Letter pubs.acs.org/NanoLett
Polarization-Resolved Near-Field Mapping of Plasmonic Aperture Emission by a Dual-SNOM System Angela E. Klein,*,† Norik Janunts,† Michael Steinert,† Andreas Tünnermann,†,‡ and Thomas Pertsch† †
Institute of Applied Physics, Abbe Center of Photonics, Friedrich-Schiller-Universität Jena, Jena, Germany Fraunhofer Institute of Applied Optics and Precision Engineering, Jena, Germany
‡
S Supporting Information *
ABSTRACT: We study the polarization characteristics of light emission and collection in the near field by the tips of a Dual-SNOM (two scanning near-field optical microscopes) setup. We find that cantilevered fiber probes can serve as emitters of polarized light, or as polarization-sensitive detectors. The polarization characteristics depend on the fiber type used for tip fabrication. In Dual-SNOM measurements, we demonstrate mapping of different field components of the plasmonic dipole pattern emitted by an aperture probe. KEYWORDS: Near-field optical microscopy, surface plasmon polariton, scanning probe microscopy, plasmonics, dipole
R
selectively map different components of the plasmonic nearfields. Numerous experimental and theoretical studies have been aimed at elucidating the orientation of the electromagnetic field components in the near-field of an aperture SNOM tip and their role in image formation.8−13 In contrast to radiative fields in homogeneous media that can be described in terms of transverse waves, the electromagnetic near-fields around subwavelength structures (including the aperture of a SNOM tip) are fully vectorial in nature, so that the notion of polarization is generally not applicable in the near-field regime.14−18 However, typical aperture SNOM tips with circular, subwavelength apertures have been shown to probe only the vector components of the electromagnetic field that are transverse to the tip axis, that is, parallel to the aperture plane. In contrast, bare fiber tips,18 scattering tips,16,17,19 splitring type aperture tips,20,21 and tips with apertures approximately a full wavelength in diameter13 can probe longitudinal field components as well. In 1996, it was demonstrated that by coupling linearly polarized light into a straight aperture tip and by placing this tip on a metal film sample, SPPs with dipole characteristics can be excited,22 meaning that the predominant orientation of the electric and magnetic fields is preserved in the near-field of the aperture. This feature of the fiber tip is a prerequisite for polarization contrast scanning near-field optical microscopy with aperture tips, which has been demonstrated both in collection mode23−25 and in illumination mode.26−28 Different techniques have been implemented in order to control the orientation of an illumination tip’s electromagnetic
ecently, scanning near-field optical microscopes (SNOM) with two aperture tips make it possible to excite nano- and microoptical systems in the optical near field and to map their optical near field response at the same time.1−4 The sample is optically excited by a first tip that serves as a subwavelength light source. A second tip is used to collect some of the light that is present at the sample surface and to transport this light to a detector. Conceptually, the intensity maps obtained by this experimental technique are related to maps of selected components of the dyadic Green’s function of the system, as the excitation SNOM tip resembles a pointlike source,5−7 and the near-field response of the sample is mapped by the collection tip. While this technique opens up exciting new possibilities for near-field characterization, the involvement of two aperture tips makes the measurement process quite complex and poses new challenges for the interpretation of measurement results. In previous publications on Dual-SNOM, the excitation tip was either assumed to act as a point dipole1 or as a randomly polarized source.4 The obtained images were compared with intensity distributions. Thus, polarization-insensitive detection was implicitly assumed. We claim that by carefully analyzing both the excitation and the collection process, the image formation in a Dual-SNOM measurement can be described in more detail. In particular, the polarization characteristics of both the excitation and the collection tip play a crucial role in understanding the experimental results. We corroborate this claim by DualSNOM measurements of surface plasmon polaritons (SPPs) with dipolar characteristics that are excited on an unstructured gold film by a first aperture tip and mapped with a second one. By precharacterizing the tips in far-field experiments and using different experimental configurations, we succeeded to © 2014 American Chemical Society
Received: April 17, 2014 Revised: July 24, 2014 Published: August 4, 2014 5010
dx.doi.org/10.1021/nl501431y | Nano Lett. 2014, 14, 5010−5015
Nano Letters
Letter
near-field, either by modifying the aperture29−31 or by influencing the polarization state of light that reaches the aperture. The latter can be achieved by either tuning the polarization state inside the fiber32 with free-space or fiber polarizers, or by using polarization-maintaining fibers.33,34 In some cases, the orientation of the transverse fields is preserved over the transmission through the fiber and the tip;27,33,35 other tip designs act as linear polarizers.29−31,36,37 While Dual-SNOM measurements require an understanding of the polarization characteristics of the excitation as well as the detection process, they can also be used as a tool to assess and confirm these characteristics. In our experiments, we position an excitation SNOM tip on an unstructured gold film. Light is guided through the fiber and excites SPPs whose propagation directions are related to the orientation of the tip’s near-field. These SPPs are then mapped by the collection SNOM tip. Depending on the collection tip’s characteristics, different components of the SPP’s electromagnetic field contribute to the detected images. Our Dual-SNOM setup consists of two identical commercial SNOM systems (MV-4000, Nanonics Imaging, Ltd.). A tuningfork based feedback mechanism operating in tapping mode controls the tip−sample distance. Two types of tips were used: tips fabricated from single-mode (SM) fibers (Newport F-SV, NA = 0.1 to 0.14) and tips fabricated from multi-mode (MM) fibers (Newport F-MSD, NA = 0.2). The fiber tips (Nanonics Imaging Ltd.) were fabricated by heating and pulling the optical fiber and subsequently applying a gold coating, leaving an approximately circular aperture (see Supporting Information) with a diameter of ≈150 nm at the apex. The fibers have a 60° bend within the tapered region, so that the fiber runs horizontal and the tip itself is tilted by around 30° with respect to the sample surface normal. The inclinations of the tips make it possible to bring both tips within very close distance of each other without one tip shaft touching the other. Collisions between both tips were effectively prevented by a proximity detection mechanism based on mechanical interaction between both tips.38 The minimum distance between both tips is reached when both tip coatings touch. As a light source, we used a diode laser with a free-space wavelength of 663 nm in all measurements presented here. We started our study by investigating the polarization selectivity of the cantilevered fiber tips in the far-field. In a first series of experiments, we illuminated the tips with a linearly polarized laser beam to study their polarization sensitivity in collection mode. By recording the collected signal at the end of the fiber for different polarization directions of the illumination beam, the polarization-dependent collection efficiencies of the tips could be quantified. The experimental setup is shown in Figure 1a. The expanded laser beam first passes a halfwavelength plate and is then focused by a 20×-objective (NA = 0.22) of an inverted microscope. The fiber tips, mounted in a SNOM head, are positioned with the apex on the microscope’s optical axis and a few tens of m above the focus. For each tip, the polarization direction ϕ of the illuminating beam is rotated in steps of 10° and the corresponding collection signals are detected by a fiber-coupled avalanche photodiode (PerkinElmer SPCM-AQR). The polarization-resolved collection efficiency of each tip has a two-lobe pattern [see the example in Figure 1b]. Light of a distinct polarization direction θ is collected most efficiently. In a simple model based on a vectorial decomposition of the electrical field of the illuminating light
Figure 1. (a) Schematic drawing of the experimental setup used to measure the polarization sensitivity of aperture SNOM tips. A freespace laser beam passes a half-wavelength plate and is then focused by a microscope objective. The tip is positioned above the laser focus. The polarization direction ϕ of the beam is rotated stepwise, and the collected light intensity is recorded by a detector. (b) Typical polar plot of the measured intensities (blue crosses) and the fitted curve (black line) according to eq 1, for the highly polarization-sensitive tip used in the near-field experiments. The angle θ denotes the polarization direction of light that is collected most efficiently by the tip.
into the components Ea, along θ, and Eb, orthogonal to θ, the collected signal is then given by I(ϕ) = a cos2(ϕ − θ ) + b sin 2(ϕ − θ )
(1)
where a and b represent the maximum and minimum collection efficiency, multiplied by the illumination intensity, respectively. We fitted expression (1) to the measured values (corrected for the detector’s dark count), obtaining a, b, and θ for each tip [Figure 1b]. In analogy to the definition of the degree of polarization, 39 we define the polarization sensitivity Qc = (a − b)/(a + b), which can take values between 0 (polarization-insensitive collection efficiency) and 1 (detection of a single polarization component). We determined these values for 14 tips, 7 of which were fabricated from SM fiber and 7 from MM fiber. We found polarization sensitivities in the range between 0.33 and 0.86. Generally, the polarization sensitivities of SM fiber based tips were higher than the polarization sensitivities of MM fiber based tips. The angles θ are not randomly distributed, but they are found in two separate angular regions, around 0° and around ±90° with respect to the plane of the fiber bend. A detailed analysis of the data can be found in the Supporting Information. The possible sources of polarization sensitivity are (1) the individual aperture shape of each tip, (2) the inclination of the tip’s aperture with respect to the beam’s transverse direction (30°), and (3) polarization-dependent loss, mode coupling, and birefringence induced by the fiber bend. The clustering of θ values around 0° and ±90° rules out the first possible cause, as it suggests that the polarizationsensitivity is related to the symmetry-break induced by the fiber bend and the aperture’s inclination. This claim is also supported by the fact that some of the apertures are almost perfectly circular. To estimate the influence of the aperture’s tilt, we measured the polarization sensitivity of several tips when the illuminating beam was tilted by ±30° with respect to the vertical direction and found that this did not significantly reduce the polarization sensitivity. 5011
dx.doi.org/10.1021/nl501431y | Nano Lett. 2014, 14, 5010−5015
Nano Letters
Letter
fibers. Both tips feature similar bending angles and bending radii. It has been reported that the orientation of the transverse electromagnetic field components in the near-field of an illumination- mode aperture tip might be different from the farfield polarization.11 We studied the near-field emission properties of the SM tip with Qe = 0.96 by using it as a source of SPPs on a gold film and observing their propagation directions. A focused ion beam was used to mill a circular grating (“bull’s eye” structure) into a homogeneous, 150 nm thick gold film deposited on a glass substrate. The seven concentric circular slits (diameter of the inner circle: 15 μm) were milled through the complete thickness of the gold film. The periodicity corresponded to the SPP wavelength (λSPP = 639 nm), and the filling factor was 0.5. This grating design ensures that SPPs excited in the middle of the circle are diffracted perpendicularly out of the sample plane. This light emission is recorded by a microscope objective focused onto the structure from below [Figure 2a]. Figure 2b shows a
In fiber optics, it is well-known that fiber bends can lead to birefringence, bend-induced mode coupling and to polarizationdependent loss.40−43 Depending on the bend radius, either the TE or the TM mode can suffer stronger losses in a fiber bend.41 While these observations have been made in untapered fibers with large bend radii in the cm range, which are different from the metal-coated fiber tapers (diameter at the bend: approximately 50 μm) used in our experiments, it is entirely conceivable that similar effects lead to the observed polarization filtering. Moreover, as we can deduce from the fibers’ numerical apertures, the refractive index contrast in the MM fibers is higher than in the SM fibers. This could explain why the bend loss, and consequently, the polarization filtering, are less pronounced in MM fiber tips. The observed preferential directions of 90° and 0° coincide with the characteristic directions of the symmetry break induced by the fiber bend. Therefore, we attribute the observed polarization filtering to polarization-dependent loss introduced by the fiber bend. For two of the tips, we additionally studied the degree of polarization Qe of light emitted from the apertures when the laser was coupled into the fiber. In this experiment, the laser beam passes through a half-wavelength plate, which allows us to turn its polarization direction and is subsequently coupled into the fiber by an aspheric lens. The tip apex is imaged onto a detector by the inverted microscope. By inserting a thin-film polarizer into the microscope’s beam path, the linear polarization components of light emitted from the aperture can be measured. For the first tip, fabricated from a SM fiber, we had previously determined a polarization sensitivity Qc (in collection mode) of 0.86 and θ was −12° [see Figure 1b]. The light emitted by this tip was polarized primarily along the direction 0°(±15°). This direction was independent of the input polarization. We varied the polarization of the input beam by turning the half-wavelength plate (before coupling the beam into the fiber). The throughput through the tip varied as a function of the input polarization. The maximum throughput was approximately 5 times higher than the minimum throughput. For each polarization of the input beam, we measured the degree of polarization Qe = (I0° − I90°)/(I0° + I90°) of the light emitted from the aperture. Qe varied depending on the input polarization. A maximum value of Qe = 0.96 was achieved and coincided with the maximum throughput through the tip. The minimum degree of polarization for the orthogonal input polarization was 0.81. The bend of the tapered fiber with a bending radius of around 100 μm and a bending angle of around 60° appears to act as a polarization filter. In subsequent measurements, we used this tip for near-field illumination, maintaining the optimum setting of the half-wavelength-plate, that is, Qe = 0.96. The second tip was fabricated from a MM-fiber and had a polarization sensitivity Qc (in collection) of only 0.35. Like in the previous measurement of the emission from a SM fiber tip, the throughput of the MM fiber tip varied by a factor of approximately 5 depending on the polarization of light coupled into the fiber end. The polarization characteristics of light emitted from the aperture were quite different from the previous measurement of the single-mode fiber tip. Generally, the degree of polarization emitted from its aperture was low, with a maximum degree of polarization of Qe ≈ 0.3. The “polarization-filtering” effect that could be clearly observed with the single-mode fiber tip is much less pronounced in the multimode fiber tip, which we attribute to differences in mode propagation and intermode coupling between the different
Figure 2. (a) Sketch of the “bull’s eye” setup to determine the angular characteristics of SPP excitation by an aperture SNOM tip. The tip is placed in the middle of a circular diffraction grating. The diffracted light is collected by a microscope focused on the sample plane. (b) Sketch of the grating. The width of the trenches and ridges is half the SPP wavelength (λSPP) each, leading to a radial “periodicity” of λSPP. (c) Resulting microscope image (bottom view). The positions of the tip and the grating are indicated by dashed lines.
bottom view of the grating that is illuminated by the tip, which has been placed in the middle of the structure, with the feedback mechanism keeping the tip−sample distance constant. The emission from the grating region originates from diffraction of SPPs excited by the tip. The two bright regions demonstrate that the plasmon excitation is strongly anisotropic and SPPs propagate mainly along the x-axis. As SPPs propagating along the x-axis have an electric field component along x and a magnetic field along y, this matches the previously measured far-field polarization direction. We conclude that the orientations of the transverse components of the excitation tip’s electromagnetic near-field essentially correspond to the far-field polarization. In order to study the near-field emission pattern of the tip in more detail, we placed the tip on an unstructured gold film (thickness = 55 nm) and performed a SNOM measurement of the excited SPPs in the region around the excitation tip with a second, collecting SNOM tip. As the signals in Dual-SNOM experiments are generally low, we found it necessary to use MM tips for collection, which showed higher collection efficiencies than SM tips both in near-field measurements and 5012
dx.doi.org/10.1021/nl501431y | Nano Lett. 2014, 14, 5010−5015
Nano Letters
Letter
For comparison, a simulated collection-mode image is shown in Figure 3d. We first calculated the electric and magnetic fields of SPPs excited by a magnetic point-dipole44 oriented along the vertical axis of the image (y). The collection-mode measurement process was then modeled by assuming that the electromagnetic field components (Ea, Ha) that correspond to a polarization along θ contribute stronger to the image than those polarized orthogonal to this preferred direction (Eb, Hb). The relative contributions a and b were determined from the previously measured polarization sensitivity Qc. As the collection tip is tilted by around 30° with respect to the sample surface normal, we introduced a tip coordinate system (x, y′, z′) depicted in Figure 3b, which is rotated around the xaxis so that the z′-axis is parallel to the tip axis. The calculation takes account of the field components which are orthogonal to the tip, that is, the field components in the xy′-plane. Although the collection tip has a finite resolution, for clarity we modeled it as a pointlike detector. In Figure 3d, the simulated intensity image, based on the magnetic fields, is shown according to I ∝ a|Ha|2 + b|Hb|2. The question whether aperture SNOM images represent primarily the H-fields or the E-fields, and equivalently, whether in illumination mode the sample is excited primarily via the electric or the magnetic fields of the tip, is not entirely settled20,45−48 and is beyond the scope of this Letter. In our experiments, the spatial distributions of the transverse magnetic fields can not be distinguished from the electric ones. In other words, the spatial field profiles of Hx could be replaced by Ey and Hy by Ex without changing the results of our work. It should also be noted that a similar excitation pattern can be obtained by modeling the excitation tip as a combination of two opposing dipoles along the z direction.49 The similarity of the simulated image [Figure 3d] and the Dual-SNOM measurement [Figure 3c] indicates that the excitation tip can indeed be considered as a dipole source for SPPs with the dipole orientation corresponding to the far-field polarization of emitted light, and the collection tip can be considered as a detector of the tangential field components Hx and Hy′ (or, equivalently, Ex and Ey′). To our knowledge, Figure 3c also constitutes the first near-field mapping of the unperturbed dipole emission pattern of an aperture SNOM tip. Fujimoto et al.4 performed a similar Dual-SNOM experiment on a planar silver film. However, the light emitted from their straight silver-coated tip was randomly polarized. Consequently, they observed a radially symmetric SPP emission pattern. In another Dual-SNOM experiment, we demonstrate that apart from acting as dipole-like emitters cantilevered SNOM tips can serve as polarization-sensitive near-field detectors. For these measurements, we used the “perpendicular” experimental configuration shown in Figure 4a, where the in-plane angle between the tips was changed to 90°, and a different MM-fiber based tip was used for collection. For this collection tip, Qc and θ could not be measured because the aperture suffered damage after the near-field measurements. The measured near-field pattern [Figure 4c] exhibits three lobes that extend diagonally from the excitation location toward three corners of the image. The fourth quadrant could not be mapped due to the geometric constraints of the setup. The measured pattern is different from the dipole pattern in Figure 3c and appears to have a 4-fold symmetry. To understand this image, we compare it to calculated images of the different components of the H-field with respect to the tip’s coordinate
in the far-field characterization experiment described above. We speculate that some of the light that is transmitted by the aperture is subsequently, within the tapered region, coupled to cladding modes, which do not contribute to the detected signal. In SM fibers, this effect may be much more pronounced than in MM fibers because of the different numerical apertures and core diameters of both fiber types. Furthermore, as detailed above, the losses in the fiber bend may be lower for MM fiber tips. In MM fibers, the polarization of light is not well preserved. Therefore, we did not use any polarizing elements in the detection path. We chose the SM-tip described above with a high degree of polarization (Qe = 0.96) for excitation and a MM-tip with a polarization sensitivity of Qc = 0.42 and an angle θ = −9° for collection. In our first experiment, the in-plane angle between the tips was 180° [see Figure 3a], that is, both
Figure 3. (a) Dual-SNOM setup in “parallel” configuration for measuring the SPP excitation pattern of an aperture tip. (b) Visualization of the (collection-)tip’s coordinate system (x, y′, z′), which is tilted versus the sample coordinate system by 30° in the yzplane, so that z′ is parallel to the tip axis. (c) Measured collectionmode SNOM image of SPPs excited on a homogeneous gold film by an aperture SNOM tip. The excitation tip is located in the region indicated by the black rectangle near the top edge of the image, its fiber axis is along the y-axis of the image (see the dashed outline). The image is assembled from several smaller scans, as indicated by the white dashed lines. (d) Simulated Dual-SNOM image, assuming that the excitation tip acts as a point dipole source for SPPs and the collection tip (with MM fiber) is sensitive to the components of the magnetic field which are perpendicular to the tip axis, that is, Hx and Hy′. The experimentally determined polarization sensitivity Qc and angle θ have been used to calculate the relative contributions a and b of the field components Ha (corresponding to a polarization along θ) and Hb (corresponding to the polarization orthogonal to θ). The shadowed rectangle symbolizes the area not accessible to scanning in Figure 4b.
tips opposed each other (“parallel” configuration). The power coupled into the illuminating tip was below 15 mW to avoid melting of the metal coating. The resulting SNOM image is shown in Figure 3c, clearly showing a dipole pattern. Although the ratio between the power incoupled into the excitation tip and the detected power from the collection tip is below 10−12, the count rates allow for scanning at moderate speed (around 10 ms/point) to get a good contrast in the images. 5013
dx.doi.org/10.1021/nl501431y | Nano Lett. 2014, 14, 5010−5015
Nano Letters
Letter
Figure 4. (a) Dual-SNOM setup in “perpendicular” configuration with a single-mode fiber tips for excitation and a multi-mode fiber tip for collection and an in-plane angle of 90° between tips. (b) Visualization of the (collection) tip’s coordinate system (x″, y, z″), which is tilted versus the sample coordinate system by 30° in the xz-plane, so that z″ is parallel to the tip axis. (c) Measured collection-mode SNOM image of SPPs excited by an aperture SNOM tip on a homogeneous gold film. The excitation tip is located in the black rectangle near the top edge of the image, its fiber axis is along the y-axis of the image (see dashed outline). The shown image is assembled from several smaller images, indicated by white dashed lines. The maximum of the color scale corresponds to a count rate of approximately 10 kHz. (d−f) Simulated intensity maps, assuming that the excitation tip acts as a point dipole, visualize the different magnetic field components with respect to the tip coordinate system. The three images share the same color scale. (g) Simulated intensity maps visualizing the different components of the electric fields. The three images share a common color scale. The field component along y is so weak that the corresponding intensity map has been multiplied by 10 to make the pattern visible. The numbers in the lower right corners of all simulated images are mean intensities (on a relative scale) corresponding to the respective field component.
possibly at the aperture, before polarization mixing in the fiber can lead to a degradation of the polarization contrast. On the other hand, the polarization sensitivities of cantilevered fibers vary from tip to tip with relatively low mean polarization sensitivities of around 0.7 for the 7 SM tips and around 0.4 for the 7 MM tips that we characterized (see Supporting Information). In conclusion, we have presented the first direct near-field measurements of dipole-like SPP emission from aperture SNOM tips. All SNOM tips were fabricated from standard single-mode or multimode fibers and had approximately circular apertures. By precharacterizing the polarization characteristics of tips in the far-field, their polarization sensitivity could be assessed. Even though no special measures had been taken to obtain polarization-sensitive tips, polarization-sensitive far-field behavior was found in all tips. For some tips, the degree of polarization of light emitted from the aperture into the far-field was additionally measured, reaching a degree of polarization as high as 0.96. The directionality of SPPs excited by an aperture tip confirmed that the predominant orientation of transverse field components in the near-field corresponds to the polarization in the far-field. By using two different measurement configurations, different field components of the plasmonic dipole pattern could be selectively mapped. Apart from the typical two-lobe pattern, a four-lobe pattern could be observed in a cross-polarization SNOM configuration. This cross-polarization Dual-SNOM measurement demonstrates the polarization-selective near-field properties of both tips. All the measurements have been done without the use of analyzers in the detection path and with standard fiber tips. No special measures had been taken in tip production to modify the polarization characteristics of the tips. We believe that our results can contribute to a better understanding of SNOM measurements utilizing single or multiple tips, especially SNOM measurements with cantilevered fiber tips.
system (x″, y, z″), which has been defined such that the z″-axis is parallel to the collection tip axis. Figure 4d−f shows the simulated intensity images assuming the same excitation dipole as in Figure 3d, but only a single component of the magnetic field contributes to the plotted intensity. While the x″- and z″components show a four-lobe pattern that resembles the measurement, the magnetic field along the y-axis [Figure 4e] is considerably stronger and shows a two-lobe dipole pattern. The resemblance of the measurement [Figure 4c] to the intensity map in Figure 4d can be understood as an effect of the specific tip used for collection, which acts as a polarization filter in the detection path due to its inherent polarization sensitivity. It is worth noting that as the SPPs’ magnetic fields are parallel to the sample plane the nonvanishing z″-component is due to the projection of the x-component onto the tilted z″-direction. In contrast, the electric field has a strong z-component, which contributes to Ez″ and Ex″ in the tilted coordinate system. Therefore, the polarization-filtering effect becomes even more evident when we compare the measured image to the electric field components [Figure 4g], which again share a common color scale. The three-lobe pattern in the measured SNOM image [Figure 4c], suggesting a 4-fold symmetry, clearly demonstrates that only the field components Ey and Hx″, which correspond to a single polarization, contribute to the detected image, whereas the other field components are efficiently suppressed. Hence, the measured image constitutes a cross-polarization measurement of the plasmonic dipole pattern. This polarization-filtering effect was achieved even though no analyzer was used in the detection path. We attribute this effect to a high polarization sensitivity of the specific MM-fiber that was used for collection in this measurement. In comparison to polarization-sensitive collection-mode SNOM techniques where a polarization filter is inserted between the optical fiber and the detector, our technique benefits from the fact that filtering occurs at the fiber bend, and 5014
dx.doi.org/10.1021/nl501431y | Nano Lett. 2014, 14, 5010−5015
Nano Letters
■
Letter
(21) le Feber, B.; Rotenberg, N.; van Oosten, D.; Kuipers, L. Opt. Lett. 2014, 39, 2802−2805. (22) Hecht, B.; Bielefeldt, H.; Novotny, L.; Inouye, Y.; Pohl, D. W. Phys. Rev. Lett. 1996, 77, 1889−1892. (23) Burresi, M.; Engelen, R. J. P.; Opheij, A.; van Oosten, D.; Mori, D.; Baba, T.; Kuipers, L. Phys. Rev. Lett. 2009, 102, 033902. (24) Kihm, H. W.; Koo, S. M.; Kim, Q. H.; Bao, K.; Kihm, J. E.; Bak, W. S.; Eah, S. H.; Lienau, C.; Kim, H.; Nordlander, P.; Halas, N. J.; Park, N. K.; Kim, D.-S. Nat. Commun. 2011, 2, 451. (25) Schnell, M.; Garcia-Etxarri, A.; Alkorta, J.; Aizpurua, J.; Hillenbrand, R. Nano Lett. 2010, 10, 3524−3528. (26) Lacoste, T.; Huser, T.; Prioli, R.; Heinzelmann, H. Ultramicroscopy 1998, 71, 333−340. (27) Ramoino, L.; Labardi, M.; Maghelli, N.; Pardi, L.; Allegrini, M.; Patanè, S. Rev. Sci. Instrum. 2002, 73, 2051−2056. (28) Mitsuoka, Y.; Nakajima, K.; Homma, K.; Chiba, N.; Muramatsu, H.; Ataka, T.; Sato, K. J. Appl. Phys. 1998, 83, 3998−4003. (29) Patanè, S.; Cefalì, E.; Spadaro, S.; Gardelli, R.; Albani, M.; Allegrini, M. J. Microsc. 2008, 229, 377−383. (30) Danzebrink, H. U.; Dziomba, T.; Sulzbach, T.; Ohlsson, O.; Lehrer, C.; Frey, L. J. Microsc. 1999, 194, 335−339. (31) Venugopalan, P.; Zhang, Q.; Li, X.; Gu, M. Opt. Express 2013, 21, 15247. (32) Nakajima, K.; Mitsuoka, Y.; Chiba, N.; Muramatsu, H.; Ataka, T.; Sato, K.; Fujihira, M. Ultramicroscopy 1998, 71, 257−262. (33) Mitsui, T.; Sekiguchi, T. J. Electron Microsc. (Tokyo) 2004, 53, 209−215. (34) Mitsui, T. Rev. Sci. Instrum. 2005, 76, 043703. (35) Grosjean, T.; Ibrahim, I. A.; Mivelle, M. Appl. Opt. 2010, 49, 2617−2621. (36) Adiga, V. P.; Kolb, P. W.; Evans, G. T.; Cubillos-Moraga, M. A.; Schmadel, D. C.; Dyott, R.; Drew, H. D. Appl. Opt. 2006, 45, 2597− 2600. (37) Dressler, D. H.; Landau, A.; Zaban, A.; Mastai, Y. Chem. Commun. 2007, 9, 945−947. (38) Klein, A. E.; Janunts, N.; Tünnermann, A.; Pertsch, T. Appl. Phys. B: Lasers Opt. 2012, 108, 737−741. (39) Al-Qasimi, A.; Korotkova, O.; James, D.; Wolf, E. Opt. Lett. 2007, 32, 1015−1016. (40) Schulze, C.; Lorenz, A.; Flamm, D.; Hartung, A.; Schröter, S.; Bartelt, H.; Duparré, M. Opt. Express 2013, 21, 3170−3181. (41) Wang, Q.; Rajan, G.; Wang, P.; Farrell, G. Opt. Express 2007, 15, 4909−4920. (42) Yu, H.; Wang, S.; Fu, J.; Qiu, M.; Li, Y.; Gu, F.; Tong, L. Appl. Opt. 2009, 48, 4365−4369. (43) Miyagi, M.; Yip, G. L. Opt. Quantum Electron. 1976, 8, 335−341. (44) Rotenberg, N.; Spasenović, M.; Krijger, T. L.; le Feber, B.; García de Abajo, F. J.; Kuipers, L. Phys. Rev. Lett. 2012, 108, 127402. (45) Devaux, E.; Dereux, A.; Bourillot, E.; Weeber, J.-C.; Lacroute, Y.; Goudonnet, J.-P.; Girard, C. Phys. Rev. B 2000, 62, 10504−10514. (46) Kihm, H. W.; Kim, J.; Koo, S.; Ahn, J.; Ahn, K.; Lee, K.; Park, N.; Kim, D.-S. Opt. Express 2013, 21, 5625−5633. (47) Kohlgraf-Owens, D. C.; Sukhov, S.; Dogariu, A. Opt. Lett. 2012, 37, 3606−3608. (48) Le Feber, B.; Rotenberg, N.; Beggs, D. M.; Kuipers, L. Nat. Photon 2014, 8, 43−46. (49) Dobmann, S.; Kriesch, A.; Ploss, D.; Peschel, U. Adv. Opt. Mater. 2014, DOI: 10.1002/adom.201400237.
ASSOCIATED CONTENT
S Supporting Information *
Statistical analysis of the polarizations θ which are most efficiently detected by 14 different SNOM tips; scanning electron (SEM) micrographs of aperture tips. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Phone: +49 (0)3641 947845. Fax: +49 (0)3641 947841. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors thank Sören Schmidt, Carsten Rockstuhl, Hesham Taha, Patricia Hamra, and Ulf Peschel for helpful discussions. We acknowledge financial support from the Thuringian State Government (MeMa), the German Federal Ministry of Education and Research (PhoNa) and the German Research Foundation (SPP 1391 “Ultrafast Nanooptics”). The authors declare no competing financial interest.
■
REFERENCES
(1) Dallapiccola, R.; Dubois, C.; Gopinath, A.; Stellacci, F.; Negro, L. D. Appl. Phys. Lett. 2009, 94, 243118. (2) Kaneta, A.; Hashimoto, T.; Nishimura, K.; Funato, M.; Kawakami, Y. Appl. Phys. Express 2010, 3, 102102. (3) Ren, X.; Liu, A.; Zou, C.; Wang, L.; Cai, Y.; Sun, F.; Guo, G.; Guo, G. Appl. Phys. Lett. 2011, 98, 201113. (4) Fujimoto, R.; Kaneta, A.; Okamoto, K.; Funato, M.; Kawakami, Y. Appl. Surf. Sci. 2012, 258, 7372−7376. (5) Chicanne, C.; David, T.; Quidant, R.; Weeber, J. C.; Lacroute, Y.; Bourillot, E.; Dereux, A.; Colas des Francs, G.; Girard, C. Phys. Rev. Lett. 2002, 88, 097402. (6) Colas des Francs, G.; Girard, C.; Weeber, J.-C.; Dereux, A. Chem. Phys. Lett. 2001, 345, 512−516. (7) Dereux, A.; Girard, C.; Weeber, J.-C. J. Chem. Phys. 2000, 112, 7775−7789. (8) Betzig, E.; Trautman, J. K.; Weiner, J. S.; Harris, T. D.; Wolfe, R. Appl. Opt. 1992, 31, 4563−4568. (9) Durkan, C.; Shvets, I. V. J. Appl. Phys. 1998, 83, 1837−1843. (10) Shin, D. J.; Chavez-Pirson, A.; Lee, Y. H. Opt. Lett. 2000, 25, 171−173. (11) Biagioni, P.; Polli, D.; Labardi, M.; Pucci, A.; Ruggeri, G.; Cerullo, G.; Finazzi, M.; Duò, L. Appl. Phys. Lett. 2005, 87, 223112. (12) Antosiewicz, T. J.; Szoplik, T. Opt. Express 2007, 15, 7845− 7852. (13) Ploss, D.; Kriesch, A.; Pfeifer, H.; Banzer, P.; Peschel, U. Opt. Express 2014, 22, 13744. (14) Girard, C.; Weeber, J.-C.; Dereux, A.; Martin, O. J. F.; Goudonnet, J.-P. Phys. Rev. B 1997, 55, 16487−16497.. (15) Lee, K. G.; Kihm, H. W.; Kihm, J. E.; Choi, W. J.; Kim, H.; Ropers, C.; Park, D. J.; Yoon, Y. C.; Choi, S. B.; Woo, D. H.; Kim, J.; Lee, B.; Park, Q. H.; Lienau, C.; Kim, D. S. Nat. Photonics 2007, 1, 53− 56. (16) Ahn, K. J.; Lee, K. G.; Kim, D. S. Opt. Commun. 2008, 281, 4136−4141. (17) Kihm, H. W.; Lee, K. G.; Kim, D. S.; Ahn, K. J. Opt. Commun. 2009, 282, 2442−2445. (18) Grosjean, T.; Mivelle, M.; Burr, G. W. Opt. Lett. 2010, 35, 357− 359. (19) Esteban, R.; Vogelgesang, R.; Dorfmüller, J.; Dmitriev, A.; Rockstuhl, C.; Etrich, C.; Kern, K. Nano Lett. 2008, 8, 3155−3159. (20) Burresi, M.; Oosten, D. v.; Kampfrath, T.; Schoenmaker, H.; Heideman, R.; Leinse, A.; Kuipers, L. Science 2009, 326, 550−553. 5015
dx.doi.org/10.1021/nl501431y | Nano Lett. 2014, 14, 5010−5015