† Corresponding author. E-mail:
Project supported by the 111 Project, China (Grant No. D17021) and the Changjiang Scholars and Innovative Research Team in University, China (Grant No. PCSIRT, IRT 16R07).
The plasmonics Talbot effect in metallic layer with infinite periodic grooves is presented in this study. Numerical approach based on the finite element method is employed to verify the derived Talbot carpet on the non-illumination side. The groove depth is less than the metallic layer thickness; however, for specific conditions, surface plasmons polaritons (SPPs) can penetrate through grooves, propagate under the metallic layer, and form Talbot revivals. The geometrical parameters are specified via groove width, gap size, period, and wavelength, and their proper values are determined by introducing two opening ratio parameters. To quantitatively compare different Talbot carpets, we introduce new parameters such as R-square that characterizes the periodicity of Talbot images. The higher the R-square of a carpet, the more coincident with non-paraxial approximation the Talbot distance becomes. We believe that our results can help to understand the nature of SPPs and also contribute to exploring this phenomenon in Talbot-image-based applications, including imaging, optical systems, and measurements.
The Talbot effect was discovered in 1836. The Talbot effect reveals the appearance of periodic images at serial distances on the back side of a periodic hole and slit array or on the illumination side of a grating when a structure is illuminated by normal monochromatic light.[1] The effect is attributed to nearfield diffraction and has been used in numerous fields, such as acoustics,[2] optical dispersive fiber system,[3] scanning electron microscope imaging,[4] lithography,[5] nonlinear optics,[6] microlens,[7] interferometric measurement,[8] array illumination,[9] and optical trapping.[10]
The unique features of surface plasmon polaritons (SPPs) make them a critical candidate for nanophotonics applications.[11–15] By analogy of SPPs, which are electromagnetic waves propagating along the interface between metal and dielectric, with other electromagnetic waves, we anticipate that SPPs also exhibit periodical revivals. More recently, the formation of revivals in the near field has been explored theoretically[16–20] and experimentally.[21–24] Plasmonics revivals were formed when the light wavelength was smaller than the array period.[25] In addition, their fast damping shortened their propagation length; thus, only the first and second row of Talbot revivals have received the attention. However, the Talbot distance in the plasmonics regime, was not coincident with that in the paraxial approximation that had been introduced by Rayleigh.[26] Furthermore, Oosten et al. introduced an alternative approach[22] and derived a non-paraxial formula to explain the shift.
In this paper, we utilize the tunneling nature of SPPs in a periodic groove array and prove that the Talbot carpet can be created if the separation distance is less than the SPPs penetration depth.[27] We use the finite elements method (FEM)-based numerical approach and determine the geometrical criteria for observing revivals. To achieve this goal, two ratios relating to geometrical parameters are introduced and a proper set of values is determined for each ratio. The comparison is accomplished by using the R-square, which is a fitting value in the MATLAB fitting tool. Furthermore, the contrast and size of self-images are determined. Our data will improve the understanding of SP tunneling in groove-array structure that can be used in a wide range of applications, such as in nanolithography, nonlinear optics, and image processing.
Figure
Our simulation approach is based on FEM and we used the Comsol commercial simulation platform. The boundary condition for the top and bottom were scattering boundary conditions that for normal illumination are proper conditions, whereas we used the periodic condition at each of the side boundaries. The silver dielectric function was described by Drude model (
Two basic characteristics of SPPs are its propagation length, expressed by Lsp = 1/Im(2ksp), where ksp = 2π/λsp,
The intensity profiles for different gaps are indicated in Fig.
To gain a more in-depth insight into the geometrical parameter criteria, we investigate the physical parameter variations for different values of α and γ. This is similar to the approach previously presented by Hua et al. for the nano-hole array.[30] We are interested in determining the effective working range for geometrical parameters of the groove array. Three parameters, namely, P, λ, and α, are subjected to optimization. Three different cases are considered:
Since period is constant, any change of α value can cause the thickness of grooves and steps to be modified. The value of P and λ are set to be 1220 nm and 632.8 nm, respectively. The light intensity along the central line varying with position in the z direction for α value increasing from 0.2 to 0.7 is shown in Fig.
The wavelength is still set to the same value as used previously. For each P, proper α should be determined. Moreover, it is found that light intensity profile behaves differently for various γ values. Figure
Regime 1 γ ≤ 1.25 and γ ≥ 2: Talbot image is not formed; thus, we ignore the studying of these values. When γ = 2 for small values of α, Talbot revivals do not appear. However, for the sake of compression, we assume the location of maximum intensity to be the revival point and extract the data.
Regime 2 1.25 < γ < 1.7, and 1.95 > γ > 2: despite the fact that the Talbot revivals appear, the R-square is less than 90%, or the Talbot distance is not consistent with the non-paraxial approximation. We only study the profiles with R-square larger than 90% to guarantee that their Talbot patterns are satisfactory.
Figure
Regime 3 1.7 < γ < 1.95: we refer to this range as working zone; because the profiles have an apparent periodic behavior and appeared Talbot images are in good consistency with non-paraxial approximation. The plots of R-square versus α for different γ values are shown in Fig.
For each wavelength, the proper γ and α should be recognized. However, to avoid long simulation time and large memory space and to accelerate the simulation speed, some approximations are indispensable. In the previous parts we found that for different cases, α has a constant working zone; thus, in this regime we pick up the best value for α, which is 0.54, and set it to be a constant. Figure
Regime 1 γ ≤ 1.25 and ≥ 2: our results show that the Talbot image does not appear any more.
Regime 2 1.25 < γ < 1.58: 1.97 < γ < 2: The Talbot revivals are formed but R-square is less than 90% and Talbot distance does not accord with non-paraxial predictions.
Regime 3 1.58 < γ < 1.97: The Talbot images are created with R-square is larger than 90%. Figure
The results are shown in Fig.
In this work, the geometrical parameter criteria for observing the Talbot effect in plasmonic infinite periodic metallic groove arrays are determined and analyzed numerically. Two opening ratios are introduced to investigate the structure. Furthermore, the R-square, size, and contrast for revivals are determined as well. Three different cases are imagined. For any wavelength in the visible range, the opening ratio α has a working zone located between 0.5 and 0.58. Meanwhile, the opening ratio γ has a working zone located between 1.7 and 1.95. In this range, the Talbot revivals with high contrast and R-square appear. Additionally, their Talbot distances are in good agreement with non-paraxial approximation. Our result can directly have the Talbot-image based applications, especially in imaging and lithography.
[1] | |
[2] | |
[3] | |
[4] | |
[5] | |
[6] | |
[7] | |
[8] | |
[9] | |
[10] | |
[11] | |
[12] | |
[13] | |
[14] | |
[15] | |
[16] | |
[17] | |
[18] | |
[19] | |
[20] | |
[21] | |
[22] | |
[23] | |
[24] | |
[25] | |
[26] | |
[27] | |
[28] | |
[29] | |
[30] |