Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Free-standing guided-mode resonance band-pass filters: from 1D to 2D structures

Open Access Open Access

Abstract

We study experimentally and theoretically band-pass filters based on guided-mode resonances in free-standing metal-dielectric structures with subwavelength gratings. A variety of filters are obtained: polarizing filters with 1D gratings, and unpolarized or selective filters with 2D gratings, which are shown to behave as two crossed-1D structures. In either case, a high transmission (up to ≈ 79 %) is demonstrated, which represents an eight-fold enhancement compared to the geometrical transmission of the grating. We also show that the angular sensitivity strongly depends on the rotation axis of the sample. This behavior is explained with a detailed description of the guided-mode transmission mechanism.

© 2012 Optical Society of America

1. Introduction

Guided mode resonance (GMR) filters – also named resonant grating filters – have attracted a great interest since Wang and Magnusson’s proposal in 1990 [1, 2]. This is mainly due to their spectral properties: the peak symmetry, the rejection rate and the narrow bandwidth. These structures are composed of a sub-wavelength grating associated to one or several dielectric layers. The grating is designed in order to control the excitation of eigenmodes in the layers through the grating diffracted orders. In the last fifteen years, different all-dielectric structures have been studied theoretically for notch filtering [35] or band-pass filtering [68]. Recently, other concepts based on metallo-dielectric structures have been presented [9, 10]. These filters can be used for teledetection and imaging purposes. In this context, the association of a polarizer function with the spectral filtering properties can be useful. Indeed differences between the two polarization states of reflected or scattered light can be used to enhance the image contrast and to detect the shape of objects with a better resolution [11]. This allows for example to distinguish man-made targets in natural or urban scenes [12, 13]. However, polarization independent filters are usually required. In the case of mid- or far-infrared imaging with extremely low photon flux, the 50% of incident light lost by reflection on polarizing filters is a strong drawback. Several studies have been dedicated to the design of polarization independent filters at normal [1416] and oblique incidences [1720].

In Ref. [10], we have presented an original concept of band-pass filters based on a sub-wavelength metallic grating deposited on a dielectric free-standing layer. The experimental behaviour of a one dimensional (1D) lamellar structure made of parallel slits was shown at normal and oblique incidence for a polarization where the magnetic field H is parallel to the slits and to the rotation axis of the structure. Here, the complete experimental behavior (angular, spectral, polarization) of three different structures is highlighted: the one dimensional (1D) lamellar grating made of parallel slits, and two different two-dimensional (2D) arrays made of perpendicularly crossed slits with a rectangular or a square pattern. Particularly, experimental and theoretical angularly-resolved transmission spectra of the 1D and 2D structures in oblique incidence with azimuthal polarization are presented. It is worth noticing that this study is usually neglected by papers on 1D structures and yet it is a fundamental problem for applications. We highlight also the non-trivial (especially at oblique incidence) similarity between 2D structures and two crossed-1D structures. Finally, the physical origin of the transmitted bands is demonstrated by the calculation of the guided-modes dispersion curves.

2. 1D and 2D structures for guided-mode resonances filters

The structures consist of gold gratings deposited on a free-standing silicon nitride membrane (SiNx). The gratings are either one-dimensional, or two-dimensional with rectangular or square patterns as represented in Fig. 1. The components are designed to act as band-pass filters in the mid-infrared wavelength range. Electromagnetic simulations are based on a RCWA modal method [21]. The period d should fulfil two conditions: no diffracted order in free-space, and a coupling between the first-order diffraction wave and the guided-modes in the SiNx layer at the resonance wavelength λR. Light is trapped in the SiNx waveguide, and then coupled to the zero-order transmitted beam via the gold grating (Fig. 1(a)). Thus, as the metallic grating has very narrow slits, the structure is highly reflective except at the resonance wavelength λR where a sharp transmission peak is obtained thanks to this guided-mode resonance effect. The transmission peak depends mainly on the grating period d and the waveguide thickness td. The thickness of the silicon nitride film td is set so that the structure exhibits a single transmission peak at normal incidence. For the 1D structure, the metal thickness tm is chosen sufficiently large and the slit width w sufficiently small to reflect all the transverse electric polarized waves: light is transmitted only when the magnetic field is parallel to the slits [22, 23]. For the 2D structures, the transmission occurs for both polarizations via the slits parallel to the magnetic field.

 figure: Fig. 1

Fig. 1 Bandpass filters based on sub-wavelength metallic gratings deposited on a free-standing dielectric layer (a) Guided mode transmission mechanism. tm=100 nm, td=650 nm, w =200 nm; (b) 1D grating with dx =2110 nm; 2D grating with rectangular patterns: the grating period are dx =2110 nm and dy = 3000 nm; 2D grating with square patterns: dx = dy =2110 nm).

Download Full Size | PDF

2.1. Fabrication process

The structures have been fabricated following the process described in Ref. [10]. The silicon nitride layer is deposited on a silicon substrate by plasma enhanced chemical vapor deposition. Electron-beam lithography (with a poly(methyl methacrylate) resin), Cr/Au deposition and a lift-off process are used to obtain the grating. The thin chromium layer (1 nm-thick) is used to improve the adhesion of gold on silicon nitride. Last, the substrate is chemically etched on the back-side over a 3mm × 3mm window to leave the grating and the silicon nitride film free-standing. Nominal geometric parameters for the 1D structure are w =200 nm, tm =100 nm, td =650 nm and d =2110 nm. The 2D structures were fabricated with the same nominal geometric parameters as the 1D structure for w, tm and td. For the 2D structure with the rectangular patterns, the grating periods are dx =2110 nm and dy =3000 nm whereas for the square patterns, dx = dy =2110 nm.

2.2. Optical characterization at normal incidence

Measured transmission spectra of the three structures are shown in Fig. 2 for normal incidence. They have been measured with a Fourier-transform infrared spectrometer (spectral resolution of 5 cm−1) [24]. Figure 2(a) represents the transmission through the 1D structure for the two polarization states. It confirms that the light is transmitted only when the magnetic field, H, is parallel to the slits. The transmission peak reaches 78% at 2.97 μm, which represents an eight-fold enhancement compared to the geometrical transmission of the grating ( wd0.1). The rectangular pattern structure exhibits two distinct peaks, at 3.12 μm and 3.96 μm, with respect to the polarization. Spectral position can be selected with the use of a polarizer (Fig. 2(b)). As light can be transmitted only when the magnetic field is parallel to the slits (Fig. 2(a)), a 2D structure can be considered as two crossed-1D structures. The square patterns transmission results are shown in Fig. 2(c). This structure is a particular case of rectangular patterns structure where the two transmission peaks coincide. Thus a polarization-independent behavior, with a single peak at 3.12 μm is obtained (component equivalent to two identical crossed-1D structures).

 figure: Fig. 2

Fig. 2 Transmission spectra measured at normal incidence. The light is polarized with the H field parallel to the x-axis (dashed dark line) or parallel to the y-axis (red line). (a) 1D structure ; (b) 2D structure with rectangular patterns; (c) 2D structure with square patterns. Insets: Scanning electron microscope images of the samples (right), and wavevectors of the propagative diffracted orders in the SiNx layer at the resonance wavelength (left).

Download Full Size | PDF

Fabricated with identical dx, the peak position of the three structures (red solid curves) should be nearly identical. We attribute the small discrepancies between the 1D and 2D cases to different SiNx thicknesses, due to slightly different fabrication conditions.

3. Angular behavior and mechanism of guided-mode resonances

3.1. 1D structure

Figure 3, shows the absolute transmission intensity diagrams T (σ,k//) of the 1D structure, where σ = 1/λ is the wavenumber, and k// is the component of incident wavevector parallel to the (x0y)-plane. Depending on whether the rotation axis of the sample is y or x, we have respectively: k//=kx(0)=2πsin(θx)/λ or k//=ky(0)=2πsin(θy)/λ (see sketches of Fig. 3 for (xy) orientation). Figures 3(a) and 3(b) show the measured transmission diagrams of the 1D structure. Transmission spectra have been determined experimentally by angle-resolved measurements, performed with a Fourier-transform infrared spectrometer (spectral resolution of 5 cm−1 and beam convergence angle of ±0.5° [24]). The incidence angles θx (or θy) range from 0° to 40° in 0.5° increments. Figures 3(c) and 3(d) present the simulation of the transmission calculated with RCWA modal method [21] with a SiNx refractive index of 2, and a Drude model for the refractive index of gold: ɛ(λ) = 1 − [(λp + )λp]−1 with λp = 159 nm and γ = 0.0077. Calculated and experimental transmission diagrams are in qualitative agreement, reproducing the same features and dispersion properties. The small discrepancies (position and width of transmission resonances) are attributed to opto-geometrical parameters. It will be presented elsewhere after further investigation.

 figure: Fig. 3

Fig. 3 Angle-resolved transmission through the 1D structure as a function of the wavenumber, σ =1/λ and of the incident wavevector. (a) and (b) Measurements. (c) and (d) Calculations. (a) and (c) The rotation axis is parallel to the H field: the incident wavevector is kx(0)=2πsin(θx)/λ; (b) and (d) The rotation axis is parallel to the E field: the incident wavevector is ky(0)=2πsin(θy)/λ. Solid and dashed lines: calculated dispersion curves related to the guided modes kgTM and kgTE respectively (for a continuous gold layer)

Download Full Size | PDF

In order to study the resonant transmission mechanisms, we first determine the wavevectors kgTM(λ) and kgTE(λ) of the transverse magnetic (TM) and transverse electric (TE) guided mode respectively. TM and TE polarizations are defined according to the guided wave propagation plane in the SiNx film. In a first-order approximation, these guided modes can be calculated in the case of a thick continuous gold layer deposited on a SiNx waveguide of thickness td. They are calculated by solving a modal equation for each polarization state. In the following, we show that resonances occur when a diffracted order couples to one of these modes [2, 25].

The projection of the (±1) diffracted wavevectors in the (x0y)-plane is:

k//(±1)=(kx(0)±Kx)x+(ky(0)±Ky)y
where Kx=2πdx and Ky=2πdy. For 1D structures, Ky = 0.

We first consider the case where the incidence angle θx varies only in the (x0z) plane. Hence, ky(0)=0 and Eq. (1) can be simplified as:

k//(±1)=(kx(0)±Kx)x
When θx ≠ 0, the wavevector norms of the +1 and −1 orders are different, so for a given θx, coupling with the TM mode in the SiNx film occurs at two different wavelengths:
k//(+1)=kgTM(λ1)
k//(1)=kgTM(λ2)

These two dispersion equations, Eq. (3) and (4) are represented as solid line in Fig. 3(c). The coupling occurs only with the TM-polarized mode because in this case, the magnetic field remains on the y direction whereas the propagation direction is along the x axis. We can see that the approximation of the continuous gold layer fits quite well: calculated transmission bands are parallel to the dispersion curves. The discrepancies are attributed to the perturbation on the guided waves induced by the slits of the actual gold structure.

In the other configuration, the incidence angle θy varies in the (y0z) plane: hence, kx(0)=0 and Eq. (1) can be simplified as:

k//(±1)=±Kxx+ky(0)y
In this case, the wavevector norms of the +1 and −1 orders are the same for any θy value, so they couple together to the same guided-mode (at the same wavelength). Unlike the previous case, the diffracted wavevectors have x- and y- components: this allows the coupling with both TM- and TE-polarized waveguide eigenmodes. Thus, dispersion relations for a given θy, are:
k//(+1)=k//(1)=kgTM(λ3)
k//(+1)=k//(1)=kgTE(λ4)
These equations can be rewritten as follows, for each polarization state:
(2πσg)2(λ)ky(0)(λ)2=Kx2
where σg=kg(λ)2π, which leads to hyperbolic dispersion curves. The transmission band is flatter than in the previous case and the angular tolerance of the band-pass filter is improved. Calculated dispersion curves related to the guided modes kgTM and kgTE (still for continuous gold layer) are represented as solid and dashed lines respectively in the transmission diagram of Fig. 3(d). We can notice that these lines fit quite well with the calculated transmission bands and fit even better for the TE case. This can be explained because in TE polarization, the electric field is almost parallel to the slits, so light is nearly totally reflected (as shown in section 2): the TE guided mode sees a nearly perfect mirror, which is similar to the unstructured configuration used in this simplified model. On the contrary in TM polarization, the electric field is almost perpendicular to the slits, so the waves penetrate into the slits and the reflection coefficient is modified compared to the continuous gold film.

The agreement between the calculated dispersion curves (in the case of a continuous gold layer) and the calculated transmission band shows that the transmission peaks are due to guided mode resonance effects. Although a subwavelength metallic grating is here concerned, the effects of vertical and horizontal plasmonic resonances, as defined in Ref. [26], have no influence on the high transmission of the component. On the one hand, grating is too thin to exhibit vertical plasmonic resonances (see Ref. [27]) and is not in an optically symmetric environment as in Ref. [27, 28]. On the other hand, the horizontal surface plasmon dispersion curves don’t fit the calculated transmission band (see Ref. [10], where Rayleigh anomalies dispersion curves, very near form the SPP dispersion curves, are shown). The metallic grating here aims at obtaining a structure highly reflective, except at the resonance wavelength λR.

3.2. 2D structures

Figure 4(a) and 4(b) represent experimental optical transmission diagrams T(σ,k//) obtained with a 2D-square pattern. The rotation axis of the sample is y or x (see sketches). They are similar to the diagrams of the 1D structure (Fig. 3).

 figure: Fig. 4

Fig. 4 Angle-resolved transmission measurements through a structure with 2D square patterns as a function of σ=1/λ, the wavenumber, and of the incident wavevector. (a) The incident wavevector is kx(0)=2πsin(θx)/λ; (b) The incident wave vector is ky(0)=2πsin(θy)/λ.

Download Full Size | PDF

Experimental transmission diagrams of 2D-rectangular component represented on Fig. 5 are also a combination of transmission diagrams of 1D structures with periods of dx =2110 nm and dy =3000 nm: Fig. 5(a) and 5(b) are similar to Fig. 3(a) and 4(a), except that Fig. 5(b) has a shifted resonance peak at normal incidence. Indeed in this case, the magnetic field H is parallel to the slits separated by a period of 3000 nm. This confirms the similarity between 1D and 2D structures, even at oblique incidence. These results highlight the main diffracted orders involved in the resonance process, according to the magnetic field orientation. In Fig. 2, the colors of the arrows show the orders which are mainly diffracted according to the magnetic field orientation. This is confirmed by calculations: for the square pattern, diffraction efficiency in (0, ±1) diffracted orders is 1 to 3 orders of magnitude lower than in the (±1, 0) orders. These calculations prove that diffraction is only efficient in one direction: the one perpendicular to the magnetic field orientation. The similarity between 2D and two-crossed 1D structures results from this.

 figure: Fig. 5

Fig. 5 Angle-resolved transmission measurements through a structure with 2D rectangular patterns as a function of σ=1/λ, the wavenumber, and of the incident wavevector. (a) The H field is parallel to the slits separated by a period of 2110 nm and the incident wave vector is ky(0)=2πsin(θy)/λ; (b) The H field is parallel to the slits separated by a period of 3000 nm and the incident wavevector is kx(0)=2πsin(θx)/λ.

Download Full Size | PDF

4. Conclusion

We have presented three structures which can be used as transmission band-pass filters with different polarization behaviors. The 1D structure is a filter with polarization selectivity, the 2D structure with rectangular patterns is a tunable spectral filter when used in combination with a polarizer (it exhibits two different spectral positions according to the polarization state of the incident light) and the 2D structure with square patterns is polarization independent at normal incidence. These filters have a high efficiency: the experimental transmission peak reaches 78% for the 1D structure (68% for the 2D structures), which represents an eight-fold (seven-fold respectively) enhancement compared to the geometrical transmission of the grating. The spectral position of these filters can be adapted over a wide wavelength range by changing the grating period. Moreover, for a given period, the bandwidth and transmission maximum can be adjusted with the slit width. The 1D structure has different angular sensitivity, depending on the orientation of the incident plane relative to the slits. We explained this particular behavior with a simplified model based on the guided mode resonance mechanisms. In particular, diffracted orders in the SiNx layer have different expressions depending on whether the plane of incidence is parallel or perpendicular to the slits. Besides, we highlighted the similarities between 2D structures and two crossed-1D structures. The flexibility of polarization and spectral behaviors allows to adapt these filters to numerous applications, such as sensing, dense wavelength division multiplexing networks or multispectral imaging [29].

Acknowledgments

This work was partially supported by the METAPHOTONIQUE ANR project, and the PRF Metamat ONERA project.

References and links

1. S. S. Wang, R. Magnusson, J. S. Bagby, and M. G. Moharam, “Guided-mode resonances in planar dielectric-layer diffraction gratings,” J. Opt. Soc. Am. A 7, 1470–1474 (1990). [CrossRef]  

2. R. Magnusson and S. Wang, “New principle for optical filters,” Appl. Phys. Lett. 61, 1022–1024 (1992). [CrossRef]  

3. O. Stenzel, “Resonant reflection and absorption in grating waveguide structures,” Proc. SPIE 5355, 1–13 (2004). [CrossRef]  

4. M. L. Wu, C. L. Hsu, Y. C. Liu, C. M. Wang, and J. Y. Chang, “Silicon-based and suspended-membrane-type guided-mode resonance filters with a spectrum-modifying layer design,” Opt. Lett. 31, 3333–3335 (2006). [CrossRef]   [PubMed]  

5. N. Destouches, J. C. Pommier, O. Parriaux, and T. Clausnitzer, “Narrow band resonant grating of 100% reflection under normal incidence,” Opt. Express 14, 12613–12622 (2006). [CrossRef]   [PubMed]  

6. R. Magnusson and S. S. Wang, “Transmission bandpass guided-mode resonance filters,” Appl. Opt. 34, 8106–8109 (1995). [CrossRef]   [PubMed]  

7. S. Tibuleac and R. Magnusson, “Narrow-linewidth bandpass filters with diffractive thin-film layers,” Opt. Lett. 26, 584–586 (2001). [CrossRef]  

8. Y. Ding and R. Magnusson, “Doubly resonant single-layer bandpass optical filters,” Opt. Lett. 29, 1135–1137 (2004). [CrossRef]   [PubMed]  

9. E. Moreno, L. Martin-Moreno, and F. J. Garcia-Vidal, “Extraordinary optical transmission without plasmons: the s-polarization case,” J. Opt. A: Pure Appl. Opt. 8, S94 (2006). [CrossRef]  

10. E. Sakat, G. Vincent, P. Ghenuche, N. Bardou, S. Collin, F. Pardo, J.-L. Pelouard, and R. Haïdar, “Guided mode resonance in subwavelength metallodielectric free-standing grating for bandpass filtering,” Opt. Lett. 36, 3054–3056 (2011). [CrossRef]   [PubMed]  

11. X. Zhang, Y. Jiang, J. Guo, and X. Lu, “Target classification using active laser polarimetric imaging technique,” Proc. SPIE 7850, 78502T (2010). [CrossRef]  

12. D. B. Cavanaugh, K. R. Castle, and W. Davenport, “Anomaly detection using the hyperspectral polarimetric imaging testbed,” Proc. SPIE 6233, 62331Q (2006). [CrossRef]  

13. B. M. Flusche, M. G. Gartley, and J. R. Schott, “Defining a process to fuse polarimetric and spectral data for target detection and explore the trade space via simulation,” J. Appl. Remote Sens. 4, 043550 (2010). [CrossRef]  

14. X. Fu, K. Yi, J. Shao, and Z. Fan, “Nonpolarizing guided-mode resonance filter,” Opt. Lett. 34, 124–126 (2009). [CrossRef]   [PubMed]  

15. Y. Wang, Y. Kanamori, J. Ye, H. Sameshima, and K. Hane, “Fabrication and characterization of nanoscale resonant gratings on thin silicon membrane,” Opt. Express 17, 4938–4943 (2009). [CrossRef]   [PubMed]  

16. J. L. Perchec, R. E. de Lamaestre, M. Brun, N. Rochat, O. Gravrand, G. Badano, J. Hazart, and S. Nicoletti, “High rejection bandpass optical filters based on sub-wavelength metal patch arrays,” Opt. Express 19, 15720–15731 (2011). [CrossRef]   [PubMed]  

17. A.-L Fehrembach and A. Sentenac, “Unpolarized narrow-band filtering with resonant gratings,” Appl. Phys. Lett. 86, 121105 (2005). [CrossRef]  

18. O. Boyko, F. Lemarchand, A. Talneau, A.-L. Fehrembach, and A. Sentenac, “Experimental demonstration of ultrasharp unpolarized filtering by resonant gratings at oblique incidence,” J. Opt. Soc. Am. A 26, 676–679 (2009). [CrossRef]  

19. H. Xu-Hui, G. Ke, S. Tian-Yu, and W. Dong-Min, “Polarization-independent guided-mode resonance filters under oblique incidence,” Chin. Phys. Lett. 27, 74211–74213 (2010). [CrossRef]  

20. A.-L. Fehrembach, K. C. S. Yu, A. Monmayrant, P. Arguel, A. Sentenac, and O. Gauthier-Lafaye, “Tunable, polarization independent, narrow-band filtering with one-dimensional crossed resonant gratings,” Opt. Lett. 36, 1662–1664 (2011). [CrossRef]   [PubMed]  

21. J. P. Hugonin and P. Lalanne, “Reticolo software for grating analysis,” Institut of Optics Graduates School (2005).

22. P. Lalanne, J. P. Hugonin, S. Astilean, M. Palamaru, and K. D. Möller, “One-mode model and airy-like formulae for one-dimensional metallic gratings,” J. Opt. A: Pure Appl. Opt. 2, 48 (2000). [CrossRef]  

23. S. Collin, F. Pardo, and J. L. Pelouard, “Waveguiding in nanoscale metallic apertures,” Opt. Express 15, 4310–4320 (2007). [CrossRef]   [PubMed]  

24. C. Billaudeau, S. Collin, C. Sauvan, N. Bardou, F. Pardo, and J.-L. Pelouard, “Angle-resolved transmission measurements through anisotropic two-dimensional plasmonic crystals,” Opt. Lett. 33, 165–167 (2008). [CrossRef]   [PubMed]  

25. A.-L. Fehrembach, D. Maystre, and A. Sentenac, “Phenomenological theory of filtering by resonant dielectric gratings,” J. Opt. Soc. Am. A. 19, 1136–1144 (2002). [CrossRef]  

26. S. Collin, F. Pardo, R. Teissier, and J.-L. Pelouard, “Strong discontinuities in the complex photonic band structure of transmission metallic gratings,” Phys. Rev. B. 63, 033107 (2001). [CrossRef]  

27. J.-A. Porto, F. J. Garcia-Vidal, and J. B. Pendry, “Transmission resonances on metallic gratings with very narrow slits,” Phys. Rev. Lett. 83, 2845–2848 (1999). [CrossRef]  

28. F.-J. Garcia-Vidal and L. Martin-Moreno, “Transmission and focusing of light in one-dimensional periodically nanostructured metals,” Phys. Rev. B. 66, 155412 (2002). [CrossRef]  

29. R. Haïdar, G. Vincent, S. Collin, N. Bardou, N. Guérineau, J. Deschamps, and J.-L Pelouard, “Free-standing subwavelength metallic gratings for snapshot multispectral imaging,” Appl. Phys. Lett. 96, 221104 (2010). [CrossRef]  

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (5)

Fig. 1
Fig. 1 Bandpass filters based on sub-wavelength metallic gratings deposited on a free-standing dielectric layer (a) Guided mode transmission mechanism. tm=100 nm, td=650 nm, w =200 nm; (b) 1D grating with dx =2110 nm; 2D grating with rectangular patterns: the grating period are dx =2110 nm and dy = 3000 nm; 2D grating with square patterns: dx = dy =2110 nm).
Fig. 2
Fig. 2 Transmission spectra measured at normal incidence. The light is polarized with the H field parallel to the x-axis (dashed dark line) or parallel to the y-axis (red line). (a) 1D structure ; (b) 2D structure with rectangular patterns; (c) 2D structure with square patterns. Insets: Scanning electron microscope images of the samples (right), and wavevectors of the propagative diffracted orders in the SiNx layer at the resonance wavelength (left).
Fig. 3
Fig. 3 Angle-resolved transmission through the 1D structure as a function of the wavenumber, σ =1/λ and of the incident wavevector. (a) and (b) Measurements. (c) and (d) Calculations. (a) and (c) The rotation axis is parallel to the H field: the incident wavevector is k x ( 0 ) = 2 π sin ( θ x ) / λ; (b) and (d) The rotation axis is parallel to the E field: the incident wavevector is k y ( 0 ) = 2 π sin ( θ y ) / λ. Solid and dashed lines: calculated dispersion curves related to the guided modes kgTM and kgTE respectively (for a continuous gold layer)
Fig. 4
Fig. 4 Angle-resolved transmission measurements through a structure with 2D square patterns as a function of σ=1/λ, the wavenumber, and of the incident wavevector. (a) The incident wavevector is k x ( 0 ) = 2 π sin ( θ x ) / λ; (b) The incident wave vector is k y ( 0 ) = 2 π sin ( θ y ) / λ.
Fig. 5
Fig. 5 Angle-resolved transmission measurements through a structure with 2D rectangular patterns as a function of σ=1/λ, the wavenumber, and of the incident wavevector. (a) The H field is parallel to the slits separated by a period of 2110 nm and the incident wave vector is k y ( 0 ) = 2 π sin ( θ y ) / λ; (b) The H field is parallel to the slits separated by a period of 3000 nm and the incident wavevector is k x ( 0 ) = 2 π sin ( θ x ) / λ.

Equations (8)

Equations on this page are rendered with MathJax. Learn more.

k / / ( ± 1 ) = ( k x ( 0 ) ± K x ) x + ( k y ( 0 ) ± K y ) y
k / / ( ± 1 ) = ( k x ( 0 ) ± K x ) x
k / / ( + 1 ) = k g T M ( λ 1 )
k / / ( 1 ) = k g T M ( λ 2 )
k / / ( ± 1 ) = ± K x x + k y ( 0 ) y
k / / ( + 1 ) = k / / ( 1 ) = k g T M ( λ 3 )
k / / ( + 1 ) = k / / ( 1 ) = k g T E ( λ 4 )
( 2 π σ g ) 2 ( λ ) k y ( 0 ) ( λ ) 2 = K x 2
Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.