c5nr08548a 5634..5640 -...

Nanoscale

PAPER

Cite this: Nanoscale, 2016, 8, 5634

Received 2nd December 2015,Accepted 10th February 2016

DOI: 10.1039/c5nr08548a

www.rsc.org/nanoscale

Highly efficient plasmonic tip design for plasmonnanofocusing in near-field optical microscopy

Takayuki Umakoshi, Yuika Saito and Prabhat Verma*

Near-field scanning optical microscopy (NSOM) combined with plasmon nanofocusing is a powerful

nano-analytical tool due to its attractive feature of efficient background suppression as well as light

energy compression to the nanoscale. In plasmon nanofocusing-based NSOM, the metallic tip plays an

important role in inducing plasmon nanofocusing. It is, however, very challenging to control plasmonic

properties of tips for plasmon nanofocusing with existing tip fabrication methods, even though the plas-

monic properties need to be adjusted to experimental environments such as the sample or excitation

wavelength. In this study, we propose an efficient tip design and fabrication which enable one to actively

control plasmonic properties for efficient plasmon nanofocusing. Because our method offers flexibility in

the material and structure of tips, one can easily modify the plasmonic properties depending on the

requirements. Importantly, through optimization of the plasmonic properties, we achieve almost 100%

reproducibility in plasmon nanofocusing in our experiments. This new approach of tip fabrication makes

plasmon nanofocusing-based NSOM practical and reliable, and opens doors for many scientists working

in related fields.

1. Introduction

Nanofocusing of a light field through a plasmon has receivedincreased attention because of its attractive feature of compres-sing light energy into a nanometric volume. Until now, manyefforts have been made not only for scientific studies ofplasmon behavior,1–7 but also for various applications, such asoptical sensing,8 optical signal control,9,10 nano-lithographyand optical nano-circuits.11 In particular, near-field scanningoptical microscopy (NSOM) with plasmon nanofocusing hasrecently emerged as a great combination that enhances theperformance of NSOM, not just because of light energy com-pression to the nanoscale but also because of the drastic sup-pression of background scattering.12–15 NSOM offers nanoscalespatial resolution as well as abundant information throughoptical analysis via near-field light excited at the apex of a met-allic tip.16–23 In ordinary NSOM, however, one is forced to illu-minate the tip apex with an incident far-field laser to excitenear-field light at the tip apex, and as a result, one alwayssuffers from huge background scattering by the incident laser,which has been a key issue in NSOM for a long time.

In the case of plasmon nanofocusing, on the other hand,the tip is illuminated far from the apex to generate near-fieldlight through plasmon excitation that compresses and propa-

gates to the apex. This means that there is no need to illumi-nate the apex directly. Therefore, there is no concern aboutbackground scattering by far-field laser illumination, whichleads to much better sensitivity in NSOM. It should be notedthat the plasmon nanofocusing technique has completelydifferent characteristics to existing noise reduction methodssuch as modulation techniques24,25 or far-field subtraction26

in terms of the fact that plasmon nanofocusing ideally doesnot have any background scattering, whereas the othermethods reduce the existing background scattering. SinceRaschke et al. demonstrated plasmon nanofocusing on a met-allic tip in 2007,27 an increasing number of reports aboutNSOM using plasmon nanofocusing have been published, e.g.,in optical nano-imaging12,28,29 and nano-Raman analysis.13,30

In these studies, as one can expect, the metallic tip plays avery important role in plasmon nanofocusing. Thus far, thereis only one tip fabrication method that has been proposed andoften discussed, which is electrochemical etching.10,27,31 Thismethod produces a fully metallic tip by pulling a metallic wirein chemical solution.19 One of crucial issues in this method isthat the shape and material of tips are limited, which aremostly gold tips in a conical shape. This fact has restricted theversatility of plasmon nanofocusing-based NSOM because it isnot possible to adjust the plasmonic properties of tips forefficient plasmon nanofocusing, depending on experimentalrequirements such as excitation wavelength and sample.Therefore, a tip fabrication method that offers flexible controlof plasmonic properties is highly required.

Department of Applied Physics, Osaka University, Suita, Osaka 565-0871, Japan.

E-mail: [email protected]; Tel: +81 (0)6 68794710

5634 | Nanoscale, 2016, 8, 5634–5640 This journal is © The Royal Society of Chemistry 2016

www.rsc.org/nanoscalehttp://crossmark.crossref.org/dialog/?doi=10.1039/c5nr08548a&domain=pdf&date_stamp=2016-02-25

In this work, we propose an efficient design of plasmonictips for plasmon nanofocusing, which offers multiple para-meters for the flexible, active and precise control of tip plasmo-nic properties. We believe that our tip fabrication method willlead to practical plasmon nanofocusing-based NSOM adjusta-ble to arbitrary experimental conditions. We also report that,due to the optimization of the plasmonic properties plasmonnanofocusing is highly efficient, such that our methodachieved almost 100% reproducibility in plasmon nanofocus-ing on fabricated tips, which is an important development inplasmon nanofocusing-based NSOM to further make this tech-nique practical and reliable. Our tip design is composed of adielectric body structure in a pyramidal shape and a metallicthin layer on one surface of the body structure, as shown inFig. 1(a). It has a grating structure on the metallic thin layerthat works as a plasmon coupler. Plasmon nanofocusing isinduced on the tapered metallic layer by illuminating thegrating with a laser to generate near-field light at the apex.32,33

We have also developed a technique to easily deposit an extre-mely smooth metallic layer on one surface of the tip-body viathermal evaporation, which allows one to have various optionsfor the material of the metallic thin layer such as gold, silver,

copper and aluminum, each of which shows very differentplasmonic properties, covering a wide range of plasmonic fre-quencies ranging from the ultraviolet to the infraredregions.34–36 In addition to the metallic thin layer, we can alsoselect a dielectric material for the body of the tip from variousmaterials available, such as silicon, silicon nitride, diamond,boron-doped diamond, etc., which also control plasmonbehavior through their dielectric constants. Besides, oxidiza-tion of the body material also allows to control the dielectricconstant of the material, which provides vast variety ofmaterials that one can use as body material.37,38 Since theplasmon propagation mode on the tip strongly relies on boththe metallic and dielectric materials,39,40 by properly choosingmaterials from various options, one is able to induce highlyefficient plasmon nanofocusing. Also, because there arenumerous kinds of combinations of metals and dielectricmaterials, it will greatly broaden the frequency range ofplasmon nanofocusing with high efficiency. Thickness of themetal layer is also an important parameter to control theplasmon propagation mode because it affects the interactionof plasmons on both sides of the metallic thin layer. Thethermal evaporation technique allows precise control of thethickness, which means that further optimization of theplasmon propagation mode is possible after choosing twomaterials, one metal and one dielectric. In our experiment, bycarefully choosing materials and optimizing the thickness ofthe metal, we evaluated the best conditions for plasmon nano-focusing for a particular excitation wavelength used in ourexperiments so that we could achieve high reproducibility ofinducing plasmon nanofocusing. Finite-difference time-domain (FDTD) simulation confirms that our tip designefficiently induces plasmon nanofocusing and generatesstrong near-field light at the apex, as seen in Fig. 1(b), wheresilver (thickness = 40 nm) and silicon dioxide were used as themetallic thin layer and the tip body, respectively, and the exci-tation wavelength was 642 nm.

In addition, there are a few advantages associated with ourtip design. Due to the flatness of the metallic coating, one caneasily apply well-established nano-lithographic techniques,such as electron beam lithography or focused ion beam (FIB)lithography, without losing beam focus in contrast to the caseof three-dimensional structures like electrochemically etchedconical tips. It allows one to fabricate perfect structures as aplasmon coupler or any other plasmonic elements to modifythe plasmonic properties of tips. Also, the metallic thincoating, which is deposited on only one side of the tip body,gives better efficiency compared to a fully metallic tip, becausethe direction of plasmon propagation is strictly restricted tothe metal on one side and the near-field light propagates onlytowards the tip apex, resulting in efficient energy compressioninto the apex.

Fig. 1(c) shows a scanning electron microscopy (SEM)image of a fabricated tip, where a silicon dioxide cantilever tipwas utilized as the body material. The inset shows an enlargedimage of the tip, which is coated with a very smooth silverlayer and has a grating structure away from the apex. It is

Fig. 1 (a) Schematic of the tip structure for plasmon nanofocusing,which is composed of a dielectric pyramidal structure and a metallicthin layer on one surface of the pyramidal structure. By illuminating thegrating with a laser, plasmon nanofocusing is induced to generate near-field light at the apex. (b) An intensity map of electric field in the vicinityof the tip apex during plasmon nanofocusing, simulated via an FDTDmethod. Silver and silicon dioxide were used as the metallic thin layerand the dielectric body, respectively. The thickness of silver was 40 nm.The excitation wavelength was 642 nm. (c) SEM image of a fabricatedtip, which was made on a silicon dioxide cantilever. The inset shows anenlarged image of the tip, where a grating structure was fabricated onthe thin and smooth silver layer.

Nanoscale Paper

This journal is © The Royal Society of Chemistry 2016 Nanoscale, 2016, 8, 5634–5640 | 5635

noted that our fabricated tips are based on commercially avail-able silicon cantilever tips, which can be easily operated by anyatomic force microscopy (AFM) system that uses an opticallever feedback technique, which is one of the most commonNSOM configurations.18,41,42 Since electrochemically etchedtips can only be used in tuning fork configuration or scanningtunnelling microscopy, this fact provides good opportunitiesfor scientists working with optical lever feedback AFM. Wewould like to note that all of the advantages mentioned abovecannot be achieved by conventional tip fabrication methods,and we expect the unique properties of our tip fabricationmethod to benefit scientific studies as well as applications inresearch related to plasmon nanofocusing.

2. Results and discussion2.1 Fabrication of a cantilever-based metallic tip for plasmonnanofocusing

In this study, we optimized tip parameters for an incidentlaser wavelength of 642 nm, as this is one of the commonwavelengths used in spectroscopic research. By optimizing ourFDTD simulation results, we conclude that the best metal forthe plasmonic thin layer is silver. The optimized thickness ofthe silver layer was estimated to be 40 nm for the best results.For a comparison, we found that gold thin layer resulted inabout a 5000 times weaker near-field intensity at λ = 642 nmthan that of silver. This is mostly because gold has a higherattenuation for plasmon propagation at this wavelength.Further, we examined several materials for the tip body andfound silicon dioxide to be the best for this wavelength. As anexample, our simulation showed that the near-field intensitywas reduced by about 100 times, when we considered siliconas the tip body material, which the most common materialused in AFM cantilever tips. This leads us to an interestingconclusion that the choice of the material of the tip-body alsoplays an important role in efficient plasmon nanofocusing.Through this study, it is confirmed that the optimization ofmaterials is essential, because a small difference in the opticalproperties of the materials used in tip fabrication may even-tually create a huge difference in plasmon nanofocusing.Owing to our FDTD results, we decided to oxidize silicon tipsto convert them into silicon dioxide, and to evaporate 40 nmthick silver layers on them to obtain the best plasmonic tipsfor efficient plasmon nanofocusing. Fig. 2(a) shows the pro-cedure for tip fabrication. We first oxidized a silicon cantilever(NT-MDT/CSG-01) at 1000 °C in the presence of flowing watervapor in an electric furnace for 20 minutes, which turnedsilicon into silicon dioxide for the entire tip.37 We thendeposited a smooth silver coating via thermal evaporation. Theevaporation conditions are described below in detail. Lastly,we fabricated a grating structure using FIB lithography at a dis-tance of about 8 µm away from the tip apex that works as aplasmon coupler. The height of the cantilever tips we used wasaround 15 µm, which was long enough to fabricate the gratingfar from the apex to prevent overlap between the incident laser

Fig. 2 (a) Procedure of tip fabrication for plasmon nanofocusing. (b)Dependence of the surface roughness of silver coating on evaporationdirection. As the evaporation angle becomes larger, the roughness isdrastically reduced, depicting that high-angle evaporation is better tomake a smooth coating. (c) Dependence of the surface roughness ofsilver coating on evaporation rate. The surface roughness was reducedwith increasing the evaporation rate. (d) SEM images of 40 nm-thicksilver layer on silicon dioxide substrates evaporated at 0.04 nm s−1

(upper half ) and 3 nm s−1 (lower half ). Aggregates of small grains wereobserved at 0.04 nm s−1, whereas large grains were deposited at3 nm s−1. (e) Topographic line-profile of the surface of a large grain areaevaporated at 3 nm s−1 (red curve) and across an aggregated area ofsmall grains evaporated at 0.04 nm s−1 (black curve). The coating wasatomically smooth at the area of large grain. (f ) SEM image of an oxi-dized silicon tip. (g) SEM image of the oxidized silicon tip with a smooth40 nm thick silver coating, which was evaporated at ∼1.5 nm s−1. Theinset shows a magnified image of the silver coating, where large grainsare seen, similar to what is seen in (d). (h) SEM image of the silver-coated oxidized silicon tip with a grating structure, which was fabricatedvia FIB lithography.

Paper Nanoscale


and apex. The grating period was also pre-optimized thoughFDTD simulation and was found to be 780 nm for the bestcoupling efficiency at λ = 642 nm.

For plasmon nanofocusing, the surface roughness of themetal film is an important factor as a smoother surface candrastically reduce the energy loss during plasmon propagation.In order to deposit a smooth surface, there are a few keypoints to note during the metal evaporation. One of them isthe evaporation direction. For plasmonic tips used in ordinaryNSOM, the metal is usually evaporated along the tip axis,which is at a low angle with respect to the tip surface. Oncesilver nucleus islands are deposited with low-angle evapo-ration, they create a shadow behind them, and no silver wouldlater be deposited in the shadowed areas, resulting in a roughsurface with silver grains, which is actually preferable for loca-lized plasmon resonance.38,43–45 In contrast, for plasmonnanofocusing, the evaporation direction normal to the tipsurface (high angle evaporation) results in a much smoothersurface, because for a high-angle deposition, there is no suchshadow effect. Fig. 2(b) shows surface roughness dependenceon evaporation direction, calculated from root mean square(RMS) values. We have deposited 40 nm thick silver on silicondioxide substrates at various evaporation directions, andmeasured the roughness by tapping-mode AFM. As one cansee, the smoothest film was deposited in the case of the evapo-ration direction being perpendicular to the tip surface, whichwould result in a very low energy loss of plasmon propagationand thereby efficient plasmon nanofocusing.

Another key parameter that affects surface roughness of thedeposited metal film is the evaporation rate. The evaporationprocess usually deposits silver in a granular structure, wherethe size of the grains and the average surface roughness acrossthese grains depend upon the evaporation rate. Fig. 2(c) showsroughness dependence, which is calculated by the averageRMS value obtained from the topography, on evaporation rate.We found that the roughness reduces as the evaporation rateincreases. An average roughness of about 0.5 nm was achievedat an evaporation rate of 3 nm s−1. Further, we observed thatthe grain size also increased with increasing evaporation rate.Fig. 2(d) shows SEM images of silver thin coatings evaporatedat 0.04 nm s−1 (upper half ) and 3 nm s−1 (lower half ). For thefaster evaporation rate of 3 nm s−1, larger grains (severalhundred nm) with very flat surfaces were achieved, whereas fora slower deposition rate of 0.04 nm s−1, very small grains(10–20 nm) were obtained and no larger flat area could beobserved. We assume that this is because the high kineticenergy of silver atoms causes silver vapor to diffuse around thesurface and contributes to the growth process rather than thecreation of another grain.46 Fig. 2(e) shows the topographicline profiles of a large grain (red curve) and the aggregates ofsmall grains (black curve), for the deposition rates of 3 and0.04 nm s−1, respectively. The topographic line profiles showthat the surface roughness remained within 0.5 nm on thesurface of a large grain for the deposition rate of 3 nm s−1,whereas it was several nanometers for the aggregated area ofsmall grains for the deposition rate of 0.04 nm s−1. Aggregates

of small grains lead to a significant loss during plasmonpropagation and are therefore not suitable for plasmon nano-focusing. These results clearly show that faster deposition thatcreates larger grains with smoother surfaces are preferred, asthey are expected to contribute positively to efficient plasmonpropagation with a low energy loss. In practice, there is atrade-off between roughness and accuracy of thickness control,and we optimally chose around 1.5 nm s−1 as the evaporationrate in tip fabrication, in order to make the silver coatingsmooth enough together with an acceptable thickness control.With proper optimization, we obtained an averaged roughnessof less than 1 nm. We have also investigated the surface rough-ness dependence on ambient pressure during evaporation,and have found that there is no strong dependence on thepressure if it is lower than 1 × 10−4 Pa, because most dust par-ticles that may affect silver deposition are removed at such alow pressure.46 We therefore used a pressure of around 5 ×10−5 Pa in our experiments. This optimization of the surfaceroughness also contributes to the high reproducibility of ourmethod.

Fig. 2(f )–(h) show SEM images of a tip at different fabrica-tion steps. Fig. 2(f ) shows a silicon dioxide tip. We confirmedthat there are no changes in the tip shape during the oxidiza-tion process. Fig. 2(g) shows the oxidized tip with an extremelysmooth silver thin layer. When we magnified the image, weobserved large grains deposited on the tip surface as seen inthe inset. At last, we fabricated a grating structure on the silverlayer by FIB milling (Fig. 2(h)). Since the grating period is780 nm, it does not require very high fabrication resolution. Itis, therefore, easy and highly reproducible to fabricate thegrating with an ordinary FIB lithographic technique.

2.2 Demonstration of plasmon nanofocusing on fabricatedtips

We confirmed through optical measurements that the fabri-cated tips were able to produce plasmon nanofocusing. Fig. 3(a) shows an optical image of a fabricated tip with laser illumi-nation (λ = 642 nm) at the grating, where white dotted linesrepresent the outer shape of the tip and white arrows indicateincident polarization. We observed that a bright spot appearsat the apex. The inset shows a magnified image of the brightspot at the apex. The incident polarization was perpendicularto the grating grooves. In order to confirm if the bright spotappeared because of plasmon nanofocusing, we investigatedits dependence on incident polarization. At 45° of incidentpolarization angle with respect to the grating grooves, weobserved a reduced intensity at the tip apex (Fig. 3(b)), andlight spot disappeared at the apex when the incident polari-zation was parallel to the grating grooves (Fig. 3(c)). In athorough investigation of incident polarization, as shown inFig. 3(d), we obtained a dipole-like intensity variation, whichsuggests that the bright spot is caused by far-field emissionfrom near-field light generated through plasmon nanofocus-ing.12 Here, 0° corresponds to parallel polarization, and 90°corresponds to normal polarization. To further confirm this,we performed an FDTD simulation, which is shown by the

Nanoscale Paper


green line in Fig. 3(d). The simulation results are in goodagreement with our experimental results.

We would like to note again that our fabrication methodprovides efficient plasmon nanofocusing through optimizationof the tip material and geometry, especially of the thicknessand the surface roughness, so that we observe plasmon nano-focusing in almost 100% of the fabricated tips for a sample ofmore than 50 fabricated tips. In order to emphasize the repro-ducibility of our experiments, we show some optical images ofdifferent fabricated tips in Fig. 4, where the laser is illumi-nated at the gratings. All tips show bright spots at their apexes,which are distinctly spatially separated from the laserillumination.

2.3 Near-field imaging with plasmon nanofocusing

In order to emphasize the plasmonic properties of our fabri-cated tips, we demonstrated near-field imaging with plasmonnanofocusing. In these experiments, Rayleigh scattering fromthe sample was detected to evaluate tip performance. Fig. 5(a)illustrates the experimental setup for near-field measurement.An incident laser (CW, λ = 642 nm) was focused on the gratingof our fabricated tip through a single-mode optical fiber andan objective lens (NA = 0.28, WD = 30 mm, ×20). Incidentpolarization was controlled by a polarizer right before theobjective. Illumination position was controlled by an actuator

in the x–y–z directions. Near-field light was then createdthrough plasmon nanofocusing (see the illustration in thelower left inset), and the Rayleigh scattered signal from thesample was collected through another objective (NA = 1.4, ×60)

Fig. 3 (a–c) Optical images of a fabricated tip with laser illumination atthe grating, where incident polarizations are indicated by the whitearrows. The white dotted lines represent the outer shapes of the tips.The insets show magnified images of the apex. For the case of polari-zation being perpendicular to the grating grooves in (a), a bright spotappears at the apex. The intensity is reduced in (b) where incident polari-zation is tilted to 45 degrees. The light spot completely disappears in (c)when the polarization is parallel to the grating. (d) Polar graph of lightspot intensity with respect to incident polarization. 0° corresponds toparallel polarization, and 90° corresponds to normal polarization. Thedipole-like intensity variation confirms that the bright spot is induced byplasmon nanofocusing. The experimental results given by the red curveare in good agreement with the FDTD simulation data given by thegreen curve.

Fig. 4 Some examples of optical images of different tips showingbright spots at their apexes. Almost 100% of fabricated tips showedbright spots.

Fig. 5 (a) A schematic of the experimental setup for near-field opticalimaging with plasmon nanofocusing. A pinhole and a mask were set tocut scattered light at the grating. (b) AFM image of a CNT sample. (c)Plasmon nanofocusing-based near-field image of a CNT sample. TheCNTs were imaged with a spatial resolution of ∼30 nm. (d) Near-fieldimage of the CNTs at the same area with incident polarization parallel tothe grating grooves. No sample was observed, which proved that near-field imaging was achieved through the plasmon nanofocusing in (c).

Paper Nanoscale


at the bottom. We utilized a pinhole (30 µm) and a spatialmask to block scattered light from the grating. Working as aconfocal element, the pinhole efficiently extracts a signal fromthe vicinity of the tip apex. The spatial mask allows only highNA components to pass, whereas the scattering at the grating,which exists at low NA, is blocked by the mask (see the inset).This configuration perfectly blocks Rayleigh scattering fromthe grating, and near-field scattering from the tip apex ismeasured. Finally, the signal was detected by a liquid-nitro-gen-cooled CCD camera.

Here, we chose carbon nanotubes (CNTs) as a standardsample. CNT powder (NanoIntegris, IsoNanotubes-M) wassonicated for about 30 minutes in a solution of 1,2-dichloro-ethane, and then the solution was dropped and spin-coated ona cleaned cover slip.47 The CNT sample was synchronouslyscanned using a piezo stage under AFM control together withnear-field signal detection. Fig. 5(b) shows an AFM image ofthe CNT sample. Fig. 5(c) shows a Rayleigh-scattered near-fieldoptical image of the same area, measured with plasmon nano-focusing. We observed strong scattering from the sample,where a dramatic contrast can be seen at different locations ofCNTs. Please note that no modulation technique or far-fieldsubtraction was applied for background suppression. Weclearly observed a background-free near-field image of theCNT sample owing to the plasmon nanofocusing. We exam-ined the intensity line-profile along the white dotted lines inFig. 5(c), and obtained a spatial resolution of 30 nm, farbeyond the diffraction limit. At the same time, we did not seeany image contrast in the case where the polarization was par-allel to the grating (Fig. 5(d)), which again proves successfulplasmon nanofocusing. We confirmed that our fabricated tipswere evidently reliable for near-field microscopy.

3. Conclusions

In conclusion, we propose a tip design that enables one tocontrol the plasmonic properties of tips for efficient plasmonnanofocusing, where the tip is composed of a metallic thinlayer and a dielectric pyramidal body structure. Since our fabri-cation method provides control of multiple parameters thatcan affect plasmonic properties, it allows us to flexibly controlthe plasmonic properties of tips to meet experimental require-ments, such as the incident laser, the sample and so on,which has not been possible with conventional tip fabricationmethods. In our experiments, we optimized the plasmonic pro-perties for a 642 nm excitation wavelength, where silver andsilicon dioxide were used, and the thickness and the surfaceroughness of silver coating were optimized to 40 nm andbelow 1 nm, respectively, resulting in almost 100% reproduci-bility of plasmon nanofocusing, as demonstrated throughoptical measurements. The high reproducibility is a muchappreciated development to make plasmon nanofocusing-based NSOM measurements reliable. We confirmed throughpolarization investigation that the light spot observed at thetip apex was actually produced through plasmon nanofocus-

ing. At this moment, although we focused on a certain inci-dent wavelength of 642 nm, the frequency range can be muchbroadened from ultraviolet to infrared through further optimi-zation. In addition, we would like to note that our fabricatedtip is based on cantilevers, which allows it to be easily appliedto conventional AFM systems that use optical lever feedback,and provides good opportunities for scientists working withAFM. We have also shown that our fabricated tips are reliablefor NSOM measurements, by conducting near-field imaging ofCNTs with plasmon nanofocusing at nanoscale spatial resolu-tion. Importantly, we clearly observed a background-free imageof CNTs without a modulation technique or other noise-reduction techniques. We expect that our reproducible andreliable cantilever-based tips will lead to further developmentin research related to this field.

Acknowledgements

The authors gratefully acknowledge the financial support fromthe Grant-in-Aid for Scientific Research (C) 24560028 and (B)25286008, and from the JSPS Asian Core Program. TU wouldlike to acknowledge the scholarship support from JSPS.

References

1 Kh. V. Nerkararyan, Phys. Lett. A, 1997, 237, 103.2 A. J. Babadjanyan, N. L. Margaryan and Kh. V. Nerkararyan,

J. Appl. Phys., 2000, 87, 3785.3 M. I. Stockman, Phys. Rev. Lett., 2004, 93, 137404.4 K. Tanaka, G. W. Burr, T. Grosjean, T. Maletzky and

U. C. Fischer, Appl. Phys. B, 2008, 93, 257.5 E. Verhagen, A. Polman and L. Kuipers, Opt. Express, 2008,

16, 45.6 M. W. Vogel and D. K. Gramotnev, J. Appl. Phys., 2010, 107,

044303.7 M. Schnell, P. Alonso-Gonzalez, L. Arzubiaga, F. Casanova,

L. E. Hueso, A. Chuvilin and R. Hillenbrand, Nat. Photonics,2011, 5, 283.

8 F. De Angelis, F. Gentile, F. Mecarini, G. Das, M. Moretti,P. Candeloro, M. L. Coluccio, G. Cojoc, A. Accardo,C. Liberale, R. P. Zaccaria, G. Perozziello, L. Tirinato,A. Toma, G. Cuda, R. Cingolani and E. D. Fabrizio, Nat.Photonics, 2011, 5, 682.

9 H. Shen, G. Lu, Y. He, Y. Cheng, H. Liu and Q. Gong, Nano-scale, 2014, 6, 7512.

10 J. Mihaljevic, C. Hafner and A. J. Meixner, Opt. Express,2015, 23, 18401.

11 H. Choo, M. Kim, M. Staffaroni, T. J. Seok, J. Bokor,S. Cabrini, P. J. Schuck, M. C. Wu and E. Yablonovitch, Nat.Photonics, 2012, 6, 838.

12 C. C. Neacsu, S. Berweger, R. L. Olmon, L. V. Saraf,C. Ropers and M. B. Raschke, Nano Lett., 2010, 10, 592.

13 F. D. Angelis, G. Das, P. Candeloro, M. Patrini, M. Galli,A. Bek, M. Lazzarino, I. Maksymov, C. Liberale,

Nanoscale Paper


L. C. Andreani and E. D. Fabrizio, Nat. Nanotechnol., 2010,5, 67.

14 A. Giugni, B. Torre, A. Toma, M. Francardi, M. Malerba,A. Alabastri, R. P. Zaccaria, M. I. Stockman andE. D. Fabrizio, Nat. Nanotechnol., 2013, 8, 845.

15 B. N. Tugchin, N. Janunts, A. E. Klein, M. Steinert,S. Fasold, S. Diziain, M. Sison, E. Kley, A. Tünnermann andT. Pertsch, ACS Photonics, 2015, 2, 1468.

16 Y. Inouye and S. Kawata, Opt. Lett., 1994, 19, 159.17 S. Kawata, Y. Inouye and P. Verma, Nat. Photonics, 2009, 3,

388.18 N. Hayazawa, T. Yano, H. Watanabe, Y. Inouye and

S. Kawata, Chem. Phys. Lett., 2003, 376, 174.19 A. Hartschuh, E. J. Sanchez, X. S. Xie and L. Novotny, Phys.

Rev. Lett., 2003, 90, 095503.20 T. Yano, T. Ichimura, S. Kuwahara, F. H’Dhili, K. Uetsuki,

Y. Okuno, P. Verma and S. Kawata, Nat. Commun., 2013,3592, 1.

21 Y. Okuno, Y. Saito, S. Kawata and P. Verma, Phys. Rev. Lett.,2013, 111, 216101.

22 J. Chen, M. Badioli, P. Alonso-Gonzalez, S. Thongrattanasiri,F. Huth, J. Osmond, M. Spasenovic, A. Centeno, A. Pesquera,P. Godignon, A. Z. Elorza, N. Camara, F. J. G. Abajo,R. Hillenbrand and F. H. L. Koppens, Nature, 2012, 487, 77.

23 N. Hayazawa, K. Furusawa, A. Taguchi, S. Kawata andH. Abe, Appl. Phys. Lett., 2009, 94, 193112.

24 R. Hillenbrand and F. Keilmann, Phys. Rev. Lett., 2000, 85,3029.

25 F. Zenhausern, M. P. O’Boyle and H. K. Wickramasinghea,Appl. Phys. Lett., 1994, 65, 1623.

26 J. Yu, Y. Saito, T. Ichimura, S. Kawata and P. Verma, Appl.Phys. Lett., 2013, 102, 123110.

27 C. Ropers, C. C. Neacsu, T. Elsaesser, M. Albrecht,M. B. Raschke and C. Lienau, Nano Lett., 2007, 7, 2785.

28 D. Sadiq, J. Shirdel, J. S. Lee, E. Selishcheva, N. Park andC. Lienau, Nano Lett., 2011, 11, 1609.

29 Z. Zhang, P. Ahn, B. Dong, O. Balogun and C. Sun, Sci.Rep., 2013, 2803, 1.

30 S. Berweger, J. M. Atkin, R. L. Olmon and M. B. Raschke,J. Phys. Chem. Lett., 2010, 1, 3427.

31 S. Schmidt, B. Piglosiewicz, D. Sadiq, J. Shirdel, J. S. Lee,P. Vasa, N. Park, D. Kim and C. Lienau, ACS Nano, 2012, 6,6040.

32 J. Takahara and M. Miyata, Opt. Express, 2013, 21, 27402.33 E. Verhagen, A. Polman and L. Kuipers, Opt. Express, 2008,

16, 45.34 A. Taguchi, Y. Saito, K. Watanabe, S. Yijian and S. Kawata,

Appl. Phys. Lett., 2012, 101, 081110.35 A. Taguchi, N. Hayazawa, K. Furusawa, H. Ishitobi and

S. Kawata, J. Raman Spectrosc., 2009, 40, 1324.36 G. H. Chan, J. Zhao, E. M. Hicks, G. C. Schatz and

R. P. V. Duyne, Nano Lett., 2007, 7, 1947.37 A. Ono, K. Masui, Y. Saito, T. Sakata, A. Taguchi,

M. Motohashi, T. Ichimura, H. Ishitobi, A. Tarun,N. Hayazawa, P. Verma, Y. Inouye and S. Kawata, Chem.Lett., 2008, 37, 122.

38 N. Hayazawa, T. Yano and S. Kawata, J. Raman Spectrosc.,2012, 43, 1177.

39 D. K. Gramotnev and S. I. Bozhevolnyi, Nat. Photonics,2010, 4, 83.

40 M. Miyata and J. Takahara, J. Appl. Phys., 2012, 111,053102.

41 T. Ichimura, N. Hayazawa, M. Hashimoto, Y. Inouye andS. Kawata, Phys. Rev. Lett., 2004, 92, 220801.

42 E. Bailo and V. Deckert, Angew. Chem., Int. Ed., 2008, 47,1658.

43 A. Taguchi, J. Yu, P. Verma and S. Kawata, Nanoscale, 2015,7, 17424.

44 T. Umakoshi, T. Yano, Y. Saito and P. Verma, Appl. Phys.Express, 2012, 5, 052001.

45 T. Mino, Y. Saito and P. Verma, ACS Nano, 2014, 8, 10187.46 K. M. McPeak, S. V. Jayanti, S. J. P. Kress, S. Meyer,

S. Iotti, A. Rossinelli and D. J. Norris, ACS Photonics,2015, 2, 326.

47 T. Yano, P. Verma, S. Kawata and Y. Inouye, Appl. Phys.Lett., 2006, 88, 093125.

Paper Nanoscale


Button 1:

c5nr08548a 5634..5640 -...

Documents