† Corresponding author. E-mail:
Project supported by the National Basic Research Program of China (Grant No. 2013CBA01702) and the National Natural Science Foundation of China (Grant Nos. 61377016, 61575055, 10974039, 61307072, 61308017, and 61405056).
According to the electromagnetic field distributions, there exist two kinds of coupled spoof surface plasmon polaritons (SSPPs), the symmetric and anti-symmetric modes, in the three-dimensional (3D) subwavelength spoof–insulator–spoof (SIS) waveguide. We study the dispersion and excitation of the two kinds of coupled SSPPs supported by the 3D SIS waveguide. The evolution of the dispersion with the thickness and gap width of the waveguide is numerically investigated, and we give a theoretical analysis according to the coupling mechanism. Specially, based on the coupling mechanism, we design a zipper structure, through which the excitation and propagation of the anti-symmetric coupled modes can be realized effectively.
In ultraviolet, visible, and near-infrared frequency bands, metals are usually described by the Drude model and behave like plasma.[1] Coherent oscillation of conduction electrons near the surface of a metal gives rise to surface plasmon polaritons (SPPs), which propagate along the surface and are confined in the perpendicular direction.[2] Owing to their high spatial confinement of electromagnetic energy, SPPs have attracted a great deal of interest and have been applied to many fields in the last decade.[3–5] Sandwiching a thin insulator layer with two metallic claddings forms a metal–insulator–metal (MIM) waveguide.[6–9] In such an MIM waveguide, each single interface can sustain bound SPPs. When the two metal–insulator interfaces are close enough, the interaction between the SPPs will give rise to coupled SPPs. The coupled modes can be divided into symmetric and anti-symmetric ones according to the symmetry of their electromagnetic field distributions.[1,6,8] In far-infrared, terahertz, and microwave frequency bands, however, the metals behave like perfect electrical conductors (PECs). A metal–insulator interface cannot confine the surface electromagnetic waves effectively.[10,11] Although these waves can still propagate along the MIM waveguide, high attenuation occurs and limits the propagation distance.[12] If some corrugations, such as holes and grooves, are added to the metal–insulator interface, another surface electromagnetic wave, spoof SPPs (SSPPs), could be excited and confined at the interface.[13–17]
Among those corrugated interface structures, the one-dimensional arrays of subwavelength grooves have been widely studied for their remarkable structural simplicity.[15,18–20] Two such groove-array waveguides separated by an insulator gap symmetrically create a new waveguide, which is called a spoof–insulator–spoof (SIS) waveguide for its analog to the non-corrugated MIM waveguide.[21–23] Similarly, an SIS waveguide supports two kinds of coupled SSPPs, i.e., the symmetric and anti-symmetric modes. It is worth noting that the symmetry of the coupled modes has different definitions in the literature. For the coupled SPPs supported by the MIM waveguide in the visible regime, people always define the mode symmetry according to the electric field distributions,[1,7,8] while in the microwave and terahertz regions, the mode symmetry is usually defined according to the magnetic field distributions.[21–23]
Recently, Shen et al.[24,25] proposed that the dispersion of the SSPPs propagating along a three-dimensional (3D) groove-array waveguide is insensitive to the thickness of the waveguide. Their theoretical and experimental results demonstrated that the SSPPs are well confined in the perpendicular direction and can propagate long distances along arbitrarily curved ultrathin comb-shaped waveguides. Based on those discoveries, the 3D SIS waveguide has been proposed and investigated.[23,26–30] However, current researches mainly focus on one kind of the coupled modes.[26–28] If we consider their magnetic field distributions, the modes are symmetric ones. Recently, reference [30] proposed an effective method to excite the odd modes in the SIS waveguide. Actually, the odd modes in that literature are also symmetric ones. There is still no effective technique to excite the anti-symmetric modes defined by the magnetic field. In this work, we propose a coupling scheme and design a zipper structure. The anti-symmetric coupled SSPPs can be generated effectively in this structure.
Figure
The case of coupled SSPPs propagating in a two-dimensional (2D) SIS waveguide has been analyzed in Ref. [21], which defined the symmetric/anti-symmetric modes based on the magnetic field distributions. In that paper, Kats et al. obtained rigorous dispersion relations of the two coupled modes numerically. According to the dispersion expressions, the symmetric modes converge to the zero frequency, while the anti-symmetric modes have cut-off frequencies; the anti-symmetric modes have higher eigen-frequencies than the symmetric modes.[21] In the following, we are concerned with the coupled SSPPs supported by the 3D SIS waveguide. With the purpose of comparing our results with those of the 2D case, we also define the symmetry of the coupled modes according to their magnetic field distributions. All the results are calculated using the finite integration method in this paper. We consider the microwave region and choose d = 5 mm as the length unit, then a, h, and w are fixed at 0.4d, 0.8d, and d, respectively. The metal is modeled as a PEC.
The influence of the waveguide thickness and gap width on the dispersion relations is investigated. Figure
Figure
Reference [30] designed a transducer using an asymmetric coplanar waveguide (CPW) and a slot line to excite the odd coupled SSPPs, that is, the symmetric modes, but the excitation of the anti-symmetric modes is still lacking. In the next, we introduce a zipper structure, through which the generation of the anti-symmetric coupled modes is realized. The zipper structure, designed on the base of the coupling scheme, also uses the property of the CPW.
The zipper structure proposed is divided into four regions, as shown in Fig.
In our design, regions I, II, III, and IV serve as the guided wave source, SSPPs convertor, SSPPs coupler, and propagation waveguide, respectively. An ungrounded CPW supports two guided modes. One is symmetric and the other is anti-symmetric. The symmetric (anti-symmetric) guided wave will convert to two symmetric (anti-symmetric) SSPPs lines when it transmits through region II. The gradient gratings in region II guarantee that the guided waves supported by the CPW convert to SSPPs gradually.[31,32] In region III, the two symmetric (anti-symmetric) transmission lines of SSPPs couple with each other gradually and form the symmetric (anti-aymmetric) coupled SSPPs in the end. The coupled SSPPs will propagate along the ultrathin SIS waveguide in region IV. However, it is noteworthy that the symmetric guided mode cannot transmit in the CPW naturally because its existence needs the electric symmetry condition. Only the anti-symmetric guided wave can propagate in the CPW without any symmetry conditions. It is a typical feature of the CPW. The electric symmetry condition is difficult to operate in practice. Therefore, the zipper structure is only effective to generate the anti-symmetric coupled SSPPs.
We use the finite integration method to simulate the whole excitation process. Figure
For an excitation system, it is necessary to check out its S parameters, which express the excitation efficiency. Figure
In a 3D SIS waveguide, both the symmetric and anti-symmetric modes have high confinement. But the applications of the SIS waveguide based on the anti-symmetric coupled modes are lacking because of the lack of excitation technique. We designed an ultrathin zipper structure with an inherent coupling mechanism that couples two SSPPs lines. In this structure, the anti-symmetric coupled SSPPs can be generated effectively. It is easy to conclude that the generation of the anti-symmetric coupled modes through the zipper structure is realizable in experiment. The scheme is simple, effective, and has the potential for further design of novel planar devices such as filters, resonators, and couplers.
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 | |
31 | |
32 |