Broadband wave absorption in single-layered and ... - OSA Publishing

0 downloads 0 Views 2MB Size Report
Apr 18, 2017 - The broadband absorption mechanism originates from two .... Graphene, a single-layered carbon atom arranged in a honeycomb lattice,.
Vol. 25, No. 9 | 1 May 2017 | OPTICS EXPRESS 9579

Broadband wave absorption in single-layered and nonstructured graphene based on farfield interaction effect FEI GAO,1 ZHIHONG ZHU,1,2,* WEI. XU,1 JIANFA ZHANG,1 CHUCAI GUO,1 KEN LIU,1 XIAODONG YUAN,1 AND SHIQIAO QIN1,2 1

College of Optoelectronic Science and Engineering, National University of Defense Technology, Changsha 410073, China 2 State Key Laboratory of High Performance Computing, National University of Defense Technology, Changsha 410073, China * [email protected]

Abstract: We present a wave absorption design consisting of periodical arrays of dielectric bricks on the dielectric substrate, which is coated with single-layered and nonstructured graphene, supported by a thick piece of metal. The design is demonstrated to broadband nearperfect absorption with 0.82 terahertz (THz) bandwidth of over 90% absorption and with central frequency of 1.68 THz. The broadband absorption mechanism originates from two contributions. Firstly, the periodical arrays of dielectric bricks on the nonstructured graphene can provide both a set of graphene plasmon resonances with large relative frequency interval and relative radiation rate γ/ω in the THz range. Secondly, the linewidth of each resonance can be broadened by the far-field interaction between neighboring resonators to overlap and spread over a wide frequency region in the THz range. The design in this paper is simple, and consequently facilitates the fabrication and promotes the application of broadband graphene absorbers. © 2017 Optical Society of America OCIS codes: (230.3120) Integrated optics devices; (130.5440) Polarization-selective devices; (250.5403) Plasmonics.

References and links 1.

M. Liu, X. Yin, E. Ulin-Avila, B. Geng, T. Zentgraf, L. Ju, F. Wang, and X. Zhang, “A graphene-based broadband optical modulator,” Nature 474(7349), 64–67 (2011). 2. B. Sensale-Rodriguez, R. Yan, M. M. Kelly, T. Fang, K. Tahy, W. S. Hwang, D. Jena, L. Liu, and H. G. Xing, “Broadband graphene terahertz modulators enabled by intraband transitions,” Nat. Commun. 3(1), 780–787 (2012). 3. M. Freitag, T. Low, W. Zhu, H. Yan, F. Xia, and P. Avouris, “Photocurrent in graphene harnessed by tunable intrinsic plasmons,” Nat. Commun. 4(3), 1951 (2013). 4. M. Freitag, T. Low, F. Xia, and P. Avouris, “Photoconductivity of biased graphene,” Nat. Photonics 7(1), 53–59 (2013). 5. S. H. Lee, M. Choi, T. T. Kim, S. Lee, M. Liu, X. Yin, H. K. Choi, S. S. Lee, C.-G. Choi, S.-Y. Choi, X. Zhang, and B. Min, “Switching terahertz waves with gate-controlled active graphene metamaterials,” Nat. Mater. 11(11), 936–941 (2012). 6. A. Andryieuski, A. V. Lavrinenko, and D. N. Chigrin, “Graphene hyperlens for terahertz radiation,” Phys. Rev. B 86(12), 121108 (2012). 7. A. Fallahi and J. Perruisseau-Carrier, “Design of tunable biperiodic graphene metasurfaces,” Phys. Rev. B 86(19), 195408 (2012). 8. H. Yan, X. Li, B. Chandra, G. Tulevski, Y. Wu, M. Freitag, W. Zhu, P. Avouris, and F. Xia, “Tunable infrared plasmonic devices using graphene/insulator stacks,” Nat. Nanotechnol. 7(5), 330–334 (2012). 9. M. Jablan, H. Buljan, and M. Soljacic, “Plasmonics in graphene at infrared frequencies,” Phys. Rev. B 80(24), 245435 (2009). 10. W. Xu, Z. H. Zhu, K. Liu, J. F. Zhang, X. D. Yuan, Q. S. Lu, and S. Q. Qin, “Dielectric loaded graphene plasmon waveguide,” Opt. Express 23(4), 5147–5153 (2015). 11. S. Thongrattanasiri, F. H. L. Koppens, and F. J. García de Abajo, “Complete optical absorption in periodically patterned graphene,” Phys. Rev. Lett. 108(4), 047401 (2012). 12. R. Alaee, M. Farhat, C. Rockstuhl, and F. Lederer, “A perfect absorber made of a graphene micro-ribbon metamaterial,” Opt. Express 20(27), 28017–28024 (2012).

#286897 Journal © 2017

https://doi.org/10.1364/OE.25.009579 Received 16 Feb 2017; revised 11 Apr 2017; accepted 11 Apr 2017; published 18 Apr 2017

Vol. 25, No. 9 | 1 May 2017 | OPTICS EXPRESS 9580

13. A. Y. Nikitin, F. Guinea, F. J. Garcia-Vidal, and L. Martin-Moreno, “Surface plasmon enhanced absorption and suppressed transmission in periodic arrays of graphene ribbons,” Phys. Rev. B 85(8), 081405 (2012). 14. Z. H. Zhu, C. C. Guo, K. Liu, J. F. Zhang, W. M. Ye, X. D. Yuan, and S. Q. Qin, “Electrically tunable polarizer based on anisotropic absorption of graphene ribbons,” Appl. Phys., A Mater. Sci. Process. 114(4), 1017–1021 (2014). 15. A. Y. Nikitin, F. Guinea, and L. Martin-Moreno, “Resonant plasmonic effects in periodic graphene antidot arrays,” Appl. Phys. Lett. 101(15), 151119 (2012). 16. W. R. Zhu, I. D. Rukhlenko, L. M. Si, and M. Premaratne, “Graphene-enabled tunability of optical fishnet metamaterial,” Appl. Phys. Lett. 102(12), 121911 (2013). 17. J. Zhang, C. Guo, K. Liu, Z. Zhu, W. Ye, X. Yuan, and S. Qin, “Coherent perfect absorption and transparency in a nanostructured graphene film,” Opt. Express 22(10), 12524–12532 (2014). 18. Y. Yao, R. Shankar, M. A. Kats, Y. Song, J. Kong, M. Loncar, and F. Capasso, “Electrically tunable metasurface perfect absorbers for ultrathin mid-infrared optical modulators,” Nano Lett. 14(11), 6526–6532 (2014). 19. C. Y. Ye, Z. H. Zhu, W. Xu, X. D. Yuan, and S. Q. Qin, “Electrically tunable absorber based on nonstructured graphene,” J. Opt. 17(12), 125009 (2015). 20. B. Wu, H. M. Tuncer, M. Naeem, B. Yang, M. T. Cole, W. I. Milne, and Y. Hao, “Experimental demonstration of a transparent graphene millimetre wave absorber with 28% fractional bandwidth at 140 GHz,” Sci. Rep. 4(2), 4130 (2014). 21. C. C. Guo, Z. H. Zhu, X. D. Yuan, W. M. Ye, K. Liu, J. F. Zhang, W. Xu, and S. Q. Qin, “Experimental demonstration of total absorption over 99% in the near infrared for monolayer-graphene-based subwavelength structures,” Adv. Opt. Mater. 4(12), 1955–1960 (2016). 22. F. Xiong, J. Zhang, Z. Zhu, X. Yuan, and S. Qin, “Ultrabroadband, more than one order absorption enhancement in graphene with plasmonic light trapping,” Sci. Rep. 5, 16998 (2015). 23. W. Gao, J. Shu, C. Qiu, and Q. Xu, “Excitation of plasmonic waves in graphene by guided-mode resonances,” ACS Nano 6(9), 7806–7813 (2012). 24. M. Amin, M. Farhat, and H. Bağcı, “An ultra-broadband multilayered graphene absorber,” Opt. Express 21(24), 29938–29948 (2013). 25. J. Kotakoski, D. Santos-Cottin, and A. V. Krasheninnikov, “Stability of graphene edges under electron beam: equilibrium energetics versus dynamic effects,” ACS Nano 6(1), 671–676 (2012). 26. H. Yan, T. Low, W. Zhu, Y. Wu, M. Freitag, X. Li, F. Guinea, P. Avouris, and F. Xia, “Damping pathways of mid-infrared plasmons in graphene nanostructures,” Nat. Photonics 7(5), 394–399 (2013). 27. Z. H. Zhu, C. C. Guo, J. F. Zhang, K. Liu, X. D. Yuan, and S. Q. Qin, “Broadband single-layered graphene absorber using periodic arrays of graphene ribbons with gradient width,” Appl. Phys. Express 8(1), 015102 (2015). 28. M. Pu, P. Chen, Y. Wang, Z. Zhao, C. Wang, C. Huang, C. Hu, and X. Luo, “Strong enhancement of light absorption and highly directive thermal emission in graphene,” Opt. Express 21(10), 11618–11627 (2013). 29. B. Z. Xu, C. Q. Gu, Z. Li, and Z. Y. Niu, “A novel structure for tunable terahertz absorber based on graphene,” Opt. Express 21(20), 23803–23811 (2013). 30. A. Andryieuski and A. V. Lavrinenko, “Graphene metamaterials based tunable terahertz absorber: effective surface conductivity approach,” Opt. Express 21(7), 9144–9155 (2013). 31. W. Tan, Y. Sun, Z. G. Wang, and H. Chen, “Manipulating electromagnetic responses of metal wires at the deep subwavelength scale via both near- and far-field couplings,” Appl. Phys. Lett. 104(9), 091107 (2014). 32. R. Yu, R. Alaee, F. Lederer, and C. Rockstuhl, “Manipulating the interaction between localized and delocalized surface plasmon-polaritons in graphene,” Phys. Rev. B 90(8), 085409 (2014). 33. Z. H. Zhu, C. C. Guo, K. Liu, W. M. Ye, X. D. Yuan, B. Yang, and T. Ma, “Metallic nanofilm half-wave plate based on magnetic plasmon resonance,” Opt. Lett. 37(4), 698–700 (2012).

1. Introduction Wave absorption is crucial for various types of application, including detection, imaging and modulators [1, 2]. Graphene, a single-layered carbon atom arranged in a honeycomb lattice, has attracted considerable attention as a result of its range of potential applications from optical to THz frequencies [1–10]. Recently, graphene has become one of the most promising materials in the design of wave absorbers that can operate in both THz and infrared spectral ranges due to its tunability of carrier mobility and conductivity [11–30]. The first demonstration of complete optical absorption based on graphene takes place in the arrays of doped graphene nanodisks lying on a dielectric layer coating a metal [11]. Then, graphene absorbers have been extensively studied and a variety of structures have been proposed such as graphene ribbons [12–14], anti-dots [15], fishnets [16, 17] or nonstructured graphene with comprising sub-wavelength resonant structures [18–22]. However, these absorbers operate at narrowband spectral ranges because their complete wave absorptions rely on the resonances of surface plasmon polaritons (SPPs) [23], which limits their potential applications. One mean

Vol. 25, No. 9 | 1 May 2017 | OPTICS EXPRESS 9581

of improving the bandwidth is to use multilayered graphene structures, where several patterned graphene layers biased at different voltages and backed with dielectric substrates are stacked on top of each other [24]. However, it may be difficult to put multilayered graphene structures with different biasing voltages into application. In addition, these structures are based on structured graphene which unavoidably introduce a great many truncation edges of the graphene. These graphene edges from the tailored fabrication haven't been considered in the theory but really exist in experiment, which might lead to edge effects, such as unordered diffuse scattering losses [25, 26], thus rendering it difficult to achieve the unity absorption in practice. Another method is to make use of nanostructured graphene with gradually changed geometric sizes [27]. However, the structures is still based upon the structured graphene to produce varying continuous plasmonic resonances when excited by an incident wave. In this paper, based on single-layered and nonstructured graphene, we present a broadband absorber with 0.82 THz bandwidth of over 90% absorption and with a central frequency of 1.68 THz. Furthermore, the absorption's dependence on incident angle is also weak and the broadband absorption remains high for wide range of incident angles. Here, we supply the physical insight of broadband absorption of the proposed model system firstly. Then, we run a numerical simulation at terahertz frequency. Finally, we investigate the effects of some relative parameters. 2. Structure and principle The representation of the proposed structure is shown in Fig. 1, which consists of periodical arrays of dielectric bricks on the substrate coated with single-layered graphene supported by a thick piece of metal. The structure is characterized by the periodic interval P, the width W and thickness H1 of the brick, the Fermi energy Ef of graphene and the thickness H2 of substrate. In the above configuration, the metal film and arrays of dielectric bricks form two mirrors of the asymmetric Fabry-Perot cavity [12]. The transmission channel could be suppressed as long as the bottom metallic film is thick enough. The reflection path is the sum of all multiple reflection, and the sum would minimize to zero due to the destructive interference if the characters of the structure are well designed. When both the transmission channel and the reflection path close, the only possible outcome will be absorption, which leads to the nearperfect absorption. In this structure, the near-perfect absorption is closely related to the graphene plasmon resonance. So, in order to achieve the broadband near-perfect absorption, the key is to obtain broadband graphene plasmon resonance.

Fig. 1. (a) Schematic drawing of broadband absorber, consisting of periodic bricks (color deep coffee) on dielectric substrate (color aqua) coated with graphene (color blue) supported by a thick piece of metal (color yellow). (b) Geometrical characters of a unit cell.

Vol. 25, No. 9 | 1 May 2017 | OPTICS EXPRESS 9582

First, we consider the excitation of the graphene surface plasmons with a free-space wave. It is well known that high doped graphene (Ef >  ω) in which optical loss from interband transition is suppressed can sustain strongly surface plasmons propagating along the sheet with wave vector [9, 10] kGSP (ω ) =

π 2

e2 E f

ε 0 (ε r1 + ε r 2 )ω (ω + iτ −1 ),

(1)

where e is the charge of electron,  is the reduced Planck's constant, ω is the angular frequency of graphene surface plasmons, Ef is the Fermi energy of graphene, τ is the carrier relaxation time in graphene, ε0 is permittivity of vacuum, εr1 is the effective relative permittivities of periodical arrays of dielectric bricks above the graphene film, and εr2 is the relative permittivities of the dielectric substrate. The wave vector kGSP in graphene is much larger than that in the free space. In order to excite the surface plasmons in graphene using free-space wave, the large difference between their wave vectors should be overcome. In the structure, one of the roles of periodical dielectric brick arrays is compensating the wavevector mismatches. Considering the C4v symmetry of the structures, for simplicity, we assume the incident wave to be in the y-z plane and the polarization to be along the x-direction (TE polarization). The resonant angular frequency ω is determined by the wave-vector matching equation Re(k GSP ) =

ω c

sin θ + N

2π , P

(2)

where Re(kGSP) is the real part of wave vector kGSP, c is the speed of light in vacuum, P is the array period, N is positive integer, θ is the incident angle defined as the angle between the incident wave and the Z axis. Then, by simple algebra operation, graphene plasmon resonant frequency ω for the N-th order mode can be obtained as

ωN =

e 2 E f sin θ + e4 E 2f sin 2 θ + 8π 2  2 c 2 e 2 E f ε 0 (ε r1 + ε r 2 ) N / P 2π  2 cε 0 (ε r1 + ε r 2 )

.

(3)

From Eq. (3), we can see that multiple graphene plasmon resonances can be excited in the ω N − ω N +1 structure. We can also find that the relative frequency interval (defined by ) ω ( N + ωN +1 ) 2 between two neighbor modes decreases with the increased mode order number. So we can achieve the largest relative frequency interval between the lowest order modes. For example, under normal incident excitation (θ = 0), the relative frequency interval, is expressed as ω N − ω N +1 N +1 − N =2 , between the first and second modes is 34.3%, and between (ωN + ωN +1 ) 2 N +1 + N the second and third modes is 20.2%. Exciting a set of graphene plasmon resonances with large relative frequency interval is the first step to obtain broadband near-perfect absorption. When the excited graphene plasmon modes propagate along the y direction, at the interface between the dielectric bricks and air gap, they partly reflect back into the bricks, and partly transmit into the air gap to become an outgoing wave to re-excite the graphene plasmon resonances supported by the neighboring dielectric bricks. So when the separation between neighboring dielectric bricks is smaller than wavelength, except for the near-field coupling effect, the far-field interaction must be considered [31, 32]. From the temporal coupled mode theory (CMT), the corresponding dynamics are described as follows:

Vol. 25, No. 9 | 1 May 2017 | OPTICS EXPRESS 9583

daN 1 = (iω N − γ N − Γ N )aN 1 + iκ N aN 2 + i γ N ( s+ + i γ N e − iϕ N aN 2 ), dt

(4)

daN 2 = (iωN − γ N − Γ N )aN 2 + iκ N aN 1 + i γ N ( s+ + i γ N e− iϕN aN 1 ), dt

(5)

where ωΝ is the center frequency of the N-th graphene plasmon resonance, aN1 = |aN1| e − jωN t and aN2 = |aN2| e − jωN t are the corresponding normalized amplitude in two nearest neighbor resonators, s+ is the amplitude of the incident light, ΓN is the absorption rate, and γΝ is the radiation rate, φΝ is the phase of the re-radiated wave of one resonator that reaches the nearest neighbor resonator. In the equations, the coefficient κΝ describes the near-field coupling between the two resonators, while γ N e− iϕN contributes to the far-field interaction between them (this approximation is valid when the separation between neighboring resonators is much smaller than wavelength). For the case ΓN = 0 and s+ = 0 in Eqs. (4) and (5), we obtain the quasi-eigen frequency

ω N ± = ω N + iγ N ± (κ N + iγ N e − iϕ ). N

(6)

Equation (6) can be also written as

ω N ± = ω N ± (κ N + γ N sin ϕ N ) + i (γ N ± γ N cos ϕ N ).

(7)

From Eq. (7), we can find two hybridized modes for the N-th graphene plasmon resonance is formed because of the interaction between two neighbor original resonators. The second and third terms on the right-hand side in Eq. (7) show that the near-field coupling is only responsible for the energy level splitting, while the far-field interaction alters both the linewidth and splitting. The hybridized mode with a narrow linewidth is subradiant dark mode, while the mode with broadened linewidth is superradiant bright mode. In the case of φΝ = 0 (critical condition), which can be achieved by varying the distance between two neighbor resonators, the subradiant dark mode has zero linewidth and thus cannot be excited by the incident plane wave. On the contrary, the superradiant bright mode reaches the maximal linewidth, which is two times as much as that of original graphene plasmon resonance, and it has strongly interaction with the incident plane wave and then gains imperium on role of wave absorption. On the other hand, from Eq. (1), we can see a quadratic dependence of Re(kGSP) on ωΝ, which implies the ability of light confinement provided by the graphene plasmon is proportional to ωΝ. The quantified degree of confinement is given by the ratio of graphene plasmon to free-space-light wave vectors Re(kGSP ) π  2ω N cε 0 (ε r1 + ε r 2 ) = . k0 e 2 EF

(8)

It indicates that graphene plasmon resonance in the terahertz range has much smaller ability of light confinement than in the mid- and far-infrared range. In other words, the radiation rate γN and relative radiation γN/ ωN of graphene plasmon resonance in the terahertz range is much larger than in the mid- and far-infrared range. Since the structure can provide both a set of graphene plasmon resonances with large relative frequency interval and larger relative radiation rate γN/ ωN in the terahertz range, we can expect the broadened linewidth for each resonance caused by the far-field interaction between neighboring resonators to overlap and spread over a wide frequency region in the terahertz range. The achieved broadband graphene plasmon resonances imply that the broadband near-perfect absorption may be obtained under certain conditions.

Vol. 25, No. 9 | 1 May 2017 | OPTICS EXPRESS 9584

3. Simulation and discussion

Fig. 2. (a) Calculated absorption spectra for TE polarization light under normal incident excitation. Electric field amplitude patterns on the central cutting Y-Z plane, corresponding to (b) 1.3 THz, (c) 1.6 THz and (d) 2.2 THz, respectively. The graphene is marked by solid black lines and the incident light is represented by the white arrows. (e) Calculated absorption as a function of frequency and incident angle. (f) Analytical graphene plasmon resonant frequencies as a function of incident angle by solving Eq. (3).

To verify the theoretical prediction, we conducted the fullwave numerical simulations using frequency domain solver in CST Microwave Studio. In the numerical simulation, the graphene is modeled as an anisotropic effective media of thickness tg = 1 nm with the in-plane component of relative complex permittivity as εrx(ω) = εry(ω) = 2.5 + iσ(ω)/(ωε0tg) and the surface normal component as εrz(ω) = 2.5, where σ(ω) is the optical conductivity and calculated in the terahertz range as

σ (ω ) =

Ef 2e 2 k BT i ln[2 cos h( )], 2 −1 π  ω + iτ 2k B T

(9)

where τ = μ E f e v 2f ( v f = 106 m/s is the Fermi energy level, and μ = 10000 cm2/(V·s) is the DC mobility), and T = 300 K is the temperature. In our simulation, the dielectric material and substrate material are considered to be non-dispersive with relative permittivity εdielectric = 12 and εsubstrate = 4, respectively. The metallic material is gold with thickness 1 μm and is treated as a dispersive medium following the Drude model. The relative permittivity is ε (ω ) = ε ∞ − ω p2 / (ω 2 + iωγ ) . The value of ε ∞ , ωp and γ are 1.0, 1.38 × 1016 rad•s−1 and 1.23 × 1013s−1 [33], respectively. In this work, the absorption A is calculated by A = 1-R-T, where R and T represent reflection and transmission respectively.

Vol. 25, No. 9 | 1 May 2017 | OPTICS EXPRESS 9585

Fig. 3. (a) Calculated absorption as a function of frequency and thickness of substrate H2. (b) Calculated absorption as a function of frequency and width W of dielectric. (c) Calculated absorption for different Fermi energy. (d) Analytical graphene plasmon resonant frequencies as a function of Fermi level by solving Eq. (3).

First, we consider the case with P = 40 μm, W = 29 μm, H1 = 14 μm H2 = 22 μm, T = 300 K and the graphene with Ef = 0.5 eV. Figure 2(a) shows the calculated absorption spectra for TE polarization under normal incident excitation. From Fig. 2(a), we find the 90% absorption bandwidth is as high as 0.82 THz with a central frequency of 1.68 THz. The electric field amplitude patterns on the central cutting y-z plane for three representational frequencies 1.3, 1.6, and 2.2 THz, corresponding to three absorption bands labeled by A, B and C, are calculated and shown in Figs. 2(b)–2(d), respectively. Figures 2(b)–2(d) clearly show that the three absorption bands correspond to the first, second and third order graphene plasmon resonances, respectively. In addition, we can also see that, in the air gap between two neighboring dielectric bricks, the electric fields are composed of propagating waves that correspond to the far-field interaction and evanescent waves that correspond to the near-field interaction effect. In order to investigate the absorption sensitivity to the oblique incident wave, we vary the incident angle θ from 0 to 89 degree while maintaining the incidence wave in the y-z plane and the electric field vector along the x-direction with all other parameters kept the same as before. Figure 2(e) shows the absorption as a function of frequency and incident angle. From Fig. 2(e), three absorption peaks are always observed and they are only slightly sensitive to the incident angle when the incident angle varies between 0 and 80 degree. By solving Eq. (3), we can obtain the analytical graphene plasmon resonant frequencies as a function of incident angle for the first, second and third modes, as shown in Fig. 2(f). Comparing Figs. 2(e) and 2(f), we can see that the three absorption peak frequencies are in agreement with the analytical graphene plasmon resonant frequencies, which slightly

Vol. 25, No. 9 | 1 May 2017 | OPTICS EXPRESS 9586

increase with increased incident angle. As the incident angle increases beyond 80 degree, the absorption decreases rapidly. The reason corresponds to the fact that the near-unity absorption of this type of the structure is closely related to destructive interference. However, under conditions of glancing incidence, the reflection amplitude from graphene plasmon resonances and periodical arrays of dielectric bricks is much larger than that from thick metal, and the destructive interference condition is destroyed. Next, we consider the situation where the Fermi level is fixed at 0.50 eV and the thickness H2, and width W are varied. Figure 3(a) shows the calculated absorption as a function of frequency and thickness H2 under normal incident excitation with P, W and H1 fixed at 40 μm, 29 μm and 14 μm, respectively. From Fig. 3(a), we can see that the absorption is sensitive to the thickness of the substrate. It results from the fact that the phase-matching condition of destructive interference is dependent on the phase delay provided by the thickness of the substrate material. Figure 3(b) shows the calculated absorption as a function of frequency and width W with P, H1 and H2 fixed at 40 μm, 14 μm and 22 μm, respectively. The width W varies between 0 and P. From Fig. 3(b), we can find the width W has a great influence on the bandwidth of the absorption and have an optimized value for obtaining maximal broadband absorption. This can be understood through the following qualitative analysis. The φΝ is tightly related with the distance between two neighbor resonators, and the critical condition of φΝ = 0 can be achieved for the optimized value, where the superradiant bright mode reaches the maximal linewidth while the subradiant dark mode has zero linewidth. When the distance deviates from the critical condition, the linewidth of the superradiant bright mode decreases while that of subradiant dark mode increases and the situation becomes quite different. For the limit case of W = 0, the structure consists of graphene, dielectric substrate and metallic mirror, and the consequent F-P effect (arising from multiple reflections at the substrate ends) enhances the absorption, resulting in a 55% maximum absorption at 2.4 THz. For W = P, the F-P effect and absorption enhancement is similar, however the effective refraction index is higher than the case of W = 0 due to the higher-index material on the top, and so the maximum absorption occurs at the lower frequency (about 1.0 THz). In the two limit cases (W = 0 and W = P), the field interaction disappears and the absorption is narrow. Then, we study the tunable effects of this absorption by varying the Fermi level. Figure 3(c) is the calculated absorption spectra for Fermi energy 0.4eV, 0.5eV and 0.6eV. From Fig. 3(c), we can observe a blue shift of the working frequency range at large absorption with the increased Fermi level, which implies that the operation frequency can be tuned by the Fermi level. This is because the broadband near-unity absorption of this type of structure is closely related to the lowest order graphene plasmon resonances, and simultaneously, these resonant frequencies increase with increased Fermi level, which is confirmed by Fig. 3(d) that shows the analytical graphene plasmon resonant frequencies as a function of Fermi level by solving Eq. (3). 4. Conclusion

In summary, we design a broadband near-perfect absorber based on single-layered and nonstructured graphene. Numerical simulations demonstrate that the working bandwidth of 90% absorption is as high as 0.82 THz, with a central frequency of 1.68 THz. The broadband mechanism originates from two key contributions. One is the multiple graphene plasmon resonances with large relative frequency interval and radiant rate. The other is the far-field interactions between the neighboring resonators. The absorber is based on nonstructured graphene, which can avoid destroying the unique properties of graphene and be beneficial to the development of many related applications. In additon, it is simple and thus facilitates the application of the broadband graphene absorbers. Funding

Natural National Science Foundation of China (NSFC) (11674396, 1304389, 61404174).