Sub-100 nm Focusing of Short Wavelength Plasmons in

Sep 2, 2014 - Near-field scanning optical microscopy (NSOM),(3-5) in which a sharp ... or breaking down; structured illumination microscopy (SIM)(9) r...
0 downloads 0 Views 3MB Size
Download PDF
Letter pubs.acs.org/NanoLett

Sub-100 nm Focusing of Short Wavelength Plasmons in Homogeneous 2D Space B. Gjonaj,# A. David,# Y. Blau, G. Spektor, M. Orenstein, S. Dolev, and G. Bartal* Department of Electrical Engineering, Technion - Israel Institute of Technology, Technion City, 32000, Haifa, Israel S Supporting Information *

ABSTRACT: We present a direct measurement of shortwavelength plasmons focused into a sub-100 nm spot in homogeneous (translation invariant) 2D space. The shortwavelength (SW) surface plasmon polaritons (SPP) are achieved in metal−insulator−insulator (MII) platform consisting of silver, silicon nitride, and air. This platform is homogeneous in two spatial directions and supports SPP at wavelength more than two times shorter than that in free space yet interacts with the outer world through the evanescent tail in air. We use an apertureless (scattering) near-field scanning optical microscope (NSOM) to map directly the amplitude and phase of these SW-SPP and show they can be focused to under 70 nm without structurally assisted confinement such as nanoantennas or nanofocusing. This, along with the use of visible light at 532 nm which is suitable for optical microscopy, can open new directions in direct biological and medical imaging at the sub-100 nm resolution regime. KEYWORDS: Super-resolution, nanofocusing, near-field, plasmonics, microscopy

R

dyes molecular photochemistry and also presents challenges. For example, PALM7 and STORM8 microscopes require several minutes of scan times and complex optics and scanning that is sensitive to misalignment or breaking down; structured illumination microscopy (SIM)9 requires extremely sensitive alignment and powerful image processing softwares. Stimulation emission depletion (STED) microscopy10 requires two well-aligned laser pulses at extremely high power (Watts) which can lead to fast bleaching of all but the most robust dyes. In general, most super-resolution techniques are characterized with a high degree of complexity. Alternative and less complex super-resolution approaches have been recently proposed, including plasmonics11 which aims at tight focusing of electromagnetic light waves and offers an alternative route to control light fields at subwavelength scales with metallic nanostructures.12−14 Surface plasmon polaritons (SPPs), evanescent surface waves propagating along metal−dielectric interfaces, are ideal candidates for resolution improvements as they are able to exhibit sub-100 nm wavelength for optical frequencies,15−17 hence can be utilized for microscopy at this scale.18−20 Plasmonic superresolution has been widely investigated in the past decade introducing the plasmonic super lens21 and hyperlens,22,23 while plasmonic focusing on a metal−air interface was limited to about half the wavelength in vacuum.24−27

ecent technological demands for subwavelength resolved information have motivated the development of modern super-resolution microscopy techniques, aiming to improve the resolution limits of a conventional optical microscope which is limited by the wavelength of the illuminating light source. This limit translates in several hundred nanometers for optical wavelengths as was first described by Abbe over 150 years ago.1 This restriction has a major impact on biological and biomedical research which mostly utilizes confocal microscopy and total internal reflection fluorescence microscopy (TIRFM) which are limited to nearly 200 nm resolution for visible light. Objects and features smaller than this length scale appear blurry, limiting the amount of useful information that can be gleaned. According to a recent review article,2 this resolution “is inadequate for dissecting the inner architecture of many subcellular structures” and thus limits the ability of researchers, diagnosticians, and drug companies to assess the information, be it for fundamental research, diagnosing disease, or drug testing. Modern super-resolution microscopy techniques, including both near-field and far-field methods, can partially overcome these limits and supported a large scientific advance. However, these techniques have their own limitations and constraints with the most powerful ones being expensive, slow, and/or temperamental. Near-field scanning optical microscopy (NSOM),3−5 in which a sharp metallic or dielectric tip is scanned over a surface, is not amenable to either fast scans or for imaging soft surfaces or interfaces such as cells.6 The state of art for far-field techniques is based on dye labeling of the samples to achieve controllable fluorescent contrast from the © 2014 American Chemical Society

Received: June 4, 2014 Revised: August 25, 2014 Published: September 2, 2014 5598

dx.doi.org/10.1021/nl502080n | Nano Lett. 2014, 14, 5598−5602

Nano Letters

Letter

The main limitation of plasmonics are the ohmic losses which are inevitable in metals and impose a critical relation between optical confinement and wave propagation; not only do they reduce the propagation length of those surface waves, they also inhibit true short wavelength plasmons owing to a large imaginary component in the dielectric permittivity, smearing the surface plasmon resonance which is associated with the ultrashort wavelength. Therefore, sub-100 nm focusing of plasmons was only directly observed using structural confinement such as nanofocusing28 or nanoantenna.29 As far as applicative bio applications go, shining visible (i.e., blue) light onto watery solutions of biosamples at the interface of silver (Ag being the least lossy metal available), the SPP wavelength of Ag−H2O in the visible band is longer than 300 nm which is not sufficient for super-resolution. Furthermore, the need for near-field interaction between the biological sample and the metal hinders the use of high refractive index materials (e.g., TiO2) which can, in principle, improve the system’s resolution. In view of these constraints, plasmonics, despite its huge potential, has so far found almost no robust application in bioimaging. Here, we present direct observation of sub-100 nm focusing of short-wavelength plasmons at visible light frequency in a homogeneous (translation invariant) 2D platform. We use amplitude and phase resolved measurements to map the wavelength of the plasmonic mode and find it to be 240 nm for 532 nm illumination wavelength. Utilizing a planar metal− dielectric−dielectric structure consisting of silver−silicon nitride−air, we achieve both tight confinement from the high refractive index nitride and the ability to perform direct measurements in air with an apertureless scattering near field microscope. Curved slits, carved through the layers, provide both light-to-SPP coupling and focusing of these shortwavelength plasmons to a focal spot of 70 nm fwhm. This 70 nm resolution, along with its translation invariance, exemplifies a new microscopy platform that can be combined with nonmechanical scanning techniques (for example, wavefront shaping30) for super-resolution in vivo bio imaging. Our plasmonic platform consists of tandem metal− insulator−insulator (MII) layered structure made of a 200 nm thick layer silver protected by a 50 nm thick silicon nitride (Si3N4) membrane, as depicted in Figure 1. This platform constitutes a planar system, homogeneous in two transverse dimensions where the outermost insulator is the air above the nitride. Such sample supports a single propagating TM SPP mode with each dielectric layer having its role: while the high refractive index layer (SiN: n = 2) interfacing the metal is responsible for short plasmonic wavelengths (i.e., “confining” layer), the evanescent tale in the air provides the interaction with the environment (i.e., “interacting” layer). This interacting layer can be replaced by different media, for example by aqueous solution containing bio samples for future biological microscopy applications. A diagram of our platform and of the measurement scheme is shown in Figure 1. The sample is illuminated from underneath with green laser light of wavelength λ0 = 532 nm. Curved subwavelength slits were milled all the way through the silver layer using a focused ion beam to provide the coupling of light into the propagating SPP waves, as well as to focus such waves into a tight spot. These plasmonic waves, propagating mainly on the Ag−SiN interface, have their evanescent tail extended well beyond the SiN layer. The calculated field distribution, shown as a vertical profile through the layers in Figure 1,

Figure 1. Diagram of the measurement scheme overlaid on the calculated field distribution. The plasmonic platform consists of a planar Ag (200 nm)−SiN (50 nm)−air system and supports the propagation of a single mode of short-wavelength plasmon. Coupling of light to this SPP mode and plasmonic focusing is achieved by launching a weakly focused 532 nm wavelength laser beam through curved slits of nearly 1 μm radius and 75 nm width. The x−y map of the focused plasmonic electric field [Ez(x,y)] is added to the figure as a surface graph, while its vertical profile is shown in yellow. The evanescent tail in air, seen in the vertical profile, is scattered by a nearfield probe while scanning the sample to produce a 2D map of the electric field. A pseudoheterodyne unit resolves the information into amplitude and phase maps.

exhibits a strong evanescent tail in the air above the SiN layer that can be used to interact with external scatterers (e.g., NSOM tip) for near-field measurement. We spatially map this near-field of the plasmonic mode using a transmission mode apertureless NSOM: an oscillating metallic (Pt−Ir) tip scatters the normal electric field component Ez into a pseudo heterodyne detection unit used to provide amplitude and phase resolved information on the measured field.31 Higher harmonics of the tip oscillations are used to filter out the possible far-field background (we present third harmonic measurements). The map is created by raster scanning the sample, and the resolution of the acquired near field image is between 10 and 20 nm depending on the tip’s apex. These near-field maps contain the information regarding the plasmon wavelength and focus size. The phase-resolved measurements, revealing the plasmonic wavelength, are shown in Figure 2. The excitation of SPPs and their focus is provided by a single half circle slit of 1 μm radius and 75 nm width (Figure 2a). A 2 × 2 μm region of the sample has been selected and raster-scanned with the near-field probe. The measured amplitude of the propagating plasmonic wave is shown in Figure 2b, confirming the creation of a 1D plamonic focus confined to 119 nm fwhm which is nearly half the SPP wavelength. This is significantly smaller than the 300 nm fwhm focusing previously shown for Au−air interface plasmons. The fringes of the amplitude are due to interference of the signal with the left-over background that was not perfectly filtered by the near-field measurement system (see Supporting Information, SI, for more details). The plasmonic propagation length is calculated to be LS = 0.9 μm. In collecting the phase of plasmonic electric field oscillation, shown in Figure 2d, we determine the plasmonic wavelength as the period of the oscillation. Using precise Fourier analysis of the electric field, the plasmonic wavelength was found to be λS = 240 ± 20 nm. The Fourier transform of the electric field also provides additional information regarding the numerical aperture (NA) of the plasmonic lens and the quality of the near-field 5599

dx.doi.org/10.1021/nl502080n | Nano Lett. 2014, 14, 5598−5602

Nano Letters

Letter

Figure 2. Phase-resolved near-field measurements of short wavelength plasmons. (a) SEM image of the plasmonic lens, a single half circle with 1 μm radius. The dashed square of 2 μm side is the scan area. (b−c) Experimental and calculated near-field maps of the amplitude of Ez, respectively. (d− e) Near-field maps of the measured and calculated phase, respectively. The wavelength of the SPP oscillations is found to be λS = 240 ± 20 nm. (f− g) Fourier transforms of the electric fields for measurements and calculations, respectively. The pronounced outer ring corresponds to the plasmonic wavelength and provides a measure of the numerical (angular) aperture of the lens. Residual far field background of the measurement resides in the weak inner circle close to the origin, confirming that the background has been properly filtered. The theoretical results also confirm that at such thickness of the SiN layer no photonic guided mode is supported in the dielectric slab. (h) Averaged line cut of the focus (intensity) and fit.

microscopy where a two-dimensional image has to be created. Super-resolution applications require a plasmonic focus confined in both planar x and y dimensions. It is worthy to compare a plasmonic lens with a conventional one; a conventional lens is intrinsically 3D and requires a 2D surface manipulation to provide a confining wavefront for a twodimensional focus. The depth of focus is the extension of this focus in the third dimension. A plasmonic lens is intrinsically two-dimensional and requires a line shaped wavefront to confine electromagnetic energy in one of the planar dimensions, with the depth of focus being the extension along the other planar dimension. This property of a plasmonic lens is due to their evanescent decay in the out-of-plane dimension. Thus, to provide 2D confinement with a plasmonic lens, it is critical to shape also the depth of focus. Conventional microscopy already provides a solution for controlling the

measurements. The radius of the outer ring in the Fourier image of Figure 2f corresponds to the plasmonic wave vector kS = 2π/λS from which the SPP wavelength is determined, while the angular aperture of the ring provides the NA of the curved slit and its corresponding focusing abilities. The additional inner circle present in the Fourier transforms represents the residual far field background which oscillates at the laser wavelength, by which we deduce our signal-to-background contrast, which is a critical merit of near-field measurements (see SI). We find that the near-field versus background contrast can get to nearly 2 orders of magnitude. The theoretical predictions of amplitude and phase depicted respectively in Figures 2c and e, confirm the good agreement between experiments and theory. Plasmonic focusing in one dimension only, while highly relevant for its fundamental properties,32 cannot find use in 5600

dx.doi.org/10.1021/nl502080n | Nano Lett. 2014, 14, 5598−5602

Nano Letters

Letter

Figure 3. Plasmonic focusing. (a) SEM image of the high NA plasmonic lens: two concentric arcs with radii of 1 and 1.124 μm provide a halfwavelength mismatch so as to achieve improved focusing of Ez. (b, d) Measured near-field maps of the intensity and phase of the normal E-field (Ez), respectively. A sharp focus is observed in the center, in close agreement with the respective calculations of (c) and (e). Close ups of the focus are shown in (f) and (g). Line cuts of the focus (h) and their relative Gaussian fits determine the focus size to be 66 ± 20 nm and 104 ± 20 nm, within the confidence interval of our predictions λS/4 = 60 nm and λS/2 = 120 nm, respectively.

Figure 3b, exhibits a sharp confinement of optical power due to short SPP wavelength and optimized geometry. A 2D focus is created in the center as a result of two counter propagating SPP waves, similarly to the theoretical calculations of Figure 3c. The focus is located in the center of the measured phase map of Figure 3d. Close-ups of this focus (Figure 3f) confirm the subdiffractive confinement: the scanned area is comparable with the laser wavelength of 532 nm. Line cuts of the focus and their relative Gaussian fits are shown in Figure 3h. The focus is asymmetric with fwhm along the vertical and horizontal axis of 66 ± 20 nm and 104 ± 20 nm, respectively. These measurements are in close agreement with the theoretical prediction for the focus. Parametrically, theory predicts a plasmonic diffraction limit of λS/2 = 120 nm with a focal depth

depth of focus in the form of 4Pi microscopy where two objective lenses positioned one against the other create sharp counter-propagating fringes of intensity. Adopting this concept in plasmonic microscopy offers an additional advantage: the simplicity of carving two line shaped plasmonic lenses one against the other. Our main results, showing 2D focusing of short wavelength plasmons obtained in an optimized way from two plasmonic focusing slits facing each other,33,34 are shown in Figure 3. The radius mismatch between the two focusing slits, half the SPP wavelength, guarantees focusing of the normal component of the electric field, Ez, whose intensity is 2 orders of magnitude higher than the other field components. The near-field intensity (square of the measured normal component Ez), shown in 5601

dx.doi.org/10.1021/nl502080n | Nano Lett. 2014, 14, 5598−5602

Nano Letters

Letter

of λS/4 = 60 nm. If a symmetric focus was to be obtained, for example using a spiral slit and circular polarization, the fwhm of such focus would be expected to be around 85 nm. In conclusion, we have presented the observation of a sub100 nm plasmonic focus obtained from phase-resolved nearfield measurements. Such focus was achieved on metal− insulator−insulator plasmonic platform that provided the means to confine light significantly below the optical diffraction limit, yet leaving open channels for direct interaction. From an applicative point of view, these experiments are fully compatible with the state of art in fluorescence microscopy, where visible light in watery solutions of samples have to be used. Therefore, establishing a plasmonic platform that can provide optical confinement below 70 nm with the ability to maintain a perturbational interaction with the surrounding environment is of great importance. While in this work the interaction is in the form of a near-field tip, it can be extended to, e.g., a fluorescent molecule labeling a biological sample to provide critical mass for future applications of in vivo bio imaging using short wavelength plasmons.

■

(11) Barnes, W. L.; Dereux, A.; Ebbesen, T. W. Nature 2003, 424, 824−830. (12) Aeschlimann, M.; Bauer, M.; Bayer, D.; Brixner, T.; de Abajo, F. J. G.; Pfeiffer, W.; Rohmer, M.; Spindler, C.; Steeb, F. Nature 2007, 446, 301−304. (13) Li, X.; Stockman, M. I. Phys. Rev. B 2008, 77, 195109. (14) Volpe, G.; Cherukulappurath, S.; Parramon, R. J.; MolinaTerriza, G.; Quidant, R. Nano Lett. 2009, 9, 3608−3611. (15) Miyazaki, H. T.; Kurokawa, Y. Phys. Rev. Lett. 2006, 96, 097401. (16) Maier, S. A. Plasmonics fundamentals and applications; Springer: New York, 2007. (17) Verhagen, E.; Dionne, J. A.; Kuipers, L. K.; Atwater, H. A.; Polman, A. Nano Lett. 2008, 8, 2925−2929. (18) Smolyaninov, I. I.; Elliott, J.; Zayats, A. V.; Davis, C. C. Phys. Rev. Lett. 2005, 94, 057401. (19) Durach, M.; Rusina, A.; Stockman, M. I.; Nelson, K. Nano Lett. 2007, 7, 3145−3149. (20) Kao, T. S.; Jenkins, S. D.; Ruostekoski, J.; Zheludev, N. I. Phys. Rev. Lett. 2011, 106, 085501. (21) Fang, N.; Lee, H.; Sun, C.; Zhang, X. Science 2005, 308, 534− 537. (22) Liu, Z.; Lee, H.; Xiong, Y.; Sun, C.; Zhang, X. Science 2007, 315, 1686−1686. (23) Smolyaninov, I. I.; Hung, Y.-J.; Davis, C. C. Science 2007, 315, 1699−1701. (24) Lerman, G. M.; Yanai, A.; Levy, U. Nano Lett. 2009, 9, 2139− 2143. (25) Liu, Z.; Steele, J. M.; Srituravanich, W.; Pikus, Y.; Sun, C.; Zhang, X. Nano Lett. 2005, 5, 1726−1729. (26) Yin, L.; Vlasko-Vlasov, V. K.; Pearson, J.; Hiller, J. M.; Hua, J.; Welp, U.; Brown, D. E.; Kimball, C. W. Nano Lett. 2005, 5, 1399− 1402. (27) Chen, W.; Abeysinghe, D. C.; Nelson, R. L.; Zhan, Q. Nano Lett. 2009, 9, 4320−4325. (28) Choo, H.; Kim, M.-K.; Staffaroni, M.; Seok, T. J.; Bokor, J.; Cabrini, S.; Schuck, P. J.; Wu, M. C.; Yablonovitch, E. Nat. Photon 2012, 6, 838−844. (29) Alonso-González, P.; Albella, P.; Schnell, M.; Chen, J.; Huth, F.; Garca-Etxarri, A.; Casanova, F.; Golmar, F.; Arzubiaga, L.; Hueso, L. E.; Aizpurua, J.; Hillenbrand, R. Nat. Commun. 2012, 3, 684. (30) Gjonaj, B.; Aulbach, J.; Johnson, P. M.; Mosk, A. P.; Kuipers, L.; Lagendijk, A. Phys. Rev. Lett. 2013, 110, 266804. (31) Ocelic, N.; Huber, A.; Hillenbrand, R. Appl. Phys. Lett. 2006, 89, 101124. (32) Alonso-González, P.; Nikitin, A. Y.; Golmar, F.; Centeno, A.; Pesquera, A.; Vélez, S.; Chen, J.; Navickaite, G.; Koppens, F.; Zurutuza, A.; Casanova, F.; Hueso, L. E.; Hillenbrand, R. Science 2014, 344, 1369−1373. (33) Fang, Z.; Peng, Q.; Song, W.; Hao, F.; Wang, J.; Nordlander, P.; Zhu, X. Nano Lett. 2010, 11, 893−897. (34) Spektor, G.; David, A.; Gjonaj, B.; Gal, L.; Orenstein, M. arXiv [physics] 2014, 1406.0115.

ASSOCIATED CONTENT

S Supporting Information *

Setup details, role of the tip’s amplitude oscillation, and chronological procedure for reproducing data. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Author Contributions #

B.G. and A.D. contributed equally to the presented work.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was supported by The Ministry of Science and Technology of Israel and by the Israeli Nanotechnology Focal Technology Area on “Nanophotonics for Detection”. B.G. acknowledges that: The research leading to these results has received funding from the European Union’s Seventh Framework Programme (FP7/2007−2013) under grant agreement n° 626812, MC-MultiSPLASH.

■

REFERENCES

(1) Abbe, E. J. R. Microsc. Soc. 1882, 2, 300−309. (2) Huang, B.; Babcock, H.; Zhuang, X. Cell 2010, 143, 1047−1058. (3) Novotny, L.; Hecht, B. Principles of Nano-Optics; Cambridge University Press: New York, 2006; p 558. (4) Balistreri, M. L. M.; Gersen, H.; Korterik, J. P.; Kuipers, L.; van Hulst, N. F. Science 2001, 294, 1080−1082. (5) Hillenbrand, R.; Taubner, T.; Keilmann, F. Nature 2002, 418, 159−162. (6) van Zanten, T. S.; Gomez, J.; Manzo, C.; Cambi, A.; Buceta, J.; Reigada, R.; Garcia-Parajo, M. F. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 15437−15442. (7) Betzig, E.; Patterson, G. H.; Sougrat, R.; Lindwasser, O. W.; Olenych, S.; Bonifacino, J. S.; Davidson, M. W.; Lippincott-Schwartz, J.; Hess, H. F. Science 2006, 313, 1642−1645. (8) Rust, M. J.; Bates, M.; Zhuang, X. Nat. Methods 2006, 3, 793− 796. (9) Gustafsson, M. G. L. Proc. Natl. Acad. U.S.A. 2005, 102, 13081− 13086. (10) Dyba, M.; Hell, S. W. Phys. Rev. Lett. 2002, 88, 163901. 5602

dx.doi.org/10.1021/nl502080n | Nano Lett. 2014, 14, 5598−5602