Photonic–Plasmonic Coupling of GaAs Single Nanowires to Optical

Apr 17, 2014 - The single NW is a well-known optical system, accurately modeled by analytical ... (38) In the FDTD calculations the simulation grid si...
0 downloads 5 Views 4MB Size
Letter pubs.acs.org/NanoLett

Photonic−Plasmonic Coupling of GaAs Single Nanowires to Optical Nanoantennas Alberto Casadei,†,¶ Emanuele F. Pecora,‡,#,¶ Jacob Trevino,§,⊥,¶ Carlo Forestiere,‡ Daniel Rüffer,† Eleonora Russo-Averchi,† Federico Matteini,† Gozde Tutuncuoglu,† Martin Heiss,† Anna Fontcuberta i Morral,† and Luca Dal Negro*,‡,§ †

Laboratoire des Matériaux Semiconducteurs Ecole, Polytechnique Fédérale de Lausanne, 1015 Lausanne, Switzerland Department of Electrical and Computer Engineering & Photonics Center, Boston University, 8 Saint Mary Street, Boston, Massachusetts 02215, United States § Division of Materials Science and Engineering, Boston University, 15 Saint Mary’s Street, Brookline, Massachusetts 02446, United States ‡

S Supporting Information *

ABSTRACT: We successfully demonstrate the plasmonic coupling between metal nanoantennas and individual GaAs nanowires (NWs). In particular, by using dark-field scattering and second harmonic excitation spectroscopy in partnership with analytical and full-vector FDTD modeling, we demonstrate controlled electromagnetic coupling between individual NWs and plasmonic nanoantennas with gap sizes varied between 90 and 500 nm. The significant electric field enhancement values (up to 20×) achieved inside the NWnanoantennas gap regions allowed us to tailor the nonlinear optical response of NWs by engineering the plasmonic near-field coupling regime. These findings represent an initial step toward the development of coupled metal−semiconductor resonant nanostructures for the realization of next generation solar cells, detectors, and nonlinear optical devices with reduced footprints and energy consumption. KEYWORDS: Semiconductor nanowires, plasmonics, near-field optics, light coupling

T

systems, little is currently known on their optical coupling regime and synergistic properties. Recently, colloidal arrays of plasmonic nanoparticles and lithographically defined metallic nanocylinders coupled to semiconductor NWs have been explored as novel metal−semiconductor interacting systems that enhance the optical response of their individual components.22,23 The combination of the mature semiconductor NWs platform with the nanoplasmonics technology could potentially open the way to novel technological applications that leverage strongly confined optical fields in order to manipulate the linear and nonlinear optical responses (i.e., scattering, absorption, emission, harmonic generation) of resonant semiconductor structures at the nanoscale. In particular, semiconductor NWs optically coupled to plasmonic nanoantennas with lithographically defined morphologies may become the basic building blocks for future high-efficiency solar cells, ultrafast optical switches, and modulators and nanoscale photodetectors with dramatically reduced energy consumption. In this paper, we investigate the resonant coupling of semiconductor NWs and plasmonic antennas. In particular, we

he optical properties of semiconductor nanowires (NWs) are currently at the center of an intense research effort due to their potential applications in a number of nanoscale optoelectronic devices, such as tunable and enhanced light sources,1−3 solar cells4−8 and photodetectors,9,10 optical switches,11 and nonlinear devices and modulators.12 NWs with engineered composition, size, and morphology offer the possibility to control the electronic structure and the linear and nonlinear optical properties of semiconductor materials.13 Recently, the engineering of metal−semiconductor NWs that support distinctive structural resonances, such as the ones predicted by the classical Mie theory,14 has been proven as a convenient pathway to enhance light-matter coupling.15 Moreover, resonant metallic nanostructures supporting traveling or localized SSPs (i.e., collective oscillations of free electrons confined in one or more spatial dimensions at the nanoscale) have been thoroughly investigated as a powerful approach to manipulate optical radiation at the subwavelength scale.16−20 In particular, plasmonic nanoparticle arrays and nanoantennas have shown the ability to strongly concentrate and increase the intensity of local electromagnetic fields over engineered nanoscale spatial domains and spectral bandwidths.21 However, although a significant amount of work has been devoted to understand and manipulate the optical responses of individual semiconductor NWs and plasmonic © 2014 American Chemical Society

Received: November 16, 2013 Revised: March 31, 2014 Published: April 17, 2014 2271

dx.doi.org/10.1021/nl404253x | Nano Lett. 2014, 14, 2271−2278

Nano Letters

Letter

min. A few drops of this solution were then released onto a prepatterned fused silica substrate. We used an optical microscope to localize the NWs position on the wafer and the nanoantennas have been designed around the NWs using a custom-made fitting software.35 The pattern encodes the relative position on the wafer and uses circular shapes for deep submicrometer alignment. The position is determined via shape recognition algorithms leading to an accuracy of approximately 50 nm. We used electron beam lithography (Vistec EBPG5000) to position the nanoantennas in close proximity to the NW surface and along its axis. Before the exposure, an MMA and PMMA resist layer were spin coated on the substrate and the evaporation of 40 nm of Cr was needed in order to dissipate the electrons coming from the electron beam. The Cr was then etched away and the evaporation of 5 nm of Ti and 30 nm of Au was performed after the developing of the resist. Figure 1a displays a schematic of the structure of a representative plasmonic antenna-coupled NW sample inves-

use Au dimer nanoantennas to couple light into GaAs NWs. The precise control of the nanoantennas position leads to the opportunity of engineering coupled systems able to enhance the absorption of determined wavelength and polarization. GaAs NWs and BaTiO3 nanoparticles have been recently studied with second-harmonic generation (SHG) as single systems.24,25 The distinctive optical resonances of NWnanoantennas system enhance the pumping electric field at specific wavelengths in the near-infrared range. As a consequence, the second harmonic generation in the visible spectral range can be strongly enhanced. We probe the system resonances by dark-field scattering and second harmonic excitation (SH-E) spectroscopy.26,27 We demonstrate resonant near-field optical concentration of radiation into individual NWs in good agreement with full-vector three-dimensional (3D) electromagnetic modeling based on the finite difference time domain (FDTD) method. In particular, the metal nanoantennas enhance the electric field outside and on the surface of the NW associated with the component of the laser pumping light having transverse polarization with respect to the NW axis. This ability introduces the possibility to control the absorption of different polarizations by a single NW, which constitutes the first step in the realization of novel photodetectors with designed spectral and polarization responses at the nanoscale. Thanks to the recent advancements in semiconductor growth and processing technology, semiconductor NWs can be obtained with excellent control on morphology and doping concentration.28,29 In this study, GaAs nanowires were grown on Si(111) undoped substrates by the Ga-assisted method in a DCA P600 solid source MBE system,30,31 under a rotation of 7 rpm, a flux of Ga equivalent to a planar growth rate of 0.3 A/s, a substrate temperature of 640 °C, and a V/III beam equivalent pressure ratio of 50. These conditions lead to a diameter of around 55 nm and a length of 12 μm. After the growth of the NW core, 4 nm AlxGa1−xAs and 3 nm intrinsic GaAs shells are obtained by turning growth conditions to those corresponding to two-dimensional epitaxy,32 leading to an overall NW diameter of around 70 nm. The concentration of Al (x) in the AlGaAs shell is around 0.3. This passivation layer avoids the detrimental effect of nonradiative surface trap states on the GaAs NW, offering opportunities to investigate the intrinsic properties of the material.33,34 We also investigated p-type Si doped GaAs NWs. In this case, the core has been grown with the same parameters described above and with an additional flux of Si that leads to a doping concentration of around 5 × 1018 cm−3. These NWs have no AlGaAs capping shell, they are 12 μm long and their diameter measures 70 nm. In our work we compared both types of NWs and obtained identical outcome. This comparison was important in views of in laterstage experiments where the NWs will be electrically contacted. In this case, it is only possible to obtain ohmic contacts if the NWs are doped. The NWs grow along the (111) direction, with six facets belonging to the {110} family. They present mainly a zinc blende crystal structure with 0−4 twins per micrometer as shown in the high resolution transmission electron microscopy in Supporting Information Figure SI1. The zinc blende structure guarantee there is no orientation dependence in the SH-E, and in the experiments, we also use circular polarized light in order to avoid the effect of the directional polarization. The NWs were removed from the as-grown substrate dissolving them in isopropanol in an ultrasound bath for 1

Figure 1. (a) Schematic of GaAs NW coupled to nanoantenna with all the geometrical parameters. SEM micrographs of (b) NW-coupled antenna with d = 500 nm, (c) NW-coupled antenna with d = 200 nm, (d) NW-coupled antenna with d = 100 nm, and (e) NW-coupled antenna with d = 90 nm.

tigated in this study. The individual nanoantennas consist of dimers of Au cylindrical nanodisks with 200 nm diameter and 30 nm height. The nanoantenna feed-gap regions host the individual NWs. We varied the gap distance from 500 to 90 nm in order to investigate the effect of tunable plasmonic near-field coupling at the level of an individual NW. It is important to realize that for the 90 nm gap, the average distance between the nanodisks in the dimer antennas and the NWs is of the order of just few nanometers, requiring an extremely precise position control. Actually, the real devices present small deviations from the ideal geometry. These discrepancies mostly come from the uncertainty in the NW sizes and the alignment between wire and nanoantenna arrays. In particular, the distance between a nanoantenna disk and the NW can change up to 40 nm from 2272

dx.doi.org/10.1021/nl404253x | Nano Lett. 2014, 14, 2271−2278

Nano Letters

Letter

Figure 2. The left panel shows FDTD simulated scattering cross section for a single NW (a) and NW-nanoantennas coupled system with d of 500 nm (b), 200 nm (c), and 90 nm (d). In panel a, dashed line is the scattering cross-section by the analytical Mie theory. In the right panel, correspondent experimentally measured dark-field scattering (e), (f), (g) and (h). The dashed line in panel a is the analytical calculation of the Mie scattering resonances in an isolated wire of infinite length. Dashed lines in panels (f-h) represent the deconvolution of the experimental spectra into Gaussian functions. Note that the x scale is different for the numerical and experimental data.

optical dispersion of Au nanocylinders is modeled using experimental data as reported in literature.38 In the FDTD calculations the simulation grid size surrounding the Au nanoparticles and the NW was set to 2 nm in all three dimensions. Plane wave excitation at normal incidence (i.e., along the z axis) was used, as labeled in Figure 1a. Perfectly matched layer (PML) boundary conditions were used to ensure perfect absorption of electromagnetic radiation at the simulation boundaries. The calculated scattering cross sections, which are obtained by the average of two calculations carried out under incident light longitudinally and transversely polarized with respect to the NW axis, were compared with experimentally measured dark-field scattering spectra obtained under unpolarized light on the fabricated devices. Dark-field scattering spectra were acquired using an Olympus IX71 microscope with a 50× long-working distance objective (N.A 0.5) for collection under dark field illumination. The samples were illuminated with a 100 W halogen lamp which was focused on the sample plane using a dark field condenser (N.A 0.92−0.80). Scattered light from nanostructures was collected and analyzed with an Andor Shamrock 750 mm focal length spectrometer using a 600 lines/mm grating blazed at 500 nm. The spectra were recorded with an Andor iDus CCD camera (DU420A), then background corrected, and finally corrected for the excitation profile of the lamp and grating efficiency. Additionally, spatial filtering at the CCD detector was used for background noise reduction. Because of our system limitations, we show experimental data in the range 400−1000 nm, whereas simulations extend down to 1400 nm. In Figure 2 (left panels) the calculated scattering cross sections by the analytical Mie theory and the 3D FDTD simulations are plotted for the isolated NW (a), as well as the antenna coupled NWs (b−d). In the right panels, we show the corresponding experimental data. We notice first that the analytical solution (panel a, dashed line) for the single NW features two main resonances in the investigated range. A similar behavior is observed in the FDTD results (solid line), which appear slightly red-shifted due

the ideal geometry leading in some cases to the superposition of the gold disk and the nanowire (Supporting Information Figure SI2). Nevertheless, these NW misalignments do not affect our experimental measurements. In fact, in our SH-E setup, we avoid any local deviation by focusing the laser light on the center of each system. Misalignments, if any, are negligible at these locations. Moreover, some deviations arise from the actual nanowire size, as we show in the Supporting Information. In fact, we observed a slight tapering of our nanowires (diameter variation of about 5 nm from one side to the other of the wire) and a statistical distribution of the nanowire diameter. In our optical measurements, we analyzed nanowires having diameter in the range between 61 and 77 nm. The longitudinal spacing (i.e., x direction) of the nanoantennas along the length of the NWs (i.e., the array grating constant Λx) is kept fixed (Λx = 840 nm). In Figure 1 we also show the SEM images of the devices analyzed as a function of the gap distance: d = 500 nm (b), d = 200 nm (c), d = 100 nm (d), and d = 90 nm (e). In order to understand the electromagnetic response of our devices, we first investigated dark field scattering of individual NWs as well as antenna coupled NWs, both numerically and experimentally. The single NW is a well-known optical system, accurately modeled by analytical scattering theory in the limit of large aspect ratio. The scattering cross sections of individual NWs have been calculated using analytical Mie theory under plane wave normal incidence (i.e., perpendicular to the NW axis) considering a NW of circular cross section and incident unpolarized radiation. In this approach, which we have employed to better understand the physical origin of the scattering spectra of NWs, the fused silica substrate is considered as an effective host medium with average dielectric constant. On the other hand, the responses of NWs on an actual silica substrate as well as of the complex antenna-coupled NW structures were modeled by 3D FDTD analysis using the commercial software package Lumerical FDTD Solutions.36 The material dispersion parameters of the GaAs NWs and the fused silica substrate are derived from Palik,37 whereas the 2273

dx.doi.org/10.1021/nl404253x | Nano Lett. 2014, 14, 2271−2278

Nano Letters

Letter

Figure 3. Calculated electric near-field distribution of the nanoantenna coupled NW systems plotted in the horizontal plane of the array (halfway up the cylinder in the z direction). Electric field intensity for the longitudinal (a) and transverse (b) polarization with respect to the wire for d = 500 nm and λ = 850 nm. In (c), (d), (e), (f), and (g) is represented the longitudinal and transverse near-field distribution for d = 90 nm. In the case of (c) and (d), λ is 735 nm, and for (e) and (f), λ is 1165 nm. The magnified region denoted by the dashed box in panel f is plotted in panel g.

refer to this situation as the “weak plasmonic coupling regime”. On the other hand, for the closest dimers separations, a new peak at longer wavelengths clearly emerges. This trend is confirmed by the experimental data, where for d = 500 nm (Figure 2f) we observe only one peak, but for d = 90 nm (Figure 2g and h), a broad band starts to appear in the infrared region. Because of the limited grating efficiency of our setup in this spectral region, we can only appreciate the onset of this new peak in Figure 2. In order to better understand the coupling behavior of the complex antenna/NW system, we performed 3D FDTD simulations of the spatial distribution of the electric near-field amplitude for the single NW and the different configurations of the coupled system. The FDTD electric near-field distributions calculated in the horizontal plane of the array are plotted in Figure 3 for the case of NW/antennas system with d = 500 nm at the peak of the calculated scattering (λ = 850 nm) and d = 90 nm for λ = 735 nm and λ = 1165 nm. The field is plotted for (a, c, e) longitudinal and (b, d, f, g) transverse polarizations with respect to the NW, as indicated in the figure by the white arrows. Because of the particular geometry of the NWs, several micrometers long in one direction and with diameter of subwavelengths dimension, the coupling with the incident light is very polarization sensitive. In high aspect ratio NWs, the longitudinal polarization is always more efficient then the transverse polarization in establishing the internal EM field, and the smallest is the NW diameter the largest is the anisotropy in the absorption behavior.40,41 In the case of the weak plasmonic coupling regime (Figure 3a and b), the nanoantennas have only a small influence on the NW response for both longitudinal and transverse polarizations. However, we can observe that EM field is concentrated into the NW only in the case of longitudinal polarization, in agreement with the fact that in this regime, the nanoantennas are too separated from the NWs to alter significantly their polarization sensitive optical behavior. A weak field modulation due to the nanoantennas can be appreciated in the NW region in between the two metal

to the influence of the substrate polarization. At shorter wavelengths, a narrow peak arises at 450 nm, which corresponds to a resonant feature in the GaAs optical constants.37 On the other hand, a broad resonant peak centered on 850 nm is also observed in Figure 2a, which is due to the NW leaky modes excited by the longitudinal light polarization. We notice that the experimental data (Figure 2e) exhibit significantly broadened spectral features compared to the calculated spectra due to the finite cone of excitation angles used in the experimental dark-field configuration. Additionally, we observe a shift of the main peak. There are many possible sources of nonideality in the samples that cannot be fully considered in the simulations, although they may cause a shift of the semiconductor structural resonances. In particular, we used the nominal (bulk) optical constant values, which can be different at the nanoscale for the fabricated NW and the underlying substrate may display spatial inhomogeneity of the refractive index. Moreover, we considered in the simulations NWs with circular cross sections, whereas the fabricated NWs have hexagonal ones and exhibit nanoscale sections of crystallographic phases39 as well as surface states. Finally, the radius of the fabricated NW may slightly differ (few nanometers) from the nominal values used in the simulations. However, the experimental scattering spectrum of the single NW still qualitatively matches the theoretical one obtained by the Mie theory and the numerical results by the FDTD. In Figure 2, we also investigate the NW response in the presence of the plasmonic nanoantennas. In particular, we notice that by decreasing the antennas gap distance d, we can gradually tune the electromagnetic response of a single NW and achieve strong near-field plasmon (i.e., quasistatic) coupling. The numerical calculations show additional bands appearing in the presence of the nanoantennas, which are related to the resonances of the antenna array as well as the coupling behavior of the combined antenna/NW system. In particular, in Figure 2b, one prominent peak emerges at around 850 nm for d = 500 nm, where the NW-nanoantenna coupling is weak. We 2274

dx.doi.org/10.1021/nl404253x | Nano Lett. 2014, 14, 2271−2278

Nano Letters

Letter

Figure 4. Cross-sectional average (red) and maximum (black) electric field enhancement values inside the region of the NW under incident light with both transverse and longitudinal polarization. The graphs refer to (a) isolated NW, (b) isolated dimer pair with d = 90 nm, (c) antenna coupled NW with d = 500 nm, and (c) antenna coupled NW with d = 90 nm. Insets are electric field cross sections, plotted at 1165 nm in log-scale for each configuration.

the maximum electric field value (black) and the spatially averaged electric field (red) calculated inside the NW region as a function of the incident wavelength (Figure 4a), an isolated dimer antenna with d = 90 nm (Figure 4b), the coupled NW/ antennas system with d = 500 nm (Figure 4c), and with d = 90 nm (Figure 4d). For the case of isolated NW, the electric field intensity inside the NW is very low for all wavelengths, around 0.5 of the incident light intensity. In Figure 4b, the dimer antenna shows a plasmonic resonance and concentrates the electric field within its small feed-gap region. In the cross sectional plane analyzed the nanoantennas are contributing to enhance only the electric field induced by the transverse polarized light as we demonstrate in Figure 3. For the coupled system with d = 500 nm, the NW and the nanoantennas show only weak coupling and the field is slightly within the NW. On the other hand, for the coupled system with d = 90 nm the field peaks inside the NW at 1165 nm and its polarization-averaged value reaches an enhancements of 5.5 times (with respect to the incident field) close to the NW surface. The electric field spatial intensity profiles are also shown in the insets in Figure 4 at 1165 nm. These results show that the designed nanostructures offer the possibility to tailor the electromagnetic coupling regime between NWs and nanoantennas. Moreover, we have shown that the dimer antennas can strongly concentrate incident light with transverse polarization, which is normally only weakly absorbed, into the individual NWs, offering the possibility to manipulate their polarization response by tuning the gaps of the nanoantennas. This new functionality can make these coupled systems ideal candidates for engineered nanophotodetectors and more efficient NW-based solar cells. To fully support this picture with additional experimental results extended over a broader spectral range, we performed

disks. In these spatial regions the field is less intense than in the other NW regions and the effect becomes stronger in the case of the strong plasmonic coupling regime (Figure 3c, d, e, f). A close-up view of calculated electric near-field distribution of the nanoantenna coupled NW systems is shown in Supporting Information Figure SI8. Figure 3c and d plots the near-field distribution for the case of strong coupling d = 90 nm at the first scattering resonance located at λ = 735 nm. Here, the nanoantennas clearly couple to the NW for both polarizations. In the longitudinal polarization, the plasmonic antennas locally decrease the electric field intensity that was previously concentrated into the NW. This interference behavior leads to a confinement of the electric field between one dimer and the other, which is clearly visible in Figure 3c. On the other hand, for the transverse polarization, the field is strongly localized by the nanoantennas into their feed-gap regions and not along the NW as for the longitudinal polarization. In the case of longer wavelength, λ = 1150 nm, the near-field distribution for transverse polarization has a maximum peak that corresponds to an enhancement of the electric field intensity of approximately 20 times compared to the incident one (Figure 3f, g). The coupling here is so strong that the electric field, usually confined outside the NW surface for this polarization, begins to penetrate into the NWs, as we can see from the magnified view in Figure 3g. This leads to a new coupled mode of the system, confirmed by the second harmonic measurements that will be discussed later in the article. To further investigate the electric field enhancement in the NW due to incident light with both transverse and longitudinal polarizations, we calculated the cross-sectional (z−y plane) electric field distribution along the dimer axis. Figure 4 shows 2275

dx.doi.org/10.1021/nl404253x | Nano Lett. 2014, 14, 2271−2278

Nano Letters

Letter

second harmonic excitation (SH-E) spectroscopy.26,27 This technique provides a unique method to experimentally probe and analyze the near-field resonances of the coupled NW/ nanoantennas systems and enables us to extend this study to the NIR region, which is not fully accessible by our dark-field scattering experiments. In the SH-E experiment, ultrafast laser pulses are generated by a tunable mode locked Ti:sapphire laser with a pulse duration of 100 fs and a duty cycle of 82 MHz. The beam is modulated at 67 Hz by a mechanical chopper with a 50% duty cycle and is steered with Ag mirrors. We use the Ti:sapphire output in the range 740−1040 nm. We further extend the investigated range down to 1300 nm by the use of a tunable optical parametric oscillator system (SpectraPhysics Inspire) pumped by the Ti:sapphire laser operated at 820 nm. The OPO output has the same temporal width and repetition rate of the pumping laser and is tunable between 340 and 2200 nm. A 700 nm long pass filter prior to the sample removes any SHG signal produced by the optical components. Additionally, the linearly polarized beam is passed through a quarter-wave plate, creating a circularly polarized beam to allow for consistent pumping conditions regardless of the NW’s orientation on the substrate. The excitation beam is then focused on the nanostructure to a spot size of approximately 5 μm by a 50× objective (NA 0.5) from the backside of the sample. The SHG is then collected in transmission mode with a 100× objective (NA 0.8) and is detected with a photon multiplier tube (PMT) and a lock-in amplifier (Oriel Merlin) after passing through a monochromatic (Cornerstone 260 1/4 m, 1200 lines/mm, 500 nm blaze). This defines an acceptance angle of ±30°, which enables the collection of intense SHG signal in our experimental configuration. The excitation beam is removed with a 670 nm short pass filter prior to the monochromator. A CCD coupled to a reflection microscope is used for alignment. A removable mirror is placed in front of the camera to direct the SHG signal to the detector. The time averaged excitation power is kept constant at 5 mW for all measurements. Absolute values of the SH power have been measured using the procedure already discussed elsewhere.42,43 Figure 5a illustrates a schematic of the experimental setup utilized for the SH-E analysis. Data are collected from single systems. As a consequence, an estimation of the statistical error bars is not possible. However, we repeated several times the same measurements on each system, always obtaining consistent results, as shown in Supporting Information Figure SI6. The second harmonic spectra are collected for varying pumping wavelengths and are shown in Figure 5b in the case of the sample with the closest distance between the wire and the disks (d = 90 nm). In Figure 5b, we plot for this representative sample the peak wavelengths of the measured SHG spectra as a function of the pump wavelength. The linear fit to these data features a slope of 0.5, as expected for the SHG process. The SH-E spectra measured for the investigated samples are plotted in Figure 6. First, the spectrum of an isolated NW features a prominent peak localized at 880 nm (Figure 6a). This is identified with the two-photon resonance of the E1 interband transition in GaAs bulk material.44,45 No other resonances are observed in the near-IR wavelengths. We note here that some SH emission related to the coupled NW-nanoantennas systems could in principle be generated by the gold nanodisks.46−48 However, in our structure, by comparing the SH-E of an isolated NW (Figure 6a) and the coupled system for d = 500 nm (Figure 6 b), it becomes evident that the SH-E contribution

Figure 5. (a) Diagram of experimental setup used in second harmonic excitation spectroscopy. (b) Second harmonic emission spectra from single GaAs NW. Each color represents a different pump wavelength. (c) Peak wavelength of second harmonic emission spectra as a function of pump wavelength with a linear fit (red line, slope =0.5), highlighting the second harmonic process.

Figure 6. Experimentally measured second harmonic excitation spectra from NW only and NW-coupled antennas with d = 500 nm, d = 200 nm, d = 100 nm, and d = 90 nm, as labeled in the figure.

of gold nanoantenna is negligible. In our system, the role of metal nanoparticles is to change the distribution of the electric field in the nanowire and therefore to modify its SHG. We can consider the effect of the electric field localization due to the nanoantennas as composed of two main contributions: one is related to transverse polarized light and the other to the longitudinal polarized light. Because of the large aspect ratio, an isolated NW of subwavelength diameter can be efficiently excited only for the longitudinal polarization.10 The nanoantennas investigated in this work are efficiently focusing transverse polarized light into the NW and on its surface. As it 2276

dx.doi.org/10.1021/nl404253x | Nano Lett. 2014, 14, 2271−2278

Nano Letters

Letter

tion. This material is available free of charge via the Internet at http://pubs.acs.org.

is shown from the simulation in Figure 4d, at 1165 nm, a maximum field enhancement of 5.5 is observed at the nanowire surface. At this wavelength, as we decrease the nanoantennas feed-gap distances, novel SH-E peaks appear. In particular, the measured intensity scales with the gap d as follows: I(Single NW)= 0.8 μW, I(d=500) = 0.3 μW, I(d=200) = 11 μW, I(d=100) = 19 μW, and I(d=90) = 23 μW. In order to estimate the electric field enhancement measured by SH-E, we assume that the SH intensity scales as E,4 where E is the maximum value of the local electric field on the wire. By comparing the isolated NW and the coupled system for d = 90 nm, we estimate a maximum enhancement of 2.3, which is in qualitative agreement with the simulation data shown in Figure 4. In Figure 6e, we also observe a peak at 950 nm, which reflects the additional structure shown in Figure 4d. However, the intensity of this peak does not scale as predicted, possibly due to small deviations from the ideal system geometry. Such small perturbations in the system become more significant when the distance between the NW and the nanoantennas is of the order of few nanometers. For transverse polarization, a SH-E peak at 880 nm appears for the isolated NW (Figure 6a and Supporting Information Figure SI7). By adding nanoantennas in the system, we observe a reduction in the intensity of the transverse peak. This reduction is consistent with increased absorption for transverse polarized light confined in the nano antenna gap region. On the other hand, the absorption of longitudinal polarized light is reduced due to the perturbing effect of the gold nanoantennas on the local filed distribution along the nanowire (see Figure 3c). In fact, we notice that the distance between two nanoantennas along the NW is 840 nm which approximately corresponds to the wavelength of maximum drop in the SH-E intensity (from 162 μW in the isolated NW to 0.01 μW in the coupled system with d = 90 nm). Our experimental data demonstrate that the distance between nano antennas along the NW also plays a significant role in determining the overall local field distribution and the resulting SH-E intensity in the coupled system. This provides an additional degree of freedom that needs to be taken into account for the design of such complex nano systems which could be useful for solar cells and photodetectors applications. In conclusion, we have demonstrated photonic-plasmonic resonances of Au nanoantennas coupled to individual GaAs NW in a controlled and reproducible way. By using dark-field scattering and SH-E spectroscopy in partnership with analytical and full-vector FDTD modeling, we demonstrate controlled electromagnetic coupling and near-field concentration between plasmonic nanoantennas and resonant NWs. Moreover, we demonstrate the ability to tailor the nonlinear optical response of individual GaAs NWs by engineering their near-field coupling regime. These findings represent an initial and very promising step toward the development of novel metal− semiconductor resonant nanostructures that leverage the plasmonic properties of strongly coupled metallic nanoantennas and the mature semiconductor NWs technology for the realization of high-efficiency solar cells, high resolution nanodetectors, and nonlinear optical devices.





AUTHOR INFORMATION

Corresponding Author

*L. Dal Negro. E-mail: [email protected]. Present Addresses #

Department of Materials Science and Engineering, Stanford University, Stanford, CA 94305, United States. ⊥ Advanced Science Research Center, The City University of New York, New York, NY 10017, United States. Author Contributions

E.R.A, F.M, G.T., and M.H contributed to the NWs growth. D.R. and M.H. designed the procedure for aligning nanoparticles close to the NWs. A.C. fabricated the samples. E.F.P., J.T., and A.C. performed the experiments and analyzed data. J.T. and L.D.N. carried out the simulations. A.C., E.F.P, J.T., A.F.iM., and L.D.N. wrote the manuscript in collaboration with all the authors. L.D.N. conceived the experiment. L.D.N. and A.F.iM. planned the set of experiments. All authors have given approval to the final version of the manuscript. Author Contributions ¶

These authors contributed equally.

Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank Esther Alarcon-Llado for insightful discussions and Sonia Conesa Boj for TEM analyses in Supporting Information. This work was supported by the AFOSR program “Emitters for High Density Information Processing using Photonic-Plasmonic Coupling in Coaxial Nanopillars” under Award No. FA9550-13-1-0011; the Swiss National Science Foundation under Grant No. 2000021-121758/1 and 129775/1; NCCR QSIT; the European Research Council under Grant “Upcon”.



REFERENCES

(1) Duan, X.; Huang, Y.; Agarwal, R.; Lieber, C. M. Nature 2003, 421, 241−245. (2) Borgstrom, M. T.; Zwiller, V.; Muller, E.; Imamoglu, A. Nano Lett. 2005, 5, 1439−1443. (3) Irrera, A.; Artoni, P.; Iacona, F.; Pecora, E. F.; Franzo, G.; Galli, M.; Fazio, B.; Priolo, F. Nanotechnology 2012, 23, 075204. (4) Tian, B.; Kempa, T. J.; Fang, Y.; Yu, N.; Yu, G.; Huang, J.; Lieber, C. M. Nature 2007, 449, 885. (5) Tian, B.; Kempa, T. J.; Lieber, C. M. Chem. Soc. Rev. 2009, 38, 16. (6) Krogstrup, P.; Jorgensen, H. I.; Heiss, M.; Demichel, O.; Holm, J. V.; Aagesen, M.; Nygard, J.; Fontcuberta i Morral, A. Nat. Photonics 2013, 7, 306−310. (7) Wallentin, J.; Anttu, N.; Asoli, D.; Huffman, M.; Aberg, I.; Magnusson, M. H.; Siefer, G.; Fuss-Kailuweit, P.; Dimroth, F.; Witzigmann, B.; Xu, X. Q.; Samuelson, L.; Deppert, K.; Borgstrom, M. T. Science 2013, 339, 1057−1060. (8) Kelzenberg, M. D.; Boettcher, S. W.; Petykiewicz, J. A.; TurnerEvans, D. B.; Putnam, M. C.; Warren, E. L.; Spurgeon, J. M.; Briggs, R. M.; Lewis, N. S.; Atwater, H. A. Nat. Mater. 2010, 9, 368. (9) Soci, C.; Zhang, A.; Xiang, B.; Dayeh, S. A.; Aplin, D. P. R.; Park, J.; Bao, X. Y.; Lo, Y. H.; Wang, D. Nano Lett. 2007, 7, 1003−1009. (10) Wang, J.; Gudiksen, M. S.; Duan, X.; Cui, Y.; Lieber, C. M. Science 2001, 293, 1455−1457. (11) Winkelmann, C. B.; Ionica, I.; Chevalier, X.; Royal, G.; Bucher, C.; Bouchiat, V. Nano Lett. 2007, 7, 1454−1458. (12) Greytak, A. B.; Barrelet, C. J.; Li, Y.; Lieber, C. M. Appl. Phys. Lett. 2005, 87, 151103.

ASSOCIATED CONTENT

S Supporting Information *

TEM images, deviations, scattering efficiency, refractive index, second harmonic excitation spectra, electric near-field distribu2277

dx.doi.org/10.1021/nl404253x | Nano Lett. 2014, 14, 2271−2278

Nano Letters

Letter

(13) Yan, R.; Gargas, D.; Yang, P. Nat. Phot. 2009, 3, 569. (14) Bohren, C. F.; Huffman, D. R. Absorption and Scattering of Light by Small Particles; Wiley: Hoboken, NJ, 2004. (15) Mokkapati, S.; Saxena, D.; Jiang, N.; Parkinson, P.; WongLeung, J.; Gao, Q.; Tan, H. H.; Jagadish, C. Nano Lett. 2012, 12, 6428−6431. (16) Zia, R.; Schuller, J. A.; Brongersma, M. L. Phys. Rev. B 2006, 74, 165415. (17) Fan, P.; Colombo, C.; Huang, K. C. Y.; Krogstrup, P.; Nygard, J.; Fontcuberta i Morral, A.; Brongersma, M. L. Nano Lett. 2011, 12, 4943−4947. (18) Maier, S. A. Plasmonics: Fundamentals and Applications; Springer: New York, 2007. (19) Dal Negro, L.; Boriskina, S. Laser Photonics Rev. 2012, 6, 178− 218. (20) Schuller, J. A.; Barnard, E. S.; Cai, W.; Jun, Y. C.; White, J. S.; Brongersma, M. L. Nat. Mater. 2010, 9, 193. (21) Hyun, J. K.; Lauhon, L. Nano Lett. 2011, 11, 2731−2734. (22) Colombo, C.; Krogstrup, P.; Nygard, J.; Brongersma, M. L.; Fontcuberta i Morral, A. New J. Phys. 2011, 13, 123026. (23) Knight, M. W.; Grady, N. K.; Bardhan, R.; Hao, F.; Nordlander, P.; Halas, N. J. Nano Lett. 2007, 7, 2346. (24) Grange, R.; Brönstrup, G.; Kiometzis, M.; Sergeyev, A.; Richter, J.; Leiterer, C.; Fritzsche, W.; Gutsche, C.; Lysov, A.; Prost, W.; Tegude, F. J.; Pertsch, T.; Tunnermann, A.; Christiansen, S. Nano Lett. 2012, 12, 5412−5417. (25) Kim, E.; Steinbruck, A.; Buscaglia, M. T.; Buscaglia, V.; Pertsch, T.; Grange, R. ACS Nano 2013, 7 (6), 5343−5349. (26) Walsh, G. F.; Dal Negro, L. Nano Lett. 2013, 13, 3111−3117. (27) Trevino, J.; Walsh, G. F.; Pecora, E. F.; Boriskina, S. V.; Dal Negro, L. Opt. Lett. 2013, 38, 4861. (28) Dufouleur, J.; Colombo, C.; Garma, T.; Ketterer, B.; Uccelli, E.; Nicotra, M.; Fontcuberta i Morral, A. Nano Lett. 2010, 10, 1734−1740. (29) Casadei, A.; Krogstrup, P.; Heiss, M.; Rohr, J. A.; Colombo, C.; Ruelle, T.; Upadhyay, S.; Sorensen, C. B.; Nygard, J.; Fontcuberta i Morral, A. Appl. Phys. Lett. 2013, 102, 013117. (30) Russo-Averchi, E.; Heiss, M.; Michelet, L.; Krogstrup, P.; Nygard, J.; Magen, C.; Morante, J. R.; Uccelli, E.; Arbiol, J.; Fontcuberta i Morral, A. Nanoscale 2012, 1, 1486−1490. (31) Uccelli, E.; Arbiol, J.; Magen, C.; Krogstrup, P.; Russo-Averchi, E.; Heiss, M.; Mugny, G.; Morier-Genoud, F.; Nygard, J.; Morante, J. R.; Fontcuberta i Morral, A. Nano Lett. 2011, 11, 3827−3832. (32) Heigoldt, M.; Arbiol, J.; Spirkoska, D.; Rebled, J. M.; ConesaBoj, S.; Abstreiter, G.; Peiró, F.; Morantece, J. R.; Fontcuberta i Morral, A. J. Mater. Chem. 2009, 19, 840. (33) Demichel, O.; Heiss, M.; Bleuse, J.; Mariette, H.; Fontcuberta i Morral, A. Appl. Phys. Lett. 2010, 97, 201907. (34) Casadei, A.; Schwender, J.; Russo-Averchi, E.; Ruffer, D.; Heiss, M.; Alarcon-Llado, E.; Jabeen, F.; Ramezani, M.; Nielsch, K.; Fontcuberta i Morral, A. Phys. Status Solidi RRL 2013, DOI: 10.1002/pssr.201307162. (35) qstarter by NCCR QSIT. http://www.qstarter.ch/projects/ automated-contacting-of-random-microstructures (accessed Jan 2014). (36) https://www.lumerical.com/. (37) Palik, E. D. Handbook of Optical Constants of Solids; Academic Press: Salt Lake City, UT, 1991. (38) Johnson, P. B.; Christy, R. W. Phys. Rev. B 1972, 6, 4370. (39) Spirkoska, D.; Arbiol, J.; Gustafsson, A.; Conesa-Boj, S.; Glas, F.; Zardo, I.; Heigoldt, M.; Gass, M. H.; Bleloch, A. L.; Estrade, S.; Kaniber, M.; Rossler, J.; Peiro, F.; Morante, J. R.; Abstreiter, G.; Samuelson, L.; Fontcuberta i Morral, A. Phys. Rev. B 2009, 80, 245325. (40) Chen, G.; Wu, J.; Lu, Q.; Gutierrez, H. R.; Xiong, Q.; Pellen, M. E.; Petko, J. S.; Werner, D. H.; Eklund, P. C. Nano Lett. 2008, 8, 1341−1346. (41) Bronstrup, G.; Leiterer, C.; Jahr, N.; Gutsche, C.; Lysov, A.; Regolin, I.; Prost, W.; Tegude, F. J.; Fritzsche, W.; Christianse, S. Nanotechnology 2011, 22, 385201. (42) Pecora, E. F.; Capretti, A.; Miano, G.; Dal Negro, L. Appl. Phys. Lett. 2013, 102, 141114.

(43) Pecora, E. F.; Walsh, G. F.; Forestiere, C.; Handin, A.; RussoAverchi, E.; Dalmau-Mallorqui, A.; Canales-Mundet, I.; Fontcuberta i Morral, A.; Dal Negro, L. Nanoscale 2013, 5, 10163. (44) Lautenschlager, P.; Garrig, M.; Logothetidis, S.; Cardona, M. Phys. Rev. B 1987, 35, 17. (45) Bergfeld, S.; Daum, W. Phys. Rev. Lett. 2003, 90, 3. (46) Canfield, B. K.; Husu, H.; Laukkanen, J.; Bai, B. F.; Kuittinen, M.; Turunen, J.; Kauranen, M. Nano Lett. 2007, 7, 1251. (47) Dadap, J. I. Phys. Rev. B 2008, 78, 121. (48) Heinz T. In Modern problems in condensed matter sciences; Ponath H., Stegeman G., Eds.; Elsevier B.V.: Amsterdam, 1991; Ch. 5, p 353.

2278

dx.doi.org/10.1021/nl404253x | Nano Lett. 2014, 14, 2271−2278