Introduction
Surface plasmon resonance (SPR) results from a coherent fluctuation of free electrons on the surface of a thin-film metal layer [1]. Light with a transverse magnetic (TM) polarization meeting the matching conditions of the surface plasmon waves (SPWs) is absorbed and creates a resonance absorption feature in the reflected angular or wavelength spectrum. Since the resonance condition is highly sensitive to the variations of the medium surrounding the metal surface, a small change in refractive index on the metal surface can be quantitatively analyzed by measuring the spectral shift of the resonance dip. This effect is widely used in highly sensitive sensors [2-4].
Most SPR-based sensors work as reflection types in the sense that a photodetector measures reflected light from the sensor surface by use of a standard Kretschmann configuration [5]. Such a setup uses a photodetector on the same side of a light source with respect to the metal film, which allows an extremely compact sensing scheme. However, the detectable range of the reflection-type SPR sensor is limited to a penetration depth in the 100-200nm range because the surface plasmon waves propagate along the metal surface and decay exponentially into both media. So a transmission-type SPR structure might be useful to investigate thick targets such as in cell analysis, in contrast to a traditional reflection-type structure [6]. In a transmission-type SPR sensor, surface plasmons are outcoupled into radiation modes by use of a waveguide layer in which the permittivity of the waveguide layer is periodically modulated. The change of the target environment can be investigated by measuring the resonance shift of the transmittance peak.
The coupling of surface plasmons with a diffraction grating has been reported in several other studies [7-9]. Most of these studies consider a corrugated thin-film metal to generate surface plasmons and for outcoupling in radiation modes. As mentioned by Park et al. [10], however, conversing of surface plasmons into only reflection modes may not have enough to connect the radiation modes from the diffract gratings directly to other external optical devices. To overcome this limitation, they utilized dielectric diffraction gratings for the efficient outcoupling of surface plasmons to transmission modes that propagate in free space. An outcoupling efficiency of 50% was presented and proved experimentally by the use of a conventional Kretschmann configuration and a dielectric grating on a silver film. From the report of Lenaerts et al. [11], transmittance of 68% was obtained for a modified structure in which a waveguide grating was added between a metal surface and air. Such gratings are composed of two kinds of dielectric material, which will intent the complexity of the fabrication process. Moreover, enhanced transmittance of up to 72% was predicted numerically by Shen et al. [12] through a thin metal slab with dielectric grating.
In this paper, we report the efficient outcoupling of the surface plasmons to transmission modes by the use of a dielectric grating. Numerical investigation results show that the peak transmission efficiency of 72.829% was obtained when the thickness and period of grating were optimized. The configuration discussed in this paper can be used as a gas or biochemical sensor, which presents extremely linear sensing characteristics. Furthermore, the influence of the refractive index of the prism on the sensing characteristics was investigated. The results presented herein have been carried out by the use of an algorithm based on rigorous coupled wave analysis [13-16]. The algorithm has been widely used for the calculation of the optical properties of periodic or aperiodic structures with a dimension less than the wavelength of the incident light.
Structure
A schematic diagram of the proposed transmission-type SPR sensor is shown in Fig. 1, where dielectric grating is represented as a one-dimensional array sitting on a 40-nm-thick silver layer (dm=40nm). In Fig. 1, dg represents the height of the dielectric grating and Λ is the period of the dielectric gratings. The permittivity of BK7 glass prism, dielectric gratings and metal layer were determined as εp=2.2958, εg=2.25 corresponding to polymethyl methacrylate (PMMA), and εm=-18+0.5i for silver at λ=633nm, respectively.
When the wave vector of the incident TM-polarized electromagnetic wave parallel to the prism-metal interface matches with that of the surface plasmon, a strong SPW is generated at the metal-dielectric interface. This resonance phenomenon can be mathematically expressed by equating the x-component of the input wave vector, kx, with the real part of the wave vector of the SPW, ksp, as shown in the following equations [1]:
At the resonance,
where kSPR and k0 denote the wave vectors of the surface plasmon and the incident light, and εs is the dielectric constant of the medium on top of a metal film. At a particular angle of incidence,θ1, when the value is equal to the resonance angle,θSPR, the resonance condition of Eq. (3) is satisfied and SPR is observed. The surface plasmon is then diffracted strongly by the dielectric gratings to airspace. The diffracted angle,θr, is given by
where k is the magnitude of the wave vector of diffracted light, q is the diffraction order, and K (=2π/Λ) is the grating vector. The grating period, Λ, is designed to be small enough that only the low diffraction orders can propagate in airspace.
Using rigorous coupled wave analysis we numerically estimate the diffraction efficiency of the radiation mode in the proposed structure considering an incident TM wave with a wavelength of 633nm. The period Λ is defined as 600nm when the line width is 300 nm (VF=0.5). This is almost equal to the wavelength of an incident light, so that only the low diffraction orders can transmit into air. The calculated reflection and transmission curves for the case of dg=110nm are shown in Fig. 2. It is obvious that the -1 diffraction order (-1T) is to be selected as the target under interrogation instead of 0T, since -1T is the indicator of a radiation mode.
Optimization
The intensity of the diffraction modes can be described by the relation I∝fQ, where f is the field-enhancement factor of the surface plasmon and Q is the coupling coefficient of the surface plasmon with the diffraction grating. f and Q depend strongly on the dielectric constant of the grating, but Q is much more sensitive to other grating parameters of the surface-relief profile, such as period, aspect ratio, and modulation depth [10].
Firstly, the height of the grating must now be determined. Figure 3(a) presents the transmittance of the -1 diffraction order (-1T) and corresponding resonance angles, as the grating thickness dg varies from 30 to 300nm. The transmittance and resonance characteristics exhibit strong dependence on the grating thickness. The transmittance of -1 diffraction order rapidly increases first to 61.46% until dg reaches 100nm, and then it becomes nearly constant. As the thickness exceeds 260nm, a slight decrease in transmittance appears since other diffraction orders become significant. The peak efficiency obtained was Tmax=65.263% at a thickness of dg=180nm. Figure 3(b) shows the resonance characteristics for the case of dg=180nm. Also, the resonance angle increases monotonically with grating thickness as the change in dg causes the local effective index to increase effectively. Results of approximately 68% of transmittance can be found in the literature for similar systems [11].
A new study has been carried out to optimize the device to reach at least the results published in the literature. Thus, the period of the dielectric gratings is still to be determined. Figure 4(a) presents the transmittance of the -1 diffraction order (-1T) and corresponding resonance angles, as the grating period Λ varies from 500 to 800nm. The outcoupling of the electromagnetic energy toward the air space is maximum if the period of grating is Λ=560nm. With the period of grating, a transitivity of 72.829% is obtained at the θ1=55.88º. Figure 4(b) shows the resonance characteristics for the case of Λ=560nm.
Fig.3 Influence of thickness of dielectric gratings on resonance characteristics (Λ=600nm, VF=0.5, and dm=40nm). (a) Transmittance efficiency (-1T) and resonance angle plotted as a function of thickness of dielectric gratings; (b) reflectance (0R) and transmittance (-1T) plotted as a function of incidence angle for the case of dg=180nm at θ1=55.96º and Tmax=65.263% |
Fig.4 Influence of period of dielectric gratings on resonance characteristics (dg=180nm, VF=0.5, and dm=40nm). (a) Transmittance efficiency (-1T) and resonance angle plotted as a function of period of dielectric gratings; (b) reflectance (0R) and transmittance (-1T) plotted as a function of incidence angle for the case of Λ=560nm at θ1=55.88º and Tmax=72.829% |
Sensing application
The proposed transmission-type SPR structure based on optimal dielectric gratings designed to provide maximum transmittance is now applied as a gas or biochemical sensor, the working principle of which is to use the sample medium as the superstrate. The resonance angle is sensitive to the local change in the environmental supersrate on the grating-metal film. By measuring the diffraction characteristics in an enhanced transmission mode, real-time monitoring can easily be achieved.
Figure 5 shows that transmittance spectra of –1 diffraction order for varying refractive indices of superstrate (increasing from 1.00 to 1.05 in steps of 0.01) with three different prism materials. The transmittance peak of -1 diffraction order can be found fine tuned by scanning of the incidence angle. With εp=3.19 (LaSFN21), the resonance angle shifts from 44.67º to 46.35º as the refractive index of the superstrate varies from 1.00 to 1.05 [Fig. 5(a)]. With εp=2.2958 (BK7 glass), the resonance angle shifts from 55.88º to 58.43º [Fig. 5(b)]. With εp=2.1199 (synthesized quartz), the resonance angle shifts from 59.43º to 62.38º [Fig. 5(c)], and the peak transmission efficiencies of -1 diffraction order are all larger than 0.7 for three prisms.
The resonance of each curve was obtained and plotted as a function of ns as shown in Fig. 6. The shifts of the resonance angle are almost linear over the range of refractive indices for each prism; the slopes of the straight lines denote the angular sensitivity. From linear fit analysis, the slopes of the lines were found to be 33.71º/RIU (R=0.99989), 51.2º/RIU (R=0.99982), and 59.09º/RIU (R=0.99978) with permittivity values of 3.19, 2.2958, and 2.1199, respectively, where RIU is refractive index unit and R is the correlation coefficient that denotes the linearity obtainable in the sensor performance. It can be seen that the angular sensitivity increases with the decrease in εp, which indicates that the choice of prism material proves to be a vital design parameter for optimizing the sensor’s performance.
Conclusion
We have presented a transmission-type SPR sensor based on resonant coupling surface plasmon to radiation modes by use of a dielectric grating, which might be useful to investigate thick targets compared with conventional reflection-type SPR sensors. Optimal grating thickness and period were determined in consideration of transmission efficiency by using an algorithm based on the rigorous coupled wave analysis. The results show that peak transmission efficiency of 72.829% was obtained for the configuration. This value is higher than those provided in the literature for this kind of system. For gas or biochemical sensing applications, the devices present extremely linear sensing characteristics with an angular sensitivity of 51.2º/RIU. In addition, the angular sensitivity of the SPR sensor can be improved by use of a lower refractive index glass prism.