Abstract
Graphene has emerged as a promising platform for THz plasmonics, allowing high confinement, long lifetimes and fast electrical tunability. Here, we predict a strong magnetic dipole response by graphene split nanorings at THz frequencies, allowing the attainment of metamaterials with a high degree of field confinement (approximately one hundredth of the excitation wavelength) that is not reachable with conventional noble metals. The magnetic response of highly doped graphene splitrings in the farinfrared is much stronger than that displayed by gold structures of similar thicknesses. We further explored stacked graphene layers as a practical way of producing highfrequency magnetism in thin, electrically tunable metamaterials. Our results support the great potential of using graphene to achieve electrically tunable magnetic metamaterials.
Introduction
The last decade has witnessed a vast body of literature on optical metamaterial designs that defy common intuition with achievements such as negative refraction^{1} and magnetic response at high frequencies,^{2,3} as well as promising applications for perfect lensing,^{4} invisibility cloaks^{5} and magnetic resonance imaging.^{6} In many studies, splitrings have become a central element because they are capable of supporting strong induced currents, leading to a resonant magnetic response down to the near infrared.^{7} Splitring resonances can be driven by the excitation of plasmons that propagate along the ring circumference, particularly when its length is of the order of half the plasmon wavelength. This condition leads to the formation of a standing wave that is similar to that of a dipole antenna, as recently demonstrated by electron energyloss spatially resolved spectroscopy.^{8}
The performances of metamaterials are hampered by optical losses and by the lack of fast means to tune their spectral response. In this context, the potential of graphene as a novel plasmonic material has been considered^{9,10} because of two important properties that are advantageous for metamaterial design: (i) graphene plasmons have short wavelengths compared to the excitation wavelength^{11,12} and (ii) the frequencies of these plasmons can be electrically tuned by injecting charge carriers into the carbon sheet.^{13,14,15,16,17} Experimental evidence of these two properties has been recently obtained by spatially mapping localized optical modes in graphene ribbons.^{18,19} Thus, graphene splitrings (GSRs) are expected to display resonances for small sizes compared with traditional noblemetal splitrings. In addition, metamaterials formed by GSRs inherit the tunability of their atomically thin fabric.
In this paper, we investigate the magnetic response of GSRs. Specifically, we show that GSRs of suitable nanoscale dimensions can be used to produce strong magnetic responses under excitation by the electric field of an incident electromagnetic wave in the THz regime. Our results are supported by analytical theory and numerical electromagnetic simulations with the behavior of graphene described through its dynamic sheet conductivity. The latter can be approximated with the Drude model, which is robust in the lowfrequency and strongdoping regimes,^{18}
where is a constant with units of length/(time)^{2} and is proportional to the Fermi energy E_{F}. The latter is defined with respect to the socalled Dirac point and depends on the charge carrier concentration n as . Here, υ_{F}≈10^{6} m s^{−1} is the Fermi velocity. The damping rate γ is estimated from the DC mobility μ as (e.g., ℏγ=1.3 meV for μ=10 000 cm^{2} V^{−1} s^{−1} and E_{F}=0.5 eV). Lastly, due to the relatively large size of the GSR resonators (GSRRs), the scattering effects at the edges can be neglected.^{20}
Methods
Figure 1a shows a characteristic GSR of diameter D=1 µm. We focused on normalincidence excitation where the electric field is polarized across the GSR gap (y direction). This configuration produces an electric dipole along the external field direction and a current circulating around the ring. The latter generates a magnetic dipole perpendicular to the ring (z direction). The coupling to the ring is maximum at a wavelength ∼110 µm (>>D), as indicated by the GSR resonance feature observed in the simulated transmission and absorption spectra of Figure 1b, which were obtained by using a finiteelement commercial code. Despite the small (compared to the excitation wavelength) size of the GSR resonators and the atomic thickness, the metamaterial presented clear resonant spectral features with transmission declining to 94% and absorption reaching 6%.
To understand the GSR response, we derived analytical expressions for the polarizability of a thin splitring. Considering the small size of this structure compared with the excitation wavelength, we could work within the electrostatic approximation in which the selfconsistent potential acting on the graphene is given by
where is the externally applied potential, f is 1 in the graphene area and 0 elsewhere,^{21} and the integral term represents the Coulomb potential produced by the induced charge. The latter is from the continuity equation in terms of the divergence of the induced current, and this is in turn given by the gradient of . We used dimensionless spatial units in the plane of the graphene. Equation (2) is also useful to study inhomogeneously doped graphene, with f proportional to the local Fermi energy.^{21} We considered normally incident farinfrared radiation linearly polarized along y so that the external potential became . Importantly, the frequency dependence and the response of the graphene were fully contained inside the dimensionless parameter η=iσ(ω)/ωD. Incidentally, when graphene was placed at the interface between two dielectrics of permittivities ε_{1} and ε_{2}, these expressions remain valid if we multiplied η by 2/(ε_{1}+ε_{2}).^{22} In matrix notation, Equation (2) reduced to , where is a Hermitian operator.^{23} It is convenient to expand the potential in terms of the eigenvectors of M plasmon resonances that emerge under the condition that 1/η equals one of the real eigenvalues of this operator, 1/η_{p}. Using Equation (1), the corresponding plasmon energy was ω_{p}−iγ/2, where
The spectral width of the plasmon features was then identical with the intrinsic damping rate γ. This result is consistent with the resonance observed in Figure 1b, which has a quality factor Q=ω_{p}/γ≈λ/Δλ≈11. Lastly, we obtained the relevant elements of the polarizability tensor from the expressions
where j^{ind} was the induced surface current density in the graphene and the superscripts E and M stand for the electric and magnetic components, respectively. The induced current distribution is shown in Figure 2a at the resonance frequency, with the maximum values in the region opposite of the gap. Inserting the noted eigenmode expansion in Equation (4) and assuming a dominant splitring plasmon mode, we find
where A^{E,M} and B^{E,M} are real dimensionless constants that only depend on the geometrical aspect ratios (i.e., g/D and W/D, see Figure 1a) and ω_{p} is given by Equation (3). The resonances beyond the frequency range of interest contributed to a smooth background that was captured by B^{E,M}. From the reciprocity theorem, we found that . All of the other components of α were nonresonant; therefore, we dismissed them. In particular, the polarization produced by an external field directed perpendicular to the gap [along x, see Figure 1a] only produced marginal nonresonant absorption at levels <0.1%. To quantify the strength of the magnetic and electric responses of the GSR resonator array, we introduced the corresponding electric and magnetic polarizability densities defined as:
Results and discussion
The results obtained by Equation (5) are shown in Figure 2b and 2c (dashed curves), together with the results of the numerical simulations for the polarizability densities (solid curves). The fitted parameters A^{E,M} and B^{E,M} were independent of frequency, ring size, graphene quality, and doping; therefore, we used Equation (5) to analyze a wide range of GSR conditions. The fitting values of A^{E,M}, B^{E,M} and η_{p} are given in Table 1 for GSRs with their representative geometrical aspect ratios.
We were interested in assessing the capabilities of graphene compared to gold, which is generally regarded as an excellent standard for metamaterial structures. We considered the performance of splitrings fabricated on four different types of thin films: singlelayer and stacked (in which layers can be separated by dielectric spacers^{24}) graphene versus singlelayer (one (111) atomic layer) and thinfilm (20 nm) gold, respectively. For sufficiently thin structures (thickness t<<D), the Drude conductivity given by Equation (1) in combination with Equation (5) remains applicable. Expressing the parameter a in terms of the bulk plasma frequency as is convenient. In stacked graphene, an effective bulk frequency was obtained by simply multiplying the graphene conductivity by the number of graphene layers. Here, d_{z} is the distance between those layers. Singlelayer graphene is retrieved for t=d_{z}. Considering d_{z}=2 nm and E_{F}=0.5 eV, we find that ℏω_{bulk}≈1.2 eV, compared with ∼9 eV for gold.^{25} Obviously, there is some flexibility to reach lower ω_{bulk} values for graphene by simply increasing the interlayer spacing d_{z}. In addition, the intrinsic damping of gold is ℏγ≈70 meV, compared with ℏγ≈1 meV in graphene for the levels of doping and mobility under consideration. Therefore, graphene should allow us to operate well at lower energies than gold.
We compared the performances of these types of thin films (Figure 3). In particular, we examined the degree of confinement (Figure 3a), quantified through the ratio of the resonance wavelength to the ring diameter (λ/D), the quality factor Q=ω_{p}/γ (Figure 3b), and the ratio of the magnetic polarizability density of monolayer (stacked) graphene over that of monolayer (thin) gold. We set the aspect ratios to W/D=g/D=0.1, and we used these expressions with the ring diameter required to obtain the resonance frequency ω_{p} running along the horizontal axis (Equation (3)). Namely,
Considering the realistic mobilities assumed here (μ≈13 000 cm^{2} V^{−1} s^{−1}),^{26,27} graphene offers a suitable alternative to fabricate resonant splitrings with Q>1 down to the THz regime (Figure 3b) with a degree of confinement that is greater than what is achieved by noble metals for similar thicknesses and frequencies (Figure 3a). In fact, at farIR wavelengths (longer than 100 µm), the confinement factor for thin gold is smaller than one. This finding indicates that in contrast to stacked graphene, thin gold SRR arrays lie in the diffraction rather than in the metamaterial regime, whereas the corresponding quality factor is much smaller than 1. A comparison between the magnetic response of singlelayer gold and singlelayer graphene reveals that doped singlelayer graphene can provide a stronger magnetic response than singlelayer gold at frequencies higher than ∼1 THz. The difference between these two materials nearly reaches an order of magnitude when considering strongly doped graphene. The magnetic responses of monolayer and stacked graphene are considerably weaker than the response of thin gold. However, this picture could significantly improve with the availability of higher mobility graphene^{28} subjected to stronger doping.^{29,30}
Conclusions
In conclusion, graphene has emerged as a tunable material that allows the fabrication of ultrathin metamaterials. A central element of these metamaterials (splitrings) is shown here, displaying resonances with unprecedented levels of field confinement. Thus, graphene monolayers can be used to reach the true homogenization limit (D<<λ) with the additional advantage of having an electrically tunable optical response, which can be beneficial for the fabrication of compact, versatile metamaterials down to the terahertz regime.
References
 1
Smith DR, Pendry JB, Wiltshire MC . Metamaterials and negative refractive index. Science 2004; 305: 788–792.
 2
Yen TJ, Padilla WJ, Fang N, Vier DC, Smith DR et al. Terahertz magnetic response from artificial materials. Science 2004; 303: 1494–1496.
 3
Grigorenko AN, Geim AK, Gleeson HF, Zhang Y, Firsov AA et al. Nanofabricated media with negative permeability at visible frequencies. Nature 2005; 438: 335–338.
 4
Pendry JB . Negative refraction makes a perfect lens. Phys Rev Lett 2000; 85: 3966–3969.
 5
Schurig D, Mock JJ, Justice BJ, Cummer SA, Pendry JB et al. Metamaterial electromagnetic cloak at microwave frequencies. Science 2006; 314: 977–980.
 6
Wiltshire MC, Pendry JB, Young IR, Larkman DJ, Gilderdale DJ et al. Microstructured magnetic materials for RF flux guides in magnetic resonance imaging. Science 2001; 291: 849–851.
 7
Enkrich C, Wegener M, Linden S, Burger S, Zschiedrich L et al. Magnetic metamaterials at telecommunication and visible frequencies. Phys Rev Lett 2005; 95: 203901.
 8
Boudarham G, Feth N, Myroshnychenko V, Linden S, García de Abajo FJ et al. Spectral imaging of individual splitring resonators. Phys Rev Lett 2010; 105: 255501.
 9
West P, Ishii S, Naik GV, Emani NK, Shalaev VM et al. Searching for better plasmonics materials. Laser Photonics Rev 2010; 4: 795–808.
 10
Tassin P, Koschny T, Kafesaki M, Soukoulis CM . A comparison of graphene, superconductors and metals as conductors for metamaterials and plasmonics. Nat Photon 2012; 6: 259–264.
 11
Jablan M, Buljan H, Soljacic M . Plasmonics in graphene at infrared frequencies. Phys Rev B 2009; 80: 245435.
 12
Koppens FH, Chang DE, García de Abajo FJ . Graphene plasmonics: a platform for strong light–matter interactions. Nano Lett 2011; 11: 3370–3377.
 13
Wunsch B, Stauber T, Sols F, Guinea F . Dynamical polarization of graphene at finite doping. New J Phys 2006; 8: 318.
 14
Hwang EH, Das Sarma S . Dielectric function, screening, and plasmons in twodimensional graphene. Phys Rev B 2007; 75: 205418.
 15
Fei Z, Andreev GO, Bao W, Zhang LM, McLeod AS et al. Infrared nanoscopy of Dirac plasmons at the graphene–SiO2 interface. Nano Lett 2011; 11: 4701–4705.
 16
Ju L, Geng B, Horng J, Girit C, Martin M et al. Graphene plasmonics for tunable terahertz metamaterials. Nat Nanotech 2011; 6: 630–634.
 17
Fang ZY, Thongrattanasiri S, Schlather A, Liu Z, Ma LL et al. Gated tunability and hybridization of localized plasmons in nanostructured graphene. ACS Nano 2013; 7: 2388–2395.
 18
Chen J, Badioli M, AlonsoGonzález P, Thongrattanasiri S, Huth F et al. Optical nanoimaging of gatetunable graphene plasmons. Nature 2012; 487: 77–81.
 19
Fei Z, Rodin AS, Andreev GO, Bao W, McLeod AS et al. Gatetuning of graphene plasmons revealed by infrared nanoimaging. Nature 2012; 487: 82–85.
 20
Thongrattanasiri S, Manjavacas A, García de Abajo FJ . Quantum finitesize effects in graphene plasmons. ACS Nano 2012; 6: 1766–1775.
 21
Thongrattanasiri S, Silveiro I, García de Abajo FJ . Plasmons in electrostatically doped graphene. Appl Phys Lett 2012; 100: 201105.
 22
Christensen J, Manjavacas A, Thongrattanasiri S, Koppens FHL, García de Abajo FJ . Graphene plasmon waveguiding and hybridization in individual and paired nanoribbons. ACS Nano 2012; 6: 431–440.
 23
Ouyang F, Isaacson M . Surface plasmon excitation of objects with arbitrary shape and dielectric constant. Philos Mag B 1989; 60: 481–492.
 24
Yan H, Li X, Chandra B, Tulevski G, Wu Y et al. Tunable infrared plasmonic devices using graphene/insulator stacks. Nat Nanotech 2012; 7: 330–334.
 25
Johnson PB, Christy RW . Optical constants of the noble metals. Phys Rev B 1972; 6: 4370–4379.
 26
Novoselov KS, Geim AK, Morozov SV, Jiang D, Zhang Y et al. Electric field effect in atomically thin carbon films. Science 2004; 306: 666–669.
 27
Novoselov KS, Geim AK, Morozov SV, Jiang D, Katsnelson MI et al. Twodimensional gas of massless Dirac fermions in graphene. Nature 2005; 438: 197–200.
 28
Bolotin KI, Sikes KJ, Jiang Z, Klima M, Fudenberg G et al. Ultrahigh electron mobility in suspended graphene. Solid State Commun 2008; 146: 351–355.
 29
Chen CF, Park CH, Boudouris BW, Horng J, Geng B et al. Controlling inelastic light scattering quantum pathways in graphene. Nature 2011; 471: 617–620.
 30
Khrapach I, Withers F, Bointon TH, Polyushkin DK, Barnes WL et al. Novel highly conductive and transparent graphenebased conductors. Adv Mater 2012; 24: 2844–2849.
Acknowledgements
This work has been supported by the Spanish MICINN (MAT201014885 and Consolider NanoLight.es), the European Commission (FP7ICT20094248909LIMA and FP7ICT20094248855N4E), the EPSRC (UK) and the Leverhulme Trust and the MOE Singapore (grant MOE2011T31005).
Author information
Affiliations
Corresponding authors
Rights and permissions
This work is licensed under a Creative Commons AttributionNonCommercialShare Alike 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/byncsa/3.0/
About this article
Cite this article
Papasimakis, N., Thongrattanasiri, S., Zheludev, N. et al. The magnetic response of graphene splitring metamaterials. Light Sci Appl 2, e78 (2013). https://doi.org/10.1038/lsa.2013.34
Received:
Revised:
Accepted:
Published:
Issue Date:
Keywords
 artificial magnetism
 graphene
 graphene plasmons
 metamaterials
 plasmonics
Further reading

Template free Ndoped 3D porous carbon@Co3O4: Towards highly efficient catalysis for peroxymonosulfate degradation of antibiotics
Optical Materials (2021)

Multipolar Plasmonic Resonances of Aluminum Nanoantenna Tuned by Graphene
Nanomaterials (2021)

Highly tunable dual bound states in the continuum in bulk Dirac semimetal metasurface
Applied Physics Express (2021)

Far‐Field Excitation of Acoustic Graphene Plasmons with a Metamaterial Absorber
Advanced Photonics Research (2021)

Unconventional Approach to Fabricating a TiO2 Nanoring with Precise Dimension Control Based on Starlike Polymeric Nanoreactors
The Journal of Physical Chemistry Letters (2021)