We demonstrate the use of plasmonic Bragg reflectors (PBRs) to enhance the extraordinary optical transmission (EOT) from an array of sub-wavelength apertures in a gold film. Arrays of partially milled lines and dimples are placed at the edges of an array of nano-holes in a gold film. These PBR structures, with half the pitch of the array, capture light scattered away from the array by Bragg reflection. By appropriate positioning of the PBR, the light is reflected in-phase with the EOT light and thereby doubles the EOT without shifting the wavelength of the resonant transmission peak. Furthermore, the PBR structures show strong polarization dependence that is also strongly dependent on the structure of the PBR, as explained in terms of scattering of the surface waves.
© 2007 Optical Society of America
Extraordinary optical transmission (EOT) has been demonstrated for arrays of nano-holes in noble metal films . EOT is also possible from a single hole or slit by using encompassing periodic surface structures [2, 3, 4]. The periodicity of those surface structures was chosen to efficiently capture normally incident light from around the aperture and then resonantly transmit it through the aperture. As a result, the periodicity is similar to the wavelength of EOT, with differences coming from the properties of the metal (including allowing for surface plasmon polaritons, or SPPs) and the details of the scattering structure. Those past structures, flanked with periodic surface structures, still suffer from losses of the incident light that travels away from the apertures. The propagation of SPPs transmitted away from nano-hole arrays has been observed directly . This is an important consideration because it introduces both loss and cross-talk between a series of EOT structures on the same film.
In this work, we propose the use of plasmonic Bragg reflectors (PBRs) to recapture the surface waves emitted away from an array of nano-holes and thereby further enhance the EOT. (The term surface wave is used to mean waves at the surface from a full electromagnetic mode expansion, including SPPs). To meet the Bragg reflection condition, the periodicity of the PBRs is chosen to be half the periodicity of the array. To reflect in-phase with the EOT light, the PBR is placed a half period away from the array. We study the polarization dependence with the PBR structures placed on only two edges of the nano-hole array, at the positive and negative x-directions. The PBR reflects surface waves that have a component that is normally incident to the arrays. This means that the (1,0) resonance (horizontal, or x-polarization) is enhanced by the PBRs, but the (0,1) resonance (vertical, or y-polarization) is not enhanced. For the (1,1) resonance, a PBR made up of lines shows enhancement, whereas a PBR made up of dimples does not, because the dimple does not scatter light efficiently in the direction perpendicular to the direction of incidence. We also present the transmission of light from an array flanked on all four sides by PBRs. These do not exhibit the polarization effects of the arrays flanked on two sides by the PBRs. We have reported our preliminary results on PBRs, without polarization dependence or a detailed study of the PBR structure’s influence . A subsequent work has used PBRs in a resonant cavity configuration ; however, that work did not study dimple arrays or polarization dependence. Furthermore, in that work, the transmission spectrum was distorted through interplay between the resonant cavity and the PBR array, and the enhancement at the EOT wavelength of the bare array was only 50%. Here, the enhancement is doubled and the EOT resonance wavelength is not shifted by the PBR structures.
2. Focused ion beam fabrication
Figure 1 shows scanning electron micrographs of the array of circular holes flanked by partially milled dimples constituting the PBR layers and the array flanked by partially milled line PBRs. A bare hole-array was fabricated, as well as a series of PBRs without the hole-array (not shown). The arrays were fabricated on a gold film of thickness 100 nm evaporated on a glass substrate. A chromium layer of thickness 5 nm is present to provide adhesion between the gold and glass layers.
The arrays of circular holes of diameter 150 nm and periodicity 800 nm were fabricated using focused ion beam milling. (Periodicities of 600 nm and 900 nm were also fabricated and measured; however, the key results are represented by the 800 nm periodicity array and only those measurements are presented here). A gallium beam current of 30 pA at 30 kV was used for milling. The gallium beam spot size was 7 nm. A beam dwell time of 15 μs was required to mill completely through the gold and chromium layers. The array covered an area of 15 μm×30 μm, and 15 μm×15 μm to study the effects of having PBRs on two sides, and all four sides.
The PBR structures consisted either of lines of width 150 nm, or dimples similar to the holes. The PBR layers were milled through a depth of approximately 25 nm. The periodicity of the reflectors is equal to one half the periodicity of the array (i.e., 400 nm) to provide Bragg reflection in the plane of the surface at the wavelength of EOT. The PBRs each covered an area of 7.5 μm×30 μm and flanked the arrays either on two or all four sides. The PBRs were placed a half-period (400 nm) away from the array to ensure that the normally incident light scattered from the edge of the PBR is 90 degrees out of phase with the resonant EOT light, and thereby does not interfere to modify the transmission resonances. This spacing also provides in-phase reflection of the surface waves with the EOT light so that the EOT is enhanced but the resonant wavelength is not shifted.
3. Transmission measurement setup
The samples were each illuminated by a broad-band halogen source. A 20× microscope objective was used to focus the beam onto the sample with normal incidence, with an iris to provide an aperture. After the combination of iris and microscope objective, the spot size was approximately 30 μm. A polarizer was used to modify the linear polarization of the light incident on the array. The zeroth order transmitted light was collected using a fiber-coupled (200μm core) UV-visible spectrometer at a distance of 5 mm from the sample. PBR structures fabricated without the arrays, showed no apparent change in the transmission spectrum with respect to the non-patterned gold film.
4. Enhanced EOT and polarization results: Arrays with dimpled and lined PBRs
Figure 2 shows the transmission spectra of the three structures for incident light polarized horizontally, or along the x-direction. [Vertical polarization is shown with an arrow in Fig. 1(a)]. The array has a (1, 0) EOT resonance at 870 nm and a broad (1,1) EOT resonance around 700 nm. Here we are referring to the EOT resonances associated with the gold-air interface since the resonances from the glass-gold interface are diminished due to the lossy Cr adhesion layer . Although the (1, 0) peak is narrow, it sits on top of the broad enhanced transmission resonances from the higher-order resonances. Therefore, we consider the amplitude of that peak from the region of the transmission minimum at 850 nm (commonly attributed to the Wood’s anomaly ), as shown with arrows in Fig. 2.
For horizontal polarization, the (1,0) EOT resonance is doubled by the introduction of lined and dimpled PBRs. There was negligible change in the location of the resonances due to the quarter-wavelength position of the PBR away from the array. While the array with dimpled PBRs showed slightly lower enhancement, it is comparable to the array with lined PBRs; therefore the reflection of the dimpled PBR is similar to the lined PBR for the horizontal polarization. For the (1,1) resonance, the array with dimpled PBRs showed less enhancement in the EOT than the array with lined PBRs. Transmission measurements on the PBR structures without the arrays of holes showed that the PBRs alone do not allow transmission.
There is a small blue-shift in the (1,0) peak of the line PBR with respect to the dimpled PBR and the bare hole-array (without a PBR). Considering the experimental error from noise and systematic variations in the FIB spot-size (barely visible in Fig. 1), this blue-shift, which is less than 10 nm, may be attributed to the difference in the reflection characteristics of the line PBR with respect to the dimpled PBR.
There has been no observed influence on the transmission spectrum from the illumination of the PBRs. In samples with only the PBRs, and no hole-array, no transmission was observed. We have varied the aperture of the beam to within the PBR, as well as repeated the experiment with different focusing objectives. The illumination of the PBR is not phased-matched in order to couple light that is incident normal to the surface towards the array.
Figure 3 shows the transmission spectra of the arrays and arrays flanked by PBRs for vertically polarized light, along the y-direction. The PBRs show a smaller (0, 1) resonance for this polarization because the arrays are half as wide in the x-direction. Furthermore, the (0,1) resonance peak of the EOT does not show any significant modifications with the addition of the PBRs. For the (1,1) resonance, the array with dimpled PBRs show almost no enhancement in the EOT with respect to the bare array, while the array with lined PBRs still show a significant enhancement in the EOT.
Figure 4 shows the variation of the transmission at resonant wavelength of 867 nm for arrays with the PBRs placed at different spacings from the edge of the array, for horizontally polarized light. Maximum enhancement is found to occur when the PBR layers are half-period i.e. 400 nm away from the array. At half-period, the waves reflected by the PBR layers are in phase with the surface waves propagating from the holes in the array, leading to maximum constructive interference, and hence maximum transmission.
Figure 5 shows the non-polarized transmission spectra of 15×15 μm2, without PBRs and with PBRs on all four sides. The resonance at 870 nm is enhanced by 1.5 times due to the PBRs, and the resonance at 800 nm is enhanced by 1.9 times due to the PBRs. Gratings milled on the gold film without the arrays do not show any transmission. Due to the fact that the array is smaller than the other arrays discussed, a few differences in the spectrum are observed, and the resonance at 870 nm is not as pronounced as in larger arrays. As in the case with the structures with PBRs on two sides, no appreciable change in the position of the resonances was observed due to the PBRs. Due to the symmetry of these arrays, they did not present the polarization dependence that was observed in the structures with PBRs on only two opposite sides.
The transmission measurements show interesting polarization and structural dependence of the enhanced EOT from arrays with PBRs. These results can be explained by considering how light scatters from arrays of holes, dimples and lines in the gold film under polarized normal incidence and for surface waves.
The SPPs generated at the edge of a nano-hole forms a dipole-like scattering pattern, where the SPP propagates in the direction of the electric field polarization of normal-incidence light [9,10]. The important point is that there is no light scattered perpendicular to the direction of the electric field polarization and that the maximum scattering occurs in the direction of the electric field polarization. A similar pattern is expected from the dimples. The lines scatter most efficiently for polarization perpendicular to its edge. These considerations are now used to explain the observations.
The enhanced (1,0) EOT for horizontal polarization comes form surface waves being generated at the nano-holes (leading to EOT) and then propagating in the direction of the electric field of the normally incident light. These surface waves are normally incident on the PBRs and they are reflected off the dimples or the lines. Subsequently, the reflected surface waves are transmitted through the hole-array, in-phase with the EOT. The dimples give strong reflection because they have strong back-scattering of surface waves; therefore, the enhanced EOT is strong for both the arrays with the lined and the dimpled PBRs. For the (1,1) transmission and horizontal polarization, the dimples do not have as strong scattering at the 45 degree angle of this resonance, so their enhancement is reduced.
For the vertical polarization, the (0, 1) EOT resonance sees no enhancement since the surface waves propagate in the vertical direction after the normally incident light is scattered from the holes, and so they do not impinge upon the PBRs. For the (1, 1) resonance, there is a component of the surface waves at 45 degrees which propagates towards the PBRs and its reflected component is at -45 degrees. For the dimples, this reflected component is perpendicular to the incident wave, and so it does not exist in the dipole approximation. Therefore the dimples do not show enhanced EOT for the (1,1) resonance and y-polarization. For the lines, there is still a reflected component and so enhanced EOT is observed.
While the dimpled PBR has periodic-variation in the y-direction, the spacing is different from the x-direction. Therefore, if the dimpled PBR has a corresponding (1,1) resonance, it would be at a different wavelength to the (1,1) resonances of the hole-array. Therefore, this is not expected to enhance the EOT, and negligible enhancement was actually observed.
From Fig. 4, maximum transmission occurs when the PBR layers are placed half-period away from the array, while minimum occurs close to when the PBRs are a full period away. An approximately sinusoidal trend is observed in the peak transmission with different spacings of the PBRs from the array, with period equal to the period of the array. The intensity of the peak transmission is modulated by the separation, although the spectral position of the peak remains the same to within a few nanometers.
We believe that the additional peaks, in Fig. 2, at 620 nm and 830 nm are part of the resonances at 700 nm and 870 nm resonances, and this is the reason that they are also enhanced by the PBRs. It is possible to speculate that the origin of this peak is from the finite extent of the array, which could lead to differences observed between Figs. 2 and 3 since the array is rectangular. Furthermore, the smaller array in Fig. 5 has a different transmission spectrum. These features are apparent in the arrays even without the PBRs, and so a detailed explanation is not within the scope of this work.
PBRs have been shown to enhance the EOT from arrays of nano-holes. We have shown that the EOT can be doubled from arrays tens of microns in dimension, thereby increasing the efficiency of plasmonic devices. With appropriate positioning of the PBRs so that they produce constructive interference, enhanced EOT may be observed without any shift in the resonance wavelengths or significant changes in the spectral features. We have also demonstrated that there is a strong polarization dependence of the enhanced EOT, which has a different influence on the various EOT resonances. There is also a strong dependence of the various polarizations and EOT resonances on the structure of the PBR used. The polarization and structural dependencies were explained in terms of polarization-dependent excitation of surface waves and surface-wave scattering off of the various PBR structures. Polarization independent enhancement was also observed in the transmission when arrays were flanked with PBRs on all four sides. The enhancements were consistent with those seen for polarized light in structures flanked by gratings perpendicular to the direction of polarization.
These results may be extended to improve the efficiency of a number of other EOT structures, such as concentric grooves. They may also be used to eliminate cross-talk between adjacent EOT structures, such as nano-hole arrays with different phases or wavelengths.
The authors would like to thank Karen L. Kavanagh for use of Simon Fraser University’s Nano-Imaging Facility. The authors acknowledge funding from CIPI, NSERC, CFI and BCKDF.
References and links
1. T. W. Ebbesen, H. H. Lezec, H. F. Ghaemi, T. Thio, and P. A. Wolff “Extraordinary optical transmission through sub-wavelength hole arrays,” Nature 391, 667–669 (1998). [CrossRef]
2. T. Thio, K. M Pellerin, R. A. Linke, H. J. Lezec, and T. W. Ebbesen “Enhanced light transmission through a single sub-wavelength aperture,” Opt. Lett. 26, 1972–1974 (2001). [CrossRef]
3. L. Martín-Moreno, F. J. García-Vidal, H. J. Lezec, A. Degiron, and T. W. Ebbesen “Theory of highly directional emission from a single sub-wavelength aperture surrounded by surface corrugations,” Phys. Rev. Lett. 90, 167401-1–167401-4 (2003). [CrossRef] [PubMed]
4. F. J. García-Vidal, H. J. Lezec, T. W. Ebbesen, and L. Martín-Moreno “Multiple paths to enhance optical transmission through a single sub-wavelength slit,” Phys. Rev. Lett. 90, 213901-1–213901-4 (2003). [CrossRef] [PubMed]
5. D. S. Kim, S. C. Hohng, V. Malyarchuk, Y. C. Yoon, Y. H. Ahn, K. J. Yee, J. W. Park, J. Kim, Q. H. Park, and C. Lienau “Microscopic origin of surface plasmon radiation in plasmonic band-gap nanostructures,” Phys. Rev. Lett. 91, 143901-1–143901-4 (2003). [CrossRef] [PubMed]
6. P. Marthandam and R. Gordon “Plasmonic Bragg reflectors for sub-wavelength hole arrays in a metal film,” 20th Annual Meeting of IEEE-LEOS, Lake Buena Vista, Florida, 21–25 Oct. 2007.
7. N. C. Lindquist, A. Lesuffleur, and S-H Oh “Periodic modulcation of extraordinary optical transmission through sub-wavelength hole arrays using surrounding Bragg Mirrors,” arXiv preprint Server: http://arxiv.org/abs/0708.1314 (2007).
8. R. Gordon, M. Hughes, B. Leathem, K. L. Kavanagh, and A. G. Brolo “Basis and lattice polarization mechanisms for light transmission through nanohole arrays in a metal film,” Nano Lett. 5, 1243–1246 (2005). [CrossRef] [PubMed]
9. C. Sönnichsen, A. C. Durch, G. Steininger, M. Koch, G. von Plassen, and J. Feldmann “Launching surface plasmons into nanoholes in metal films,” Appl. Phys. Lett. 76, 140–142 (2000). [CrossRef]
10. M. Brun, A. Drezet, H. Mariette, N. Chevalier, J. C. Woehl, and S. Huant “Remote optical addressing of single nano-objects,” Europhys. Lett. 64, 634–640, (2003). [CrossRef]