Polarization insensitive, broadband terahertzmetamaterial absorber

James Grant, Yong Ma, Shimul Saha, Ata Khalid, and David R. S. Cumming*School of Engineering, University of Glasgow, Glasgow, United Kingdom, G12 8LT

*Corresponding author: david.cumming.2@glasgow.ac.uk

Received May 5, 2011; revised August 1, 2011; accepted August 15, 2011;posted August 16, 2011 (Doc. ID 147121); published August 31, 2011

We present the simulation, implementation, and measurement of a polarization insensitive broadband resonantterahertz metamaterial absorber. By stacking metal-insulator layers with differing structural dimensions, three clo-sely positioned resonant peaks are merged into one broadband absorption spectrum. Greater than 60% absorption isobtained across a frequency range of 1:86THz where the central resonance frequency is 5THz. The FWHM of thedevice is 48%, which is two and half times greater than the FWHM of a single layer structure. Such metamaterialsare promising candidates as absorbing elements for bolometric terahertz imaging. © 2011 Optical Society ofAmericaOCIS codes: 160.3918, 040.2235, 250.5403.

Metamaterials (MMs) have been widely explored sincethe first theoretical [1] and experimental demonstration[2] of their unique properties. MMs can provide a highlycontrollable electromagnetic response, and to date MMshave been demonstrated in every technologically rele-vant spectral range including the optical [3], near-IR[4], mid-IR [5], terahertz [6], millimeter-wave [7], micro-wave [8], and radio [9] bands. A branch of MMs currentlyprovoking extensive interest is MM perfect absorbers. Bymanipulating the effective electrical permittivity, ε, andmagnetic permeability, μ, absorption close to unity ispossible [6]. The ideal MM absorber should have anabsorption magnitude close to unity and be polarizationinsensitive and omnidirectional. The concept of a MM ab-sorber is especially important at terahertz frequencieswhere it is difficult to find strong frequency selectiveterahertz absorbers [10]. MM absorbers have been devel-oped in order to overcome the thickness limitation oftraditional quarter wavelength devices. In such MM ab-sorbers, the thickness of the structure is typicallyλ=20. MM absorbers naturally lend themselves to tera-hertz detection applications, such as thermal sensorswhich, if integrated with suitable terahertz sources(e.g., quantum cascade lasers), could lead to compact,highly sensitive, low cost terahertz imaging systems.Terahertz MM absorber designs are based on strong

electromagnetic (EM) resonances to absorb the incidentlight and consequently the bandwidth of this resonant ab-sorption is narrow, typically no more than 20% of the cen-ter frequency. Both single-band [11,12] and dual-band[13] absorbers have been demonstrated, and in the lattercase, both resonances are narrowband. Broadband ab-sorbers have been studied in the 1–20GHz frequency re-gime [14,15], where the fabrication of such devices is farsimpler than for the terahertz. This Letter describes thedesign, simulation, fabrication, and characterization oftwo types of broadband terahertz MM absorbers. Thefirst device, shown in Figs. 1(a) and 1(b), consists ofan alternating stack of metal crosses and dielectric layerson top of a metal ground plane. The cross-shaped reso-nator is an example of an electric ring resonator (ERR)[16,17] and couples strongly to uniform electric fields,but negligibly to a magnetic field. By pairing the ERR

with a ground plane, the magnetic component of the in-cident terahertz wave induces a current in the sections ofthe ERR that are parallel to the direction of the E-field.The electric and magnetic response can then be tunedindependently and the impedance of the structurematched to free space by varying the geometry of theERR and the distance between the twometallic elements.The ERR determines the electric response while the di-electric type and thickness between the ERR and theground plane determines the magnetic response. To max-imize the absorption magnitude, the impedance (Z ¼ffiffiffiffiffiffiffiffi

μ=εp

) of the MM structure must be matched to thatof free space, i.e., ε ¼ μ ⇒ Z ¼ 1 (for normalized coeffi-cients of ε and μ) [14,18]. The resonant frequency of theabsorption peak is mainly determined by the arm lengthof the cross that is parallel to the electric field. Therefore,in different layers we design crosses of differing lengths(L1–L3) in order to support several resonant modes clo-sely positioned together in the absorption spectrum. Bytuning the dielectric thickness (h1–h3), the multilayerstructure can be impedance-matched to free space at

Fig. 1. (Color online) (a) Plan view of the three-layer MMabsorber and (b) cross section of the complete devive.(c) SEM image of nine unit cells of the multilayer absorberand (d) SEM image of a single superpixel broadband absorber.The orientation for TE polarization is shown in the inset.

3476 OPTICS LETTERS / Vol. 36, No. 17 / September 1, 2011

0146-9592/11/173476-03$15.00/0 © 2011 Optical Society of America

each resonant frequency and broadband absorption ob-tained. We have also designed a simplified broadbandterahertz absorber. This device consists of a four ERRsincorporated into a four color superpixel onto a singledielectric layer, i.e., ground plane/dielectric/metallicERRs. Such a device is much simpler to fabricate than themultilayer absorber. The cross length of each ERR is var-ied to produce several resonances.The optimized multilayer MM absorber structure was

obtained through finite-difference time-domain (FDTD)simulations (Lumerical Solutions, Inc.). Ye et al. [19] havepresented simulation data of a similar structure and haveshown that, due to its inherent symmetry, the absorptioncharacteristics are omnidirectional for azimuthal anglesless than 80°, implying there is minimal scattering, andpolarization insensitive. Our three-dimensional simula-tions were performed with a plane wave source incidentin the z direction on the multilayer unit cell. Periodicboundary conditions were used for the xy plane alongwith a mesh step size of Δx ¼ Δy ¼ 0:2 μm and Δz ¼0:05 μm. The metallic sections of the absorber were mod-eled as Au with a frequency independent conductivity of4 × 107 Sm−1. A refractive index of 1:8þ 0:06i was usedfor the dielectric. Reflection and transmission spectrawere recorded at planes 100 μm above and 100 μm belowthe top ERR. The optimum values of the parametersshown in Figs. 1(a) and 1(b) are listed in Table 1. Thewidth of each cross arm is 6 μm, while the thicknessof each ERR and the ground plane is 200 nm. The unitcell has a period of 22 μm. Figure 1(c) shows a scanningelectron microscope (SEM) image of the multilayerabsorber.Standard metal evaporation, spin coating, and e-beam

lithography techniques were used to fabricate the multi-layer absorber. First, a 20=180 nm Ti/Au metallic groundplane was evaporated onto a silicon substrate. The di-electrics layers were deposited by spin coating a liquidpolyimide (HD Microsystems PI2545). The thickness ofeach dielectric layer was modified by varying the spinspeed/duration and by using multiple coatings. Afterspinning the required thickness of polyimide, the layerwas cured in a nitrogen purged oven for at least 2 h at180 °C. Each ERR was defined in a bilayer of polymethylmethacrylate e-beam resist using a Vistec VB6 e-beamtool and 20=180 nm film of Ti/Au evaporated and liftedoff. Standard e-beam registration was used to ensure ac-curate alignment of each ERR to approximately 50 nm.Samples were characterized under vacuum in a Bruker

IFS 66v/S Fourier transform IR spectrometer in transmis-sion mode at normal incidence and in reflection mode at30° incidence. Reflection measurements at other incidentangles were not possible due the limitations of the mea-surement setup. The measured transmission spectrawere normalized with respect to the signal measuredfrom a 7mm diameter open aperture, and the reflectionspectra were normalized to that of a gold mirror. The

resulting absorption, A, was therefore calculated usingAðωÞ ¼ 1 − RðωÞ − TðωÞ, where R is the reflection coeffi-cient and T is the transmission coefficient. As the thick-ness of the ground plane is much larger than the typicalskin depth in the terahertz regime, the reflection is theonly factor limiting absorption. Experimental measure-ments were also performed on samples with no ERRlayers to confirm that absorption was a consequence ofthe MM structure and not of the dielectric. The 5 μm thickpolyimide sample with no ERR structure had a maximumabsorption of 5% across the frequency range of interest,see Fig. 2(a), thereby verifying that at the resonance fre-quency absorption was a result of the MM structure.

The experimentally obtained absorption spectra andthe simulated data for the multilayer MM absorber areshown in Fig. 2(a). Also plotted is the experimentally ob-tained absorption spectrum for a single ERR of armlength 17 μm and dielectric thickness of 2 μm. As ex-pected, the absorption spectra of both devices wereindependent of the polarization angle of the incident ter-ahertz beam (not shown). The one layer structure has asingle resonance peak at 5:42THz, where 78% of the EMradiation is absorbed. In contrast, the three-layer devicehas three resonances at 4.32, 5.31, and 5:71THz with ab-sorption magnitudes of 66%, 77%, and 80%, respectively.Owing to these three closely position resonant peaks weobtain a wide frequency band, from 4.08 to 5:94THz,where the absorption is greater than 60%. Taking the cen-tral frequency of the three-layer structure to be 5:01THz,the FWHM of the absorption is 48% of the central fre-quency. This is almost two and a half times the FWHMof the single layer structure (FWHM of the single layeris 20%). The experimental data is in reasonable agree-ment with the simulated spectrum. The discrepancy be-tween the curves can be attributed to several factors; themost significant two are the possibility of a discrepancyin the real and assumed value of the polyimide refractive

Table 1. Geometric Parametersof the Multilayer Absorber

Parameter p L1 L2 L3 h1 h2 h3

Value (μm) 22 17 15.4 15 0.7 1.2 2.0

Fig. 2. (Color online) (a) Experimental and simulated (FDTD)data of the multilayer absorber. Also plotted is the experimentalabsorption spectrum for a single layer absorber. (b) Absorptionspectra for the superpixel broadband absorber.

September 1, 2011 / Vol. 36, No. 17 / OPTICS LETTERS 3477

index and potential thickness nonuniformity of the threepolyimide layers.To understand the origin of the spectral characteris-

tics, the simulated absorption distributions in thex-z plane of the three resonances are plotted inFigs. 3(a)–3(c). The resonance at 4:84THz is primarilyassociated with excitation of the bottom ERR layer,while the resonances at 5.16 and 5:70THz are mainly aconsequence of excitation of the middle and top ERRlayers, respectively.These distributions clearly reveal that each ERR

contributes to the broadband absorption.An SEM image of a superpixel terahertz absorber is

shown in Fig. 1(d). Figure 2 shows the simulated and ex-perimental absorption spectra for a superpixel with armlengths of 17, 15, 13, and 11 μm and arm widths of 6 μm.The pixel period is 44 μm, while the polyimide thicknessis 2 μm. An SEM image of a superpixel is shown inFig. 1(d). Four resonances are observed in both the simu-lation and experimental data. The disadvantage of such asuperpixel structure is that, as shown in Fig. 2(b), there issome polarization dependence. For both polarizations,the superpixel absorber has greater than 50% absorptionbetween 5.08 and 7:27THz, with a range of 2:19THz. TheFWHM for TE polarization is 37% while it is 41% for TMpolarization, representing double the FWHM of thesingle pixel.

In conclusion, we have demonstrated two types ofbroadband terahertz MM absorbers. The multilayer ab-sorber is polarization insensitive and has an absorptionbandwidth two and a half times that of a narrowbandstructure. Greater than 60% absorption is obtained overa frequency range of 1:86THz. The superpixel absorberdevice is simple to fabricate and has greater than 50%absorption over a frequency range of 2:18THz. Suchbroadband MM absorbers may be integrated with bolo-metric sensors allowing sensitive terahertz detection.

References

1. J. B. Pendry, A. J. Holden, D. J. Robbins, and W. J. Stewart,IEEE Trans. Microwave Theory Tech. 47, 2075 (1999).

2. D. R. Smith, W. J. Padilla, D. C. Vier, S. C. Nemat-Nasser,and S. Schultz, Phys. Rev. Lett. 84, 4184 (2000).

3. G. Dolling, M. Wegener, and S. Linden, Opt. Lett. 32,551 (2007).

4. S. Zhang, W. J. Fan, N. C. Panoiu, K. J. Malloy, R. M. Osgood,and S. R. J. Brueck, Phys. Rev. Lett. 95, 137404 (2005).

5. S. Linden, C. Enkrich, M. Wegener, J. F. Zhou, T. Koschny,and C. M. Soukoulis, Science 306, 1351 (2004).

6. N. I. Landy, C. M. Bingham, T. Tyler, N. Jokerst, D. R. Smith,and W. J. Padilla, Phys. Rev. B 79, 125104 (2009).

7. M. Gokkavas, K. Guven, I. Bulu, K. Aydin, R. S. Penciu, M.Kafesaki, C. M. Soukoulis, and E. Ozbay, Phys. Rev. B 73,193103 (2006).

8. D. R. Smith and N. Kroll, Phys. Rev. Lett. 85, 2933 (2000).9. M. C. K. Wiltshire, J. B. Pendry, I. R. Young, D. J. Larkman,

D. J. Gilderdale, and J. V. Hajnal, Science 291, 849 (2001).10. M. Tonouchi, Nat. Photon. 1, 97 (2007).11. H. Tao, C. M. Bingham, A. C. Strikwerda, D. Pilon, D.

Shrekenhamer, N. I. Landy, K. Fan, X. Zhang, W. J. Padilla,and R. D. Averitt, Phys. Rev. B 78, 241103 (2008).

12. J. Grant, Y. Ma, S. Saha, L. B. Lok, A. Khalid, and D. R. S.Cumming, Opt. Lett. 36, 1524 (2011).

13. Y. Ma, Q. Chen, J. Grant, S. C. Saha, A. Khalid, and D. R. S.Cumming, Opt. Lett. 36, 945 (2011).

14. S. Gu, J. P. Barrett, T. H. Hand, B. I. Popa, and S. A.Cummer, J. Appl. Phys. 108, 064913 (2010).

15. H. Wakatsuchi, S. Greedy, C. Christopoulos, and J. Paul,Opt. Express 18, 22187 (2010).

16. D. Schurig, B. J. Justice, S. A. Cummer, J. B. Pendry, A. F.Starr, and D. R. Smith, Science 314, 977 (2006).

17. W. J. Padilla, M. T. Aronsson, C. Highstrete, M. Lee, A. J.Taylor, and R. D. Averitt, Phys. Rev. B 75, 041102 (2007).

18. H. Tao, N. I. Landy, C. M. Bingham, X. Zhang, R. D. Averitt,and W. J. Padilla, Opt. Express 16, 7181 (2008).

19. Y. Q. Ye, Y. Jin, and S. L. He, J. Opt. Soc. Am. B 27,498 (2010).

Fig. 3. (Color online) (a) Absorption distribution in the x-zplane at y ¼ 0 μm at the three resonant frequencies. The hori-zontal white lines denote Au layers.

3478 OPTICS LETTERS / Vol. 36, No. 17 / September 1, 2011