

FIGURE 2. (a) PL spectra for the CMS NWs with variation of excitation power density. (b) Curve fitting of typical PL spectrum. The excitation power density was 0.1 kW/cm2. Opened white circles show experimental data, dashed lines are fitted curves for GaAs NW. Red (P1), blue (P2), and black (P3) solid lines are for shell layers. The dashed lines show Gaussian line shape from peak deconvolution. (c) Excitation power dependence of PL peak positions for P1 (closed red circles), P2 (closed blue circles), and P3 (closed black circles).

peaks are 11 and 18 meV/dec Note that each plots in Figure 2c were calculated from the deconvolution for each PL spectra same as Figure 2b. The value of the energy shift for P1 corresponds to that of a Zn-related donor-acceptor pair (DAP) transition37 and P2 is thought to be the Si-C related DAP transition in the AlGaAs cladding layer.38 Although the values of the peak shift of P2 are slightly higher than those previously reported, these shift reasonably explain the value when the fluctuation of the carrier concentration for individual NWs in the NW ensemble is considered. The PL intensities of P1 and P2 vary almost linearly with increased the excitation power density (see Supporting Information, Figure S1). When the transition occurs via deep luminescent centers such as DX or vacancy-related centers, the emission intensity shows sublinear dependence on the excitation power.37 Thus, the P1 and P2 are thought to be DAP luminescence. The Al composition in the AlGaAs shell is estimated at 12%, taking into account the donor and acceptor levels. The PL from the GaAs quantum well (QW) is clearly observed at 1.574 eV for weak excitation, as shown in the inset of Figure 2b. The fwhm from the peak deconvolution for the PL is 8 meV. This narrow line width reflects exciton emission from the GaAs QW. The P3 hindered in the other intense P1 and P2 peaks under strong excitation powers as shown in Figure 2a. The PL deconvolution suggested that the fwhm of the P3 increased with the inclement of the excitation power (see Supporting Information, Figure S2). This is because the free electron-hole pair PL began to appear at higher excitation power density.39 Moreover, the peak position of P3 was constant for the variation of the excitation power. From the PL peak energy, thickness of the GaAs QW is estimated to be 7 nm. The estimation was performed using Schro¨dinger’s equation with simple rectangularshaped well. Also, we used that the ratio of conduction and valence band offset in GaAs/AlGaAs system is 60:40 for the calculation. Next, LEDs were fabricated using the DH CMS NWs on Si. First, SiO2 (50 nm)/Al2O3 (20 nm) were deposited on the NW surface to avoid reactive-ion-etching (RIE)-induced dam© 2010 American Chemical Society

age as well as to make slit around the NW sidewalls. The SiO2 was deposited by RF-sputtering and the Al2O3 was formed using atomic-layer deposition (ALD). Next, the sample was buried with poly resin and etched with RIE to expose the top parts of the NWs for metal contacts. After the RIE, SiO2/Al2O3 was removed by BHF solution and 70 nm wide slit was formed around the CMS-NWs. The depth of the slit was 1.5 µm. Then, Cr/Au metal was deposited onto the top parts of the RIE-etched NWs. In the metal deposition, we used the rotating sample holder to effectively coat the metals on the sidewalls of the CMS NWs. The rotation per minute was 50. Also, the wafer was tilted to about 5°. After deposition of the Cr/Au metal, the top parts of the NWs were mechanically ground to make wrapped metal contacts on NW sidewalls only (The process are illustrated in Supporting Information, Figure S3). We used Ti(20 nm)/Au(100 nm) metal and silver for backside electrode. Figure 3a illustrates the NW-based LED structure. Two types of CMS-NWs with different Cr/Au thicknesses were fabricated to investigate electronic properties: one is Cr(10 nm)/Au(130 nm) with 25 nm thick n- and p-AlGaAs shell layer (Sample A), and the other is Cr(2 nm)/Au(3 nm) semitransparent metal with 25 nm thick n- and p-AlGaAs shell layer (Sample B). The latter sample was used for extracting as many photons generated from the sidewalls of the CMS NWs as possible. In these LED structures, approximately 2 × 105 NWs are connected in parallel. The total junction area is estimated to be 1.3 × 10-3 cm2. To determine the recombination mechanism in the CMS-NWbased LED on Si, we measured the current of the CMS-NW devices at voltages from -4 to 4 V using an HP 4156B parameter analyzer. In the measurements, the Cr/Au metal was positively biased, and the Ti/Au backside electrode was grounded. Figure 3b shows the current-voltage (I-V) curves for these structures. The curves for all devices show moderate rectifying properties. The inset of the Figure 3b shows a semilogarithm plot of the I-V properties. In this structure, the depletion width is estimated to be 23 nm and the total thickness of this CMS layer was thicker than the depletion 1641

DOI: 10.1021/nl9041774 | Nano Lett. 2010, 10, 1639-–1644

higher than conventional GaAs-based diode (n ∼ 1.1-1.5) and also are higher than the thermal diffusion or recombination expected from the Sah-Noyce Schottky model.41 Extremely high ideal factors are usually observed for AlGaNbased diodes.42,43 The reason for such high value of the ideal factor is thought to be the carrier tunneling across the junction. The high RS for Sample B was caused by the current injection was not distributed with uniformity of metal films. This is because a difficulty of filling the thin Cr/Au metal into the deep slit, that is, the sidewalls of the NWs was not sufficiently covered with the thin Au/Cr metal films. The current in Sample B is, therefore, lower than that in Sample A. The I-V curve in Sample B indicates leakage current in the negative-bias region. It should be noted that this behavior was not observed for CMS-NWs on GaAs(111)B substrate (not shown here). This is because of the band discontinuity across the GaAs/Si junction. Heterojunction of the III-V and Si usually forms band discontinuity. Several reports have investigated the valence and conduction band offset with photoemission spectroscopy.44,45 Potential barrier caused from the band offset leads to such Schottky properties. The Schottky properties of III-V NW on Si have been reported.15,17,20,28 The discontinuity would be affected by several factors, such as misfit dislocations, interface states, and conductance of Si and III-V NWs. Further investigations are required to clarify this discontinuity across the III-V NWs/Si junctions and to further improve the performance of III-V NWs-based devices on Si. Figure 4a shows the typical EL spectra with several current conditions for Sample A. The threshold current for EL is 0.5 mA (current density is 0.3 A/cm2) with 1.9 V. The EL peak position is around 1.48 eV. The EL peak position is shifted to 60 meV from that of the GaAs bandgap at room temperature (Eg ) 1.42 eV). This EL came from the DH structures, because the estimated width of the GaAs QW tube is 7 nm, which is consistent with the value estimated from the low-temperature PL spectra shown in Figure 2a.

FIGURE 3. (a) Illustration of CMS NW-based LED structure. (b) I-V curves for Sample A (solid red line) and Sample B (solid blue line). Inset shows the semilogarithmic plots for the I-V curves.

width. The rectifying curves for the CMS NW-based LED array with thick AlGaAs layers (see Sample A and B in Figure 3b) indicate gradual linear curves resulting from unexpected resistance. The resistances originated from series and contact resistance across the NWs and Si substrate. Also, premature turn-on behavior40 originating from surface states occurred. Considering these resistance (RS) with the CMS NW-based LED and shunt resistance to be infinity, the current equation is expressed as J(n) ) J0 exp[(V-RSI)/nkT], where n is the ideal factor of the p-n junction. From this equation, the n is estimated to be 3.8 for Samples A and B, and the RS is calculated to be 45.3 Ω for Sample A and 2.0 kΩ for Sample B. The ideal factors for Sample A and B are

FIGURE 4. (a) Electroluminescence (EL) spectra under several current injections at room temperature. Solid lines indicate the EL, and the dashed line spectrum depicts the PL at room temperature. (b) Semilogarithm plot of EL intensity as a function of injected current. Inset depicts linear-plot of the EL intensity. © 2010 American Chemical Society

1642

DOI: 10.1021/nl9041774 | Nano Lett. 2010, 10, 1639-–1644

The fwhm of the EL is 130 meV, which is larger than the theoretical line width (1.8 kT) in LED spectra. This is because the EL spectra in Figure 4a contain several luminescence centers resulting from DAP transitions from Si, C, and Zn impurities in the AlGaAs layers. PL spectra at room temperature (see spectra with dashed line in Figure 4a) also show large fwhm. In case of Sample B, the threshold current was 0.15 µA with 2.0 V. The EL spectra in Sample B were the same as Sample A, and the EL intensity in Sample B was higher than that of Sample A. The EL intensity in Sample B was, however, unstable with current injection variations. This is because current injections around the sidewalls of the NWs were unstable due to the thickness of the thin metal. The EL intensity increased superlinearly with the current injection, and it was saturated from I ) 1.5 mA (1.2 A/cm2). The superlinear characteristic in the EL intensity at low current injection indicates the CMS NW-based LED is similar to that of superluminescence LED.46 This is thought to be because the AlGaAs shell layer surrounded by Cr/Au metal and oxides acts as a waveguide and because photon emitting into the waveguides increased with the current injections. The origin of the saturation was the carrier overflow as observed in surface-emitting LEDs. This is because the volume of the active region in the CMS NWs is very thin and parasitic resistance (contact resistance, series resistance, and so on) in the NWs is slightly high. Moreover, no shift of the EL peak position resulting from Joule heating in the junction temperature was observed with further increasing the current injections. This is because the Si substrate acts as a heatsink due to its good thermal conductivity as compared to III-V compound semiconductors. This heterogeneous integration can, therefore, lead to thermally stable driving. The EL intensity was 2 × 104 counts per second for 1.5 mA-current injections. Generally, GaAs-based LEDs fabricated on Si without buffering do not achieve such bright EL because of threading dislocations resulting from the thermal coefficient difference. Reduction of threading dislocations using selective-area growth47 has been reported, and we also found that InAs and GaAs NWs grown on Si contained no threading dislocations.22,26 The nanometer-scaled selective area technique, therefore, could control the generation of threading dislocations and produce better performing CMS NW-based LEDs directly grown on Si surfaces without buffering techniques. In summary, vertically aligned AlGaAs/GaAs/AlGaAs CMS NWs with p/n junctions and CMS NW-based LED arrays were successfully integrated on Si substrate. Appropriate rectifying properties and strong EL emission were observed from the CMS NWs directly grown on Si. The threshold current for the EL was 0.5 mA at 1.9 V. The EL intensity increased superlinearly with current injection and saturated with further increased current injection. The nonlinear increase in EL intensity was similar to the superluminescence behavior, and the saturation of the EL intensity was resulted from the carrier overflow effect due © 2010 American Chemical Society

to the thin optical region and low barrier height. The CMS NW-based LED array showed no shift of the EL peaks with further increasing current injection, which indicates the luminescence behavior is thermally stable. This technique is expected for applying to on-chip integration as well as Si photonics based on III-V NWs on Si. Acknowledgment. This work was financially supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan. One of the authors (K.T.) would like to acknowledge the financial support provided in part by a grant from the Global Center of Excellence (GCOE) Program from the MEXT, Research Foundation for Opto-Science and Technology, and PRESTO, Japan Science and Technology Agency (JST). Supporting Information Available. Figure S1 Excitation power dependence of PL intensity for peak P1 (closed red circles) and P2 (closed blue circles). Figure S2 (a) Peak deconvolution for PL under high excitation (11 kW/cm2) Inset shows PL under weak excitation under weak excitation power denisity (0.1 kW/cm2) (b) Excitation power dependence of Peak P3. Figure S3 Schematic illustrations for fabrication process of vertical NW-based LEDs: (a) selectivearea growth of core-multishell NWs with p-n junction on Si, (b) deposition of Al2O3 (20 nm) by atomic layer deposition (ALD), and SiO2 (50 nm) by RF sputtering, (c) Spin-coating of polymer, (d) RIE etching for top parts of the NWs, (e) wet etching for Al2O3/SiO2 by buffered HF solution, (f) Metal (Cr/ Au) deposition, and (g) mechanical polishing for top parts of the metal-coated NWs. (h), (i), and (j) show SEM images for Figure S3(d), (f), and (g), respectively. This material is available free of charge via the Internet at http://pubs.acs.org. REFERENCES AND NOTES (1) (2) (3)

(4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)

1643

Jalali, B.; Fathpour, S. J. Lightwave Technol. 2006, 24, 4600. Rong, H.; Xu, S.; Kuo, Y.-H.; Sih, V.; Cohen, O.; Raday, O.; Paniccia, M. Nat. Photonics 2007, 1, 232. Kang, Y.; Liu, H.-D.; Morse, M.; Paniccia, M. J.; Zadka, M.; Litski, S.; Sarid, G.; Pauchard, A.; Kuo, Y.-H.; Chen, H.-W.; Zaoui, W. S.; Bowers, J. E.; Beling, A.; McIntosh, D. C.; Zheng, X.; Campbell, J. C. Nat. Photonics 2009, 3, 59. Corcoran, B.; Monat, C.; Grillet, C.; Moss, D. J.; Eggleton, B. J.; White, T. P.; O’Faolain, L.; Krauss, T. F. Nat. Photonics 2009, 3, 206. Groenert, M. E.; Leitz, C. W.; P, A. J.; Yang, V.; Lee, H.; Ram, R. J.; Fitzgerald, E. A. J. Appl. Phys. 2003, 93, 362. Yonezu, H.; Fukukawa, Y.; Wakahara, A. J. Cryst. Growth 2008, 310, 4757. Liu, W. K.; et al. J. Cryst. Growth 2009, 311, 1979. Hiruma, K.; Yazawa, M.; Katsuyama, T.; Ogawa, K.; Haraguchi, K.; Koguchi, M.; Kakibayashi, H. J. Appl. Phys. 1995, 77, 447. Huang, M. H.; Mao, S.; Feick, H.; Yan, H.; Wu, Y.; Kind, H.; Weber, E.; Russo, R.; Yang, P. Science 2001, 292, 1897. Huang, Y.; Duan, X.; Cui, Y.; Lauhon, L. J.; Kim, K.-H.; Lieber, C. M. Science 2001, 294, 1313. Gudiksen, M. S.; Lauhon, L. J.; Wang, J.; Smith, D. C.; Lieber, C. M. Nature 2002, 415, 617. Hayden, O.; Agarwal, R.; Lieber, C. M. Nat. Mater. 2006, 5, 352. Bakkers, E. P. A. M.; Borgstro¨m, M. T.; Verheijen, M. A. MRS Bull. 2007, 32, 117. Rehnstedt, C.; Mårtensson, T.; Thelander, C.; Samuelson, L.; Wernersson, L.-E. IEEE Trans. Electron Devices 2008, 55, 3037. DOI: 10.1021/nl9041774 | Nano Lett. 2010, 10, 1639-–1644

(29) Lee, C.-H.; Yoo, J.; Hong, Y. J.; Cho, J.; Kim, Y.-J.; Jeon, S.-R.; Baek, J. H.; Yi, G.-C. Appl. Phys. Lett. 2009, 94, 213101. (30) Qian, F.; Gradecˇak, S.; Li, Y.; Wen, C.-Y.; Lieber, C. M. Nano Lett. 2005, 5, 2287. (31) An, S. J.; Chae, J. H.; Yi, G.-C.; Park, G. H. Appl. Phys. Lett. 2008, 92, 121108. (32) Yang, Y.; Sun, X. W.; Tay, B. K.; You, G. F.; Tan, S. T.; Teo, K. L. Appl. Phys. Lett. 2008, 93, 253107. (33) Lai, E.; Kim, W.; Yang, P. Nano Res. 2008, 1, 123. (34) Kikuchi, A.; Kawai, M.; Tada, M.; Kishino, A. Jpn. J. Appl. Phys. 2004, 43, L1524. (35) Sekiguchi, H.; Kishino, K.; Kikuchi, K. Electron. Lett. 2008, 44, 151. (36) Goto, H.; Nosaki, K.; Tomioka, K.; Hara, S.; Hiruma, K.; Motohisa, J.; Fukui, T. Appl. Phys. Exp. 2009, 2, 035004. (37) Xu, Z. Y.; Kreismanis, V. G.; Tang, C. L. J. Appl. Phys. 1983, 54, 4536. (38) Ballingall, J. M.; Collins, D. M. J. Appl. Phys. 1983, 54, 341. (39) Pal, B.; Gopal, A. V.; Prabhu, S. S.; Vengurlekar, A. S. Phys. Rev. B 2002, 65, 045312. (40) Schubert, E. F. Light-Emitting Diodes, 2nd ed.; Cambridge University Press: Cambridge, 2006. (41) Sah, C. T.; Noyce, R. N.; Shockley, W. Proc. IRE 1957, 45, 1228. (42) Casey, H. C., Jr.; Muth, J.; Kishnankutty, S.; Zavada, J. M. Appl. Phys. Lett. 1996, 68, 2867. (43) Perlin, P.; Osinski, M.; Eliseev, P. G.; Smagley, V. A.; Mu, J.; Banas, M.; Sartori, P. Appl. Phys. Lett. 1996, 69, 1680. (44) Katnani, A. D.; Margaritondo, G. Phys. Rev. B 1983, 28, 1944. (45) Zhou, Y. C.; Zhu, Z. H.; Crouse, D.; Lo, Y. H. Appl. Phys. Lett. 1998, 73, 2337. (46) Ray, S. K.; Groom, K. M.; Liu, H. Y.; Hopkins, M.; Hogg, R. A. Jpn. J. Appl. Phys. 2006, 45, 2542. (47) Nishinaga, T.; Nakano, T.; Zhang, S. Jpn. J. Appl. Phys. 1988, 27, L964.

(15) Tanaka, T.; Tomioka, K.; Hara, S.; Motohisa, J.; Sano, E.; Fukui, T. Appl. Phys. Exp. 2010, 3, 025003. (16) Svensson, C. P. T.; Mårtensson, T.; Tra¨gårdh, J.; Larsson, C.; Rask, M.; Hessman, D.; Samuelson, L.; Ohlsson, J. Nanotechnology 2008, 19, 305201. (17) Wei, W.; Bao, X.-Y.; Soci, C.; Ding, Y.; Wang, Z.-L.; Wang, D. Nano Lett. 2009, 9, 2926. (18) Ma˚rtensson, A. T.; Svesson, C. P.; Wacaser, B. A.; Larsson, M. W.; Seifert, W.; Deppert, K.; Gustafsson, A.; Wallenberg, L. R.; Samuelson, L. Nano Lett. 2004, 4, 1987. (19) Roest, A. L.; Verheijen, M. A.; Wunnicke, O.; Serafin, S.; Wondergem, H.; Bakkers, E. P. A. M. Nanotechnology 2006, 17, S271. (20) Ihn, S.-S.; Song, J.-I.; Kim, T.-W.; Leem, D.-S.; Lee, T.; Lee, S.-G.; Koh, E. K.; Song, K. Nano Lett. 2007, 7, 39. (21) Ma˚rtensson, T.; Wagner, J. B.; Hilner, E.; Mikkelsen, A.; Thelander, C.; Stangl, J.; Ohlsson, B. J.; Gustafsson, S.; Lundgren, E.; Samuelson, L.; Seifert, W. Adv. Mater. 2007, 19, 1801. (22) Tomioka, K.; Motohisa, J.; Hara, S.; Fukui, T. Nano Lett. 2008, 8, 3475. (23) Bao, X.-Y.; Soci, C.; Susac, D.; Bratvold, J.; Aplin, D. P. R.; Wei, W.; Chen, C.-Y.; Dayeh, S. A.; Kavanagh, K. L.; Wand, D. Nano Lett. 2008, 8, 3755. (24) Dick, K. A.; Deppert, K.; Samuelson, L.; Wallenberg, R.; Ross, F. M. Nano Lett. 2008, 8, 4087. (25) Woo, R. L.; Xiao, R.; Kobayashi, Y.; Gao, L.; Goel, L.; Hudait, M. K.; Mallouk, T. E.; Hicks, R. F. Nano Lett. 2008, 8, 4664. (26) Tomioka, K.; Kobayashi, Y.; Motohisa, J.; Hara, S.; Fukui, T. Nanotechnology 2009, 20, 145302. (27) Gao, L.; Woo, R. L.; Linag, B.; Pozuelo, M.; Prikhodko, S.; Jackson, M.; Goel, N.; Hudait, M. K.; Huffaker, D. L.; Goorsky, M. S.; Kodambaka, S.; Hicks, R. F. Nano Lett. 2009, 9, 2223. (28) Garnett, E. C.; Yang, P. J. Am. Chem. Soc. 2008, 130, 9224.

© 2010 American Chemical Society

1644

DOI: 10.1021/nl9041774 | Nano Lett. 2010, 10, 1639-–1644