applied sciences

Article A -Dependent Frequency-Selective Absorber with Multiple Absorption Peaks

Guangsheng Deng, Tianyu Xia, Yong Fang, Jun Yang and Zhiping Yin *

Academy of Photoelectric Technology, Hefei University of Technology, Key Lab of Special Display Technology, Ministry of Education, No. 193, Tunxi Road, Hefei 230009, China; [email protected] (G.D.); [email protected] (T.X.); [email protected] (Y.F.); [email protected] (J.Y.) * Correspondence: [email protected]; Tel.: +86-551-6290-2791

Academic Editor: Kenneth Chau Received: 19 April 2017; Accepted: 31 May 2017; Published: 4 June 2017

Abstract: A polarization-dependent, frequency-selective metamaterial (MM) absorber based on a single-layer patterned resonant structure intended for F frequency band is proposed. The design, fabrication, and measurement for the proposed absorber are presented. The absorber’s absorption properties at resonant frequencies have unique characteristics of a single-band, dual-band, or triple-band absorption for different polarization of the incident . The calculated surface current distributions and power loss distribution provide further understanding of physical mechanism of resonance absorption. Moreover, a high absorption for a wide range of TE-polarized oblique incidence was achieved. Hence, the MM structure realized on a highly flexible polyimide film, makingthe absorber suitable for conformal geometry applications. The proposed absorber has great potential in the development of polarization detectors and .

Keywords: absorber; metamaterial; multi-band absorption; polarization-dependent

1. Introduction Electromagnetic , which represent a class of artificial structures whose electric and magnetic response can be controlled freely, have attracted a significant amount of attention [1–3]. Due to their unique benefits, metamaterials have provided many important effects, such as invisible cloaking [4], negative [5], and super-lens [6]. Since the metamaterial absorber (MA) was presented in 2008 by Landy et al. [7], metamaterial has attracted many researchers because of its advantages, such as size minimization and suitable thickness, for applications using electromagnetic wave absorbers [8–10]. In recent years, various structures, such as split rings [11], closed rings [12], and fishnet structures [13], have been proposed for MAs. A certain progress has been made in broadband MAs developing [14,15]. Meanwhile, some studies on tunable absorbers [16,17] and multiband absorbers [18–20] have been performed. Currently, research is focused on absorbers intended for application in polarization imaging and polarization detecting. Therefore, the polarization-dependent frequency-selective absorbers are highly demanded. For most single patterned resonant structures, the number of absorption peaks was fixed or the absorbers were polarization-independent [21,22]. Hokmabadi et al. proposed a stereometamaterial perfect absorber, wherein absorption frequency is related to polarization angles [23]. Hu et al. presented a polarization-dependent MA with different resonant frequencies in two orthogonal directions [24]. However, these structures have only one polarization-dependent absorption peak, which limits their application in spectroscopic imaging and detecting, wherein multiple absorption peaks are required [25,26].

Appl. Sci. 2017, 7, 580; doi:10.3390/app7060580 www.mdpi.com/journal/applsci Appl. Sci. 2017, 7, 580 2 of 9

In this paper, the focus is on the F frequency band (F-band), and a simple polarization- dependent Appl. Sci. 2017, 7, 580 2 of 9 frequency-selective metamaterial absorber is proposed. The proposed structure has three separate absorptionIn peaksthis paper, with the high focus absorption is on the for F TE-polarized frequency band and (F TM-polarized‐band), and a simple normal polarization incident .‐ Furthermore,dependent the frequency absorption‐selective peaks metamaterial can be dynamically absorber is controlled proposed. byThe adjusting proposed ofstructure a polarization has three angle of theseparate incident absorption wave. Comparedpeaks with high to the absorption previously for TE reported‐polarized MAs, and TM the‐polarized MA proposed normal inincident this paper has severalwaves. advantages.Furthermore, Firstly,the absorption compared peaks to complex can be resonancedynamically structures controlled with by multipleadjusting absorption of a peaks,polarization such as all-in-one angle of the packed incident super-unit wave. Compared cell [27 ,to28 the] and previously stacked reported layers [ 29MAs,], the the proposed MA proposed absorber has onlyin this one paper patterned has several resonant advantages. structure, Firstly, which compared meets currentto complex demands resonance on miniaturization structures with and multiple absorption peaks, such as all‐in‐one packed super‐unit cell [27,28] and stacked layers [29], simplification. Secondly, the multiple absorption peaks can be dynamically tuned by changing of the proposed absorber has only one patterned resonant structure, which meets current demands on polarizationminiaturization angles and of the simplification. incident wave. Secondly, Lastly, the the multiple proposed absorption MA structure peaks can is realizedbe dynamically on a highly flexibletuned polyimide by changing film of without polarization a rigid angles substrate, of the incident so the absorber wave. Lastly, can the be easilyproposed used MA in structure the non-planar is and conformal-geometryrealized on a highly flexible applications. polyimide film without a rigid substrate, so the absorber can be easily used in the non‐planar and conformal‐geometry applications. 2. Structure and Design 2. Structure and Design The proposed MA structure and the unit cell structure of the copper pattern are presented in Figure1.The Unlike proposed previously MA structure reported and results, the unit where cell structure multiple of the different-sized copper pattern metal are presented patterns in were usedFigure to obtain 1. Unlike multi-band previously absorption, reported results, here, where the simple multiple etching different of‐sized a C-shaped metal patterns slot intowere aused circular patch,to provides obtain multi a triple-band‐band absorption, absorption. here, the The simple geometrical etching dimensions of a C‐shaped of slot the copperinto a circular pattern patch, shown in provides a triple‐band absorption. The geometrical dimensions of the copper pattern shown in Figure Figure1b are a = 820 µm, d = 60 µm, r1 = 112.5 µm, r2 = 161 µm, and r3 = 367.5 µm. The copper pattern 1b are a = 820 μm, d = 60 μm, r1 = 112.5 μm, r2 = 161 μm, and r3 = 367.5 μm. The copper pattern with a with a thickness of 0.5 µm and an electric conductivity of (5.8×107) S/m was printed on the top layer thickness of 0.5 μm and an electric conductivity of (5.8×107) S/m was printed on the top layer of a of a thickthick dielectricdielectric substrate substrate with with a athickness thickness of of50 μ 50mµ mthat that was was made made of polyimide of polyimide film filmwith witha relative a relative permittivitypermittivity of 3.6 of 3.6 and and a dielectric a dielectric loss loss tangent tangent of 0.02. Moreover, Moreover, a thick a thick copper copper plate plate that thatacts as acts a as a groundground plane plane was was printed printed on on the the bottom bottom of of the the dielectricdielectric substrate. substrate.

FigureFigure 1. (a 1.) The (a) The proposed proposed metamaterial metamaterial absorber absorber (MA) (MA) unit unit cell cell structure; structure; (b) (layoutb) layout of the of unit the cell unit cell structure and its optimal values given in micrometers: a = 820, d = 60, r1 = 112.5, r2 = 161, and r3 = 367.5. structure and its optimal values given in micrometers: a = 820, d = 60, r1 = 112.5, r2 = 161, and r3 = 367.5.

The absorptivity can be calculated by The absorptivity can be calculated by A = 1 − |S11|2 − |S21|2 (1) 2 2 where A, S11, and S21 are the absorptivity,A = 1 the− |Sreflection11| − coefficient,|S21| and the transmission coefficient, (1) respectively. In this case, the transmission coefficient S21 is zero because the thickness of the metal wheregroundA, S11 plane, and isS 21muchare larger the absorptivity, than its skin depth, the reflection so the absorptivity coefficient, is anddetermined the transmission by A = 1 − |S coefficient,11|2. respectively. In this case, the transmission coefficient S21 is zero because the thickness of the metal 3. SimulatedResultsandDiscussion 2 ground plane is much larger than its skin depth, so the absorptivity is determined by A = 1 − |S11| . The simulations were conducted in CST Studio 2014 (Darmstadt, Germany), using 3. SimulatedResultsandDiscussionthe finite‐element frequency‐domain method, and the unit cell boundary conditions were utilized in Theboth simulationsx and y directions. were Meanwhile, conducted the in Floquet CST Microwave port condition Studio was employed 2014 (Darmstadt, in the z‐direction. Germany), Since using the proposed absorber has similar absorption characteristics for TE‐polarized and TM‐polarized the finite-element frequency-domain method, and the unit cell boundary conditions were utilized normal incident waves, only the electromagnetic responses for the TE‐polarized incident wave, in both x and y directions. Meanwhile, the Floquet port condition was employed in the z-direction. whose is parallel to the x‐axis, were calculated for the normal incidence angle. Three Since the proposed absorber has similar absorption characteristics for TE-polarized and TM-polarized Appl. Sci. 2017, 7, 580 3 of 9 normalAppl. Sci. incident2017, 7, 580 waves, only the electromagnetic responses for the TE-polarized incident wave, whose3 of 9 electric field is parallel to the x-axis, were calculated for the normal incidence angle. Three distinct absorptiondistinct absorption peaks, which peaks, are which at 92.4 are GHz,at 92.4 102.0 GHz, GHz, 102.0 and GHz, 130.1 and GHz 130.1 with GHz absorptivities with absorptivities of 88.1%, of 98.2%,88.1%, 98.2%, and 94.6%, and 94.6%, respectively, respectively, are presented are presented in Figure in 2Figure. In addition, 2. In addition, three resonance three resonance peaks peaks were were labeled as mode f1, mode f2, and mode f3 starting from low to high frequencies. From Figure 2, labeled as mode f 1, mode f 2, and mode f 3 starting from low to high frequencies. From Figure2, the separationthe separation of the of first the twofirst adjacent two adjacent resonant resonant peaks ofpeaks the MA of the is not MA high is not enough, high whichenough, may which hinder may its applicationshinder its applications in the multiband in the detection. multiband This detection. can be improved This can by be reconstructing improved by the reconstructing geometries of the proposedgeometries structure. of the proposed structure.

Figure 2.2. Absorption spectra of thethe proposedproposed metamaterialmetamaterial absorberabsorber forfor thethe TE-polarizedTE‐polarized normalnormal incident wave. wave.

In order toto showshow thethe intrinsicintrinsic mechanismmechanism of thethe proposedproposed metamaterialmetamaterial absorber, thethe surfacesurface currents distributed on the metal pattern and the metal ground plane that correspond to three absorption peaks were calculatedcalculated andand thethe obtainedobtained resultsresults areare presentedpresented inin FigureFigure3 .3. As shown in Figure 3a, for mode f1, the surface currents on the metal pattern diverge from inner As shown in Figure3a, for mode f 1, the surface currents on the metal pattern diverge from inner circular patch to two arms of the closed ring, and then converge to the lower left corner of the ring, which means that mode f1 represents a typical dipole resonance. Meanwhile, the surface currents on which means that mode f 1 represents a typical dipole resonance. Meanwhile, the surface currents on the metal ground plane, shown in Figure3 3b,b, are are reversed, reversed, and and circulating circulating current current loops loops are are formed formed because the current onon twotwo metalmetal layerslayers areare anti-parallel,anti‐parallel, soso thethe magneticmagnetic responseresponse isis excited. excited. Similarly, for mode f2, the electric response is excited by surface currents on the metal pattern Similarly, for mode f 2, the electric response is excited by surface currents on the metal pattern that flowthat flow along along two arms two ofarms the closedof the ring,closed forming ring, forming a dipole a resonance dipole resonance (Figure3 c).(Figure However, 3c). However, compare compare with mode f1, the surface currents on the closed ring in mode f2 diverge from the lower right with mode f 1, the surface currents on the closed ring in mode f 2 diverge from the lower right corner andcorner converge and converge to the upper to the left upper corner. left Accordingcorner. According to results to presented results presented in Figure 3ind, Figure the magnetic 3d, the magnetic response of mode f2 is also excited by circulating currents between two metal layers. In the response of mode f 2 is also excited by circulating currents between two metal layers. In the case of case of mode f3, the surface currents are mainly concentrated on the connecting bar between a circular mode f 3, the surface currents are mainly concentrated on the connecting bar between a circular patch patch and a closed ring (Figure 3e), and mode f3 represents a dipole resonance that mainly depends and a closed ring (Figure3e), and mode f 3 represents a dipole resonance that mainly depends on the connectingon the connecting bar dimensions. bar dimensions. In order to demonstrate the absorption mechanism, the power loss density in the dielectric layer and the ohmic losses at resonator surface for mode f1 − f3 were determined, and the obtained results and the ohmic losses at resonator surface for mode f 1 − f 3 were determined, and the obtained results are illustratedillustrated in in Figures Figures4 and 45 ,and respectively. 5, respectively. Dielectric Dielectric losses of thelosses absorber of the are absorber mainly concentrated are mainly inconcentrated areas with in a strong areas with electric a strong field, while electric the field, ohmic while losses the at ohmic resonator losses surface at resonator are determined surface are by surfacedetermined currents. by surface In addition, currents. it has In beenaddition, observed it has that been the observed power losses that inthe the power dielectric losses layer in arethe moredielectric than layer dissipated are more on the than resonator dissipated surface, on the and resonator the polyimide surface, film and has the a polyimide very important film has role a in very the absorptionimportant role of the in incidentthe absorption waves. of the incident waves. Appl. Sci. 2017, 7, 580 4 of 9 Appl.Appl. Sci. Sci. 2017 2017, ,7 7, ,580 580 44 of of 9 9

FigureFigure 3. 3. Simulated SimulatedSimulated surface surface surface current current distribution distribution on on ( ( aa)) a a metal metal pattern pattern at at 92.4 92.4 GHz; GHz; ( ( bb)) a a metal metal metal ground ground planeplane atat 92.4GHz;92.4GHz; (( cc)) aa metalmetal pattern pattern atat 102.0102.0 GHz;GHz; (( dd)) aa metalmetal groundground planeplane atat 102.0102.0 GHz;GHz; (( ee)) aa metalmetal patternpattern at130.1 at130.1 GHz GHz and and ( (ff)) a a metal metal ground ground plane plane at at 130.1 130.1130.1 GHz. GHz.

FigureFigure 4. 4. Distribution DistributionDistribution of of dielectric dielectric power power losses losses on on polyimide polyimide film film film at at ( ( aa)) 92.4 92.4 GHz; GHz; ( ( bb)) 102.0 102.0 GHz, GHz, and and ((cc)) 130.1 130.1 GHz. GHz. Appl. Sci. 2017, 7, 580 5 of 9

Appl. Sci. 2017, 7, 580 5 of 9 Appl. Sci.Figure 2017 5., 7 Distribution, 580 of ohmic losses on resonator surface at (a) 92.4 GHz; (b) 102.0 GHz and (c) 130.1 GHz. 5 of 9

4. MA Absorption Dependence on the Wave Incident Angle and Polarization The absorption characteristics of the proposed structure for oblique incident wave with TE polarization were simulated, and the obtained results are shown in Figure 6a. During the calculation, E‐field was in the x‐direction, and H‐field and directions varied simultaneously for angle θ.As shown in Figure 6a, when the oblique incident angle increases from 0 to 60°, the absorption peaks at the three resonance frequencies are about 90% or greater, which is caused by a resonator structure, wherein for TE polarization a strong electric resonance can be excited with a relatively small field strength. For the TM‐polarized wave, whose H‐field is in the x‐direction, the absorption characteristics for the oblique incident wave were also simulated, and the obtained results are shown in Figure 6b. However, as can be seen in Figure 6b, when the incident angle increases, the absorption characteristic FigureFigure 5. 5. DistributionDistribution of ohmic of ohmic losses losses on resonator on resonator surface at surface (a) 92.4 at GHz; (a) ( 92.4b) 102.0 GHz; GHz (b and) 102.0 (c) 130.1 GHz GHz. and for TM(c)‐ 130.1polarized GHz. oblique incidence deteriorates rapidly at all resonance modes. Nonetheless, we were focused on absorption characteristics for different polarization angles for 4. MA Absorption Dependence on the Wave Incident Angle and Polarization the4. MA normal Absorption incident Dependence wave. For TE on and the TM Wave polarization, Incident Angle E‐field and changes Polarization for angle φ with respect to the xThe‐and absorption y‐direction, characteristics respectively. ofThe the simulated proposed responses structure forfor TEoblique and TMincident waves wave for differentwith TE The absorption characteristics of the proposed structure for oblique incident wave with TE polarization wereangles simulated, are presented and thein Figure obtained 7a andresults Figure are shown7b, respectively. in Figure 6a.As Duringshown inthe Figure calculation, 7a, for polarization were simulated, and the obtained results are shown in Figure6a. During the calculation, Eresonant‐field was modes in the f1 andx‐direction, f3, the absorption and H‐field is high and whenwave φpropagation is in the range directions 0–45°. However, varied simultaneously the resonance E-field was in the x-direction, and H-field and wave propagation directions varied simultaneously for forof mode angle f 2 θdecreases.As shown gradually in Figure when 6a, φwhen changes the fromoblique 0 to incident45°. According angle toincreases results presentedfrom 0 to in 60°, Figure the angle θ.As shown in Figure6a, when the oblique incident angle increases from 0 to 60 ◦, the absorption absorption7a, it can be peaks concluded at the thatthree the resonance resonance frequencies of mode f 2are vanishes about at90% φ =or 45° greater, because which the surface is caused current by a peaks at the three resonance frequencies are about 90% or greater, which is caused by a resonator resonatoron the resonator structure, surface wherein in mode for TE f2 (Figure polarization 3c) mainly a strong flows electric perpendicularly resonance canto the be directionexcited with of the a structure, wherein for TE polarization a strong electric resonance can be excited with a relatively small relativelyelectric field small vector. field Therefore, strength. the electric response cannot be excited. fieldFor strength. the TM‐polarized wave, whose H‐field is in the x‐direction, the absorption characteristics for the oblique incident wave were also simulated, and the obtained results are shown in Figure 6b. However, as can be seen in Figure 6b, when the incident angle increases, the absorption characteristic for TM‐polarized oblique incidence deteriorates rapidly at all resonance modes. Nonetheless, we were focused on absorption characteristics for different polarization angles for the normal incident wave. For TE and TM polarization, E‐field changes for angle φ with respect to the x‐and y‐direction, respectively. The simulated responses for TE and TM waves for different polarization angles are presented in Figure 7a and Figure 7b, respectively. As shown in Figure 7a, for resonant modes f1 and f3, the absorption is high when φ is in the range 0–45°. However, the resonance of mode f2 decreases gradually when φ changes from 0 to 45°. According to results presented in Figure 7a, it can be concluded that the resonance of mode f2 vanishes at φ = 45° because the surface current on the resonator surface in mode f2 (Figure 3c) mainly flows perpendicularly to the direction of the electric field vector. Therefore, the electric response cannot be excited.

Figure 6. Simulated responses of the structure for different incident angles θθ for (a) the TE-polarizedTE‐polarized wave and (b) the TM-polarizedTM‐polarized wave.

For the TM-polarized wave, whose H-field is in the x-direction, the absorption characteristics for the oblique incident wave were also simulated, and the obtained results are shown in Figure6b. However, as can be seen in Figure6b, when the incident angle increases, the absorption characteristic for TM-polarized oblique incidence deteriorates rapidly at all resonance modes. Nonetheless, we were focused on absorption characteristics for different polarization angles for the normal incident wave. For TE and TM polarization, E-field changes for angle ϕ with respect to the x-and y-direction, respectively. The simulated responses for TE and TM waves for different polarization angles are presented in Figures7a and7b, respectively. As shown in Figure7a, for resonant modes f and f , the absorption is high when ϕ is in the range 0–45◦. However, the resonance of mode f 1 3 2 decreases gradually when ϕ changes from 0 to 45◦. According to results presented in Figure7a, it can ◦ be concludedFigure 6. Simulated that the resonanceresponses of of the mode structuref 2 vanishes for different at ϕ incident= 45 because angles θ for the (a surface) the TE current‐polarized on the resonatorwave and surface (b) the in TM mode‐polarizedf 2 (Figure wave.3c) mainly flows perpendicularly to the direction of the electric field vector. Therefore, the electric response cannot be excited. Appl. Sci. 2017, 7, 580 6 of 9 Appl. Sci. 2017, 7, 580 6 of 9

Figure 7. Simulated absorption characteristic of the proposed structure for different polarization Figure 7.7. SimulatedSimulated absorption absorption characteristic characteristic of theof the proposed proposed structure structure for different for different polarization polarization angles angles φ for (a) the TE‐polarized wave and (b) the TM‐polarized wave. anglesϕ for ( aφ) for the ( TE-polarizeda) the TE‐polarized wave and wave (b )and the ( TM-polarizedb) the TM‐polarized wave. wave.

Similarly, when TM polarization angle is changed, the absorption peaks for mode f2 are above Similarly, when TM polarization angle is changed, the absorption peaks for mode f2 are above Similarly, when TM polarization angle is changed, the absorption peaks for mode f 2 are above 80%, while the absorptivities of mode f1 and f3 decrease gradually when φ changes from 0 to 45°. 80%, while the absorptivities of mode f1 and ff3 decrease gradually when φ changes from 0 to 45°.◦ 80%,Therefore, while the the proposed absorptivities structure of mode provides1 and both3 singledecrease‐band gradually and multi when‐bandϕ absorptionchanges from for different 0 to 45 . Therefore, the proposed structure provides both single‐band and multi‐band absorption for different Therefore,polarization the angles. proposed structure provides both single-band and multi-band absorption for different polarization angles. 5. Experimental Results 5. Experimental Experimental Results The MA was fabricated using standard photolithography. Firstly, the copper with a thickness of The MA MA was was fabricated fabricated using using standard standard photolithography. photolithography. Firstly, Firstly, the the copper copper with with a thickness a thickness of 0.5 μm was deposited on the bottom surface of the polyimide film. Then, the photolithography was 0.5of 0.5 μmµ wasm was deposited deposited on onthe thebottom bottom surface surface of the of thepolyimide polyimide film. film. Then, Then, the photolithography the photolithography was used to define the metallic patterned resonator at the top film surface. Lastly, the copper was usedwas used to define to define the themetallic metallic patterned patterned resonator resonator at atthe the top top film film surface. surface. Lastly, Lastly, the the copper copper was evaporated, and a metal lift‐off process was used to complete the pattern transfer. The fabricated evaporated, and and a metal lift lift-off‐off process was used to complete the pattern transfer. The fabricated sample composed by 18×18 unit‐cells with a dimension of 1.5 cm×1.5 cm is displayed in Figure 8. sample composed by 18×18 18×18 unit unit-cells‐cells with with a a dimension dimension of of 1.5 1.5 cm×1.5 cm×1.5 cm cm is is displayed displayed in in Figure Figure 8.8 .

Figure 8. Fabricated metamaterial absorber: (a) real size; (b) enlarged portion. Figure 8. Fabricated metamaterial absorber: (a) real size; (b) enlarged portion. Figure 8. Fabricated metamaterial absorber: (a) real size; (b) enlarged portion. The free space measurement method was adopted in the experiment.experiment. A pair of horn antennas The free space measurement method was adopted in the experiment. A pair of horn antennas were connectedconnected to to the the Agilent Agilent N5224A N5224A vector vector network network analyzer analyzer (Agilent (Agilent Technologies Technologies Inc, Santa Inc, Clara,Santa were connected to the Agilent N5224A vector network analyzer (Agilent Technologies Inc, Santa CA,Clara, USA) CA, via USA) VNA via extenders VNA extenders VDI VNAX600 VDI VNAX600 (Virginia (Virginia Diodes Inc, Diodes Charlottesville, Inc, Charlottesville, VA, USA) VA, that USA) were Clara, CA, USA) via VNA extenders VDI VNAX600 (Virginia Diodes Inc, Charlottesville, VA, USA) usedthat were for frequency used for rangefrequency 90–140 range GHz. 90–140 In this GHz. case, In one this case, one acted antenna as a transmitting acted as a antennatransmitting and that were used for frequency range 90–140 GHz. In this case, one antenna acted as a transmitting otherantenna as aand receiving other as antenna. a receiving In order antenna. to avoid In order the reflections to avoid fromthe reflections the environment, from the the environment, calibration was the antenna and other as a receiving antenna. In order to avoid the reflections from the environment, the performedcalibration aswas follows. performed The as reflections follows. fromThe reflections an identical from copper an identical plate, which copper acts plate, as an which ideal acts reflector, as an calibration was performed as follows. The reflections from an identical copper plate, which acts as an wereideal firstlyreflector, measured. were firstly Then, measured. the actual Then, reflections the actual from reflections the sample from were the sample calculated were by calculated comparison by ideal reflector, were firstly measured. Then, the actual reflections from the sample were calculated by ofcomparison the reflection of the from reflection the structure from the and structure the reflection and the from reflection the copper from plate.the copper It should plate. be It noted should that, be comparison of the reflection from the structure and the reflection from the copper plate. It should be fornoted the that, given for sample the given dimensions, sample dimensions, a minimum a far-fieldminimum distance far‐field of distance 40 cm must of 40 be cm met. must Moreover, be met. noted that, for the given sample dimensions, a minimum far‐field distance of 40 cm must be met. Appl. Sci. 2017 7 Appl. Sci. 2017, 7, 580 7 of 9

Moreover, the measurement of responses for different incident angles has not been carried out due the measurement of responses for different incident angles has not been carried out due to the large to the large of this finite sample at wide incident angles [30]. scattering of this finite sample at wide incident angles [30]. The measured MA absorption for various polarization angles (φ) for both TE‐polarized and TM‐ The measured MA absorption for various polarization angles (ϕ) for both TE-polarized and polarized normal incident waves is presented in Figure 9. For TE‐polarization, a dual‐band TM-polarized normal incident waves is presented in Figure9. For TE-polarization, a dual-band absorption is achieved for a polarization angle of 45°. However, only one absorption peak with absorption is achieved for a polarization angle of 45◦. However, only one absorption peak with absorptivity of above 70% is achieved for polarization angle of 45° in the case of TM‐polarized absorptivity of above 70% is achieved for polarization angle of 45◦ in the case of TM-polarized incident incident wave, which is in agreement with the simulated results. wave, which is in agreement with the simulated results.

Figure 9. a Figure 9. MeasuredMeasured MA MA absorption absorption for for different different polarization polarization angles angles for for ( ) the(a) the TE-polarized TE‐polarized wave wave and (b) the TM-polarized wave. and (b) the TM‐polarized wave.

6. Conclusions A simplesimple designdesign ofof aa polarization-dependentpolarization‐dependent frequency-selectivefrequency‐selective metamaterialmetamaterial absorber based on a a circular circular patch patch with with a C‐ ashaped C-shaped slot is slot presented. is presented. In contrast In to contrast previous to studies previous on compression studies on compressionof multi resonators, of multi in resonators, the proposed in the design, proposed a multi design,‐band a multi-band absorption absorption is obtained is using obtained a simple using aresonant simple resonantstructure. structure. Furthermore, Furthermore, changing the changing polarization the polarization angle of the angle incident of the wave, incident a triple wave,‐band, a triple-band,a dual‐band, a and dual-band, a single‐ andband a absorptions single-band can absorptions be achieved, can bewhich achieved, is required which in is the required polarization in the‐ polarization-dependentdependent detection. In detection. order to In demonstrate order to demonstrate the absorption the absorption mechanism, mechanism, the electric the electric field fielddistribution, distribution, the surface the surface current current distribution, distribution, and the and power the power-loss‐loss distributions distributions were were obtained. obtained. The Theproposed proposed design design was was verified verified both both by simulations by simulations and andexperiments, experiments, which which were were performed performed using using the thefree free space space measurement measurement method. method. A Agood good agreement agreement between between simulated simulated and and measured measured results results was achieved, which proves aa highhigh absorptionabsorption performanceperformance ofof thethe proposedproposed MA.MA.

Acknowledgments: This work was supported by the National Natural Science Foundation of China (No. 51607050) Acknowledgments: This work was supported by the National Natural Science Foundation of China (No. and the National Defense Pre-Research Foundation of China (Grant No. 6140239010106). 51607050) and the National Defense Pre‐Research Foundation of China (Grant No. 6140239010106). Author Contributions: Guangsheng Deng and Zhiping Yin conceived and wrote the paper; Tianyu Xia and YongAuthor Fang Contributions: performed the Guangsheng experiments; Deng Jun Yangand Zhiping guided the Yin experimental conceived and design wrote and the analyzed paper; theTianyu data. Xia and Yong Fang performed the experiments; Jun Yang guided the experimental design and analyzed the data. Conflicts of Interest: The authors declare no conflict of interest. Conflicts of Interest: The authors declare no conflict of interest. References References 1. Smith, D.; Pendry, J.; Wiltshire, M. Metamaterials and negative refractive index. Science 2004, 305, 788–792. 1. [Smith,CrossRef D.;][ Pendry,PubMed J.;] Wiltshire, M. Metamaterials and negative refractive index. Science 2004, 305, 788–792. 2. Chen, H.T.; Padilla, W.J.; Zide, J.M.; Gossard, A.C.; Taylor, A.J.; Averitt, R.D.R.D. Active devices. Nature 2006, 444, 597–600. [CrossRef][PubMed] 3. Poddubny, A.; A.; Iorsh, Iorsh, I.; I.; Belov, Belov, P.; P.; Kivshar, Kivshar, Y. Hyperbolic Y. Hyperbolic Metamaterials, Metamaterials. Nat. PhotonicsNat. Photonics 2013, 7,2013 948–957., 7, 948–957. 4. [Chen,CrossRef H.S.;] Wu, B.I.; Zhang, B.; Kong, J.A. Electromagnetic wave interactions with a metamaterial cloak. 4. Chen,Phys. Rev. H.S.; Lett. Wu, 2007 B.I.;, 99 Zhang,, 063903. B.; Kong, J.A. Electromagnetic wave interactions with a metamaterial cloak. 5. Phys.Valentine, Rev. Lett.J.; Zhang,2007, 99S.;, Zentgraf, 063903. [CrossRef T.; Ulin‐][Avila,PubMed E.; Genov,] D.A.; Bartal, G.; Zhang, X. Three‐dimensional optical metamaterial with a negative refractive index. Nature 2008, 455, 376–379. Appl. Sci. 2017, 7, 580 8 of 9

5. Valentine, J.; Zhang, S.; Zentgraf, T.; Ulin-Avila, E.; Genov, D.A.; Bartal, G.; Zhang, X. Three-dimensional optical metamaterial with a negative refractive index. Nature 2008, 455, 376–379. [CrossRef][PubMed] 6. Kaina, N.; Lemoult, F.; Fink, M.; Lerosey, G. Negative refractive index and acoustic from multiple scattering in single negative metamaterials. Nature 2015, 525, 77–81. [CrossRef][PubMed] 7. Landy, N.; Sajuyigbe, S.; Mock, J.; Smith, D.; Padilla, W. Perfect metamaterial absorber. Phys. Rev. Lett. 2008, 100, 207402. [CrossRef][PubMed] 8. Yin, S.; Zhu, J.F.; Xu, W.D.; Jiang, W.; Yuan, J.; Yin, G.; Xie, L.J.; Ying, Y.B.; Ma, Y.G. High-performance terahertz wave absorbers made of silicon-based metamaterials. Appl. Phys. Lett. 2015, 107, 073903. [CrossRef] 9. Huang, L.; Chowdhury, D.R.; Ramani, S.; Reiten, M.T.; Luo, S.N.; Taylor, A.J.; Chen, H.T. Experimental demonstration of terahertz metamaterial absorbers with a broad and flat high absorption band. Opt. Lett. 2012, 37, 154–156. [CrossRef][PubMed] 10. Shi, C.; Zang, X.F.; Wang, Y.Q.; Chen, L.; Cai, B.; Zhu, Y.M. A polarization-independent broadband terahertz absorber. Appl. Phy. Lett. 2014, 105, 031104. [CrossRef] 11. Zhu, W.R.; Huang, Y.J.; Rukhlenko, I.D.; Wen, G.J.; Premaratne, M. Configurable metamaterial absorber with pseudo wideband spectrum. Opt. Express 2012, 20, 6612–6621. [CrossRef][PubMed] 12. Wang, B.X.; Zhai, X.; Wang, G.Z.; Huang, W.Q.; Wang, L.L. Design of a Four-Band and Polarization-Insensitive Terahertz Metamaterial Absorber. IEEE Photonics J. 2015, 7, 4600108. [CrossRef] 13. Faraji, M.; Morawej-Farshi, M.K.; Yousefi, L. Tunable THz perfect absorber using graphene-based metamaterials. Opt. Commun. 2015, 355, 352–355. [CrossRef] 14. Zhu, J.F.; Ma, Z.F.; , W.J.; Ding, F.; He, Q.; Zhou, L.; Ma, Y.G. Ultra-broadband terahertz metamaterial absorber. Appl. Phys. Lett. 2014, 105, 021102. [CrossRef] 15. He, X.J.; Yan, S.T.; Ma, Q.X.; Zhang, Q.F.; Jia, P.; Wu, F.M.; Jiang, J.X. Broadband and polarization-insensitive terahertz absorber based on multilayer metamaterial. Opt. Commun. 2015, 340, 44–49. [CrossRef] 16. Shrekenhamer, D.; Chen, W.C.; Padilla, W.J. Liquid Absorber. Phys. Rev. Lett. 2013, 110, 177403. [CrossRef][PubMed] 17. Zhang, Y.; Feng, Y.; Zhu, B. Graphene based tunable metamaterial absorber and polarization modulation in terahertz frequency. Opt. Express 2014, 22, 22743–22752. [CrossRef][PubMed] 18. Yahiaoui, R.; Guillet, J.P.; de Miollis, F.; Mounaix, P. Ultra-flexible multiband terahertz metamaterial absorber for conformal geometry applications. Opt. Lett. 2013, 38, 4988–4990. [CrossRef][PubMed] 19. Yahiaoui, R.; Tan, S.Y.; Cong, L.Q.; Singh, R.; Yan, F.P.; Zhang, W.L. Multispectral terahertz sensing with highly flexible ultrathin metamaterial absorber. J. Appl. Phys. 2015, 118, 083103. [CrossRef] 20. Yahiaoui, R.; Hanai, K.; Takano, K.; Nishida, T.; Miyamaru, F.; Nakajima, M.; Hangyo, M. Trapping waves with terahertz metamaterial absorber based on isotropic Mie resonators. Opt. Lett. 2015, 40, 3197–3200. [CrossRef][PubMed] 21. Singh, P.K.; Korolev, K.A.; Afsar, M.N.; Sonkusale, S. Single and dual band 77/95/110 GHz metamaterial absorbers on flexible polyimide substrate. Appl. Phys. Lett. 2011, 99, 264101. [CrossRef] 22. Zhou, W.; Wang, P.; Wang, N.; Jiang, W.; Dong, X.; Hu, S. Microwave metamaterial absorber based on multiple square ring structures. AIP Adv. 2015, 5, 117109. [CrossRef] 23. Hokmabadi, M.P.; Wilbert, D.S.; Kung, P.; Kim, S.M. Polarization-Dependent, Frequency-Selective THz Stereometamaterial Perfect Absorber. Phys. Rev. Appl. 2014, 1, 044003. [CrossRef] 24. Hu, F.R.; Zou, T.B.; Quan, B.G.; Xu, X.L.; Bo, S.H.; Chen, T.; Wang, L.; Gu, C.Z.; Li, J.J. Polarization-dependent terahertz metamaterial absorber with high absorption in two orthogonal directions. Opt. Commun. 2015, 332, 321–326. [CrossRef] 25. Wang, B.X.; Wang, G.Z.; Sang, T. Simple design of novel triple-band terahertz metamaterial absorber for sensing application. J. Phys. D Appl. Phys. 2016, 49, 165307. [CrossRef] 26. Mao, Z.W.; Liu, S.B.; Bian, B.R.; Wang, B.Y.; Ma, B.; Chen, L.; Xu, J.Y. Multi-band polarization-insensitive metamaterial absorber based on Chinese ancient coin-shaped structures. J. Appl. Phys. 2014, 115, 204505. [CrossRef] 27. Arezoomanda, A.S.; Zarrabib, F.B.; Heydaria, S.; Gandjic, N.P. Independent polarization and multi-band THz absorber base on Jerusalem cross. Opt. Commun. 2015, 352, 121–126. [CrossRef] 28. Bhattacharyya, S.; Ghosh, S.; Srivastava, K.V. Triple band polarization-independent metamaterial absorber with bandwidth enhancement at X-band. J. Appl. Phys. 2013, 114, 094514. [CrossRef] Appl. Sci. 2017, 7, 580 9 of 9

29. Hu, F.R.; Wang, L.; Quan, B.G.; Xu, X.L.; Li, Z.; Wu, Z.A.; Pan, X.C. Design of a polarization insensitive multiband terahertz metamaterial absorber. J. Phys. D Appl. Phys. 2013, 46, 195103. [CrossRef] 30. Álvarez, H.F.; Gómez, M.E.; Las-Heras, F. Angular stability of metasurfaces: Challenges regarding reflectivity measurements. IEEE Antennas Propag. Mag. 2016, 58, 74–81. [CrossRef]

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).