p-Type InN Nanowires

Share Embed


Descrição do Produto

Letter pubs.acs.org/NanoLett

p‑Type InN Nanowires S. Zhao,† B. H. Le,† D. P. Liu,‡ X. D. Liu,‡ M. G. Kibria,† T. Szkopek,† H. Guo,‡ and Z. Mi*,† †

Department of Electrical and Computer Engineering, McGill University 3480 University Street, Montreal, Quebec H3A 0E9, Canada Department of Physics, McGill University, 3600 University Street, Montreal, Quebec, H3A 2T8, Canada



S Supporting Information *

ABSTRACT: In this Letter, we demonstrate that with the merit of nanowire structure and a self-catalytic growth process p-type InN can be realized for the first time by “direct” magnesium (Mg) doping. The presence of Mg acceptor energy levels in InN is confirmed by photoluminescence experiments, and a direct evidence of p-type conduction is demonstrated unambiguously by studying the transfer characteristics of InN nanowire field effect transistors. Moreover, the nearsurface Fermi-level of InN can be tuned from nearly intrinsic to p-type degenerate by controlling Mg dopant incorporation, which is in contrast to the commonly observed electron accumulation on the grown surfaces of Mg-doped InN films. First-principle calculation using the VASP electronic package further shows that the p-type surface formed on Mg-doped InN nanowires is highly stable energetically. KEYWORDS: InN, nanowire, Mg doping, p-type very large residual electron density (on the order of 1018 cm−3, or higher) and high density of accumulated electrons on the lateral nonpolar surfaces12,15,16 due to the formation of extensive n-type defects and their preferential incorporation into the near-surface region.17 Furthermore, recent studies18 have shown that defect species such as VN and ON are present (e.g., by self-compensation) and stable in p-type materials, further enhancing the residual electron density. These factors and the resulting strong n-type characteristics underscore the difficulty in realizing p-type conductivity in InN nanowires. For similar reasons “direct” p-type doping has not been possible in any of the technologically important narrow-bandgap semiconducting nanowires, including InN, InAs, and InSb. Recently, we have demonstrated that intrinsic InN nanowires can be achieved by an improved molecular beam epitaxy (MBE) technique17,19 that involves the use of an in situ deposited In seeding layer to promote the nucleation and formation of InN nanowires. A self-catalytic growth process, together with an ultrahigh vacuum MBE environment, can largely eliminate any undesirable impurity atom incorporation thus further minimize the formation of extensive surface defects. Consequently, the resulting InN nanowires possess extremely low residual electron densities (in the range of 1013 to 1015 cm−3)20 and are absent of any surface electron accumulation on the lateral nonpolar surfaces.17,21,22 This progress has made the realization of p-type doping into InN nanowires possible. In this Letter, we investigate the direct

III-nitride semiconductors are critical for a wide range of applications including solid-state lighting, ultraviolet photonics, photovoltaics, high-power electronics, and biosensors; the IIInitrides are regarded as the next Si.1,2 One grand challenge for III-nitride semiconductors is the realization of p-type InN, which severely limits their device applications. To date, the study of p-type doping into InN is limited to InN thin films3−7 and a direct evidence for the presence of free holes is still lacking due to the commonly measured surface electron accumulation on the grown surfaces of InN thin films,8,9 which greatly overwhelms the p-type conduction. Recently, by cleaving an InN thin film the absence of surface electron accumulation was observed on nonpolar planes,10,11 which sheds light on the possibility to measure p-type conduction directly. However, obtaining nonpolar grown surfaces without surface electron accumulation has not been possible mostly due to the lack of suitable substrates. In this regard, resorting to other low-dimensional structures with their optical and electrical properties being largely determined by nonpolar planes may provide a feasible solution. It has been found that the lateral surfaces of InN nanowires grown directly on Si substrates are typically nonpolar m-planes,12,13 which largely determine the optical and electrical properties of InN nanowires because these nonpolar planes form the majority of the nanowire surface area. The nanowire approach can therefore offer a promising route to achieving p-type conduction. However, InN nanowires grown by conventional methods, such as the vapor−liquid−solid (VLS) process14 and spontaneous formation process,13 typically exhibit tapered surface morphology, and thus poor optical and electrical properties. These nominally nondoped InN nanowires possess © XXXX American Chemical Society

Received: August 17, 2013 Revised: September 12, 2013

A

dx.doi.org/10.1021/nl4030819 | Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

Letter

Figure 1. Electron microscopies of Mg-doped InN nanowires. (a) An SEM image of Mg-doped InN nanowires taken with a 85° angle. The nanowires are oriented along the c-axis with their sidewalls being nonpolar m-planes. (b,c) High-resolution TEM images from the top region and root region of a typical Mg-doped InN nanowire, respectively. The arrows denote the growth direction.

along the c-axis.23 The nearly defect-free nanowire root region stems partly from the absence of an epitaxial relationship between III-nitride nanowires and Si substrates by the present MBE process.19,24,25 The presence of Mg acceptors is directly measured by microphotoluminescence (PL) experiments, which were performed in a homemade system consisting of a semiconductor diode laser (λ = 635 nm), a 100× objective, a highresolution spectrometer, a liquid-nitrogen-cooled InGaAs detector (with a cutoff wavelength of ∼2.2 μm), and a single channel lockin amplifier. Figure 2 shows the PL spectra of Mg-

incorporation of Mg into InN nanowires grown by the aforementioned MBE technique, and we demonstrate that ptype InN can be realized for the first time by direct magnesium (Mg) doping. Moreover, the nanowire surface can be tuned from nearly intrinsic to p-type degenerate, which is in contrast to the commonly observed electron accumulation on grown surfaces of InN thin films. This work overcomes the major roadblock to extending the spectral range of contemporary IIInitride optoelectronic devices from the ultraviolet and visible to the infrared and holds great promise to the realization of complementary ultrahigh speed electronic devices based on IIInitride semiconductors. Mg-doped InN nanowires were grown on Si(111) substrates by radio frequency plasma-assisted MBE under nitrogen-rich conditions. A thin (∼0.6 nm) In seeding layer was deposited on Si substrates before introducing nitrogen. The In seeding layer forms nanoscale droplets at high temperatures, which can promote the subsequent formation and nucleation of InN nanowires. The growth conditions for Mg-doped InN nanowires included a substrate temperature of ∼480 °C, an In beam equivalent pressure of ∼6 × 10−8 Torr, a nitrogen flow rate of ∼1.0 sccm, and a RF plasma forward power of ∼350 W. The Mg cell temperatures were 190, 210, 220, 230, and 240 °C, which corresponds to Mg beam equivalent pressures of ∼7 × 10−12, 3.1 × 10−11, 6.1 × 10−11, 1.2 × 10−10, and 2.6 × 10−10 Torr, respectively. Figure 1a shows a typical scanning electron microscope (SEM) image of Mg-doped InN nanowires, taken using a Hitachi S-4700 system with a 85° angle. As seen, the Mg-doped InN nanowires exhibit nontapered surface morphology with a well-defined hexagonal structure. Moreover, high-resolution transmission electron microscope (TEM) images, that were taken with a Tecnai G2 F20 S/TEM system equipped with a Gatan 4k × 4k CCD camera, from the top (Figure 1b) and root (Figure 1c) regions of a typical Mg-doped InN nanowire indicate that such Mg-doped InN nanowires are free of stacking faults. A detailed examination further suggests that the nanowire is free of misfit dislocations. The interplanar spacing is found to be 0.293 nm, confirming that the growth direction is

Figure 2. The PL spectra of Mg-doped InN nanowires measured at 7 K under a 9 mW optical excitation. The PL spectrum of nondoped InN nanowires measured at 20 K under the same power is also shown for a comparison. The spectra were normalized by the PL peak intensity of the nondoped InN nanowires and were shifted for display purpose. The dotted green curve shows the PL spectrum measured under an excitation of 200 μW from the Mg-doped sample with a Mg cell temperature of 190 °C; and the inset shows the PL peak intensity ratio of PPLL over PPLH as a function of the excitation power. B

dx.doi.org/10.1021/nl4030819 | Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

Letter

doped InN nanowires measured at 7 K under an excitation of 9 mW. The PL spectrum of nondoped InN nanowires measured at 20 K under the same excitation power is also shown for a comparison. For the nondoped InN nanowires, only a single PL peak appears at ∼0.69 eV, which is consistent with the PL emission of high-quality InN under such high excitation conditions.22,26 When the Mg acceptors are incorporated (with a Mg cell temperature of 190 °C), two PL emission peaks can be clearly distinguished with peak energies EPLL ∼ 0.61 eV and EPLH ∼ 0.67 eV, respectively. The energy separation between EPLH and EPLL is ∼60 meV, which is consistent with the Mg ionization energy in InN.3,4 Detailed power dependent studies further indicate that the EPLL peak dominates under low excitation conditions. As shown by the dotted curve, only the EPLL peak can be clearly observed when a low 200 μW excitation was used. At low excitations, primarily electrons in the conduction band and holes at the Mg-acceptor energy level participate in radiative recombination. The power dependent PL peak intensity ratio (PPLL over PPLH, where PPLL and PPLH are the PL peak intensity of the EPLL peak and the EPLH peak, respectively) is shown in the inset of Figure 2, confirming that the EPLL peak arises from the Mg-acceptor related transition rather than other effects such as being a phonon replica of the EPLH peak. With increasing Mg doping (with a Mg cell temperature of 210 °C), only a low-energy peak can be observed, which is further redshifted to EPL3 ∼ 0.59 eV as compared with EPLL ∼ 0.61 eV. This redshift can be largely ascribed to the bandgap renormalization with increased Mg doping,27 and the drastic reduction of the PL peak intensity with increasing Mg concentration has been commonly measured in Mg-doped InN, which can be ascribed to the enhanced defect incorporation such as nitrogen vacancies or Mg related defects. Such PL emission characteristics have been previously observed in InN thin films.3 However, a direct measurement of p-type conduction and/or p-type surface has not been possible in previous studies due to the accumulated electrons on grown surfaces. In what follows, we demonstrate that with relatively high Mg dopant incorporation and optimized growth conditions, p-type conduction including p-type surface with tunable near-surface Fermi-level, can be realized for the first time in InN. To perform electrical measurements, single nanowire field effect transistors were fabricated. The as-grown Mg-doped InN nanowires were first dispersed to patterned Si chips coated with a SiOx dielectric layer. Then EL11/PMMA2 was coated to serve as the electron beam resist for the subsequent e-beam lithography process. Multilayer electrical contact (Ni/Au/Ni/ Al/Ni/Au) was deposited by e-beam evaporation. The single nanowire transistors were annealed at 400 °C for 1 min in N2 gas before electrical measurements. The devices were measured under a direct current (dc) bias at room temperature in vacuum. A single nanowire field effect transistor fabricated from a Mg-doped InN nanowire with a Mg cell temperature of 220 °C is schematically shown in Figure 3a. An SEM image of a device is shown in Figure 3b. The nanowire has a radius r of ∼150 nm with a channel length L of ∼1 μm. At room temperature, it can be seen from Figure 3c that ISD increases linearly with VSD in the measured voltage range, indicating the formation of Ohmic contacts. Moreover, with a negative backgate voltage applied (VGD < 0), the nanowire conductance ISD/ VSD (where ISD is the source-drain current and VSD is the source-drain voltage) increases dramatically. The increase in

Figure 3. Characterization of a Mg-doped InN nanowire field effect transistor. (a) A schematic plot of the measurement configuration. (b) An SEM image of a Mg-doped InN nanowire field effect transistor. (c) ISD−VSD characteristics. (d) ISD−VGD dependence under VSD = 0.05 V. The arrow denotes the minimum ISD position. The solid red line is a linear fit.

conductance ISD/VSD with increasingly negative VGD is the unambiguous signature of p-type conduction in Mg-doped InN nanowires, as has been commonly observed in other p-type nanowire field effect transistors.28−30 Such a direct measurement of p-type conduction from the Mg-doped InN nanowires is a natural consequence of the absence of surface electron accumulation on the nanowire nonpolar sidewalls (see Figure 4d). For a comparison, surface electron accumulation has been near-universally observed on the grown surfaces of InN films, which prevents a direct measurement of p-type conduction. The ISD−VGD dependence at a fixed VSD (= 0.05 V) is plotted in Figure 3d, clearly exhibiting an increase in ISD for more negative VGD. The minimum ISD occurs at a small positive VGD ∼ 0.1 V at which point free holes are depleted, indicating that the conduction is p-type at zero gate bias. By the ISD−VGD dependence, the field effect hole mobility can be derived via μ = gmL2/(CGVSD), where gm = dISD/dVGD and CG is the total gate capacitance estimated by CG = 2πεε0L/cosh−1[(r + tox)/r] with a cylinder-on-plate model.29 For an oxide thickness tox = 100 nm, a nanowire radius r = 150 nm, a source-drain distance L = 1 μm, and the measured transconductance gm from the slope of the linear part in Figure 3d, the field effect hole mobility is derived to be ∼100 cm2/V·s. The mobility is comparable to recent theoretical calculations by the ensemble Monte Carlo method.7 With this hole mobility, the hole concentration (at VGD = 0 V) can be approximated to be ∼5 × 1015 cm−3 via 1/ (Rtotal − Rcontact)[L/(πr2)] = neμ, where the contact resistance was estimated by three-point measurements (see Supporting Information). The p-type field effect transistor behavior was measured in three devices. The field effect hole mobility and hole concentration are in the ranges of 64 to 130 cm2/V·s and 2 × 1015 to 6 × 1015 cm−3, respectively. The contact resistance takes about 50−60% of the total resistance for all the devices (see Supporting Information). The maximum VSD in these experiments was 50 mV, beyond which the devices tend to burn out. Subsequently we demonstrate that p-type surface can be realized. The valence band spectra on the lateral nonpolar C

dx.doi.org/10.1021/nl4030819 | Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

Letter

Figure 4. X-ray photoelectron spectroscopies of Mg-doped InN nanowires with different Mg doping concentrations. Panels (b−f) correspond to Mg cell temperatures of 190, 210, 220, 230, and 240 °C, respectively. The spectrum of a nondoped sample is also shown in (a) for a comparison.

Figure 5. Surface charge properties of Mg-doped InN nanowires. (a) The derived near-surface Fermi-level from the XPS results as a function of the Mg cell temperature. In this plot, EV is set to be at 0 eV for all the samples. For Mg cell temperatures below 220 °C, Mg desorption dominates; while for Mg cell temperatures above 220 °C, Mg incorporation, which is captured by the first-principle calculation as shown in (b), dominates. (b) In-substitutional Mg formation energy along the nanowire radial direction. The inset shows the structure used for calculation. The Mg position index 0 represents the surface absorption position of Mg. The index 1 represents a transient state from the surface absorption to the surface doping, that is, the surface In-replacement. The indices from 2 to 11 represent the InN layer from the surface to the deeper layer.

surfaces of such Mg-doped InN nanowires were measured by the angle-resolved X-ray photoelectron spectroscopy (XPS) using a Thermo Scientific K-Alpha system, as illustrated in Figure 4. In this experiment, an X-ray beam was impinged upon the nanowires at a 60 degree angle with respect to the nanowire c-axis, ensuring the majority of the signal was derived from nanowire sidewalls. The valence band spectra were calibrated by both Au-4f peak (84 eV) and C-1s peak (285 eV). The spectrum of nondoped InN nanowires is also shown for a comparison. A distinct feature shown in Figure 4 is that for all Mg-doped nanowire samples, no surface electron accumulation is observed, in contrast to the commonly measured surface electron accumulation from Mg-doped InN thin films.3,5 In addition, the near-surface Fermi-level of the nondoped InN nanowires and Mg-doped InN nanowires with Mg cell temperatures below 220 °C are approximately the same: the energy separation between the near-surface Fermi-level and the valence band maximum (VBM) is in the range of 0.4 to 0.5 eV, suggesting that the surface is nearly intrinsic (the intrinsic Fermi-level is estimated to be 0.38 eV above the VBM, see Supporting Information). However, with further increasing Mg cell temperatures shown in Figure 4e,f, the near-surface Fermilevels shift noticeably toward the VBM. The nanowire surfaces become completely p-type for Mg cell temperatures of 230 and 240 °C with the corresponding near-surface Fermi-levels positioned at 0.25 and 0.10 eV above the VBM, respectively, which are well below the intrinsic Fermi-level of InN. The measured near-surface Fermi-level as a function of the Mg cell temperature is further summarized in Figure 5a. The conduction band edge (EC), intrinsic Fermi-level (EFi), and valence band edge (EV, which is set to be 0 eV in this plot) in the near-surface region are also illustrated in Figure 5a. It is clearly seen that with increasing Mg doping the near-surface Fermi-level shifts toward the VBM and the surface becomes completely p-type. To further understand the Mg atom incorporation mechanism in InN nanowires, we have performed the first-principle calculation on m-planes (nanowire sidewalls) using the VASP electronic package including the plane wave basis (with a cutoff energy of 450 eV),

pseudopotential approach (with a nonlinear core correction), and the local spin density approximation (LSDA). The calculations were performed on a 3 × 2 × 1 grid k-point sampling with a 1.52 nm × 1.70 nm × 3.12 nm slab-supercell, and atomic positions were fully relaxed until the forces on each atom converged to less than 1 × 10−2 eV/Ang. Twelve atomic InN layers were included in the slab-supercell and the convergence of work function for the slab were tested. The transition states between on-surface and in-surface Mg dopant were calculated by the nudged elastic band (NEB) method.31 As illustrated in Figure 5b, the In-substitutional Mg-doping has significantly lower surface formation energy compared to that in the bulk region, which thus leads to the preferential incorporation of Mg dopants into the near-surface region. However, the direct doping of Mg atoms into InN nanowires suffers considerably from the large surface desorption of Mg at elevated growth temperature (see Supporting Information). This can largely explain the nearly intrinsic surface measured for relatively low Mg-doped nanowires, e.g., the Mg-doped InN nanowires grown with Mg cell temperatures below 220 °C in this study (illustrated in Figure 5a). With increasing Mgdoping, enhanced acceptor incorporation in the near-surface region due to the lower formation energy of Mg can balance to a certain extent the large surface desorption of Mg atoms during the epitaxial growth process; in other words, Mg incorporation dominates (as shown in Figure 5a for Mg cell temperatures larger than 220 °C). This could lead to the achievement of p-type surfaces. D

dx.doi.org/10.1021/nl4030819 | Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

Letter

(9) Linhart, W. M.; Veal, T. D.; King, P. D. C.; Koblmüller, G.; Gallinat, C. S.; Speck, J. S.; McConville, C. F. Appl. Phys. Lett. 2010, 97, 112103. (10) Wu, C.-L.; Lee, H.-M.; Kuo, C.-T.; Chen, C.-H.; Gwo, S. Phys. Rev. Lett. 2008, 101, 106803. (11) Ebert, P.; Schaafhausen, S.; Lenz, A.; Sabitova, A.; Ivanova, L.; Dähne, M.; Hong, Y. L.; Gwo, S.; Eisele, H. Appl. Phys. Lett. 2011, 98, 062103. (12) Stoica, T.; Meijers, R. J.; Calarco, R.; Richter, T.; Sutter, E.; Lueth, H. Nano Lett. 2006, 6, 1541. (13) Calleja, E.; Ristić, J.; Fernández-Garrido, S.; Cerutti, L.; Sánchez-García, M. A.; Grandal, J.; Trampert, A.; Jahn, U.; Sánchez, G.; Griol, A.; Sánchez, B. Phys. Status Solidi B 2007, 244, 2816. (14) Zhang, J.; Liu, H.; Huang, R.; Kong, T.; Cheng, G. J. Nanoeng. Nanomanuf. 2012, 2, 112. (15) Richter, T.; Luth, H.; Schapers, T.; Meijers, R.; Jeganathan, K.; Estevez Hernandez, S.; Calarco, R.; Marso, M. Nanotechnology 2009, 20, 405206. (16) Segura-Ruiz, J.; Molina-Sánchez, A.; Garro, N.; García-Cristóbal, A.; Cantarero, A.; Iikawa, F.; Denker, C.; Malindretos, J.; Rizzi, A. Phys. Rev. B 2010, 82, 125319. (17) Zhao, S.; Fathololoumi, S.; Bevan, K. H.; Liu, D. P.; Kibria, M. G.; Li, Q.; Wang, G. T.; Guo, H.; Mi, Z. Nano Lett. 2012, 12, 2877. (18) Duan, X.; Stampfl, C. Phys. Rev. B 2009, 79, 035207. (19) Chang, Y. L.; Li, F.; Fatehi, A.; Mi, Z. Nanotechnology 2009, 20, 345203. (20) Zhao, S.; Salehzadeh, O.; Alagha, S.; Kavanagh, K. L.; Watkins, S. P.; Mi, Z. Appl. Phys. Lett. 2013, 102, 073102. (21) Zhao, S.; Wang, Q.; Mi, Z.; Fathololoumi, S.; Gonzalez, T.; Andrews, M. P. Nanotechnology 2012, 23, 415706. (22) Zhao, S.; Mi, Z.; Kibria, M. G.; Li, Q.; Wang, G. T. Phys. Rev. B 2012, 85, 245313. (23) Tang, T.; Han, S.; Jin, W.; Liu, X.; Li, C.; Zhang, D.; Zhou, C.; Chen, B.; Han, J.; Meyyapan, M. J. Mater. Res. 2004, 19, 423. (24) Consonni, V.; Knelangen, M.; Jahn, U.; Trampert, A.; Geelhaar, L.; Riechert, H. Appl. Phys. Lett. 2009, 95, 241910. (25) Zhao, S.; Kibria, M. G.; Wang, Q.; Nguyen, H. P. T.; Mi, Z. Nanoscale 2013, 5, 5283. (26) Holtz, M. E.; Gherasoiu, I.; Kuryatkov, V.; Nikishin, S. A.; Bernussi, A. A.; Holtz, M. W. J. Appl. Phys. 2009, 105, 063702. (27) Kudrawiec, R.; Suski, T.; Misiewicz, J.; Muto, D.; Nanishi, Y. Phys. Status Solidi C 2009, 6, S739. (28) Zhong, Z.; Qian, F.; Wang, D.; Lieber, C. M. Nano Lett. 2003, 3, 343. (29) Wang, D.; Wang, Q.; Javey, A.; Tu, R.; Dai, H.; Kim, H.; McIntyre, P. C.; Krishnamohan, T.; Saraswat, K. C. Appl. Phys. Lett. 2003, 83, 2432. (30) Colinge, J.-P.; Lee, C.-W.; Afzalian, A.; Akhavan, N. D.; Yan, R.; Ferain, I.; Razavi, P.; O’Neill, B.; Blake, A.; White, M.; Kelleher, A.-M.; McCarthy, B.; Murphy, R. Nat. Nanotechnol. 2010, 5, 225. (31) Sheppard, D.; Henkelman, G. J. Comput. Chem. 2011, 32, 1769. (32) Li, H.-Y.; Wunnicke, O.; Borgstrom, M. T.; Immink, W. G. G.; Weert, M. H. M. v.; Verheijen, M. A.; Bakkers, E. P. A. M. Nano Lett. 2007, 7, 1144. (33) Ho, J. C.; Yerushalmi, R.; Jacobson, Z. A.; Fan, Z.; Alley, R. L.; Javey, A. Nat. Mater. 2008, 7, 62.

In this regard, a careful control of competing processes (e.g., Mg acceptor desorption and adsorption) through variations of the growth parameters including the substrate temperature and Mg flux enables a wide range of tunability of the Fermi-level from nearly intrinsic to p-type degenerate on InN nanowire surfaces. Moreover, there exists a relatively large energy barrier (∼250 meV) to remove any In-substitutional Mg in the nearsurface region after such substitution occurs, illustrated by the formation energy difference between the Mg position indices 1 and 2 in Figure 5b. This energy barrier is significantly higher than the thermal energy at both the growth temperature (∼480 °C) and the room temperature. Therefore, the p-type surface formed on Mg-doped InN nanowires is expected to be highly stable energetically. This work is the first realization of p-type conduction in any In-containing narrow-bandgap semiconductor nanowires by direct doping, in contrast to previously reported indirect techniques such as remote doping32 and surface doping.33 The control over surface electronic properties will have tremendous impact on emerging nanoscale Si-integrated photonic, electronic, and biochemical sensing devices and systems, to the call of Moore’s law. For instance, the precise control of Fermi-level on narrow bandgap nanowire surfaces opens the door to complementary ultrahigh speed field effect transistor circuits with significantly reduced power consumption.



ASSOCIATED CONTENT

S Supporting Information *

Additional information, figures, and references. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: 1-514-398-7114. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the Natural Sciences and Engineering Research Council of Canada, the Fonds de recherché sur la nature et les technologies, and U.S. Army Research Office under Grant W911NF-12-1-0477. Part of the work was performed in the McGill Nanotools-Microfab facility. SEM and TEM studies were performed in the Facility for Electron Microscopy Research, McGill University.



REFERENCES

(1) Strite, S.; Morkoc, H. J. Vac. Sci. Technol., B 1992, 10, 1237. (2) Wu, J. J. Appl. Phys. 2009, 106, 011101. (3) Wang, X.; Che, S.-B.; Ishitani, Y.; Yoshikawa, A. Appl. Phys. Lett. 2007, 90, 201913. (4) Khan, N.; Nepal, N.; Sedhain, A.; Lin, J. Y.; Jiang, H. X. Appl. Phys. Lett. 2007, 91, 012101. (5) Wang, K.; Miller, N.; Iwamoto, R.; Yamaguchi, T.; Mayer, M. A.; Araki, T.; Nanishi, Y.; Yu, K. M.; Haller, E. E.; Walukiewicz, W.; Ager, J. W. Appl. Phys. Lett. 2011, 98, 042104. (6) Mayer, M. A.; Choi, S.; Bierwagen, O.; Smith, H. M.; Haller, E. E.; Speck, J. S.; Walukiewicz, W. J. Appl. Phys. 2011, 110, 123707. (7) Ma, N.; Wang, X. Q.; Liu, S. T.; Chen, G.; Pan, J. H.; Feng, L.; Xu, F. J.; Tang, N.; Shen, B. Appl. Phys. Lett. 2011, 98, 192114. (8) Kuo, C.-T.; Lin, S.-C.; Chang, K.-K.; Shiu, H.-W.; Chang, L.-Y.; Chen, C.-H.; Tang, S.-J.; Gwo, S. Appl. Phys. Lett. 2011, 98, 052101. E

dx.doi.org/10.1021/nl4030819 | Nano Lett. XXXX, XXX, XXX−XXX

Lihat lebih banyak...

Comentários

Copyright © 2017 DADOSPDF Inc.