A kind of plasmonic open waveguide, which is a periodic subwavelength metallic Domino array, is investigated both theoretically and experimentally in this paper. Based on the guiding mechanism of spoof surface plasmon polaritions (spoof SPPs), the transmission properties of this waveguide are controllable by altering the geometric parameters of the periodic structure. Microwave experimental results verify the high efficiency of wave guiding in such open waveguide, as predicted in theoretic analysis.

Surface plasmon polaritons (SPPs) are electromagnetic (EM) excitations propagating along the metal-dielectric interface, whose electromagnetic fields can be strongly confined to the near vicinity of the interface [

The proposed plasmonic open waveguide is constituted by periodic subwavelength Domino structures, as shown in

In experiment, the waveguide adaptor we used is in the X band of 8.2 GHz - 12.4 GHz. From the above analysis, when the height of Domino structure h increases, the asymptotic frequency f_{s} decreases and the working frequency band widens. On the other hand, from Ref [9,11], the asymptotic frequency f_{s} increases as the period d becomes smaller. For the case h = 0.4 mm and d = 10 mm as in

sequently, we can obtain a band pass filter by designing these parameters of the Domino structure. The transmission of the above structure is lossless as the metal is assumed to be PEC, but it is lossy for practical metal due to finite conductance. Utilizing the perturbation method [12- 14], we can calculate the loss of such waveguide made by Aluminum (Al) with finite conductance. The tangential electric field will vanish on the PEC surface, while for the conductor with finite conductivity, there will exist a slight tangential electric field, and a tangential magnetic field, they have the relation ofhere is the characteristic impedance, is the unit vector normal to the metallic surface and σ is the conductance of metal. The dissipated power in the metal can be calculated as

The integral area is the whole metallic surface for a Domino lattice unit. We get the tangential magnetic field under the assumption of PEC. By the power conservation principle, the decrease of the total power P_{f} of the spoof SPPs must be equal to the power dissipated, and the attenuation constant a of the spoof SPPs can be calculated as

Here P_{f} represents the total time-average power, d is the lattice constant of the plasmonic waveguide and k is the complex propagation constant. _{s} are 11.6352 GHz and 11.233 GHz for these two cases. It can be seen that the attenuation constant a is ultra-small at lower frequencies, and it greatly increases as approaching the asymptotic frequency. For example, for the case h = 4.5 mm, when f ≤ 11 GHz, a is smaller than 0.128 m^{−}^{1}, when f ≈ f_{s}, a ≈ 45.19 m^{−}^{1}, and for the case h = 5.5 mm, when f = 11.233 GHz, a ≈ 59.787 m^{−}^{1}. This is because the modal field is highly restricted near the metal surface near the asymptotic frequency to generate much larger loss. As a result, we can adjust the plasmonic open waveguide to work at the frequency a little lower than the asymptotic frequency f_{s}, which has the advantages of both low loss and high field confinement.

To experimentally verify the plasmonic waveguide above, we design and fabricate a metallic structure as shown in

As a result, the field distribution from the rectangular waveguide is gradually transformed into the spoof SPP

field distribution along the transition region, reducing the reflection in the junction between the plasmonic waveguide and the rectangular waveguide adapter. First, consider the case of h = 4 mm, d = 10 mm, a = 0.5 d, L = 5 mm, w = 75 mm, and t = 375 mm with dispersion line shown in solid black line in _{21} increases from −13.4 dB at 8 GHz to −2.49 dB at 11.24 GHz, and it falls below −12 dB close to the asymptotic frequency 11.505 GHz, as indicated dash red line. The simulation and experimental results agree well with each other, the slight magnitude difference is due to the loss in experimental metal which is assumed to be PEC in simulation. The measurement of S_{21} for metallic plane without the Domino structure is also shown in dotted blue line, which proves that the transmission property is induced by the spoof SPP mechanism of the Domino structure. For the case of h = 4.8 mm, d = 5.5 mm with dispersion line shown in solid black line in _{21} increases from −2.9 dB at 8 GHz to −1.478 dB at 9.645 GHz, then decrease to −9.94 dB at 10.98 GHz, and falls below −24.9 dB above the asymptotic frequency 11.233 GHz. From the previous analysis, the plasmonic waveguide with these geometric parameters has high modal field confinement, as a result, the EM field from the input can be effectively guided to the output by such Domino structure. The slight difference between the experimental and simulation is caused by the limit of fabrication precision.

In conclusion, we have studied in numerical and experiment the plasmonic open waveguide constructed by periodic subwavelength Domino structure, which supports spoof SPPs with high modal field confinement. Such waveguide has metallic open structure, whose geometric parameters can be adjusted to control the working frequency bandwidth. We utilized waveguide adaptor in our experiment, so that the S parameters of the Domino structure can be obtained to analyze the transmission properties. This kind of the waveguide structure can be applied in high power microwave transmission system.

The financial supports by the National Science Council of ROC under Grant Nos. NSC 100-2112-M-216-002, NSC 100-2221-E-216-015 and the National Natural Science Foundation of China under Grant No. 60971062 are gratefully acknowledged.