Resonant Antenna Probes for Tip-Enhanced Infrared Near-Field

Letter pubs.acs.org/NanoLett

Resonant Antenna Probes for Tip-Enhanced Infrared Near-Field Microscopy Florian Huth,†,‡ Andrey Chuvilin,†,§ Martin Schnell,† Iban Amenabar,† Roman Krutokhvostov,† Sergei Lopatin,∥ and Rainer Hillenbrand*,†,§ †

CIC nanoGUNE Consolider, 20018 Donostia−San Sebastián, Spain Neaspec GmbH, 82152 Martinsried, Germany § IKERBASQUE, Basque Foundation for Science, 48011 Bilbao, Spain ∥ FEI Electron Optics, 5651 GG Eindhoven, Netherlands ‡

ABSTRACT: We report the development of infrared-resonant antenna probes for tip-enhanced optical microscopy. We employ focused-ion-beam machining to fabricate high-aspect ratio gold cones, which replace the standard tip of a commercial Si-based atomic force microscopy cantilever. Calculations show large field enhancements at the tip apex due to geometrical antenna resonances in the cones, which can be precisely tuned throughout a broad spectral range from visible to terahertz frequencies by adjusting the cone length. Spectroscopic analysis of these probes by electron energy loss spectroscopy, Fourier transform infrared spectroscopy, and Fourier transform infrared nearfield spectroscopy corroborates their functionality as resonant antennas and verifies the broad tunability. By employing the novel probes in a scattering-type near-field microscope and imaging a single tobacco mosaic virus (TMV), we experimentally demonstrate highperformance mid-infrared nanoimaging of molecular absorption. Our probes offer excellent perspectives for optical nanoimaging and nanospectroscopy, pushing the detection and resolution limits in many applications, including nanoscale infrared mapping of organic, molecular, and biological materials, nanocomposites, or nanodevices. KEYWORDS: Near-field microscopy, s-SNOM, resonant near-field probes, infrared antennas, infrared nanospectroscopy, FIB

O

dium oxide,14,15 of chemical composition of polymers16−19 and biological objects,20,21 of strain in ceramics,22 or of plasmons in graphene.23,24 s-SNOM offers an excellent optical resolution in the 10 nm range, independent of the wavelength.9 It is typically based on atomic force microscopy (AFM) where the tip is illuminated with a focused laser beam and the tip-scattered light is detected simultaneously to topography. Using metallic tips, the strong optical near-field interaction between tip and sample modifies the scattered light allowing for probing the local dielectric properties with nanoscale resolution. Unavoidable background contributions are suppressed by vertical tip oscillation at frequency Ω and subsequent higher-harmonic demodulation of the detector signal at nΩ with n ≥ 2 (ref 9). Combining this higher harmonic demodulation with interferometric detection, background-free near-field optical amplitude sn and phase φn contrast imaging is possible. In typical IR s-SNOM studies, standard metalized AFM tips are used. The infrared antenna performance of these tips,25 however, is widely unexplored and antenna concepts26−40 have

ptical spectroscopy, particularly in the infrared (IR) and terahertz (THz) frequency range, has tremendous merit in chemical and structural analysis of materials and in assessment of free-carrier properties. IR and THz radiation addresses a rich variety of light-matter interactions because photons in this low energy range can excite molecular vibrations and phonons, as well as plasmons and electrons of nonmetallic conductors,1−4 currently motivating major efforts in the development of IR and THz imaging systems.5,6 However, the diffraction-limited spatial resolution has been preventing applications in micro- and nanoscale material and device analysis, industrial failure analysis, and quality control. Recently, tip-enhanced near-field optical microscopy has become a valuable method for nanoscale material characterization, which enables optical spectroscopies such as fluorescence, Raman, infrared, or terahertz to be performed with nanoscale spatial resolution.7,8 The promising and wide application potential of this microscopy principle has been demonstrated in numerous studies. At infrared frequencies, scattering-type scanning near-field optical microscopy (sSNOM)9 based on field enhancement at the apex of sharp metal tips enables, for example, the nanoscale mapping of free carriers in transistors,10 of single electrons in semiconductor quantum dots,11 of semiconductor nanowires12,13 and vana© 2013 American Chemical Society

Received: November 20, 2012 Revised: January 16, 2013 Published: January 30, 2013 1065

dx.doi.org/10.1021/nl304289g | Nano Lett. 2013, 13, 1065−1072

Nano Letters

Letter

not yet been applied to optimize near-field probes for the infrared spectral range. Here we demonstrate the successful fabrication of infrared-resonant antenna tips on standard Si cantilevers, by employing focused ion beam (FIB) machining. Characterization of these tips by electron energy loss spectroscopy (EELS), Fourier transform infrared (FTIR) spectroscopy, and Fourier transform infrared near-field spectroscopy (nano-FTIR) clearly reveals geometrical antenna resonances in the tips, which are found to be in good agreement with numerical calculations. We verify the probe performance of these tips by infrared near-field imaging of tobacco mosaic viruses of only 18 nm diameter. The design of the anticipated near-field probes is illustrated in Figure 1a. A high-aspect ratio gold cone with a sharp tip apex replaces the tip of a standard Si AFM cantilever. We first study numerically the antenna function of such tips (using a commercial FDTD software package, Lumerical). To that end we consider a tip geometry as illustrated in Figure 1b, where we assume a plane wave illumination (Einc) with p-polarization at an angle α with respect to the tip axis. Figure 1c shows the calculated near-field enhancement at the apex of three differently long antenna tips for α = 0°. For the 0.5 μm long antenna (shown in red) we find a pronounced resonance peak at a frequency ωres,1 = 5120 cm−1 (λ ≈ 1.95 μm), corresponding to the fundamental dipolar mode (l = 1) and a second, weaker peak at ωres,3 = 13 420 cm−1 (λ ≈ 0.75 μm), corresponding to the second order dipolar mode (l = 3).41 With increasing rod length L, the resonance positions shift to lower frequencies ω, from the near-IR to the mid-infrared to the THz spectral range. Compared to classical radio wave antennas,42 where the resonance length Lres = λ/2 is about half the excitation wavelength λ, the tip resonances occur at much shorter lengths. We find Lres = λ/3.9, λ/2.7 and λ/2.4 for the near-IR, mid-IR and THz spectral range, respectively, which we assign to the tip geometry, the finite frequency-dependent conductivity of the metal and the Si base. The latter acts as a capacitive load, which typically redshifts the antenna resonance.43,44 An interesting aspect is the significant increase of the field enhancement at the tip apex, amounting to more than 100, when the tip length increases, that is, when the resonance shifts to the THz spectral range. We assign this finding to the increased sharpness of the tip relative to the wavelength and the rod-length. In Figure 1d we show calculated near-field spectra for α = 30°, which is a typical situation in s-SNOM. Additional to the dipolar modes (l = 1, 3), the plane wave efficiently couples to the odd modes (l = 2, 4),45 due to retardation along the tip axis. Under normal incident the odd modes are “dark”, owing to the vanishing total dipole moments because of a centrosymmetric charge distribution. The proposed antenna tips promise precise tuning and optimization of the local field enhancement at the tip-apex throughout a broad spectral region from visible to THz frequencies. Tilted illumination thereby excites both even and odd modes, which increases the usable spectral bandwidth of the tips, although the maximum field enhancement decreases slightly. For comparison, we show in Figure 1e the calculated field enhancement at the tip apex of a low aspect ratio Au pyramid with a length of 10 μm and a tip angle of 60° (green solid line). This geometry resembles that of commercially available AFM tips, which are typically used in s-SNOM experiments. Several resonances can be seen, which are broader and weaker than the resonances observed for the high-aspect ratio gold cones (red and blue curves, same as in Figure 1d). This can be understood

Figure 1. Resonant antenna probes for near-field microscopy. (a) Sketch of a Au-rod placed on a Si-based cantilever. (b) Model to simulate the near-field enhancement at the tip apex for rods of different lengths L placed on a Si base of 1 μm length. (c) Calculated near-field (ENF) at 10 nm below the tip apex (marked with “x” in panel b), normalized to the incident field (Einc), which is linear polarized at an angle α = 0° with respect to the tip axis. For better visibility, the spectra of the 3.3 μm (blue) and 0.5 μm (red) long antenna tip have been scaled by a factor of 2 and 5, respectively. (d) Same as panel c but for a polarization tilted by α = 30° with respect to the tip axis. (e) Calculated near-field enhancement (ENF) 10 nm below the apex of a pyramidal shaped full metal tip with a length L = 10 μm and a tip angle of 60° (green spectrum). Incident polarization is under an angle of α = 30° relative to the tip axis. For comparison, spectra from panel d are shown (blue and red spectra). The tip radius for all calculations was set to 20 nm.

in analogy to conical radio wave antennas. Acting as broadband antennas,42 the low aspect ratio pyramids provide enhanced fields in a broad spectral range but with lower maximum values on resonance (green spectrum). In the mid-IR spectral range, the high aspect ratio gold cones (black, blue, and red spectra) thus offer an improvement of the resonant field enhancement by factor up to 3 (corresponding to an intensity enhancement factor of 9), which is even higher in the far-IR (THz) and visible spectral regions. The improvement of the field enhancement by a factor of 4 in the near-IR spectral region could push the Raman signal in tip-enhanced Raman spectroscopy (TERS) by a factor of more than 250, as the 1066


Nano Letters

Letter

Figure 2. Fabrication of antenna tips by focused-ion-beam (FIB). (a) Several rods cut from a 25 μm thick Au wire. (b) Single Au rod attached to the cantilever. (c) Intermediate step of sharpening the tip with FIB. The ion beam (red arrow) is moved along the red dashed line. (d) Tip sharpening by circular ion milling (as indicated by the red dashed circle). (e) Final antenna tip.

Raman signal scales with the forth power of the field enhancement.8 Figure 2 shows the fabrication procedure of the proposed antenna tips by focused ion beam (FIB) machining. The fabrication was done with a Helios600 DualBeam (FEI, Netherlands) electron microscope. We use standard Si atomic force microscopy (AFM) cantilevers (Nanosensors, Switzerland), where we first cut the tip apex. Then a several micrometer long rod is milled out of a 25 μm thick Au wire (Figure 2a) and attached to the cut Si tip by focused ion beam induced deposition (FIBID) of SiOx (Figure 2b). As illustrated in Figure 2c,d, the Au rod is shaped by the ion beam until a sharp high aspect ratio cone is formed (Figure 2e). By adjusting the milling parameters, such as the ion beam current and the ion beam focusing, the conical angle of the tip can be varied. Simultaneous monitoring of the ion milling process with the electron beam allows for controlling the desired length of the Au cone with a precision of ±10 nm. The length can be varied between 200 nm and several tens of micrometers, yielding resonant antennas for the visible, IR, and THz spectral ranges. Remarkably, the radius of curvature of the tip apex can be as small as 10 nm (see Figure 3a). We first study the antenna tips by EELS. This method allows for both local plasmon spectroscopy and spatial plasmon mapping, the latter directly revealing the plasmonic eigenmodes of the antennas.46−49 We use a Titan 60-300 (FEI, Netherlands) electron microscope equipped with a gun monochromator and an EELS spectrometer (GIF Quantum, Gatan, U.S.A.). For an accelerating voltage of 80 kV, the energy filter can reach an energy resolution of approximately 100 meV. This gives access to electron energy losses starting from 400 meV (corresponding to ω = 3200 cm−1), which allows for the study of plasmon resonances in the visible and near-IR spectral range. We use EELS spectrum imaging (SI) in scanning transmission electron microscope (STEM) mode to acquire spatially resolved EELS spectra. This method acquires a full resolution EELS spectrum at each pixel of the image. The recorded data cube is further processed to obtain either cumulative EELS spectra over a defined area or a spatial intensity distribution of

Figure 3. EELS of FIB-fabricated antenna tips. (a) STEM image of a 0.5 μm long antenna tip. (b) EELS spectrum recorded at tip apex. (c) EELS maps at the peak positions in the energy loss spectrum.

particular energy losses (e.g., plasmon maps). In Figure 3, we show the results obtained for a 0.5 μm long antenna tip. The STEM image (Figure 3a) shows the tip with an apex radius of about 10 nm. Recording an EELS spectrum (Figure 3b) at the tip apex reveals maxima at 0.75, 1.33, and 1.8 eV. By mapping the energy loss at these energies (Figure 3c), we can unambiguously assign the antenna modes.41,46−49 The two bright spots in the 0.75 eV map clearly reveal the l = 1 (fundamental dipolar) plasmon mode. At 1.33 eV we observe three maxima, visualizing the l = 2 (“dark”) mode of the tips. The l = 3 (second order dipolar) mode can be identified by the four maxima seen in the map at 1.8 eV. The EELS spectrum is in good agreement with the numerically calculated optical spectrum of the 0.5 μm long antenna tip shown in Figure 1, 1067


Nano Letters

Letter

concentrated infrared spot at the tip apex. To that end, we recorded the near-field amplitude signals sn (for the harmonics n = 1, 2, 3, and 4) as a function of the distance z between the tip and a silicon surface (so-called approach curve). Figure 5b

experimentally verifying the function of the tips as resonant optical antennas. The tip fabrication process offers great tunability of the antenna resonance, by simply adjusting the tip length. With lengths in the micrometer range we obtain resonances in the mid-infrared spectral range (4−20 μm wavelength), which is difficult to access by EELS. We thus test the mid-infrared performance of these tips by directly employing them as nearfield probes in our near-field microscopy and spectroscopy setup shown in Figure 4a. It is based on a commercial s-SNOM

Figure 5. AFM and s-SNOM performance of antenna tips. (a) SEM image of a 4.8 μm long antenna tip. (b) Approach curves on a flat Ausurface showing the IR amplitude signal sn as a function of tip−sample distance. The amplitude of the tip vibration was set to 20 nm. The tip was illuminated with the QCL operated at 1660 cm−1 and the laser power in the focus was 1.5 mW. (c) AFM-topography image of a single tobacco mosaic virus. (d) Simultaneously acquired near-field amplitude image s3 (tip-illumination set to 1660 cm−1). (e) Near-field phase images φ3 for tip-illumination set to 1660 (right) and 1725 cm−1 (left).

Figure 4. Near-field imaging and spectroscopy setup. (a) Schematics showing an s-SNOM employing a tunable single line laser (quantum cascade laser, QCL) or a broadband plasma source (high-temperature argon arc source, HTAAS) for tip illumination. The illumination source can be changed with a flip mirror (FM). The light backscattered from the tip is analyzed with a Fourier transform spectrometer, comprising a beamsplitter (BS, uncoated ZnSe), a reference mirror (RM), and a detector. (b) Output spectrum of the plasma source. The available frequencies are limited by the cutoff of the detector at the lower frequency side, and a low-pass filter (AntiReflex coated Ge window, not shown) at the high-frequency side. The dips in the spectrum are due to atmospheric absorption of air (e.g., water vapor, CO2), which the IR beam has to pass.

shows the approach curves we obtained with a 4.8 μm long tip (shown in Figure 5a) that was illuminated with light of 6 μm wavelength (ω = 1660 cm−1). We find that already for n > 1 the amplitude signals rapidly decay to the noise level when z increases, revealing strong near-field confinement at the tip apex and that the amplitude signals are essentially background free. We note that the illumination wavelength has been chosen in order to match an expected resonance of the 4.8 μm long antenna tip, which is verified below in Figure 6. The performance of the antenna tips for topography and near-field imaging is demonstrated in Figure 5c−e. For a test object, we have chosen a tobacco mosaic virus (TMV) because of its well-defined geometry and infrared response. TMVs have a diameter of 18 nm and consist of pure protein, exhibiting a well-known molecular vibrational resonance at 1660 cm−1 (amide I band). In the topography image (Figure 5c), the virus can be easily identified as a 450 nm long rod with a height of about 18 nm. In the amplitude image s3 (Figure 5d), a decrease of the signal can be seen on the virus, which can be assigned to the lower refractive index of the protein compared to the Si substrate. The phase image φ3 recorded at 1660 cm−1 (Figure 5e, right) shows a distinct contrast between the virus and the Si substrate, amounting to approximately 20°. This

(Neaspec, Germany), where cantilevered AFM tips are used as near-field probes. Operating the AFM in tapping mode, the tip is vertically vibrating with an amplitude of about 50 nm at the mechanical resonance frequency Ω of the cantilever, in this work at about Ω ∼ 120 kHz. We perform mid-IR near-field imaging by illuminating the FIB-fabricated antenna tips with a quantum cascade laser (QCL, Daylight Solutions, U.S.A.), tunable between 1550 and 1750 cm−1 (Figure 4a). A Michelson interferometer is used to analyze the tip-scattered light. Unavoidable background signals can be efficiently suppressed by demodulating the detector signal at a higher harmonic of the tapping frequency, nΩ (ref 9). Employing a pseudoheterodyne detection scheme50 (where the reference mirror RM oscillates at a frequency M) enables the simultaneous detection of both amplitude sn and phase φn signals. We first validate the mid-infrared antenna function of the tips, that is, the conversion of the incident light into a highly 1068


Nano Letters

Letter

To study the spectral near-field characteristics of the tips, we employ the Michelson interferometer shown in Figure 4a as Fourier transform infrared (FTIR) spectrometer, thus operating the s-SNOM in nano-FTIR mode. For tip illumination we use the radiation from a current-induced high-temperature argon arc source (HTAAS, in the following referred to as plasma source).51 Its output spectrum corresponds to that of a blackbody emitter of about 10 000 K temperature, providing a spectral energy density comparable to synchrotron radiation.52 Figure 4b shows the output spectrum, which spans the whole mid-infrared region from 700 up to 2700 cm−1. FTIR spectroscopy of the light scattered by the tip is performed by recording the near-field amplitude signal s2 as a function of the reference mirror (RM) position d. Fourier transformation of the interferogram s2(d) subsequently yields the near-field spectrum s2(ω) with ω being the infrared frequency. We previously used this setup with standard AFM tips and a broadband laser source for nanospectroscopy of samples exhibiting infrared vibrational resonance.19 In order to obtain the infrared response of the antenna tips, we perform spectroscopy of the scattered light when the tip is in nearfield interaction with a gold surface, which exhibits a flat spectral response in the mid-infrared spectral range. Figure 6 shows representative infrared near-field (nanoFTIR) spectra of antenna tips with different lengths, varying between L = 3.3 and 5.2 μm. In the spectrum of the 3.7 μm long antenna tip we find two peaks, one around ωres,1 = 950 cm−1 (corresponding to a wavelength of λ ≈ 10.5 μm) and one around ωres,2 = 2100 cm−1 (λ ≈ 4.8 μm). Comparing the peak positions with the calculations in Figure 1 (blue curve), we can assign the peak around 950 cm−1 to the fundamental dipolar mode (l = 1) and the second peak to the second order resonance (l = 2). With increasing antenna length (to L = 4.8 and 5.2 μm), the resonance peaks of the first and second order modes (l = 1 and l = 2) shift to lower frequencies. This clearly demonstrates the possibility to tune the resonances by simple adjustment of the tip length. Interestingly, we observe the l = 2 mode, as predicted in Figure 1c. To corroborate this mode assignment, we calculate the field distribution around a L = 3.7 μm long antenna tip (as illustrated in Figure 1b), situated above a metal surface. It is illuminated with p-polarized light at an illumination angle of 30° relative to the surface. Figure 6b shows the field distributions for illumination at ωres,1 = 950 cm−1 and ωres,2 = 2100 cm−1, the two frequencies where we observe the experimental resonance peaks in Figure 6a (L = 3.7 μm). The two spatial field maxima observed at 950 cm−1 show the fundamental dipole mode (l = 1), while the 3 maxima at 2100 cm−1 clearly reveal the l = 2 mode. For both modes, we find the strongest field concentration at the tip apex, which is the key to nanoscale resolved infrared mapping, as demonstrated in Figure 5c−e. Because of the pronounced resonances of the tips, the near-field signals will be enhanced only in certain spectral regions, which, however, can be readily addressed by adjustment of the tip length. For comparison, Figure 6c shows the near-field spectra of three Au-coated AFM tips of the same type (Nanosensors, PPP-NCST-Au). We observe a spectrally oscillating near-field signal, which tends to lower values with increasing frequency. The calculations shown in Figure 1e confirm such a behavior qualitatively. However, the details of the three spectra differ significantly, although the tips have the same geometry and nominal tip length. We assign this to variations in the length of

Figure 6. Infrared near-field response of antenna tips. (a) Nano-FTIR spectra of antenna tips with different lengths L in contact with a flat Au surface. The spectra are normalized to the input spectrum of the plasma source and the maximum value of each spectrum is set to 1. The small dips or peaks around 2350 cm−1 we assign to changing atmospheric CO2 absorption between the measurement of the tip spectrum and the input spectrum. (b) Calculated field distribution (logarithmic color scale) around a 3.7 μm long antenna tip placed 20 nm above a metal surface, assuming p-polarized plane wave illumination under an angle of 60° relative to the tip axis. Illumination as indicated by the white arrow (at the given frequencies). (c) NanoFTIR spectra of three commercial Au-tips (PPP-NCST-Au, Nanosensors). Experimental conditions and representation are the same as in (a).

phase contrast can be explained by the molecular absorption of the sample at this illumination frequency.17,19,20 As expected, the phase contrast vanishes when the illumination is tuned to 1725 cm−1, where the protein does not absorb (Figure 5e, left).20 1069


Nano Letters

Letter

optical and infrared frequencies we have to consider the finite conductivity of metals, plasmons, as well as aspect ratios that are much smaller than 103. On the other hand, Novotny54 showed that antenna designs can be transferred to the optical frequency regime by replacing λ by a linearly scaled effective wavelength λeff = n1 + n2λ/λp, with λp being the plasmon wavelength and n1, n2 being coefficients that depend on geometry and material properties. In this case, the resonance wavelength of the mode l can be found according to

the commercial tips, which have been observed by electron microscopy.53 We additionally verify the infrared resonances of the FIB fabricated tips by recording relative transmission spectra with a standard FTIR microspectrometer (Bruker, Hyperion 2000 microscope coupled to a Vertex 70 FTIR spectrometer). The tips are illuminated with light polarized either parallel or perpendicular to the tip axis. A small aperture (15 μm × 15 μm) is employed to isolate the infrared response of the tip. Figure 7 shows two representative spectra, obtained with

λres, l =

2L l

−n1 n2

λp

(1)

At infrared frequencies and for micrometer long gold antennas with high aspect ratio (>10), n1 becomes very small and eq 1 becomes λres,l = (2λp)/(n2l) L = blL, with bl = (2λp)/ (n2l), which for the case of bl = 2/l yields the relation for an ideal antenna. Fitting the data of Figure 8 with eq 1, we find b1 = 2.9 for the nano-FTIR (blue) and b1 = 2.4 for the FTIR experiments (green). Obviously, the nano-FTIR resonances are red shifted compared to FTIR. We assign this to the near-field interaction between the antenna tip and the gold sample, which is absent in the case of the FTIR measurements. For the second order resonance (l = 2) we obtain b2 = 1.3 ≈ b1/2, which confirms our mode assignment. Altogether Figure 8 clearly corroborates that the optical response of the tips is governed by geometrical antenna resonances, which can be understood and described within the framework of optical antenna theory. We finally compare a resonant tip of 1.92 μm length (Figure 9a) with a standard commercial Au tip (Figure 9b). To that end

Figure 7. Relative IR transmittance spectra of antenna tips with a length L = 4.5 μm and L = 4.8 μm.

antenna tips of L = 4.5 and 4.8 μm length and for polarization parallel to the tip axis. Both show a clear dip in the relative transmission, appearing at around ωres,1 = 850 and 920 cm−1, respectively. Comparing with the calculations of Figure 1, we assign these dips to the fundamental dipolar antenna mode (l = 1). In a control experiment, we recorded spectra of the transmitted light when illuminating the tip with light polarized perpendicular to the tip axis (not shown). The dips in the transmission vanish, corroborating that the dips in Figure 7 can be assigned to resonant infrared excitations along the tip axis. To summarize our results, we plot in Figure 8 the resonance wavelengths λres,l ∝ 1/ωres,l obtained by EELS (red symbols),

Figure 9. Comparison of the performance of resonant and standard tips under the same conditions. (a) Results obtained with a 1.92 μm long tip, resonant at 1660 cm−1. (Top) Topography (left) and phase image φ3 (right) of a TMV virus. (Bottom) Approach curves on a flat Si substrate. (b) Results obtained with a standard tip (Nanosensors, PPP-NCST-Au). (Top) Topography (left) and phase image φ3 (right) of the same TMV virus as in (a). (Bottom) Approach curves on a flat Si substrate. The approach curves of the 1.92 μm long tip have been normalized to the signal at z = 0 nm. The approach curves of the standard tip have been normalized to the signal at z = 0 nm obtained with the resonant tip at the corresponding harmonic n. For imaging, both tips were illuminated with 0.3 mW optical power at 1660 cm−1 and both tips were oscillating with 20 nm tapping amplitude. For approach curves, both tips were illuminated with 0.15 mW optical power at 1660 cm−1 and both tips were oscillating with 50 nm tapping amplitude.

Figure 8. Resonance wavelength λres as a function of the tip length L, measured by EELS (red symbols, data extracted from Figure 3b), nano-FTIR (blue symbols, data extracted from Figure 6) and FTIR (green symbols, data extracted from Figure 7). The lines represent linear fits to the corresponding data points.

nano-FTIR (blue symbols) and FTIR (green symbols) as a function of the tip length L. We find a linear behavior for the nano-FTIR and FTIR results, as expected from antenna theory.42 However, the simple relation λres,l = 2L/l of antenna theory does not fit our data. This is because antenna theory assumes ideal metals and very large aspect ratios (the antenna length compared to antenna width is typically >103),42 while at 1070


Nano Letters we imaged one and the same TMV virus and recorded approach curves on a Si substrate with both tips employing the same imaging parameters. The length of the FIB fabricated tip was chosen in order to exhibit the fundamental resonance at 1660 cm−1 (see Figure 8). We applied an illumination power of only 0.3 and 0.15 mW for imaging and approach curves, respectively, in order to better see the signal-to-noise ratio in the images and approach curves and to demonstrate the performance of the tips using the typically weak power of broadband infrared sources.17,19,55−57 After maximizing the near-field signals for each tip, we obtain about two times more amplitude signal with the 1.92 μm long tip, in reasonable agreement with the calculations shown in Figure 1e. We note that the absolute near-field signals depend on the alignment of the laser focus and the interferometer, which lets us consider this comparison as an estimate rather than an exact quantitative evaluation. We find that the phase contrast φ3 of the virus obtained with the resonant tip (Figure 9a, upper row) is about two-fold enhanced compared to the image obtained with the standard tip (Figure 9b, upper row), while the noise is reduced. Recording approach curves (amplitude sn vs tip−sample distance z), we also find that the near-field confinement at the tip apex is better for the antenna tip. The amplitude signal s3 decays to half its value at a tip−sample distance of zd = 4.6 nm (Figure 9a), while for the standard tip we measure zd = 8.6 nm. The improved confinement at the apex of the antenna tip we assign to the apex geometry and an eventually smaller tip radius. This might also explain the enhanced phase contrast of the virus, however, further studies are needed to separate the different contributions of apex geometry and tip resonance on the image contrasts. In summary, we have demonstrated FIB-assisted fabrication of micrometer-long antenna tips on conventional AFM cantilevers. As verified by EELS, FTIR, and nano-FTIR, the optical resonances of these tips can be precisely tuned from visible to infrared frequencies by simply adjusting the tip length. Employing these tips for near-field infrared imaging of individual viruses, we verified state-of-the-art AFM performance, accompanied by an excellent optical performance. Because of their well-defined geometry, these antenna tips will significantly ease the qualitative description of the tip− sample near-field interaction, which will be essential for quantitative measurements of the local sample properties such as dielectric function and molecular absorption. We furthermore envision a wide application potential of these tips in infrared nanoimaging and nanospectroscopy. Optimization of the tips could further enhance the resolution and sensitivity, eventually enabling single molecule infrared spectroscopy in the future. Shorter tips with resonances at visible and near-infrared frequencies could be applied to push the field enhancement and sensitivity in tip-enhanced Raman spectroscopy (TERS).

■

■

ACKNOWLEDGMENTS

■

REFERENCES

Letter

We acknowledge discussions with T. Taubner (Univ. Aachen). This work was financially supported by an ERC Starting Grant (TERATOMO) and the National Project MAT2009-08398 from the Spanish Ministerio de Ciencia e Innovacion. We also acknowledge financial support from FEI Company (The Netherlands) within the framework of a collaborative project.

(1) Kuzmany, H. Solid-State Spectroscopy; Springer: Berlin, 1998. (2) Ferguson, B.; Zhang, X. C. Nat. Mater. 2002, 1, 26−33. (3) Mittleman, D. Sensing with THz radiation; Springer Verlag: Berlin, 2003. (4) Tonouchi, M. Nat. Photonics 2007, 1, 97−105. (5) Chan, W. L.; Deibel, J.; Mittleman, D. M. Rep. Prog. Phys. 2007, 70, 1325−1379. (6) Withayachumnankul, W.; Png, G. M.; Yin, X. X.; Atakaramians, S.; Jones, I.; Lin, H. Y.; Ung, B. S. Y.; Balakrishnan, J.; Ng, B. W. H.; Ferguson, B.; Mickan, S. P.; Fischer, B. M.; Abbott, D. Proc. IEEE 2007, 95, 1528−1558. (7) Novotny, L.; Stranick, S. Annu. Rev. Phys. Chem. 2006, 57, 303− 331. (8) Novotny, L.; Hecht, B. Principles of Nano-Optics; Cambridge University Press: Cambridge, 2006. (9) Keilmann, F.; Hillenbrand, R. Nano-Optics and Near-Field Optical Microscopy, Artech House: Boston, 2008. (10) Huber, A. J.; Keilmann, F.; Wittborn, J.; Aizpurua, J.; Hillenbrand, R. Nano Lett. 2008, 8 (11), 3766−3770. (11) Jacob, R.; Winnerl, S.; Fehrenbacher, M.; Bhattacharyya, J.; Schneider, H.; Wenzel, M. T.; Ribbeck, H.-G. v.; Eng, L. M.; Atkinson, P.; Schmidt, O. G.; Helm, M. Nano Lett. 2012, 12 (8), 4336−4340. (12) Stiegler, J. M.; Huber, A. J.; Diedenhofen, S. L.; Gómez Rivas, J.; Algra, R. E.; Bakkers, E. P. A. M.; Hillenbrand, R. Nano Lett. 2010, 10 (4), 1387−1392. (13) Stiegler, J. M.; Tena-Zaera, R.; Idigoras, O.; Chuvilin, A.; Hillenbrand, R. Nat. Commun. 2012, 3, 1131. (14) Qazilbash, M. M.; Brehm, M.; Chae, B. G.; Ho, P. C.; Andreev, G. O.; Kim, B. J.; Yun, S. J.; Balatsky, A. V.; Maple, M. B.; Keilmann, F.; Kim, H. T.; Basov, D. N. Science 2007, 318, 1750−1753. (15) Jones, A. C.; Berweger, S.; Wei, J.; Cobden, D.; Raschke, M. B. Nano Lett. 2010, 10 (5), 1574−1581. (16) Taubner, T.; Hillenbrand, R.; Keilmann, F. Appl. Phys. Lett. 2004, 85, 5064−5066. (17) Xu, X.; Rang, M.; Craig, I.; Raschke, M. J. Phys. Chem. Lett. 2012, 3 (13), 1836−1841. (18) Akhremitchev, B. B.; Pollack, S.; Walker, G. C. Langmuir 2001, 17, 2774−2781. (19) Huth, F.; Govyadinov, A.; Amarie, S.; Nuansing, W.; Keilmann, F.; Hillenbrand, R. Nano Lett. 2012, 12 (8), 3973−3978. (20) Brehm, M.; Taubner, T.; Hillenbrand, R.; Keilmann, F. Nano Lett. 2006, 6, 1307−1310. (21) Paulite, M.; Fakhraai, Z.; Li, I. T. S.; Gunari, N.; Tanur, A. E.; Walker, G. C. J. Am. Chem. Soc. 2011, 133 (19), 7376−7383. (22) Huber, A. J.; Ziegler, A.; Köck, T.; Hillenbrand, R. Nat. Nanotechnol. 2009, 4 (3), 153−157. (23) Chen, J.; Badioli, M.; Alonso-González, P.; Thongrattanasiri, S.; Huth, F.; Osmond, J.; Spasenović, M.; Centeno, A.; Pesquera, A.; Godignon, P.; Zurutuza Elorza, A.; Camara, N.; Garcı ́a de Abajo, F. J.; Hillenbrand, R.; Koppens, F. H. L. Nature 2012, 487 (7405), 77−81. (24) Fei, Z.; Rodin, A. S.; Andreev, G. O.; Bao, W.; McLeod, A. S.; Wagner, M.; Zhang, L. M.; Zhao, Z.; Thiemens, M.; Dominguez, G.; Fogler, M. M.; Castro Neto, A. H.; Lau, C. N.; Keilmann, F.; Basov, D. N. Nature 2012, 487 (7405), 82−85. (25) Brehm, M.; Schliesser, A.; Cajko, F.; Tsukerman, I.; Keilmann, F. Opt. Express 2008, 16 (15), 11203−11215. (26) Frey, H. G.; Keilmann, F.; Kriele, A.; Guckenberger, R. Appl. Phys. Lett. 2002, 81, 5030−5032.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare the following competing financial interest(s): R.H. is co-founder of Neaspec GmbH, a company producing scattering-type scanning near-field optical microscope systems, such as the one used in this study. All other authors declare no competing financial interests. 1071


Nano Letters

Letter

(27) Kalkbrenner, T.; Ramstein, M.; Mlynek, J.; Sandoghdar, V. J. Microsc. 2001, 202, 72−76. (28) Kalkbrenner, T.; Hakanson, U.; Schädle, A.; Burger, S.; Henkel, C.; Sandoghdar, V. Phys. Rev. Lett. 2005, 95, 200801. (29) Zou, Y.; Steinvurzel, P.; Yang, T.; Crozier, K. B. Appl. Phys. Lett. 2009, 94 (17), 171107−3. (30) Taminiau, T. H.; Moerland, R. J.; Segerink, F. B.; Kuipers, L.; van Hulst, N. F. Nano Lett. 2007, 7 (1), 28−33. (31) Mivelle, M.; van Zanten, T. S.; Neumann, L.; van Hulst, N. F.; Garcia-Parajo, M. F. Nano Lett. 2012, 12 (11), 5972−5978. (32) Taminiau, T. H.; Stefani, F. D.; Segerink, F. B.; van Hulst, N. F. Nat. Photonics 2008, 2 (4), 234−237. (33) Höppener, C.; Novotny, L. Nano Lett. 2008, 8 (2), 642−646. (34) Johnson, T. W.; Lapin, Z. J.; Beams, R.; Lindquist, N. C.; Rodrigo, S. G.; Novotny, L.; Oh, S.-H. ACS Nano 2012, 6 (10), 9168− 9174. (35) Weber-Bargioni, A.; Schwartzberg, A.; Cornaglia, M.; Ismach, A.; Urban, J. J.; Pang, Y.; Gordon, R.; Bokor, J.; Salmeron, M. B.; Ogletree, D. F.; Ashby, P.; Cabrini, S.; Schuck, P. J. Nano Lett. 2011, 11 (3), 1201−1207. (36) Fleischer, M. Nanotechnol. Rev. 2012, 1 (4), 313−338. (37) Fleischer, M.; Weber-Bargioni, A.; Altoe, M. V. P.; Schwartzberg, A. M.; Schuck, P. J.; Cabrini, S.; Kern, D. P. ACS Nano 2011, 5 (4), 2570−2579. (38) Fleischer, M.; Stanciu, C.; Stade, F.; Stadler, J.; Braun, K.; Heeren, A.; Häffner, M.; Kern, D. P.; Meixner, A. J. Appl. Phys. Lett. 2008, 93 (11), 111114−3. (39) Farahani, J. N.; Pohl, D. W.; Eisler, H.-J.; Hecht, B. Phys. Rev. Lett. 2005, 95 (2), 017402. (40) Berweger, S.; Atkin, J. M.; Olmon, R. L.; Raschke, M. B. J. Phys. Chem. Lett. 2012, 3 (7), 945−952. (41) Aizpurua, J.; Bryant, G. W.; Richter, L. J.; Garcı ́a de Abajo, F. J.; Kelley, B. K.; Mallouk, T. Phys. Rev. B 2005, 71 (23), 235420. (42) Balanis, C. A. Antenna Theory: Analysis and Design; John Wiley & Sons, Inc.: Hoboken, NJ, 2005. (43) Alù, A.; Engheta, N. Phys. Rev. Lett. 2008, 101, 043901. (44) Alù, A.; Engheta, N. Nat. Photonics 2008, 2 (5), 307−310. (45) Dorfmüller, J.; Vogelgesang, R.; Weitz, R. T.; Rockstuhl, C.; Etrich, C.; Pertsch, T.; Lederer, F.; Kern, K. Nano Lett. 2009, 9 (6), 2372−2377. (46) Nelayah, J.; Kociak, M.; Stéphan, O.; Garcı ́a de Abajo, F. J.; Tencé, M.; Henrard, L.; Taverna, D.; Pastoriza-Santos, I.; Liz-Marzán, L. M.; Colliex, C. Nat. Phys. 2007, 3 (5), 348−353. (47) Bosman, M.; Keast, V. J.; Watanabe, M.; Maaroof, A. I.; Cortie, M. B. Nanotechnology 2007, 18 (16), 165505. (48) Rossouw, D.; Couillard, M.; Vickery, J.; Kumacheva, E.; Botton, G. A. Nano Lett. 2011, 11 (4), 1499−1504. (49) Koh, A. L.; Fernández-Domínguez, A. I.; McComb, D. W.; Maier, S. A.; Yang, J. K. W. Nano Lett. 2011, 11 (3), 1323−1330. (50) Ocelic, N.; Huber, A.; Hillenbrand, R. Appl. Phys. Lett. 2006, 89, 101124. (51) Bridges, J. M.; Migdall, A. L. Metrologia 1995/1996, 32 (6), 625−628. (52) McIntosh, A. L.; Wofford, B. A.; Lucchese, R. R.; Bevan, J. W. Infrared Phys. Technol. 2001, 42 (6), 509−514. (53) Personal communication with T. Taubner, University of Aachen. (54) Novotny, L. Phys. Rev. Lett. 2007, 98 (26), 266802. (55) Huth, F.; Schnell, M.; Wittborn, J.; Ocelic, N.; Hillenbrand, R. Nat. Mater. 2011, 10 (5), 352−356. (56) Amarie, S.; Keilmann, F. Phys. Rev. B 2011, 83 (4), 045404. (57) Amarie, S.; Zaslansky, P.; Kajihara, Y.; Griesshaber, E.; Schmahl, W. W.; Keilmann, F. Beilstein J. Nanotechnol 2012, 3, 312−323.

1072


Resonant Antenna Probes for Tip-Enhanced Infrared Near-Field

Recommend Documents