Tuning the Surface Charge Properties of Epitaxial InN Nanowires

Letter pubs.acs.org/NanoLett

Tuning the Surface Charge Properties of Epitaxial InN Nanowires S. Zhao,† S. Fathololoumi,† K. H. Bevan,‡ D. P. Liu,§ M. G. Kibria,† Q. Li,∥ G. T. Wang,∥ Hong Guo,§ and Z. Mi*,† †

Department of Electrical and Computer Engineering, ‡Department of Mining and Materials Engineering, and §Department of Physics, McGill University, 3480 University Street, Montreal, QC, H3A 2A7, Canada ∥ Advanced Materials Sciences, Sandia National Laboratories, Albuquerque, New Mexico 87185, United States S Supporting Information *

ABSTRACT: We have investigated the correlated surface electronic and optical properties of [0001]-oriented epitaxial InN nanowires grown directly on silicon. By dramatically improving the epitaxial growth process, we have achieved, for the first time, intrinsic InN both within the bulk and at nonpolar InN surfaces. The near-surface Fermi-level was measured to be ∼0.55 eV above the valence band maximum for undoped InN nanowires, suggesting the absence of surface electron accumulation and Fermi-level pinning. This result is in direct contrast to the problematic degenerate two-dimensional electron gas universally observed on grown surfaces of n-type degenerate InN. We have further demonstrated that the surface charge properties of InN nanowires, including the formation of twodimensional electron gas and the optical emission characteristics can be precisely tuned through controlled n-type doping. At relatively high doping levels in this study, the near-surface Fermi-level was found to be pinned at ∼0.95−1.3 eV above the valence band maximum. Through these trends, well captured by the effective mass and ab initio materials modeling, we have unambiguously identified the definitive role of surface doping in tuning the surface charge properties of InN. KEYWORDS: Nanowire, InN, electron accumulation, two-dimensional electron gas, X-ray photoelectron spectroscopy, photoluminescence 1014 cm−2) forms at both the polar and nonpolar grown surfaces of such InN nanowires. Consequently, the surface Fermi-level pins deep into the conduction band,13−16 leading to severe downward band bending in the near-surface region and a highly inhomogeneous electron distribution.21,25 This produces distinctive signatures in the photoluminescence (PL) emission and micro-Raman spectra of InN nanowires.21,22,26 The resulting uncontrolled electronic and optical properties severely limit the practical applications of nanowire devices. In order to minimize the n-type characteristics of InN nanowires, we have recently developed a special growth technique, with the use of an in situ deposited In seeding layer, for the self-organized growth of nontapered InN nanowires directly on Si (111).27,28 In the previous report,25 near-intrinsic InN nanowires were demonstrated with a narrow PL spectral line width (∼13 meV) and a low residual electron density (∼2 × 1016 cm−3). However, the achievement of intrinsic InN surfaces, i.e., direct evidence for the absence of surface band bending and electron accumulation as well as tuning of the surface charge properties had not been possible. In this work, we have investigated the correlated surface electronic and optical properties of [0001]-oriented nontapered

D

islocation-free semiconductor nanowire structure devices are an extremely promising route toward compound semiconductor integration with silicon technology.1−8 However, precise control over nanowire doping as well as the surface charge properties has remained a near universal material challenge to date.3,5,9 In the technologically important indiumcontaining compounds (such as InN, InAs, and In2O3), the conduction band minimum (CBM) lies well below the charge neutrality level (CNL) due to the low In-5s orbital energy.10−12 Moreover, in the extreme case of InN, which has the lowest CBM of any reported semiconductor, the CNL is positioned at ∼1.2 eV above the CBM.12 Hence, defects and impurities in these compounds are generally donor-like, leading to extremely high electron densities even in nominally undoped structures. Common to these materials is the near-universal observation of a two-dimensional electron gas (2DEG) at the grown surfaces of both thin film and nanowire structures, the origins of which have remained a subject of intense debate.10,13−20 For these reasons, the achievement of intrinsic and p-type InN as well as precise control over their surface charge properties has remained an elusive goal. To date, uncontrolled surface charge properties represent one of the most fundamental challenges in nanowire electronics and photonics technology development. Conventional epitaxially grown InN nanowires generally

Received: February 5, 2012 Revised: April 18, 2012

exhibit an uncontrolled tapered morphology and are degenerately n-type.21−24 It has been observed that a 2DEG (∼1013− © XXXX American Chemical Society

A

dx.doi.org/10.1021/nl300476d | Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

Letter

Figure 1. SEM images of nontapered undoped InN nanowires grown on Si (111) substrate. (a) Short growth duration nanowires (lengths ∼0.7 μm). (b) Relatively long growth duration nanowires (lengths ∼4 μm).

1a is the scanning electron microscopy (SEM) image of InN nanowire arrays on Si (111) taken with a 45° angle, which exhibit a well-defined hexagonal morphology. This near-perfect morphology and phase purity can also be achieved for wires with relatively large dimensions (see Figure 1b). InN nanowires in this study have lengths in the range of ∼0.7 to ∼4 μm, depending on the growth duration. The nanowires are oriented along the [0001] polar direction, with their sidewalls being nonpolar m-planes. Such nontapered nanowires are nearly free of structural defects (i.e., dislocations and stacking faults),27,28 thereby minimizing unintentional n-type doping due to defects. This could be the prerequisite for one to tune the surface charge properties through controlled n-type doping, which can be analyzed by studying the electronic and optical properties of these pristine nanowires. Micro-PL measurements were performed at both low and room temperatures. The nanowires were optically excited using a semiconductor diode laser (λ = 635 nm) with a 100× objective and a beam size of ∼5 μm. The emitted light from nanowires (through the same 100× objective) was spectrally resolved by a high-resolution spectrometer and detected by a liquid nitrogen cooled InGaAs detector (with a cutoff wavelength of ∼2.2 μm). The signal was collected with a single channel lock-in amplifier. The nontapered undoped InN nanowires, in this work, can exhibit an extremely narrow PL spectral line width of ∼9 meV under an excitation power of ∼0.5 μW at 10 K, shown in Figure 2a. This line width is nearly a factor of 5−10 times smaller, compared to the commonly reported values for the conventional n-type degenerate InN nanowires or films,22,23,26,29−31 suggesting the superior quality of such nontapered InN nanowires. The peak energy position at ∼0.675 eV also agrees reasonably well with the band gap of InN in this temperature range.32−34 Recent studies have suggested that the residual doping of InN can be approximately derived by analyzing the PL spectral line width at low temperatures.35 The same approach was employed here. To first order, for such InN nanowires with a PL spectral line width (Γ) of ∼9 meV, the total electron density, n, at 10 K was estimated to be ∼2.5 × 1016 cm−3 (when the inhomogeneous broadening is neglected).35 However, by considering the inhomogeneous broadening, Γih, a lower value of n can be derived from n0(1 − Γih/Γ)1/α, where α ≈ 0.5 and n0 is the residual dopant concentration. Assuming an inhomogeneous broadening of ∼5 meV (which represents a lower limit on the values normally measured for III−V compound semiconducting nanowires grown on Si) for the presented InN nanowires, we obtain n ∼ 4 × 1015 cm−3.22,36

InN nanowires grown on silicon using angle-resolved X-ray photoelectron spectroscopy (XPS) and PL spectroscopy. We have demonstrated, for the first time, that intrinsic InN can be achieved both within the bulk and at nonpolar InN surfaces. The near-surface Fermi-level was measured to be ∼0.55 eV above the valence band maximum (VBM) for undoped InN nanowires, suggesting the absence of surface electron accumulation and Fermi-level pinning. This result is in direct contrast to the problematic degenerate 2DEG universally observed on grown surfaces of n-type degenerate InN, which arguably presents one of the most fundamental obstacles in nitride device technology development. Furthermore, we have demonstrated that InN nanowire surface electronic properties, including the band structure, Fermi-level, and carrier concentration as well as the associated optical characteristics can be precisely tuned through controlled n-type doping. At relatively high doping levels in this study, the near-surface Fermi-level was found to be pinned at ∼0.95−1.3 eV above the VBM. Through these trends, well captured by the effective mass and ab initio materials modeling, we have unambiguously identified the definitive role of surface doping (by either controlled donor incorporation or surface-defect donors controlled by growth temperature) in tuning the surface charge properties of InN nanowires. This unprecedented tunability at InN surfaces, which had not yet been achieved in any nanowire or film structures, is crucial for many emerging nanoscale electronic and photonic devices. In this experiment, InN nanowires were grown on Si (111) substrates by a Veeco Gen-II radio frequency plasma-assisted molecular beam epitaxial growth system under nitrogen-rich conditions. Prior to loading into the MBE system, the Si substrates were cleaned by hydrofluoric acid and standard solvent solutions. The surface oxide was thermally desorbed at ∼770 °C in situ. In this experiment, a thin (∼ 0.6 nm) In layer was deposited on the substrate surface before introducing nitrogen,27,28 compared to the conventional spontaneous formation of InN nanowires.22 The In layer forms nanoscale droplets at elevated temperatures, which can promote the subsequent formation and nucleation of InN nanowires. The growth conditions for InN nanowires included: a substrate temperature of ∼480 °C, an In flux of ∼6 × 10−8 Torr, a nitrogen flow rate of ∼1.0 sccm, and a RF plasma forward power of ∼350 W. Si cell temperatures of ∼1250 and ∼1350 °C were used for Si-doped InN nanowires. Under optimum growth conditions, InN nanowires can exhibit a nontapered wurtzite structure (i.e., identical top and bottom sizes28 for both undoped and Si-doped InN nanowires). Shown in Figure B

dx.doi.org/10.1021/nl300476d | Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

Letter

ensures that a majority of the signal is derived from nanowire sidewalls. Illustrated in Figure 3a is the measurement of the

Figure 2. Micro-PL spectra of intrinsic InN nanowires with lengths ∼4 μm. (a) Spectrum measured at 10 K under an excitation power of ∼0.5 μW. (b) Power-dependent PL spectra of the same intrinsic InN nanowires measured at 10 K. (c) Excitation power-dependent integrated PL intensity at 10 K derived from (b) (displayed in logarithm scale). Each spectrum in (a) and (b) was normalized by its individual peak intensity and shifted vertically for display purposes.

Since the estimated value may include residual doping, n0, and photogenerated electrons, n − n0, we therefore conclude that such nontapered undoped InN nanowires exhibit an extremely low (less than 4 × 1015 cm−3) residual electron density. This value is nearly 2−3 orders of magnitude smaller than the commonly reported values for InN nanowires and films.22,26 It is also approximately 10 times smaller than the previous report by Chang et al.25 and is the smallest value ever reported in any InN structures. Nevertheless, the derived carrier concentrations need to be further correlated with direct electrical measurements, such as those via field-effect transistor devices37 or space charge limited transport using nanoprobes,38 which are currently under investigation. The high quality of such undoped InN nanowires is further evidenced by the presence of exciton-mediated PL emission. As shown in Figure 2b, the PL peak energy position is nearly independent of the optical excitation power for well over four decades of intensity variation, and the emission spectra are highly symmetric. This is fundamentally distinct from the Mahan excitons in n-type degenerate InN, which generally leads to a PL peak energy significantly larger than the band gap of InN.39 Detailed power-dependent integrated PL intensity study suggests the PL emission is mediated by excitons at low temperatures, i.e., if the PL spectra are mediated by exciton-line emission, then one would expect a linear power dependence in the integrated PL intensity.40 This is indeed the case as shown in Figure 2c, where the integrated PL intensity exhibits a linear dependence with respect to the excitation power over three orders of magnitude. (The deviation from linearity under very high excitation conditions can be attributed to optical heating.) Nevertheless, to clarify the underlying physics of exciton-line emission in InN, further work is needed. The surface charge properties of such undoped InN nanowires were investigated directly by angle-resolved XPS. In this experiment, an X-ray beam impinged upon the nanowire samples with the resulting photoelectrons being collected at a near-zero takeoff angle. This measurement configuration

Figure 3. Angle-resolved XPS spectra and ab initio calculations. Angleresolved XPS spectra were measured from the lateral surfaces (mplane) of [0001]-oriented nontapered (a) undoped and (d) Si-doped InN nanowires, showing energy separations of ∼0.55 eV in (a) and 0.95 eV in (d) between the valence band maximum and near-surface Fermi-level, respectively. The inset of (a) shows the photoelectron spectra of In 3d5/2 and N 1s orbitals measured from the lateral surfaces of undoped nanowire samples, demonstrating the absence of In−O related bonds. The surface band structure and electron distribution are modeled by performing self-consistent effective mass calculations of a 200 nm diameter InN nanowire. Intrinsic nanowires are characterized by the absence of surface band bending and surface 2DEG (shown in (b) and (c), respectively). The Si-doped nanowires, on the other hand, show ∼0.3 eV surface band bending and 2DEG formation (shown in (e) and (f), respectively), due to a surface donor density of 6.7 × 1012 cm−2 in the first two atomic layers. (g) An ab initio calculation showing that the formation energy for In-substitutional Si doping near the InN m-plane surface is significantly lower than that in the bulk.

near-surface Fermi-level relative to the position of the VBM. It can be seen that the near-surface Fermi-level lies at ∼0.55 eV above the VBM, suggesting minimal downward band bending and negligible electron accumulation on the InN nanowire sidewalls. Furthermore, highly symmetric In-3d5/2 and N-1s XPS spectra peaks (shown in the inset of Figure 3a) suggest negligible levels of impurity bonding (such as In−O, N−H, or N−C) associated with electron accumulation.41 To the current knowledge, this is the first demonstration that surface 2DEG formation and Fermi-level pinning are absent at the grown C

dx.doi.org/10.1021/nl300476d | Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

Letter

doping is much lower at the InN m-plane surfacespecifically within the first 0.5 nm. Consequently, Si-dopants are preferentially incorporated in the near-surface region. Similar surface dopant segregation has been widely observed in other nanowires, e.g., refs 2 and 5. We attribute this to the reduced lattice strain imposed by surface donors with respect to bulk donors. The tuning of surface charge properties can also be demonstrated by the PL emission characteristics. We systematically measured the changes in the PL spectra of InN nanowires with controlled Si-doping. Two Si-doped InN nanowire samples grown at Si cell temperatures of ∼1250 and ∼1350 °C were investigated. As the Si-dopants are incorporated, the emission spectra change drastically. Illustrated in Figure 4a, the PL spectra of intrinsic and Si-doped InN

nonpolar surfaces of InN since this prediction was made by Van de Walle et al..18,42 This behavior can be further captured through a selfconsistent effective mass model,43 where the time-independent Schrödinger equation was solved self-consistently assuming an effective mass of 0.055 me, an InN relative dielectric constant of 13, and cylindrical symmetry. As shown in Figure 3b,c, under uniform near-intrinsic doping level, less than 5 × 1015 cm−3, InN surface band bending does not occur at the m-plane (shown in Figure 3b, juxtaposed against the XPS data in Figure 3a). This is illustrated by the spatial distribution of the long wavelength electrons that occupy the CBM under near-intrinsic doping level. As shown in Figure 3c, electrons at the CBM pull away from the surface due to quantum confinement and do not form a surface 2DEG but rather partially deplete the near surface region (Figure 3b). This observation is in direct contrast to the commonly reported 2DEG that forms at the surfaces of conventional n-type degenerate InN nanowires and thin films.13,44,45 Significantly, the achievement of intrinsic InN nanowires makes it possible for one to precisely tune the surface charge properties as well as their optical emission characteristics over a wide range of doping concentrations from intrinsic to the degenerate n-type. This tuning of surface charge properties was first demonstrated by the XPS experiments on Si-doped InN nanowires. We investigated the surface charge properties of nontapered Si-doped InN nanowires grown at a Si cell temperature of ∼1250 °C (an estimated average doping concentration of ∼5 × 1017 cm−3 based on the secondary ion mass spectroscopy measurements on GaN films grown with similar growth rate and with this Si cell temperature). For such Si-doped nanowires it was observed that the near-surface Fermi-level lies ∼0.95 eV above the VBM, shown in Figure 3d. Therefore, given an energy band gap of ∼0.65 eV at room temperature, the near-surface Fermi-level of Si-doped InN nanowires is located ∼0.3 eV above the CBM. This indicates the presence of a high-density 2DEG (>6 × 1012 cm−2), which is significantly larger than the bulk doping concentration. This observation is similar to the commonly reported 2DEG that forms at the surfaces of unintentionally n-type doped InN nanowires and thin films;13,44,45 here, however, it was obtained by the controlled Si-doping. The high surface electron density (discussed above) most likely arises from the segregation of Si donors at the m-plane. This phenomenon is demonstrated through the effective mass calculations presented in Figure 3e,f. Upon raising the bulk doping to ∼5 × 1017 cm−3 (which lies in the range of the average doping concentration obtained at a Si cell temperature of ∼1250 °C), a surface 2DEG arises to screen the segregated surface donors (having a concentration of ∼6.7 × 1012 cm−2) resulting in ∼0.3 eV of surface band bending (which correlates well with the XPS data in Figure 3d). This donor segregation model is strongly supported by atomistic ab initio calculations, where density functional theory (DFT) calculations on a 3 × 3 InN slab 12 atomic layers (16.2 Å) thick were performed within VASP.46 At this thickness the slab work function converged to the bulk value. A plane wave basis at a cutoff energy of 450 eV was employed. Pseudopotentials and nonlinear core corrections were applied within the local spin density approximation (LSDA). Computations were performed on a 3 × 2 × 1 k-point grid within a 15.2 × 17.0 × 31.2 Å3 supercell. The forces on all atoms were converged to 0.01 eV/Å. As shown in Figure 3g, it can be seen that the formation energy for In-substitutional Si-

Figure 4. Micro-PL spectra of intrinsic InN nanowires and wires with various Si-doping levels. The nanowire lengths are ∼0.7 μm. (a) Measured at 20 K under an excitation power of ∼50 μW. (b) Measured at room temperature under an excitation power of ∼9 mW. The spectra of Si-doped InN nanowires in (a) and (b) were normalized by the peak intensity of the intrinsic InN nanowires and were shifted vertically for display purposes.

nanowires are measured at 20 K under an excitation power of ∼50 μW. With increasing doping, it was observed that the PL peak energy shifted from ∼0.68 to ∼0.73 eV and the spectral line width broadened from ∼19 to ∼51 meV. These changes in emission characteristics can be understood as follows. The significant blue-shift at 20 K with increasing Si-doping can be explained by Mahan excitons in n-type degenerate InN, i.e., radiative recombination primarily involving electrons at the Fermi-level in the conduction band and localized photogenerated holes.39,47 Furthermore, the photogenerated electrons and holes are spatially separated by the downward band bending at surfaces (see Figure 3e), which leads to a drastic reduction in the radiative recombination efficiency and significantly reduced PL intensity in Si-doped InN nanowires. Additionally, the downward band bending at surfaces will generate a large variation in the energetic spacing between the Fermi-level and the CBM in the near-surface region (see Figure D

dx.doi.org/10.1021/nl300476d | Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

Letter

by Compute Canada through CFI. Some of the XPS studies were performed at Sandia and funded by the U.S. Department of Energy Office of Basic Energy Sciences, Division of Materials Science and Engineering. Sandia National Laboratories is a multiprogram laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-AC0494AL85000

3e), which subsequently gives rise to the observed PL spectral broadening with increased Si-doping. The strong correlation between PL emission spectra and doping suggests that the primary radiative recombination process evolves from the nanowire bulk to the near-surface region with enhanced surface doping, which is further supported by PL spectra obtained under relatively high excitation conditions at room temperature. Illustrated in Figure 4b, the intrinsic InN nanowires exhibit a PL spectrum with a well-defined single peak, while the Si-doped InN nanowires grown at a Si cell temperature of ∼1250 °C exhibit two emission peaks centered at ∼0.65 and ∼0.73 eV, respectively; with further increasing Si-doping (i.e., nanowires grown at Si cell temperature of ∼1350 °C), the emission peak at ∼0.65 eV is greatly suppressed, and the emission peak at ∼0.73 eV dominates. The transition to two emission peaks can be attributed to radiative carrier recombination in both the inner core and the near-surface regions of the nanowires. These regions are respectively characterized by the presence of moderate and very high electron densities, due to the enhanced surface doping (as illustrated in Figure 3e,f). As final remarks, we have further demonstrated, both experimentally and theoretically, that the tuning of the surface charge properties can be achieved by varying the nanowire morphology through controlling growth parameters (see Supporting Information). Hence, the present study provides unambiguous evidence that the commonly measured large surface electron accumulation and Fermi-level pinning is not a fundamental property of InN. It reconciles, to a great extent, the intensive debate on the origins of the 2DEG commonly observed at the nonpolar surfaces of InN. Most importantly, we have shown, for the first time, that the surface charge properties (such as the 2DEG formation) as well as the optical emission characteristics of a semiconducting nanowire can be precisely tuned through controlled n-type doping and wire morphology. The unprecedented tunability of nanowire surface electronic structure, by designing materials at the atomic scale, promises an entirely new avenue for the development of silicon integrated nanoscale electronic, photonic, and biochemical devices.

■

■

(1) Ho, J. C.; Yerushalmi, R.; Jacobson, Z. A.; Fan, Z.; Alley, R. L.; Javey, A. Nat. Mater. 2008, 7, 62. (2) Perea, D. E.; Hemesath, E. R.; Schwalbach, E. J.; Lensch-Falk, J. L.; Voorhees, P. W.; Lauhon, L. J. Nat. Nanotechnol. 2009, 4, 315. (3) Allen, J. E.; Perea, D. E.; Hemesath, E. R.; Lauhon, L. J. Adv. Mater. 2009, 21, 3067. (4) Koren, E.; Berkovitch, N.; Rosenwaks, Y. Nano Lett. 2010, 10, 1163. (5) Xie, P.; Hu, Y. J.; Fang, Y.; Huang, J. L.; Lieber, C. M. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 15254. (6) Lu, W.; Lieber, C. M. J. Phys. D: Appl. Phys. 2006, 39, R387. (7) Shin, J. C.; Kim, K. H.; Yu, K. J.; Hu, H. F.; Yin, L. J.; Ning, C. Z.; Rogers, J. A.; Zuo, J. M.; Li, X. L. Nano Lett. 2011, 11, 4831. (8) Chen, R.; Tran, T. T. D.; Ng, K. W.; Ko, W. S.; Chuang, L. C.; Sedgwick, F. G.; Chang-Hasnain, C. Nat. Photonics 2011, 5, 170. (9) Fukata, N.; Ishida, S.; Yokono, S.; Takiguchi, R.; Chen, J.; Sekiguchi, T.; Murakami, K. Nano Lett. 2011, 11, 651. (10) King, P. D. C.; Veal, T. D.; Payne, D. J.; Bourlange, A.; Egdell, R. G.; McConville, C. F. Phys. Rev. Lett. 2008, 101, 116808. (11) Piper, L. F. J.; Colakerol, L.; King, P. D. C.; Schleife, A.; ZunigaPerez, J.; Glans, P. A.; Learmonth, T.; Federov, A.; Veal, T. D.; Fuchs, F.; Munoz-Sanjose, V.; Bechstedt, F.; McConville, C. F.; Smith, K. E. Phys. Rev. B 2008, 78, 165127. (12) Van de Walle, C. G.; Neugebauer, J. Nature 2003, 423, 626. (13) Mahboob, I.; Veal, T.; McConville, C.; Lu, H.; Schaff, W. Phys. Rev. Lett. 2004, 92, 036804. (14) King, P. D. C.; Veal, T. D.; McConville, C. F.; Fuchs, F.; Furthmuller, J.; Bechstedt, F.; Schley, P.; Goldhahn, R.; Schormann, J.; As, D. J.; Lischka, K.; Muto, D.; Naoi, H.; Nanishi, Y.; Lu, H.; Schaff, W. J. Appl. Phys. Lett. 2007, 91, 092101. (15) Darakchieva, V.; Schubert, M.; Hofmann, T.; Monemar, B.; Hsiao, C. L.; Liu, T. W.; Chen, L. C.; Schaff, W. J.; Takagi, Y.; Nanishi, Y. Appl. Phys. Lett. 2009, 95, 022103. (16) Colakerol, L.; Piper, L. F. J.; Fedorov, A.; Chen, T. C.; Moustakas, T. D.; Smith, K. E. Euro. Phys. Lett. 2008, 83, 47003. (17) Noguchi, M.; Hirakawa, K.; Ikoma, T. Phys. Rev. Lett. 1991, 66, 2243. (18) Van de Walle, C. G.; Segev, D. J. Appl. Phys. 2007, 101, 081704. (19) Wu, C. L.; Lee, H. M.; Kuo, C. T.; Chen, C. H.; Gwo, S. Phys. Rev. Lett. 2008, 101, 106803. (20) Ebert, P.; Schaafhausen, S.; Lenz, A.; Sabitova, A.; Ivanova, L.; Dahne, M.; Hong, Y. L.; Gwo, S.; Eisele, H. Appl. Phys. Lett. 2011, 98, 062103. (21) Lazic, S.; Gallardo, E.; Calleja, J. M.; Agullo-Rueda, F.; Grandal, J.; Sanchez-Garcia, M. A.; Calleja, E.; Luna, E.; Trampert, A. Phys. Rev. B 2007, 76, 205319. (22) Stoica, T.; Meijers, R.; Calarco, R.; Richter, T.; Sutter, E.; Luth, H. Nano Lett. 2006, 6, 1541. (23) Shen, C.; Chen, H.; Lin, H.; Gwo, S.; Klochikhin, A.; Davydov, V. Appl. Phys. Lett. 2006, 88, 253104. (24) Kuykendall, T.; Ulrich, P.; Aloni, S.; Yang, P. Nat. Mater. 2007, 6, 951. (25) Liu, J.; Cai, Z. H.; Koley, G. J. Appl. Phys. 2009, 106, 124907. (26) Segura-Ruiz, J.; Garro, N.; Cantarero, A.; Denker, C.; Malindretos, J.; Rizzi, A. Phys. Rev. B 2009, 79, 115305. (27) Chang, Y. L.; Mi, Z. T.; Li, F. Adv. Funct. Mater. 2010, 20, 4146.

ASSOCIATED CONTENT

S Supporting Information *

Tuning the surface charge properties by controlling InN nanowire morphology. This material is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] 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, Canada Foundation for Innovation (CFI), U.S. Army Research Office, and McGill University. Part of the work was performed in the Microfabrication Facility at McGill University. Computational resources were provided by the U.S. National Science Foundation Network for Computational Nanotechnology and E

dx.doi.org/10.1021/nl300476d | Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

Letter

(28) Chang, Y. L.; Li, F.; Fatehi, A.; Mi, Z. T. Nanotechnology 2009, 20, 345203. (29) Arnaudov, B.; Paskova, T.; Paskov, P. P.; Magnusson, B.; Valcheva, E.; Monemar, B.; Lu, H.; Schaff, W. J.; Amano, H.; Akasaki, I. Phys. Rev. B 2004, 69, 115216. (30) Fu, S. P.; Chen, T. T.; Chen, Y. F. Semicond. Sci. Technol. 2006, 21, 244. (31) Wu, J. Q. J. Appl. Phys. 2009, 106, 011101. (32) Wu, J.; Walukiewicz, W.; Yu, K.; Ager, J.; Haller, E.; Lu, H.; Schaff, W. Appl. Phys. Lett. 2002, 80, 4741. (33) Davydov, V.; Klochikhin, A.; Seisyan, R.; Emtsev, V.; Ivanov, S.; Bechstedt, F.; Furthmuller, J.; Harima, H.; Mudryi, V.; Aderhold, J.; Semchinova, O.; Graul, J. Phys. Status Solidi B 2002, 229, R1. (34) Johnson, M.; Lee, C.; Bourret-Courchesne, E.; Konsek, S.; Aloni, S.; Han, W.; Zettl, A. Appl. Phys. Lett. 2004, 85, 5670. (35) Moret, M.; Ruffenach, S.; Briot, O.; Gil, B. Appl. Phys. Lett. 2009, 95, 031910. (36) Schlager, J. B.; Sanford, N. A.; Bertness, K. A.; Barker, J. M.; Roshko, A.; Blanchard, P. T. Appl. Phys. Lett. 2006, 88, 213106. (37) Huang, Y.; Duan, X.; Cui, Y.; Lieber, C. Nano Lett. 2002, 2, 101. (38) Leonard, F.; Talin, A. A.; Swartzentruber, B. S.; Picraux, S. T. Phys. Rev. Lett. 2009, 102, 106805. (39) Feneberg, M.; Daubler, J.; Thonke, K.; Sauer, R.; Schley, P.; Goldhahn, R. Phys. Rev. B 2008, 77, 245207. (40) Schmidt, T.; Lischka, K.; Zulehner, W. Phys. Rev. B 1992, 45, 8989. (41) Nagata, T.; Koblmuller, G.; Bierwagen, O.; Gallinat, C. S.; Speck, J. S. Appl. Phys. Lett. 2009, 95, 132104. (42) Mahboob, I.; Veal, T. D.; Piper, L. F. J.; McConville, C. F.; Lu, H.; Schaff, W. J.; Furthmuller, J.; Bechstedt, F. Phys. Rev. B 2004, 69, R201307. (43) Assad, F.; Banoo, K.; Lundstrom, M. Solid-State Electron. 1998, 42, 283. (44) Linhart, W. M.; Veal, T. D.; King, P. D. C.; Koblmuller, G.; Gallinat, C. S.; Speck, J. S.; McConville, C. F. Appl. Phys. Lett. 2010, 97, 112103. (45) Segura-Ruiz, J.; Molina-Sanchez, A.; Garro, N.; Garcia-Cristobal, A.; Cantarero, A.; Iikawa, F.; Denker, C.; Malindretos, J.; Rizzi, A. Phys. Rev. B 2010, 82, 125319. (46) Kresse, G.; Joubert, D. Phys. Rev. B 1999, 59, 1758. (47) Mahan, G. D. Phys. Rev. 1967, 153, 882.

F

dx.doi.org/10.1021/nl300476d | Nano Lett. XXXX, XXX, XXX−XXX