We report a new optical near-field transducer comprised of a metallic nano-antenna extending from the ridge of a C-shaped metallic nano-aperture. Finite-difference time domain simulations predict that the C-aperture nano-tip (CAN-Tip) provides high intensity (650x), high optical resolution (~λ/60), and background-free near-field illumination at a wavelength of 980 nm. The CAN-Tip has an aperture resonance and tip antenna resonance which may be tuned independently, so the structure can be made resonant at ultraviolet wavelengths without being unduly small. This near-field optical resolution of 16.1 nm has been experimentally confirmed by employing the CAN-Tip as an NSOM probe.
© 2011 OSA
All near-field scanning optical microscopy (NSOM) techniques suffer a performance tradeoff between high resolution, intensity and reduction of stray background illumination. In this letter, we propose a hybrid scheme to obtain high near-field intensity and high optical resolution without background illumination. A C-aperture nano-tip (CAN-Tip) is composed of a resonant C-shaped nano-aperture with a nano-antenna attached to the ridge on the exit side. The highly efficient transmission of the C-aperture provides a high concentration of charges at the corners of the ridge on the exit surface . Due to the lightning-rod effect, surface charges are further driven to the tip of the nano-antenna and a high intensity ultra-small optical spot can be obtained close to the tip. Finite-difference time-domain (FDTD) simulations show that the presence of the tip red-shifts the thickness resonances  of the C-aperture. By tuning the dimensions of the antenna, the combined resonance due to the aperture thickness and the tip length can be made to merge with the C-aperture surface resonance  and this combination leads to the near-field intensity being enhanced 11.6x over that of a planar C-aperture for a total enhancement of > 650x. We also fabricated an optical fiber based NSOM probe using a CAN-Tip and demonstrated a λ/60 optical resolution under 980 nm illumination.
Light passing through a nano-aperture is laterally confined to the dimensions of that aperture, providing a means to concentrate optical energy to spots with areas far smaller than a square wavelength; this was the first method proposed to overcome the “diffraction limit” of ray optics . However, a decrease in aperture size below than λ/2 causes a fourth order reduction in its power throughput [4,5]. The weak power transmitted through a round (or a square) nano-aperture limits its practical applications. The C-shaped nano-aperture first proposed by our group , as well as other ridge apertures later proposed [7,8] takes advantage of its non-convex shape to red-shift its cutoff frequency, concentrating light at higher intensities in a nano-scale cross-section. In a ridge nano-aperture, surface charges concentrate along the ridge at the exit surface of the aperture, and a strong optical near-field can be obtained close to the ridge. Such apertures generally provide two to three orders of magnitude higher power throughput than round or square apertures while maintaining a similar subwavelength optical spot size. Due to the limitations of present fabrication technology, these optical spot sizes are normally limited to 40-70 nm .
By employing a sharp metallic tip to concentrate light, apertureless NSOM can achieve much higher resolution, but the measured signal is contaminated by background illumination. Two common designs use atomic force microscope (AFM) tips  and bowtie nano-antennas . Diffraction-limited focused light illuminating a metallic AFM tip or nano-antenna generates high optical intensities close to the sharpest features of the structure due to the “lightning-rod effect” and resonances (if any) of the nano-antenna. Because the optical spot size of the apertureless structure is determined by the sharpness of the tip, it can be smaller than 10 nm . However, the external illumination cannot be blocked, so there will always be substantial background optical energy in addition to the high intensity light concentrated by the tip. The background signal contribution can be decoupled from the near-field signal by applying modulation techniques ; however, the extraneous illumination can still lead to unwanted photo-bleaching in certain fluorescence experiments .
To gain the benefits of both apertured and apertureless NSOM techniques, a nano-scale optical source generated by combining an aperture and an antenna was devised by Frey et al. . Their system combined a nano-aperture that eliminated the background illumination and a nano-antenna that worked as a scattering object, concentrating optical energy at the tip . Because the design used a round aperture, it provided only a weak evanescent source to excite the antenna due to its fundamentally low transmission. Although the antenna length can be tuned to resonance, it only offered a slight intensity enhancement .
2. Design of the C-aperture nano-tip and FDTD simulations
A natural improvement to this scheme is to put the nano-antenna at a location where there is strong localized field, such as on the ridge of a C-aperture. This C-aperture nano-tip (CAN-Tip) design is illustrated in Fig. 1(a) . Naturally, the behavior of this compound structure is not analytically derivable, so moving from concept to design requires computational support. We employed an in-house FDTD code to simulate the structure and its interactions with objects in an NSOM configuration. To tune the CAN-tip to resonate at 980 nm, we set the characteristic dimension α of the C-aperture to 40 nm, which determines the surface current resonance condition . A 100 nm (td) thick silicon nitride membrane forms the substrate, above which the C-shaped aperture is cut in a 150 nm gold layer. The tip length h is tuned over a small range. The simulation cell sizes are 2.5 nm in x, y, and z with perfectly matched layer boundary conditions on each side. The gold is modeled as a Drude material with ε∞ = 11.1952, ωp = 1.4358 × 1016 s−1, and τc = 8.6697 × 10−15 s. Figure 1(b) shows spectra of near-field enhancement of CAN-Tips with tip lengths of 0 nm (planar C-aperture), 30 nm, 50 nm, 70 nm, and 90 nm. The enhancement factors are measured 6 nm away from the tip, which is the typical tip-to-sample distance in our experiment.
Unlike the antenna used in conjunction with a round aperture (previously reported in ), the antenna of the CAN-Tip is not required to be exactly λ/4 in length; rather, the total resonance is determined by the combined geometry of the C-aperture and the tip. The CAN-Tip has an optimal near-field intensity which occurs when the two peaks in Fig. 1(b) merge together. This can be understood in terms of the resonance mechanisms of the C-aperture. Upon illumination of the entrance surface of the C-aperture with polarization parallel to the ridge, induced surface current flows back and forth between the ridge and the back of the C-aperture and charges accumulate at the two ends. This surface current couples to the waveguide modes and ultimately induces similar surface currents on the exit surface. The oscillating charges, like an oscillating dipole, at the exit surface of the C-aperture radiate to the far-field and result in enhanced transmission if the incidence wavelength is about twice the path length of the surface current. In real metals, the resonance wavelength of nano-apertures can be modified by hybridization with surface plasmon modes [16–18]. The total transmission can be further enhanced by thickness resonances [2,19]. The planar C-aperture shown in Fig. 1(b) has an aperture surface resonance peak at 960 nm (the main peak), and a less prominent thickness resonance peak at 652 nm. Sun et al. predicted that a C-shaped waveguide (thick C-aperture) would perform best when the aperture surface resonance and the thickness resonance occur at the same frequency . In planar ridge apertures, in order to red-shift the thickness resonance towards the aperture surface resonance, the thickness of the aperture has to be increased; however, this also reduces the transmission of the aperture surface mode because it is evanescently coupled.
The CAN-Tip provides an additional parameter, the antenna length, to tune the thickness resonance peak. The tip elongates the path that the exit surface charges traverse as they flow between the ridge of the aperture and the opposite interior face on each optical cycle. As a result, the less prominent peak, inherited from the thickness resonance of the aperture, red-shifts with the increase of the tip length (Fig. 1(b)) because the surface current along the ridge goes to the tip. The surface mode enhancement is reduced because the presence of a short tip (30 nm) breaks the symmetry of the resonances on the entrance and the exit surfaces. When the surface resonance peak and hybrid antenna resonance peak start to overlap, both peaks increase. The hybrid antenna resonance peak merges with the main aperture surface resonance peak when the tip length is 90 nm. This combined resonance shows a very large near-field enhancement (654x), some 11.6x higher than that of a planar C-aperture (56x). Thus, tuning the tip length provides a way to merge the surface resonance and the thickness resonance of ridge nano-apertures. Figure 1(c) and 1(d) show the corresponding near-field profiles of a planar C-aperture and a CAN-Tip with 90 nm tip length. The tip reduces the spot size from 15.37 nm x 48.09 nm (for a planar aperture) to 18.36 nm x 18.36 nm, which is roughly the size of the tip. The spot size can be made even smaller and the enhancement factor even larger if the tip can be further sharpened; such as by helium-ion milling.
3. NSOM experiments and discussions
In order to verify the shape and brightness of the optical near-field spot generated by the CAN-Tip, we fabricated CAN-Tips, mounted a CAN-Tip probe in a NSOM configuration, and scanned a test structure with vertical step contours. We determined the optical resolution by examination of the optical response. Figure 2(a) shows a custom-built optical fiber-based NSOM  with a CAN-Tip probe. The probe was fabricated on a 100 nm thick pyramid-shaped silicon nitride membrane with a 400 nm thick pre-deposited gold film and a 5 nm thick chromium adhesion layer. The Cr adhesion layer was not incorporated in the FDTD simulations because it was too thin to be adequately resolved by the 2.5 nm cell size. As explained in reference , the effect of an adhesion layer is a reduction in optical efficiency; the agreement between our experimental and simulation results indicate that this layer has no effect on optical resolution. The tip and the C-aperture were milled using an FEI Strata DB 235 FIB operating at 1pA. Figure 2(b) and 2(c) show scanning electron microscope (SEM) images of a probe with α ~40 nm, tm ~130 nm, h ~150 nm, and a ~15-20 nm tip radius. The probe and its supporting silicon nitride membrane were aligned and glued to the end of a cleaved optical fiber (Corning HI-980) with UV-cured optical adhesive. The sample used for the NSOM scan consisted of a series of 20 nm thick Cr disks of different sizes on a 0.17 mm thick glass substrate and was fabricated using e-beam lithography and metal lift-off. To illuminate the CAN-tip, a 980 nm CW laser source was coupled to the optical fiber with its polarization parallel to the ridge of the aperture. The tip-sample gap was held constant using tuning fork based shear-force feedback . The scattered light was collected with a 0.4 NA microscope objective lens at the bottom of the sample substrate and recorded by a femto-watt photodetector (New Focus Model 2151).
While scanning, the tip of the probe simultaneously obtains optical information and topography information (AFM response). The noise in the optical signal predominantly comes from the noise of the femto-watt photodetector and results in a signal-to-noise ratio of 8.9. The AFM response (Fig. 3(a) , top) delineates the geometry of the Cr disk along the line scan. When the probe is scanned across the smaller disks (Fig. 3(a), left), enhancement in transmission is observed. Interestingly, when scanning across a larger disk (Fig. 3(a), right), the transmission exhibits a ripple at a distance from the disk edge. The reason for this behavior is not immediately obvious. FDTD was used to further illuminate these experimental results, simulating the structures enclosed within the dotted lines of Fig. 2(a). The simulation parameters are predominantly unchanged from before. The additional chromium layer is modeled as a Drude material with ε∞ = 2.2629, ωp = 1.2535 × 1017 s−1, and τc = 4.6854 × 10−18 s. Three line scans across a 20 nm tall round chromium disk with a diameter of 282.5 nm are done according to Fig. 4(a) ; line 1 bisects the disk and line scans 2 and 3 are 70 nm and 105 nm off-center respectively. To calculate the magnitude of scattered light measured through the microscope objective lens, the far-field transmission is predicted with a near-field to far-field transformation , and integrated over a NA 0.4 solid angle (Fig. 4(b)). Visualization of the near-field around the disks (Fig. 4(c) and 4(d)) helps to explain the transmission ripple from Fig. 3(a). Strong surface plasmons are excited around the circumference of the disk when the probe is over the top of the Cr disk. As the CAN-Tip probe is close to the middle of a disk along lines 1 and 2, the drop in the far-field transmission matches the experimental observation from Fig. 3(a). This can be better understood by reference to the radiation pattern shown in Fig. 4(e) and 4(f). In Fig. 4(e), the radiation is predominately directed along the z-axis when the CAN-Tip is above the edge of the Cr disk. However, in Fig. 4(f), when the probe is above the center, the radiation mainly propagates away from the z-axis at angles outside the NA = 0.4 collection cone. The drop in intensity at the middle of the scan is the direct result.
In Fig. 3, the closely matched sharp simultaneous responses of the AFM and NSOM prove that the optical resolution of the CAN-Tip is mainly determined by the shape of the tip; no background illumination was present in the result. The predictions of the AFM and NSOM responses assume that the Cr disks have vertical step contours and that the optical near-field spot generated by the CAN-Tip has a Gaussian intensity distribution (shown as red dashed lines in Fig. 3(b)). A ~16.1 nm optical resolution and ~28.5 nm AFM resolution were obtained from the scan.
In summary, we have designed a new near-field probe, a nano-antenna attached to a resonant C-aperture, which provides strong, background-free optical near-field intensity and ultra-high resolution. The tip of the CAN-tip increases the effective thickness of the C-aperture and enables 11.6x higher intensity enhancement than a planar C-aperture by matching the aperture surface resonance and hybrid waveguide-tip resonance. An NSOM scan over Cr structures experimentally proves that a λ/60 near-field spot size can be generated with a CAN-Tip in the near-IR regime. The near-field spot of a CANTip can be further scaled down by employing a sharper tip, and higher near-field intensity can be obtained consequently. The CAN-Tip design may be modified with other ridge nano-apertures, such as “bowtie” nano-apertures, H- (or I-) nano-apertures, and fractal nano-apertures (“frapertures”) . The CAN-Tip enables a range of near-field technologies that require high optical intensity with nano-scale resolution, such as near-field optical recording, heat-assisted magnetic recording (HAMR), tip-enhanced Raman scattering (TERS), and nano-scale electron source generation.
We acknowledge the funding support by the CBIS seed funding at Stanford University.
References and links
1. X. Shi and L. Hesselink, “Design of a C aperture to achieve λ/10 resolution and resonant transmission,” J. Opt. Soc. Am. B 21(7), 1305–1317 (2004). [CrossRef]
3. E. H. Synge, “A suggested method for extending microscopic resolution into the ultra-microscopic region,” Philos. Mag. 6, 356–362 (1928).
4. H. A. Bethe, “Theory of diffraction by small holes,” Phys. Rev. 66(7-8), 163–182 (1944). [CrossRef]
5. X. Shi and L. Hesselink, “Mechanisms for enhancing power throughput from planar nano-apertures for near-field optical data storage,” Jpn. J. Appl. Phys. 41(Part 1, No. 3B), 1632–1635 (2002). [CrossRef]
6. X. Shi, R. L. Thornton, and L. Hesselink, “A nano-aperture with 1000× power throughput enhancement for very small aperture laser system (VSAL),” Proc. SPIE 4342, 320–327 (2002). [CrossRef]
7. Z. Rao, L. Hesselink, and J. S. Harris, “High-intensity bowtie-shaped nano-aperture vertical-cavity surface-emitting laser for near-field optics,” Opt. Lett. 32(14), 1995–1997 (2007). [CrossRef] [PubMed]
8. K. Tanaka and M. Tanaka, “Optimized computer-aided design of I-shaped subwavelength aperture for high intensity and small spot size,” Opt. Commun. 233(4-6), 231–244 (2004). [CrossRef]
9. J. B. Leen, P. Hansen, Y. Cheng, A. Gibby, and L. Hesselink, “Near-field optical data storage using C-apertures,” Appl. Phys. Lett. 97(7), 073111 (2010). [CrossRef]
10. U. C. Fischer, J. Koglin, and H. Fuchs, “The tetrahedral tip as a probe for scanning near-field optical microscopy at 30 nm resolution,” J. Microsc. 176, 231–237 (1994). [CrossRef]
11. D. P. Fromm, A. Sundaramurthy, P. J. Schuck, G. Kino, and W. E. Moerner, “Gap-dependent optical coupling of single “bowtie” nanoantennas resonant in the visible,” Nano Lett. 4(5), 957–961 (2004). [CrossRef]
13. T. H. Taminiau, R. J. Moerland, F. B. Segerink, L. Kuipers, and N. F. van Hulst, “λ/4 resonance of an optical monopole antenna probed by single molecule fluorescence,” Nano Lett. 7(1), 28–33 (2007). [CrossRef] [PubMed]
14. H. G. Frey, F. Keilmann, A. Kriele, and R. Guckenberger, “Enhancing the resolution of scanning near-field optical microscopy by a metal tip grown on an aperture probe,” Appl. Phys. Lett. 81(26), 5030–5032 (2002). [CrossRef]
15. K. Tanaka, M. Tanaka, and T. Sugiyama, “Metallic tip probe providing high intensity and small spot size with a small background light in near-field optics,” Appl. Phys. Lett. 87(15), 151116 (2005). [CrossRef]
18. B. Hou, X. Q. Liao, and J. K. S. Poon, “Resonant infrared transmission and effective medium response of subwavelength H-fractal apertures,” Opt. Express 18(4), 3946–3951 (2010). [CrossRef] [PubMed]
20. H. Aouani, J. Wenger, D. Gérard, H. Rigneault, E. Devaux, T. W. Ebbesen, F. Mahdavi, T. Xu, and S. Blair, “Crucial role of the adhesion layer on the plasmonic fluorescence enhancement,” ACS Nano 3(7), 2043–2048 (2009). [CrossRef] [PubMed]
21. K. Karrai and R. D. Grober, “Piezoelectric tip-sample distance control for near field optical microscopes,” Appl. Phys. Lett. 66(14), 1842–1844 (1995). [CrossRef]
22. A. Taflove and S. C. Hagness, Computational Electrodynamics: The Finite-Difference Time-Domain Method (Artech House, 2005), Chap. 8.